NTN TR82 en

For New Technology Network
Special Issue; Composite Material,

Technology and Products
for Industrial Machines
October 2014
ISSN 0915-0528 OSAKA, JAPAN

Introduction of “WIND LAB.” - Test equipment
for ultra-large main shaft bearings for wind power generators
NTN has developed the world's largest large-bearing-test apparatus (“WIND LAB.”) which allows evaluation of off-shore
wind power generator main shaft bearings with an outer diameter of up to 4.2m. This will enhance our ability to develop
technology to cope with increasing demand around the world. NTN introduced the equipment at the Advanced Technology
R&D Center (Kuwana-shi, Mie-ken).
“WIND LAB.” applies loading to the bearings with a total of 6 hydraulic cylinders, in vertical and horizontal directions, to
simulate the actual conditions experienced by a wind turbine. This enables detailed technological analysis and specification
review of bearings selected for large wind power generator applications. In addition to enabling evaluation of the load
conditions in the stationary or extreme state, it can also support various types of bearings, such as self-aligning roller bearings
and tapered roller bearings, so that the customers can conduct tests tailored to their needs.
NTN develops and provides all kinds of bearings used in the components of wind power generators including main bearings,
final gears, step-up gears, and generators. Wind power generators, including those for off-shore use, are increasing in size, and
main shaft bearings require advanced technology to support rotation of blades, which can exceed 100 m in diameter.
Therefore, evaluating the actual ultra-large bearings will help accelerate development and improve bearing reliability. In
addition, we will aim to monitor the bearing conditions during the test by integrating the condition monitoring system (CMS)
for wind power generators developed by NTN (“WindDoctor ®”- page 74). Our objective is to improve the functionality of
the bearings while continuing to further enhance the equipment diagnostic technology of CMS.
We will strive to enhance our advantage as a bearing manufacturer by using "WIND LAB." and further grow our business
for wind power generation equipment.
Hydraulic cylinder
Bearings under test
▲ “WIND LAB.”
test equipment for ultra-large wind
power generator main shaft bearings
▲ An example of an ultra-large
main shaft bearing
R
TECHNICAL REVIEW
Special Issue: Composite Material,
No.82 Technology and Products
for Industrial Machines
corporation
NTN TECHNICAL REVIEW No.82
CONTENTS
Preface Natsuhiko MORI 1
Contribution Advanced Powders Processing – Recent Powder Metallurgy Techniques and It’s Trend –
2
Hideshi MIURA Professor, Department of Mechanical Engineering, Kyushu University
● Special Issue for Composite Material Products and Technology

Product Introduction and Complex Technology of Resin,Sintered Metal and Magnet for Growth Markets
12
Yoshio OKI
The Improvement in Strength of Sintered Machine Parts
21
Takahiro OKUNO and Naoki YASHIRO
High Performance Magnetic Core for Induction Hardening Devices 26
Takuji HARANO
The Copper-Iron-based Material Equivalent to Bronze-based Material for Sintered, Oil-impregnated Bearings 30
Tomonori YAMASHITA
High Load Sliding Bearing Unit For Use in Copier and Printer Fusers 34
Kei HATTORI and Takuya ISHII
● Special Issue for Manufacturing Technology

Machine Tool Main Spindle Bearings with Air Cooling Spacer 38
Yushi ONDA, Kikuo FUKADA, Yohei YAMAMOTO and Masato YOSHINO
New Engineering Plastic Cage of High Speed Angular Contact Ball Bearings for Machine Tools 44
Wakana INOUE, Mineo KOYAMA and Soichiro ONITSUKA
Bearing Technical Calculation System for Machine Tool Spindles 49
Keiichi UEDA, Yushi ONDA and Shouhei HASHIZUME
Estimation Method for Friction Torque of Air-oil Lubricated Angular Contact Ball Bearings 54
Hiroki FUJIWARA
Real-time Monitoring of the Superfinishing Process
61
Kohei HIGASHI, Hajime ISA, Takakazu KITAGAWA, Takashi ONISHI and Kazuhito OHASHI
Lightweight Bearing for Robot 68

Sebastien BRISSON
●Special Issue for CMS Supertechnology New Products

Development of Condition Monitoring System 74
Keisuke HASHIZUME, Akitoshi TAKEUCHI and Yuzuru TANAKA
Introduction of "ULTAGE" series : High Functionality Roller Bearings
78
Shogo TABATA, Kouhei TODA and Yasuhiko SHIMIZU
Technical Trend of Aircraft Bearings 83
Takashi NISHIKAWA, Nao HAYASHI and Akiko HAYAKAWA
● Technical Papers
The Influence of the Humidity and Quantity of Hydrogen into Steel under the Abrasive Wear
88
Motohiro ITO
Evaluation of Shear Fatigue Properties of Medium Carbon Steels and
Rolling Bearing Steels up to Giga-cycle Regimes 94
Noriaki MIWA
● Our Line of Award Winning Products

“2013 ‘CHO’ MONODZUKURI Innovate Components Awards” NIPPON Brand Award
Parallel Link High Speed Angle Control Equipment 98
Hiroshi ISOBE, Yukihiro NISHIO, Seigo SAKATA, Naoya KONAGAI and Hiroyuki YAMADA
“The Japan Powder Metallurgy Association Awards 2013” New Materials
Development of the Oil Seal Made of Sintered Metal for Variable Valve Timing Device 99
Yosuke SUGAI and Toshihiko MOURI
“The Japan Powder Metallurgy Association Awards 2013” Effort Prize
Sintered Hydrodynamic Bearing for Super Thin Fan Motor 100
Tadahiro UCHIYAMA and Shinji KOMATSUBARA
“The Japanese Society of Tribologists 2013” Encouragement Award
Development of Ultrasonic Torsional Fatigue Tester for Rapid Shear Fatigue 101
Property Evaluation of Rolling Bearing Steels
Noriaki MIWA
Our Line of New Products 102

NTN TECHNICAL REVIEW No.82（2014）
[ Preface ]
A Message for the Special Issue:

Composite Material, Technology and Products
for Industrial Machine
Natsuhiko MORI
Executive Officer
As the manufacturing industry is globalized and both the market and demand of products
are expanding throughout the world, maximizing the effective use of the minimum resources
and paying attention to the global environment has become an urgent issue. In Japan,
development and utilization of advanced technologies and materials are accelerating in order
to maintain sustainable and constructive competitiveness. Therefore, methods for utilizing
materials and some manufacturing methods are about to change significantly. Some
examples are the expanded application of CFRP and use of 3D printers.
In view of these circumstances, NTN has formed a new division, the Composite Material
Product Division in April, 2011. The purpose of this division is to consolidate and enhance the
resin, sintered metal and magnetic material technologies to develop and introduce composite
products and unit module products into the market. Then, they can contribute to the fuel
efficiency and electric drive of automobiles and high-performance industrial machine
components.
At the 27th Japan International Machine Tool Fair (JIMTOF2014), which took place from
October 30 to November 4, with the theme “’Mono-Zukuri DNA’ connects to the Future and
the World”, NTN Technical Review No. 82 was published featuring NTN’s composite
material, technology and products for industrial machines.
In this Technical Review, we had a contribution entitled, “Advanced Powders Processing –
Recent Powder Metallurgy Techniques and Its Trend” from Professor Hideshi Miura.
Professor Hideshi Miura is from the Department of Mechanical Engineering, Kyushu
University, and is an authority on powder and powder metallurgy.
In the “Special Issue for Composite Material Products and Technology”, we introduce high-
strength sintered metal, magnetic material and resin as well as hybrid products. In the
“Special Issue for Manufacturing Technology”, we introduce technology and products that
were exhibited at JIMTOF2014, manufactured with our proprietary technology. Finally, in
“Special Issue for CMS Supertechnology New Products” we introduce base
technology/products and monitoring technology for industrial machines.
At NTN we will celebrate our 100th anniversary in fiscal year 2017. Following our theme,
“For New Technology Network: We shall contribute to international society through creating
new technologies and developing new products,” we are striving to contribute to
development of the industry by utilizing our proprietary technology development capabilities.
-1-
[ Contribution ]
Advanced Powders Processing

– Recent Powder Metallurgy Techniques and It’s Trend –
Hideshi MIURA
Professor, Department of Mechanical Engineering, Kyushu University
Advanced powder processing techniques such as metal injection molding (MIM)

are hoped to be suitable for fabricating complex shaped components with low cost.
So far, we have applied MIM process to produce various types of industrial
materials such as ferrous and Ti alloys, especially heat treated Fe-6Ni-0.4C
showed the super high strengthened properties of 2000MPa tensile strength and
8% elongation which are superior to wrought materials. On the other hand, direct
laser forming (DLF) technique is another advanced powder processing technique.
Recently, much effort has been made to apply this process to automobile,
aerospace and medical fields due to the potential of producing complex shaped Ti
alloy structures. In this paper, the above two techniques are introduced as recent
powder metallurgy techniques.
processing, as well as development of new technology

1. Introduction are well underway for P/M products with high density
Powder metallurgy (P/M) is a metal processing and high performance. However, recent P/M
method using a phenomenon called sintering, a technologies are focused on machining and composite
process which bonds particles by heat (temperature processing technologies in the boundary areas with
lower than the melting point) for extracting particular other material processing technologies, among others,
properties from the metal and ceramics powder and as shown in Table 11). For example, the powder
plays an important role as one of the manufacturing injection molding is a combination of P/M and plastic
processes of materials and products in high- molding, and powder lamination 3D printing is a
industrialized society. Liquid-phase sintering also combination of P/M and laser machining
exists depending on the components. The biggest (CAD/CAM).Acquiring knowledge and applications of
advantage of P/M is to be able to shape the powder these new technologies should be very useful for
directly into the end product (near net or net shaping) manufacturing/development of new materials and
having a significant degree of freedom in material products for NTN. In this paper, I will focus on the
properties, composition, thermal processing and above two innovative and unique molding/sintering
microstructure.This provides properties that cannot be methods, providing an overview of the technologies
achieved by the melting method. In addition, there is and introduce you to the recent outcomes of our
an economical benefit allowing volume production. research as an alternative to providing you with a
With these properties, P/M is used to produce various perspective into the future.
components for machine structural use such as gears,
bearings and connecting rods (for automobiles) P/M is
2. Metal Injection Molding
also used to produce cutting/abrasion resistant tooling
materials such as cemented carbide tip and molding, A molding technique that allows both a high degree
high melting point metal materials such as W and Mo, of freedom for shaping and high density in a relatively
porous materials such as filters and implants for living easy process has been desirable in P/M. One such
bodies, electric contacts/power collecting materials technique is the metal injection molding (MIM) process
such as switches and materials for pantograph contact with binder developed in the 1970s.
strips, and magnetic materials such as magnetic cores MIM process is used in many forms however the
and sensor rings. basic steps are similar to the ones shown in Fig. 11).
Improvement of conventional processes from The typical raw material is nearly round particles of
production of powder, shaping, sintering and post- carbonyl and powder produced from oxide reduction
-2-
Advanced Powders Processing – Recent Powder Metallurgy Techniques and Its Trend –
Table 1 Recent applications of various P/M processings
Use Required properties Process and materials

Anticorrosive component True density uniform alloy HIP, titanium
Thermal process equipment Carburizing resistance and creep resistance Reaction sintering, extrusion, nickel aluminide
Fine jewelry/accessories Good appearance, complex shape Injection molding, precious metal, titanium
Pipe, thick plate, rod, plate True density, net shape Spray forming, steel, super alloy
Magnetic component Magnetic functionality, ambient resistance Sintering, injection molding, rapidly solidified powder
Covering material Oxidation resistance and corrosion resistance Spray forming, intermetallic compound
Microelectronics board Adjustment of thermal property, absorption of microwave Injection molding, W-Cu, Mo-Cu, Al-SiC, Al-AlN
High melting-point metal Complex shape Complete densification method such as HIP, special alloy
Aircraft structural material High relative strength Complete densification method, rapidly solidified Al alloy
Structural material Fatigue strength and fracture strength Powder forging, iron alloy
Dynamic membranel filter Small pore size, corrosion resistance CIP, stainless steel
Abrasion-resistant structural material Lightweight, high abrasion resistance Hot press, boride, Al based composite material
equivalent to approximately 6 wt% in case of ferrous

metal, although the percentage depends on the filling
characteristics of the powder. The best shaping can
Powder Binder
be attained when the viscosity of the kneaded material
is 100 Pa・s or less however, the viscosity depends
Preliminary not only on the viscosity of the binder itself but also the
kneading
temperature of the kneaded material, shear rate,
volume of the solids and type of surface wetting agents
included in the binder. Good uniformity is required for
Kneading and granulation the kneaded material to maintain good process control.
(particles) The viscosity, which is sensitive to the composition and
variations may also prevent good flow into a mold
cavity. Usually, the viscosity of the kneaded material
Injection can be kept within the desirable range by slightly
molding
adding more binder.
The device used for molding is the same as those
Solvent Heated degreasing/
degreasing preliminary sintering used for plastic injection molding. The reciprocal-type
screw driven by a motor is used for mixing/kneading
the raw material so that the kneaded material becomes
uniform, producing the required pressure to fill the
kneaded material into the mold. The raw material, in
the form of pulverized particles at normal temperature,
Sintering
is introduced from the charging hopper and heated
Fig.1 Conceptual sequence of MIM steps over the melting temperature of the binder as it passes
the barrel. The kneaded material of melted powder and
binder is pushed forward to instantaneously fill in the
and gas atomization with the average grain size of 0.5 mold cavity. After being sufficiently cooled down, the
to 15 μm to facilitate densification during sintering. compact is removed and the process is repeated. For
Generally, thermoplastic polymer materials are used as better molding, higher pressure or lower viscosity is
the binder however; water and diverse inorganic required. Since the pressure that can be used for the
materials are also often used. The typical binder molding machine is limited and the viscosity is
consists of 70% of paraffin wax and 30% of governed by the temperature, the main control factors
polypropylene with an appropriate lubrication and are temperature and pressure for the molding process.
wetting agent to provide adhesiveness to the binder After molding, the binder is removed from the
and the powder. This type of binder completely melts compact by a process called debinding. Many
at approximately 150˚C. The amount of the binder is debinding techniques and processes related to
about 40 vol% of the kneaded material, which is different binders have been developed which, together
-3-
with the composition of the binders, account for the Table 2 Various debinding techniques
majority of the patents. Table 2 2) shows various Process name Binder content Debinding condition
debinding techniques.
Heated degreasing
In the industry, thermal decomposition and solvent
MACPHERSON process PE, camphor Vacuum
degreasing are mainly used, however the former still
WITEC process WAX，PE Turbulence, absorber
has issues such as long debinding time (slow heating
up to 600˚C) and possible deformation of the product. VI process WAX，PE，PP High vacuum, evaporation
The latter has significantly overcome those issues but, (volatile)

many of the solvents such as acetone and ethylene are RIVERS process Water, methyl cellulose Anhydration within the mold
hazardous to humans and the environment and their QUICKSET process Water, PEG Freeze dry
handling involves additional problems. Therefore, new Solvent extraction
binders soluble in ethanol and water are actively under WITEC process PE，PS，PEG Water, methylene chloride
development. At any rate, long debinding time is MACPHERSON process PE，PS Trichlene
disadvantageous for production, which considerably WAX，PE，PP
limits the permissible thickness of the MIM products. AMAX process Methylene chloride
Peanut oil
Up until a few years ago, it was considered that a Chemical decomposition
thickness of around 10 mm was the limit from the BASF process Modified Polyacetal Nitric acid, oxalic acid vapor
economical point of view however, with the evolution of UV decomposition process WAX, acryl UV irradiation
debinding technology it is now possible to allow a PE: polyethylene, PP: polypropylene, PEG: Polyethylene glycol,
thickness of around 25 mm. In addition, there are other PS: polystyrene, POM: polyacetal
special techniques such as debinding by catalyst 3) and
a kind of water freezing technique called Quickset Table 3 Mechanical properties of various alloy steels
process 4). by different material processings
The next step, sintering, is processed under inert, or Mechanical Melting
Steel type MIM P/M
reduction, atmosphere or vacuum atmosphere, similar properties method
to the conventional P/M technique. Sintering creates High-speed steel
Deflecting strength (MPa) 3200 2500 2500
SKH10
strong bonding of particles and voids are removed by (tempered material)
Hardness (HRC) 70 71 67
densification. Isotropic powder filling results in uniform
Maraging steel Tensile strength (MPa) 1640 1500 1800
contraction (15–20%) as can be predicted. Therefore, 18Ni-8Co-5Mo Elongation (%) 2∼3 1∼2 8
initial compacts are made a little larger so that the final (aged material) Hardness (HRC) 47 35
compacts fit the desirable size. After sintering, the Martensitic SUS410
stainless steel Tensile strength (MPa) 1600 900 1950
compacts show superior strength and uniform
SUS440C (17Cr-1C) Elongation (%) 1∼2 4 2
composition compared to the characteristics possible (tempered material) Hardness (HRC) 53 30 57
with many other manufacturing processes. Table 3 17-4PH stainless steel Tensile strength (MPa) 1340 970 1370
shows an example of the comparison of mechanical SUS630 Elongation (%) 11 2 14
properties of various ferrous sintered materials with (aged material) Hardness (HRC) 44 24 45
this process and other manufacturing processes. The
mechanical properties of the MIM process not only
exceed the properties of the conventional P/M powder technique which indicates extremely high
materials but are also equivalent or superior to the mechanical properties (equivalent to sinter forging
molten materials in any steel types 5)-9). This indicates material) of 1800 MPa tensile strength and elongation
that the MIM process is effective for forming materials of 3-4% which cannot be achieved by the conventional
with hard machining properties, as well as for P/M technique (tensile strength of ordinary sintered low
improving the material properties. The following alloy steel by press molding is around 1000 MPa and
indicates some of the typical examples from the results elongation is 1-2%) 10). We conclude that this is a
of our research. manifestation of superior properties of fine
heterogeneous structure with locally gradient nature of
2.1 Ultra-high strengthening of ferrous sintered contents unique to sintering material of mixed
materials by MIM process poweder. This structure is the martensite phase (in
The author group, from our research, proposed a case of high concentration, retained austenitic phase),
new precise controlling method of the amount of which appears according to the Ni concentration
carbon by atmosphere control in the MIM process. gradient, particularly due to insufficient diffusion of
This developed sintered low alloy steel (Fe-2%Ni-0.5% aggregated Ni powder in the Fe matrix powder, and is
Mo-0.4%C thermal processed material) by the mixed surrounded in mesh by the tempered martensite
-4-
phase. As the above properties were only obtained by in any structure, highlighted by the arrows and are
using the natural phenomenon of aggregation of Ni surrounded by tempered martensite phases, seen as a
powder and composition of Fe-2% Ni-0.5%Mo, further network structure. It is observed that the bigger the Ni
superior properties (targeting more than 2000 MPa of powder grain size and the more the Ni content, the
tensile strength, more than 5% of elongation, and more larger the Ni-rich phases.
than 600 MPa of rotational bend fatigue strength) could Fig. 3 shows the distribution of Ni by EPMA. As it
be sufficiently expected by artificially changing the size shows the areas of high and low Ni concentration, it is
and content of the Ni powder or sintering/thermal clear that this is the heterogeneous structure with Ni
processing conditions based on this steel type. content gradient. It was found that when Ni powder of
Therefore, we examined the optimum structure on the small grain size is used, Ni is dispersed within an Fe
micro-level such as the composition and percentage of matrix resulting in a smaller difference in concentration.
each phase that composes the heterogeneous It was also revealed that when Ni content is large, the
structure and their dispersion state. difference in concentration also becomes larger.
Fig. 4 shows the tensile strength obtained from the
2.2 Method of Experiment tensile test. All the test pieces show a very high value
We used carbonyl iron powder (Fe-OM, average of around 2000 MPa.
grain size: 4.4 μm, made by Fukuda Metal Foil &
Powder Co., Ltd.) as the powder for this experiment.
We also used water atomization powder (AKT-Ni,
made by Mitsubishi Steel Mfg. Co., Ltd.) as the Ni to be (a)
added, classified into 3 grain sizes. The average grain
sizes were 6, 16 and 24 μm, which we called fine,
medium and coarse, respectively, and added 4, 6 and
8 mass% in the mixture. As the binder, we blended 69
mass% of paraffin wax, 20 mass% of atactic
polypropylene, 10 mass% of carnauba wax and 1
(b)
mass% of stearin acid. The powder and the binder
were kneaded so that the powder ratio became 65
vol% and a tensile test plate the shape of a dumbbell
was created by injection molding using the obtained
raw material. Paraffin wax was first removed from the
compact using the solvent degreasing method in the 50µm
heptane gas phase, then heated degreasing and

sintering were conducted successively. Sintering was Fig.2 Representative of optical microstructures for various
Fe-Ni steel compacts sintered at 1250 ˚C for 1 hour in vacuum
conducted in a vacuum atmosphere of 10-1 Pa and a atmosphere. The arrow shows the bright martensite region.
sintering temperature of 1200-1350˚C was maintained (a) As-tempered compacts of different Ni particle size,
for 1 hour. The sintered compact was thermal (b) As-tempered compacts of different Ni content (mass%)
processed in the Ar atmosphere.
After maintaining 900˚C for 30 minutes, quenching
was performed in oil and then after maintaining 200˚C (a)
for 120 minutes, tempering was performed by air
cooling. The test piece was then tested for relative
density measured by the Archimedes method, carbon
content, structure observation, and Ni distribution by
EPMA and tensile test.
(b)
2.3 Test Results and Observations
The density of the obtained sintered compacts was
95% or more of the relative density in any Ni powder
grain sizes and Ni contents. The carbon content of any
thermal processed materials was about 0.4 mass%.
Fig. 2 shows the pictures of the structure of the typical
Fig.3 Representative of Ni element mappings by EPMA
thermal processed materials. analysis for as-tempered Fe-Ni steel compacts changing
The specimens show that white Ni-rich phases exist (a) Ni mean particle size, (b) Ni content (mass%)
-5-
Higher strength and elongation are seen in samples However, hardness of the Ni-rich phase increased
with smaller Ni powder grain size and the test pieces nearly 20% from before the test showing an even
with Ni content of 6% show the maximum values. Fig. 5 higher value than the tempered martensite phase. On
shows the elongation, showing that the smaller the Ni the other hand, little change was observed in the
powder grain size, the greater the elongation and the hardness of the martensite phase.
Ni content of 6% again shows the largest value. The We also attempted to quantify the retained austenite
test piece with fine Ni and 6% content showed a very before and after the tensile test by X-ray diffractometry
strong property of 2040 MPa of tensile strength and (XRD). Fig. 7 shows the retained austenite amount
8.1% elongation. This is the highest level among the before and after the tensile test. The retained austenite
reported sintered low alloy steels 11-14) by MIM. It is existed nearly 17% before the test and decreased to
concluded that such ultra high strength properties were 6% after the test. From the above, it is concluded that
obtained because of the fine heterogeneous structure. the stress-induced martensitic transformation (TRIP)
Fig. 6 shows the hardness of the test pieces of fine occurred in the Ni-rich phase resulting in improved
powder with Ni content of 6% before and after the hardness. It was found that this also caused ultra-
tensile test. The Ni-rich phase showed a lower value strong tensile properties in the thermal processed
than the tempered martensite phase before the tensile materials of sintered low alloy steel with
test. heterogeneous structure 15).
2500 680
Tensile strength, MPa
660
2000 640
Hardness, HV
620
1500 600
580
1500
560
500 Fine 540
520 Tempered α’
0 Medium 500
4 Coarse Ni rich phase
6 Before
8 After
Ni content, % Tensile test
Fig.4 Tensile strength of different Ni content and Fig.6 Micro vickers hardness of each region at before
particle size and after tensile testing
18
10
16
8
Retained austenite, %
14
Elongation, %
6 12
10
4
8
2 Fine 6
0
Medium
4
4 Coarse 2
6
8 0
Before After
Ni content, %
Fig.5 Elongation of different Ni content and particle size Fig.7 Retained austenite at before and after tensile testing
-6-
controlling laser output. In the case of Fe-Ni mixed

3. Powder Lamination 3D Printing powder, development of improving dimensional
CAD has become a common practice for accuracy is also conducted by sealing the remaining
development and design of components and products. pores with shot peening however, those methods
Particularly, as component shapes are stored as 3D require many processes and the strength is not
CAD data, a new additive manufacturing (AM) has sufficient.
been developed to shape the actual body from the Therefore, a method to directly irradiate by the laser
data. The most representative method is only metallic powder for melting and binding to build
stereolithography which forms thin layers of light curing products is under research and development. Its
liquid resin by laser light scanning, to shape the principle is shown in Fig. 8 17). First, a layer of powder
products by adding layers as needed. This is applied to is placed on the manufacturing table (cylinder) and the
the metallic and ceramic powder in a technique called laser irradiates the powder based on the CAD data so
powder bed fusion (PBF; the former is called “laser that the powder melts and solidifies. Next, the following
forming” hereinafter). The following methods are layer is placed on the manufacturing table from the
available as AM: powder supplying cylinder with a roller then the laser
● Material extrusion (ME) method: The most popular irradiated the sample, fusing and solidifying the
method of additive manufacturing uses continuous powder. By repeating this lamination process, the
extrusion of melted thermal plastic resin from a desired 3D shape can be obtained. In this case, the
nozzle. dimensional accuracy and surface roughness are not
● Sheet lamination (SL) method: An additive at the industrially satisfactory level. Due to this, metal
manufacturing method that uses adhering sheet-like stereolithography composite machining technology
material such as paper and film. was developed where accurate cutting technology is
● Directed energy deposition (DED): An additive added as an additional machining process to produce
manufacturing method that injects droplets of molding with short delivery time and low cost.
metallic powder liquefied by plasma or laser from a The following describes the recent outcome of
nozzle. research of the author group 17-22), as they have been
Recently, material development and manufacturing focusing on biomaterial and Ti as aircraft component
techniques have evolved so that AM is not only used material and conducting basic research of laser
simply for creating a mold to check the shape design forming using Ti powder in the degas/Ar gas
but also for manufacturing products of rich functionality atmosphere as the net shape machining technology of
with a certain degree of precision. We are mainly complex structure.
focusing on laser forming here.
Several techniques for supplying powder for laser
forming have been researched for laser forming using
powder 16). Initial techniques consisted of the use of Fiber laser
binder content mixed with or coated on the powder
where only the binder was melted and used for
binding. Some require further sintering and others do
Chamber
not. The former required degreasing and both Roller
(Ar, Vacuum)
frequently used impregnation. With a larger amount of
resin binder, the forming speed is faster. Therefore, Base plate
Powder
with green body forming, the time is significantly
reduced. However, in instances where the resin binder
is not used, the forming time tends to be longer. In the
former case, the processes are shortened from 2
Sintering table
processes of degreasing/sintering and infiltration to
one process of degreasing/sintering/infiltration by Powder supply
changing blending of resin (thermosetting phenol and table
paraffin wax) and the coating method, as well as the Fig.8 Sketch of laser forming equipment
property of the base material. There are also
improvements of strength and dimensional accuracy by
HIP processing instead of infiltration, or by applying
impregnation and infiltration after liquid phase sintering
of nickel phosphor bronze and completely melting it
-7-
3.1 Ti Powder Laser Forming The energy density is the energy input on the unit
An important aspect of laser 3D forming is the high area, P [W] is the laser output, s [mm] is the scanning
density and high accuracy of the product. In our interval and v [mm/s] is the scanning speed. The
research, we have improved the relative density of the surface roughness decreases as the energy density
compact up to 97.8% through the improvement of the increases and converges at around Sa=10 μm. It is
powder supply process. In this research, we have reported that the surface tension of the melted metal
created high density compacts by improving laser decreases as the temperature increases 23), therefore,
conditions, evaluating their mechanical properties and it is considered that the wettability of the melted part
examining the impact of the powder grain size on the increases on the board as the temperature increases
side surface accuracy. This is done with the intent of along with the increase of energy density to form a
creating compacts of even higher density and higher smooth layer on the surface.
accuracy. Next, we conducted a lamination experiment. We
decided that the shape of the laser formed compacts
3.2 Experiment Method would be the same as the test pieces for the tensile
We used Ti-6Al-4V alloy powder (TILOP64 made by test (14B test piece). For the forming parameters, we
Osaka Titanium technologies Co., Ltd.) by gas- used 260 W and 80 mm/s, which were the conditions
atomized technique as the raw material powder. The to make the lowest surface roughness in the single
grain size is under 45 μm and the average grain size is layer manufacturing experiment in the previous
33.4 μm. In addition, we prepared finer powder with the section.
grain size under 25 μm (average grain size of 18.8 μm) The annealing process was applied to some laser
to evaluate the impact of the powder grain size on the formed compacts. Annealing consisted of maintaining
side surface accuracy. We used ytterbium fiber laser of the pieces at 700˚C for 2 hours then cooling them in a
the continuous oscillation method with maximum furnace. Fig. 10 shows the cross-section image of a
output of 330 W for the laser oscillator. The laser beam laser formed compact. This is a high density compact
from this oscillator scans the powder with a final beam with a relative density of 99.8% with almost no voids.
diameter of 50 μm using the scanning system including Fig. 11 shows the corrosion structure. Generation of
galvanometer mirror. The manufacturing chamber is acicular martensite by laser forming can be observed.
filled with high-purity argon (600 Pa) after being In addition, epitaxial growth where the said structure
degassed. grew vertically on the direction of formation was also
observed. This structure did not change between
3.3 Test Results and Observations before and after the annealing process.
First, a single layer was created changing laser Fig. 12 shows the tensile test results of the high
parameters and the surface roughness was measured density compacts. The JIS Standard for 60 types of
with a laser microscope (OLS-4000 from Olympus) in titanium alloy specifies that the tensile strength and
order to evaluate the impact of laser parameters on the elongation to e 895 MPa and 10% respectively. The
surface roughness. Fig. 9 shows the measurement of
the surface roughness of the single layer compact over
energy density Ep [J/mm2] indicated by equation (1).
Ep=P/(s･v) ……………………………(1)
80 Fig.10 Cross sectional microstructure of laser formed

Scan
Surface roughness Sa, µm
speed,
Ti compact
mm/s
60 50
60
40 (a) (b)
80
20 100
160
0
0 200 400 600
Energy density, J/mm2
Fig.9 Surface roughness of single layers in function of Fig.11 Etched microstructures of laser formed Ti compacts
volumetric laser energy density (a) As – formed (b) After heat treatment
-8-
value. One of the causes was assumed to be the

1200 impact of small voids still present within the material
1217 14.0
1131 even with high density however; the details are to be
Tensile strength, MPa
examined in the foregoing review.
Elongation, %
800
8.88 10.0 When laser scan is applied linearly on the evenly
placed metallic powder layer, the metallic powder
6.89
forms one linear compact as the metallic powder melts
400 6.0
Tensile strength and solidifies as shown in Fig. 14. At that time,
granular residues are generated around the formed
Elongation
0 2.0 compact. Fig. 15 shows the graph indicating the
(a) (b) diameter of the residues. It was found that the diameter
of residues became smaller by increasing the laser
Fig.12 Tensile properties of laser formed Ti compacts
(a) As – formed (b) After heat treatment output and was halved when fine powder was used.
Fig. 16 shows the side surface roughness of the laser
formed compacts. Improvement of around 60% in
surface roughness was confirmed in the compact using
tensile strength of the laser formed compacts is a fine powder. A change in side surface properties was
sufficiently higher value than the standard. verified by using fine powder even with the same
The elongation indicates a lower value, but it is forming parameters however, the stable supply of
improved by the annealing process by over 10% in uniform powder layer is difficult with fine powder.
some samples. The fusion and solidification of the Therefore, further study is required for the
surface of the compacts is repeated by the laser improvement of side surface accuracy.
scanning during laser forming, which generates
residual stress within the formed compacts. This
residual stress is considered to be one of the causes of
the reduced elongation. We assume that by applying
the annealing process, the residual stress was
removed and the elongation was improved with a
closer value to the JIS Standard. Fig.14 Appearance of a laser formed line
Next, we also evaluated the dynamic mechanical
properties. We introduced fatigue using the pulsating
tension test with a frequency of 30 Hz and stress ratio 100
of R=0.1 using an electro-hydraulic servo testing Ti64_33.4µm
Diameter, mm
80 Ti64_18.2µm
machine. Fig. 13 shows the fatigue test results. The
vertical axis indicates fatigue strength under pulsating 60
tension (twice the stress amplitude). The fatigue
40
strength estimated from the tensile test result is more
than 600 MPa and the fatigue strength of this test was 20
100 150 200 250 300
260 MPa, a significantly lower value than the estimated
Laser power, W
Fig.15 Diameter of adhered particles
600
Stress amplitude*2, MPa
100
500
Under 45µm
Roughness Rz, µm
400 80
Under 25µm
300 60
200
40
100
1.E+05 1.E+06 1.E+07 1.E+08
20
Number of cycle, N Right side Left side
Fig.13 S-N curve of laser formed Ti compacts Fig.16 Side surface roughness of laser formed Ti
after heat treatment compacts using different Ti particle powders
-9-
References
4. Conclusion 1) R. M. German, Powder metallurgy science, translated by Hideshi
Miura et al. Uchida Rokakuho (1996).
Metal injection molding (MIM) and powder lamination 2) Kazuo Okamura et al. Plastic molding technology, 27 (1996) 235.
3D printing by laser are introduced as new powder 3) D Weinand et al. Plastic molding technology, 13 (1996) 19.
forming methods from our research and the former 4) C. Quichand：Proc. of Powder Met．World Congress，EPMA，2
(1996) 1101.
MIM is already recognized as an industrially useful 5) Hideshi Miura et al. Powder and powder metallurgy, 40 (1993) 393.
forming method. However, creating large products is 6) Hideshi Miura et al. Powder and powder metallurgy, 42 (1995) 353.
still difficult because of deformation due to their own 7) Hideshi Miura et al. Powder and powder metallurgy, 41 (1994) 1071.
8) Takeharu Baba et al. Powder and powder metallurgy, 42 (1995)
weight during the debinding process and long 1119.
processing time. If these problems are resolved, the 9) Hideshi Miura et al. Powder and powder metallurgy, 40 (1993) 988.
scope of application is expected to be greatly 10) Hideshi Miura et al. Powder and powder metallurgy, 42 (1995) 378.
11) H. Miura, M. Matsuda, Material Transactions, 43 (2002) 343.
expanded.
12) Materials Standards for PM Structural Parts, 2009 Edition, MPIF,
On the other hand, metallic powder lamination 3D Princeton, NJ, (2009) 48-49.
printing has just begun with the ability of forming 13) Materials Standards for Metal Injection Molded Parts, 2007 Edition,
complex structure on the outside, as well as the inside. MPIF, Princeton, NJ, (2007) 16-17.
14) Metals Handbook, 10th ed., 1, ASM Int., Materials Park, OH, (1990)
However, as the powder is melted and laminated, 430-448 and 793-800.
particularly roughness of the side surface and the 15) W. Harun, H. Miura et al. J. Japan Soc. Pow. & Pow. Met., 59 (2012)
entire dimensional accuracy need to be improved. 677.
16) Masato Imamura，Powder and powder metallurgy，48 (2001) 415.
There are also issues associated with volume 17) Hideshi Miura et al. Powder and powder metallurgy, 53 (2006) 740.
production but, this is unique from the manufacturing 18) Hideshi Miura et al. Powder and powder metallurgy, 54 (2007) 707.
viewpoint and attracting a lot of attention. Based on 19) Hideshi Miura et al. Powder and powder metallurgy, 55 (2008) 738.
20) H. Miura, et al. Proceedings of the MPIF/APMI International
this, further development is very much expected.
Conference on Powder Metallurgy & Particulate Materials (CD-
ROM), (2009) 708.
21) H. Miura, et al. Materials Science Forum, 654-656, (2010) 2057.
22) H. Miura, et al. Proceedings of the 2012 Powder Metallurgy World
Congress & Exhibition(CD-ROM), (2012).
23) Akio Sasama, et al. : Journal of The Japan Institute of Metals，40
(1976) 1030.
<Author biography>
Hideshi Miura
Professor, Doctor of Engineering, Department of Mechanical Engineering, Kyushu University
1997 Steel Metallurgy Course (Master) completed, Department of Mechanical Engineering, Kyushu University, Assistant
1985 Doctor of Engineering received from Kyushu University, Assistant Professor, Department of Materials Development Engineering,
Kumamoto University
1992-1993 Visiting Researcher, Rensseaer Polytechnic Institute, USA
1995 Professor, Faculty of Engineering, Kumamoto University
2004 Professor, Mechanical Engineering, Graduate School of Engineering, Kyushu University
2011- present President, Education and Research Center of Manufacturing, Kyushu University
[Specialty]
Material processing, powder metallurgy, metallic injection molding, laser powder lamination 3D printing
[Academic society and committee affiliations]
Japan Society of Powder and Powder Metallurgy: Board Director (1996-2010), Vice President (2003-2010), President (2012-2014), Adviser (2014-)
The Japan Institute of Metals and Materials: Subcommittee V, Manager (1995, 1996), Vice Chairman (1997, 1998), Chairman (1999, 2000)
The Iron and Steel Institute of Japan: Division Steering Committee Member (Powder/Particle Forum Chair) (1998-2012),
Division Vice Chair (2012-2014), Division Chair/Board Director (2014-2016)
American Powder Metallurgy Institute: Powder Injection Molding International Symposium Organization Committee Member (1995, 1996, 1998-)
P/M Sci. & Tech. Briefs Editorial/Review Member (1999-2005), International Conference Program
Committee Member (2004-)
The Japan Society of Mechanical Engineering: Materials and Processing Division, 81st Period Steering Committee Member (2nd Technical
Committee Chair) (2003-2004), Division Vice Chairman (2005-2006), Division Chair (2006-2007),
Fellow (2010-)
The Japan Society for Technology of Plasticity: Board Director/Publishing Business Committee Chair (2007-2009), Kyushu District Head (2010-2012)
Conference Steering Committee Chair, Publication Editorial/Review Committee Member, Award Selection Committee Member, Representative,
Council Member, Program Committee Member of Powder Metallurgy International Conferences, Organizing Committee Member, Chair of the above
academic societies
-10-
-11-
[ Technical Article ]
Product Introduction and Complex Technology of Resin,

