Handbook of Casting Defects

HANDBOOK ON CASTING
AND OTHER DEFECTS

In Gold Jewellery Manufacture
HANDBOOK ON CASTING
AND OTHER DEFECTS
In Gold Jewellery Manufacture
by Dieter Ott
Copyright© 1997 by World Gold Council, London
Publication Date: November 1997

Reprinted 2001
Published by World Gold Council, Industrial Division,

Times Place, 45 Pall Mall, London SW1Y 5JG
Telephone: +44 (0)20 7930 5171. Fax: +44 (0)20 7839 6561
E-mail:[email protected]
Produced by Dieter Ott, FEM, Schwäbisch Gmünd, Germany

Editor: Dr Christopher W. Corti
Originated and printed by Trait Design
NOTE: Whilst every care has been taken in the preparation of this publication, World Gold
Council cannot be responsible for the accuracy of any statement or representation made or
the consequences arising from the use of information contained in it. The Handbook is
intended solely as a general resource for practising professionals in the field and specialist
advice should be obtained wherever necessary.
It is always important to use appropriate and approved health and safety procedures.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without prior permission in writing of the copyright holder.
P R E FA C E
Gold has an image of preciousness - colour, rarity, indestructability, value, etc. - and this is
reflected in the ever growing desire to own gold jewellery. Thus, the consumer’s perception
of gold as a unique, attractive and valuable metal should be reflected in their expectation of
a quality-made jewellery product. As manufacturers of gold jewellery are well aware, this
expectation is not always so easily fulfilled in practice and production of gold jewellery, as
in all metal fabrication, is prone to the occurrence of defects which can be difficult and
expensive to rectify. Many progressive jewellery manufacturers are refocusing their
businesses to a Total Quality approach. The old maxim ‘prevention is better than cure’ has
strong economic benefits for the producer. In an increasingly international and competitive
industry, high quality at low cost is essential.
Investment casting is now the dominant mass production process in gold jewellery
manufacture and is especially prone to defect formation. The nature and causes of such
defects are generally not well understood in the industry and thus their prevention during
manufacturing is not easily attained. In recent years, however, much light has been shed
on these problems and we now have a good scientific understanding of the principal
defects and their causes; this knowledge has been translated into manufacturing ‘best
practice’ which, if followed, should lead to a significant reduction in the probability of
defect formation.
World Gold Council provides technical support to the gold jewellery industry in terms of
disseminating technical information on manufacturing technology and ‘best practice’
through its journal, Gold Technology, its International Technology Symposia held annually
in Vicenza, Italy, and through a series of local symposia and technical seminar programmes
held in the major jewellery production centres worldwide. Increasingly, these latter are
being supported by a series of technical publications. The Technical Manual is a practical
basic guide to manufacturing best practice and this is supplemented by more specialised
publications on specific aspects such as Investment Casting and Finishing. This Handbook
complements these and describes typical common manufacturing defects, their causes and
ways to prevent them. It is the first major publication on this topic and takes a systematic,
user-friendly, technology-based approach and, I believe, it will prove to be essential reading
for all serious jewellery producers interested in quality.
Inevitably, a Handbook such as this cannot cover all defects that can and do occur in
jewellery manufacture but we believe that most of the common ones are listed. Much of
the material in this Handbook is based on a comprehensive case history study of actual
production defects carried out at the German Precious Metals Research Institute, FEM, in
Schwäbisch Gmünd, supported by the Santa Fe Symposium on Jewelry Manufacturing and
World Gold Council. It was our concern that this valuable reference information source
should be available to the industry in a practical and useful way; hence, the production of
this Handbook. We would like to continue to add to this ‘defect database’ and FEM would
be pleased to receive other defects in jewellery for investigation to provide further
information that can be shared with others.
Christopher W. Corti
Manager, Technical Information & Development, World Gold Council.
CONTENTS
Preface 3
Introduction 6
Part A: Survey of Defect Types 9
Part B: Case Histories: Description of Defects 15

1 Investment Casting 16
1.1 Porosity 17
1.2 Form filling 25
1.3 Surface quality and investment 28
1.4 Inclusions 37
2 Ingot Casting and Continuous Casting 45

2.1 Cracks and material separation 45
3 Defects caused by Alloy Composition 49

3.1 Low melting impurities 49
3.2 Inclusions 55
4 Corrosion, Tarnishing, Discolouration 57

4.1 Discolouration 57
4.2 Stress corrosion and intergranular corrosion 60
5 Soldering 63
5.1 Fracture 63
6 Heat Treatment 65
6.1 Age hardening 65
6.2 Blistering 67
Part C: Basic Aspects of the Metallurgy of Carat Golds

and the Investment Casting Process 71
1 General Properties of Metals 72
1.1 Physical properties 72
1.2 Structure 72
2 Metallurgical Effects 73
2.1 Melting and solidification 73
2.2 Deformation and mechanical properties 75
2.3 Annealing and heat treatment 76
2.4 Low melting components 78
2.5 Grain refiners 79
3 Investment Casting 80
3.1 Shrinkage porosity 80
3.2 Gas porosity 81
3.3 Influence of investment on casting quality 81
3.4 The investing process (gypsum bonded investment) 82
3.5 Behaviour of gypsum-bonded investment during casting 86
Further Reading 88
Acknowledgements 91
a
INTRODUCTION
Just as in any production process, jewellery fabrication can give rise to
defects which lead to rejects and customer dissatisfaction. The defect rate
depends strongly on the type of production process. The focus of this
Handbook is on investment casting, which is known to be a critical
process. This might be explained by two facts:
a) Investment casting is widely used for the production of a great
variety of jewellery items in an equally large number of casting
shops with levels of equipment, expertise and experience that vary
widely.
b) Investment casting is a complex set of processes which is
determined and influenced by numerous physical, metallurgical and
chemical factors. Mastering and controlling all of the process steps is
difficult, if not almost impossible.
Of course, other production methods are also susceptible to defect
formation, such as wire-drawing, rolling, heat treatment and surface
treatment. Corresponding cases are also included in this Handbook.
Naturally, if a defect occurs, measures have to be adopted to eliminate
its cause and prevent its repetition. Of course, this is only possible if the
defect can be identified and its cause determined. Therefore the following
procedure should be followed:
• identify the nature of the defect
• seek the cause
• develop procedures to prevent the cause from recurring.
The first step is not as easy as it seems. Some types of defects may
appear almost identical although they have completely different causes.
A good example is porosity. Pores visible on the surface of a polished
jewellery item may look very similar, independent of their origin.
The problem can only be solved by a two-pronged approach:
• exploring the ‘history’ of the defective item
• performing a metallurgical and/or chemical investigation.
Identification of the nature of a defect is at least half the way towards
finding its cause. Understanding the cause is the first and most important
step towards avoiding repetition of the defect in future. This Handbook
will provide some help in performing these tasks.
Part A is a structured survey of defect types which, with the aid of
diagrams and photographs, should act as an orientation guide. The visual
appearance of a defect, together with some simple information, should
guide the user of the handbook to case descriptions which might fit the
actual case. However, confirmation by extended investigation will be
necessary in most cases.
A number of case histories of actual production defects have been
investigated and are described in Part B. A rough classification is made in
terms of causes and types of defects. It is not unambiguous.
The description of each defect case follows a constant pattern. Firstly,
the defect is described as it appears with visual inspection (use of a
magnifying glass or a stereomicroscope can help a lot). Photographs
support the description as far as possible. Additional information is
6
mentioned in “Alloys” and “Manufacturing method” (so far as necessary
and available).
Unfortunately, it is generally not possible to identify the type and the
cause of each defect only by visual inspection. Frequently, additional
investigation is necessary. We are conscious that this may constitute a
disadvantage for the straight use of this Handbook in the workshop.
However, there is no easy way to overcome this difficulty. Therefore,
in addition, other parts which follow the defect description (“Influence
on properties”, etc) present the results of such metallurgical investigation.
The “Brief explanation” provides initial information about the kind and
cause of the defect. Additionally, more technical and metallurgical
information is provided in “Extended explanation”, where appropriate.
Last but not least, advice on preventing the defect is given under
“Recommendations for avoidance”.
More information on the metallurgical and technical background to
the processes and causes of defects is supplied in Part C. The purpose of
this part is twofold:
a) To improve the understanding of the investment casting process
and, through this, help to avoid defects.
b) Many defects have a very similar metallurgical background; to
explain in one place saves much repetition. In each case history, a
reference to the relevant section is given under “Further reading”.
‘Part C: Basic aspects of metallurgy’ focuses primarily on the
investment casting process and some other common problems in
jewellery fabrication. It is not a textbook about metallurgy in general or
precious metals in particular. It is an effort to produce an understandable
manual for practising craftsmen and technicians who have not been
technically trained in metallurgy or chemistry.
One very important aspect has been disregarded: alloy phase diagrams.
Understanding phase diagrams needs a more extended explanation which
is not appropriate in this chapter.
Hopefully, this Handbook will help jewellery producers to improve the
quality of their jewellery fabrication. However, it cannot be completely
comprehensive as it is based on the investigation of a limited number of
jewellery defects submitted to the author, mainly through the auspices of
the Santa Fe Symposium, with support of World Gold Council.
The author would like to continue to add to our defect database and
requests that jewellery producers continue to send him their defects for
investigation. Hopefully, a supplement to this Handbook can be
published at a later date.
Dieter Ott
Schwäbisch Gmünd
October 1997
7
PA R T A :
SURVEY OF
DEFECT TYPES
Surface defects
Kind of Production Typical for: Defect
Defect Method (e.g.) Alloy Case(s)
Investment A
Casting General Gas pores 1,2
Reaction
Small with
Investment investment B
pores on Casting and
General Gas pores 1,2
surface remelting
Investment polluted alloy C
Casting General Dendritic pores 3
A B C
Surface defects
Investment Casting, Remelting of A

polluted material polluted material Pin holes 1
Pores and Remelting of B

General
dendrites Casting in polluted material Dendritic pores 2
reducing
Atmosphere Reaction with C
investment Dendritic surface 3
A B C
10
Surface defects
Investment Casting, Polluted crucible, A

too much flux General
melting procedure Slag (‘cauliflower’)13
Irregular B
Investment Casting, Remelting polluted Zinc containing
surface melting procedure metal, oxidation alloys Inclusion of zinc oxide
structure (‘crow’s feet’) 12
Investment Casting, High heating rate C
at burn-out General
burnout cycle Wrinkled surface 9
A B C
Defects caused by investment or investing process

A
Quality of Sandy surface 7
investment powder, Investment was
investing procedure, too weak
Rough Centrifugal casting General
B
surface Fins 7
Investing Wrong investing C

procedure procedure Water marks 8
A B C
11
Inclusions
Investment Inclusions of A
Casting grain refiner 20
Abundant if grain
Hard refiners are used
Nests of B
Rolled sheet grain refiner 20
spots
Bad quality of gold C
Rolled sheet Impurities in gold used for alloying 20
A B C
Fracture
Casting, alloy Low melting Yellow gold with A

Brittle silicon
composition components Silicon phase 17
Brittle Casting in reducing Reaction with Jewellery alloys B

Formation of
fracture Atmosphere investment in general sulphide 19
Heat treatment, Age hardening in Red and pink C

Embrittlement by
Soldering red/pink gold alloys soldering 24
A B C
12
Fracture
Investment Shrinkage General, 14ct alloys A

Casting Porosity are more critical 4
Fracture,
Investment Cold B
material Casting shuts 6
separation General
Ingot casting, Lamination caused C
rolling by ingot pipe 16
A B C
13
PA R T B :
CASE HISTORIES:
DESCRIPTION OF DEFECTS
NOTE: The photographs in this section have been reproduced from slides and are magnified by approximately 150%.
Therefore, where magnifications are given, they are not exact but should be considered as approximate.
1
1 INVESTMENT CASTING
INTRODUCTORY REMARKS
Investment casting is a very useful technique in medium- to large-scale

jewellery production. However, it can also be a source of many defects,
with up to 10 - 20% of defective items not unusual.
A large proportion of rejected items is due to porosity. There are two
different types of porosity - shrinkage and gas porosity - which, in their
most characteristic shape, can normally be distinguished easily. However,
in practice their appearance can vary widely, and all intermediate shapes
between small spherical holes (gas porosity) and dendritic pores
(shrinkage porosity) are possible.
The origin of porosity is also variable. Shrinkage porosity is caused by a
purely physical process (the decrease in metal volume at solidification)
whereas gas porosity was thought originally to be caused solely by
chemical decomposition of the mould investment, either directly or via
sulphide formation (see Part C: Basic Aspects, section 3). Whilst often the
case, other causes can also lead to gas porosity, such as oxidised and
contaminated metal.
In the following case histories, several defects are illustrated that have
similar origins but different appearances.
Another common cause of defects arising during investment casting is
connected to the investing process, which commences with slurry
preparation and ends with the flask (mould) burn-out process. Poor
quality investment powder may add some extra difficulties.
Incomplete form-filling (mould-filling) is another common defect that
should be mentioned but which is often not well understood or properly
investigated. The causes of poor form-filling can be detected relatively
easily.
Other reasons for defects in investment casting are not so common.
Cracks and fractures can often be related to porosity or alloy properties.
Such defects induced by alloy properties or alloyed impurities are
reported in chapter 3.
16
1.1 POROSITY
1
CASE 1: PIN HOLES ON THE SURFACE - CAUSED BY GAS POROSITY
Similar to
Cases 2 and 3
Key Words
Gas porosity; static casting; contaminated/polluted material
Fig. 1.1 The polished surface
Description of Defect
The polished surface of a flat jewellery item shows numerous, uniformly
distributed round pores.
Visual Appearance
The pores are like pinholes and rather randomly scattered on the surface.
The type of porosity and its origin cannot be identified only by its visual
external appearance. Examination of a micro-section is necessary.
Alloy: 18ct yellow gold (AuAgCu).
Manufacturing Method: Static casting in argon atmosphere, starting Fig. 1.2 Microsection through a piece of
pressure 0 bar, final pressure 2-3 bar. jewellery with gas porosity. 50 x
Influence on Properties
Surface Quality
The pinholes are detrimental to the quality of a polished surface. In most
cases, they cannot be removed by extended grinding or polishing; on the
contrary, even more pinholes (lying below the surface) will be exposed.
Mechanical Properties
Pinholes usually have little influence on the strength of the item, in
contrast to other types of porosity which can result in fracture.
Microstructure
Typically, the gas pores are roughly spherical and are often situated in a
discrete layer just beneath the surface, indicating their origin from reaction
of the investment at casting. Alternatively, as in this case, they may be
distributed across the entire cross-section, indicating their likely origin as a
reaction of contaminated (polluted) alloy. The gas pores are small but
abundant. Only a few pores are exposed and visible on the surface.
Obviously, attempting to remove these pinholes by polishing them away is
unrewarding because other, subsurface pores will then become exposed.
Brief Explanation
The almost homogeneous distribution across the cross-section and the
smooth surface indicates that, in this particular case, the porosity is probably
not caused by reaction (decomposition) of the investment during pouring of
the melt into the flask. More probably, the substance producing the pores is
introduced with the metal. In the present case, the use of remelted scrap
material containing sulphides, together with new metal (made from fine
gold and a master alloy), is responsible. The master alloy contained copper
oxide which reacted with the sulphide to form sulphur dioxide gas, resulting
in this typical gas porosity. Note that contaminated material can also cause
porosity with a different appearance (for example, see Case 3)
17
1 Recommendation for Avoidance
Two measures should be taken to prevent the occurrence of the defect:
a) Use oxide-free, clean master alloys (or oxygen-free silver and copper)
for preparing a new alloy melt. Note: Fine silver can have a high
oxygen content. It can be removed by melting the silver separately
under reducing conditions prior to alloying.
b) If used alloy scrap such as feeders is remelted, it must be thoroughly
cleaned to remove any mould investment material as this can react
Fig. 1.3 Typical layer of gas pores just
beneath the surface 9 x with the alloy, leading to the formation of sulphides. Used alloy
from castings with severe porosity should not be remelted again but
refined.
Extended Explanation: Copper oxide(s) can exist in pure copper (use
oxygen-free ‘high conductivity’ copper for preference) and in silver-copper master
alloys up to a high concentration. Silver can also contain significant amounts of
oxygen. In yellow carat gold alloys, any copper oxide present can be reduced to
copper with formation and release of oxygen gas under certain conditions.
Sulphides of silver and copper can be formed on remelting dirty scrap from
casting when remains of the old investment are present. The gypsum (calcium
Fig. 1.4 Detail of fig.1.3, gas pores at sulphate) binder in the investment can be decomposed, reacting with copper
higher magnification 50 x
and/or silver during remelting to form sulphides. When sulphide and oxide (or
oxygen) are present at the same time, they can react chemically during melting
with the generation of sulphur dioxide gas, thus causing gas porosity.
(The simultaneous presence of sulphide and oxide has been proved in several
investigations).
When the gas porosity only extends to a layer below the surface, the cause is
more usually due to the decomposition of the gypsum binder during pouring of
the molten metal. The temperature at which this occurs is lowered in the
presence of carbon (hence the need for a clean burn-out).
Another Example
This example, figs 1.3,1.4, of a discrete layer at the surface of a 14 carat
yellow gold is more typical of gas porosity due to reaction of the
investment.
Further Reading
Part C: Basic Aspects, section 3.2, Gas porosity.
18
CASE 2: GAS POROSITY - CAUSED BY IMPURITIES (A FURTHER EXAMPLE)
1
Similar to
Case 1
Key Words
Investment casting; surface porosity; contaminated/polluted material
Description of the Defect

Numerous small pores on the surface of rings
Fig. 2.1
Visual Appearance
The surface looks as if it is perforated by small pores and holes.
This appearance is typical for any kind of gas porosity and also for
shrinkage porosity. Polishing evens out the typical differences in the
appearance of both kinds of porosity. Hence, for reliable information on
type and cause of porosity, microsections have to be prepared.
The pinholes are small. Without use of a microscope, the defect often
looks like cloudy spots. At higher magnification, the irregular shape of
pores can be recognised.
Alloys: Independent of alloys. Examples are shown of 10ct and 18ct
yellow gold alloys.
Fig. 2.2 Another example of the same kind
Manufacturing Method: The main source for this type of porosity is of defect
remelting contaminated metal (see below). Melting and casting under
oxidising conditions are another reason. These defects are often seen
when casting is done in an open system without a protective atmosphere.
Microstructure
The structure shows severe porosity, fig 2.5. The pores, which are close to
the surface, are of two types of pores: small, almost spherical gas pores
and irregularly shaped larger pores. At higher magnification, small
inclusions are visible which are probably copper oxide (CuO). Again
caused by impurities, the mixture of spherical gas pores and irregular
Fig. 2.3 Typical example at higher
pores, resembling shrinkage porosity, can be recognised. Most cases with magnification 100 x
really severe porosity are caused by this type of defect.
Brief Explanation
Several cases of defects involving severe porosity have been found which
could not be attributed unambiguously to just shrinkage or gas porosity.
The porosity comprises both smooth spherical shaped pores, typical of
gas porosity, and very irregular, large ones. This kind of defect is often
found in cases where oxidised metal has been used and/or no precautions
taken to prevent uptake of oxygen during melting. Typically, oxide
inclusions are found in the microsections.
The cause of this type of defect is not well understood. A possible
Fig. 2.4 Another example at higher
explanation is that abundant inclusions of copper oxide decompose when magnification 100 x
the melt is about to solidify, releasing oxygen and forming gas pores.
Whether other factors such as the presence of reducing agents (old
investment/sulphides) play a role is not clear. Further study is necessary.
19
1 However, it can be stated for certain that polluted, oxidised material plays
a major role in this defect type.
Recommendation for Avoidance

This defect can be avoided if only clean, unoxidised material is used and
the casting conditions are controlled to avoid extensive oxidation. The
amount of scrap used in a new melt charge should be limited - normally
up to 30 - 40%, depending on the quality (cleanliness) of the metal.
Continued re-use of scrap will also lead to a build-up of deleterious
Fig. 2.5 Microstructure of a ring with a impurities in the alloy.
porous surface. 50 x Extended Explanation: The explanation given above is not yet fully
proven and is based on practical experience. In the system gold-copper-oxygen-
(silver), the stability of copper oxide is temperature dependent. In alloys strongly
contaminated with oxides, the equilibrium between copper in solid solution,
oxygen and copper oxide will change during solidification causing evolution of
gaseous oxygen.
A distinction between this kind of porosity and that caused by sulphide
inclusions (from reaction with investment) is very difficult.
Further Reading
Part C: Basic Aspects, section 3.2, Gas porosity.
Fig. 2.6 Another example. 200 x
20
CASE 3: NESTS OF SMALL PORES ON THE SURFACE
1
Similar to
Cases 1 and 2.
Key Words
Surface defect; dendritic porosity; contaminated/polluted material

A ring shows nests of pores on the surface. The shank is covered with
Fig. 3.1 9 x
dark brownish spots. The surface cannot be improved by stronger
polishing. It is a characteristic feature of this kind of defect that extended
polishing tends to increase the incidence of the defect rather than
improve the surface quality.
Visual Appearance
The shank shows many dark spots. Distinct pores can be recognised,
using a pocket magnifying lens. The defect is not restricted to critical
places such as necks, or transition zones between thick and thin cross-
sections, etc.
Alloy: 14 ct yellow gold
Manufacturing Method: Only a few details were given about the
Fig. 3.2 50 x
investment casting conditions. However, some very important details
were noted: (1) the casting temperature was in the range 980 - 1010°C
and (2) a melt charge comprising 40% new metal/60% scrap was used.
There was no actual complaint about fracture. However, fracture might
occur if the ring was to be expanded (“sized”).
Microstructure
Both micro-graphs show the cross-section of the shank at different
magnifications. At low magnification, nests of dendritic pores are visible,
randomly distributed all over the cross-section. Only a few of them reach Fig. 3.3 200 x
the surface, causing dark spots on the polished shank. The microsection
makes clear that further polishing would not improve the surface; on the
contrary, more pores would be exposed.
At higher magnification, some interesting details are revealed: the
pores are a mixture of roughly spherical, small gas pores and dendritic
porosity. Real shrinkage porosity can be excluded as a cause; another
explanation has to be found.
Brief Explanation
Several instances have been investigated showing defects similar to those
shown in this case which, at first sight, appear to be attributable to
dendritic shrinkage porosity. However, more detailed examination leads
to the conclusion that the main cause is impurities, probably copper
oxide, introduced by use of dirty scrap material or improper melting
methods (for further explanation, see Cases 1 and 2).
In this case, the use of 60% of recycled metal in the new charge is
21
1 relatively high. Such a high proportion can only be acceptable if the
reused scrap is very clean. Otherwise, in the course of the time, impurities
(oxides, inclusions, embrittling metals, etc.) will accumulate to damaging
levels.
The other critical point is the comparatively low metal casting
temperature. Depending on the casting method, this temperature is
probably sufficient for form-filling. However, if impurities (especially
oxides) are present, they are unlikely to be separated from the melt at low
melt temperatures and short melting times. Increasing the melt
Fig. 3.4 temperature to 1050 - 1100°C would help to minimise the defect.

The only safe way to avoid this kind of defect is the use of clean materials
and melting in a clean crucible under protective gas. The proportion of
(clean) recycled scrap material should be restricted to no more than 30-
40% of the total melt charge.
Sometimes, increasing the casting temperature will also help to remove
oxides (if melting is performed in a protective atmosphere).
Another example
The connection between oxide inclusions and porosity can be
demonstrated with the following example, a broken 18ct item.
The microstructure, fig 3.4, shows heavy porosity. The pores are partly
filled with oxides (identified as copper and zinc oxides). Around these
pores, many small inclusions of oxides are visible.
Further Reading
Part C: Basic Aspects, section 3.2, Gas porosity
22
CASE 4: FRACTURE - CAUSED BY SHRINKAGE POROSITY
1
Key Words
Fracture; shrinkage porosity

A petal of a flower-like jewellery item broke off after final polishing. The
fracture occurred at a low applied stress. The fracture surface looks brittle
and oxidised.
Fig. 4.1
Visual Appearance
A flower was cast in 18ct. yellow gold. One of the petals broke off when
the item was bent by a very small amount. No deformation of the
material is visible. The only remarkable detail is a brownish (‘oxidised’)
discolouration of the fracture surface (which is not revealed on the
photograph). Visual inspection does not disclose the cause of the defect.
Examination of the microstructure is essential (see figures 4.2 and 4.3)
Alloy: This particular defect was observed in 18ct yellow gold alloy
(75.0 Au, 15.0 Ag 9.3 Cu, 0.7 Zn). However, the same defect can also
occur in other jewellery alloys.
Manufacturing Method: The flower was vacuum cast; casting and
flask temperatures are not reported. In another example, fig. 4.3, casting Fig. 4.2 50 x
of a 14ct alloy was by a combined vacuum and pressure assisted method
with a very low flask temperature (200°C) and a high melt casting
temperature (1200°C).
Microstructure
A microsection through the fracture of the petal (ref. fig. 4.1) discloses a
typical morphology. The petal, with its relatively thick cross-section, is
connected to the calyx of the flower by a thin neck. The fracture has
occurred at this neck, the thinnest part of the item. Only in this region,
a spongy structure of dendrites is formed, with voids inbetween.
This example shows the microstructure of the centre of a cast ring in Fig. 4.3 50 x
14ct yellow gold. The spongy shape of the structure is identical with the
example shown above.
It is evident that such a spongy porous structure in a critical narrow
connection will drastically reduce the strength locally. The part will break
off with only a little force.
Others
The same defect can also be manifested at the surface. The dendritic structure
to the pores should be recognisable. However, if the surface is sand blasted or
tumble polished, the typical appearance is masked and examination of
microsections is necessary to reveal the true nature of the porosity. As before,
any attempts to remove the pores by polishing will not be successful.
Brief Explanation
The defect is a exemplary case of shrinkage porosity in terms of
appearance and the cause. It must be noted that liquid (molten) metals
23
1 shrink on solidification, causing a contraction in volume, and that heat is
lost from the molten metal through the mould walls.
At the beginning of solidification, solid ‘crystallites’ grow in the melt
in the typical dendritic form (it can be compared with the structure of
branches of a fir tree). With continued solidification, the dendrites will
grow and eventually meet each other; the spaces between them arise from
the shrinkage on solidification and should be filled with solidifying
material, if liquid metal can continue to be fed into these voids.
In this case, solidification starts and proceeds faster at the thin walled
neck (since heat is more quickly extracted here) compared to the adjacent
thicker parts. The flow of liquid metal to the thicker petal to feed the
voids arising from solidification shrinkage becomes difficult due to the
resistance to flow in the spongy neck region and any remaining liquid in
this region is sucked into the petal region, leaving the neck with
shrinkage porosity.
The situation with the ring is somewhat different. Here, solidification
started very quickly at the surface due to the low flask temperature, as did
the gate and other thin walled parts of the item. These parts completely
solidify while the core of the ring is still liquid and so interrupting further
supply of liquid metal to the core region. When the core starts
solidifying, the supply of melt necessary to compensate the shrinkage in
the core is no longer possible.
Extended Explanation and Further Reading

See Part C: Basic Aspects, section 3.1, Shrinkage porosity.