Sintered Metal and Magnet for Growth Markets
Yoshio OKI*
NTN Composite Material Products Division has developed the

mechanical parts and bearings, and the unit module products by a wide
range of materials and technology such as resin material, sintered metal,
magnetic material and fluid dynamic pressure technology.
In order to meet the market to be diversified, NTN is working to develop the future markets while
creating new material properties by promoting fusion of each technology and Composite Materials which
are applying tribology, bearing material, high-precision processing technology and magnetic materials.
This paper reports the Developed Products of Composite Products for the further growth markets.
1. Introduction 2. Product line and production system of

Composite Material Products Division
NTN’s Composite Material Products Division
consolidates and integrates fluid dynamic pressure NTN’s composite material products include mainly
technology and technologies of NTN’s related resin, sintered metal, and magnetic materials. These
companies, which include NTN Engineering Plastics products are produced by the NTN’s four related
Corp., NTN Powder Metal Corporation and Nippon companies shown in Fig. 1. Key products developed by
Kagaku Yakin Co., Ltd., to develop unit module the NTN companies are introduced in Section 3 and the
products to be used in the automotive and industrial subsequent sections. Various technologies/products
industries. Currently, NTN is working specifically on the are also covered by different articles.
following activities, focusing on the growing markets
such as electric vehicles and energy:
BEAREE®
NTN Engineering Plastics Corp.
Composite Material Products Division
1) Development and proposal of not only components

Resin sliding bearings, injection molding products,
but unique unit products by combining resin and resin machining, coating components
sintered metal
BEARPHITE®
2) Development of light and compact sliding bearings
NTN Powder Metal Corporation
as an alternative to rolling element bearings Sintered oil-impregnated bearings,
3) Sintering of machine parts which require high sintered machine parts
strength NIKAMET®
4) Product development for industrial and automotive Nippon Kagaku Yakin Co., Ltd.
industries applying magnetic material technologies Magnetic material products, resin molding products,
sintered machine parts
5) Proposal of “only one” technology including
Hydrodynamic BEARPHITE
amorphous magnetic material into growing markets
NTN Manufacturing (Thailand) Co.,Ltd
Hydrodynamic bearing units
Fig. 1 Products and production system of composite
material products
* Composite Material Products Division
-12-
Product Introduction and Complex Technology of Resin, Sintered Metal and Magnet for Growth Markets
inferior crack resistance against impact. The cost is also

3. Resin material products a challenge. Since BEAREE AS5704 bearings are
Resin components have a high degree of freedom for mainly made of PPS resin, they have superior
design and are used in diverse fields as alternatives to properties in self-lubricity, chemical resistance, and a
metal components since they are inexpensive and can high degree of freedom in design using injection
be made lightweight, compact, and corrosion resistant. molding. It is easy to provide lubrication grooves on the
NTN develops and designs resin materials, specifically internal surface of the bearing bores, end faces, and
working toward higher product functionality by retainers (D-cut, projection, etc.) on the outer surface of
leveraging the resin properties and improving bearing the bearings which can be molded into a single piece
load properties by creating composite material with without the machining process.
metal. The following is the recent resin material:
(1) Resin sliding bearings1) for electric water pump Water

Use of electric and hybrid vehicles is increasing Coil
Impeller
because of to the reduction in CO2 emissions. Magnet
Conventional vehicles driven by internal combustion
Thrust pad Thrust pad
engines are equipped with centrifugal pumps for
circulation of radiator water for engine cooling. These
are driven by the engine rotation which is transmitted to Water
the pump shafts through the engine belts. Contrary to
this, electric vehicles do not have engines and hybrid Stationary shaft
vehicles stop engines with a start/stop function.
Therefore, electric pumps are required for inverters or Bearing
engine cooling systems. These electric water pumps Fig. 2 Structure of magnet drive centrifugal pump
are mainly high-efficient and compact centrifugal pumps
with magnetic drives.
Fig. 2 shows the structure of the magnetic-drive
centrifugal pump. This structure allows fluid, such as
antifreeze liquid, to flow by rotating impellers integrated
with magnetic casing using the magnetic force of the
magnet installed on the motor shaft as it rotates. As
shown in Fig.3, BEAREE AS5704 made of
polyphenylene sulfide (PPS) resin with special filler
blended in is used as the bearings in the center of the
impeller. These bearings have sliding contacts
between the internal bore of the bearing and the
Fig. 3 Plastic sliding bearings for electric water pump
stationary shaft and the bearing side face and the thrust
(BEAREE AS5704)
pad. These sliding contacts support the radial and axial
loads respectively.
[Features] Table 1 Comparison of features various bearings
(1) Friction in the water is 1/5 or less compared with Types of BEAREE Phenol resin PTFE
bearings AS5704 Carbon (graphite blended) (graphite blended)
sliding bearings made of universal PPS resin
Machining Injection Injection/
(2) Reduced friction damage to mating material, Machining compression Machining
method molding molding
such as stainless steel
Wear resistance
(3) Larger degree of freedom for design due to (under water) ◎ ○ △ ×
injection molding Chemical
◎ ◎ △ ◎
resistance
(4) Compatible with antifreeze, acid, and alkaline
Water absorption
fluid dimensional stability ○ × △ ○
Crack by impact ○ × × ○
Table 1 shows a comparison of performance among Deformation by
impact ○ ○ ○ ×
typical sliding bearings used in fluid. Bearings made of
Degree of freedom
carbon excel in wear resistance in fluid; however, they for design ◎ × ○ ×
show a lower degree of freedom in shape because they Price ◎ × ○ ×
are machined from formed materials and possess ◎ Excellent ○ Good △ Fair × Poor
-13-
Fig. 4 shows the specific wear rate of BEAREE results in a longer life when compared to general-
AS5704 and various PPS resin materials with glass purpose resin sliding screws.
fiber, carbon fiber, and PTFE resin blended in. NTN has recently developed resin sliding screws for
AS5704 is superior in wear resistance when high load, which have been limited to lower load
compared to other PPS materials because it has applications, and proposing them for a cleaner
special filler materials added to the PPS resin. production environment and increased load capacity.
NTN has created a composite nut applying PPS resin
sliding material onto the surface of threads of a brass
500 nut as indicated in Fig. 6, by injection molding, and
increased the heat transfer and load capacity. The
Specific wear rate
400 brass has a special surface treatment to improve

300
adhesion with the resin. This composite structure
improved not only the bearing load capacity but also
200 durability.
×10-8mm3 100 The resin sliding screws for high load have twice the
N・m axial load capacity and 9-times higher nut static
0 breaking load when compared to the conventional
BEAREE PPS resin PPS resin PPS resin sliding screws. In addition, the life under high load
AS5704 + + +
glass fiber carbon fiber PTFE resin
condition is 3 times longer than the conventional resin
Fig. 4 Wear factor of PPS bearings sliding screws.
[Features]
Comparison with conventional products (NTN
(2) Resin sliding screws for high load 2) resin sliding screw)
Feed screws, which convert rotational motion of the (1) Permissible axial load: 2 times
motor to linear motion, are used for carrying units of (2) Nut static breaking load: 9 times
medical and food equipment. Feed screws include ball (3) Friction: 1/3
screws and resin sliding screws.
Table 2 shows the comparison of their performance.
Ball screws are superior in load capacity and screw
efficiency, although they are an expensive option. In
addition, as they are lubricated with grease, they are Resin nut
not suitable for use where dispersion of grease and
deterioration due to high temperatures are a concern.
On the other hand, resin sliding screws have lower load
capacity but they do not require lubrication allowing for Screw shaft
use in a broader range of applications, including
vacuum or high-temperature ambient. They are also
less expensive and quieter during operation.
NTN’s sliding screws made of precision resin consist Fig. 5 Resin sliding screws
of PPS resin nuts and stainless steel screw shafts as
shown in Fig. 5. Due to the PPS resin having improved Composite nut
wear/friction property by blending solid lubricant, it also
Bronze PPS resin
Table 2 Comparison of features various sliding screws

Resin sliding Resin sliding screw
Item Ball screw screw for high load
Not required Not required
Lubrication Required (can also be used (can also be used
(grease) with lubrication) with lubrication)
Load resistance ◎ △ ○
Screw shaft
Screw efficiency ◎ △∼○ ○
Noise △ ○∼◎ ◎
Cross section structure
Heat resistance △ △∼○ ○
◎ Excellent ○ Good △ Fair Fig. 6 Resin composite sliding screws for high load
-14-
(3) Hybrid PEEK bearings 2) The hybrid PEEK bearings have the following features
The majority of power consumption of an air compared to conventional PTFE resin wrapped bushings:
conditioner is occupied by the compressor. Therefore, [Features]
higher efficiency is required to reduce energy Compared to conventional products (PTFE resin
consumption of an air conditioner. Fig. 7 shows the wrapped bushing)
structure of a compressor and the installation positions (1) Friction coefficient: 2/3, wear: 1/3
of the bearings. The bearings are required to have (2) Heat resistance: more than 5 times
improved friction/wear properties and resistance against
heat (the refrigerant liquefied by over-compression can
4. Sintered metal products
make the refrigerant oil leak from friction).
Therefore, NTN is developing hybrid PEEK bearings NTN Powder Metal Corporation works on high
with improved friction/wear properties and resistance functionality of conventional oil-impregnated sintered
against being burnt using composite material of bearings and develops high-strength machine parts
sintered alloy and PEEK resin. mainly for automotive applications. In the automotive
Fig. 8 shows the structure of hybrid PEEK bearings. applications, sintered products are increasingly used as
Different from polytetrafluoroethylene resin wrapped a part of the component cost reduction, which accounts
bushings made by the impregnation/sintering process for approximately 60% of the sintering market or
on the sintered intermediate layer of a steel plate, these approximately 9.2 kg per vehicle. The following is the
bearings are made with a layer of 0.5 mm thick PEEK introduction of high-strength sintered materials and
resin applied on the internal surface of the sintered alloy various types of sintered bearing products developed
bushing by injection molding. PEEK resin is suitable for for this growing market:
injection molding and excels in heat, wear, chemical,
and fatigue resistance. In addition, it also considerably (1) High-strength sintered materials 4)
improved heat transfer properties in the composite Sintered machine parts for automotive applications
structure with metal compared to the bearings with only include engines, drive components, and in-vehicle oil
PEEK resin material. pumps and starter motors. Fig. 9 shows an example of
sintered gears. In addition, Fig. 10 shows the results
from a comparison test of ring compressive fatigue
strength of high-strength sintered materials. The fatigue
strength of the developed products is about 2.3 times
Fixed scroll that of the conventional products. The sintered machine
parts require not only high-strength but also support for
Revolving scroll
high precision and the ability for complex shaping.
Thrust bearing
Additionally, we are optimizing material composition,
manufacturing method, and thermal processing to
Bearing expand applications starting from replacement of cold
Motor
forging materials.
[Features]
(1) True density ratio: achievement of 96%
Main shaft
(2) Fatigue strength: 2.3 times compared to the
conventional products
(3) Young’s modulus: achievement of 180 GPa
Fig. 7 Structure of air conditioner compressors
Sintered alloy bush

0.5mm
Example : Thinner gear

Left : Development
PEEK resin
Fig. 8 Structure of hybrid PEEK sliding bearing Fig. 9 High strength sintered gear
-15-
Fatigue strength MPa [Features]

True density ratio % (1) Simultaneous forming of two types of powder
8.0 600 materials
7.8 340
(2) Different properties between the inner layer
Fatigue limit
Density
7.6 96 500 400

150 and outer layer
7.4 100
91
200
g/cm3 7.2
7.0 Fig 12 shows comparison of the friction coefficients
MPa
6.8 0 between this newly developed product and
Conventional Developed Bearing
product product steel conventional product (bearings that use the inner layer
Density g/cm3 7.1 7.5 7.8 material of the newly developed product only).
True density ratio % 91 96 100
This developed product uses the high-wear resistant
Fatigue limit MPa 150 340 500
material in the inner layer only but has the equivalent
Fig. 10 Ring compressive fatigue strength friction property as the conventional product, which
uses the same material for the entire product.
(2) Multi-layer sintered oil-impregnated bearings 5)

With increased infrastructure development in the 0.30
world including emerging markets, demand for Developed bearings
0.25
construction machines, such as hydraulic shovels, is Conventional bearings
Friction coefficient
also increasing. For joints of these hydraulic shovels, 0.20
use of sintered oil-impregnated bearings is increasing 0.15

for longer oiling intervals. Higher strength is also
0.10
required due to severe operating conditions.
Conventionally, oil content is secured in low density to 0.05
prolong oiling intervals while strength and hardness are 0

0 1 2 3 4 5
improved by thermal processing, then dimensional
accuracy is achieved by machining the bearings. Time h
The multi-layer sintered oil-impregnated bearings Fig. 12 Friction coefficient measurements

shown in Fig. 11 have two types of powders with
different properties molded simultaneously into one
piece so that the inner layer has sliding and wear (3) Corrosion resistant sintered bearings 4)
resistant properties while the outer layer has a high Bearings for valves in fuel pumps and exhaust gas
strength property. This process can eliminate the need recirculation (EGR) systems require high corrosion
for thermal processing such as carburized hardening resistance as they are in contact with fuel, etc. The
and the machining process to achieve a significant typical material for this is cupronickel-based materials.
reduction of the manufacturing workload. However, this material contains a rare metal, nickel,
with the requirement for high corrosion resistant
bearings without rare metal increasing. NTN has
focused on aluminum bronze which exhibits good
corrosion resistance resulting in corrosion resistant
bearings with better corrosion resistance and sliding
property than cupronickel-based sintered bearings.
Fig. 13 shows an example of bearings for a fuel
pump.
[Features]
Outer layer: high-strength structure material (1) Corrosion resistance equivalent or superior to
cupronickel-based material
(2) Improved corrosion resistance without using
rare metal
(3) Significant improvement of wear resistance
over cupronickel based material
Inner layer: bearing part; high wear resistance
Fig. 11 Multi-layer BEARPHITE for joints Table 3 shows the test results of corrosion resistance
of hydraulic excavators against organic acid and sulfur. The test piece was
-16-
acceleration/deceleration, which requires step-up

Fuel pump
(amplification) of the dedicated battery output voltage.
Therefore, step-up reactors are used for the system
circuit components, which should support large
to engine current/high frequency as well as have a compact size
for reducing electric cost. The current hybrid vehicles
use Fe-Si (ferrite-silicon alloy powder) with the
compression manufacturing method. In addition, for
larger current to be applied with higher EV output, Fe
Fuel tank (pure iron powder) products with high saturation
magnetic flux density manufactured using the
Fig. 13 Sintered bearing for fuel pump
compression method are being considered. However,
these materials have large iron loss with large heat
Table 3 Result of anti-corrosion examination
generation and the step-up reactor tends to become
Weight change rate % larger in order to avoid this.
Developed Cupronickel
material based material Fig. 14 shows the performance comparison of
(1) Corrosion resistance test different core material for step-up reactors. The Fe-Si
against organic acid 0.14 0.15
and Fe products made by the above mentioned
(2) Corrosion resistance test
against sulfur 0.10 0.40 compression method exhibit large iron loss. In
comparison, the amorphous products by Nippon
Kagaku Yakin Co., Ltd. have smaller iron loss and can
immersed in test fluid at the set temperature and time be manufactured by injection and compression molding.
and the variation was calculated from the weight The newly developed “hybrid amorphous core”, in
difference before and after test. particular, is a combination of injection amorphous core
<Test condition> and compression amorphous core. Fig. 15 shows the
(1) Corrosion resistance test against organic acid structure of hybrid amorphous core. Compression
・Test fluid: Organic acid concentration 2%
・Temperature: 50˚C
・Test time: 100 hours Compressed
Fe material
50 Compressed
(2) Corrosion resistance test against sulfur µ :Relative permeability Fe-Si material (µ100 ∼ 300)
(µ60 ∼ 100)
・Test fluid: Gasoline with 300 ppm of sulfur Injected amorphous
40 (µ10)
・Temperature: 80˚C
Iron loss
・Test time: 300 hours 30 Injected/compressed

Hybrid amorphous
(µ20 ∼ 40)
The comparison between the newly developed material 20
kW <Reference> ferrite
and cupronickel based material revealed favorable (µ700 ∼ 10,000)
m3 10 Compressed amorphous
results, with an equivalent friction coefficient and the
(µ60)
specific wear rate approximately 1/10 of the cupronickel 0
based material. 0.0 0.5 1.0 1.5 2.0 2.5
Saturated magnetic flux density T
5. Magnetic material products Fig. 14 Comparison of features various core material
Nippon Kagaku Yakin Co., Ltd. develops and

promotes magnetic material products aiming at the Compression amorphous core Injection amorphous core
converting mechanical components into functional
components. 6)
In this article, we are introducing products for growing
the market, particularly by applying its unique
technologies.
(1) Reactors for automobiles 7)

The motor drive of the recent hybrid vehicles use
dedicated batteries. These vehicles have achieved
drive performance equivalent to fuel-driven engines in Fig. 15 Structure of hybrid core
-17-
amorphous with small iron loss is used for the area of (2) Dual track magnetic encoder
electric-magnetic conversion for lower heat generation Industrial machine robotics require high positioning
and injection amorphous with superior shaping flexibility accuracy and use high-resolution absolute angle
is used for the magnetic conduit for retransmitting the sensors which requires a highly accurate encoder.
converted magnet to the coil. Fig. 16 shows an NTN is introducing our newly developed dual track
analytical example of heat generation of the hybrid magnetic encoder. Fig. 17 shows the structure of this
core, which indicates appropriate placement of the newly developed product. Hard magnetic material is
respective cores. In addition, Table 4 shows a injected on the core metal of compressed ferrous soft
comparison of hybrid core performance against magnetic material. This diagram shows an axial type
injection and compression amorphous cores, dual track encoder with inner and outer ring patterns of
respectively. The hybrid core exhibited favorable results a different number of accurately magnetized poles. The
in any property enabling manufacturing of pod-shaped following are the features of the developed product:
core, which is difficult with compression amorphous [Features]
core. (1) Supports opposite angle detection
[Features] (2) Angle accuracy: ±0.2˚
(1) Can be used in applications where large (3) Supports hollow structure
current or high frequency is required
(2) Compact and low heat generation by In addition, Fig. 18 shows an absolute angle
composite structure measurement method using this developed dual track
(3) Higher degree of freedom for reactor shape magnetic encoder. Two sensors facing inside and
design by injection molding outside magnetized rings, respectively, detect the
signals. The absolute angle can be accurately
measured by calculating the relative displacement of
the patterns from the magnetic sensor signals, using a
Compressed core Injected core principle called caliper or Vernier principle.
125℃
Hard magnetic material:

dual track magnetization
Coil
Fig. 16 Temperature distribution analysis of hybrid core Ferrous soft

magnetic mandrel
Fig. 17 Structure of dual-track magnetic encoder
Table 4 Comparison of features various reactor core
Hybrid core coil
Single core coil (newly developed New sensor IC (off the shelf product)
(conventional product) product) Example of detected signals
Sensor 1 (NS poles can be precisely divided)
Compression core Injection core Hybrid core
8 periods
Sensor 2
For large △ ○
(large press required) ○ 7 periods
current (DC bias: good)
0° 90°
For high Rotation of shaft
frequency ○ △ ○
(low loss)
Calculate angle position from the
○ × displacement (phase difference) of two
Compact ○ magnetic patterns (caliper principle,
(Permeability: high) (Permeability: low)
Vernier principle)
○ Dual track magnetic encoder
Heat
suppressed (Heat transfer: good)
× ○ Consists of two rings, inner and outer;
Different number of poles for each ring
Flexible △ ○
shaping ○
(CNC press required) (Injection molding: allowed)
Fig. 18 Absolute angle measurement system
-18-
(3) Choke coil for MRI 6) This is used in this developed product and the
The MRI shown in Fig. 19 is a medical device used generally-used ferrite/iron-silicon (Fe-6.5Si) soft
to provide images of sections of the human body using magnetic material.
large magnetic energy. The choke coil integrated in this In addition, Fig. 23 shows the frequency response of
device is an electronic component to generate large the inductance when weak current is applied. From
current for creating a gradient magnetic field during these properties, the following features can be pointed
testing. out and this choke coil is suitable for this application
Fig. 20 shows the components used for the developed which requires large current and stable high-frequency
choke coils. The coil, covered by an insulating case, is response characteristics:
incorporated in the injection amorphous core to make the [Features]
choke coil shown in Fig. 21. (i) Inductance deviation rate is small even when
Fig. 22 shows the inductance deviation rate (DC bias high current is applied compared to the
characteristic), which serves as a measure of the conventional ferrite and Fe-6.5Si.
magnetic field generation capability when applied (ii) Change rate of inductance in the high
current is swept for amorphous soft magnetic material. frequency range is smaller than ferrite.
120 Amorphous
Rate of change of inductance %
Fe-6.5Si
100
Ferrite
80
60
40
20
Fig. 19 Medical magnetic resonance imaging

0
0 100 200 300
Current A
Fig. 22 DC bias characteristics
120
Rate of change of inductance %
100
a）Coil b）Case c）Injection amorphous
Fig. 20 Parts of choke coil 80
60
40
Amorphous
20 Fe-6.5Si
Ferrite
0
1 10 100 1000
Frequency kHz
Fig. 23 Frequency characteristics
Fig. 21 Choke coil for MRI power supply
-19-
6. Conclusion
In this article, new products with composite materials
for growing markets are mainly introduced.
With increased fuel efficiency and use of electric
drive components in automobiles, being lightweight is a
focal point in the development of components and
composite material products.
In addition, NTN is also proposing new products to
contribute to the development of our customers in the
broad field of industrial machines, including medical
devices, food equipment, robotics, etc.
NTN is poised to actively develop products by
integrating a wide range of materials and technologies
in response to the trend of the consistently growing
markets and contributing to these markets.
References
1) Norio Itoh, Takuya Ishii: Plastic Bearings of Electric
Pump for the Next Generation Cell, NTN TECHINICAL
REVIEW, No. 77 (2009) 51-55.
2) Takuya Ishii, Yoshihide Himeno: Functional
Improvement of Resin Sliding Bearing Through the
Combination with Metal, Journal of the Japan Society of
Polymer Processing, 25 (2), 73, 2013
3) Takuya Ishii, Ken Yasuda: Hybrid PEEK Sliding
Bearing, NTN TECHNICAL REVIEW, No. 79 (2011)
125-129.
4) Tomonori Yamashita, Tomokazu Sonozaki: The
Introduction of Sintered New Products for Automobile,
NTN TECHNICAL REVIEW, No. 81 (2013) 74-77.
5) Yosuke Sugai, Toshihiko Mouri: Multi Layer
BEARPHITE, NTN TECHNICAL REVIEW, No. 80
(2012) 83-86.
6) Takuji, Harano, Shinji Miyazaki, Hajime Katsuura:
Introduction of Magnetic Material Products, NTN
TECHNICAL REVIEW No. 80 (2012) 87-91
7) Takayuki Oda, Shinji Miyazaki, Eiichiro Shimazu: The
Reactor Core for HEV Boost Converter, NTN
Photo of author
Yoshio OKI
Composite Material
Products Division
-20-
Takahiro OKUNO*
Naoki YASHIRO*
Sintered machine parts are used in many industrial fields. Especially,

these are widely used for automotive parts. It is essential for the
increasing use to improve mechanical properties of sintered parts.
However, the pores in the sintered body are dominant factor for the
strength of sintered parts. Therefore, various kinds of trial to decrease
pores in the sintered body, or to increase sintered density were reported.
This article introduces the method of improvement of density and strength, and the mechanical
properties of high-density sintered machine parts.
1. Introduction 2. Sintered alloy steel

Sintered machine parts have been adopted in diverse Sintered alloy steel is used in many sintered machine
fields as they can be made into complex shapes and parts 2).
are suitable for inexpensive, large volume production. Powders for sintered alloy steel have a particle size
The production volume of sintered machine parts in of approx. 300 μm or less and include alloy elements
Japan amounted to approx. 90,000 tons in 2013 and such as copper (Cu), nickel (Ni), chromium (Cr) and
more than 90% of that volume was for transportation molybdenum (Mo), in addition to iron (Fe). Other
machines. elements include those specified in JIS Z 2550, which
Recently, in particular, sintered machine parts have complies with the ISO Standards that specify their
been actively used in the automotive industry, with 9.0 chemical contents. The specification for sintered alloy
kg used per vehicle; 50% of these are for the engine. 1) steel is different from molten steel. For example, Cu,
In order to expand the scope of sintered machine which is treated as an impurity in molten steel, is
part applications, improvement in their mechanical actively used in sintered alloy steel for improving
properties is vital. However, since sintered compact is sintering properties at low temperature and enhancing
porous, ordinary sintered machine parts are not as sliding properties. Also, carbon (C) is added as graphite
strong as molten steel parts. Therefore, reduction of powder to be solidified into iron during sintering. This is
pores in the sintered compact, namely, by methods to to avoid significant deterioration of formability if graphite
increase density, is being attempted to improve is solidified into iron in the powder form as the powder
mechanical properties. becomes hardened. Material manufacturers are
In this paper, we will present methods to make the developing proprietary powders outside the
sintered machine parts higher density/higher strength specification so that the optimum materials can be
through materials or compacting and machining selected depending on the intended use.
processes and improvement in mechanical properties. The base metal powder is either mixed powder with
all the additive elements, completely alloyed powder
with additive elements pre-alloyed and homogenized, or
diffusion alloyed powder which uses pre-alloyed
additive elements with no negative effect of compaction
and the remaining additive elements diffusion bonded
by thermal diffusion.
* Advanced Technology R&D Center
-21-
Table 1 shows the classification of metal powder of 2.1 Fe-Cu-C base

typical sintered alloy steel; Table 2 shows the features Fe-Cu-C based material is widely used as sintered
of completely alloyed powder and diffusion alloyed material for machine parts. Cu has a melting point of
powder; Fig. 1 shows SEM micrographs. The following 1085˚C and melts under the general sintering
are typical sintered alloy steels used for machine parts. conditions of sintered alloy steel.
Therefore, Cu exists in the liquid phase between iron
particles and is alloyed as sintering progresses. This
results in high-strength. In addition, Cu is known for
Table 1 The classification of metal powder of typical
sintered alloy steel volume expansion when sintered with Fe powder 3). It
can be used to control size change since the volume
Powder type Sintered alloy steel
generally shrinks with sintering. Mixed powder is
Mixed powder Fe-Cu-C base
frequently used mixing pure iron powder with pure
Completely alloyed Fe-Ni-Mo-C base
powder Fe-Cr-Mo-C base copper powder or graphite, etc.
Diffusion alloyed Fe-Ni-Mo-C base
powder Fe-Cu-Ni-Mo-C base 2.2 Fe-Ni-Mo-C base
Fe-Ni-Mo-C based material is used to obtain high-
strength sintered compact by various tempering and
Table 2 Feature of completely alloyed powder and annealing processes after sintering. Wear resistance
diffusion-alloyed powder can also be obtained by applying a surface hardening
Completely alloyed Diffusion alloyed process in addition to high-strength. Improvement of
powder powder steel toughness can also be expected by Ni. Ni is
Homogeneity of powder High Low frequently used with completely alloyed powder or
Compaction property after diffusion alloyed powder.
forming process Low High
2.3 Fe-Cu-Ni-Mo-C base

Fe-Cu-Ni-Mo-C based material is made by diffusion
bonding of fine powders of Cu, Ni, Mo, etc. to pure iron
powders. This is done to combine the features of the
above two previous materials. It is currently widely used
as high-strength sintering material since it has both high
compaction and excellent mechanical properties.
2.4 Fe-Cr-Mo-C base

Cr, as an element for alloy, is generally used as
(a) Completely alloyed powder machine structural parts in molten steel 4). On the other
hand, when used as sintering material, it is difficult to
reduce oxygen content as it has high affinity with
oxygen. This means that high compaction and high
density are difficult to obtain. However, with recent
technological progress by material manufacturers,
powders with low oxygen content are developed with
improved compaction properties. Fe-Cr-Mo-C based
material is used in completely alloyed powder due to
(b) Diffusion alloyed powder the above mentioned affinity with oxygen.
Fig. 1 SEM micrographs of completely alloyed powder
and diffusion-alloyed powder
-22-
3.4 Infiltration
3. Methods for densification/strengthening Infiltration is a method to infiltrate melted material of a
Densification methods can be performed during low-melting point into the pores of a sintered compact.
molding, sintering and after sintering. Table 3 shows Material is communicated from the surface using
those densification methods in each process. capillary phenomenon to fill in the pores 10). Specifically,
the sintered compact and the green compact of low-
melting point material are placed in contact with each
Table 2 The method of densification at the production process other and heated to a temperature more than the
Process Method melting point of the green compact and less than the
Warm forming melting point of the sintered compact for infiltration.
During molding
Warm forming with mold lubrication, etc. Since it is possible to close the pores in the sintered
2 forming 2 sintering compact, it is used in the applications where high air-
During sintering
Infiltration, etc.
tightness is required such as compressor components
After sintering Sinter forging, etc.
or hydraulic pressure resistant components.
3.5 Sinter forging

3.1 Warm forming Sinter forging is a method to obtain product by
Warm forming is a method to obtain high-density forming and sintering metallic powder, and inserting the
green compact by heating the powder and mold to 80- heated sintered compact into a mold for closed die
300˚C during molding of metal powder. This reduces forging 10). A regular forging method, such as hot
the yield strength of metal powder and increases forging of sintered compact, with a shape close to the
compaction property 5). formed product, is also adopted. Since it is possible to
Since this method is lower in cost compared with the obtain a value close to the true density, it is used in
2-forming 2-sintering method and sinter forging method connecting rods of vehicle engines, etc.
described later, it is used for sprockets of vehicle
engines, etc. 6)
4. Development of high density sintered
machine parts
3.2 Warm forming with mold lubrication
In general, metallic soap, such as Zn and Ca, or a NTN is working on development of high density
resin based lubricant is added/mixed to reduce friction sintered machine parts with potential for various
between the metal powder and mold when metal applications. The following is an example of mechanical
powder is molded; however, since the lubricant properties of the developed materials.
disappears during sintering, it inhibits densification.
Warm forming with mold lubrication is a method to 4.1 Method for making test piece
reduce lubricant by pre-applying the lubricant to the We made a 23.2 mm OD×16.4 mm ID×7 mm H
mold, in addition to the above warm forming method, to ring-shaped test piece by warm forming of 80˚C from
obtain high density green compact 7). Since it is the Fe-Ni-Mo-C based material. This was made of
possible to obtain higher density sintered compact than diffusion alloyed powder of Fe and diffusion bonded Ni
the warm forming method, it is used in reactors of and Mo with a predetermined amount of graphite mixed
hybrid vehicles, etc. 8) with lubricant, using floating die method. Next, the
green compact was sintered in a mixed atmosphere of
3.3 2-forming 2-sintering N2 and H2, then carburized tempering/annealing was
Also called 2P2S (2Press-2Sinter). This is a method applied. In addition, we prepared Fe-Cu-Ni-Mo-C based
to obtain high density sintered compacts by first forming material as conventional high-density sintered compact
powder under a predetermined pressure and applying as a control.
preliminary sintering with a relatively low temperature of
1000˚C or below. Then the compact is formed under a 4.2 Mechanical properties
predetermined pressure again and sintered with a Fig. 2 shows sintered density and radial crushing
higher temperature than the preliminary sintering. This strength. The sintering density of the developed
is used in synchronizer hubs of vehicle transmission, material is 7.63 g/cm3 which is approx. 97% of true
etc. where strength is required 9). density of 7.8 g/cm3. The radial crushing strength is
2180 MPa which is an improvement of 30% or more
from the 1630 MPa of the control. Fig. 3 shows the
cross section micrograph of the sintered compact. It
-23-
8.5 2500
Radial crushing strength MPa

8.0 2000
Sintered density g/cm3
Repetitive load
7.5 1500
7.0 1000
6.5 500
6.0 0
Fig. 4 Schematic the method of measuring ring
Developed Control compressive fatigue strength
material
Fig. 2 Sintered density and radial crushing strength
σ
σmax
Stress amplitude
σmin
Average stress
Fig. 5 The stress wave form at fatigue test
600
(a) Developed material (b) Control Developed material
Density 7.63g/cm3 Density 7.01g/cm3
Stress amplitude MPa
500 Control
Fig. 3 Optical micrographs of sintered part
400
can be observed that the number and size of the 300
internal pores are both reduced.