In the case of the flower, the best and safest method for avoiding the
defect is changing the design. Thin necks between adjacent thicker parts
are always a likely cause of such problems and should be avoided, where
practicable. Alternatively, repositioning of the gates or the addition of
extra gates to feed the thicker regions may overcome the problem.
Another approach for reducing the shrinkage porosity is to change
flask and/or casting temperatures. However, which direction to change
them to yield improved results is difficult to predict. In the present case,
it is probable that increasing the flask temperature will improve the
quality.
In the case of the ring, however, a way to effect an improvement is
much easier to predict. The flask temperature is definitely too low and
should be at least 400°C (still relatively low) and the casting temperature
should be decreased.
24
1.2 FORM-FILLING
1
CASE 5 SHRINKAGE POROSITY - CAUSED BY IMPROPER GATING
Key Words
Shrinkage porosity; surface defects; gating, sprueing

A part of a tie clip shows small pores on the surface which are concentrated
on the tapered end of the item. At higher magnification the dendritic Fig. 5.1 Tie Clip
structure of the pores is revealed, figs. 5.2, 5.3. Polishing does not remove
these defects. Some of the cast items broke near the porous end.
Visual Appearance
The pores, which are concentrated on the tapered end, are too small to be
visible in this figure. With the naked eye they appear as dark clouds.
At higher magnification, the dendritic structure of the surface defects
can be recognised.
Alloys: This particular defect was observed in a 14ct yellow gold alloy
(58.6 Au-17.0 Ag-19.5 Cu- 0.5 Zn). However, the same defect can occur
with other jewellery alloys.
Manufacturing Method: The clip was cast by the combined pressure Fig. 5.2
and vacuum assisted method. An important detail is the ‘gating’ of the

item (often erroneously called ‘sprueing’). Fig 5.4 shows the wax model
together with the gate which attaches it to the central sprue. Note that
the gate itself is relatively thick and short. It is attached to the hook-
shaped end of the tie clip.
Microstructure
A microsection through the porous end of the clip, at the opposite end to
the gate, fig 5.5, discloses a typical dendritic appearance of the pores,
resulting in a spongy metal structure with dendritic pores.
Mechanical Properties Fig. 5.3
Such a pronounced dendritic porous structure drastically reduces the

strength.; the part may break with little effort.
Brief Explanation
The defect is an other exemplary case of shrinkage porosity. However, in
this case, the gating of the wax is a critical point. The gate itself is short
and sufficiently thick for supplying melt during solidification. However,
the positioning of the gate on the wax model has to be criticised. The
molten metal, fed to one end of the clip, has to be transported through
thin bridges to the other, thicker end. The melt freezes on its way through
the thin cross-sectioned part, resulting in an insufficient supply to the
other end of the clip and, hence, severe shrinkage porosity occurs there. Fig. 5.4 Wax model with attached gate
The gate might be better attached to this thicker part of the clip, thus
alleviating this problem. However, there is a high probability that shrinkage
porosity will now occur at the opposite end for the same reason. There is
25
1 no ideal solution for gating. A forked gate with two branches attached at
opposite ends of the clip will help. However, in practice the branches
would be of considerable length. A positive effect cannot be guaranteed.
Fig. 5.6 shows a similar case with a divided gate. Of course, a good result
was not achieved in this way. Many trials in this case resulted in the use of a
three branched gate (fixed to both ends and the middle) whereby the
diameter of gate was almost the same as the needle itself. Shrinkage porosity
could be avoided with this method, but material consumption is large.
Fig. 5.5 80 x
The design of the clip is very unfavourable for investment casting. There is no
place for applying a single gate to ensure an item free of shrinkage porosity. A
solution would be the use of a two branched gate, attached to each end but
the branches must have such a sufficient diameter that a practical appliance
will not be realistic. Two separate gates might be a preferred solution.
Experimenting with melt and flask temperatures might give a limited
possibility to reduce shrinkage porosity. However, no firm recommendation
can be given in this case.
More about gating

Figures 5.7 and 5.8 show favourable and unfavourable ways of gating.
Fig. 5.6 Wax model of a needle with two
branches on a sprue.
The subjects on the left hand side show wax models with gates designed
and positioned by the jewellery designers; these resulted in incomplete
filling of the mould. Those on the right hand side demonstrates what the
casters did in repositioning and sizing the gates to improve the quality,
i.e. to achieve good form-filling.
A part of a brooch (fig. 5.7, top) is an excellent example. The left side
model has extremely thin gates. Of course, removing the gates from the
cast item can be done with a minimum of work, but porosity would be a
serious problem in this case. The problem is solved using a three
branched, relatively thick gate (shown on the right side).
The dog’s head (middle, left) is gated on the thinnest part of the item.
Fig 5.7
Even if the gate is sufficiently thick and short (as in this case), premature
solidification in the thin parts interrupts melt supply and causes
shrinkage porosity. The solution is seen on the right side.
The gate on the item, left bottom, is fairly long and somewhat tapered
towards its end; it is fixed to the wax model in a way which produces
problems in cleaning the casting. On the right side, the gate has been
shortened, but has a neck. Necks should be avoided if possible. Investment
particles can break off in these places, obstructing melt flow. The best way
of attaching the gate would be midway between these extremes.
The three examples in figure 5.8 show a good gate design, with regard
to length and diameter. However, the gate of the ear-ring (top) is not
attached properly. The tangential fixation of both branches provides only
a small connection area, similar to a neck. As already mentioned necks
Fig. 5.8 should be avoided.
Further Reading
Part C: Basic Apects, section 3.1, Shrinkage porosity.
26
CASE 6: COLD SHUTS - CAUSED BY LOW MELT AND/OR FLASK
1
TEMPERATURES
Key Words
Cold shut; investment casting; form-filling.
The surface of a flat thin piece shows traces which resembles cracks.
However, in contrast to cracks, they have smooth edges and the surface
in the surrounding of the traces is slightly wrinkled.
Fig. 6.1
Visual Appearance
The defect was observed on a segment of a gramophone record used as a
test specimen for testing casting conditions. The thin flat shape of the
record is very susceptible to cold shuts. Grooves and a wrinkled surface
structure are characteristic of a defect called a ‘cold shut’:
Alloy: This defect can happen with any alloy
Manufacturing Method: Cold shuts can occur with any casting
method. Low casting or/and flask temperature usually are crucial factors.
Cold shuts can become cracks when stress is applied to the jewellery piece.
Surface quality
The surface is severely disrupted. Polishing or even grinding will not help
to produce a good finish.
Microstructure
The microsection reveals that, in the defective region, the material is
partly separated. Two flows of melt met without fully fusing together.
Brief Explanation
The melt enters the hollow space of the mould with some turbulence.
Usually, the turbulence ‘disappears’ during solidification with no sign of it
visible in the solid state. Sometimes, however, the ‘tongue’ of the
turbulent melt solidifies on the mould wall before filling is complete. The
turbulent structure is ‘frozen’ and the subsequent molten metal filling the
mould cavity does not fuse into it, causing the formation of a cold shut.
Several filigree pieces on the same tree may be reproduced incompletely.
A very similar defect can occur in zinc-containing alloys. Zinc oxide,
formed during melting and casting, can form a thin membrane that floats
in the melt and prevents the adjacent parts of the melt from fusing
together on solidification.

Formation of cold shuts can be avoided by increasing the casting and/or the
flask temperatures but restricted to the minimum amount necessary to avoid
the occurrence of other kinds of defects (e.g. resulting from reactions with
the investment). Material separation by films of zinc oxide can only be
avoided by using clean material and a protective atmosphere.
27
1 1.3 SURFACE QUALITY AND INVESTMENT
CASE 7: ROUGH SANDY SURFACE AND FINS - TYPICAL DEFECTS
CAUSED BY WEAK INVESTMENT
Key Words
Investment casting; surface defects; sandy surface; fins; investment failure.
This case includes two different defects with a common cause. They often
Fig. 7.1 Sandy surface on a relatively
occur together and more usually on large, heavy flat items. The surface of
heavy test piece. the casting appears rough and ‘sandy’. It seems as if a wet surface has
been gritted with fine grained sand. On the edges of cast items, thin ‘fins’
may be formed. Fins are thin foils of material with irregular shape
adhering to the edges of the casting.
Commonly, both types of defect are obvious after removing the
investment from the cast tree. Sand blasting can partially remove the
defects and alter the appearance. In this case, it is more difficult to
identify the type and origin of the defects.
Visual Appearance
Centrifugally cast, the part of the tree most affected by this defect was at the
Fig. 7.2 Fins on a test tree. flask bottom (opposite end to the melt button). Small fins can also be seen.
Typically the fins occur on edges of heavy cross-sectioned parts, at the
flask bottom (opposite end to the melt button).
Alloys: In general, high carat alloys may have an increased propensity
for this defect because of their higher density, placing a greater force on
the investment.
Manufacturing Methods: The main cause of the defect is an
incorrect investing procedure and/or a bad batch of investment (stored
under the wrong conditions or for too long).
However, the casting method has also an influence. The higher the
pressure in forcing the melt into the flask, the higher is the danger for
getting fins or a sandy surface. With static casting, the pressure is limited
Fig. 7.3 Microsection through the edge of a by several factors. Centrifugal casting machines enable a much higher
piece with fins. 50 x pressure to be exerted, depending on rotational speed. Thus, this kind of
defect is more often reported when centrifugal casting is used.
Typically, the faults are more abundant at the bottom of the flask (end
away from the melt button). The highest pressure exerted on the mould
walls, in both centrifugal and static casting, occurs at the flask bottom.
Both types of defects are restricted to the surface. Neither mechanical
properties nor microstructure are influenced. The main disadvantage is a
higher level of rejects and/or a greater effort in finishing (polishing) the
jewellery items. Deterioration of fine surface details causes rejects.
Brief Explanation
The main cause for both kinds of defects is a structural weakness of the
mould investment. The investment cannot resist the pressure and
abrasion of the incoming melt. A weak, friable mould surface, due to poor
28
investment or, possibly, too rapid burn-out or casting conditions, leads to
1
a rough, sandy surface structure after solidification.
In addition, on edges of heavy cross-sectioned parts, the investment
breaks or cracks, allowing melt to penetrate the mould wall and resulting
in the formation of fins.
The main reasons for the failure of investment are:
a) A bad batch of investment was used (investment powder too old or
stored in humid surroundings - causing premature setting of the
binder)
b) The investment was mixed with a too high a quantity of water.
c) Sometimes, accelerated heating of the flask in the burn-out cycle
causes a break-down of the investment.
The tendency to induce the defect increases with increasing pressure
during casting. Use of too high a speed of rotation in centrifugal casting
is critical. In extreme cases, even sound investment can break. The force
acting on the investment wall is highest in the case of large, thick walled
items; therefore, the defects are more likely to be encountered with such
jewellery items.