200
We evaluated the fatigue properties of the high-
R=0.1
density sintered compact with the ring compression 100
fatigue test. The ring compression fatigue test evaluates 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08
fatigue strength by repetitively applying compressive Number of repetitive loads

load on the ring outer surface which produces tensile
Fig. 6 Stress-endurance diagram
stress on the inner surface of the ring in the direction of
the load. Fig. 4 shows an overview of the measuring
method. The test condition was set with a stress ratio
R=0.1 and load frequency 50 Hz. The fatigue strength 4.3 Improvement of mechanical properties
was defined as the stress amplitude when the repetitive by cold working
load of 107 times did not break the test piece. Here, the We attempted to improve mechanical properties of
stress ratio is a ratio of the minimum stress over the the sintered compact made in Section 4.1 by applying
maximum stress and the stress amplitude is the cold working to reduce pores. The cold working was
amplitude of the variable stress waveform shown in applied after sintering and before thermal processing.
Fig. 5. Fig. 6 shows the stress-endurance diagram of Fig. 7 shows an overview of cold working. The sintered
the fatigue test. The fatigue strength of the developed compact is placed between the outer restraint jig and
material was 405 MPa, which was 2.25 times the the compression jig. The sintered compact was rolled
control (180 MPa). The fatigue strength of the by 0.25 mm, pressing the sintered compact by rolling the
developed product is equivalent to molten steel, compression jig; then carburized tempering/annealing
SCM440 thermal refined steel. was applied.
-24-
Fig. 8 shows the cross section micrograph. By this References

cold working, the compact was densified in the area 0.5 1) Kazunori Arai, Toru Shimizu: SOKEIZAI Industrial
mm from the surface, which is twice the rolling amount Yearbook 2013, 5. Powder Metallurgy, Sokeizai, 55, 5
with porosity of less than 0.5%. The fatigue strength (2014) 52-58.
after cold working was 450 MPa, which is more than a 2) Edited by: Japan Society of Powder and Powder
Metallurgy: A Handbook of Powder and Powder
10% improvement from 405 MPa before cold working.
Metallurgy, Uchida Rokakuho (2010) 236.
3) Hidenori Kuroki, Shinya Fujisawa: The Mechanism of
Dimensional Changes of Iron-Copper Powder Compacts
during Sintering, Powder and Powder Metallurgy, 38, 5
(1991) 571-576.
4) Edited by The Japan Society for Heat Treatment:
Guidebook for Heat Treatment, Taiga Publishing (2008)
Load
13.
5) Hironori Suzuki, Yoshikazu Seki, Kazuhisa Fujisawa:
Properties of "KHD-Segless" Warm Compaction Steel
Powder, R&D Kobe Steel Engineering Reports, 52, 3
(2002) 70-73.
Compression jig
6) Akira Fujiki, Yukihiro Maekawa, Yoshimi Sugaya,
Outer restrained jig
Takashi Shibano: Development of Warm Compacted
Test piece P/M Sprocket for Automotive Engines, Journal of the
Fig. 7 Schematic cold working Japan Society of Powder and Powder Metallurgy, 51, 7
(2004) 507-514.
7) Mikio Kondoh: Development of High Accuracy, High
Density Powder Compaction Technology, Journal of the
Japan Society of Powder and Powder Metallurgy, 54, 7
(2007) 506-512.
8) Takeshi Hattori, Shin Tajima, Mikio Kondoh, Masaki
Sugiyama, Toshiya Yamaguchi, Hidefumi Kishimoto,
Tomo Okochi, Takanobu Saito, Satoshi Takemoto,
Yuichiro Fujita: Development of High-Density and Low-
loss Powder Magnetic Core for Reactor in Hybrid
Vehicles, (1) Si volume of Fe_Si based compact core
and impact on molding pressure, Proceedings of Japan
Society of Powder and Powder Metallurgy, Autumn
2009 Conference, Nagoya 2009, (2009) 99.
9) Edited by: Japan Society of Powder and Powder
(a) Before cold working (b) After cold working Metallurgy: A Handbook of Powder and Powder
Fig. 8 Optical micrographs of sintered part Metallurgy, Uchida Rokakuho (2010) 245.
10) Edited by: Japan Society of Powder and Powder
Metallurgy: A Handbook of Powder and Powder
Metallurgy, Uchida Rokakuho (2010) 154.
5. Conclusion
In this paper, we presented
densification/strengthening methods of sintered
machine parts and mechanical properties of high-
density sintered machine parts. We have obtained high- Photo of authors
density sintered compacts by applying optimum
methods in each process and we have obtained fatigue
strength equivalent to molten steel SCM440 thermal
refined steel. In addition, it is possible to further improve
strength by applying cold working, etc.
We are poised to expand application of this
developed material to the machine parts that require
high strength, as well as to continue developing
materials with even higher performance and expanding Tadahiro OKUNO Naoki YASHIRO
Advanced Technology Advanced Technology
their applications as well. R&D Center R&D Center
-25-
[ New Product ]
High Performance Magnetic Core

for Induction Hardening Devices
Takuji HARANO*
Nihon Kagaku Yakin Co., Ltd. of the NTN group develops various
magnetic materials satisfying demand properties. In this report, it is
attached to the heating coil part of the induction hardening device and
introduces about a high-performance magnetic core in a magnetism
characteristic and a mechanical characteristic to be used for control of
the quenching depth.
1. Introduction 2. Magnetic core for induction

hardening heating coils
Machine structural parts, such as bearings and
gears, are strengthened by the thermal process for One of the induction hardening parts for automobiles
improving reliability. The drive parts for automobiles are is a drive shaft. Fig. 1 shows some examples of its
required to have strength in certain areas of complex application.
shapes; therefore, the best thermal process is applied A magnetic core installed on the back of the heating
to the required area with induction hardening. coil of induction hardening equipment can accelerate
Magnetic material called "core" is attached to the induction heating by concentrating magnetic flux on the
heating coil of induction hardening equipment work.
("magnetic core"). In this paper, we are presenting a On the other hand, if it is installed on the front of the
high performance magnetic core that exhibits superior heating coil, it can block the magnetic flux to prevent
magnetic characteristics as well as mechanical strength heating where hardening is not needed. Therefore, it is
developed for induction hardening. an essential component for heating coil of induction
hardening equipment.
When the target work has a complex shape and the
depth of hardening area needs to be adjusted, induction
heating can be completed by adjusting the shape, size,
Drive shaft
Inner rolling surface

of outer ring Stem
Shaft
Induction hardened part

Fig. 1 Example of induction hardened parts (Drive shaft)
* Engineering Dept., Nihon Kagaku Yakin Co, Ltd.
-26-
High Performance Magnetic Core for Induction Hardening Devices
number, direction and position of the installed magnetic 3.3 Features and performance of developed
cores. This controls the hardening depth of the work. magnetic cores
Nihon Kagaku Yakin Co., Ltd. developed high 3.3.1 Structure and features
performance magnetic core by combining iron powder Fig. 3 shows the structure of the developed high
and thermosetting resin as shown in Fig. 2. performance magnetic cores.
We used iron powder for powder metallurgy as the
magnetic powder, added a small amount of
thermosetting resin with high adhesive effect, and
applied insulating layer on the iron powder surface by
granulation process. This is a composite material which
reduces the damage to the insulation layer by applying
compaction molding at low pressure and thermal
hardening.
This process brought a magnetic material that is
strong yet has low contacts among magnetic powders.
It has low iron loss, even in the high frequency range,
so that the material can be used in a broad frequency
range.
Fig. 2 High performance magnetic core
The features of the developed high performance
magnetic cores are as follows:
・Saturation magnetic flux density: 1300 mT
3. Development of high performance ・Relative permeability: 54
magnetic core ・Iron loss [10KHz/200mT]: 1480 kW/m3
3.1 Background of the development ・Radial crushing strength: 150 MPa
Commonly used ferrite based magnetic core has
relatively small iron loss even in the high frequency
range. However, a larger size may be required to Pore
maintain hardening performance due to its low
saturation magnetic flux density. Magnetic steel sheet
cannot be used, even if it has high magnetic flux
density, because the iron loss becomes larger in the
higher frequency range.
Commercially available cores or base material of
powder compact based cores manufactured by powder
metallurgy are frequently used as magnetic cores
installed on heating coils of induction hardening
equipment. They are expensive, exhibit low material Resin
Iron powder
strength, and are challenging to machine and handle.
Therefore, our task was to develop magnetic cores Fig. 3 Structure of the high performance magnetic core
(Image)
that have better characteristics than the conventional
material, can be produced with a simple manufacturing
process, and are suitable for volume production.
3.2 Development concept

Since the magnetic cores for heating coils of
induction hardening equipment are used in the high
frequency range (several kHz to 100 kHz), the following
characteristics are required:
(1) High saturation magnetic flux density
(2) High relative permeability
(3) Good frequency response (small variation of
inductance against variation of frequencies)
(4) Low iron loss
(5) Superior mechanical strength
-27-
3.3.2 Performance (ii) Variation of inductance

(1) Magnetic characteristics Fig. 5 shows the rate of inductance variation, which
(i) Saturation magnetic flux density, serves as a measure of the magnetic field generation
relative permeability capability in the high frequency range, among the
Table 1 shows a comparison of the saturation frequency characteristics of the magnetic material.
magnetic flux density and relative permeability of the The rate of inductance variation of the developed
developed and conventional products. Fig. 4 shows the material is equivalent to the conventional material in all
graph indicating the relation between the magnetic flux ranges, including the high frequency range, with both
density B and magnetic field strength H (B-H curve). materials showing small inductance variation up to
The developed product improved approximately 10% about 100 kHz.
in saturation magnetic flux density and approximately As a result, it was revealed that this characteristic is
30% in relative permeability over the conventional stable up to the high frequency range.
product. In addition, the rise of the magnetic flux density
B is faster over the magnetic field strength H indicating
its high efficiency. When the developed magnetic core 120
is installed on the back of the heating coil of the
Rate of inductance variation %

induction hardening equipment introduced at the 100
beginning of this paper, it can accelerate induction

80
heating by concentrating magnetic flux on the work.
60
40
Table 1 Saturation magnetic flux density and relative
permeability Developed material
20
Saturation magnetic Relative Conventional material
Material flux density
mT
permeability 0
1 10 100 1000
Developed material 1300 54
Frequency kHz
Conventional material 1200 40
Fig. 5 Frequency characteristics
1400
(iii) Iron loss
Fig. 6 shows iron loss of the developed and
1200
conventional materials in the high frequency range (10
Flux density B mT
1000 kHz) equivalent to the induction hardening condition.

The iron loss of the developed product at 200 mT
800
was approximately 15% smaller than the conventional
600 material indicating less energy loss; a favorable result.
400
200
Developed material
Conventional material
Frequency : 10kHz
0 3000
0 5000 10000 15000 20000 25000 30000
Magnetic field strength H A/m Developed material
2500
Conventional material
Iron loss kW/m3
Fig. 4 B-H curve 2000
1500
1000
500
0
0 50 100 150 200 250
Flux density mT
Fig. 6 Iron loss
-28-
High Performance Magnetic Core for Induction Hardening Devices
(2) Mechanical strength

The developed magnetic core is a strong material, 4. Conclusion
where iron powders are adhered with thermosetting In this paper, we presented the high performance
resin and hardened, as mentioned above. The magnetic core with superior magnetic and mechanical
mechanical strengths of the developed and characteristics to be installed on the heating coil of
conventional materials were compared by obtaining inductance hardening equipment.
radial crushing strength (Fig. 7), applying compressive The developed magnetic core showed superior
stress on the annular test piece, and using the following magnetic characteristics compared to the conventional
equation: product and significantly improved mechanical strength.
By applying this high performance magnetic core to
K ：Radial crushing strength (MPa)
inductance hardening equipment, we believe we can
F (D−e) F ：Maximum load at break (N)
achieve not only stable quality and improved reliability
K= L ：Length of hollow cylinder (mm)
L・e2 of inductance hardened components but also reduced
D ：Outer diameter of hollow cylinder (mm)
size of these materials through enhanced strength and
e ：Wall thickness of hollow cylinder (mm)
hardness of the structural components. This will
The result is shown in Table 2. The radial crushing contribute to manufacturing lighter vehicles.
strength of the developed material is approximately 5
times stronger than the conventional material.
References
Load 1) Takayuki Oda, Shinji Miyazaki, Eiichiro Shimazu: The
Reactor Core for HEV Boost Converter, NTN
2) Takuji, Harano, Shinji Miyazaki, Hajime Katsuura:
Introduction of Magnetic Material Products, NTN
3) Shin Tomogami: Technical Trends in Constant Velocity
Universal Joints and the Development of Related
Products, NTN TECHNICAL REVIEW, No. 75 (2007)
10-15.
4) Japan Electronics and Information Technology
Industries Association: Soft Magnetic Metal Materials
Fig. 7 Method for radial crushing strength measuremen
Table 2 Mechanical strength
Density Radial crushing strength

Material g/cm3 MPa
Developed material 6.1 150
Conventional material 6.4 30
Photo of author
Takuji HARANO
Engineering Dept.,
Nihon Kagaku Yakin Co, Ltd.
-29-
[ New Product ]
The Copper-Iron-based Material Equivalent to Bronze-based Material

for Sintered, Oil-impregnated Bearings
Tomonori YAMASHITA*
The materials of the sintered oil-impregnated bearing are classified into

bronze-based, copper iron-based and iron-based in general. Bronze based
materials are superior in sliding an acoustic feature, the problem is the
copper powder is high cost compared with iron powder, so it was necessary
to reduce its quantity. By adopting special copper powder, we developed
new materials having same or surpassing sliding and acoustic features as in
the existing bronze-based materials, even with the small amount of copper
inside. This report introduces the sintered oil-impregnated bearing using
developed materials.
between sliding properties, durability, and cost, from the

1. Introduction base copper-iron-based material.
Sintered, oil-impregnated bearings are widely used in
automobiles and industrial machines and their materials
3. Features of Special Copper Powder
are broadly classified into bronze-based, iron-based
and copper-iron-based, which combines the former two Table 1 shows the composition of the developed
materials. When superior sliding performance is material and Fig. 1 shows the cross section of a
required in a bearing application, such as electric parts bearing using a special copper powder based on this
(power windows, fan motors, etc.), copying machines, developed material composition. The upper side of the
or laser printers, bronze-based material is suitable. figure is the bearing bore surface, where a thin layer of
However, because of the recent increase in the cost of copper can be observed. The special copper powder
copper, NTN has developed a new material as an tends to aggregate on the surface of the powder
alternative to bronze for sintered bearings. In this paper, compact; therefore, a green compact can be produced
we present the properties of this copper-iron-based in the regular molding process without needing any
material equivalent to bronze sintered material for special processes, such as plating. On the other hand,
sintered, oil-impregnated bearings. many general copper and iron powders added in the
blending process are contained inside the compact.
Fig. 2 shows the bearing bore surface. Copper is
2. Material of Sintered, Oil-impregnated exposed on approx. 60% or more of the bore surface
Bearings which contributes to the excellent sliding properties.
Among sintered, oil-impregnated bearings, bronze-
based material is superior in its sliding properties, iron-
based material is superior in its durability, and copper-
iron-based material has both properties.
On the other hand, since the price of iron is less
expensive than copper, the sliding properties, durability,
and cost are in a trade-off situation. NTN determined
the correct composition of copper and iron based on the
applicable use and conditions considering the balance
*Engineering Dept., NTN Powder Metal Corporation
-30-
The Copper-Iron-based Material Equivalent to Bronze-based Material for Sintered, Oil-impregnated Bearings
Table 1 Chemical components of the developed product The friction coefficient of the developed material was
Chemical components wt% the lowest at the beginning of operation and it was the
Developed Cu Sn C Fe same as the bronze-based material at the time before
material 15∼22 0.5∼2.5 0.5∼2.5 Remainder stabilization. In addition, the friction coefficient 10
minutes later was at the same level as the bronze-
based material and even lower than the copper-iron-
Copper thin layer Upper layer of the based material containing an equivalent amount of
bearing bore surface copper.
From these results, it was found that the developed
product has excellent initial fitness and that its friction
properties are equivalent or superior to the bronze-
based material.
<Test condition>
・Peripheral speed: 75 m/min
Iron ・Contact pressure: 1.2 MPa
Copper ・Bearing size: φ6×φ12×6J
・Shaft material: SUS420J2
Fig. 1 Section of the developed material
・Test temperature: Room temperature
・Test time: 10 min.
Iron 0.14
0.13
Graphite
0.12
0.11
0.10
Copper
0.09
0.08
Fig. 2 Macrophotograph of bearing bore 0.07

0 1 2 3 4 5 6 7 8 9 10
Time min
Copper-iron-based Developed
Bronze-based
(20% Cu) material
Fig. 3 Result of initial fitness evaluation

4. Performance of the Developed Bearings
We considered initial fitness, limiting PV value, wear
resistance, and low temperature properties when
evaluating the performance of the developed products. 4.2 Limiting PV Value
PV value, which is an index to determine the
4.1 Initial Fitness operation limit of the bearings, is indicated by the
The sintered, oil-impregnated bearings have a high product of the contact pressure applied to the bearings
friction coefficient at the beginning of operation which P and peripheral speed V.
gradually decreases. This phenomenon is called initial The larger the value, the tougher the operation
fitness. condition.
Since motors are required to have stable operation Fig. 4 shows the measurement results of the limiting
characteristics, a low friction coefficient from the PV value of different bearings. This test is conducted by
beginning of operation and early stabilization is maintaining a constant rotational speed and applying a
required. Therefore, we evaluated the initial fitness of load on the bearings. After a predetermined time, if the
the developed product. friction coefficient becomes stable, the weight of the
In addition to the developed product, we also load is increased. This is repeated. If the friction
evaluated bronze-based material and copper-iron- coefficient does not stabilize after the predetermined
based material containing 20% copper, equivalent to time, the PV value at that time is considered as the
the developed product. The result is shown in Fig. 3. limiting PV value.
-31-
The friction coefficient of the developed material 10

behaved the same way as the bronze-based material 9
Bore surface wear µm

8
and was lower than the copper-iron-based material
7
containing an equivalent amount of copper. In addition, 6
the limiting PV value of the developed material was 5
significantly higher than the bronze-based product and 4
equivalent with the copper-iron-based material. 3
2
From this result, it can be determined that the
1
developed material has an equivalent friction coefficient
0
as the bronze-based material in the low PV value range Bronze- Copper-iron- Developed
based based material
and maintains a low friction coefficient in the high PV (20% Cu)
value range, as well. Therefore, it can be used in the Fig. 5 Result of wear resistant evaluation
broad PV value range.
<Test condition> 4.4 Low Temperature Properties

・Shaft material: SUS420J2 When used in an automotive application, a noise may
・Bearing size: φ6×φ12×6 be produced at the beginning of operation in low
・Test temperature: Room temperature ambient temperatures - (abnormal noise at a low
temperature). The cause of this noise is considered to
be the metal-on-metal contact between the shaft and
0.16 the bearings.
0.14 With sintered, oil-impregnated bearings, the
0.12 rotational shaft is supported by the oil film formed

0.10 between the shaft and the bearings which may contract
0.08
in low temperatures, causing less lubrication and
0.06
resulting in metal-on-metal contact.
0.04
0.02
Fig. 6 shows the measurement result of oil film
0 forming rates in low temperature, as an evaluation of oil
0 50 100 150 200 250 300 350 400 450 500 550 600
film forming behavior. As a comparison, the
PV value MPa・m/min measurement result in room temperature is also shown.
Copper-iron-based Developed
Bronze-based The oil film forming rate of 100% indicates non-contact
(20% Cu) material
and 0% indicates full contact.
Fig. 4 Result of limiting PV value measurement In Fig. 6 (1), it was found that the oil film forming rate
of the bronze-based product was poor, indicating metal-
on-metal contact even 30 minutes later. On the other
hand, the developed product showed a good oil film
4.3 Wear Resistance forming rate right after the beginning of the test,
For the evaluation of wear resistance, the wear was transitioning to a non-contact state in an early stage. As
estimated from the difference of the bearing bore a result, we consider that the developed product has
diameter before and after the operation test. Fig. 5 superior oil film forming properties to bronze-based
shows the evaluation results. product in low temperatures, contributing to low noise
It was found that the wear amount of the developed and vibration in a low-temperature environment. In
product is less than that of the bronze-based material addition, as shown in Fig. 6 (2), the developed product
and copper-iron-based material, presenting excellent indicates better oil film forming properties than the
wear resistance. bronze-based product suggesting superior operational
life and wear resistance.
<Test condition>
・Peripheral speed: 38 m/min <Test condition>
・Contact pressure: 4.0 MPa ・Shaft material: S45C
・Bearing size: φ6×φ12×6 ・Contact pressure: 0.51 MPa
・Shaft material: SUS420J2 ・Peripheral speed: 93.4 m/min
・Test temperature: Room temperature ・Shaft material: SUS420J2
・Test time: 8 hours ・Test time: 30min.
・Test temperature: room temperature, -40°C
-32-
The Copper-Iron-based Material Equivalent to Bronze-based Material for Sintered, Oil-impregnated Bearings
100 5. Expansion of Applications of the

Oil film forming rate %
80
Developed Product
The features of the developed product as observed in
60
the evaluation results of the previous section are as
40 follows:
(1) Equivalent friction properties as the bronze-
20
based product
0 (2) Equivalent limiting PV value properties as the
0 5 10 15 20 25 30
copper-iron-based product
Test time min
(3) Superior wear resistance
Bronze-based Developed material (4) High oil film forming properties in low
Fig. 6 (1) Result of oil film forming evaluation (-40°C) temperatures
Because of these features, the application of the
developed material can be expanded into diverse areas
as an alternative to bronze-based material. In particular,
100 its application in electric motors, such as the starter
Oil film forming rate %
80
motor, is expected due to its good low-temperature
characteristics.
60 In industrial machines, it can also be applied to the
small diameter motors where large loads are applied,
40
such as vibration motors and stepping motors.
%
20
0 6. Conclusion
0 5 10 15 20 25 30
Test time min In this paper, we presented the newly developed
Bronze-based Developed material copper-iron-based material equivalent to bronze
sintered material for sintered, oil-impregnated bearings.
Fig. 6 (2) Result of oil film forming evaluation We are poised to expand our sales by leveraging the
(Room temperature)
unique features of these products and continue our
development aiming at further enhancement.
Photo of author
Tomonori YAMASHITA
Engineering Dept.,
NTN Powder Metal Corporation
-33-
[ New Product ]
High Load Sliding Bearing Unit For Use

in Copier and Printer Fusers
Kei HATTORI*
Takuya ISHII**
As one of the cost reduction for business machine typified in copiers and printers,
rolling bearings is required replaceable from sliding bearings. Because of using
bearing conditions at fuser roller of medium and high-speed models are very hard, it
is difficult to apply the sliding bearing. Therefore, NTN developed “High Load Sliding
Bearing Unit”, that combined the metal inner ring and PPS resin outer ring, has
twice load capacity,lower friction and wear properties in comparison with
conventional resin sliding bearings. This article introduces the characteristic and
performance of “High Load Sliding Bearing Unit”.
into the paper. The heat from the heating roller is

1. Introduction transferred to the fuser roller through the belt. This
With the increased demand for lower cost copiers makes the fuser roller hotter than the pressure roller.
and printers manufacturers are examining replacing The bearings support both the fuser roller and the
rolling bearings in the fuser roller with sliding bearings.
However the operating conditions in the fuser roller is
Pressure Fuser roller
demanding and the use of sliding bearings is not
roller Optical unit
Heating roller
always possible. The use of either rolling bearings or
resin sliding bearings 1) is determined depending on
the specification of the model. Since resin sliding
bearings have inferior load carrying capacity when
compared with rolling bearings, the use of resin sliding
bearings is limited to relatively light-load models.
In this paper, we will present a “High Load Sliding
Bearing unit” with improved load carrying capacity. This Print paper
bearing combines an outer ring made of polyphenylen

sulfide (PPS) resin and an inner ring made of metal. Fuser assembly Transfer belt Development,
exposure unit
Fig. 1 Laser beam printer
2. Bearing specifications and issues for
the fuser roller Fuser roller
2.1 Bearing specifications for the fuser roller Fuser roller bearing
Fig. 1 shows the design of a laser beam printer while
Toner
Fig. 2 shows the design of a fuser assembly. The
image captured in the optical unit is replicated onto the
transfer belt with toner in the development and Separating claw
exposure unit. The replicated image is then moved from Shaft Paper
the transfer belt and fixed onto paper at the fuser roller. Pressure roller bearing
Pressure roller
During this process, the paper passes between the
fuser roller and the pressure roller, is heated to Pressure
approximately 200°C and has the toner/image pressed Fig. 2 Fuser assembly
**Composite Material Engineering Dept., Composite Material Products Division

**Engineering Dept. NTN Engineer Plastic Corp.
-34-
High Load Sliding Bearing Unit For Use in Copier and Printer Fusers
pressure roller are supported by. and In operation the properties. The material for the inner ring is common
pressure roller is pressed onto the fuser roller, usually bearing steel while a fluoride based grease is used.
by springs that transmit the force through the bearings. With this design and materials, we have improved
Therefore, the bearings of the fuser roller and pressure friction/wear properties of the bearing, made it usable
roller are always pressed in one direction and are used under the high load conditions, and given the bearing a
under shaft rotation. self-aligning property, as well.
Table 3 provides a comparison between resin sliding
2.2 Issues for bearings in the fuser roller bearings, a high load sliding bearing unit and rolling
Resin sliding bearings are widely used in low-speed bearings. The high load sliding bearing unit, which has
machines with maximum printing speeds of 30 sheets a hybrid structure, has the following features compared
per minute, since the load on the pressure roller is with the resin sliding bearings:
relatively low. However as the printing speed increases
both the temperature and pressure of the fuser roller
increases in order to quickly print the toner. It becomes
Flange Grease
difficult for sliding bearings to survive under these (any shape allowed) (applied between inner
and outer rings)
operating conditions necessitating the move to, rolling Slit
bearings that possess higher load capacities.
There are other issues that a designer needs to
consider when using resin sliding bearing in the Rotation stop
(any shape allowed)
application such as different friction/wear characteristics
depending on the material of the bearing, the shaft fit,
shaft surface roughness and uneven wear due to
deflection and/or misalignment of the shaft.
Outer ring Inner ring
Fig. 3 Structure of high load sliding bearing unit

3. Features of the high load sliding
bearing unit
Aligning
Fig. 3 shows the design of the high load sliding
bearing unit while Fig. 4 shows a partial cross section.
Outer ring
The high load sliding bearing unit is composed of two
components, a PPS resin based outer ring and a metal
based inner ring. The two components contact between
the inner convex spherical surface of the outer ring and
the outer concave spherical surface of the inner ring. Inner ring
Operating clearance is provided between the outer ring
and the inner ring to accomodate thermal expansion
while lubrication is supplied by grease applied to the Fig. 4 Cross section of high load sliding bearing unit
sliding surface. By making the outer concave spherical
surface of the inner ring larger than the inner convex Table 1 Constituent material
spherical surface of the outer ring the contact area (and Component Non-conductive type Conductive type
friction) is reduced. The clearance in the bearing allows BEAREE AS5056 BEAREE AS5965
Outer ring
for the formation of a grease sump resulting in stable (non-conductive) (conductive)
friction condition. Inner ring Bearing steel Bearing steel
Table 1 shows the materials of the outer ring, the Non-conductive Conductive fluoride
Grease fluoride based grease based grease
inner ring and the grease. The electrical characteristics
required for bearings mounted in copiers and printers
vary depending on the design. To accommodate these Table 2 Basic properties of outer ring resin material
different designs there are two types of high load sliding Item Unit BEAREE AS5056 BEAREE AS5965
bearing units each with different materials for the outer Volume resistivity Ω・cm ×1014or more ×103
ring and for the grease, namely, non-conductive and Specific gravity − 1.58 1.62
conductive. The respective outer rings use the Tensile strength MPa 58 43
BEAREE AS materials shown in Table 2. These are Coefficient of thermal expansion ×10-5/˚C 8.7 7.3
PPS resins with superior high temperature Thermal conductivity W/(m・K) 0.32 0.38
propertiesand special additives to improve friction/wear Note: The above properties (room temperature) are the typical values.
-35-
(Features) Coupling Housing Mating shaft

Bearing Thermocouple
(1) A superior friction/wear property in a high Motor under test
temperature environment
Heater
(2) Twice the load carrying capacity
(3) High degree of freedom for designing shafts
(shaft material and surface roughness does not
impact bearing performance)
(4) Consistent sliding property with aligning design
Load
Table 3 Comparison of features various bearings Fig. 5 Radial test machine

Item Resin sliding High load sliding Rolling bearing
bearing bearing unit
Table 4 Test condition
Resin based outer ring Metal basedouter ring
Structure Single resin Metal based inner ring Metal based inner ring
Steel ball Item Description
Load carrying capacity △ ○ ◎ Shaft rotational speed 16.5 m/min (210min-1)
Friction property △ ○ ◎ Contact pressure 0.95 MPa (166N)
Wear property △ ○ ◎ Mating shaft Aluminum alloy
Degree of freedom Temperature 180˚C (surface of mating shaft)
for shaft design × ○ ○
Time 20h
Aligning property × ○ △
Thermal insulation ◎ ◎ △
Cost ◎ ○ △ 0.2
Dynamic friction coefficient

◎ Excellent ○ Good △ Fair × Poor
0.1
Wear amount mm
0.2 0
4. Performance of high load sliding
bearing unit 0.1
4.1 Friction/wear property
Fig. 5 shows the test machine used for measuring 0
Conventional Developed
the friction coefficient and wear amount. The bearing product product
under test was installed in the housing, to which load Fig. 6 Friction and wear property
was applied upward from underneath the housing. Then
the shaft was rotated to measure the friction coefficient
and resulting wear. The bearing under test had an inner 4.2 Limiting PV property
diameter of 25 mm and width of 7 mm. The inside of the One important operating condition of sliding bearings
shaft was heated and the surface temperature was is known as the PV value. This is defined as a product
controlled using a thermocouple. of the contact pressure at bearing sliding surface P and
Table 4 shows the test conditions. The friction the sliding velocity V. The limiting PV values were
coefficient and amount of wear for both the high load evaluated under the test condition shown in Table 5.
sliding bearing unit (developed product) and the resin We defined allowable contact pressure as the
sliding bearing (conventional product) were measured. maximum contact pressure where the increase of the
Both the developed product and the conventional operating clearance between the inner and outer rings
product are of the non-conductive type. The from the test was 0.1 mm or less, through a number of
conventional product is a sliding bearing made of PPS tests with different loads at the predetermined rotation
resin with polytetrafluoroethylene resin blended in. speeds. Here, the contact pressure is a value of load
Fig. 6 shows the amount of wear and dynamic divided by the product of the inner diameter and width
friction coefficient forthe bearings. We defined the of the bearing.
amount of wear in the developed product by measuring Fig. 7 shows the P-V curve, which shows the relation
the increase in the operating clearance between the between the rotation speed of the shaft and allowable
inner and outer ring from the test. The amount of wear contact pressure of the developed product and
and the dynamic friction coefficient of the developed compares this with the PV curve for conventional
product were 1/3 and 1/2, respectively, when compared product. This test revealed that the developed product
to the conventional product. is superior in load carrying capacity as the allowable
-36-
High Load Sliding Bearing Unit For Use in Copier and Printer Fusers
contact pressure was twice that of the conventional dimensions similar to those of equivalent rolling bearings.
product. However it should be noted that it is also possible to
Fig. 8 shows the time-dependent change of dynamic provide thinner bearing units with smaller outer
friction coefficient of the developed product with the shaft diameters. This is achieved by reducing the thickness of
rotation speed of 10 m/min and contact pressure of 1.66 both the outer and inner rings. In addition, as the outer
MPa.This revealed that the dynamic friction coefficient ring is manufactured by injection molding, the size and
was low and consistent at approximately 0.06. shape of the flange and rotation stop can be arbitrarily
We also evaluated bearings of the conductive type. changed and the number of peripheral components,
The conductive type bearings showed equivalent such as the retaining ring, may be reduced.
friction/wear results and limiting PV values as the However, since the resin outer ring has a slit, a
conductive type. rotation stop needs to be provided on the outer ring .
Table 5 Test condition Table 6 Dimensions of high load sliding bearing unit
Item Description Main dimensions mm Allowable
Flange aligning angle
5m/min (64min-1) Inner diameter Outer diameter outer diameter Width mm/mm
Shaft rotational
10m/min (128min-1)
speed 20 32 34.8 7 1/100
16.5m/min (210min-1)
Mating shaft 25 37 39.8 7 1/100
Aluminum alloy
200˚C (surface of mating shaft) 30 42 44.8 7 1/100
Temperature
35 47 49.8 7 1/100
Time 100h
Notice : Allowable misalignment is a reference level to vary according to condition of use.
4
Allowable contact pressure MPa
Developed product
6. Conclusion
3 Conventional product
We have developed a high load sliding bearing unit
2 with significant improvement on the load carrying
capacity and friction/wear properties in a high-
temperature environment. Additionally the new bearing
1
allows for a high degree of freedom in designing the
shaft and alignment accuracy. With this development,
0
0 5 10 15 20 we are able to provide an alternative to rolling bearings
Shaft rotational speed m/min for the fuser rollers of copiers and printers.
Fig. 7 Limiting P-V curve In addition, the high load sliding bearing unit can be
applied to various other applications by changing
0.40 materials of the components. NTN is looking at
Shaft rotational speed：10 m/min expanding the use into food processing equipment and
0.35
Contact pressure：1.66 MPa
Dynamic friction coefficient
medical equipment.
0.30 Temperature : 200 ˚C
0.25 References
0.20 1) Yoshio OKI, Takumi HAYASHI, Takuya ISHII: Characteristics
and Applications of Super Engineering Plastics, Journal of
0.15
Japanese Society of Tribologists 49 (7), (2004) 31.
0.10
0.05
Photo of authors
0
0 20 40 60 80 100
Time h
Fig. 8 Coefficient of dynamic friction
5. Dimension of bearings
Table 6 shows an example of dimensions for a high
load sliding bearing unit mainly used in the fuser rollers Kei HATTORI Takuya ISHII
Composite Material Engineering Dept., Engineering Dept.
for copiers and printers. This table shows bearings with Composite Material Products Division NTN Engineer Plastic Corp.
-37-
[ New Product ]
Machine Tool Main Spindle Bearings with Air Cooling Spacer
Yushi ONDA*
Kikuo FUKADA*
Yohei YAMAMOTO**
Masato YOSHINO***
NTN developed “Machine Tool Main Shaft Bearing with Air Cooling
Spacer” with the air cooling technology for machine tool main spindle
bearings, which realizes high-speed and high-rigidity at the same time with
higher level than ever. In this paper, bearing cooling effect through the
technology and technical activities for practical use are introduced below
with tests and analysis results.
now developed a “cooling spacer 3),” by applying cooling

1. Introduction technology for bearings as a new approach. In this
For machine tool main spindles, high rigidity and high paper, we present the structure and mechanism of
precision are required for machining difficult-to-cut “machine tool main spindle bearings with air cooling
materials and parts with complex shapes. The spacer” which is an improved version of this
characteristics of the machine tool main spindles vary development to the practical level.
depending on the machining content and the target
workpiece because, for example, the machining of
2. Structure
molding for machine parts and components for medical
equipment requires high-speed and high-precision Fig. 1 shows the structure of the bearings with air
rotation. Therefore, the demand for equipment with cooling spacer. NTN’s proprietary environment friendly
multiple machining capabilities such as 5-axis air oil nozzle 4),, which enables reduction of air volume,
processing machines and field-assisted machining oil quantity and noise, is applied to the outer ring spacer
equipment is increasing. The main spindles and installed between back-to-back angular contact ball
bearings are required to have high-speed performance bearings (DB arrangement), in addition, air cooling
and high rigidity. nozzles are installed independent from the air oil
To increase rigidity of the bearings, it is effective to nozzle.
raise preload during assembly and to increase the The air cooling nozzles are arranged offset from the
payload capacity of the bearings. However, those center of the shaft at three locations on the
measures may cause high heat during operation, circumference toward the direction of rotation. The
preventing high speed operation and degrading room temperature air (hereafter, “cooling air”) injected
machining accuracy due to thermal expansion of the from the air-cooling nozzles passes between the inner
main spindles. Therefore, it is important to reduce heat spacer and outer spacer and through the bearings
during operation in order to achieve high-speed swirling toward the direction of the rotation to cool the
operation with increased rigidity of the bearings. bearings.
NTN has been working on higher speed, higher
rigidity and lower noise for the bearings of the main
spindles by improving air oil lubrication 1), 2). We have
***Machining Tool/Aerospace Engineering Dept., Industrial Machinery Division

***Industrial Machinery Engineering Dept., Industrial Machinery Division
***Advanced Technology R&D Center
-38-
Cooling air nozzle

Air cooling spacer Cooling air nozzle
Bearings with (outer ring spacer)
air cooling spacer Offset
Air cooling spacer

Cooling air (inlet) (outer ring spacer)
Inner ring spacer

Environment friendly air oil
lubrication nozzle 4)
Inner ring spacer
Direction of rotation
(inner ring)
Cooling air (outlet) Cooling air (outlet) Arrow A
A
Fig. 1 Structure of the bearing with air cooling spacer
result is shown in Fig. 2.