In the first place, the quality of investment powder and the correctness of
the investing process has to be checked. The easiest way for doing this is
measuring the ‘gloss-off’ time. After checking this point, the burn-out
cycle has to be carried out in a correct, controlled manner as does the
subsequent casting procedure, especially if using centrifugal casting.
Static casting is not so critical.
The following schematic diagram shows the sequence of steps to be
followed to determine the causes and, hence, the measures necessary for
avoidance of the defect(s). Usually, quality of investment powder or/ and
the investing process are the decisive factors. The burn-out cycle depends
on the flask size, type and size of the furnace and the furnace filling.
Therefore, any recommendations can only provide initial guidance. For
safety, an additional hold at 450°C for 2 hours during burn-out should be
included. The investment manufacturers recommendations should be
followed.
Extended Explanation
The strength of the investment is, in the first place, determined by the quality of
the binder, gypsum (calcium sulphate). Partially dehydrated calcium sulphate
takes up water during the investing process and forms a network of (hydrated)
gypsum crystals which bind the refractory silica (quartz and cristobalite) grains
together. Investment powder can lose this binding property when it is stored in a
humid atmosphere or for a long time (premature uptake of humidity). In a
slurry with excess water, the gypsum needles cannot develop the bonding
network, with the consequence that the investment has a reduced strength.
During the burn-out process, gypsum looses a part of its water in a temperature
range 190- and 200°C. The bonding strength can only be maintained if this
process occurs slowly.
Decomposition of calcium sulphate also has to be avoided as this can also
give rise to sandy surfaces. When mixed with silica, this can occur at
29
1 temperatures as low as 750°C, especially when carbon is present (which can
arise from incomplete burn-out of wax absorbed on the mould wall). These
temperatures can be easily attained at the mould surface during pouring of
molten metal. Higher caratage alloys and palladium white golds have high
melting ranges.
Another example
In this case, there are at least three causes that have conspired to produce
the defective pendant:-
Fig. 7.4 Pendant in 14ct gold a) The investing procedure was incorrect. Instead of an anticipated
‘gloss-off’ time of 12 minutes, a time of only 6 minutes was
measured, probably due to an incorrect powder/water mixing ratio.
b) The burn-out procedure was wrong. The temperature-time cycle was
programmed correctly. However, the furnace contained too many
flasks. Some flasks were placed very close to the heating elements
and were heated too fast on this facing side, resulting in a weak
investment locally.
c) The burn-out cycle was interrupted during the night. The flasks
cooled down to about 100°C and were re-heated to working
temperature (350°C) the following day. Cooling down the flask below
approximately 250°C weakens the investment.
A bad, rough sandy surface and the formation of fins on the edges of
the flat pendant were the result of this mistreatment.
Further Reading
Part C: Basic Aspects, section 3.3, Influence of investment on casting
quality.
30
1
Testing gloss-off time
Check burn-out cycle

Time within limits Heating rate slow
(recommendations yes resp. holding time at
of supplier or former about 250°C, max.
experience) temp. below 750°C
no
Cycle within
Check mixing ratio limits?*
(by weight) and
water quality/
temperature
no yes
Ratio
corresponds to
yes recommendations,
no measuring
failure
Discharge
batch of
investment
no Reducing
Change maximum speed
burn-out cycle of rotation
(centrifugal
casting)**
Remarks:
Use correct * Usually more than 4hrs in the range
conditions up to 250°C, depending strongly on
flask size and furnace
** Pressure difference in static casting
has no significant influence
Trouble shooting for investing and burn-out process
31
1 CASE 8: ROUGH SURFACE AND ‘WATERMARKS’ - CAUSED BY INCORRECT
INVESTING PROCEDURE OR BAD QUALITY INVESTMENT
Similar to
Case 7
Key Words
Investment casting; investing process; surface defects; watermarks.
Fig. 8.1 approx. 20 x Raised traces are visible, particularly on flat surfaces. The orientation of
the traces on the item run approximately from top to bottom of the tree.
Visual Appearance
The examples are severe cases. Not only raised traces are formed but also a
complete destruction of the fine surface structure of the record has occurred.
Alloys: Independent of alloy composition.
Manufacturing Method: The cause is an incorrect investing process
(as explained later). The casting method has little influence.
The only property affected is the surface quality. Slight traces on plain,
Fig. 8.2 Watermarks on test piece (piece of
flat surfaces can be removed by polishing (or gentle grinding).
a record) approx. 10 x
More pronounced watermarks, especially on structured surfaces, are
irrecoverable and should be rejected.
Brief Explanation
After mixing, the investment slurry contains an excess of free water. When
the setting process commences, the water is taken up by the binder (partially
dehydrated gypsum). If the slurry is left motionless for several minutes before
setting starts, the water and solids can separate. Some of the separated water
can collect on the surface of the wax models and may run together, forming
the characteristic watermarks. The traces left by the water remain on the
mould wall after dewaxing and are eventually reproduced on the casting.
Flat pieces are more sensitive to this fault than small, sculptured items.
Such defects occur frequently when the slurry is made with too much water.

As a first requirement, the investment setting time has to be known. The
best way to determine this is by measuring the ‘gloss-off’ time. A standard
value is approximately 12 min. for most quality investments. However,
this value can vary with brand of the investment and working conditions.
The manufacturers recommendations should be followed.
The ‘working time’ is one or two minutes less than the gloss-off time.
During the working time, the slurry should kept in motion. The working
time is normally taken up by mixing, degassing, filling the flask,
degassing again and vibrating. Where vacuum investing machines are
used, a separate degassing is not necessary and filling the flask
automatically is fast. Therefore, the mixing procedure has to be extended
to keep the slurry in motion for a reasonable time to prevent separation.
32
Note: Mixing the investment slurry in vacuum investing equipment
1
with a good vacuum can also have an adverse effect. Water evaporates in
high vacuum very easily, the slurry can become too viscous and the
working time is reduced. This has a detrimental influence on surface
quality and even cracks in the investment can occur when the working
period is longer than the reduced working time.
Note: At high altitudes, less water than the recommended amount is
needed.
Further Reading
Part C: Basic Aspects, section 3.3, Influence of investment on casting quality
33
1 CASE 9: WRINKLED SURFACE - CAUSED BY FAST HEATING DURING
THE BURN-OUT PROCESS
Key Words
Investment casting; invest defect; burn-out process; surface defect.
The surface of cast items has a wrinkled, scarred appearance. The defect
occurs only in small number of flasks out of a furnace fill. Often it is
noticed only on pieces from one side of a flask.
Visual Appearance
Fig. 9.1 Wrinkled surface of 14ct yellow
The defect was restricted to some flasks of a ‘furnace fill’. Furthermore, it
gold ring 6 x
appears mainly on parts positioned on one side of the flask.
Alloys: The defect is independent of alloy composition.
Manufacturing Method: The burn-out furnace was overfilled with
flasks to the top. Some flasks stood very close to the furnace walls.
Not influenced
Microstructure
The microstructure is not affected and no porosity or pitting is observed.
The cross-section, fig 9.2, shows relatively flat surface irregularities.
Fig. 9.2 100 x
Surface quality
For a smooth surface, more grinding and polishing has to be applied.
Brief Explanation
The burn-out furnace was overfilled with flasks, some stood very close to
the heated walls of the furnace. The sides of these flasks facing the wall
heated up very quickly and so the surplus water in the investment could
not evaporate gently. Violent boiling occurred and the wax softened,
causing the surface of the mould cavities to deteriorate; these defects were
reproduced on the casting.

The burn-out furnace should not be filled to the top and close to the sides
with flasks. This prevents uniform heat distribution. Depending on the
furnace, the temperature within the chamber can vary widely (50 -75°C is
not untypical). Usually, near the front door (and the back wall) the
temperature remains low. Flasks in this position might not be heated
sufficiently. In contrast, the heated side walls are usually the hottest places in
the furnace. Flasks placed near them will heat up too fast. To avoid such
problems, flasks should be positioned at a safe distance.
The heating rate should be sufficiently slow, at least up to 400°C, to allow
water to evaporate away gently and allow phase transformations in the silica
refractory to be completed without causing cracks. A holding time period at
250°C is recommended. Investment manufacturer’s recommendations should
be followed.
Further Reading
Part C: Basic Aspect, section 3.4.3, Burn-out cycle
34
CASE 10: DENDRITIC SURFACE - DUE TO SHRINKAGE AND REACTION
1
WITH THE INVESTMENT
Key Words
Dendritic structure; investment casting; reaction with investment;
flask temperature.
The surface of heavy cross-sectioned items is rough with a typical
dendritic structure. This structure often appears on the central sprue.
Fig. 10.1 Surface of central sprue
VISUAL APPEARANCE
The central sprue shows a typical rough structure. Even with the naked
eye, the dendritic appearance can be recognised.
With higher magnification, the typical dendritic structure is better
resolved. Removing any remaining investment from the tree by sand
blasting hides the defect. Only a rough, irregular surface remains, with
pores visible. The real reason for the defect is hidden.
Alloys: Any yellow carat alloy can show this kind of defect but 14ct
alloys are especially sensitive to this surface structure.
Manufacturing Method: In general, the defect tends to appear on
investment casting under a protective atmosphere and with a relatively
Fig. 10.2
high casting and/or flask temperature.
In one actual case, a 21ct alloy was cast with a flask temperature of
720°C and a melt temperature of 1150°C under a protective atmosphere.
The reason for using such high temperatures was to obtain sufficient
form-filling. Vacuum (static) casting was used under conditions where
there was no great pressure difference.
In most cases, the mechanical properties are not significantly influenced,
despite shrinkage porosity and porosity caused by some impurities.
Microstructure
Fig. 10.3 The dendritic surface structure as
The microsection, Fig. 10.4, shows a rough, degraded surface. revealed in a scanning electron microscope
The structure itself is not affected and the casting shows no porosity. (SEM).
In most cases, the defect is restricted to a small surface zone, as in this

case. Sometimes, however, it will extend to include both shrinkage and
gas porosity. A dendritic surface should be seen as a warning that porosity
might have occurred too.
Other
The roughness of the surface is considerably increased, so extending the
grinding and polishing necessary to obtain a smooth, polished surface.
Brief Explanation
Growth of crystallites at the onset of solidification manifests itself as a
dendritic shape. The residual melt remains in the inter-dendritic space. Fig. 10.4
If the melt does not wet the investment at the mould wall and
decomposition of the investment gypsum binder causes the formation of
sulphur dioxide gas, the residual melt is pushed away from the surface,
35
1 leaving a skeleton of dendrites. Thus, the typical dendritic surface
structure appears. These conditions are preferentially fulfilled in a neutral,
protective atmosphere and with high casting and/or flask temperatures.
Usually, this defect is restricted to heavy cross-section parts or the
central sprue. Both introduce a great quantity of heat, increasing the
temperature on the surface of the investment and increasing the
likelihood of gypsum decomposition. A dendritic surface is less likely
when the casting is done in air. In this case, the wetting behaviour of the
melt is enhanced by the formation of oxides.

If thick-walled items show this effect, the casting temperature and/or
flask temperature should be reduced. However, the form-filling of any
filigee items on the same tree, for example, may become worse.
Improving form-filling by other methods (e.g. by increasing the pressure
difference in pressure or vacuum assisted casting) should be done to
compensate the negative effect of lower temperatures. Use of an alloy
with small zinc additions should also reduce the roughness caused by the
dendritic surface structure.
A dendritic structure on the central sprue is not necessarily a bad sign.
In the case of casting thin-walled items, it can be neglected. However,
if heavy parts are on the tree, a dendritic surface on the central sprue has
to be taken as a hint that casting conditions are not optimum.
Further reading
Part C: Basic Aspects, section 2.1, Melting and solidification and section
3.5, Behaviour of gypsum-bonded investment during casting.
36
1.4 INCLUSIONS
1
Inclusions can be divided into two classes:-
a) Non-metallic particles, which do not originate in the alloy.
Investment particles, eroded particles from worn crucibles, slags and
oxides are examples and are caused by improper casting procedures
and, therefore, are treated in the following chapter.
b) Metallic and non-metallic inclusions which originate from a
contaminated (polluted) alloy. These include segregations of
insoluble grain refiners but, for convenience, these are considered in
chapter 3 on ‘Defects caused by Alloy Composition’, even though
the defect may occur during investment casting.
CASE 11: PORES AND INCLUSIONS ON THE SURFACE - CAUSED BY

INVESTMENT
Key Words
Investment Casting; surface defect; inclusion; investment.

Scattered, relatively large pores and inclusions are visible on the surface of
a cast item. Usually, the defect is not visible until the jewellery is polished.
Sometimes, attempts to remove the defect by grinding and/or polishing
Fig. 11.1
may be successful. However, in other cases the pores will be enlarged by
this procedure.
Visual Appearance
The pores are randomly scattered all over the surface. The pores can be empty
or filled with obviously non-metallic material. Probably, an empty pore
originally was filled by an inclusion which has been removed by the
subsequent surface treatment, e.g. de-vesting, pickling, grinding and polishing.
Alloys: Independent of alloy composition.
Manufacturing Method: Investment casting in general, although
centrifugal casting is more susceptible to this defect as the force acting on
the investment is higher. More important factors are the quality of the
Fig. 11.2 40 x
investment, the investing procedure and the way of building up the tree
(see later).
The mechanical properties are not usually affected.
Microstructure
Many pores are visible on the surface, fig 11.2, and just beneath the
surface. Inclusions are not apparent; perhaps, the defects are too close to
the surface and any inclusion was removed by surface treatment or
sample preparation.
Figure 11.3 shows a pore at higher magnification. Only a small part of
Fig. 11.3 400 x
the defect volume reaches the surface. More polishing would enlarge the
defect.
37
1 An example from another defect case, fig 11.4, reveals an investment
inclusion of considerable size. Such examples, with the inclusion
completely retained after surface treatment and sample preparation, are
not very usual.
Brief Explanation
The incoming flow of liquid metal during pouring breaks off parts of the
investment out of the mould which then become entrained in the melt
as it fills the mould cavity. These eroded investment particles (mainly
Fig. 11.4
silica) are the inclusions. They are frequently found near the surface of a
jewellery item.
Reasons for this defect can be manifold: Firstly, a weak investment; the
strength of the investment will be low if a bad batch - a batch too old or
stored on a humid place - is used or if the powder/water ratio was wrong.
Secondly, improper setting up of the wax tree. Sharp edges, especially at
the connection of the wax model/gate to the central sprue, or elsewhere
on the tree or in the design of larger jewellery items, can break when they
are impacted by the melt as it is poured.
The danger of investment particles breaking off increases with increasing
velocity of melt flow, particularly where the flow is turbulent. Thus,
centrifugal casting tends to cause more of these defects than static casting.
Alternatively, inclusions may originate as eroded particles of graphite or
ceramic refractory from old, worn melting crucibles, from old investment
adhering to unclean recycled scrap or, simply, from airborne dusts.

Investment: The first step is the control of the investment quality and the
investing procedure. If the ‘gloss-off’ time becomes unusually long, the
danger for getting such (and other) defects increases considerably.
Provided that the mixing ratio is correct, a bad powder quality is the
cause. The powder originates from a bad supplied batch or the powder
was stored for too long a time and/or in a humid place. Investment must
be stored in an air-tight container.
Wax tree design: Make sure that wax model/gate-sprue joints are smooth
and profiled and that larger cast items do not have sharp edges which can
break off when impacted by the melt.
Casting: Do not use too high a speed or acceleration during centrifugal
casting.
Other: Do not use old worn melting crucibles or unclean recycled scrap.
The latter should be preferably remelted under a flux cover and grained.
Further information
Part C: Basic Aspects, section 3, Investment casting.
38
CASE 12: ‘CROWS FOOT’-LIKE SURFACE DEFECTS - CAUSED BY ZINC
1
OXIDE INCLUSIONS
Key Words
Investment casting; surface defects; porosity; inclusions; impurities.
Porosity is claimed. The pores appear predominantly after finishing and
polishing. In the case of low carat alloys, the area surrounding the pores
is often tarnished.
Fig. 12.1 Surface porosity caused by
impurities.
Visual Appearance
The example shows a 10ct alloy which contains many oxide inclusions.
At higher magnification, figs 12.2, 12.3, the typical form of such ‘crows
feet’-like surface porosity can be seen. Unfortunately, sometimes
shrinkage porosity has a very similar appearance. Identification requires
detailed information on casting conditions and examination of a
microsection taken at the site of the defect.
Alloys: Alloys with a higher zinc content. Generally, with few
exceptions, 8 -10 carat alloys have a high zinc content. However, even 14
and 18 carat alloys can contain a considerable, high zinc content.
Manufacturing Method: The defect occurs with any casting method Fig. 12.2 Typical surface porosity caused by
where there is not sufficient protection against oxidation. In the same oxide impurities in a high zinc-containing
alloy.
way, use of remelted contaminated material is critical.
In extreme cases, embrittlement can occur; the items break when a low
force is applied if the structure is infiltrated by a large quantity of oxides.
Microstructure
Typically, zinc oxide forms inclusions in the shape of membranes or
films. In microsections, the membranes are visible as filament-like traces,
fig 12.4. Usually a high magnification and good metallographic
preparation of the microsection are necessary to reveal the inclusions.
Where the inclusions reach the surface (usually after polishing the Fig. 12.3
jewellery items), pores in the shape of a ‘crows foot’ appear on the
surface.
Other
Low carat alloys may be tarnished around the porous area. Chemical
treatment, such as pickling, electrolytic stripping or cleaning, or under
the influence of human sweat or other aggressive media such as sprays,
the oxides will be etched away. Such aggressive media can remain
absorbed in these micro-cavities, causing local corrosion and tarnishing.
Brief Explanation
Zinc is a good deoxidiser in gold alloys and readily forms zinc oxide in
the presence of oxygen in preference to copper oxide formation, thus Fig. 12.4
producing brighter castings. In contrast to copper oxide, it is not possible
to remove zinc oxide by melting such contaminated alloy under reducing
conditions. Unlike copper oxide, zinc oxide does not tend to float up to
39
1 the surface of the melt but remains within the bulk of the melt. After
solidification, zinc oxide often forms thin membranes, situated mainly on
grain boundaries (and appearing as filament-like inclusions in
microsections). The result is that zinc oxide, once formed, persists in the
material, causing porosity and surface defects.
The usual sources for zinc oxide inclusions are: a) Melting under
oxidising conditions. Where melting is not done in closed chamber,
shielding of the melt with a protective atmosphere is important. Purging
an open crucible with a flow of argon or nitrogen may not be sufficient.
b) Remelting contaminated scrap material. This is generally thought to be
the main cause for the defect.

Avoid melting without a protective gas atmosphere. At the very least, the
crucible should have a protective cover of gas (preferably ‘forming’ gas,
blown on the surface of the melt). Where melting is done in a closed
chamber, a clean oxygen-free neutral atmosphere such as nitrogen or
argon is recommended
Scrap material used for remelting has to be clean. In contrast to copper
oxide, zinc oxide, once formed, is very difficult to remove by melting
procedures.
40
CASE 13: SURFACE ‘CAULIFLOWER’ DEFECTS - CAUSED BY SLAG
1
INCLUSIONS.
Key Words
Investment casting; surface defect; inclusions; slag.

The cast item has a partly very rough surface with a cauliflower-like
structure. Some inclusions are visible on surface.
Fig. 13.1 8 x
Visual Appearance
The surface has a cauliflower-like structure caused by liquid slag or flux
introduced into the flask with the melt. The slag collects on the surface and
is mostly removed on removing the investment and cleaning the casting.
Alloys: This actual case is an 18ct yellow gold; however, the defect is
independent of alloy composition.
Manufacturing Method: Investment casting by the vacuum assisted
method, with bottom pouring from a ceramic crucible.
The mechanical properties are not usually affected.
Fig.13.2
Microstructure
The alloy microstructure is not influenced by the defect. Inclusions near
the surface can be frequently observed. In this actual case, different kinds
of inclusions were found: Some had relatively sharp edges, fig. 13.3,
which were identified as pure silica. Other particles had smooth edges
and an apparently complicated structure, fig. 13.4, which analysis showed
contained zinc, silicon and a small amount of nickel.
The sharp-edged inclusions, fig 13.3, were identified as silica particles.
The relatively coarse size indicates that the inclusion did not come from
the investment.
Fig 13.5 shows what is probably a slag particle with a maximum
diameter of approximately 1mm. It can easily be detected and Fig. 13.3 SEM-micrograph
characterised with a magnifying glass as an irregular structure with
rounded edges.
Other
It is not possible to obtain a bright surface by polishing. Inclusions tend
to break out causing scratches.
Brief Explanation
The cause of the defect are external impurities introduced into the melt. At
least some of the impurities were liquid (slag) at the moment of casting and
are swept into the flask where some float to the surface of a cast item while
the metal was still liquid. After solidification, this typical cauliflower-like
structure of the metal surface remains, while the slag particles themselves Fig. 13.4 Inclusion of slag with smooth
are removed during the cleaning of the casting. However, not all the slag edges, suggesting that the particle was
probably liquid at the moment of casting
(or inclusions) collect on the surface. Some of it forms inclusions close to
the surface. The shape of the inclusions and their analysis by SEM-EDX can
provide reliable information on the origin of defect.
41
1 The roughly spherical shape of the particles points to their origin as
slag. It consists mainly of zinc silicate with a small amount of nickel
oxide or silicate. These compounds can accumulate in the melt crucible if
dirty scrap material, contaminated with some used investment, is
remelted as part of a new melt. Boric acid, used as flux or for
‘conditioning’ the crucible, can also be involved. It decreases the melting
range of the slag and increases the fluidity.
The fairly sharp-edged inclusion was introduced to the melt in the
solid state. Size and shape exclude investment powder as a probable
Fig. 13.5 Another example of a silicate source. It could be a refractory particle from the crucible or possibly ‘dirt’
inclusion. introduced from outside.
In summary, the evidence points to either a dirty, worn crucible with
an excess of flux (used to ‘condition’ the ceramic crucible) or some ‘dirt’
from the bench as the cause of the defects.

Use only clean crucibles which are not worn out. Avoid too great an
excess of flux, both in ‘conditioning’ the crucible or simply adding flux
to the melt. In principle, the quantity of flux - if ever used - should be
kept as small as possible.
42
CASE 14: POROSITY - CAUSED BY INCLUSIONS OF SLAG COMBINED
1
WITH GAS PORES
Key Words
Investment casting; surface porosity; impurities.
The surface shows a great number of pores.
VISUAL APPEARANCE
Fig. 14.1 A spring hook with surface
The surface is littered with irregular pores, fig 14.1.
porosity
Sometimes, shallow, flat hollows occur on a porous surface when slags
are involved in the defect, fig 14.2.
Alloys: The example is an 8ct yellow gold alloy, although the defect can
appear with other alloys, too. However, low carat alloys are more critical.
Manufacturing Method: Investment casting; the most important
fact is that a great quantity of flux was used to prevent oxidation and
collect any oxides, especially zinc oxide.
Fractures can be caused by excessive porosity in smaller cross-sectioned
Fig. 14.2 Surface porosity on a 14ct yellow
regions. gold casting
Microstructure
Irregular shaped large pores with dendritic edges are one of the
characteristics of this kind of defect, fig 14.3. The pores are filled in part
with a glassy substance which can be identified as slag.
This example, fig 14.4, shows large pores in a cast 14ct alloy. The thin
lines between pores are a characteristic feature, caused by thin oxide
membranes.
Brief Explanation
Sometimes material is melted and remelted in air (often in a worn-out
clay-graphite crucible). To remove any remaining old investment and
Fig. 14.3 200 x
oxides and to prevent further oxidation during melting, a flux (e.g. borax)
is added in considerable quantity. The flux forms a viscous surface slag
and may accumulate with increasing number of castings performed with
the same crucible, if the crucible is not sufficiently cleaned after each
casting. Some of the slag containing oxides are poured with the alloy
melt into the flask and form inclusions in the casting. Zinc oxide will
form thin membranes and copper oxide is suspected of inducing severe
gas porosity.
Even if the reaction mechanism which leads to such pronounced
porosity is not completely understood, the appearance of this type of
defect is a positive hint that it is caused by a considerable amount of
impurities. In most cases, the defect occurs with zinc-containing alloys. Fig. 14.4 50 x
Zinc oxide is readily formed and is not easily removed from the melt. Too
much flux can exacerbate the defect considerably.
43
There are only a few, effective rules for avoiding this defect:
a) Only clean material should be melted.
b) The portion of used material should be restricted to about a third of
the total quantity.
c) Any remaining used investment and other impurities must be
completely removed.
d) Melting should be performed in a protective atmosphere, making
addition of flux unnecessary.
e) Use of flux should be avoided in investment casting or, at least,
reduced to a minimum. Any remaining flux and slag in the crucible
should be removed after each casting, preferably whilst the crucible is
still hot and the slag/flux is still pasty.
f) Avoid using an alloy with a high zinc concentration.
44
2
2 INGOT CASTING AND
CONTINUOUS CASTING
Relatively few cases of defects have been reported that are attributable to
these production methods. One reason might be that defects in a cast bar,
rod or tube are only revealed many production steps later. Tracking back
their causes to the origin is difficult, if not impossible.
2.1 CRACKS AND MATERIAL SEPARATION

Fig. 15.1 Surface of the ring approx 20 x
CASE 15: CRACKS ON THE SURFACE OF A RING - CAUSED BY A BAD
SURFACE ON A CONTINUOUSLY CAST TUBE
Key Words
Continuous casting; surface defects; over-stressing; cracks.
Small cracks are visible on the surface of a ring and are repeated.
Apparently, the cracks are restricted to the surface.
Visual Appearance
The surface of the ring, fig 15.1, is split open in many cracks, some
Fig. 15.2 SEM image of the surface
small. Despite the cracks, the ring has not broken.
The SEM photograph, fig 15.2, shows the cracks more clearly: the surface
has a scaly structure. Apparently the scales or other surface defects have
broken open.
Alloys: The defect appeared in a 14ct yellow gold containing 6% zinc.
Manufacturing Method: A continuously cast tube was put through a
pilger mill and further reduced in diameter and wall thickness by
drawing. Small ring blanks were cut off for ring manufacture.
Whilst fracture of the rings was not reported, it could occur if a ring is
Fig. 15.3 250 x
enlarged by stretching.
Microstructure
Both figures 15.3 and 15.4 show incipient cracks. The cracks are not very
deep and have the typical structure of laminations or laps. There is no
sign of embrittlement. Impurities, e.g. segregations on grain boundaries
or oxides, cannot be detected.
Surface Quality
The defect cannot be removed by polishing as the cracks are too pronounced.
Brief Explanation
The cause of the defect is probably a bad surface on the tube produced
during continuous casting. The rough, scaly (laminated) surface of the
Fig. 15.4 250 x
cast tube will tend to be levelled when passed through a pilger mill,
although the laminations are hidden beneath the new smooth surface.
Subsequent drawing lengthens the laminations.
Subsequent rolling of the ring blank deforms the material in a radial
45
2 direction. This deformation with the applied radial tensile stress causes
the laminations to break up, resulting in cracks and flakes visible on the
surface. Overstressing the material during rolling increases the defect
considerably. Even small defects have a notch effect, causing cracks when
the material is towards the end of its workability, i.e. in a severely work
hardened condition. In the present case, the hardness was HV 260, a high
value indicating that the material was highly strained by deformation.
Therefore, surface defects produced on continuously cast tube,
followed by heavy deformation without any intermediate soft annealing,
can be blamed for the occurrence of the defect on the rings.