3. Cooling mechanism If the air cooling nozzles are not offset from the
The bearings with air cooling spacer take the heat spindle center, the cooling air diffuses toward the axial
generated from the bearings from the surface of the direction after colliding against the outer surface of the
inner ring spacer with the cooling air. When the cooling inner ring spacer, and then the air is exhausted out of
air stays longer in the spacer, the heat exchanging time the spacer. Therefore, the bearing cooling effect from
between the cooling air and the surface of the inner ring the cooling air is small.
spacer increases, which improves the cooling effect of On the other hand, when the nozzles are offset, the
the bearings. In addition, faster air flow increases the cooling air flows along the circumference of the outer
heat taken out of the surface of the inner ring spacer surface of the inner ring spacer after its collision to the
per hour, increasing cooling efficiency of the inner ring surface in high speed, remaining on the surface for a
spacer and the bearing inner ring. longer time.
We have modeled the cooling air flow inside the Therefore, the bearing cooling effect from the cooling
bearings with air cooling spacer by fluid analysis. The air is large.
Cooling air nozzle Cooling air nozzle Outer ring spacer

inner surface
Offset Inner ring spacer Cooling air nozzle

Cooling air diffuses outer surface
after colliding with
the inner ring spacer
and leaves the surface Direction of rotation
Direction of rotation
of the inner ring spacer (inner ring) Cooling air flows along
(inner ring) outer surface. the inner ring spacer
outer surface in high
Cooling air nozzle speed.
Cooling air nozzle

Cooling air nozzle
Slow Flow Fast Slow Flow Fast
When cooling air nozzles are not offset When cooling air nozzles are offset
Fig. 2 Cooling air flow in bearing spacer (Fluid analysis results)
-39-
In case of more practical cooling air flow of 100

4. Cooling effect NL/min, the temperature difference between the inner
We have conducted a high speed operation test of and outer rings can be reduced by approx. 4˚C and the
the bearings to verify the cooling effect of the bearings maximum contact pressure of the inner ring runway
with air cooling spacer. Fig. 3 shows the structure of the surface can be reduced by approx. 6% at 23,000 min-1
tester, Table 1 shows the test conditions, and Fig. 4 (dmn value of 2,070,000), therefore, the maximum
and Fig. 5 show the rise of temperature and noise level rotation speed of the bearings can be increased by
of the bearings, respectively. The offset level of the approx. 5%.
nozzles is determined to be 80% of the outer radius of The bearing preload during assembly of the main
the inner ring spacer based on past test results 1). spindle of 0 N can be increased up to approximately
In Fig. 4, the temperature rise of the inner ring is 150 N.
small when cooling air is supplied, compared with no As a result, by adopting this technology we can
cooling air. From this result, we verified that the achieve both high-speed and high-rigidity.
difference of the temperature between the inner and On the other hand, in Fig. 5, we verified that the
outer rings at the rotation speed of 23,000 min-1 (dmn noise becomes loud in the medium/low speed range,
value of 2,070,000) decreases approx. 14˚C when the when the cooling air flow is 300 NL/min. We discuss
cooling air flow of 300 NL/min is supplied. this reduction of noise in the following section.
By this reduction of temperature difference between
the inner and outer rings, the maximum contact
pressure of the inner ring runway surface from inside 30
the bearings during operation can be reduced by No cooling air
25 Cooling air: 100 NL/min
approx. 35% and the maximum rotational speed can be
Temperature difference between
Cooling air: 200 NL/min

inner ring and outer ring ˚C
increased by approx. 20%. In addition, the bearing

preload during assembly of the main spindle of 0 N can
be increased up to approx. 1,300 N. 15
10
Bearings with
5
air cooling spacer
Cooling air (inlet)
0
-5
0 5000 10000 15000 20000 25000
Driving Speed of rotation min -1
motor
Fig. 4 Cooling air amount and temperature difference
between inner ring and outer ring
Cooling air (outlet) Cooling air (outlet)

100
No cooling air
Fig. 3 Structure of test machine Cooling air: 100 NL/min

Noise level dBA

Table 1 Test conditions
φ70×φ110×20 90
Bearings under test Equivalent to 5S-2LA-HSL014
Fixed location pressurization
Pressurization method (pressurization after assembly ON)
Speed of rotation 0∼23000min-1 85
Lubrication method Air oil lubrication
Oil level 0.03ml/10min
Lubricant ISO VG32 80
0 5000 10000 15000 20000 25000
Lubrication air flow 30Nl/min
-1
Outer cylinder cooling Yes, synchronized with room temperature (21±1˚C) Speed of rotation min
Spindle position Horizontal Fig. 5 Cooling air amount and noise level
-40-
into this bearing part, we provided a space between the

5 Initiatives of noise reduction inner ring spacer and outer ring spacer which extends
In order to investigate the cause of the louder noise to the circumference. Fig. 7 shows the fluid analysis
with increased cooling air flow, we modeled the cooling result of this design. The cooling air injected to this
air flow by fluid analysis. space flows to the circumferential direction without
The analysis result when the air cooling spacer width diffusing immediately after being injected, then flows
of 22 mm is shown in Fig. 6. A part of the cooling air into the bearing space slowing down the speed. In this
diffuses to the axial direction also, immediately after case, the collision between the cooling air flown in and
being injected from the air cooling nozzles flowing into the air curtain is reduced compared with the case of
the bearings maintaining the fast flow speed. It is Fig. 6; therefore, the noise becomes low.
considered that this cooling air and the air curtain In order to verify the above space effect, we
produced around the rotating rolling element collide conducted an operational test. This result is shown in
strongly with each other, which caused a lot of noise. Fig. 8. We verified reduction of noise by provision of a
In order to reduce the speed of the cooling air flowing space between the inner and outer ring spacers.
22 Cooling air nozzle 22

Cooling air nozzle
Air cooling Space

Air cooling spacer
spacer
Diffuses horizontally, as Cooling air injected into this

well, immediately after space does not diffuse in
being injected from the the axial direction but in the
cooling air nozzle circumferential direction
immediately after being injected
Location where Location where

air curtain is air curtain is
produced in the produced in the
bearings Loud noise produced bearings
as the air flow main-
taining fast speed into
Inner ring
bearings collides with Inner ring
tapered surface air curtain tapered surface Lower noise due to low
speed when air flow
Inner ring spacer Inner ring spacer collides with air curtain
outer surface outer surface
Inner ring Location where air Inner ring Location where air
tapered surface curtain is produced tapered surface curtain is produced
in the bearings in the bearings
Slow Location where air curtain is produced in the bearings Fast Slow Location where air curtain is produced in the bearings Fast
Fig. 6 Cooling air flow near inner ring spacer surface Fig. 7 Cooling air flow near inner ring spacer surface
(Fluid analysis results) for the case of special space
(Fluid analysis results)
-41-
100 66
No cooling air
No space, cooling air: 300 NL/min
With space, cooling air: 300 NL/min
95
Noise level dBA
Cooling air nozzle

90
Air cooling spacer
85
Fig. 9 Bearing with air cooling spacer of 66mm width
80
0 5000 10000 15000 20000 25000 30
Speed of rotation min-1 No cooling air
25
Temperature difference between

inner ring and outer ring ˚C

Fig. 8 Special space of air cooling spacer and noise level
20
15
6. Relation between the space width

10
and noise
In order to verify the relation between the spacer 5
width and noise, we conducted a test with the cooling

0
air spacer width of 66 mm, as shown in Fig. 9, which is
triple the size of the 22 mm spacer width described in -5
the previous section. Fig. 10 shows the rise of 0 5000 10000 15000 20000 25000
Speed of rotation min-1
temperature of the bearings and Fig. 11 shows the
resulting noise level. Fig. 10 Temperature difference between inner ring and
With Fig. 10, we verified that the temperature outer ring of bearings with air cooling spacer 66mm width
difference between the inner and outer rings is reduced
by approx. 7.5°C and the maximum contact pressure of
the inner raceway surface is reduced by approx. 20% at 100
the cooling air flow of 300 NL/min and with 23,000 min-1 No cooling air
(dmn value of 2,070,000). As a result, the maximum Cooling air: 100 NL/min
rotation speed of the bearings can be increased by 95 Cooling air: 200 NL/min
Noise level dBA
approx. 15% and the bearing preload during assembly

of the main spindle of 0 N can be increased up to
90
approx. 500 N.
Also, we verified that the temperature difference
between the inner and outer rings is reduced by approx.
2.5°C and the maximum contact pressure of the inner 85
raceway surface is reduced by approx. 5% at the

cooling air flow of 100 NL/min and with 23,000 min-1
80
(dmn value of 2,070,000). As a result, the maximum 0 5000 10000 15000 20000 25000
rotation speed of the bearings can be increased by Speed of rotation min-1
approx. 5% and the bearing preload during assembly of
Fig. 11 Noise level of bearings with air cooling spacer
the main spindle of 0 N can be increased up to approx. 66mm width
90 N. From the above, we verified the bearing cooling
effect also with the cooling air spacer width of 66 mm.
On the other hand, no increase of noise level was
observed with the cooling air spacer width of 66 mm,
with the cooling air flow of 300 NL/min and no space
between the inner and outer ring spacers, from Fig. 11.
-42-
In order to understand the reason, we analyzed the

cooling air flow with the cooling air spacer width of 66 7. Conclusion
mm. The result is shown in Fig. 12. With the cooling air NTN achieved a high level of speed and rigidity by
spacer width of 66 mm and no space between the inner using the air cooling technology for the bearings of
and outer ring spacers, a part of the cooling air diffuses machining tool main spindles. We strive to continue
to the axial direction immediately after being infected contributing to "Monozukuri” as we believe that the
from the cooling air nozzles. However, since the “Machine Tool Main Spindle Bearings with Air Cooling
distance to the bearing part is longer and the air flow Spacer” will contribute to higher speed, higher rigidity
speed was reduced by the air viscosity while it passed and higher reliability of the machine tools.
the space between the inner and outer ring spacers, it
is considered that the resulting noise due to the collision
with the air curtain was reduced.
References
1) Kenji Fujii, Masatsugu Mori, Yoshimi Ota: Noise
Reduction for Machine Tool Main Spindle Air Oil
A part of cooling air diffuses in the axial direction Lubrication Bearings, The Japan Society for Precision
immediately after being injected from the cooling
air nozzles Engineering, Autumn Meeting 2000 Proceedings (2000)
449.
2) Yoshinobu Akamatsu and Masatsugu Mori: Minimizing
Location where
air curtain is Lubricant Supply in an Air-Oil Lubrication System, NTN
produced in the TECHNICAL REVIEW, No 72 (2004).
bearings
3) Yushi Onda, Mamoru Mizutani and Masatsugu Mori:
Machine Tool Main Spindle Bearings with “Air Cooling
Spacer”, NTN TECHNICAL REVIEW, No 80 (2012).
4) NTN Catalog: Precision Rolling Bearings, CAT. No.
2260-V/J 12.09.01.
Because of the long distance
Inner ring to the bearings, cooling air
tapered surface speed is reduced before it
flows in the bearings due to
Inner ring spacer air viscosity as it passes
outer surface through the space
Location where air

curtain is produced
Inner ring in the bearings
tapered surface
Cooling air flow speed around the

Slow inner ring spacer surface Fast
Fig. 12 Cooling air flow in bearing 66mm spacer

(Fluid analysis results)
Photo of authors
Yushi ONDA Kikuo FUKADA Yohei YAMAMOTO Masato YOSHINO

Machining Tool/Aerospace Machining Tool/Aerospace Industrial Machinery Advanced Technology
Engineering Dept., Engineering Dept., Engineering Dept., R&D Center
Industrial Machinery Division Industrial Machinery Division Industrial Machinery Division
-43-
[ New Product ]
New Engineering Plastic Cage of High Speed

Angular Contact Ball Bearings for Machine Tools
Wakana INOUE*
Mineo KOYAMA*
Soichiro ONITSUKA**
To support the high-speed, high-rigidity and high-reliability requirements of

machine tools, NTN developed new resin cage of high speed angular contact
ball bearings for machine tools. The new cage has good characteristics in
high-temperature range and optimized pocket design. With the modifications,
the cage can be applied higher speed range than conventional one. In this
document, we introduce the design points and test data.
properties, as well as a special surface modification

1. Introduction technique for internal design of inner and outer rings to
The speed of spindle rotation for machine tools has achieve higher speed, rigidity and reliability. By this
been increasing for improving quality of the machined adoption, higher speed operation was made possible
surface and machining efficiency as represented by the while maintaining preload after assembly of the main
machining centers for mold manufacturing. In order to spindles allowing broad use in the main spindles of
respond to these requirements, NTN has introduced high-speed lathes and machining centers.
high speed angular ball bearings, ULTAGE Series *1 It is possible to allow use of steel balls or ceramic
HSE type, applying special materials and special balls as rolling elements and to select contact angles of
surface treatment onto the inner and outer rings of the inner and outer rings (15°, 20° and 25°) depending on
bearings, as well as optimum design for the internal the operational conditions of the main spindles.
components.
The HSE bearings adopt machined cages made of
phenol resin as standard; however, for mid/low speed Cage
main spindles of dmn value (an index of the speed of · High speed main spindle: Machined cage with phenol
resin
rotation) of 1,000,000 or lower, cages made of · Mid/low speed main spindle: Cage with polyamide resin
polyamide resin are widely used considering the cost. (different shape)
We have designed cages made of newly developed
polyamide resin applicable for a higher speed range Raceway rings
(inner ring, outer ring)
where machined cages made of phenol resin have
· Special material
been traditionally used. + surface treatment
In this paper, we present features and evaluation
results of the developed cages. Contact angle
(15°, 20° and 25°)
Rolling element
2. High speed angular contact ball · High speed main spindle: Ceramic ball
bearings - HSE type · Mid/low speed main spindle: Steel ball
The high speed angular contact ball bearings - HSE Fig. 1 High speed angular contact ball bearings HSE type
type (shown in Fig.1) employs a special material
featuring improved wear resistance and anti-seizure
**Machining Tool/Aerospace Engineering Dept., Industrial Machinery Division

**Industrial Machinery Engineering Dept., Industrial Machinery Division
※1 ULTAGE is a name created from the combination of "ultimate" and "stage” signifying NTN’s intention to pursue the ultimate performance of precision bearings.
-44-
New Engineering Plastic Cage of High Speed Angular Contact Ball Bearings for Machine Tools
3. Features of cages with new resin High speed operation of main spindle
3.1 Issues with cages made of the conventiona

polyamide resin Rolling heat generation Centrifugal force
between rolling element applied to the inner ring
Fig. 2 shows the conditions of the cage pocket after and raceway and rolling element
high speed operation of bearings with the cage made of
conventional polyamide resin at over 1,000,000 of dmn Large rise of temperature
value. The contact area of the cage pocket and the for the bearings
(large difference of
rolling element locally melts and wears depending on temperature between the
operating conditions. This damages progresses as inner and outer rings)
tough operation continues, and may cause abnormal
rise of temperature or seizure of the bearings. Increase of preload of bearing
(increase of load on the rolling element)
The damage process of the cage pocket is shown in
Fig. 3. High speed sliding contact is observed between
the cage pocket and rolling element in high speed Slide contact between Strong contact between
the cage pocket and the cage pocket and
operation of main spindles of machine tools. Fig. 4 rolling element rolling element
shows the relation between the speed of the main
spindle rotation and the peripheral speed of the rolling
element. In case of main spindles of machining center Heat generation at the cage pocket contact area
with the main spindle diameter of 70 mm and speed of
rotation 15,000 min-1, the peripheral speed of the
Poor lubrication at the cage pocket contact area
rotating element can reach approximately 35 m/s, and (further heat generation)
over 40 m/s with higher speed spindles.
In addition, the preload of the bearings increases with
Wear/melt of cage pocket
speed, due to the temperature difference between the
inner and outer rings, which expands because of the Fig. 3 Cage damage process in high speed operation
heat generated between the rolling element and
raceway and the centrifugal force applied to the inner
ring and the rolling element. This leads to the increase
HSE014 (φ70×φ110×20)
Rolling element surface velocity m/s
of the load on the rolling element, which causes tighter 50

contact between the cage pocket and the rolling Expansion of applications
by making the cage for
element. 40 higher speed operation
As a result, the heat generated in the cage pocket
30
can locally increase, causing potentially poor Main spindle for high speed
Operation
range of the machining center (φ70×15000min-1)
lubrication, melting, and wear. 20
conventional
cage
10
Initial wear/melt 0
0 5000 10000 15000 20000
Bearing rotating speed min-1
0 0.5 1.0 1.5 2.0

dmn value ×106
Fig. 4 Relationship between bearing rotating speed and

rolling element surface velocity
New After high speed operation

Fig. 2 Conventional cage pocket after high speed
operation
-45-
3.2 Adoption of polyamide resin material with rolling element due to the decrease of dynamic modulus
a high melting point of elasticity and the resulting heating can be reduced.
As mentioned in the previous section, it is necessary To evaluate the developed cage material, we made a
to apply the materials with a high glass transition point cage with the same shape as the conventional cages
and high melting point, in particular, to prevent local for comparison.
wear and melt of the cage pocket, as shown in Table 1, Fig. 6 shows the test spindle, Fig. 7 shows the test
for resin cages to allow higher speed of rotation. result with air oil lubrication, and Fig. 8 shows the test
Table 2 shows the characteristics of the developed result with grease lubrication.
cage material. We adopted a high melting point semi- Even if no difference in rise of temperature was
aromatic polyamide as the material for the developed observed between two cages, the resin of the
cage. It has a higher melting point and glass transition developed cages with both air oil and grease lubrication
point compared with the conventional polyamide used demonstrated durability for applications of higher speed
for cages, with smaller strength deterioration and than the conventional polyamide resin.
dimensional change due to high temperature.
Fig. 5 shows the relationship between the dynamic
modulus of elasticity*2 to temperature.
×1010
The dynamic modulus of elasticity of polyamide for 1.2
Pa
Developed cage
the conventional cages starts decreasing from approx. Dynamic modulus of elasticity 1.0 material
50°C, around the glass transition temperature; however,
0.8
polyamide with a high melting point for the developed
cages has a higher glass transition temperature, 0.6
maintaining a gradual decrease of dynamic modulus of Conventional
0.4 cage material
elasticity up to around 120°C. Therefore, when the cage
pocket is heated by the high speed slide contact with 0.2
the rolling element, increase of the contact area with the
0
-50 0 50 100 150 200 250
Temperature ˚C
Table 1 The necessary characteristic of cage material Fig. 5 Relationship between the dynamic modulus of
for high speed operation elasticity to temperature
Required properties Purpose

· Low water absorption and
Small dimensional dispersion,
1 dimensional change allows stable operation Bearings
· Low linear expansion coefficient under test
· High tensile strength and flexural Little deformation in high Supporting
2 strength speed operation bearing
· High glass transition temperature

3 Little wear/melt in high speed
· High melting point
slide contact
4 · Superior sliding property
5 · Good injection moldability Stable quality
6 · Low cost Broader applicability without

cost restriction
Motor side
Table 2 Characteristic of cage material

Conventional cage Developed cage
Item material material
High melting point
Fig. 6 Test spindle
Resin type Polyamide polyamide
Melting temperature Tm ˚C 260 300
Glass transition temperature Tg ˚C 58 125
Tensile strength MPa 160 175
Water absorption ratio ％ 2∼3 0.2
※2 Dynamic modulus of elasticity is obtained from the response of the sinusoidal amplitude given to the test piece.
-46-
New Engineering Plastic Cage of High Speed Angular Contact Ball Bearings for Machine Tools
3.3 Shape optimization of cages

【Test condition】
The shape of the pocket of the developed cage is
Bearing 5S-2LA-HSE014DB (contact angle of 20˚)
under test φ70×φ110×20 back-to-back assembly
shown in Fig. 9. We have provided space for lubrication
close to the contact area between the rolling element
Fixed position preload
Preload and cage pocket for the developed cage. As a result, in
Preload during assembly: 300 N
Air oil Supply amount: 0.03mℓ/8 min case of air oil lubrication, smooth supply/drain of
lubrication Air flow: 40 Nℓ/min lubricating oil was made possible at the contact point of
External cylinder the rolling element and the cage pocket. In case of
cooling Yes
grease lubrication, it also contributed to the retention of
grease in the proximity of the contact point increasing
35
Rise in temperature of outer ring ˚C
Rise in lubrication reliability in high speed operation.

Developed cage material temperature
30
Conventional cage material In addition, the guiding surfaces, which are the
25 contact area of the developed cage with the rolling
20 element in rotational and axial directions, are made flat
15
(ref. (1) and (2) of Fig. 9) to reduce generation of heat
by minimizing the contact area with the rolling element
10
compared with the circular hole contact area of the
5 conventional cages with polyamide resin.
0
0 10000 20000 30000
Plain (1) guide to
Circular hole contact area rotational direction
0 0.5 1 1.5 2 2.5 3 with rolling element
Space
dmn value ×106
Fig. 7 Test result under air-oil lubrication with ceramic ball
Rotational
direction
Bearing 2LA-HSE014DB (contact angle of 20˚)
under test φ70×φ110×20 back-to-back assembly
Fixed position preload Plain (2) guide to
Preload
Preload during assembly: 300 N axial direction Axial direction Rolling
element
Grease
lubrication NTN SE-1 Conventional cage Developed cage
Fig. 9 Cage pocket shape
External cylinder
cooling Yes
30
Developed cage material
Rise in
25
temperature 4. Performance of the developed cage
Conventional cage material Rise in
temperature
We evaluated the HSE bearings using ceramic balls
20
with air oil lubrication and grease lubrication, and using
15
steel balls with grease lubrication in high speed
operation with the test spindle shown in Fig. 6.
10 The results are shown in Fig. 10, Fig. 11 and Fig. 12.
We verified that the developed cage experienced
5
reduced the rise of temperature enabling higher speed
0 operation.
0 5000 10000 15000 20000
0 0.5 1 1.5
dmn value ×106
Fig. 8 Test result under grease lubrication with steel ball
-47-
【Test condition】【Test condition】

Bearing 5S-2LA-HSE014DB (contact angle of 20˚) Bearing 2LA-HSE014DB (contact angle of 20˚)
under test under test φ70×φ110×20 back-to-back assembly
φ70×φ110×20 back-to-back assembly
Fixed position preload Fixed position preload
Preload Preload during assembly: 300 N
Grease
Air oil Supply amount: 0.03mℓ/8 min lubrication NTN SE-1
lubrication Air flow: 40 Nℓ/min
External cylinder
External cylinder cooling Yes
Yes
cooling
35

35 Developed cage material Rise in
Rise in temperature
Developed cage material temperature 30
30 Conventional cage material
Conventional cage material 25 Rise in
25 temperature
20 20
15 15
10 10
5
5
0
0 10000 20000 30000 0
0 5000 10000 15000 20000
0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2

dmn value ×106 dmn value ×106
Fig. 10 Test result under air-oil lubrication with ceramic ball Fig. 12 Test result under grease lubrication with steel ball
5. Conclusion
Bearing 5S-2LA-HSE014DB (contact angle of 20˚)
under test φ70×φ110×20 back-to-back assembly We have improved the high speed performance of
Fixed position preload
cages for high speed angular contact ball bearings for
Grease machine tools over the conventional cages made of
lubrication NTN SE-1 polyamide resin by adopting high melting point
External cylinder
cooling Yes polyamide resin and a new cage shape.
30
We have enabled cages made of newly developed
Rise in
Developed cage material temperature polyamide resin applicable for a higher speed range
25 Conventional cage material where machined cages made of phenol resin have
Rise in
temperature been traditionally used. We will introduce this cage
20
using ceramic balls with air oil lubrication for HSE
bearings to be used with dmn value of 1,600,000.
15
We strive to work on further improvement of
10 performance of the bearings for main spindles
contributing to higher speed, rigidity and reliability of
5
machine tools.
0
0 5000 10000 15000 20000 25000
Photo of authors
0 0.5 1 1.5 2
dmn value ×106
Fig. 11 Test result under grease lubrication with ceramic ball
Wakana INOUE Mineo KOYAMA Soichiro ONITSUKA

Machining Tool/Aerospace Machining Tool/Aerospace Industrial Machinery
Engineering Dept., Engineering Dept., Engineering Dept.,
-48-
Bearing Technical Calculation System

for Machine Tool Spindles
Customer Keiichi UEDA*

Data sent
by email Yushi ONDA*
Shouhei HASHIZUME*
Data received
24 hour/7 day
by email
support
In recent years, machine tool and spindle manufactures have been developing various products which can
meet the new customer demands. To support the activities, NTN developed the bearing technical calculation
system for machine tool spindles. Machine tool and spindle manufactures can calculate various
characteristics like spindle rigidities and fatigue life of spindle bearings through the system in short time.
Features, mechanism and functions of the system are introduced below.
Therefore, NTN developed the Bearing Technical

1. Introduction Calculation System for Machine Tool Main Spindles that
The main spindles of machine tools are an essential customers can use to automatically calculate
element in determining the machining accuracy and characteristics of the bearings and main spindles in a
productivity of machine tools. Their critical properties, short time. This new tool will help our customers with
such as rigidity, are determined by the type, their designs and contribute to reductions in lead time
arrangement, placement, preload, etc., of the bearings. for development of machine tools.
Therefore, when the manufacturers of machine tools In this paper, we present the features, configuration,
and manufacturers specialized in spindles (hereafter, and functions of this system.
"customers") design main spindles, they need to obtain
specifications of the bearings and properties of the main
2. Overview
spindles by calculation and verify if they satisfy the
target objectives. The technical calculation system we developed can
Since the these calculations are generally performed calculate characteristics of 10 main spindle bearing
by the bearing manufacturers, they are required to layouts of different types, arrangements and preload
calculate the properties of multiple main spindle design methods. The overview is shown in Fig. 1.
candidates quickly and in detail based on the The customer enters data required for calculation
information from the customers’ operating conditions, such as the bearing type and operating conditions into
etc. In addition, materials, machining processes, and the dedicated Excel ®※1 file and sends it to NTN’s
machine tools users have become increasingly dedicated server by email. The server automatically
diversified in the recent years with additional performs the calculation based on the input information
requirements such as complex functionalities, and immediately responds back to the customer with
downsizing, and environment friendliness in addition to the calculation results such as the rigidity of the main
the conventional requirements of high speed, high spindle and bearing life. Traditionally, NTN required at
rigidity and high precision. In order to design main least 1 day ※2 from receipt of the calculation request
spindles that meet these requirements, more reviews from the customer until the report with the calculation
and propositions are required with increasingly frequent results was sent back to the customer. However, with
design opportunities. this system, the response is sent back to the customer
*Machining Tool/Aerospace Engineering Dept., Industrial Machinery Division

※1 Excel ® is a trademark or registered mark of Microsoft Corporation in the United States of America or other countries.
※2 Required time for technical calculation of bearings for machine tool main spindles, in general, and may vary depending on the content
of bearing calculation.
-49-
within approx. 10 minutes of the customer’s 3.2 Configuration

emailrequest. This system is shown in Fig. 1 and consists of the
In this system, the server handles all processes from following elements ① – ③ . Any customer with Excel ®
receipt of the email requests to automatically sending (Excel 97-2003 or later) and an email system can use
the email responses. Therefore, the time required for this system. No special settings, such as installation of
transferring information between the customers and dedicated programs, is required for customers: one of
NTN can be significantly reduced. Additionally, the benefits of this system.
customers are free to submit as many calculations as
they desire, making it possible to narrow down main ① Dedicated server (installed at NTN)
spindle design plans very quickly. Furthermore, as the ② Dedicated program installed in ①
server is able to work on a 24 hour, 7 days a week ③ Dedicated Excel ® file for exchanging input/output
basis, customers can use the system without no time information between the customer and ①
restrictions.
Though the design process leading to the finalization 3.3 Flow
can be shortened by this system, contributing to The system flow is as follows: The sequence ① – ⑤
significant development lead time reductions for main corresponds to ① – ⑤ in Fig. 1. After step ⑤ , the
spindles and machine tools, the final specifications of specification of the bearing is finalized in a discussion
bearings must still be determined by discussions with the customer and NTN.
between the customer and NTN.
3. System features, configuration, and flow ① The customer enters required information into the dedicated
Excel ® file.
3.1 Features
Industry-first technical calculation system with
② The customer attaches the file of ① to an email and sends
automatic processing it to the NTN dedicated server.
(1) Fast technical calculation of bearings for main
spindles
③ The NTN dedicated server automatically calculates the data.
10 minutes after sending a request email
(traditionally, at least 1 day)
④ Approx. 10 minutes after ②, an email is sent from the NTN
(2) Technical calculation of many types of main dedicated server to the customer. The Excel ® file attached
spindle layouts to the email contains the calculation result (specifications
10 types of main spindle layouts can be processed of the bearings and properties of the main spindle
determined by them).
(3) Automatic calculation within the dedicated
server
⑤ The customer confirms the calculation results and evaluates
24×7 operations if they satisfy the objective. If they do not satisfy the
(4) Simple communication of input/output data objective, the customer can repeat ① to ④ as many times
as the customer wants.
Adoption of a practical Excel ® file
Send calculation results:

approx. 10 minutes
Customer
①Create entry sheet ② Send Bearing Technical

¡Main spindle layout
¡Bearing
input sheet Calculation System
model number for Machine Tool Spindles
③Automatic calculation
support
¡Receive email
¡Automatic calculation
¡Send calculation Automatic
⑤Design main spindle result to customer calculation
based on output sheet
④ Receive
¡Bearing property output sheet
¡Main spindle property 24 hour/ 7 day support
Fig. 1 System flow
-50-
Bearing Technical Calculation System for Machine Tool Spindles
Table 1 Input and output items

4. System input/output
[Items regarding main spindle]
4.1 Input/Output items · Main spindle layout
Table 1 shows input/output items for this system. Select from 10 layouts
As described in the previous section, the customer · Operating conditions
Speed of rotation, external load, etc.
enters information in the predefined Excel ® file and the
Input item
system calculates and responds with the calculation [Items regarding bearings]
results (output) to the customer. · Bearing model number, bearing position
In the input, the items and operating conditions · Preload, space
· Engagement
regarding the main spindles are entered. 10 types of · Amount of adjustment of inner/outer ring
layouts can be selected for main spindles, as shown in spacer, etc.
Fig. 2, so various design candidates for main spindles
[Bearing properties]
can be created. · Preload, spring constant, rolling fatigue life
Preload, spring constant, and rolling contact fatigue Output item
[Main spindle rigidity]
life of bearings, as well as rigidity during assembly
· Rigidity during assembly (radial, axial)
(radial and axial direction) to the main spindles are
obtained as output. The customer verifies the result and
evaluates if it meets their objectives.
Regarding the main spindle layouts, NTN is
expanding the number of types moving forward.
Bearings on the front side

Bearings on the
back side
Layout 1 Layout 6
Constant preload spring
Layout 2 Layout 7
Layout 3 Layout 8
Layout 4 Layout 9
Layout 5 Layout 10
Fig. 2 Spindle layout
-51-
4.2 Input/Output screen examples to the output screen of the file in Fig. 4, and
The customer selects layouts that need to be automatically sends the file to the customer.
examined from the main spindle layout group shown in The customer can receive the calculation results
Fig. 2 and enters the required information into the approx. 10 minutes after sending email to the server.
shaded cells of the dedicated Excel ® file shown in Fig. 3.
The file has a function to check the entered items;
5. Calculation examples
therefore, the customer can easily check if any entered
information is missing or wrong. As a part of the output examples shown in Fig. 4, the
After entry of the required information, the customer calculation results of main spindle axial rigidity when
sends the file to NTN's dedicated server by email. bearings, arrangements and preload methods in Fig. 5
The server which receives this email automatically are examined against the main spindle layout 1 of Fig. 2
calculates based on the input information and enters are shown in Fig. 6.
the calculation results, such as the main spindle rigidity, The time needed to narrow down design candidates
Fig. 3 Example of input sheet (Spindle layout 1, extract)
Fig. 4 Example of output sheet (Spindle layout 1, extract)
-52-
Bearing Technical Calculation System for Machine Tool Spindles
Front side Rear side Front side Rear side

High speed angular ball Cylinder rolling High speed angular ball Cylinder rolling
Bearing contact bearings bearings Bearing contact bearings bearings
5S-2LA-HSE020 N1018HSK 5S-2LA-HSE020 N1018HSK
Arrangement Two-row back-to-back (DB) ― Arrangement Four-row back-to-back (DTBT) ―
Preload method Fixed position (GL preload) ― Preload method Fixed position (GL preload) ―
Front side Front side Rear side

Rear side
Fig. 5 Technical study example of spindle layout 1 Fig. 7 Technical study example of spindle layout 4
18 18
Axial displacement at origin µm

Axial displacement at origin µm
16 16
14 14
12 12
10 10
8 8
6 6
4 4
2 2
0 0
0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3
Axial force kN Axial force kN
Axial rigidity of spindle Axial rigidity of spindle
Fig. 6 Example of calculation result Fig. 8 Example of calculation result

(Axial rigidity of spindle layout 1) (Axial rigidity of spindle layout 4)
can be reduced, given that the calculation of properties

of multiple main spindle design candidates can be
6. Conclusion
performed quickly. and compared NTN developed the Bearing Technical Calculation
For example, the rigidity can be quickly compared System for Machine Tool Spindles which can
when double-row back-to-back (DB) bearings are automatically calculate properties of bearings and main
applied to the main spindle of shaft diameter 100 mm spindles in a short time, for precision rolling bearings to
on the front side shown in Fig. 5 and four-row back-to- be used for the main spindles of machine tools.This
back (DTBT) bearings in Fig. 7 are applied (main system will support diverse product development of
spindle layout 4 ). In this case, it is observed that the machine tool manufacturers and spindle manufacturers.
four-row back-to-back bearings shown in Fig. 8 have With this system, NTN is poised to provide faster
higher rigidity than the two-row back-to-back bearings technical service and obtain higher customer
shown in Fig. 6. satisfaction, as well as to contribute to reduced lead
time for machine tool development.
NTN will explore easy-to-use, useful, and user friendly
systems by the further improvement of this system.
Photo of authors
Keiichi UEDA Yushi ONDA Shouhei HASHIZUME