In the first place, a good smooth surface should be achieved on the
continuously cast tube. The melt should be clean and free from oxides and
slag. A protective atmosphere will help to prevent oxidation and avoid the
need for too much flux. The advance of the emerging tube during
continuous casting and the cooling conditions should be perfectly
matched to ensure a good surface. Any surface defects should be removed,
e.g. by grinding, from the continuous cast tube before advancing to other
production steps to avoid the occurrence of defects on the rings.
Furthermore, overstraining (excessive work hardening) must be
avoided by an intermediate soft annealing of the material. Annealing is
recommended after pilger milling the tube and before ring rolling.
Further Reading
See Part C: Basic Aspects, section 2.2, Deformation and mechanical
properties and section 2.3.1, Soft annealing.
46
CASE 16: LAMINATIONS ON A RING
2
Key Words
Ingot casting; laminations; sink hole; primary pipe; shrinkage.
Wedding rings show lamination or crack-like separations of the material
on the side edges.
Fig. 16.1
Visual Appearance
The ring shows a crack-like defect on the side, fig 16.1. Several rings from
the same batch have the similar defects. The length of cracks is variable;
grinding or polishing does not remove them.
At higher magnification, fig 16.2, the material separation (lamination)
is more obvious.
Alloy: The defect appeared in a 14ct yellow gold with 6% zinc, but is
independent of alloy composition.
Manufacturing Method: The ring was punched from a rolled bar.
Mechanical properties and surface quality
Fig 16.2 50 x
The material lamination is permanent and can lead to complete
separation of the ring into two pieces if it is further deformed to widen it.
Microstructure
The material separation is obvious, fig 16.3. Non-metallic inclusions
(oxides, slag) are often seen accumulated on the parting plane.
Brief Explanation
On casting a bar vertically, a deep sink hole - or primary pipe - is formed on
the top surface. In extreme cases, this sink hole can extend to a small and
deep axial porosity - a secondary pipe. These pipes should be removed (by
‘cropping’ the ingot) prior to rolling the bar down to a smaller section.
Otherwise, they can lead to a material separation in the rolled material.
Fig. 16.3 200 x
Even a small remnant of the pipe is extended to a considerable length
during rolling and the defect remains undetected until final processing,
e.g. by cutting or bending, deep drawing etc. Also, after rolling to plate or
sheet, the material may split or delaminate into separate sheets.
The present case shows these separations on the cut edge of the
punched wedding rings.

The only way to avoid this defect is by removing the pipe from the cast
bar by cropping prior to rolling. Hidden axial porosity - a secondary pipe -
might exist, which is not easily removed because of its depth, even if the
primary pipe is small or virtually absent. To obtain a flat sink hole or
pipe, the melt and mould temperature should be adjusted. Usually, an
increased mould temperature will reduce the depth of the pipe.
Further Reading
Part C: Basic Aspects, section 3.1, Shrinkage porosity.
47
3
3 DEFECTS CAUSED BY ALLOY
COMPOSITION
In this context, defects are considered that owe their origin primarily to
the alloy composition, i.e. the consequences of deliberate alloying
additions, and also to impurities insofar as they cannot be considered as
non-metallic inclusions.
3.1 LOW MELTING IMPURITIES

Fig. 17.1
CASE 17: CRACKS - CAUSED BY SEGREGATION OF LOW MELTING
SILICON COMPOUNDS.
Key Words
Investment casting; cracks; low melting compounds, silicon.

A ring fractured during de-investing (or shortly afterwards) without being
subjected to any significant strain or deformation. The fracture follows
either a straight line or a jagged trace like a lightning path.
This and other very similar defects are now being reported more
frequently, a consequence of the fact that silicon additions are now used
Fig. 17.2
more often to improve casting properties.
Visual Appearance
A 14ct yellow ring fractured during channel setting, the cracks occurring
preferentially in the surrounding region of the hammered area. The ring
has broken in a straight line. More jagged cracks like lightning paths are
visible. Cracks of this shape invariably indicate that the material is brittle,
i.e. it breaks without any visible deformation. In most cases, this happens
with only a small force applied or even without any external stress.
In general, there are two reasons for the occurrence of such behaviour:
the presence of either low melting alloying components or similar
impurities. (Also, a special kind of corrosion causes a similar brittle defect Fig. 17.3
in low carat alloys - stress corrosion cracking)
The figures 17.3 and 17.4 show another example. This item was cast in
18ct yellow gold with an addition of a silicon-containing ‘grain refiner’.
The castings were found to be already broken after removing the
investment.
The brittle nature of the fracture is obvious.
Alloys: The defect is not strictly restricted to a particular caratage or
fineness level, but the presence of some alloying elements or impurities is
essential. Alloys of higher caratage and alloys with higher silver content
are more sensitive to this kind of defect than lower caratage alloys with
high copper content.
Manufacturing Method: The defect is not primarily a casting defect; Fig. 17.4 SEM of fracture surface
the composition of the alloy is the decisive factor. However, it occurs
almost exclusively with some specially formulated casting alloys.
49
3 Influence on Properties
The material breaks easily without suffering any visible amount of
deformation. A characteristic of this type of embrittlement is that it happens
even if the alloy hardness is very low. In this case, the hardness was only HV
123, i.e. at the lower end of possible values for 14ct yellow gold.
Microstructure
At low magnification, only the path of the crack, along the grain
boundaries, is visible whereas, at higher magnification (500x), greyish
Fig. 17.5 Inclusions in a broken 14ct inclusions are revealed on the grain boundaries.
yellow gold sample The grey inclusions, fig 17.5, were identified as a silicon-compound.
Thin segregations of this compound along grain boundaries can cause
embrittlement and cracks.
Figures 17.6 and 17.7 show the microstructure of a silicon-containing
18ct alloy. Two phenomena can be seen: the grain coarsening effect of
silicon additions and very tiny segregations of a low melting phase on
grain boundaries connected to cracks and crack-like pores.
Brief Explanation
Silicon additions, which cause the brittle behaviour, are used in 14ct
yellow golds for preference to improve the castability of the alloy.
Fig. 17.6 However, silicon at too high a concentration forms a low melting
component (‘eutectic’) which collects on the grain boundaries during
solidification. This creates a brittle film on the grain boundaries which
completely destroys the material’s properties. Cracks can occur due to
internal stresses generated during solidification or on cooling of the
casting as well as later on further processing (polishing, setting etc).
Silicon has a negligible solubility both in gold and in silver. A low melting
compound (eutectic) is formed. The solidus temperature in the binary Au-Si
system is 363°C, and 830°C in the Ag-Si system. The low melting phase is
brittle and collects on grain boundaries during solidification and weakens the
Fig. 17.7 material. The copper content of yellow gold enhances the solubility of silicon.
For this reason, a small amount of silicon can be tolerated for improving
castability.
The silicon concentration which can be used without causing problems
depends strongly on the ratio Au/Ag/Cu and decreases with increasing (gold +
silver) content. Thus, high carat golds are more susceptible to embrittlement
than low carat alloys and, for a given caratage, silver-rich (pale) yellow gold
will tolerate less silicon than a copper rich (pink) alloy.
In addition, silicon can cause a pronounced coarse grained structure which is
not desirable. This multiplies the embrittling effect of even very small
concentrations of silicon and other low melting compounds (see also figs. 17.8 -
17.10 below). Because of this fact, the critical level of a deleterious addition
depends not only on the type of alloy but also on its microstructure.
Similar critical embrittling components include sulphur, lead, bismuth and
phosphorus.
50
3
Silicon in yellow gold is a critical addition. To obtain the benefits of this
element in the casting of carat golds without the danger of embrittlement
and cracking, the concentration has to be strictly limited, according to
the composition of the alloy. For example, a 14ct yellow alloy should not
contain more than 0.1% silicon; an 18ct alloy can only tolerate 0.05%
silicon at the most. Higher carat golds should not use silicon additions.
However, the silver to copper ratio and any heat treatment has also to be
considered.
Fig. 17.8 A pendant, broken after de-
Another example vesting
In fig 17.8, the alloy is a 14ct white gold with 0.04% silicon and 0.001%
boron additions. Silicon is considered as a ‘grain refiner’.
The microstructure of this sample, figs 17.9 and 17.10, shows very thin
traces of segregations on the boundaries of (large) grains. A very small
concentration of a low melting compound is sufficient to form a brittle
film on the grain boundaries in this coarse grained structure. The grains
are effectively insulated from each other by this brittle substance. The
detrimental effect of these compounds would be not so prominent in a
fine grain structure (where the grain boundary area is much larger). The
volume of the compound would probably not be sufficient to cover the
increased surface area of the grains.
Fig. 17.9 100 x
Further Reading
Part C: Basic Aspects, section 2.4, Low melting components
Fig. 17.10 200 x
51
3 CASE 18: BRITTLE FRACTURE - CAUSED BY LEAD AS AN IMPURITY
Key Words
Embrittlement, fracture; lead.
A piece of wire was completely brittle and fractured.
Visual Appearance
Alloy: 18ct yellow gold with a small impurity of lead (0.02%)
Fig. 18.1
Manufacturing Method: The wire was manufactured in the usual
manner without any difficulty. However, it was finally annealed at about
300°C. In this state the embrittlement occurs.
The wire has broken with only a small deformation. It cannot withstand
any strain; it is brittle.
Microstructure
The intergranular fracture (‘grainy’ structure), fig 18.2, is typical of a
brittle fracture caused by impurities. The impurity, lead, segregated to the
surface of the grains and degraded the bond between the grains (because
Fig.18.2 SEM: fracture surface of a lead-
containing 18ct alloy the segregation itself has little strength) and resulting in the alloy
fracturing along the weakened grain boundaries.
Brief Explanation
Lead is a dangerous impurity in gold alloys. Its solubility and, therefore,
its critical concentration is very dependent on the composition of the
alloy (components: gold, copper and silver). The higher the gold and
silver concentration, the lower is the tolerable lead concentration.
Furthermore, the detrimental effect of lead is also dependent on the alloy
heat treatment. Annealing between 300 and 400°C is the most critical
treatment. In this condition, even 0.02% lead can cause complete
embrittlement of an alloy. This is well demonstrated in this case.
The alloy was ductile prior to this critical heat treatment, otherwise it
would not have been possible to draw a thin wire. Only after annealing
did the material become very brittle.

The only way to avoid the embrittlement is to use lead-free material.
Sources for lead impurity include: a) recycling of old or scrap material
repaired with soft solder (tin-lead alloy), b) the use of lead-containing
(‘free-machining’) brass as a master alloy for zinc additions to the gold
alloy (use 70/30 brass as brass with less 67% copper can contain some
lead; brass with less than 60% Cu usually has more than 2% lead) and
c) use of a lead support for working on jewellery items. Remnants of lead
will diffuse into the material on subsequent annealing.
Further Reading
Part C: Basic Aspects, section 2.4, Low melting components
52
CASE 19: CRACKS - CAUSED BY SULPHIDE INCLUSIONS IN AN
3
INVESTMENT CAST ITEM
Key Words
Investment casting; reaction with investment; cracks; sulphides; low
melting components.