Machining Tool/Aerospace Machining Tool/Aerospace Machining Tool/Aerospace
Engineering Dept., Engineering Dept., Engineering Dept.,
-53-
[ Technical Paper ]
Estimation Method for Friction Torque

of Air-oil Lubricated Angular Contact Ball Bearings
Hiroki FUJIWARA*
Angular contact ball bearings for high-speed spindles are lubricated by air-oil
lubrication. A main factor of bearing friction torque is rolling viscous resistance.
On the air-oil lubricated bearings, oil starvation leads to decrease in the rolling
viscous resistance. In this article, the expression of deduction factor of rolling
viscous resistance under the driving condition is proposed. It is derived by
comparing between rolling viscous resistances estimated from measured friction
torques of angular contact ball bearings under the air-oil lubrication and calculated
rolling viscous resistances under the full lubrication. The friction torque of the bearing is computed with
consideration for differential slip, spin, elastic hysteresis loss, cage guiding friction, and the starved rolling
viscous resistance. The starved rolling viscous resistance is obtained for product of the deduction factor and
the rolling viscous resistance by the NTN original regression formula under the full lubrication.
The friction torque of rolling bearings using oil-bath

1. Introduction lubrication can be calculated by considering differential
High speed spindles are required to contain heat in slip, spin, rolling viscous resistance, shear resistance of
order to reduce thermal expansion of the main spindle. the oil film between the rolling element and cage
Air-oil lubrication has been adopted to reduce torque of pocket, etc. 1).
the bearings, therefore reducing heat generation. Air-oil However, this calculation method is applicable only to
lubrication enables accurate control of lubricating oil the case where lubricating oil is abundant and cannot
quantity by adjusting the plunger pump to supply the be applied to air-oil lubrication as its friction torque is
minimum required lubricating oil directly to the raceway 1/10 to 1/5 that of bearings with sufficient lubrication.
surface by suspending the oil in air. Through this There are no reports that examine the estimation of
operation, stirring resistance of the lubricating oil can be friction torque for air-oil lubrication. Although there are
almost eliminated and the lubricating oil quantity attempts to theoretically estimate the starvation status
between the rolling element and raceway is reduced to of the ball pushing through the lubricating oil on the
the bare minimum. raceway, given certain assumptions of lubricating oil
Lubricating oil above a specific amount at the inlet of quantity 2), 3), it is difficult to accurately find the
the elastohydro dynamic lubrication (hereafter, EHL) lubricating oil quantity on the raceway in air-oil
contact between the rolling element and raceway lubrication. Therefore their application to practical use is
presents sufficient lubrication where the thickness of the challenging.
lubricating oil film does not increase anymore. However, In this paper, we propose an estimation method for
from the viewpoint of damage control, sufficient friction torque of air-oil lubricated angular contact ball
lubrication is not necessarily required. Only oil film large bearings in their unique starvation status. The friction
enough to separate the two surfaces from contact is torque of these bearings consists of differential slip,
required. Since the rolling viscous resistance, which is spin, rolling viscous resistance, elastic hysteresis loss
the main element of friction torque, is affected by the and friction between the cage guiding surface and the
lubricating oil quantity at the EHL contact inlet, the rolling element.
rolling viscous resistance can be reduced by actively We considered that the lubricating oil film of these
creating insufficient oil quantity, called "starvation", to bearings produced a level of no mixed boundary
the level of having no oil film between the rolling lubrication, assumed that the friction torques due to
element and raceway breaks. differential slip and spin are not affected by the
*Advanced Technology R&D Center
-54-
Estimation Method for Friction Torque of Air-oil Lubricated Angular Contact Ball Bearings
starvation, and applied an experimentally obtained T0 ：Standard value on the ball pass cycle
reduction factor to the rolling viscous resistance. This U ：Dowson's dimensionless parameter on velocity
reduction factor is proposed as an equation W ：Dowson’s dimensionless parameter on the
experimentally identified comparing the experimental point contact load
value of the friction torque with the calculated value of α0 ：Viscosity-pressure factor of lubricating oil
sufficient lubrication. α ：Contact angle
β ：Inclination of the ball’s axis of rotation
δ ：Distance from the pure rolling point to the
2. Symbols
contact point
a ：Long axial radius of contact ellipse μt ：Coefficient of traction
a0 ：Standard value on the long axial radius of contact μt max ：Maximum coefficient of traction
ellipse ν ：Kinematic viscosity of lubricating oil
C ：Reduction factor constant of rolling viscous resistance ν0 ：Standard value on the kinematic viscosity of
(includes standard values of the parameters) lubricating oil
C’ ：Reduction factor constant of rolling viscous resistance τ ：Shear stress due to Couette flow of contact area
dp ：Pitch circle diameter of a ball between the ball and raceway ring
F! ：Traction generated inside differential slip φr ：Reduction factor of rolling viscous resistance
F@ ：Traction generated outside differential slip φre ：Reduction factor of rolling viscous resistance
FHD ：Dimensionless rolling viscous resistance (experimental value)
Fr ：Rolling viscous resistance ω ：Angular velocity
Fs ：Traction ω0 ：Standard value on angular velocity
G ：Dowson's dimensionless material parameter ωb ：Rotational angle of the ball
k ：Ellipticity of contact ellipse Subscripts
l! ：Length of the contact point and pure rolling point i ：Inner ring
in the radius direction inside the differential slip o ：Outer ring
l@ ：Length of the contact point and pure rolling point
in the radius direction outside the differential slip
M ：Moment around rotational axis on the ball 3. Static analysis of angular contact ball
Mr ：Friction torque of the bearing due to friction between bearings with only preload
the ball and raceway in the rolling direction High speed spindle angular contact ball bearings are
m ：Moment due to F! and F@ typically arranged with 2-rows or 4-rows on the front
Pb ：Component force of the oil film reaction force on side and used with fixed position or constant pressure
the ball in the rolling direction preload. Although minor radial load may be applied
Pe ：Component force of the oil film reaction force in the during operation depending on the use of spindles, we
rolling direction in the equivalent system ignored the effect of the radial load and assumed pure
P1 ：Component force of the oil film reaction force applied axial load with only preload applied to discuss the
on object 1 (ball) in the rolling direction motion of a ball. Regarding the static analysis of angular
P2 ：Component force of the oil film reaction force applied contact ball bearings, Fujii has provided a detailed
on object 2 (raceway) in the rolling direction review 4) and Fujii has determined rotational axis of the
r1 ：Radius of object 1 (ball) ball based on Jones' control ring theory 5). With Jones’
r2 ：Radius of object 2 (raceway) theory, the ball only rolls on the raceway of the ring
rb ：Radius of the ball which has larger friction and only spin occurs on the
rb' ：Distance from the center of the ball to the pure other raceway. However, in actuality, the rotational axis
rolling point of the ball should be determined by the way that the
re ：Equivalent radius force and moment are balanced and the consumed
S! ：Inside area of differential slip of inner ring contact energy is minimized. In the following sections, we will
S@ ：Outside area of differential slip of inner ring contact discuss hydrostatic movement of the ball considering
s ：Slip ratio additional force due to slippage.
sm ：Slip ratio to give the maximum coefficient of traction
sT ：Index on the ball pass cycle 3.1 The force and moment produced due to
sν ：Index on the kinematic viscosity of lubricating oil slippage between the ball and raceway
sα ：Index on the long axial radius of contact ellipse Fig. 1 shows the schematic diagram of an axial cross
sω ：Index on angular velocity section of one ball and the inner/outer ring indicating the
T ：Pass cycle of the ball force due to slippage within the contact ellipse.
-55-
When a tangent force is applied to the ball as shown The moment of the ball around the pure rolling point
in Fig. 1, the force can be obtained with the following due to slippage can be calculated with the following
equation: equation:
∫ ∫
Fs,i = τi dS@− τi dS! ∫ ∫
mi = τi l!dS!＋ τi l@dS@
…………………… (3.1) …………………… (3.2)
Fs,o =∫τo dS −∫
@ τo dS ! mo =∫τo l! dS!＋∫τo l @ dS@
The moment of the ball around its rotational axis due

to slippage can be expressed as follows:
mo
F@,o F!,o ∫
Mi = (r’b＋δ)τi dS = rb Fs,i−mi
…………… (3.3)
Mo =∫(r’b＋δ)τo dS = rb Fs,o−mo
ωb
From Fig. 2, the peripheral force equilibrium on the
F!,i F@,i ball is:
mi Fs,i−Fr,i＋Pb,i−Fs,o＋Fr,o−Pb,o = 0 ………… (3.4)
and as Zhou-Hoeprich 6) states, the rolling viscous
resistance has the following relation:
Fig. 1 Direction of forces and moments from slip
Pe re = Pl rl = P2 r2 = 2Fr re …………………… (3.5)
Therefore, it can be rewritten as follows:

re,i re,o
The moment of the ball around the pure rolling point Fs,i−Fr,i＋2 r Fr,i−Fs,o＋Fr,o−2 r Fr,o= 0 … (3.6)
b b
due to slippage can be calculated with the following
equation: The equilibrium of the moment around the rotational
axis of the ball can be expressed,
∫ ∫
mi = τi l!dS!＋ τi l@dS@
…………………… (3.2) rb cos (αi−β) (−Fr,i＋Fs,i )＋rb cos (αo−β) (−Fr,o＋Fs,o )
mo =∫τo l dS! ＋∫τo l dS@
! @
−mi cos (αi−β) −mo cos (αo−β) = 0 ……… (3.7)
The moment of the ball around its rotational axis due and from equation (3.3), it can be rewritten as follows:
to slippage can be expressed as follows:
rb cos (αi−β) (−Fr,i )＋rb cos (αo−β) (−Fr,o )
∫
Mi = (r’b＋δ)τi dS = rb Fs,i−mi ＋Mi cos (αi−β) ＋Mo cos (αo−β) = 0 ……… (3.8)
…………… (3.3)
Mo =∫(r’b＋δ)τo dS = rb Fs,o−mo
F@,o F!,o
mo
Fr,o
Pb,o
Fr,i Pb,i ωb
F@,i
mi
F!,i ωi β αo
αi
Fig. 2 Forces and moments on a ball
-56-
The moment applied to the outer ring of Fig. 3 can 4. Starvation effect to rolling viscous
be expressed by Equation (3.9): resistance
Mr =( d2 ＋ r cosα ) (F ＋F )−m cosα
p
b o r,o s,o o o
4.1 Theoretical regression of rolling viscous resistance
The main undertaking of EHL is to study oil film
=( ＋ r cosα ) F ＋
d p d p thickness. Hamrock-Dowson 8), Chittenden et al. 9) and
b o F ＋M cosα
r,o s,o o o
2 2 others are proposing many theoretical regression
…………………… (3.9) equations. On the other hand, only a few reports exist
This is the friction torque of the bearings generated on rolling viscous resistance, and with regard to line
by the contact between the ball and raceway. contact, no equations based on EHL theory are found
except for Zhou-Hoeprich's equation 6). Also, with
Fr,o regard to point contact, few reports exist on equations
Fs,o = F@,o−F!,o to easily obtain rolling viscous resistance. The equation
proposed by Houpert 10) lacks material parameter and no
mo consideration was given to a viscosity-pressure factor. It
is natural to think that the rolling viscous resistance is
affected by the viscosity-pressure and therefore we do
not adopt Houpert’s equation in this paper.
Fujiwara is proposing the theoretical regression
equation regarding point-contact rolling viscous
resistance by defining dimensionless rolling viscous
resistance with Equation (4.1), and summarizing it as
Equation (4.2) using a point-contact dimensionless
Fig. 3 Forces and moment on the outer race used to number of Dowson, etc. 11).
calculate outer ring friction torque
Frαo
FHD = ………………………………… (4.1)
re2
FHD = 44.6U 0.694G 0.961W 0.401｛1−0.962exp (−0.818k)｝
The rotational speed, orbital speed and the inclination ………………………………… (4.2)
of the rotational axis of the ball are determined so that We used the equation for rolling viscous resistance
the force and moment of the ball are balanced and the under sufficient lubrication to obtain the rolling viscous
friction torque is minimized. resistance under air oil lubrication by multiplying a
By determining the behavior of the ball in a certain reduction factor to account for the lubricant starvation.
state, the slip distribution and rolling viscous resistance
can be calculated. The tangent force is the contact 4.2 Reduction factor of rolling viscous resistance
pressure at the observation point multiplied by the When a ball passes on the raceway, lubricating oil on
coefficient of traction. The coefficient of traction can be the raceway is pushed away by the ball but moves back
given as the function of slip ratio. to the center of the raceway before the next ball
We adopted Lee-Hamrock’s circular model 7) shown passes, thus the lubricating oil quantity at the EHL oil
in Equation (3.10). film inlet is recovered. It is considered that the recovery
quantity of lubricating oil is affected mainly by the ball
s / sm passage cycle T, kinematic viscosityν, and long axial
μt = =μt max ……………… (3.10)
radius of contact ellipse a . The longer the passage
1＋ (s / sm ) 2
cycle of ball T, the larger the recovery amount. The
In this model, the maximum coefficient of traction smaller the kinematic viscosityν, the greater the
μmax and the slip ratio sm must be assumed so we have recovery amount because lubricating oil can move
set μt max=0.05 and sm=0.003. more easily. The smaller the contact ellipse long axial
The rolling viscous resistance is calculated radius of the ball, the greater the recovery amount
considering the starvation effect which is discussed in because the distance the lubricating oil has to move is
detail in the next section. The numerical calculation shorter. It is also possible that centrifugal force on the
considers the forces due to this slip and rolling viscous lubricating oil may also affect the recovery amount. With
resistance for a convergent calculation. a large centrifugal force, lubricating oil moves outward
on the raceway and therefore the recovery amount is
reduced on the contact area on the inner ring side.
-57-
However, the recovery amount is increased on the In addition to rolling viscous resistance, other factors
outer ring side because the lubricating oil is forced to to impact the friction of rolling bearings are differential
move toward the center of the raceway. Since it is not slip, spin and elastic hysteresis loss. The behavior of
possible to separate the torque produced at the inner the ball is determined by the balance between traction
ring and outer ring by the experiment, the impact of the due to slip produced by various frictions and the
reduction of torque at the inner ring side offset by the resistance against those frictions. Therefore, resistance
increase of torque at the outer ring side is considered cannot be separated into these various factors.
using angular velocity ω. Although lubricating oil However, we consider that the sum of the friction
quantity affects starvation, in the case air oil lubrication, torques given by each factor is the friction torque of the
the impact was determined to be relatively small as a bearings. In this study, in order to identify the reduction
result of the experiment shown in Fig. 4, so we ignore factor of rolling viscous resistance, we compared the
will it in this discussion. friction torque obtained from the experiment removing
the calculated friction torques of differential slip, spin,
Bearing: inner diameter 100 mm, ceramic ball specification
elastic hysteresis loss and friction within the cage
Preload: 2360 N guiding surface, and the friction torque due to rolling
Lubricating oil: ISO VG32
0.4
viscous resistance obtained by calculation assuming
sufficient lubrication. We defined the value of the rolling
viscous resistance obtained from the experimental data
Friction torque N・m
0.3
divided by the calculated value of sufficient lubrication
as the experimental reduction factor of rolling viscous
0.2 resistanceφre . We determine C, sT, sν, sa, sω of
Equation (4.4) so that the square of the difference
20000 min-1
0.1 betweenφr andφre becomes the smallest. We can
15000 min-1
then obtain mathematically precise values if we conduct
10000 min-1
0
multiple linear regression. However, for the purpose of
0 0.5 1 1.5 2 this report, due to the accuracy of the experiment and
Lubricating oil quantity mL/h
assumptions given, we do not have to define accurate
Fig. 4 Effect of oil quantity on friction torque
values. Instead integers or similar values such as
0, ±1/2, ±1, ±2, …can be selected for sT, sν, sa and sω.
The reduction factor of the rolling viscous resistance For C, it is sufficient to round to two significant figures
is a function of the amount of lubricating oil recovered on the average of ratios betweenφr andφre when C=1
and therefore can be considered to be as a function of is assumed.
the ball passage cycle T, kinematic viscosityν and Skipping the details of the data, we identifiedφr under
long axial radius of contact ellipse a . The reduction the above considerations using experimental data of
factor for the rolling viscous resistance can have a friction torque by varying the bearing size, ball material,
value between 0 and 1 and it is adequate to consider speed of rotation, preload and lubricating oil viscosity,
that each parameter has a standard value when we could obtain the following equation:
determining the starvation factor 1. Therefore, we
T 0.5
define the reduction factor of the rolling viscous φr = 3.5 × 10-11 ………………………… (4.5)
resistance in the following equation: νa 2
S Sν Sa Sω
This matches with the physical perception that the
T Tν a ω longer ball passage cycle T makes largerφr because it
φr = C’ …………………… (4.3)
T0 ν0 a0 ω0 allows longer time for recovery of lubricating oil, smaller
Where, when T＞T0, T = T0 kinematic viscosity ν makes largerφr because it allows
whenν＞ν0, ν=ν0 lubricating oil to move quickly for recovery, and smaller
when a＞ a0, a = a0 contact ellipse long axial radius a makes largerφr
when ω＞ω0 ω = ω0 because the distance lubricating oil needs to move for
However, since it is difficult to obtain the standard recovery is shorter. In addition, ω which we introduced
values of T0， to consider the impact of centrifugal force was
ν0，a0 andω0, by experiments, Equation (4.3)
is converted to Equation (4.4) to identify C, ST, Sν, Sa, Sω. negligible. Since the coefficient 3.5 x 10-11 includes T0,
ν0 and a0, it has a dimension of [m4/s1.5].
C’ ST Sa Sω
νν a ω
S
φr = S Sν Sa Sω T
T0 T ν0 a0 ω0
ω Sa
=C T T ν a ω ………………………… (4.4)
ν S S S
-58-
friction torques, and added them to the final friction

5. Comparison of the calculated value torque. We referred to Kakuta’s analysis 12) for the
and experimental value of friction calculation of elastic hysteresis loss and calculated the
torque considering starvation friction within the cage guiding surface as the boundary
We calculated the rolling viscous resistance using the friction caused by the centrifugal force. The effect of the
reduction factor obtained in section 4.2 and obtained elastic hysteresis loss and the friction within the cage
the friction torque of the rolling bearings considering guiding surface is small. The friction torque produced
impact of slip, etc. We compared this calculated value between the ball and raceway ring accounts for 60 -
with the value from the experiment. Although the friction 90% of the overall friction torque.
force of elastic hysteresis loss and friction within the Fig. 5 shows the experimental and calculated values
cage guiding surface should also be included in the of friction torque for comparison. Fig. 5 (a) shows the
simultaneous equation, their impact is relatively small. results of angular contact ball bearings with an inner
So we obtained them individually, converted to the diameter of 70 mm, using ceramic balls, a standard
Bearing: inner diameter 70mm Bearing: inner diameter 70mm

Ceramic ball specification Ceramic ball specification
Preload: 1470N Lubricating oil: ISO VG32 Preload: 1470N Lubricating oil: ISO VG10
0.5 0.5
Experimental Experimental
0.4 Calculated 0.4 Calculated
Friction torque N・m 0.3

0.3
0.2 0.2
0.1 0.1
0 0
0 5000 10000 15000 20000 25000 0 5000 10000 15000 20000 25000
Rotational Speed min-1 Rotational Speed min-1
(a) Comparative example 1 (b) Comparative example 2
Bearing: inner diameter 70mm Bearing: inner diameter 70mm

Ceramic ball specification Steel ball specification
Preload: 1470N Lubricating oil: ISO VG68 Preload: 1470N Lubricating oil: ISO VG32
0.5 0.5
Experimental Experimental
Calculated Calculated
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0 5000 10000 15000 20000 25000 0 5000 10000 15000 20000 25000
Rotational Speed min-1 Rotational Speed min-1
(c) Comparative example 3 (d) Comparative example 4
Bearing: inner diameter 100mm Bearing: inner diameter 70mm Ceramic ball specification
Ceramic ball specification Rotational speed: 1800min-1
Preload: 2360N Lubricating oil: ISO VG32 Lubricating oil: ISO VG32
0.5 0.5
Experimental
0.4 Calculated
0.4
0.3 0.3
0.2 0.2
0.1 Experimental 0.1

Calculated
0 0
0 5000 10000 15000 20000 25000 0 1000 2000 3000
Rotational Speed min-1 Preload N
(e) Comparative example 5 (f) Comparative example 6
Fig. 5 Comparative example of experimental results to calculated values of friction torque
-59-
preload applied, and lubricated with ISOVG32 References

lubricating oil. The calculated values are a little smaller 1) Hiroki Fujiwara, Kenji Fujii: Rolling Bearing Torque in Oil
than the experimental values however they are roughly Bath Lubrication, Japan Society for Precision
consistent. Fig. 5 (b) and (c) are the results with Engineering Spring Conference Academic Lecture
different kinematic viscosity of lubricating oil. Fig. 5 (d) Proceedings (2002), 219.
2) Takashi Nogi: An Analysis of Starved EHL Point
shows the results with steel balls and Fig. 5 (e) is the
Contacts with Reflow, Tribologist, 59 (2014) 239-250.
result with a different bearing size. Fig. 5 (f) shows the
3) Kenichi Shibasaki, Masato Taniguchi, Marie Oshima:
comparison with a modified preload. In all cases, the
Development of Numerical Method for Coupled
calculated values duplicated the experimental values Simulation of Starved EHL and Macro Flow,
with accuracy sufficient for practical application. Proceedings of Japan Tribology Conference Tokyo
(2010-5), 245-246
4) Kenji Fujii: Research on High-Speed Rolling Bearings,
6. Conclusion
Meiji University, Doctoral Dissertation (2001).
We proposed an estimation method for friction torque 5) A. B. Jones: A General Theory for Elastically
of air-oil lubricated angular contact ball bearings. Since Constrained Ball and Radial Roller Bearings under
air-oil lubrication does not achieve a sufficient Arbitrary Load and Speed Conditions，Trans. ASME J.
lubrication condition, we used experimental values for Basic Eng., 82 (1960) 309-320．
6) R. S. Zhou and M. R. Hoeprich: Torque of Tapered
friction torque to identify the reduction factor of the
Roller Bearings, Trans. ASME, J．Tribol., 113 (1991)
rolling viscous resistance. This was done by comparing
590-597．
the rolling viscous resistance estimated from the
7) R. T. Lee and B. J. Hamrock: A Circular Non-Newtonian
experimental values of friction torque and the calculated Fluid Model: Part I – Used in Elastohydrodynamic
values of rolling viscous resistance of sufficient Lubrication, Trans．
lubrication. The friction torque of air-oil lubrication can ASME, J. Tribol., 112 (1990) 486-496．
be calculated by solving the equilibrium of forces and 8) B. J. Hamrock and D. Dowson: Isothermal
moments considering the rolling viscous resistance Elastohydrodynamic Lubrication of Point Contacts, Part
multiplied by the reduction factors, differential slip, 2 - Ellipticity Parameter Results, Trans. ASME, J. Lub.
traction produced by the spin, and adding the elastic Tech., 98 (1976) 375-383．
hysteresis loss and friction within the cage. 9) R. J. Chittenden, D. Dowson, J. F. Dunn and C. M.
This technology can be used not only for improving Taylor: A Theoretical Analysis of the Isothermal
Elastohydrodynamic Lubrication of Concentrated
accuracy of the selection of high-speed bearings and
Contacts II. General Case, with Lubricant Entrainment
preloads, but also for contributing to further
along Either Principal Axis of the Hertzian Contact
development and improvement of high-speed bearing
Ellipse or at Some Intermediate Angle, Proc. R. Soc．
and spindle design. Lond., A 397 (1985) 271-294．
10) L. Houpert: Piezoviscous-Rigid Rolling and Sliding
Traction Forces, Application: The Rolling Element-Cage
Pocket Contact, Trans. ASME, J. Tribol., 109 (1987)
363-371．
11) Hiroki Fujiwara: Rolling Viscous Resistance of Point
Contact EHL, Proceedings of Japan Tribology
Conference Tokyo (2009-5), 119-120.
12) Kazuo Kakuta: Friction Moment of Radial Ball Bearings
under Thrust Load, Transactions of the Japan Society of
Mechanical Engineers 27-178, 3 (1961) 945-956.
Photo of author
Hiroki FUJIWARA
Advanced Technology
R&D Center
-60-
[ Technical Paper ]
Kohei HIGASHI*
Hajime ISA*
Takakazu KITAGAWA*
Takashi ONISHI**
Kazuhito OHASHI**
The advantage of superfinishing is that, the fine surface of workpiece can be

obtained by the transition of machining states from cutting action to finishing
one, while the condition of superfinishing stone surface changes self-
sharpening to wear during one processing cycle. In conventional superfinishing
process, to obtain the fine surface, the superfinishing stone, machining
condition and timer control have been determined by empirical laws. In this study, we have monitored the
machining force during machining process, and have judged the roughness of workpiece by calculating
the machining force ratio.
In view of these issues, the authors developed a

1. Introduction technology to monitor the superfinishing process in real-
Rolling bearings (hereafter, bearings) are said to be time 1), 2). In this study the in-process monitoring
the “central core” of the machine industry and to have technology was controlled to determine the finish
supported the growth of Japan's manufacturing industry accuracy over a preset threshold, focusing on the
(such as automobile and machine tools) with high machining force ratio, which is the ratio of the thrust and
performance and consistent quality. Bearings are principal forces applied to the grinding stone. In
always required to provide high quality, low cost and addition, the improvement over a conventional
short lead time as machine components for high- machining cycle based on fixed time was examined and
volume production. This is the reason for the is included in this report.
designation “central core” and production technologies
to achieve these requirements are also required.
2. Characteristics of superfinishing and
A high level of performance is required for these
machining models
bearings from three basic elements: low torque, low
noise and long life. Superfinishing machining 2.1 Characteristics of superfinishing
technology (hereafter, “superfinishing"), which is Superfinishing is a machining process in which a
presented in this paper, is a very important machining stick-shaped, fine-grained stone with a relatively low
process in bearing manufacturing, has a significant hardness is applied to a work piece under a certain
impact on these three elements of bearing pressure with oscillating motion, rotational motion and
performance. feed motion between the stone and the work piece and
Conventional superfinishing has typically involved a large amount of processing solution is supplied 3).
selecting grinding stones and machine conditions based Fig. 1 shows the transition cycles of the stone during
on experience and determining machining time based superfinishing. An important characteristic of
on desired accuracy. In addition, the machining superfinishing involves the initial contact between the
requirements are affected by the accuracy of the prior work piece and the stone and subsequent relative
process and the condition of the grinding stones. motion which creates new cutting edges. This action is
**Production Engineering R&D Center

**Graduate School of Natural Science and Technology, Okayama University, Division of Industrial Innovation Sciences
-61-
called self-dressing of the grinding stone. 2.2 Superfinishing model

Fig. 1 (a): The initial stage of the cycle is called the Fig. 2 shows a general model of the superfinishing
grinding state, where the stone removes the top rough process. Pressure is applied vertically to the grinding
surface of the work piece while the stone also self- stone which oscillates perpendicularly to the rotation of
dresses, smoothing the work piece. the cylindrical work piece. When focusing on the
Fig. 1 (b): The final stage of the cycle is called the machining action of the grinding stone, it moves in a
polishing state, where self-dressing of the stone sinusoidal motion on the surface of the work piece. The
concludes and the stone becomes loaded and there is maximum velocity vector on the below sinusoidal curve
a rapid improvement in the roughness of the work is called "grinding directional angleθ" and is expressed
piece. by Equation (1).θis an important indicator of
Fig. 1 (c): A new work piece is introduced which superfinishing processing conditions which is commonly
causes rapid self-dressing of the grinding stone due to used.
the roughness of the new work piece. This overall cycle
repeats. Although required in normal grinding, the a ： Grind stone oscillation amplitude (mm)
grinding stone is not dressing in the superfinishing f ： Grind stone oscillation frequency (c/min)
process. D： Work piece diameter (mm)
N： Work piece rotational speed (min-1)
For machining accuracy, superfinishing produces a
good non-directional surface roughness due to the θ=tan-1 (af/DN ) ………………………………（1）
smaller amplitude motion and lower machining force
than ordinary cutting or grinding. In addition, since In theory, a stronger grinding action results from a
superfinishing generates less heat during machining, it largerθand a stronger polishing action results from a
minimizes the affected work layer. As a result this smallerθ.
process ensures the machined raceway surface is However, in reality, since there are countless grains
acceptable for bearings with high wear resistance. on the grinding stone, a smooth and quality finish can
be obtained.
Oscillation φ
Grind stone Pressure πa fcosφ
Pressure
Work piece πa f
f
(a) Grinding state a

Stone
Loaded θ
Polishing surface Work piece
Vs
Vmax Vw
(b) Polishing state
Fig. 2 Superfinishing model

Pre-processed
Loaded surface
(c) Self-dressing
Fig. 1 Mechanism of superfinishing cycle
-62-
2.3 A superfinishing example Coolant

The raceway, in contact with the bearing rolling Sizing sensor nozzle
elements (balls, rollers, etc.), is required to have good
surface roughness and quality surface properties, Three-component
Work piece force sensor
particularly with no surface protrusions, to minimize
noise.
As an example, Fig. 3 shows a comparison of the
Oscillation
inner ring raceway surfaces of a deep-groove ball device
bearing after grinding and after superfinishing (part z
Grind stone
number 6203 with envelope dimensions of: ID = 17 mm, Laser
Stone holder displacement
OD = 40 mm & width = 12 mm). An after grinding sensor
surface roughness of Ra 0.3 was improved to Ra 0.03 x Air cylinder
y
after superfinishing with good surface properties close
Fig. 4 Experimental setup
and no significant protrusions.
Table 1 Sensing device and evaluation points

Sensing device Displacement Evaluation item
(1) Sizing sensor Dimension change Stock removal
(2) Laser displacement Advancement of
grind stone table Wear of stone
sensor
Thrust force
(3) Three-component Grinding resistance Principal force
force sensor against grind wheel Force on the oscillating direction
Ra 0.3 Ra 0.03
4. Investigation of superfinishing
machining characteristics
4.1 Test conditions
After grinding process After superfinishing process To understand the machining characteristics of
Fig. 3 Improvement of surface roughness superfinishing and to estimate the formation process of
the finished surface, an experiment was conducted
under the test conditions shown in Table 2.
The test used a grinding stone wheel (WA1200) with
3. Experimental setup
a thickness of 5 mm and the work piece was made from
Fig. 4 shows the experimental apparatus used for bearing steel pre-processed to a cylindrical shape with
this study. In order to model the setup shown in Fig. 2, a surface roughness of Ra 0.3 – 0.4. In addition, water-
a cylindrical stepped round bar was selected as the insoluble grinding fluid was used as the coolant. The
work piece. The work piece is supported on both sides grinding directional angleθselected was 3.2˚ and
and rotated by the lathe shaft. The grinding stone is based on experience is positioned between grinding
held by a fixture and pushed horizontally against the and polishing processing, allowing for continuous
work piece by an air cylinder in a direction transitions from the grinding state to the polishing state.
perpendicular to the work piece. The grinding stone, The machining time was set at 20 seconds max. and
holding fixture and air cylinder are attached to a table allowed an arbitrary 2 – 20 seconds to find the
which oscillates axially with respect to the work piece. machining state during the process.
Table 1 lists the sensing devices used in this study to
monitor the machining process and the items that were Table 2 Experiment condition in one step
evaluated. They include a sizing sensor to measure the Spec：WA1200 20A
Grinding stone
work piece diameter, a laser displacement sensor to Size：t5×W30 mm
measure the movement of the grinding stone table as Work piece SUJ2 φ20（HV770）
Grinding stone oscillating amplitude 0.75mm
an indication of the stone wear, and a three-component
Cooling flow 2.0L/min (Water-insoluble)
force sensor, installed between the stone and the air
Grinding stone oscillating frequency 1870 c/min
cylinder, to measure the machining force applied to the
Shaft rotational speed 1230 min‐1
stone. In this study, these sensors were used to Grinding direction angle 3.2˚
determine the finish accuracy by monitoring the Pressure on grinding stone 0．54 MPa
machining force applied to the grinding stone. Machining time 2∼20s
-63-
4.2 Test results Work piece

Fig. 5 shows a diagram of the sensing devices in the
machining area including forces and distances. Fig. 6
Grinding stone
shows the monitoring data over 20 seconds of Laser
machining under the test conditions defined in Table 2 displacement
with the elapsed time on the horizontal axis. The top
R/2 sensor
figure shows a plot of the advancement of the grind

stone stand d and stock removal R. The bottom figure In-process P
shows the plot of the thrust force P and principal force Q gauge
applied to the grinding stone. Q d
The movement of dimension d barely changes after 3
seconds from the beginning of machining. Therefore, it
Fig. 5 Diagram of sensing devices in machining area
is theorized that the grinding stone became loaded and
the machining state transitioned to polishing. Looking at
the change in the machining force, the thrust force P is
0.4 10
constant at 80 N as the pressure to the stone is
grinding wheel mm
Stock removal µm
Advancement of
constant; however, we could verify that the principal Stock removal
force Q gradually decreases after 3 seconds. 0.2 5
Focusing on this change, the machining force ratio Advancement of grinding wheel
P/Q (hereafter “force ratio”) was added to the plot. It 0

0
was observed that after a brief decrease, the force ratio 100 6
Machining force ratio P/Q

gradually increased over time. This trend is theorized to
Machining resistance N
80 5
be related to the roughness of the finishing surface. Thrust force P
60 4
4.3 Formation process of a finished surface 40 3

In order to understand the roughness formation P/Q
20 2
process on a finished surface, the relationship between Principal force Q
force ratio and roughness was examined. Experiments 10 1
0 5 10 15 20
were conducted under 7 different conditions with Time s
machining times of 2 – 20 seconds under the test
Fig. 6 Machining characteristics during one stage processing
conditions shown in Table 2. As a result, the
relationship between the machining force ratio (after
machining is completed) and the roughness of the
finished surface was determined. The results from each Correlation coefficient：-0.97
0.4 5
Roughness of finished surface µm
experiment are shown in Fig. 7. As the machining time

increased, the machining force ratio increased while the 4
0.3
Machining force ratio
roughness of the finished surface reduced. By Machining force ratio

calculating the correlation coefficient between the force 0.2 3
ratio and the surface roughness at 7 points, a very high
correlation of -0.97 was obtained. Therefore, it is highly 0.1 2
probable that roughness of the finished surface can be
determined by monitoring the machining force ratio. Roughness of finished surface
0 1
0 5 10 15 20
Time s
Fig. 7 Relationship between machining force ratio and

surface roughness
-64-
the end of the machining cycle (20 seconds total), the

5. Machining characteristics under force ratio was slightly below 4 and the roughness of
forced state transition the finishing surface (Ra) was 0.052.
5.1 Test conditions
When using the superfinishing process during
normal production, it is necessary to remove the work 0.4 10
grinding wheel mm
Stock removal µm
Advancement of
piece’s preprocess surface (stock removal) and
Advancement of grinding wheel
achieve final finish accuracy. To meet these 0.2 5
requirements, it is common to program machining
conditions for both grinding and polishing and have Stock removal
0 0
the cycle changed sequentially by the timer. That way 100 6
Thrust force P
stock removal is ensured and the stone is forced to be

80 5
Machining force N
loaded for polishing.
Therefore, the conditions were set for the test 60 4
P/Q
machine grinding and polishing processes in two stages
40 3
and the machining characteristics were investigated. Principal forceQ
Table 3 shows the experimental conditions. The 20 2
grinding stone, coolant and work piece were equivalent 10 Grinding Polishing 1
0 5 10 15 20
to those shown in Table 2 and the wheel swing speed
Time s
was set to 670 min-1 for both grinding and polishing.
The grinding directional angle was changed by reducing
the stone oscillation frequency from 1,870 c/min (cycles
per minute) for grinding to 500 c/min for polishing to
constitute the forced transition condition from grinding to
polishing. The machining time for both grinding and
1µm
polishing each lasted 10 seconds. Ra 0.052

1mm Rz 0.649
Table 3 Test conditions in two steps Fig. 8 Machining characteristics during two stage processing
Spec：WA1200 20A
Stone
Size：t5×W30 mm
Work piece SUJ2 φ20mm（HV770）
Stone oscillating amplitude 0.75mm
6. Determination of finishing accuracy
Coolant flow 2．0L/min (Water-insoluble)
via threshold
Machining condition Grinding (1st step) Polishing (2nd step) 6.1 Method for determining threshold
Stone oscillating frequency 1870 c/min 500 c/min A high correlation between the machining force ratio
Main shaft rotational speed 670 min-1 670 min-1 P/Q (force ratio) and the finish accuracy (roughness)
Grinding direction angle 6.0˚ 1.6˚ was verified by the results shown in Fig. 6 and 7.
Stone pressure 0.54 MPa 0.54 MPa Therefore, to estimate the finish accuracy (roughness)
Machining time 10s 10s in real-time during the machining process, the force
ratio was monitored during the machining process and it
was compared to a threshold. Fig. 9 shows a diagram
5.2 Test results of the signal processing system. The output signals
Fig. 8 shows the monitoring data from the machining from the three component force sensor are transferred
test conditions listed in Table 3. For the first 10 to the processing unit after the A/D conversion. The
seconds, the large machining directional angle is set to processing unit has an output terminal which sends the
the grinding state. During this time, the grinding wheel signal to the air cylinder which in turn applies pressure
and stock removal are both progressing at a constant to the stone.
speed and the machining force of the grinding stone Fig. 10 shows the flowchart used for roughness
and the machining force ratio (P/Q or force ratio) are determination. The two-step machining conditions that
constant. After 10 seconds, when the grinding were used are listed in Table 3. At the beginning of the
directional angle is reduced to transition into the machining process, the part is ground for 10 seconds
polishing state, both stone advancement and stock for pre-process stock removal. Next, after switching to
removal cease and the force ratio starts to increase. At polishing conditions, the force ratio is continuously
-65-
Three-component calculated and the result is compared to a

force sensor predetermined threshold at regular intervals. The scan
A/D
Thrust force P conversion period was set to a 0.1 second interval. When the force
ratio becomes larger than the predetermined threshold,
Air cylinder
Principal force Q CPU the finishing process is deemed complete and a signal
processing is sent to the air cylinder to retract the grinding wheel
away from the work piece.
Fig. 9 Signal processing system
6.2 Threshold determination results
The plots in Fig. 11 show the measurement results
Start of machining of P, Q and P/Q when the thresholds are set to 3.4 and
3.7, respectively, as well as the surface roughness
Rotation of work piece/ before and after the machining process. When the
oscillation of stone
threshold was set to 3.4, P/Q exceeded 3.4 after 5
Advancement of grinding wheel seconds of the polishing cycle. Roughness of the
finished surface (Ra) was 0.064 but small irregularities
1st step (grinding) condition such as grinding marks were found along the
Timer 10 seconds
roughness curve. When the threshold was set to 3.7,
P/Qexceeded 3.7 after 17 seconds of the polishing
2nd step (polishing) condition cycle. However, this time the roughness of the finished
surface was 0.037 and no grinding marks were
0.1 sec scanning observed along the roughness curve. Based on these
results, it was verified that setting a higher threshold for
P/Q：Threshold
P/Q results in a longer polishing time for improved
roughness of the finished surface and no grinding
Forced setback of stone marks.
Stop work piece/ 6.3 Relationship between the force ratio (P/Q )
stone operation
and surface roughness
Machining ends In order to obtain a straightforward relationship
between the force ratio (P/Q) threshold and roughness,
Fig. 10 Judgement signal flow chart roughness and polishing time were plotted against P/Q
End of machining with force ratio P/Q=3.4 End of machining with force ratio P/Q=3.7
100 6 100 6
Thrust forceP
Thrust forceP
Machining force N
Machining force N
80 5 80 5
Threshold
60 4 60
Threshold
4
40 3 40 3
P/Q P/Q
20 2 20 2
Principal force Q Principal force Q
10 1 10 1
0 10 20 30 0 10 20 30
Time s Time s
Before After threshold Before After threshold
machining reached machining reached
1µm
1µm
1mm 1mm
Ra 0.364 Ra 0.064 Ra 0.340 Ra 0.037

Rz 2.717 Rz 0.814 Rz 2.492 Rz 0.459
Fig. 11 Comparison of surface roughness after processing using threshold force ratio
-66-
thresholds set at 0.1 intervals (shown in Fig. 12 ). As

the threshold is increased, the polishing time lengthens 7. Conclusions
and roughness improves. Technology was created to monitor the superfinishing
However, when the threshold exceeds 3.7, there is process via an in-process gauge and the following
no significant improvement in roughness even though results were obtained:
the machining time continues to increase. (1) A method was proposed to automatically determine
To compare the superiority of this targeted machining the finish accuracy by in-process monitoring of
process with a conventional method using fixed machining force ratio, thus monitoring the machining
machining time, additional testing was performed. In state in the superfinishing process.
order to obtain a roughness (Ra) of 0.052 (indicated by (2) Under polishing conditions, the machining force ratio
the red dotted line in Fig. 12 ), which is equivalent to P/Q , which is calculated from the thrust force P and
the finishing surface roughness of the conventional the principal force Q applied to the stone, was found
method (as shown in Fig. 2 ), a threshold of 3.6 was set to increase as the machined surface roughness
for this test. The time required for the polishing step was became lower.
6 seconds, which is a 40% reduction from the 10 (3) The machining cycle time can be reduced by ending
seconds used by the conventional machining method. the machining process as soon as the machining
From the above results, it is possible to control the force ratio P/Q exceeds a preset threshold. This is
roughness of the finishing surface by monitoring the compared to the conventional machining method of
machining force ratio and governing the process using setting machining conditions of all parts for the same
the force threshold. As a result the productivity can be amount of time.
improved by reducing the polishing time compared to a
conventional cycle which machines parts for a fixed References
amount of time. 1) Takahiro Iguchi, Taskashi Onishi, Kazuhito Ohashi,
Kohei Higashi, Shinya Tsukamoto: Development of
Monitoring Technology for Superfinishing Process, The
Japan Society for Abrasive Technology, Proceedings of
0.20 20 Abrasive Technology Conference (2011) 297.
Roughness of finished surface µm
2) Shingo Yamashita, Takashi Onishi, Kazuhito Ohashi,

Kohei Higashi, Takahiro Iguchi, Ryosuke Yamamoto,
0.15 15
Polishing time s
Shinya Tsukamoto: In-Process Monitoring of Machining

State in Superfinishing – Determination of Machining
0.10 10
Roughness of State by Machining force ratio -, The Japan Society for
finished surface
Abrasive Technology, Proceedings of Abrasive
0.05 5 Technology Conference (2012) 45.
Polishing time 3) Toshikatsu Nakajima, Norihiko Narutaki: Kikai Kakogaku
0 0 (1984) 195.
3.0 3.4 3.5 3.6 3.7 3.8
4) T. Onishi, K. Ohashi, K. Higashi, S. Yamashita, T. Iguchi
Grinding
state Threshold against P/Q and S. Tsukamoto, In-Process Monitoring of Machining
State in Superfinishing, Proceeding of the 7th
Fig. 12 Relationship between finished surface roughness
International Conference on Leading Edge
and polishing time at different threshold force ratios
Manufacturing in 21st Century (LEM21), (2013) 89-92．
Photo of authors (titles are at the time of release)
Kohei HIGASHI Hajime ISA Takakazu KITAGAWA Takashi ONISHI Kazuhito OHASHI
Production Engineering Production Engineering Production Engineering Graduate School of Natural Graduate School of Natural
R&D Center R&D Center R&D Center Science and Technology, Science and Technology,
Okayama University, Division of Okayama University, Division of
Industrial Innovation Sciences Industrial Innovation Sciences
-67-
[ New Product ]
Lightweight Bearing for Robot
Sebastien BRISSON*
Robotic arms do have many qualities, but intelligence is not their best point.
It is necessary and important to be programmed move by move. They work in
a XYZ coordinates world associated to θX θY θZ for rotations. Moreover,
programming and maintenance are matter for specialists. Their size and
potential tremendous force imply large secured areas and heavy procedures
when an incident occurs.
Some trends can be found such as Cobotics that have new technologies of
motion control and can coexist with surrounding humans.
Based on its know-how, experience and Market understanding, NTN-SNR prototyped a bearing concept
that aims to accompany the Robotics Megatrends. Making the bearing lighter, less expensive by tuning its
stiffness according to real needs of coming technologies.
process, when parts are conveyed by robots, the robots

1. Introduction have to grab the parts in the same way and transport
Many characteristics are required for industrial them to the predefined locations following the same
robotic arms; however, the following two points are trajectory every time. Therefore, the positioning of the
particularly important: robot and objects must always be accurate. This is true
• Robot arm design must conform to the demand for with welding and painting processes as well, so
several sizes and use in a broad range of applications accurate positioning and stable operation is again
while also considering the cost of the robot arm. required.
• Higher functionality of robot arms and improved Fig. 1 shows a typical perpendicular articulated 6-
operability axis robot for industrial use. The programming for
operation is achieved by expressing and modeling the
To satisfy these requirements, particularly from a accurate motion of the robots in X, Y, Z and θX, θY, θZ
functionality standpoint, robot arms use technology to axes, as shown in Fig. 2, for each process to be
determine various conditions in addition to high rigidity, accomplished.
broad operational range and easy operation. Therefore, Therefore, good performance of today’s industrial
bearings supporting the robot arms are required to be robot arms is determined by the following items:
designed to conform to these technology trends. • Accuracy of positioning
In this paper, we briefly discuss industrial robots and • Accuracy of positioning reproducibility
present the lightweight bearings for robots developed • Rigidity
by NTN-SNR. With the performance of these items, motions are
combined to establish automatic operation of an
industrial robot. In addition, sufficient rigidity to support
2. Structure and capability of robots the payload capacity of the robot is required.
The basic functionalities of today’s industrial robots Fig. 3 shows the performance required for the current
have evolved first from positioning, then to grabbing industrial robots. Today bearings for industrial robots
objects with "hands" to more detailed tasks such as are heavy due to the necessary control of axis, high
assembling, welding, painting, etc. which previously precision, high rotational speed (acceleration) and high
was the work of humans. For example, in the assembly rigidity required.
*NTN-SNR ROULEMENT Industry Innovative Developments
-68-
Fig. 4 shows an image of the industrial robot line-up.

In general, the maximum payload capacity is approx. 3. Cobotics and new technology
1,500 kg, however, the majority of applications are 20 3.1 Cobotics
kg or less. High rigidity is required not only for Cobotics derived from "cobot", an abbreviation of
maintaining accuracy of positioning, as mentioned collaborative robot, means a collaborative robot that
earlier, but also for supporting the weight of the operates in an environment surrounded by humans.
manipulator. Fig. 5 shows examples of operation. Cobots detect the
co-working humans and other robots when they
approach the cobots and work on their own without
making contact with others. Cobots also work for
automated or assisted driving of vehicles. For example,
they can sense how humans work and assist them with
their work or change their behavior accordingly.
Manipulator The payload of cobots is usually 20 kg or less and
these robots are becoming lighter and smaller.
However, these approaches are being adopted by large
units, as well.
Controller
Fig. 1 6 axis robot (manipurator and controller)
Example of autonomous cobot
Fig. 2 Modelling and programing for motion
Example of cobot to follow man’s movements

while multiplying its strength
Fig. 3 Specification of usual robots Fig. 5 Payload of industrial robots
Fig. 4 Payload of industrial robots
-69-
The cobots are mainly deployed for tasks with less

added value, time consuming tasks and repetitive tasks 4. Bearings for robots developed by
of critical process. Cobots are also aimed at reducing NTN-SNR
the load of workers and preventing injuries such as In Europe, NTN-SNR is a leading bearing
arthritis. manufacturer of reducers for robots and maintains a
high share not only for industrial robots but also for
3.2 New technologies work/motion-assist robots.
Among the technologies on robots, the following are Fig. 7 shows an example of the specifications for the
directly related to control: manipulator and bearings of common industrial robots.
• Image processing technology The rigidity required for the bottom most turning
• Force sensing technology bearings and the wrist rotating bearings at the tip of the
• Laser measurement technology arm are significantly different even for the same robot.
Fig. 6 shows some examples of movement using the Therefore, the bearings to be used in the robot must be
above technologies. These technologies enable robots individually selected depending on the task, payload,
to understand the shape of the work, support it, and position and operation of the robot.
accurately transport it after computational processing Due to the above reason, crossed roller bearings
for various movements. shown in Fig. 8 or four-point contact ball bearings are
They also assist operation with other robots and used for turning, rotating and oscillating robot joints and
adjust distance with the workers in the proximity. the bottom most bearings need to support the weight of
the manipulator.
NTN-SNR focused on developing a more lightweight
bearing while maintaining the previously mentioned
functional requirements.
Movement of vision and image treatment

End of arm Bearing
Down to 0.3kg
Down to 20 Nm/Arcmin
Base of arm Bearing

Up to 11.5kg
Up to 1800 Nm/Arcmin
Force sensing Laser measurement
Fig. 7 Manipulator and specification example of bearing
Picking Deburring Sorting

Fig. 6 Example of movement by new technologies
Fig. 8 NTN-SNR crossed roller bearing
-70-
side of the resin parts, so that no displacement on the

5. Lightweight bearings for robots outer ring raceway of both sides is created in the radial
The design concept of NTN-SNR lightweight direction.
bearings for robots is a composite material of resin and There are several ways to connect both sides of the
steel. Conventionally, inner and outer rings are specially resin parts, such as use of rivets as shown in Fig. 12 ①,
shaped and made of steel because they are attached to or to insert steel-made sleeves and fixed by bolts as
the peripheral components with bolts, etc., however, the shown in ②.
developed product uses steel only for bearing raceway
assembly and rolling elements and adopts resin for the
specially shaped part to be attached to the peripheral Outer ring
(resin made, separate
components. Sintered component pieces on both sides)
The application of resin must function without
affecting the bearing peripheral components. Outer ring
Therefore, we also considered the operating (steel made; separate
pieces on both sides)
environment of industrial robots in the selection of resin
Rolling element
material and adopted nylon resin of high-melting point
with excellent strength. Inner ring (steel part)
Also, the developed bearings can replace the
Inner ring (resin part)
conventional bearings without redesigning of robot
Fig. 10 Lightweight bearing for robot (separate type)
joints and reducers.
In the following, we present the types of lightweight
bearings for robots. Offset δ
5.1 Non-separable type

With the non-separable type, resin is injection molded
on the inner and outer rings of the deep groove ball
bearings, as shown in Fig. 9. The rigidity is increased
by reducing the internal clearance of the bearings.
Outer ring (resin part) Fig. 11 Assemble example of lightweight bearing

(separate type)
Outer ring (steel part)
Rolling element ② Steel-made sleeve ① Rivet
Inner ring (steel part)
Inner ring (resin part)

Fig. 9 Lightweight bearing for robot (non-separate type)
5.2 Separable type

The separable type is the crossed roller bearings
shown in Fig. 10, to obtain higher rigidity than the non- Fig. 12 Fastening example of lightweight bearing
separable type which adopted deep groove ball
bearings. In the assembly of the bearings, only the
inner ring is molded in a single piece with resin and the 5.3 Rigidity
resin parts of the outer rings are assembled from both The developed bearings have reduced rigidity
sides of the steel parts after the rollers are inserted. compared with the conventional bearings because the
Also, as shown in Fig. 11, the widths of the separable specially shaped part of the inner/outer rings is replaced
resin parts on each side are different. The reason is that with resin. Table 1 and Fig. 13 show an example of the
the outer rings made of steel are engaged only to one comparison of rigidity of the separable type (crossed
-71-
roller bearings) by FEM analysis.

In case of cobotics discussed in Section 3, which are 6. Conclusion
controlled by image processing and force sensors, the We have discussed the lightweight bearings for
high speed and high rigidity requirements of current robots developed by NTN-SNR. We have also
robots are not required; therefore, this reduction of summarized the recent trend and new technologies of
rigidity of the developed product should be in the the robot industry and the resulting requirements to the
permissible range. bearings in Fig. 14.
As shown in Fig. 3, the control of axis for the robot
motion has been emphasized until now. However, it is
Table 1 FEM Analysis to compare the stiffness between
developed bearing and current bearing believed that the requirements on being lightweight and
Current Developed
lower in cost will be more emphasized more in the
bearings bearings future instead of high speed and high rigidity, due to
Bearing main sizes ｍｍ φ40×φ110×26 the increasing control with image processing and
Analysis ① Axial load kN 40 support of workers by cobotics.
conditions ② Moment load Nm 1000 We developed the bearings for robots discussed in
① Axial rigidity mm this paper to cope with this trend. We will continue our
0.020 0.150
Analysis (axial movement)
development with a close eye on the trends of the robot
results ② Moment rigidity deg
0.04 0.25 industry and are poised to contribute to the
(bearing inclination angle)
development of the European and global industry by
providing this product for various industrial robots.
Example of analysis ① Example of analysis ②

(Axial rigidity) (Moment rigidity) Developed
bearings
Fig. 13 FEM analysis example
Fig. 14 Technology trend for robot and requirement of
bearing
5.4 Weight
Table 2 shows the comparison of weight between the
conventional bearings made of steel and the developed
product of the equivalent size. The developed bearings
reduced the weight by 63% for the separable type and
73% for the non-separable type compared to the
conventional bearings, resulting in a significant weight
reduction.
Photo of author
Table 2 Weight comparison between developed bearing
and current bearing
Developed Developed
Current bearings bearings bearings
Bearing type (separate (separate (non-separate,
crossed roller) crossed roller) deep groove ball)
Appearance
Bearing size mm φ40×φ110×29 Sebastien BRISSON

Weight kg 1.00 0.37 0.27 NTN-SNR ROULEMENTS
Reduction factor % ̶ 63 73 Industry Innovative Developments
-72-
Condition Monitoring System (CMS) for Wind Power Generators
Wind Doctor ® R
Data acquisition Data management/monitor/

module analysis software
NTN’s Condition Monitoring System (CMS) for Wind

Power Generators consists of a data acquisition module,
measurement sensors, and data management/monitor/
analysis software and enables monitoring of bearing
condition, etc. from a remote location. It helps early
detection of errors to prevent expansion of damages and
ensures availability of power generators.
CMS: abbreviation of Condition Monitoring System
NTN’s system products supporting stable operation

of offshore wind power generation
-73-
Development of Condition Monitoring System
Keisuke HASHIZUME*
Akitoshi TAKEUCHI*
Yuzuru TANAKA*
NTN has developed Condition Monitoring System using bearing

diagnosis technology. This article introduces CMS’s products, and
explains necessity of CMS in the market and expansion in the future.
at wind turbine sites in Japan, and have started

1. Introduction acquiring large amounts of data. In this paper, we
Aiming at the prevention of global warming, energy discuss the reason why CMS is drawing attention,
conservation and the reduction of power generation technical details of analysis through CMS, and its future
cost, demonstration and full-scale production of power technology trends.
generation devices using renewable energy are well
underway. In Japan, the feed-in tariff program for
2. Condition Monitoring System (CMS)
renewable energy (photovoltaic, wind, hydro,
geothermal and biomass power) started in July 2012 Condition Monitoring System is a system to
and their deployment is expected to accelerate. Among continuously or intermittently acquire data remotely
these, the cost of wind power generation is low among from several sensors, such as vibration sensors,
renewable energies, and many sites are actually displacement sensors and temperature sensors which
deployed as key power generation facilities in Europe, are attached to the target machinery, for early detection
which is a leading region for wind power generation. It of component anomalies based on the acquired data.
is expected that wind power generation will further By early detection of anomalies, damage progression
expand as larger wind power generation facilities are can be limited, which results in reduced maintenance
developed for off-shore deployment, and large-scale costs, and by determining the damaged area,
introduction in the Asian region is expected as well. appropriate preparation and treatment can be
NTN introduced the Condition Monitoring System performed.
(CMS) for wind power generation as a product for the
renewable energy market that supports the next-
3. Market demands of CMS
generation power industry, taking advantage of
vibration analysis technology and sensing technology. Now, we will discuss the market demands of CMS,
Through various improvements since its introduction, taking the power generation business as an example.
we now market this system as Wind Doctor ® as shown In order to reduce power generation costs, the
in Fig. 1. We are increasing the number of installations reduction of initial facility investment, the introduction of
highly efficient generation technologies and the
improvement of equipment availability are necessary.
That is, after switch-over, maintaining stable operation
is critical, and downtime must be reduced to a
minimum. In general, stable operation of power
generation is considered to be a basic requirement, and
unscheduled downtime has a significant impact on
consumers.
As a countermeasure to avoid such downtime,
Fig. 1 Data acquisition module, Wind Doctor ® periodical inspection of equipment and preventive
*CMS Engineering Dept., Industrial Machinery Division
-74-
maintenance are necessary, but it is also important to dedicated server through the Internet. Based on this
gather information beforehand in order to identify acquired data, various analyses and precise diagnoses
anomalies and their locations based on early and are done using dedicated software applications
accurate detection of changes in equipment condition In installed on the user PC. Fig. 3 shows the base screen
fact, with the gas turbine generation and steam power of this condition monitoring. When an anomaly is
generation, various sensing technologies, such as detected, an alarm is displayed on the screen, and
vibration, temperature, rotation speed, pressure, emergency emails are sent to the designated email
displacement and acoustic response, are used for addresses.
continuous monitoring and for associated maintenance This software also displays the trend view of RMS
operations. values as shown in Fig. 4 and vibration wave patterns
On the other hand, in the field of industrial machines, of acquired data as shown in Fig. 5. In addition, it also
some businesses, including those with long production has capabilities of various calculations and analyses
lines such as steel, papermaking and chemical plants, such as the envelope signal processing as shown in
are conducting preventive maintenance with immediate Fig. 6 based on vibration wave patterns as shown in
effect by introducing their own CMS with the objectives Fig. 5 and FFT spectrum analysis and display as
of maintaining supply capabilities and ensuring stable shown in Fig. 7 It also allows arbitrary setting of band-
quality. pass filters to maximize the signal-to-noise ratio (S/N
ratio), which is appropriate for displaying defect
frequencies of the bearings and step-up gears
4. Examples of equipment and applications diagnosed as damaged.
NTN is moving forward with the development and
market introduction of CMS related technology
including special sensors. In this section, we describe
some of the typical examples of NTN's commercial
products.
4.1 CMS for wind power generation

The following is the data acquisition module, Wind
Doctor ® targeted for use in wind power generation
applications as shown in Fig. 1 in the previous section.
This module works with a 100V-240VAC power
Fig. 3 Home screen of monitoring software
supply and is equipped with 17 channels of signal
inputs. It also complies with IP65 level of Japanese
Industrial Standard for electrical machinery and
apparatus protection code JIS C 0929 for its dust/water
proof performance.
In order to achieve highly stable operation, it is also
equipped with data protection and cooling capabilities
during power outages as a hedge against risk. This
Wind Doctor ® module is operated as part of a system
as shown in Fig. 2. The output data in multiple formats
from various sensors are encrypted and uploaded to a Fig. 4 Trend view of RMS value
Various Condition monitoring

sensors software Anomaly
Dection
Caution/
warning
Server
Precise
diagnostics
Reporting
Wind Doctor ®
Fig. 2 Layout of CMS for wind turbines
-75-
In addition, the raw data can be downloaded in CSV vibration data is automatically sent to the server through
format so that users can freely process the data. the Internet using a dedicated application as shown in
These operations can be run repeatedly without Fig. 9. The server determines normal/abnormal status
limitation, so they can be used as an effective data of inner rings, outer rings and rolling elements of the
analysis tool. bearings through calculation and analysis and sends
the results back to the smartphone.
Fig. 10 shows an example of the operation screen
and result display screen on the smartphone. Fig. 11
shows FFT analysis results, Fig. 12 shows the overall
values and Fig. 13 shows a trend view of the RMS
values. These data are stored on the dedicated server
Fig. 5 Wave pattern of acquired data not only on a temporary basis, but also past data can
Fig. 6 Envelope signal processing of acquired data
Fig. 8 Handy type failure detection device
Smartphone
Bluetooth
Server
Fig. 7 FFT Analysis Results
Vibration pick-up
A/D converter
4.2 Handy type failure detection device
Fig. 8 shows the handy type failure detection device
developed for condition monitoring and maintenance of
general industrial machines such as general-purpose Data display
motors and fans. Fig. 9 Layout of handy type failure detection device
This system consists of a vibration pick-up and
Operation button Judgment Results
pocket-sized A/D converter that are provided by NTN
and link to a user-provided smartphone. This product
was realized by applying bearing diagnostics
technology which has been developed over the years
with the goals of performing diagnostics “easily”,
"consistently”, “on the spot” and “in a short time”.
Conventional all-in-one type diagnostics systems
calculate and analyze the measured vibration data
within the system, and require large displays for easy Fig. 10 Operation screen and judgment screen
identification of detailed data, limits how far their size
and weight can be reduced. In contrast, by using
smartphones, displayed data can be easily recognized
by zooming in to the most detailed view.
In addition, maintenance personnel of manufacturing
equipment and production lines need to carry the
system together with tools and paperwork during their
maintenance and inspection activities.
Therefore, NTN’s handy type failure detection device
is made very compact and portable for the
measurement of vibration. In this system, the acquired Fig. 11 FFT Analysis Results
-76-
be downloaded; therefore, the data can be used for prediction based on much broader information will be
fixed-point monitoring of facilities such as displaying the enabled by integrating the data obtained from SCADA
trend view of past measurement results. with data from CMS which is useful for early detection
of anomalies.
The ultimate function required for CMS is not only
early discovery of damage, but also prediction of
remaining life. It is important to know how to minimize
the period of downtime due to anomalies; and
specification of operating limitations is frequently
requested in order to secure preparation time.
However, because the progression from anomaly
Fig. 12 View of overall values detection to failure can vary significantly depending on
the operating conditions, disturbances and
environment, predicting remaining life is fraught with
challenges. In the future it may become necessary to
employ new analysis methods and to introduce precise
diagnostics technologies incorporating not only vibration
and displacement data but also acoustic and visual
sensing data.
7. Conclusion
We have discussed why CMS is necessary and also
Fig. 13 Trend view of RMS values NTN’s products and activities. We are poised to
contribute to further improve availability and reduce
maintenance burden of not only wind power generation
but also industrial machines in general, by continually
5. Participation in the NEDO (New
advancing high functionality and high performance data
Energy and Industrial Technology
acquisition devices and providing highly accurate and
Development Organization) project
useful technology information services.
NTN is participating in the “NEDO Advanced
Practical Research and Development of Wind Power
Generation” (R&D of Smart Maintenance Technologies)
project with other organizations such as the University References
of Tokyo, to perform experimental evaluations of CMS. 1) Akitoshi Takeuchi, Takashi Haseba, Hiroshi Ikeda:
This research and development aims to increase Application of Condition Monitoring System for Wind
average availability of domestic wind power generation Turbines, NTN TECHNICAL REVIEW, No. 80 (2012)
facilities from the current level below 20% to over 23% 15-18.
by developing highly-efficient maintenance techniques
and conducting demonstration tests on actual wind
turbines. NTN will conduct demonstration tests for this
R&D project by installing Wind Doctor ® modules to the
wind power generation facilities across Japan thereby
promoting the further advancement of CMS equipment
diagnostic technology. Photo of authors
6. Future development
Power plants in general including wind power are
building remote monitoring/controlling systems
(SCADA: Supervisory Control and Data Acquisition) to
acquire operational data of their facilities and entire
sites. We believe that advanced control such as Keisuke HASHIZUME Akitoshi TAKEUCHI Yuzuru TANAKA
CMS Engineering Dept., CMS Engineering Dept., CMS Engineering Dept.,
automatic trend analysis and degradation trend Industrial Machinery Division Industrial Machinery Division Industrial Machinery Division
-77-
[ New Product ]
Introduction of "ULTAGE" series :

High Functionality Roller Bearings
Shogo TABATA*
Kouhei TODA**
Yasuhiko SHIMIZU***
When it comes to the operational demands put on equipment (such

as iron and steel machinery, or wind power generation) and industrial
machinery (such as construction and mining machinery, or gearboxes),
customers are looking to maximize output and efficiency under more
severe usage conditions, like higher vibration and impact loads. In response to these increase
demands, NTN developed the ULTAGE® Series of cylindrical ,tapered and thrust spherical roller
bearings, which deliver the highest load capacity and allowable speed by optimizing the internal
design and crowning profile. NTN will release these bearings as the ULTAGE Series which reduce
the internal friction resulting in less energy needed to operate.
1. Introduction 2. ULTAGE Cylindrical Roller Bearings

Technology development that emphasizes global The ULTAGE cylindrical roller bearing is optimized
environmental protection is well underway for every for rolling contact so that the contact pressure
industrial machine. distribution between the rolling elements and rolling
Bearings used in rotational components of industrial bearing rings becomes uniform. The features of this
machines are required to have an improved “long life”, bearing are as follows:
“high load carrying capacity”, “high speed”, and “ease of
use”. To meet this demand NTN has developed the 2.1 Features
world's-highest level of new-generation bearings known (1) Load carrying capacity (compared with the
as the “ULTAGE ®※” series. This series started with conventional cylindrical bearings)
precision rolling bearings for main spindles of machine • Basic dynamic load rating: 1.2 x
tools and was followed by standardization of new • Basic rating life: 1.8 x
spherical roller bearings, roll neck sealed type 4-row (2) Limiting speed (compared with the
tapered roller bearings for rolling mills, etc. Since then conventional cylindrical bearings)
new cylindrical roller bearings, tapered roller bearings • Max. 1.2 times under oil lubrication condition
and thrust spherical roller bearings have been added to (3) Resin cage material is applied as standard
the ULTAGE series. These bearings achieve long life, • Applicable for oil and grease lubrication
high load carrying capacity and high speed and are (4) Allowable inclination (misalignment)
discussed in the following: ULTAGE: 1/500 (0.002 rad.)
Conventional cylindrical bearing : 1/2000 (0.0005 rad.)
※ ULTAGE ® is a name created from the combination of “ultimate,” signifying refinement, and “stage,” signifying NTN’s
intention that this series of products be employed in diverse applications. This is the general name for NTN’s new
generation of bearings that are noted for their industry-leading performance.
***Industrial Machinery Engineering Dept., Industrial Machinery Division

***New Energy Engineering Department, Industrial Machinery Division
***Construction Machinery and Railway Engineering Department, Industrial Machinery Division
-78-
2.2 Structure
Stress concentration point of
Fig. 1 shows an example of the structure of NU type conventional bearing crown design
cylindrical roller bearings.
High ULTAGE
Reduction of
Conventional Bearing Design
edge stress
Contact pressure
1) Optimized internal design
Adoption of larger diameter rollers and optimized
crowning shape enabled use under tough conditions
such as larger inclination and heavy load, increasing
load carrying performance. Low
2) Resin cage material is standard
Glass fiber enhanced polyamide resin is used as the Roller length
material. Resin cages are lightweight and
characterized by low torque and low noise. Fig. 2 Contact stress distribution of rollers
Cage 2.4 High Speed Operation

Outer ring
¡Roller guiding method is adopted ( increase of allowable rotational speed)
¡Resin material cage The ULTAGE cylindrical roller bearings reduces the
※Steel-base, brass-base and machined
cages are also possible depending on contact pressure between the rollers and rolling bearing
the operating conditions
rings and improved the contact conditions between the
rollers and rib section. This improves lubrication
performance inside the bearings and reduces heat
generation allowing a 20% improvement of rotational
speed compared with the conventional bearings. Fig. 3
shows the test result of oil lubrication temperature rise.
<<Test Condition>>
Bearing part number: NU307E (φ35×φ80×21)
Radial load: 3.5 kN (5% Cor)
Bearing internal clearance: CN clearance (0.025 - 0.050)
Inner ring Roller Lubrication: Turbine oil VG32 circulation lubrication
¡Optimum crowning shape
¡Large diameter rollers are adopted
Fig. 1 Design of ULTAGE cylindrical roller bearing 80

ULTAGE
Temperature rise ˚C

2.3 High Load Carrying Capacity 60
Fig. 2 shows the stress distribution under the

following calculation conditions. 40
The ULTAGE cylindrical rolling bearing has

significantly reduced edge stress compared with the 20
conventional bearings by optimizing the roller crowning

0
shape. In addition, smoothly blending the straight area
4000 6000 8000 10000 12000 14000
and crowning area of the outer surface of the roller by
Rotational speed min-1
applying a new processing method, we have reduced
concentration of stress at the starting point of the Fig. 3 Temperature rise test results
crowning area seen in the conventional bearing crown
designs . This reduces the contact stress even under
tough conditions, resulting in improved load carrying 2.5 Standardized Series
performance and bearing life. The ULTAGE design will be the standard for
cylindrical roller bearings with an outer diameter of 180
<<Calculation Conditions>> mm or less.
Bearing part number: NU304E (φ20×φ52×15)
Radial load: 7.9 kN
Inclination: 1/500 (0.002 rad)
-79-
<<Calculation Conditions>>
3. ULTAGE Large Tapered Roller Bearings Bearing part number: 30316U (φ80×φ170×42.5)
The ULTAGE large tapered roller bearings have the Radial load: 102 kN
following features by optimizing the internal design of Inclination: 1/600 (0.0017 rad)
conventional bearings:
3.1 Features
High ULTAGE
(1) Load carrying capacity (compared with Conventional Bearing Design
Contact pressure
conventional tapered roller bearings)
• Basic dynamic load rating: 1.3 x
• Basic rating life: 3 x
(2) Limiting speed (compared with conventional
tapered roller bearings ) Low
• Max. 1.1 times under oil lubrication condition Small Large

(3) Reduction of running torque diameter diameter
Roller length
(4) Long-term dimensional variation
• 1/10 of bearing steel, 1/4 of case hardening steel Fig. 5 Contact stress distribution of rollers
(5) Allowable inclination (misalignment)
ULTAGE: 1/600 (0.0017 rad)
Conventional tapered roller bearings: 1/2000 (0.0005 rad) 3.4 High Speed Operation
( increase of allowable rotational speed)
3.2 Structure The ULTAGE large tapered roller bearings have
Fig. 4 shows the structure of this bearing. improved the allowable rotational speed by 1.1 times
compared with the conventional units by reducing the
running torque and temperature rise by optimizing the
Outer ring sliding contact area of the rollers and inner rings as
¡Bearing steel with special Cage shown in Fig. 6.
thermal treatment is adopted
Roller
2
1
1 Optimum contact point

for the roller
Inner ring Roller Rib
2 Enhanced precision of rib
¡Bearing steel with special ¡Optimum crowning shape surface finish
thermal treatment is adopted
¡Optimum rib design Fig. 6 Contact point between inner ring large end rib
face and roller large end face
Fig. 4 Design of ULTAGE taper roller bearing
3.3 Higher Load Carrying Capacity (1) Temperature Rise Test