Rings cracked on one or both sides of the shank. The cracks were already
present at the time of de-investing (removal of the investment).
Fig. 19.1 A broken shank of an 18ct
yellow gold ring
Visual Appearance
The fracture occurred without any deformation. The fracture surface and
the crack paths are jagged. The ring has fractured either during cooling
after casting or whilst de-investing the tree. The alloy is a simple gold-
silver-copper alloy and is not expected to be a problem. Only impurities
can be the cause of such a fracture like this. In this actual case, sulphide
(generated by reaction with investment) was found to be responsible.
Alloy: Conventional gold-silver-copper 18ct yellow gold.
Manufacturing Method: The investment casting was done by
induction melting in a ceramic crucible and casting by the vacuum
assisted method with bottom pouring. Forming gas (75% nitrogen, 25%
Fig. 19.2
hydrogen) was used as the protective cover.
The strength of the material is almost zero; the rings can be broken by
hand with little effort..
Microstructure
The microsection reveals some interesting details. Numerous cracks and
many gas pores are visible at low magnification. The ‘cracks’ are, in part,
gaps on grain boundaries, fig.19.2. This becomes more clear at higher
magnification in the SEM-micrograph, fig. 19.3. Another important detail
is a small sulphide particle (analysed by EDX-microanalysis) in a Fig. 19.3
‘boundary gap’ (marked by a cross in fig. 19.3). A detailed search detected
more tiny segregations of a grey compound, identified as sulphides, on
grain boundaries.
Brief Explanation
The main cause of the defect was the formation of gas pores and
sulphides at grain boundaries due to reaction of molten alloy with the
investment. Probably, remelting of already contaminated, sulphide-
containing scrap material played an active part in this case. The strong
reducing atmosphere (protective gas with high hydrogen content) aided
the reaction of melt with the investment to produce sulphides.
Gypsum (calcium sulphate), an important constituent of investment, can react
with the melt to form both sulphur dioxide and sulphides of copper and silver.
The sulphur dioxide dissolves and manifests itself as characteristic spherical
53
3 pores in the solidified metal. The sulphides constitute low melting components
and collect on grain boundaries (together with gas pores). The alloy loses its
strength completely and falls apart when only a very small load is applied.
Remelting of material already poisoned with sulphides or with adhering
investment will amplify the detrimental effect of investment reaction and
should be avoided.

For melting and casting in a closed system, the rule is: a) Never melt and
cast gold alloys in a strongly reducing atmosphere like forming gas.
Strong reaction between the melt and the gypsum-bonded investment
will occur. Neutral gas like nitrogen or (more expensive) argon is
recommended. A strong reducing atmosphere will do more harm than
good. [Note: Forming gas can be used if the material is melted and cast in
an open system (in air) for protecting the crucible and the melt from
oxidation.]. b) Remelting of sulphur-contaminated material as well as
scrap alloy with adhering remnants of old investment must be avoided.
Further Reading
Part C: Basic Aspects, section 3.2, Gas porosity and section 3.5, Behaviour
of gypsum-bonded investment during casting.
54
3
3.2 INCLUSIONS
CASE 20: HARD SPOTS - CAUSED BY INCLUSIONS OF ‘GRAIN
REFINERS’ AND GOLD IMPURITIES.
Key Words
Hard spots; inclusions; iridium; impurities;

Hard spots on the surface were detected whilst polishing the item. Visual
inspection disclosed hard, brownish particles. Fig. 20.1 14ct yellow gold with iridium as
grain refiner.
Visual Appearance
Figures 20.1–20.3 show typical inclusions in different alloys that resulted
in hard spots. They can appear as single particles of significant size or as
an agglomeration (nest) of smaller particles. Fig 20.2 is a surface of a
rolled sheet, etched with a cyanide-containing solution.
Alloy: Any alloy with iridium (or other platinum group metal) used as a
grain refiner, independent of caratage, or alloys made from impure fine gold.
Description of Manufacturing Method: The examples were found
in both investment castings and in rolled material. The defect is
independent of the manufacturing method.
Fig 20.2 22ct yellow gold with 2% iridium
Influence on Properties 50 x
Microstructure
The solubility of iridium in gold alloys is very low, even in the liquid state.
The addition is not homogeneously distributed. Nest of hard iridium
particles are formed, fig 20.4. They cause hard spots when they reach the
surface or they produce cracks on deformation due stress concentration.
Workability and mechanical properties
An excess of grain refiner and, therefore, the formation of nests of particles
decreases the workability of the alloy. The maximum degree of deformation
is reduced to about 60-65% (depending on alloy and amount of addition),
whereas an addition-free alloy can be worked to 70% or more reduction.
Wire drawing will be seriously affected by both single particles and Fig. 20.3 14ct yellow gold made from
contaminated gold
nests of inclusions. Even very small particles will cause breaks on drawing
a small diameter wire.
Polishing
A flawless surface cannot be achieved by polishing. In most cases,
improper (excessive) use of grain refiners or the presence of similar kinds
of impurities will be noticed by difficulties at polishing. Hard spots with a
characteristic ‘comet tail’ appearance occur.
Brief Explanation
High melting point metals, such as some platinum group metals, with
only a limited solubility in the solid gold alloy are sometimes used in very
small amounts as grain refiners. Such small, finely dispersed particles act Fig. 20.4 Nests of iridium inclusions in a
21ct alloy
as nuclei in the liquid metal during solidification, hence producing a fine
grained structure which has superior mechanical properties, useful, for
example, for polishing or deep drawing. Other additions such as cobalt act
to refine the grain size during annealing of cold worked material.
55
3 However, the addition of an improper compound or too high a
concentration results in clusters of hard particles. Instead of improving
the polishing procedure, a detrimental effect is caused.
Critical additions are iridium, ruthenium and - with a higher
concentration - cobalt.
Sometimes, the same detrimental effect can occur when carat golds are
made from fine gold ‘good delivery’ bars of 99.5% fineness. Some bars
may contain small amounts of metallic compounds of, typically, iridium,
ruthenium, tungsten and molybdenum. Such deleterious impurities can
only be removed by electrolytic refining to 99.99% purity.
Heavy insoluble metals such as iridium tend to sink in the molten gold
alloy under slow cooling conditions and agglomerate at the lower surface
of the ingot, hence accentuating the hardspot problem at the surface of
the finished jewellery.
RECOMMENDATION FOR AVOIDANCE

In the first instance, the effectiveness of a particular grain refiner in a
particular alloy under your production conditions has to be determined.
It makes little sense to add a grain refining element at will to any alloy
just because it is known to have a grain refining in other gold alloys.
If any such grain refiners are used, the concentration has to be limited to
about 0.01%. Melting needs some additional superheat and melting time
should extended in comparison with addition-free alloys. The grain
refiner has to be added as a master alloy, usually with copper.
As a general rule, if these conditions cannot be fulfilled with certainty,
it is better to avoid the use of a grain refiner. The ensuing difficulties will
outweigh the advantage of a grain refined alloy.
A grain refining effect in casting needs the segregation of extremely fine particles at
the beginning of solidification which act as nuclei for the formation (and
subsequent growth) of crystallites. An abundance of nuclei means a great number
of growing crystallites and, consequently, a finer (smaller) grained structure.
The conditions for such a nucleation process are: a) The grain refiner should
only have a very limited solubility in the solid alloy, but some solubility in the
liquid melt, b) The grain refiner (or a compound formed from it) should have a
relatively high solidus temperature and solidify before the main gold alloy.
The condition for a successful application of such a grain refiner is a complete
dissolution in the melt prior to casting. The melting temperature has to be high
enough and some time given for dissolution of the grain refiner. The most common
additions used as ‘grain refiners’ in jewellery alloys are ruthenium and iridium.
Both additions are difficult to use and should only be done using suitable master
alloys. The grain refining effect of iridium can be variable.
In most defect cases involving hard spots, the ‘grain refiner’ was present in
clusters of particles and very inhomogeneously distributed. In this state, no
significant grain refining effect can be expected, but difficulties with hard spots
on polishing will occur.
Further Reading
Part C: Basic Aspects, section 2.5, Grain refiners.
56
4
4 C O R R O S I O N , TA R N I S H I N G ,
D I S C O L O U R AT I O N
4.1 DISCOLOURATION
CASE 21: TARNISHING - CAUSED BY THE LINING MATERIAL OF A
STORAGE BOX.
Key Words Fig. 21.1
Tarnishing; sulphide.
Jewellery items stored in boxes showed discolouration. The tarnishing
ranges from reddish brown to black. The intensity of discolouration is
obviously influenced by the intensity of the contact which the jewellery
had with the lining material of the box
Visual Appearance
A 10ct yellow gold ring tarnished while stored in a box containing
sulphidic compounds. The colour varies between somewhat reddish to
black, due to formation of silver or copper sulphide.
Alloy: 10 ct yellow gold
Description of Manufacturing Method: Tarnishing is independent
of the manufacturing method with the exception of surface treatment by
electroplating with high carat gold.
Only the brightness and the colour of the jewellery surface is affected.
Brief Explanation
Silver and copper react very easily with many sulphur compounds, with
formation of sulphides. Very thin layers of sulphides can show a range of
colour between yellow to reddish or brown and finally black when the layer
reaches sufficient thickness.
Pure gold and high carat alloys (e.g. 18ct and higher) don’t react with
sulphur compounds and are, therefore, tarnish resistant. Usually, 14ct alloys
also resist tarnishing unless subjected to a heavy attack of hydrogen sulphide.
Low carat alloys (10ct and lower) show similar behaviour to silver or
copper and tarnishing is a naturally occurring phenomenon. Not only
sulphur compounds can cause tarnishing; other compounds discolour the
surface by oxidation.

The best way to avoid tarnishing is through use of higher carat gold
alloys. Otherwise, a practical way to prevent low carat gold from
tarnishing is by electroplating the pieces with fine gold or high carat gold
with a sufficient thickness. This is not a permanent solution, however, as
the electroplated layer will be eventually worn away. On storage,
tarnishing can be prevented by keeping the jewellery pieces in a dry
environment and avoiding contact with sulphur-containing boxes.
57
4 CASE 22: DISCOLOURATION OF YELLOW GOLD RING -
BY ACCIDENTAL AMALGAMATION WITH MERCURY.
Key Words
Discolouration; amalgamation; mercury.

The surface of a yellow gold jewellery item is discoloured. The colour
varies between brownish and dirty white or grey.
A customer complained about the defect after wearing the jewellery for
Fig. 22.1 A gold chain with discolouration.
some time. No details have been provided about the circumstances.
Investigation, using SEM-EDX analysis, showed that mercury was present
and the cause for the discolouration.
Visual Appearance
The surface should be yellow, but it’s colour has changed to a dirty white
and brownish hue.
Alloys: All gold jewellery alloys will be sensitive to this defect.
Manufacturing Method: Manufacturing method is not important.
The defect is typically produced by the customer.
Fig. 22.2 This example shows the stained
surface in more detail Mechanical Properties
Strength is often degraded. Sometimes, mercury can initiate a stress
corrosion fracture.
Microstructure
The dark spotted edge in the micro-section, fig 22.3, indicates that
mercury has penetrated into the gold alloy and covers almost the entire
cross-section. Sometimes, preferential segregation of the mercury
amalgam on grain boundaries can be detected, too.
Others: Colour
The main cause for rejection by the customer was discolouration. There is
a broad range of colours. The colour may vary between silver white/grey
Fig. 22.3
and brownish.
Brief Explanation
The discolouration is caused by uptake of mercury which is liquid and
readily forms an alloy with gold (gold amalgam) even at room
temperature. At a low gold concentration, the amalgam is liquid at room
temperature. At higher gold concentrations, it is solid. When gold or gold
alloys come into contact with mercury, a thin layer of amalgam is formed
on the surface causing a silver white or grey appearance. Thus, yellow gold
turns into a dirty ‘white gold’. Where only a small amount of mercury is
available, only a slight, somewhat brownish discolouration takes place.
Mercury can penetrate into the material preferentially along grain
Fig. 22.4 boundaries, decreasing its strength significantly.
The growth of amalgam crystallites on the surface of gold can be easily
observed in the scanning electron microscope (SEM), fig. 22.4
The SEM photograph shows amalgam crystallites grown on the surface
of 18ct yellow gold jewellery. How the jewellery came into contact with
58
mercury is not known; a broken thermometer is often the source.
4
This defect is not a problem of the manufacturer. It can only occur when
the jewellery comes into contact with mercury, normally an accident by
the customer. The source for mercury might be a broken thermometer or
mercury used in dentistry (dental amalgams).
Recovery of an amalgam-covered piece of jewellery is sometimes
possible by gently heating it up, preferentially under a vacuum, until the
mercury is evaporated off [safety note: ventilate well and extract fumes
via a fume hood; mercury vapour is very poisonous]. However, in the case
of jewellery alloys, a dull or even brownish surface can remain which has
to be polished off.
In the present case, with a deep penetration of mercury into the
material, successful recovery is unlikely. The item should be sent for
refining.
59
4 4.2 STRESS CORROSION AND INTERGRANULAR
CORROSION
CASE 23: CRACKS - CAUSED BY INTERGRANULAR CORROSION IN A
SILICON-CONTAINING 14CT ALLOY.
Key Words
Investment casting; cracks; low melting components; silicon,
intergranular corrosion.
Fig. 23.1 Fractured 14ct gold ring Description of the Defect