The ULTAGE large tapered roller bearings have Fig. 7 shows the temperature rise test results of the
improved load carrying capacity compared with the ULTAGE large tapered roller bearing compared with an
conventional bearings by optimizing the roller NTN’s conventional bearing and another
crowning. This optimized crowning reduces excessive manufacturer’s bearing .
contact stress and provides unform contact stress
distribution. This reduction of contact stress results in <<Test Condition>>
improved bearing life even under tough conditions.. Bearing part number : 32940XUE1 (φ200×φ280×51)
Fig. 5 shows the roller contact stress distribution Lubrication: Turbine oil VG56 circulation lubrication
under the following calculation conditions. Bearing internal clearance: C2 (0.100 – 0.260)
-80-
70
Dimensional change ratio

ULTAGE
60
Another manufacturer's
50 Bearing Design Conventional
(Bearing steel)
40
Case
30 hardening steel
20
ULTAGE
10
0 0 500 1000 1500 2000 2500

0 500 1000 1500 2000 (Holding temperature:120˚C)
Rotational speed min-1 Elapsed time h
Fig. 7 Temperature rise test results Fig. 9 Dimensional change test results
(2) Running Torque Test

Fig. 8 shows the running torque test results of the 4. ULTAGE Thrust Spherical Roller Bearings
ULTAGE large tapered roller bearing compared with an The ULTAGE thrust spherical roller bearings have
NTN’s conventional bearing and another the following features by optimizing the internal design
manufacturer’s unit. of conventional bearings:
<<Test Condition>> 4.1 Features

Bearing under test: 30316U (φ80×φ160×42.5) (1) Load carrying capacity (compared with the
Axial load: 6.9 kN conventional units)
Lubrication: Turbine oil VG32 • Basic dynamic load rating: 1.7 x (max.)
Temperature: Room temperature • Basic rating life: 6 x (max.)
(2) Limiting speed (compared with the conventional
units)
1200 • Allowable rotational speed: 1.2 x (max.)
ULTAGE (3) Window type punched steel cage is adopted
1000 Conventional Bearing Design
Torque N・cm
Another manufacturer's Bearing Design

• Applicable for oil and grease lubrication
800
(4) Heat resistant up to 200°C
600 • Adoption of special heat treatment
400
4.2 Structure
200
Fig. 10 shows the structure of this bearing.
0
0 100 200 300 400 500 600 700
Rotational speed min-1
Fig. 8 Running torque test results Cage Shaft rolling bearing washer
¡Window type punched ¡Optimized groove curvature
steel cage ¡Optimized rib curvature
3.5 Reduction of Long-Term Dimensional Variation

The operating long-term dimensional variation was
reduced to 1/10 or less compared with the conventional
bearings using the bearing steel (1/4 or less compared
with case hardening steel) by adopting special thermal
treated (heat resistant dimensional stability treatment)
bearing steel. This result is shown in Fig. 9.
Housing rolling bearing washer Roller
¡Optimized groove curvature ¡Large diameter rollers
3.6 Standardized Series are adopted
The ULTAGE design will be the standard for metric ¡Maximum number of
rollers
series bearings with an outer diameter of 270 mm or
more. Fig. 10 Design of ULTAGE thrust spherical roller bearing
-81-
1) Optimized internal design 50

The diameter and length of the rollers are maximized ULTAGE
and the curvature of the groove and rib is optimized. 40
2) Adoption of the window type punched steel cage 30

By adopting a punched steel cage, lubricant can
easily flow to the inside of the bearings enabling use of 20
either oil or grease as lubricant.
10
4.3 High Load Carrying Capacity 500 1000 1500 2000 2500
The ULTAGE thrust spherical roller bearings Rotational speed min-1

achieved the world's-highest level of load carrying Fig. 11 Temperature rise test results, vertical shaft
capacity by maximizing the diameter and length of the
rollers of conventional bearing . 50
ULTAGE
4.4 High speed operation 40
( increase of allowable rotational speed)
The ULTAGE thrust spherical roller bearings adopts 30
a window type punched steel cage as opposed to the
supported-ring-guided machined steel cage of the 20
conventional bearing . In addition, optimization of the rib
curvature of the shaft rolling bearing washer reduces 10
the rise of temperature which improved allowable 500 1000 1500 2000 2500
rotational speed by up to 1.2 times compared with the Rotational speed min-1
conventional bearing . Fig. 12 Temperature rise test results, horizontal shaft
Fig. 11 and 12 show the test result of temperature
rise of vertical and horizontal shaft tests. In both tests,
the rise of temperature was lower than the conventional
bearing . 5. Conclusion
Three new types of roller bearings are now added to
(1) Vertical Shaft Test the NTN next-generation bearings, the ULTAGE Series.
<<Test Condition>> As technology innovation of industrial machines for
Bearing part number: 29418E
global environmental protection is currently underway,
Axial load: 98 kN (5% Coa)
we believe that bearings featuring high load carrying
Lubrication: Turbine oil VG32 circulation lubrication
capacity and high speed will meet the market
requirements of industrial machines by being
(2) Horizontal shaft test lightweight, compact and having a long-life.
<<Test Condition>> We are poised to continue with the development of
Bearings under test: 29418E next-generation bearings with further improvement and
Axial load: 98 kN (5% Coa) higher functionality that respond to various market
Lubrication: Turbine oil VG32 circulation lubrication demands.
4.5 Standardized Series

The ULTAGE design will be the standard for the
Photo of authors
29400Series and bearings with an outer diameter of
130 - 320 mm.
Shogo TABATA Kouhei TODA Yasuhiko SHIMIZU

Industrial Machinery New Energy Construction Machinery and
Engineering Dept., Engineering Dept., Railway Engineering Dept.,
-82-
Technical Trend of Aircraft Bearings
Takashi NISHIKAWA*
Nao HAYASHI*
Akiko HAYAKAWA**
There are many different types of bearing for aircraft applications.

NTN supplies bearings to jet engine manufacturers all over the world.
Recently due to new customer requirements for reducing weight ,
environmental impact and higher performance, bearing systems are
becoming very complicated that require higher reliability and quality
controls.
In this paper, NTN’s design theory to jet engine main shaft bearings along with the industry trends
are documented.
processes occur in clean room environments, and

1. Introduction customer-authorized quality assurance representatives
Demand and sales of commercial aircraft are growing make final product release decisions.
steadily, together with rapid growth of emerging In addition, NTN’s development includes hybrid
markets, as well as the global expansion of low cost bearings with rolling elements made of ceramic and
carriers (LCC). Aircraft employ many bearings, and raceway rings made of bearing steel, for which demand
NTN also provides bearings to the key jet engine is expected to grow, nitrided bearings in which raceway
manufacturers. In this paper, we discuss NTN's rings and rolling elements are nitrided, as well as its
activities on the bearings for aircrafts, and bearings for own test rig which evaluates dry-run performance to
main shafts of jet engines which are used under determine if the bearings can be used without seizure,
particularly tough conditions. even if the supply of lubricant is interrupted for a certain
period.
NTN has technology development sites, in addition to
2. NTN’s activities on bearings for aircraft production sites, in Japan and France. Centered on
Development and manufacturing of aerospace these sites, NTN has been directly dealing with the key
products requires considerably high technology. NTN jet engine manufacturers, helicopter manufacturers and
has established bearing plants dedicated to aerospace aerospace businesses not only in Japan and France
applications at NTN Kuwana Works (Mie Prefecture) but around the world, providing a number of high quality
and NTN-SNR Argonay Works (France) to guarantee products for many years. Recently, NTN’s bearings are
high precision and high reliability with utmost quality adopted by General Electric’s GENX engine which is
control. To support this, these plants utilize customer- used in Boeing 787, Rolls Royce’s Trent XWB engine
approved special processes and certified inspectors expected to be used by Airbus A350, Pratt & Whitney’s
accredited by the international certification program GTF engine expected to be used by Mitsubishi MRJ,
NADCAP (National Aerospace and Defense etc., which are raising NTN’s presence in the aerospace
Contractors Accreditation Program). Special process industry.
operations for the aerospace industry and certification NTN’s bearings are used in various parts of aircraft in
programs of various customers may include heat addition to main shafts of jet engines, such as
treatment, surface plating & treatments, non-destructive accessory gear boxes (AGB), flight control systems
testing, etc. In addition, final assembly and inspection (FCS), landing gears and auxiliary power units (APU).
**Machining Tool/Aerospace Engineering Dept., Industrial Machinery Division

**Kuwana Works, Industrial Machinery Division
-83-
3. Introduction of bearings for main Bearings for jet engine
Inner diameter of bearings mm

Ball bearings for gas turbine engine
shaft of jet engines 200
dn 2.8×106 Cylindrical roller bearings for gas turbine engine
Ball bearings for industrial machines
dn 2.4×106 Cylindrical roller bearings for industrial machines
3.1 Types of jet engines
150 Tapered roller bearings for rail car shafts
There are four types of aircraft turbine engines:
turbojet, turbofan, turboprop and turboshaft, and the
100
former two are usually called jet engines. The most dn 2.0×106
widely used engines are two-shaft turbofan engines dn 1.6×106
50
featuring high efficiency and low noise; however, three-
shaft engines which enable optimization of compressor
design are also used. There are also GTF engines 0
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000
equipped with a planetary gear mechanism for driving Rotational speed min -1
fans with reduced speed.

Fig. 1 shows the basic structure of a turbofan engine. Fig. 2 dn value of bearings
In this configuration, the air absorbed from the front side
of the engine is compressed by the compressor, which
is sent to the combustion chamber where fuel is burned (2) High temperature
and the exhaust drives the turbine. There are two The bearings for jet engine main shafts are exposed
driving systems for the turbine, compressor drive and to temperature of approx. 200°C during operation. In
fan drive, and most of the thrust of a turbofan engine is addition, the turbine remains in the so-called heat soak
obtained from the exhaust of the fan. back state where heat is stored with no escape even
after the engine stops, which raises the temperature of
the bearings to over 300°C. Therefore, low dimensional
Fan Additional turbine
variation and high fatigue life under high temperature
Fan exhaust jet
are required.
Air inlet Turbine

exhaust jet (3) Reliability and traceability
Since aircraft safety is an absolute requirement, the
bearings are also required to have high reliability. We
maintain high reliability by conducting full inspections in
Compressor Combustion
chamber Turbine every step of bearing manufacturing process, such as
Fig. 1 Outline structure of Turbo fan engine procurement of material, forging, turning, heat
treatment, machining and surface treatment, to verify
manufacturing conformity including various certification
3.2 Characteristics required for bearings of jet contents mentioned in the previous section.
engine main shaft In the event of any possible trouble such as damaged
Key characteristics required for the bearings of jet bearings, the ability to trace all manufacturing data of
engine main shaft are: every process is also required.
• Suitable for high-speed rotation
• Operable in high temperature 3.3 Main features of bearings of jet engine main shafts
• Highly reliable and traceable
(1) Bearing types
(1) High speed Ordinary jet engines have two shafts, one for the
An index to evaluate high speed property of bearings compressor/turbine on the high speed/high pressure
includes dn value (bearing inner ring inner diameter side, and the other one for the fan/compressor/turbine
(mm) x inner ring rotational speed (min-1)). Fig. 2 shows on the low speed/low pressure side. The bearings have
dn value of the bearings of the typical use in general a role to support them.
industrial machines. The bearings of jet engine main The bearing types for jet engine main shafts are
shafts are used under the condition that exceeds dn mainly three-point contact ball bearings and cylindrical
2.0×106, therefore, they are required to have superior roller bearings as shown in Fig. 3. The three-point
fatigue life and anti-galling properties. In addition, contact ball bearings support load in the axial and radial
significant hoop stress is placed on the bearing inner directions.
ring due to centrifugal force; therefore, the fracture They can support both directions of axial loads. The
strength to withstand the force is also required. inner rings of these bearings can be split to incorporate
-84-
as many balls as possible to ensure long life under a 70

large load. The cylindrical roller bearings are used as 65
Hardness HRC
floating-side bearings in order to absorb longitudinal 60
55 M50
expansion/contraction of the engine’s main shaft while
50
supporting radial load.
45
Fig. 4 shows the bearings for main shafts used on SUJ2
40
Boeing’s passenger aircraft B767. 35
30
100 150 200 250 300 350 400 450
Temperature ˚C
Outer
ring Fig. 5 Hardness of bearing steel for high temperature
Cage Roller
Ball
350
Inner
ring 300
(a) Three-point contact (b) Cylindrical roller
Hoop Stress MPa

250
ball bearing bearing
Fig. 3 Bearing types for jet engines 200
150
100
50
0
0.0 1.0 2.0 3.0 4.0
dn x106
Fig. 6 Hoop stress on bearing inner ring
(a) Three-point contact ball bearing (b) Cylindrical roller bearing
Fig. 4 Examples of main shaft bearings of jet engine
(2) Bearing material M50NiL
The material of bearings for jet engine main shafts

requires long life, dimensional stability, high strength
M50
and high wear resistance, particularly under high
temperature. For raceway rings and rolling elements,
0 10 20 30 40 50 60
high speed steel such as molybdenum-based AISI M50
Fracture Toughness MPa・m1/2
or tungsten-based T1 is used. As shown in Fig. 5, M50
has hardness of around 58HRC at 300°C and high Fig. 7 Comparison of fracture toughness
dimensional stability.
When bearings are used under high speed rotation,
centrifugal force is applied to the inner rings and the (3) Bearing design
tensile stress on the circumferential direction of the In order to absorb vibration produced at the main
inner rings (hoop stress) will increase. In case of shaft in high speed rotation, the outer ring is unitized
through-hardening steel such as M50, once exfoliation with the vibration absorbing components with springs
starts at the inner rings, cracking may progress very which are supported by the oil damper, as shown in
fast resulting in fracture of the inner rings; therefore, Fig. 8. Currently, the vibration absorbing components
heat-resistant carburized steel AISI M50NiL may be are most likely unitized with the outer ring in order to
used. Fig. 7 shows the comparison of fracture reduce the weight and number of components.
toughness. Fig. 9 shows an example of the commercially
Cages are frequently used with SAE4340 steel available unitized bearings.
hardened to around 30HRC. The cylindrical roller bearings have low load carrying
Usually, the bearings of aerospace applications must requirements because they only support the main shaft
only use the material certified by the aircraft weight as the radial load and the force that the inner
manufacturers or engine manufacturers. ring drives to the rollers is also relatively small. Sliding
NTN also procures the material from suppliers between the rollers and the raceway may cause early
certified by the engine manufacturers. wear. Therefore, in some cases, the outer diameter is
-85-
made into an oval shape so that preload is given to the most widely used as the lubricant for jet engines. In
rollers when they are assembled into the housing. addition, lubricant oil with superior oxidation stability
under high temperature and sludge resistance is also
developed and standardized as MILPRF-23699 HTS.
Outer ring ＋ Vibration absorbing component Along with this, the conventional oil MIL-PRF-23699
Oil ring Spring was standardized as MIL-PRF-23699 STD.
The lubricant must be delivered to the rolling contact
surface of the bearings and the contact area between
the cage guiding surface and land. Therefore, forced
lubrication is required, such as oil jet lubrication where
Oil dumper the high speed lubricant is injected from the nozzle
Conventional Unitized (shown in Fig. 11).
configuration configuration
Along with the high speed rotation of the main shaft,
Fig. 8 Unitized bearings
a “wall” of air, or so-called “air curtain”, is formed due to
the rotation of the rolling elements and the cage, which
prevents sufficient delivery of lubricant oil injected from
the nozzles. In that case, under-raceway lubrication is
used to deliver oil to the required area with the
centrifugal force. Fig. 12 shows an example of under-
raceway lubrication.
The oil jet injected from the nozzle is captured by the
oil scoop and guided to inside the main shaft. Then, the
Fig. 9 Unitized model of a bearing oil is supplied into the bearing with the centrifugal force.
Since the under-raceway lubrication delivers the oil
directly into the inner ring of the bearings, the cooling
Outer surface: Ellipse effect on the bearings is higher than with oil jet
Raceway surface: lubrication.
True circle
Lubricant Outer Lubricant

inlet ring inlet
(a) Discrete state

Outer surface:True circle Land or
Oil jet shoulder
Raceway surface:
Ellipse Ball
Cage Divided
Preload Preload inner rings
Hollow shaft
(b) Assembled state

Fig. 11 Oil jet lubrication
Fig. 10 Non-true circle cylindrical roller bearing
Lubricant
outlet
(4) Lubrication
Oil scoop
Mineral oil based SAE1010 was used as lubricant of
early jet engines, however, its fast oxidative
degradation due to the use under high temperature Oil jet
affected the bearing life.

The first solution to this challenge was the
development of synthesized ester base MIL-L-7808. Oil cooling and hole for lubricant
Then, MIL-L-23699 was developed with improved load Fig. 12 Under-raceway lubrication
carrying performance and currently, MIL-PRF-23699 is
-86-
(5) Surface treatment

On the cage guide surface, sliding friction between 4. Conclusion
the cage and the land cannot be prevented; therefore, Bearings are used in various locations of aircraft,
good friction properties and wear resistance are including jet engines. NTN has been providing a
required. In order to obtain good friction properties, number of high quality bearings for many years, and
secure delivery of lubricant oil to the cage guide surface guaranteeing high precision and high reliability through
is required and the above mentioned under-raceway comprehensive quality control.
lubrication is used. In addition, the cage surface is silver Bearings for jet engines are expected to face
plated to prevent contact between the similar metals. increasingly severe operating conditions considering
The main standard for this silver plating is AMS2414. the ongoing development of high performance jet
The land surface on the rolling bearing rings shown engines. NTN will continue to advance research and
in Fig. 13 may be TiN coated, which is a type of hard development of new technologies recognizing the
ceramic, in order to improve wear resistance. importance of bearings in jet engines. At the same time,
we are poised to contribute to the aerospace industry
by providing our high technology and quality to
customers around the world through our participation in
new development projects with jet engine and
helicopter manufacturers, among others.
References
1) Masuji Matsuoka: Koku Kagaku Koza Vol. 11 Jet
Fig. 13 TiN coating Engine (Structure)
2) NTN TECHNICAL REVIEW, No.67 (1998) 7．
3) NTN Bearing Engineer, Vol. 21, Issue 1 (1972) 61.
3.4 Technology development for bearings of jet 4) NTN Naoki Matsumori, Bearings for Jet Engine Main
engine main shafts Shafts to Protect Safety of Aircrafts, Tool Engineer, Vol.
Recent jet engines are designed to operate at higher 54 No 13, Taiga Publishing Co., Ltd. (2013) 61.
temperature and pressure for lower fuel consumption
and lighter weight, and the structure of the bearings for
jet engine main shafts is becoming more complicated.
Therefore, reliability of bearings must be improved. This
important development task is being pursued on
several fronts:
1) Surface modification
Common causes of bearing degradation are exfoliation
and adhesion, which originate from surface defects such
as indentations produced by hard foreign objects.
Raceway hardening processes via surface modification
are being developed to reduce such degradation and
achieve high reliability.
2) Rolling element material Photo of authors

We are considering use of ceramics (Si3N4) which is
light with the specific gravity of approx. 40% of M50 and
exhibits good friction and wear properties for the rolling
elements in order to reduce the contact pressure between
the rolling elements and the raceway rings due to high
speed rotation. However, since high reliability is required
for its application in bearings for jet engine main shafts,
Takashi NISHIKAWA Nao HAYASHI Akiko HAYAKAWA
we are working on establishment of a high precision non-
Machining Tool/Aerospace Machining Tool/Aerospace Kuwana Works,
destructive testing method for detecting defects. Engineering Dept., Engineering Dept., Industrial Machinery Division
Industrial Machinery Division Industrial Machinery Division
-87-
[ Technical Paper ]
The Influence of the Humidity and Quantity of Hydrogen

into Steel under the Abrasive Wear
Motohiro ITO*
An early bearing damage is occurred and accompanied by unique

white microstructures in the raceway subsurfaces. The main cause is
thought to be the result of hydrogen embrittlement caused by generated
hydrogen from decomposition of lubricant and containing water. It is
considered that hydrogen embrittlement is occurred by penetrated
diffusible hydrogen. There is a method to detect diffusible hydrogen in
steel. This method measures the specimen that is worn under abrasive wear by thermal desorption
analysis. In this report, we tested abrasive wear in space that is controlled humidity and investigated
that relationship between humidity and penetrated diffusible hydrogen into steel via abrasive wear.
1. Introduction 2. Wear test

Early flaking together with distinctive white structural Fig. 1 shows an overview of the test equipment and
change observed in roller bearings is considered to be Table 1 shows the wear test conditions. In this test,
mainly caused by hydrogen 1). The mechanism of this emery paper #220 was set on the rotary surface table
phenomenon is assumed that the hydrogen produced and three SUJ2-material disks were set on the
by hydrolysis of the lubricant or water inside the specimen holder. The chamber is designed to provide
bearings penetrates the steel causing the steel to air circulation with which the temperature and humidity
become brittle and separate 2)- 4). In addition, the are controlled by an air conditioner. The load on the
hydrogen which causes the embrittlement is considered specimen is provided by a weight with a target of 85.4
to be diffusible hydrogen. kPa of contact pressure. The wear test was conducted
In order to detect the diffusible hydrogen, two applying 5 ml of test oil on the emery paper before the
methods can be used. One method is to have test and then dropping 1 ml/min of test oil during the
hydrogen electrochemically penetrate into steel and test. Both the rotary surface table and the specimen
capture the permeated hydrogen as the ionized holder rotate with the rotational speed of the rotary
current 5). The second method is to have hydrogen surface table, at the center of the specimen holder,
penetrate into steel under a corrosive environment and being 0.59 m/s and the rotational speed of the
apply thermal desorption analysis (TDA) 6). specimen holder, at the center of the specimen, being
Diffusible hydrogen derived from water or lubricant 0.03 m/s.
hydrolysis can be detected by thermal desorption Fig. 2 shows the shape of the specimens and Table
analysis after abrasive wear of steel by emery paper 2 shows the processing conditions of the specimens.
while water or oil is dropped on the steel 7). However, The specimen is SUJ2 processed with ordinary
diffusible hydrogen is also detected in the dry abrasion quenching at 850˚C and tempering at 180˚C x 2 h.
test where no water or oil is applied. Therefore vapor in Table 3 shows various wear test conditions using
the atmosphere may also contribute 7). mineral oil, water-added oil and fluorine-based oil.
In this paper, we have investigated the amount of For mineral oil, additive-free turbine oil VG100 was
hydrogen that penetrated into steel and the impact of used.
temperature on the amount through abrasive wear The temperature in the test chamber was set to 20˚C
under the environment with controlled temperature and and 30˚C and several levels of relative humidity values
humidity. were used ranging of 25 to 95%RH. The wear test
-88-
Specimen holder Intake and exhaust of duration was 40 minutes.

temperature and humidity In order to verify the impact of water mixed in oil, a
controlled air
Weight wear test was also conducted using water-added oil.
For water-added oil, additive-free turbine oil VG100 with
Gears pure water was used. The amount of added water was
2, 5, 10 and 15 wt%, with a set temperature of 30˚C and
two relative humidity conditions, 25%RH (low humidity)
and 95%RH (high humidity).
Diffusible hydrogen in steel after the wear test is
Test oil considered to be derived not only from water but also
from the decomposition of the mineral oil. In order to
Specimen Emery paper
verify the impact of hydrogen from hydrolysis, a wear
Chamber test was conducted with fluorine based oil which does
Fig. 1 Schematic of test rig not contain hydrogen in the chemical structural formula.
For fluorine-based oil, additive-free perfluoro-polyether
oil with kinematic viscosity at 40˚C adjusted to 100
Table 1 Wear test common condition mm2/s was used. A set temperature of 30˚C and
Number of specimens Three relative humidity of 25, 60 and 95%RH were used in the
No. of emery paper #220 test chamber. The wear test duration was set at 20, 30
Table rotation speed at the and 40 minutes.
center of specimen holder 0.59 m/s
Holder rotation speed at the 0.03 m/s

center of specimen 3. Thermal desorption analysis
Contact pressure 85.4 kPa
Drop rate
After the wear tests, 3 specimens were cleaned and
1 ml/min
added to the analyzer tube for thermal desorption
analysis. The rate of temperature rise was set to
180˚C/h and the final test temperature was set to
5
t1 540˚C. A thermal conductivity detector (TCD) was used
as the hydrogen detector. Fig. 3 shows an example of
R
6
TDA profiles. The diffusible hydrogen was defined as

the total amount of hydrogen emitted up to 200˚C.
Surface area：104mm2
Fig. 2 Specimen shape (unit: mm)
0.03
No test
wt-ppm
After test (30℃ 25%RH)

min
Table 2 Disposal condition of specimen

Material SUJ2
0.02
Emitted hydrogen ratio
Thermal
processing Quenching temperature: 850˚C, tempering: 180˚C×2 h Diffusible
hydrogen
0.01
Table 3 Each wear test conditions

Item Mineral oil Water added oil Fluorine-based oil
Test oil Additive-free turbine oil VG100 Perfluoro-polyether oil 0
0 100 200 300 400 500
Amount of 2, 5, 10, 15 wt%
added water 0 wt% (Pure water) 0 wt% Temperature ℃
Set temperature 20, 30 ˚C 30 ˚C
Set relative Fig. 3 An example of TDA profiles
humidity 25∼95 %RH 25, 95 %RH 25, 60, 95 %RH
Abrasive 40 min 20, 30, 40 min
wear time
-89-
0.20
4. Mineral oil
Amount of diffusible hydrogen wt-ppm

Set temperature 30℃
Fig. 4 shows the TDA profile of the specimen before Set temperature 20℃
and after the abrasive wear tests. No diffusible 0.15
hydrogen emission was detected in the “no test” case.
The emitted diffusible hydrogen increased as the
relative humidity increased at both 20˚C and 30˚C. 0.10
Fig. 5 shows the relation between the relative
humidity and amount of diffusible hydrogen. At both set
temperatures of 20˚C and 30˚C, a high correlation was 0.05
observed.
Fig. 6 shows the relation between the absolute
humidity and amount of diffusible hydrogen. 0
0 5 10 15 20 25 30
The absolute humidity was obtained based on the Absolute humidity g/m3
measured temperature and relative humidity.
There was a strong correlation between the absolute Fig. 6 Relationship between absolute humidity and
amount of diffusible hydrogen
humidity and amount of diffusible hydrogen with little
impact of temperature.
The results show that the amount of diffusible hydrogen
in steel is dependent on the amount of vapor in the unit
5. Water added oil
volume of air.
Fig. 7 shows the relation between the amount of
0.012 added water and amount of diffusible hydrogen. Both in
Set temperature 20℃ Set temperature 30℃
the low humidity and high humidity cases, as the
wt-ppm
0.010 94%RH 93%RH

min
72%RH 63%RH amount of water increased, the amount of diffusible

44%RH 40%RH
0.008 26%RH
hydrogen increased. The amount of diffusible hydrogen
No test
No test in the low humidity test, excluding the 15 wt% water

0.006 added oil, was less than the amount of diffusible
hydrogen in the high humidity test, excluding the water-
0.004
free oil.
0.002 When water added oil was used, the amount of
diffusible hydrogen increased compared to the water-
0 free oil in both high and low humidity. This indicates that
0 100 200 0 100 200
the water added to the grease as well as the small
Temperature ℃ amount of water vapor in the air both contributed to as
Fig. 4 TDA profiles of specimens before and after diffusible hydrogen sources to the new steel surface
abrasive wear testing produced during abrasion.
0.20 0.25

0.20
0.15

0.15
0.10
0.10
0.05
0.05
High humidity (set relative humidity: 95%RH)
Low humidity (set relative humidity: 25%RH)
0 0
0 20 40 60 80 100 0 5 10 15
Relative humidity %RH Amount of added water wt%
Fig. 5 Relationship between relative humidity and Fig. 7 Relationship between amount of containing water and
amount of diffusible hydrogen amount of diffusible hydrogen in cases of water containing oils
-90
However, the amount of diffusible hydrogen in the 5.1 Relationship between humidity and water
low humidity test, excluding the 15 wt% water added oil contained in oil
(0.153wt-ppm), was less than the amount of diffusible It was shown that the amount of diffusible hydrogen
hydrogen in the high humidity test, excluding water-free in steel after the abrasive wear test depends on the
oil (0.176wt-ppm). This shows that the impact on amount of vapor in the air. It was also shown that the
diffusible hydrogen into the steel after abrasion was impact of humidity is larger than the impact of added
larger for water vapor in the air than that of water added water to the lubricant. Therefore, we investigated the
directly to the test oil. water content of the oil when exposed to different
Table 4 shows the amount of water involved in the humidity conditions in order to verify the impact of
abrasive wear test and the amount of diffusible humidity on water content of the oil.
hydrogen. Since the volume of the test chamber is Table 5 shows the test conditions for measuring the
0.0104 m3, the amount of vapor contained in the test water content. The same additive-free turbine oil VG100
chamber air is 0.080g in low humidity and 0.271g in was used. A set temperature of 30˚C and relative
high humidity. When 15 wt% of water was added to the humidity of 30, 60 and 90%RH were used for the test
test oil, with a drop rate of 1 g/min and an abrasion test chamber. The container with test oil was exposed in the
duration of 40 minutes, the amount of water in the test chamber for 30 minutes, then the water content
dropped oil was 6g. The total amount of water when 15 was measured.
wt% water added oil was used in low humidity was Fig. 8 shows the measurement results of the water
6.080g (6 + 0.080g). content in the oil. As the absolute humidity increased,
On the other hand, when water-free oil was used in the water content in the oil also increased.
high humidity, the amount of water was 0.271 g. This This means that in high humidity the amount of water
shows that while the amount of water involved in the contained in oil increased.
abrasive wear test was larger in the case of 15 wt% It is considered that as humidity increases, water
water added oil in low humidity versus that of water-free contained in the oil increases and becomes the source
oil in high humidity, the amount of diffusible hydrogen of hydrogen. Therefore, diffusible hydrogen in steel
was smaller. While the reason why the amount of depends on the amount of water vapor in the air.
diffusible hydrogen varies depending on the state of
water is unclear, the test result showed humidity had a
larger impact on the amount of diffusible hydrogen in Table 5 Water content of oil test condition
steel after the abrasion test.
Test oil Additive-free turbine oil VG100
Set temperature 30˚C
Set relative humidity 30, 60, 90 %RH
Table 4 Amount of water and amount of diffusible Exposure time 30 min
hydrogenin wear test
Low humidity High humidity
(25%RH) (95%RH)
Item + +
added water no added water 0.005
(15 wt%) (0 wt%)
Volume of chamber space 約 0.0104 m3
Amount of water content wt%
0.004
Absolute humidity 7.6 g/m3 25.9 g/m3
Amount of water in the chamber 0.080 g 0.271 g
Drop rate 1 g/min 0.003
Test time 40 min

Amount of added water 15 wt% 0 wt% 0.002
Amount of water in oil 6g 0g
Total amount of water 6.080 g 0.271 g
0.001
Diffusible hydrogen 0.153 wt-ppm 0.176 wt-ppm
0
0 5 10 15 20 25 30
Absolute humidity g/m3
Fig. 8 Water content of oil
-91-
but also from decomposition of mineral oil. Therefore,

6. Fluorine-based oil we conducted an abrasive wear test using fluorine-
Fig. 9 shows the TDA profile of the specimen before based oil to obtain correlation between the increase of
and after the wear tests. After the abrasive wear test, humidity and increase of diffusible hydrogen. Since
diffusible hydrogen was detected from every specimen fluorine-based oil does not contain hydrogen in the
and the amount of diffusible hydrogen increased as the chemical structural formula, diffusible hydrogen
relative humidity increased (in their respective abrasion detected in the abrasive wear test using fluorine-based
duration). As the abrasion time increased, diffusible oil derives only from water.
hydrogen increased. In addition, since the abrasive wear tests indicate a
Fig. 10 shows the relationship between the relative linear increase of diffusible hydrogen within the test
humidity and amount of diffusible hydrogen with range, it is assumed that diffusible hydrogen is not
fluorine-based oil. There is a high correlation observed saturated in the steel but is in a transient state.
for all abrasion times. Therefore, it is possible that the impact of humidity will
Fig. 11 shows the relationship between the abrasion be more significant when abrasion takes longer time
time and amount of diffusible hydrogen. than these tests.
As the abrasion time increased, diffusible hydrogen The impact of humidity on the amount of hydrogen
increased for a given relative humidity. penetrated in steel has not been reviewed until now.
In the abrasive wear test using mineral oil, diffusible However, this result indicates that the impact of
hydrogen was detected from steel and a high humidity is a significant factor.
correlation was observed between the absolute
humidity and diffusible hydrogen. From this, it was 0.15
shown that the amount of diffusible hydrogen in steel Abrasive wear time 40min
Abrasive wear time 30min

after an abrasive wear test depends on the amount of
Abrasive wear time 20min
vapor contained in the unit volume of air. However, in
case of mineral oil, the source of diffusible hydrogen 0.10
could derive not only from the vapor contained in the air
0.012
Abrasive wear time 20min Abrasive wear time 30min
0.05
wt-ppm
91 %RH 88 %RH
min
0.010
63 %RH 63 %RH
26 %RH 22 %RH
0.008 No test No test
0.006 0
0 20 40 60 80 100
Relative humidity %RH
0.004
Fig. 10 Relationship between relative humidity and amount
0.002 of diffusible hydrogen in cases of fluorine based oil
0
0 100 200 0 100 200
0.15
Set temperature 20℃ Set temperature 20℃ High humidity
0.012 (set relative humidity: 95%RH)

Abrasive wear time 40min Middle humidity
(set relative humidity: 60%RH)
wt-ppm
93 %RH
min
0.010 Low humidity

63 %RH (set relative humidity: 25%RH)
24 %RH 0.10
0.008 No test
0.006
0.05
0.004
0.002
0 0
0 100 200
0 10 20 30 40 50
Abrasive wear time min
Fig. 9 TDA profiles of specimens before and after abrasive
wear testing in cases of fluorine based oil for wear duration Fig. 11 Relationship between wear duration and amount
of 20, 30, 40 min of diffusible hydrogen in cases of fluorine based oil
-92-
References
7. Conclusion 1) K. Tamada, and H. Tanaka, "Occurrence of brittle
We have investigated the impact of humidity on the flaking on bearings used for automotive electrical
amount of hydrogen penetrated in steel by conducting instruments and auxiliary devices" Wear, 199 (1996),
pp．245-252．
an abrasive wear test in a chamber under controlled
2) M. Kohara, T. Kawamura and M. Egami，"Study on
temperature and humidity and applying thermal
Mechanism of Hydrogen Generation from Lubricants"
desorption analysis. As a result, the following were Tribology Transactions, 49 (2006), pp. 53-60．
revealed: 3) L. Grunberg，"The Formation of Hydrogen Peroxide on
1) When the abrasive wear test was conducted using Fresh Metal Surfaces" Proc．Phys. Soc (London). B66
mineral oil, the amount of diffusible hydrogen (1953), pp. 153-161．
increased as the relative humidity increased for a 4) P. Schatzberg and I. M. Felsen, "EFEEECTS OF
given temperature. When absolute humidity was WATER AND OXYGEN ROLLING CONTACT
used instead of relative humidity in the arrangement LUBRICATION" Wear, 12 (1968), pp. 331-342．
of data, a strong correlation was observed with the 5) Shigeru Asano, Yoshio Fujishima, Namio Ohtani, “The
amount of diffusible hydrogen. Diffusivity of Hydrogen in Iron at Room Temperature”,
2) When the abrasive wear test was conducted with Journal of The Japan Institute of Metals，Vol. 37, Issue
3 (1973), pp. 301-306.
pure water added to mineral oil, the amount of
6) M. Nagumo, M. Nakamura, and K. Takai, "Hydrogen
diffusible hydrogen increased both in low and high
Thermal Desorption Relevant to Delayed-Fracture
humidity. Susceptibility of High-Strength Steels"
Within the range of ambient temperature in this test, METALLURGICAL ANDMATERIALS TRANSACTIONS
the amount of diffusible hydrogen in the low humidity A, 32A (2001),pp. 339-347．
test, excluding the 15 wt% water added to oil, was 7) Hiroyoshi Tanaka, Hiraku Tanimoto, Joichi Sugimura,
less than the amount of diffusible hydrogen in the "Observation of Hydrogen Permeation into Fresh
high humidity test, excluding no water added to oil. Bearing Steel Surface by Thermal Desorption
3) Even when fluorine-based oil was used, which does Spectroscopy”, Proceedings of Japan Tribology
not contain hydrogen atoms in the chemical Conference Tokyo (2010-5), pp. 203-204.
structural formula, diffusible hydrogen was detected
from the steel. The amount of diffusible hydrogen
increased when the abrasive wear duration was
increased.
Photo of author
Motohiro ITO
Advanced Technology
R&D Center
-93-
[ Technical Paper ]
Evaluation of Shear Fatigue Properties in Medium Carbon Steels

and Rolling Bearing Steels up to Giga-cycle Test Regimes
Noriaki MIWA*
We have developed an ultrasonic torsional fatigue tester which enables for the
rapid evaluation of shear fatigue properties of high strength steels. This report
presents the test results of shear fatigue properties of induction hardened
medium carbon steels, and rolling bearing steels, up to giga-cycle regimes as
evaluated by the developed tester.
1. Introduction 2. Impact of boron (B) on shear fatigue

properties of induction hardened medium
Repetitive torsional load is applied to the shafts of
carbon steels up to 1x109 cycles
constant velocity universal joints, which are used in
front-wheel drive and four-wheel drive vehicles and 2.1 Background
industrial machines. Therefore, it is important to Induction hardened medium carbon steel S40C is
investigate shear fatigue properties up to giga-cycle test used for the shafts of ordinary constant velocity joints.
regimes in order to guarantee long-term reliability of the When the shaft diameter is relatively small, the material
materials. As one of the stresses governing generation with increased manganese (base steel) is used to
of cracks before internal-originated cracking in rolling improve the quenching property and when the shaft
bearings , NTN can report completely reversed diameter is large, boron (B) is added to the base steel
alternative shear stress acted repeatedly on the or its equivalent (boron-added steel) to further increase
surface 1). Therefore, it is important to evaluate shear the quenching property. NTN has evaluated shear
fatigue properties in order to determine life of rolling fatigue properties of shaft materials up to 1x109 cycles
bearings. with an ultrasonic torsional fatigue tester and
The shear fatigue life can be obtained by a torsional investigated the impact of added boron on the fatigue
fatigue test. However, since the load frequency of the properties.
conventional torsional fatigue tester is low, evaluation
up to 1x109 cycles has been practically impossible. 2.2 Test method
Therefore, we developed an ultrasonic torsional fatigue Boron-added steels, where 19 ppm of boron and 420
tester (USTF tester) aimed at rapid evaluation of shear ppm of Ti are added to the base steel , were used as
fatigue properties of high strength steel. Since the test materials. Fig. 1 shows the induction hardening
commercially available torsion vibration converters have conditions and Fig. 2 shows the diagram of the
low power, we optimized the shapes of torsion sample. The gray area of Fig. 2 is heated to adjust the
amplitude magnifying horns and specimens and hardness of the hatched area including the smallest
succeeded to provide shear stress amplitude of around diameter area to 700 HV. Then, the specimen is
950 MPa max 2). In this paper, we discuss the following tempered at 170˚C x 1h and finished by grinding. To
shear fatigue property evaluation results using the eliminate the impact of surface roughness, emery
developed ultrasonic torsional fatigue tester: abrasion (#500 and #2000) and diamond wrapping
(1) Impact of B on shear fatigue properties of induction (grain size of 1 μm) were applied for a mirror finish
hardened medium carbon steels up to 1x109 cycles before the ultrasonic torsional fatigue test. Resonance
(2) Relationship between shear fatigue life and rolling frequency of 20000 Hz and stress ratio R = -1
fatigue life of bearing steels obtained by an (completely reversed) were used as the test conditions.
ultrasonic torsional fatigue test. When a shear fatigue crack is generated and
propagated in the sample, resonance frequency
-94-
Evaluation of Shear Fatigue Properties in Medium Carbon Steels and Rolling Bearing Steels up to Giga-cycle Test Regimes
decreases. 850
It was determined that the sample was damaged 800
B-added steel (dashed line)
when resonance frequency exceeded the preset range,
Shear stress amplitude MPa

750
and the test was terminated. Since continuous loads
would cause the specimen to heat up in the USTF test, 700
an intermittent load method was used, alternating loads 650

Base steel (solid line)
and pauses 3). In the case of shear stress amplitudes of 600
more than 700 MPa, 0.11 sec of load and 1.1 sec of Base steel (Shear-mode crack in axial direction)
550 Base steel (Shear-mode crack in circum. direction)
pause were used until damage was detected from B-added steel (Shear-mode crack in axial direction)
500 B-added steel
heat generation. (Shear-mode crack in circum. direction)
In the case of less than 700 MPa, 0.11 sec of load 450
105 106 107 108 109 1010
and 1.1 sec of pause were used 1x107 times, then the Nnmber of cycles
conditions were changed to 1 sec of load and 0.2 sec
Fig. 3 Shear fatigue properties of Base and B-added
of pause. When no damage was detected up to 1x109 steels evaluated by USTF tester
cycles , the test was terminated. The real load
frequency of 0.11 sec of load and 1.1 sec of pause is
1803 Hz and that of 1 sec of load and 0.2 sec of pause
is 16529 Hz 2).
1050
25kHz
Temperature ˚C
water-soluble
(a) Base steel (b) B-added steel
quenching
Fig. 4 SEM images of fracture surface at initial shear
fatigue cracking
5 10
Time sec. intergranular enhancement elements 4); its effect could
Fig. 1 Induction heat treatment condition
not be confirmed since the fracture in these
experimental conditions were transgranular.
3. Relationship between shear fatigue life and

rolling fatigue life of bearing steels obtained
Fig. 2 Specimen shape and hardened region by ultrasonic torsional fatigue testing.
3.1 Background
2.3 Results and observation Shear fatigue properties of carbon steel for machine
Fig. 3 shows the S-N curves of the base steel and parts S53C and heat-resistant carburized steel M50NiL
the Boron-added steel. S-N curves were obtained by (AMS6278) were evaluated. In addition, we investigated
applying the fatigue test results to the continuous drop the relationship between the shear fatigue properties
single logarithm curve model of the Standard and rolling contact fatigue (RCF) test results of these
Regression Method of S-N Curves from the Society of two steels and SUJ2 bearing steel 2) for which previous
Materials Science, Japan. Fatigue strength of the base test results exsist.
steel and Boron-added steel after 1x108cycles with
10% fracture probability was 636 MPa and 643 MPa, 3.2 Test method
respectively, almost equal. Table 1 shows the chemical contents of S53C and
Shear type initial cracks started at the smallest M50NiL, used for the USTF test samples Fig. 5 shows
diameter , and propagated and transitioned to tensile the cross section hardness distribution of the USTF
type. Shear type initial cracks are either axial or radial samples at minimum diameter. To eliminate the impact
and the latter are plotted with a horizontal line in Fig. 3. of surface roughness, emery abrasion (#500 and
Fig. 4 shows the SEM images of the fracture surface at #2000) and diamond wrapping (grain size of 1 μm)
initial shear fatigue cracking. Both materials exhibited were applied for a mirror finish before the ultrasonic
transgranular fracture. Although B is reported to be torsional fatigue test.
-95-
The diameter and length of the RCF samples were The USTF test conditions are the same as those in
12 mm and 22 mm, respectively. Section 2.2. When there was no damage up to 1x1010
Table 2 shows the number of lots, total number of cycles, the test was terminated.
samples , as well as the average, maximum and Fig. 6 shows the schematic diagram of the RCF test
minimum hardness measurements of samples equipment and Table 3 shows the test conditions. The
evaluated in the tests. samples are driven by the driving roller to cause rolling
fatigue with the 3/4 inch steel balls in contact under high
contact stress conditions. When damage occurs in the
Table 1 Chemical compositions of S53C and M50NiL sample , vibration of the tester increases, which was
used for USTF specimens (mass %) used as the termination signal.
C Si Mn P S Ni Cr Mo V
S53C 0.54 0.17 0.86 0.011 0.008 0.02 0.14 - - 3.3 Results
M50NiL 0.12 0.20 0.25 0.005 0.002 3.40 4.08 4.15 1.18 Fig. 7 shows the shear fatigue properties of three
types of steels, namely, the two steel types tested in
this paper and SUJ2 2) . S53C was the lowest both in
900
fatigue strength at finite life and fatigue limit (442 MPa)
800
among the three steel types evaluated so far. M50NiL
Vickers hardness HV
was the highest both in fatigue strength at finite life and

700 fatigue limit (680 MPa) among the three steel types.
The slopes of shear fatigue properties of the three steel
600
types were not significantly different within these fatigue
500
strength cycles.
S53C
M50NiL
Table 4 shows the average, maximum and minimum
400 values of 50% life of these three steel types in the RCF
0.0 0.5 1.0 1.5 2.0
tests previously conducted. The overall performance of
Depth from surface mm
50% life was S53C<SUJ2<M50NiL.
Fig. 5 Hardness distributions of S53C and M50NiL at
minimum circumference of USTF specimens
1200
Table 2 Number of lots, total number and hardness of RCF SUJ22)
specimens S53C
1000
M50NiL
Number Total number Hardness (HRC)
τw0=680MPa (M50NiL)
of lots of specimens Ave. Min.∼Max. 800
SUJ2 9 124 62.4 62.0∼62.8

600
S53C 2 34 61.0 60.5∼61.5
M50NiL 3 63 63.8 62.4∼64.5
400
τw0=577MPa (SUJ2)2)
200 τw0=442MPa (S53C)

Driving roller
0
105 106 107 108 109 1010 1011
Specimen
Number of cycles
3/4” ball
Fig. 7 Shear fatigue properties of three kinds of bearing
steels evaluated by USTF tester
Table 4 RCF test results

Guide roller
Number RCF life L50 (cycles)
Fig. 6 Schematic of RCF tester of lots Average Minimum∼Maximum
SUJ2 9 1.5×108 8.5×107∼2.1×108
Table 3 RCF test conditions S53C 2 3.5×107 3.3×107∼3.7×107
Ball 3/4” (19.05mm)×2 made of SUJ2 M50NiL 3 1.0×109 7.3×108∼1.2×109
Max. contact pressure 5.88 GPa
Loading rate 46240 c/min
Lubricant Turbine oil VG68
-96-
Evaluation of Shear Fatigue Properties in Medium Carbon Steels and Rolling Bearing Steels up to Giga-cycle Test Regimes
3.4 Observation (2) S53C was the lowest both in fatigue strength at finite
Fig. 8 shows shear fatigue life within finite life life and fatigue limit (442 MPa) of shear fatigue
(fracture probability 50%) at a stress amplitude of 700 properties among the three steel types evaluated.
MPa of three steel types and rolling fatigue life L50. A On the other hand, M50NiL was the highest both in
linear relationship was observed between the data sets. fatigue strength at finite life and fatigue limit (680
Therefore, it was implied that the USTF test may MPa).
become an alternative to the RCF test. That is, it could (3) The slopes of shear fatigue properties of the three
be an effective tool for determining relative superiority of steel types were not significantly different within
RCF life. these fatigue strength cycles.
(4) Linear correlation was found between shear fatigue
life within finite life (fracture probability 50%) at
109
stress amplitude 700 MPa and rolling fatigue life L50
from previous rolling fatigue testing.
amplitude of 700MPa cycles
Shear fatigue life at stress
108 This paper has been edited based on the original

manuscripts 5) and 6). We thank the Iron and Steel
Institute of Japan and Japanese Society of Tribologists,
107 who gave us the permission to publish it.
SUJ2
S53C
References
M50NiL
1) G. Lundberg and A. Palmgren: DYNAMIC CAPACITY OF
106
107 108 109 1010
ROLLING BEARING, Acta Polytechnica, Mechanical
Engineering Series, 1-3, (1947) 5-58.
RCF life L50 cycles
2) Noriaki Sakanaka, Yukio Matsubara, Yoshinobu
Fig. 8 Relationship between shear fatigue life and RCF life Shimamura, Hitoshi Ishii: Rapid Evaluation of Shear
Fatigue Properties of Rolling Bearing Steels for Lifespans
Up to Gigacycle Range, NTN TECHNICALREVIEW, No.
79, (2011) 104-110.
3) H. Ishii, K. Yamanaka and K. Tohgo: GIGA-CYCLE
4. Conclusion FATIGUE STRENGTH IN SOME HIGH STRENGTH
Since torsional loads are applied repeatedly to the STEELS BY ULTRASONIC FATIGUE TESTING, Material
Science Research International, Special Technical
shafts of constant velocity joints, it is important to
Publication-1, (2001) 59-63.
evaluate their shear fatigue properties. We evaluated 4) M. Yamaguchi: FIRST-PRINCIPLES STUDY ON THE
shear fatigue properties of the materials with increased GRAIN BOUNDARY EMBRITTLEMENT OF METALS BY
manganese on S40C (base steel) and boron (B) added SOLUTE SEGREGATION: PART1.
IRON (FE) -SOLUTE (B, C, P, AND S) SYSTEMS,
on the base steel or its equivalent material up to 1x106 -
Metallurgical and Materials Transaction A, 42, 2, (2011)
cycle regimes with an ultrasonic torsional fatigue tester. 319-329.
On the other hand, it is important to evaluate shear 5) Sakanaka, Matsubara: Impact of Boron on Shear Fatigue
fatigue properties in order to determine rolling fatigue Properties of Induction Hardened Medium Carbon Steels
life, since alternate shear stress is considered to be one for Lifespans Up to Gigacycle Range, CAMP-ISIJ, 26,
(2013)316.
of the stresses governing crack generation and its 6) Miwa: Relationship Between Shear Fatigue Life of Bearing
initial propagation causing internal-originated damage Steel obtained from torsional fatigue test and Rolling
of the rolling bearings. Therefore, we evaluated shear Fatigue Life, Tribology Conference 2013 Autumn
fatigue properties of carbon steel for machine parts Proceedings (2013-10) E6.
S53C and heat-resistant carburized steel M50NiL
(AMS6278). In addition, we investigated the relationship Photo of author
between shear fatigue properties and rolling fatigue life
test results of three steel types including two steel types
in this paper and bearing steel SUJ2 which had been
previously evaluated.
(1) Fatigue properties of base steel with increased

manganese to S40C and boron-added steel
(fracture probability 10%) after 108 cycles were
similar with 636 MPa and 643 MPa, respectively, Noriaki MIWA
exhibiting no difference in shear fatigue properties. Advanced Technology
R&D Center
-97-
NTN TECHNICAL REVIEW No. 82（2014）
Award Winning Product
2013 “CHO” MONODZUKURI Innovative Components Awards Nippon Brand Award
Parallel Link High Speed Angle Control Equipment
Hiroshi ISOBE, Yukihiro NISHIO, Seigo SAKATA, Naoya KONAGAI, Hiroyuki YAMADA
1. Introduction Parallel link high speed

angle control equipment
NTN developed Parallel Link High Speed Angle Control
Equipment (hereafter, equipment) applying the special parallel Stand
link mechanism of rotational 2-degrees-of-freedom, and applied
this equipment to grease dispensing. This equipment was Dispenser
recognized for its ability to achieve high speed and highly
accurate positioning in a broad operating range, with an Work
operating angle of 90˚ and angle of traverse of 360˚, and
X XY stage
received the 2013 “CHO” MONODZUKURI Innovative
Y
Components Awards Nippon Brand Award. Fig. 2 System configuration example of grease dispensing
2. Background complicated shapes by enabling application at varying angles.

The parallel link mechanism adopted in this equipment was The equipment is capable of setting commanded angles by
originally developed as a constant velocity joint for automobiles, specifying XY coordinates on the operating stage, to facilitate
achieving high operating angle with a maximum of 90°. Aiming teaching input tasks upon introduction into a new production
for applications other than automobiles, we continued line.
developing this parallel link mechanism as angle control We hope that the equipment will be used in various fields
equipment by combining actuators with it, and with an such as medical, service, entertainment, etc. in addition to
innovative driving mechanism and control methods, we could production facilities, and will contribute to further advancement
achieve smooth operation of rotational 2-degrees-of-freedom of automation and robotization where Japan is competitive.
plus high speed/high accuracy operations 1)-5). Recognized for
its high speed and superior operability, it is used in the grease
application mechanism in the production lines of automobile References
1) Keisuke Sone, Hiroshi Isobe, Koji Yamada: Wide Angle Active Link Equipment,
parts, where reduced tact time and high reliability are required.
NTN TECHNICAL REVIEW No. 71, (2003) 70-73.
2) Hiroshi Isobe, Yukihiro Nishio: Parallel Link High Speed Angle Control
Equipment (PHACE), NTN TECHNICAL REVIEW, No. 80, (2012) 42-47.
3. Implementation on grease application 3) Hiroshi Isobe: Parallel Link High Speed Angle Control Equipment – Basic
equipment Configuration and Application for Grease Dispensing, Machine and Tool,
February Issue, (2013) 83-87.
Fig. 1 shows the equipment developed for grease 4) Hiroshi Isobe, Yukihiro Nishio, Keisuke Sone, Hiroyuki Yamada, Yoshio
application, and Fig. 2 shows an example of grease application Fujikawa: Parallel Link High Speed Angle Control Equipment, Proceedings of
equipment configuration where a dispenser for contact-less the Japan Society for Precision Engineering Spring Meeting in 2013, (2013)
application is attached to the tip of this equipment. In this 809-810.
5) Hiroshi Isobe, Yukihiro Nishio, Seigo Sakata, Naoya Koganei, Hiroyuki
configuration, the equipment moves the work on the XY stage
Yamada, Yoshio Fujikawa: Parallel Link High Speed Angle Control
and controls the behavior of the dispenser. It achieves high Equipment—Implementation on Grease Application—, Proceedings of the
speed application of around 10 points per second and Japan Society for Precision Engineering Spring Meeting in 2014, (2014) 1087-
automates the grease application process on workpieces with 1088.
Driving Photo of authors (presenters)

mechanism
Parallel link
mechanism
Tip
Fig. 1 Parallel link high speed angle control equipment Hiroshi ISOBE Yukihiro NISHIO Seigo SAKATA
New Product Development New Product Development New Product Development
R&D Center R&D Center R&D Center
-98-
2013 Japan Powder Metallurgy Association, New Product Award in New Materials Division
Development of the Oil Seal Made of Sintered Metal

for Variable Valve Timing Device
Yosuke SUGAI Toshihiko MOURI
2.3 Improved machinability and accuracy by optimized sintering

1. Introduction conditions
Variable valve timing devices for automobiles are used in various In addition to the enhanced strength by adding low-melting point
types of vehicles for fuel efficiency, processing of exhaust gas, etc. metals, we have added a sizing process to improve dimensional
and demand is expected to increase particularly in the emerging accuracy. This is possible by the optimization of the sintering
markets. conditions. Fig. 3 shows the dimensional accuracy. By ensuring the
Hydraulic variable valve timing devices are equipped with multiple developed material made of almost only ferrite phase, we can achieve
oil seals, as shown in Fig. 1, to divide hydraulic chambers. excellent sizing properties and required dimensional accuracy.
Conventional oil seals are made of resin; however, we have developed
oil seals made of sintered alloy and introduced them into the market.
The characteristics required for oil seals are sealing capability and 400
Transverse rupture
strength. If dimensional accuracy is not sufficient, oil may leak from the 300
strength N
gaps with the surrounding components and affect the operating
200
responsiveness of variable valve timing devices. Insufficient strength
may cause deformation and damage to the devices. 100
The developed oil seals made of sintered metals received “The 35th
0
Japan Powder Metallurgy Association, New Product Award in New Segregation-free Developed material Resin
Materials Division” which is awarded to superior new technologies in processed powder sintered compact
powder metallurgy. In this article, we present an overview of this
development. Fig. 2 Comparison of transverse rupture strength
4.0
Hydraulic chamber Total length Cp
Oil seal Straightness Cp
Cp 2.0
0.0
FG64 Segregation-free Segregation-free Developed
processed powder processed powder material
Cam pulley + +
Camshaft sizing sizing
Fig. 1 Point-of-use of the oil seal Fig. 3 Dimensional precision (Cp value)
2. Features of the developed product 3. Achieved results from the development

2.1 Good dimensional stability due to segregation-free processing By using segregation-free graphite powder, the addition of tin and
Ordinary ferrous sintered material, FG64, has unstable dimensional optimization of sintering conditions, we can achieve good dimensional
accuracy. This is due to segregation of graphite contained in the accuracy and sealing abilities , B We will continue our development of
material which prevents proper flow of powder and leads to unstable oil seals made of sintered metal to improve the function of variable
compact weight during the molding process, resulting in reduced valve timing equipment.
dimensional accuracy.
Therefore, we used ferrous powder and segregation-free
processing to graphite to improve the powder flow and to stabilize the
compact weight. This resulted in good dimensional stability.
Photo of authors
2.2 Strength enhancement by mixing low-melting point metal
In order to ensure seal strength with sintered materials, a low
melting point metal, tin (Sn), was selected and mixed in the optimum
proportion. Since tin has good wettability with copper (Cu), when it is
sintered below the melting point of Cu, it is diffused to facilitate bonding
of ferrous powders, which improves strength of the sintered product .
Fig. 2 shows the comparison of transverse rupture strength as an
index of the strength of seals.
Transverse rupture strength evaluates bend strength of the material
by applying bending loads at three points on the test sample without
Yosuke SUGAI Toshihiko MOURI
cutouts. As a result, the developed material almost doubled in strength
Engineering Dept. Engineering Dept.
compared with the material made of resin. NTN Powder Metal Corporation NTN Powder Metal Corporation
-99-
2013 Japan Powder Metallurgy Association, Effort Prize
Sintered Hydrodynamic Bearing for Super Thin Fan Motor
Tadahiro UCHIYAMA Shinji KOMATSUBARA
2mm
Peak area
1. Introduction
As mobile devices rapidly expand, the cooling fan motors used in 1.8mm
these devices are becoming increasingly thin. The use of these
devices is becoming much more diverse and the operating Groove area
environment is also becoming much more challenging. Though the
4mm
operating environment is becoming more challenging, the
requirements for the bearings are to provide lower noise and more Fig. 2 Example of sintered hydrodynamic bearing
reliable operation than ever.
By leveraging high precision and high rigidity of the conventional Shaft
sintered hydrodynamic bearings, we have developed sintered 4mm
hydrodynamic bearings for super thin cooling fan motors (Fig. 1).
In this article, we present an overview of this product development, Housing Sintered hydrodynamic
Thrust washer bearings
which received “2013 Japan Powder Metallurgy Association, Effort
Prize.” Fig. 3 Example of a super thin fan motor application
2.2 Optimization of the hydrodynamic groove machining conditions

In order to improve hydrodynamic bearing performance, we worked
on higher precision molding for forming hydrodynamic grooves as a
way to improve groove precision. We have optimized the
hydrodynamic groove machining conditions for molding and also
Conventional Developed
product product improved the molding precision. We also optimized hydrodynamic
groove machining conditions for the bearings such as timing and
machining margin during the dynamic groove machining process to
Fig. 1 Appearance of sintered hydrodynamic bearing obtain precise and stable groove depth and shape and enabled high
precision machining of hydrodynamic grooves.
2.3 New development of lubricant oil

2. Description of development Due to the more severe operating conditions and property
Sintered hydrodynamic bearings have a structure shown in Fig. 2. requirement, we also worked on improving the lubricant. We developed
Sintered bearings have herringbone-like hydrodynamic grooves a lubricant oil which is appropriate for the bearings of super thin fan
formed on the inner surface of the sleeve-shaped bearing rings where motors considering reduced rigidity in high temperature and increased
fluid lubrication film in the bearing clearance produces pressure bearing loss in low temperature, etc.
(hydrodynamic pressure) by the rotation of the shaft to support the
rotary shaft. In addition, as the material is the sintered alloy, lubricant is
contained within the bearings resulting in a good anti-seizure property.
3. Conclusion
Fig. 3 shows an example of a super thin fan motor. The height of We have developed sintered hydrodynamic bearings for super thin
the motor is 4 mm and the length of the shaft must be 2 mm or less. fan motors by optimizing the design of the dynamic groove
When the length of the bearings is short, rigidity and supportive force specification, optimized dynamic groove machining conditions and use
are reduced; therefore, how to maintain bearing rigidity is the main of a newly developed lubricant oil. Its superior quietness and reliability
challenge. We have focused on the specification of hydrodynamic were recognized and the bearings were widely adopted by fan motor
grooves and their machining conditions. By optimizing the the manufacturers in Japan and abroad. These bearings are highly
hydrodynamic grooves, we have succeeded in development of sintered recognized by end users such as the PC manufacturers and expansion
hydrodynamic bearings for super thin fan motors with a shaft length of of further adoption can be expected.
2 mm or less.
Photo of authors
2.1 Optimization of the hydrodynamic groove specification
As mentioned earlier, with shorter bearing length comes decreased
bearing rigidity which may cause noise or short product life. Therefore,
we created our own analytical program for theoretical analysis based
on the rotor weight and motor structure, etc. which affect the bearing
performance. We then optimized the design parameters such as
hill/groove ratio of dynamic bearings, angle of groove and width of
groove for super thin fan motors. We also conducted a verification test
using motors to establish hydrodynamic bearing specification backed
by theory and evaluation of the actual units in the early stage and Tadahiro UCHIYAMA Shinji KOMATSUBARA
significantly reduced the development and design time. Engineering Dept. Fluid Dynamic Bearing Unit Dept.
NTN Powder Metal Corporation Composite Material Product Division
-100-
2013 Japanese Society of Tribologists Effort Prize
Development of Ultrasonic Torsional Fatigue Tester for

Rapid Shear Fatigue Property Evaluation of Rolling Bearing Steels
Noriaki MIWA
It is not possible to directly measure shear stress applied to

1. Introduction the smallest diameter area of the specimen. Therefore, first we
As one of the stresses of rolling bearings governing crack obtained the relationship between the maximum shear stress
generation before internal-originated flaking and its initial and torsional angle by FEM analysis, then the relationship
progress, we can point out almost completely reversed between the amplifier output and torsional angle through
alternating shear stress acting repeatedly on the surface. The experimenting, and finally, identified the relationship between
shear fatigue life can be obtained by torsional fatigue test. the amplifier output and the maximum shear stress, to control
However, since the load frequency of the conventional torsional the load stress.
fatigue tester is low, evaluation up to the giga-cycle was
practically impossible. Therefore, we developed an ultrasonic
3. Evaluation of bearing steel and estimation
torsional fatigue tester (vibration frequency of 20000 Hz,
of fatigue limit contact pressure
completely reversed) aiming at rapid evaluation of the shear
fatigue property of high strength steel for rolling bearings 1). This Fig. 2 shows shear fatigue properties of SUJ2 up to giga-
achievement was published in the journal of Japanese Society cycle regimes evaluated with the developed tester. We have
of Tribologists, “Tribologist”, and received the 2013 Japanese devised an estimation method of fatigue limit contact pressure
Society of Tribologists Effort Prize. from the shear fatigue limit which is drawn from the obtained
shear fatigue properties.
The fatigue limit contact pressure of SUJ2 estimated from the
2. Production of ultrasonic torsional fatigue shear fatigue limit was almost identical with the ISO Standard.
tester
Fig. 1 shows the configuration of the tester. The main 1200
components are a torsional vibration converter, amplifier and
torsion amplitude magnifying horn. Since the rated output of the 1000 S-N diagram
converter was small, the shape of the specimen and the
amplitude magnifying horn needed to be adjusted. The 800
specimen is dumbbell shaped and shear stress becomes
600
largest on the surface of the smallest diameter area when
torsional load is applied. Shear stress on the specimen is 400
τw0＝577 MPa
increased as the size of the specimen increases; however,
resonance became unstable and we could not continue the test. 200
Therefore, we limited the size of the specimen so that stable
resonance is provided at up to 90% of the amplifier's output and 0
105 106 107 108 109 1010 1011
increased the magnifying rate by improving the shape of the
Number of cycles
torsional magnifying horn so that high shear stress could be
stably given. Fig. 2 Shear fatigue property of SUJ2 evaluated by
developed tester
Converter of References
torsional vibration 1) Miwa, Matsubara, Shimamura, Ishii, Tribologist, 58, (2013) 658-666.
Photo of author
Specimen
Horn to magnify
torsional angle
Noriaki MIWA
Advanced Technology
Fig. 1 Schematic of ultrasonic torsional fatigue tester R&D Center
-101-
[ Our Line of New Products ]
Super Low Friction Hub Bearing

Significant reduction of hub bearing rotational friction during vehicle operation,
Results in approximately 0.3% improved fuel economy!
Features Structure
33% reduction in bearing rotational friction Minute dimples
(compared with the conventional units) in lip surface
(1) Newly developed low friction grease with improved

ingredient composition and low viscosity made Outer seal Inner seal
exclusively for bearings

(2) Minute dimples implemented to the seal lip sliding
surface
Applications
¡Axles for passenger cars
Low friction grease

exclusively for bearings
-102-
Our Line of New Products
EBW Driveshaft
EBW : Electron Beam Welding
Contributes to shorter development lead time

for automobile manufacturers with an innovative new production method!
Driveshaft on the differential gear side

Outer ring cup section Stem section
(standard design) (individual design for each model)
Welded section: Electron beam welding
Differential gear side

Tire side
Driveshaft (entire unit)
Features Structure
Contributes to shorter development lead Characteristics of electron beam welding
time for new models and higher reliability (compared with friction welding)
(1) Common outer ring cup section which had been Electron beam Friction
different in vehicles in the past (only the stem is Item welding welding
subject to individual designs) Welding time ◎ ○
(2) Secure and high precision welding in a short time with Strength of welded section ◎ ◎
electron beam welding
Concentric accuracy of welded section ○ △
Secondary processing after welding Not required Required

◎Excellent ○Good △Fair
Applications
¡Driveshafts for passenger cars
-103-
Low Section High Strength Press Pulley

By optimally shaping the design and improving press molding,
the industry's highest load resistance is achieved!
Features Structure
(1) Steel sheets 1.5 times thicker than conventional press
pulleys can be applied Pressed steel sheet
⇒Improves the pulley load resistance by two times or more
(2) Low pulley section of approximately 70% of
conventional products
⇒Improves bearing operating life and load capacity
(3) High durability under high temperature conditions and
excellent wear resistance against foreign matters
suitable for application in vehicles for emerging markets
Applications
¡Automobile engine accessory belt drive system
・ Optimal design achieved by utilizing analysis
technology
Air-conditioner compressor Low section high
strength press pulley ・ Improved press technology results in a reduced
section and thicker sheets
・ Revised thermal processing conditions for a
stronger mold, preventing mold damage and
seizure.
Alternator
Crank pulley Accessory belt
-104-
Our Line of New Products
Long-life Needle Roller Bearing Unit

More than 10 times the operating life compared to conventional products
with contaminated lubrication!
Features Structure
Long-life roller bearing unit achieving more than
Roller Shaft
10 times the operating life with contaminated
lubrication compared to conventional products
(1) Special (heat treatment) for rollers and shafts for Cage
longer operating life with contaminated lubrication
(2) Improved roller crowning shape to limit edge stress
due to bearing installation error
(3) Optimal design for the cage pocket shape to reduce Planetary gear
(partially cut away)
introduction of contaminants
(4) Added thrust washer discharge groove to improve
Thrust washer
contaminant discharge (oil discharge quantity: 6 times
the conventional method)
Gear and bearing cut away
(planetary reducer)
Applications
¡Planetary reducers for construction and
agricultural machinery
Swing drive
(Reduction gear box)
Final drive
(Reduction gear box)
Hydraulic excavator
-105-

NTN TR82 en

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NTN TR82 en

Uploaded by

Copyright:

Available Formats

For New Technology Network

Special Issue; Composite Material,

ISSN 0915-0528 OSAKA, JAPAN

Bearings under test

Preface Natsuhiko MORI 1

● Special Issue for Composite Material Products and Technology

● Special Issue for Manufacturing Technology

Lightweight Bearing for Robot 68

●Special Issue for CMS Supertechnology New Products

● Our Line of Award Winning Products

Our Line of New Products 102

A Message for the Special Issue:

Advanced Powders Processing

Advanced powder processing techniques such as metal injection molding (MIM)

processing, as well as development of new technology

Table 1 Recent applications of various P/M processings

Use Required properties Process and materials

equivalent to approximately 6 wt% in case of ferrous

The latter has significantly overcome those issues but, (volatile)

heptane gas phase, then heated degreasing and

controlling laser output. In the case of Fe-Ni mixed

80 Fig.10 Cross sectional microstructure of laser formed

value. One of the causes was assumed to be the

examined in the foregoing review.

Fig.15 Diameter of adhered particles

Product Introduction and Complex Technology of Resin,

NTN Composite Material Products Division has developed the

1. Introduction 2. Product line and production system of

1) Development and proposal of not only components

* Composite Material Products Division

inferior crack resistance against impact. The cost is also

(1) Resin sliding bearings1) for electric water pump Water

400 brass has a special surface treatment to improve

Table 2 Comparison of features various sliding screws

Sintered alloy bush

Example : Thinner gear

Fatigue strength MPa [Features]

7.6 96 500 400

(2) Multi-layer sintered oil-impregnated bearings 5)

also increasing. For joints of these hydraulic shovels, 0.20

use of sintered oil-impregnated bearings is increasing 0.15

prolong oiling intervals while strength and hardness are 0

The multi-layer sintered oil-impregnated bearings Fig. 12 Friction coefficient measurements

acceleration/deceleration, which requires step-up

・Test time: 300 hours 30 Injected/compressed

5. Magnetic material products Fig. 14 Comparison of features various core material

Nippon Kagaku Yakin Co., Ltd. develops and

(1) Reactors for automobiles 7)

Hard magnetic material:

Fig. 16 Temperature distribution analysis of hybrid core Ferrous soft

Fig. 19 Medical magnetic resonance imaging

Fig. 22 DC bias characteristics

Fig. 20 Parts of choke coil 80

Fig. 23 Frequency characteristics

Fig. 21 Choke coil for MRI power supply

The Improvement in Strength of Sintered Machine Parts

Sintered machine parts are used in many industrial fields. Especially,

1. Introduction 2. Sintered alloy steel

* Advanced Technology R&D Center

Table 1 shows the classification of metal powder of 2.1 Fe-Cu-C base

2.3 Fe-Cu-Ni-Mo-C base

2.4 Fe-Cr-Mo-C base