A 14ct cast ring is broken in the stone setting area. The fracture is brittle.
The defect was reported by a customer after approximately 7 month service.
Visual Appearance
One of the prongs is broken without any deformation (brittle fracture).
The fracture is situated at a place where two relevant factors occur
together, i.e. stress is applied and cleaning is difficult.
At higher magnification, fig 23.2, the fracture surface is intergranular,
i.e. shows a grainy structure, indicating that fracture took place along
grain boundaries. On the left edge of the figure, a dark coating is visible
which consists of silicon and some remnants of a medium (chloride,
Fig. 23.2
potassium, calcium) which is presumed to have had a corrosive effect.
Alloys: 14ct yellow gold with additions of silicon, boron and iridium.
The defect might also occur on alloys with even higher gold content if
the quantity of silicon addition lies in a critical range or other low
melting components are present. All low carat golds with such additions
are susceptible.
Manufacturing Method: Investment casting with vacuum assisted
static machine. However, the defect is presumed to be independent of
manufacturing method.
Fig. 23.3
The material has lost its strength. Low stress applied to areas, e.g. by
stone setting, causes cracks and fracture. The defect needs time to develop
and the influence of a corrosive medium.
Microstructure
A cross-section through the fractured prong clearly discloses the nature of
the fracture, fig 23.3. The alloy has cleaved along grain boundaries and
the material has fallen apart without any deformation.
This appearance is typical for two special kinds of corrosion: stress
corrosion and intergranular corrosion.
Brief Explanation
The influence of low melting components (silicon phase) and of
intergranular corrosion lead to embrittlement. The fracture has occurred
at a place where cleaning is difficult and impurities will tend to collect.
Usually, 14ct yellow gold is fairly resistant to corrosion from sweat,
cleaning agents, soap, salt water, etc. However, if additions cause
60
segregations of less noble compounds on grain boundaries, even relatively
4
weakly corrosive agents can attack these compounds preferentially. For
disintegration of the complete item, only the dissolution of this thin
grain boundary layer is necessary. Any applied stress accelerates the
process.
In the present case, corrosion took place at a location where a corrosive
medium could remain for a longer time and where the material (prongs)
is stressed.

The only safe way to avoid this defect is to change the alloy composition
to prevent segregation of undesirable compounds to the grain boundaries.
(A thick layer of electroplated high carat gold will also help, at least for a
limited time).
Stress corrosion and/or intergranular corrosion, leading to cracking along grain
boundaries, is one of the most common reasons for complaints with low carat
gold alloys (10ct and below). In gold alloys, the two mechanisms cannot be
separated unambiguously.
The principal mechanism can be explained as follows:
Grain boundary segregations have a different electrochemical behaviour (in
relation to the more noble gold-rich grain itself) and this forms the basis for a
voltaic (or galvanic) cell, where the presence of a corrosive medium (e.g. sweat,
sea water, cleaning fluids, etc) results in an electrochemical/galvanic corrosion
process. In the very small anodic area on grain boundaries, the less noble
material is preferentially dissolved. The corrosion current is concentrated on this
small area, and therefore the dissolution of this region is very fast.
Stress (tensile stress) can increase the corrosion rate to a significant extent.
The initial progress of intergranular corrosion forms notches which expose new
areas to corrosion, further extending the depth of notches and increasing the
concentration of stresses on the base of the notches and so it continues. (This
explanation of stress corrosion is somewhat simplified and cannot be applied
unmodified to other metal alloys.)
This kind of corrosion can be expected particularly in low carat alloys. It is
much less frequent in 14ct alloy. Known exceptions are the long term attack by
hydrochloric acid (used for removing an aluminium core from a hollow jewellery
item) or a corrosive attack on an alloy with less noble precipitation on grain
boundaries. Differentiating between intergranular corrosion and stress corrosion
in gold jewellery alloys is very difficult. The mechanisms are not yet completely
evaluated for this type of alloy.
61
5
5 SOLDERING
5.1 FRACTURE
CASE 24: FRACTURE - DUE TO MELTING OF LOW MELTING
COMPONENTS DURING SOLDERING.
Key Words
Soldering; brazing; fracture, white gold, low melting components.
Fig. 24.1 Broken prong on a white gold

finding
A prong has broken on a white gold finding. The fracture is located at the
junction of prong and bridge. Obviously, some soldering has been done
on the item.
Visual Appearance
The prong was broken without deformation. The fracture looks dendritic.
Remnants of solder are visible
At low magnification, the structure resembles a honeycomb, fig 24.2.
This image is not so easy to interpret; sometimes, severe gas porosity can
cause a ‘fracture’ with a similar appearance.
Alloys: 14ct white gold, based on gold-copper-nickel-zinc with
Fig. 24.2 SEM image of the fracture
additions of silicon and boron. The solidus temperature is very low; surface
melting range is 845° - 885°C (1550°F - 1625°F).
In principle, the defect can occur in any alloy containing low melting
components (e.g. silicon)
Manufacturing Method: The finding was produced by vacuum
assisted, static investment casting under normal conditions. The setting
had been soldered to a shank.
At the place where the fracture occurred, almost no physical connection
between the two soldered parts existed. Applying a very low stress has
lead to the fracture.
Microstructure
The investigation of the fracture surface in the scanning electron
microscope (SEM) reveals a dendritic structure; the dendrites have smooth
edges, suggesting that they have been molten.
The constituents of the solder used - cadmium, gold, silver and copper
- were found on the fracture surface.
Brief Explanation
The soldering temperature was too high and, probably in this case, an
additional problem is the presence of low melting components in the
alloy (mainly due to the silicon content). The low melting silicon phase
melted and became liquid in the space between the dendrites of the cast
structure, inhibiting the flow of solder into the gap which was only partly
filled with a small quantity of solder, insufficient to produce a solid bond.
Sometimes, the appearance (morphology) of this defect is very similar
63
5 to fracture arising from stress corrosion or intergranular corrosion. A
distinction between these types of defects is difficult and can only be
done with certainty if related information is provided about the history
of the defective item.

The soldering temperature has to be decreased by applying less heat
and/or use of a lower melting solder, where possible. If this is not
possible, then another alloy composition should be used which is free of
(or has reduced amounts of) low melting components.
Further Reading
Basic Aspects, section 2.4, Influence of low melting components.
64
6
6 H E AT T R E AT M E N T
6.1 AGE HARDENING
CASE 25: CHANGE IN SIZE OF A RING AND FRACTURING - DUE TO
UNWANTED AGE HARDENING.
Key Words
Cracks, fracture, age hardening.
Fig. 25.1
DESCRIPTION OF THE DEFECT
The ring grew in size during soldering and finally cracked with a brittle
fracture. The same effect happened while annealing the same material at
360°C. (The material had been pre-treated by soft annealing at about
700°C. )
Alloy: 18ct pink gold (gold-copper-silver)
Manufacturing Method: Investment casting, soft annealed at 700°C,
soldered, and age hardened at about 360°C.
The alloy is embrittled and has very low ductility.
Microstructure
The grains are very large, fig 25.1. A single grain covers almost the whole
cross-section of the ring shank. The grains show an internal structure of
series of parallel ribbons, a microstructure that is characteristic of age
hardened material.
Brief Explanation
Copper-containing gold alloys of 18 carat or lower have a very special
property. Soft annealed material (that is, material which has been
annealed at about 700°C ), if slow cooled, changes its microstructure by
an ‘ordering’ effect in which the gold and copper atoms are no longer
distributed randomly in the crystal (grain) lattice but are arranged in a
well defined order. The beneficial effect of this behaviour is substantially
increased hardness and improved wear resistance. The disadvantage is
reduced ductility leading to brittleness and a physical change in size
during the hardening process. The end result can be a change in size of
the jewellery item, important if it is a ring, for example, or even cracks
and fracture as can be seen from the present example.
This phenomenon can be controlled and used beneficially in jewellery
production by rapid cooling (water quenching) the alloy from the soft
annealing temperature (this retains the soft condition and the ductility)
and, as a final pre-polishing process, ‘ageing’ the alloy at a low
temperature, typically 260-360°C for 1 hour.
Both binary gold-copper alloys and gold-copper-silver alloys are
sensitive to the age hardening phenomenon. The extent of age hardening
depends strongly on the copper/silver ratio and the caratage. Pink and red
gold alloys with high copper contents are especially sensitive. Note: High
carat golds (21/22 carats and higher) are not affected by this phenomenon.
65
Age hardening can be avoided by water quenching the material as fast as
possible from a temperature higher than 500°C, preferably 600-700°C.
There are potentially some difficulties if subsequently soldering the
material, causing parts to be heated up in the range 300-400°C, since
ageing will occur at these temperatures. A safe solution to this problem
doesn’t exist. The best way would be to cool down the alloy from a
temperature above 500°C to room temperature as slowly as possible (in a
furnace) to enable the ordering process to occur, then carrying out
Fig. 25.2 6 x soldering which should avoid formation of any cracks. To soften
(ductilise) the material again (if desired) soft annealing at about 600°C
should follow - but only if the melting range of the solder lays well above
this temperature!
Extended Explanation and Further Reading

More detailed information is provided in Part C: Basic Aspects of Metallurgy,
section 2, Age hardening. It is strongly recommended to read this chapter if
difficulties with red or pink gold occur. See also the ‘Technical Manual for Gold
Jewellery’, chapter 6, published by World Gold Council.
Another example
Fig. 25.3 1000 x The edge of a stone setting on class ring is cracked and a piece has broken
off, fig 25.2. The ring was only worn for a short time. The bezel of the
ring was soldered on the shank. The yellow gold alloy contains silicon.
Very small segregations are visible on grain boundaries, fig 25.3. The
investigation revealed two relevant facts, both contributing to the failure: a)
The ring is age hardened with a hardness of HV 275, a high value for a
yellow gold alloy; embrittlement is the natural consequence. The age
hardening occurred during soldering. b) The material shows segregation on
grain boundaries which are very small, but they can cause embrittlement.
66
6
6.2 BLISTERING
CASE 26: BLISTERING AND POROSITY - DUE TO HEAT TREATMENT IN A
HYDROGEN-CONTAINING ATMOSPHERE.
Key Words
Blistering; gas porosity; intergranular cracking; hydrogen embrittlement;
surface exfoliation.
Description of the defect

Annealed thick sheet and watch case pressings showed surface exfoliation Fig 26.1 Polished and annealed surface,
of grains. In some instances, the (originally bright) surface had a ‘milky’ 18ct yellow gold 200 x
appearance. Close inspection under a low power microscope showed
surface blisters.
Visual appearance of the defect

Figures 26.1 and 26.2 show protrusion and exfoliation of whole grains
from the polished surface of 18ct yellow gold (3N) strip after annealing in
a cracked ammonia atmosphere at 780°C. In other cases, the surface has a
‘milky’ appearance which, on inspection under a microscope, is revealed
to be a blistered surface.
Alloys: The defects have been observed in several conventional gold
jewellery alloys: 14ct yellow gold, 18ct yellow gold (Au-Ag-Cu, 3N and Fig 26.2 Polished and annealed surface,
2N) and 21ct alloy. 18ct yellow gold 200 x
Manufacturing method: Continuously cast and rolled strip which

was interstage annealed in a hydrogen-containing atmosphere, cracked
ammonia, at a relatively high temperature of 650°C for 30 mins.
Subsequently blanked and stamped with a final anneal at 780°C in
cracked ammonia. Alloys were made from 4-9’s purity gold, electrolytic
silver and high purity, electrolytic copper.
Similar defect observed after soldering (brazing) of carat gold in a
hydrogen-containing atmosphere.
Mechanical properties Fig 26.3 Watchcase, cold pressed and
Mechanical properties usually are not significantly influenced by this annealed (650°C, 30 mins) 200 x
type of defect.
Surface quality
As already described, the surface structure is degraded by blistering and
exfoliation of grains. Polishing does not normally succeed in removing
the defect because new blisters and pores are continually revealed.
Microstructure
In these cross-sections of an 18ct (3N) yellow gold, figs 26.3 and 26.4, the
incidence of widespread porosity and intergranular cracks throughout the
section thickness is evident. The porosity is spherical and appears
predominantly at grain boundaries and is similar to gas porosity seen in
castings. The alloy is known to have had high oxygen contents, up to Fig 26.4 Watchcase, cold pressed and
annealed (650°C, 30 mins) 100 x
40-50 ppm.
Spherical gas pores in brazed joints of a pen made of a high carat alloy,
fig 26.5, are also found in the wrought material which was not molten
67
6 during brazing. Obviously, the pores were formed after rolling the
material since any sphericals void would be flattened by rolling. The only
possibility for the generation of gases is during brazing (soldering) or
during the heat treatment connected with brazing.
Brief Explanation
Blistering and porosity are characteristic of gas formation and similar
effects are common in copper and silver alloys. Coupled with particularly
high annealing temperatures in hydrogen-containing atmospheres, the
evidence is strongly suggestive of the ‘steam reaction’ whereby hydrogen
Fig. 26.5 Brazed joint of a pen made in a
high carat gold diffuses into the alloy during annealing, preferentially along the grain
boundaries, and reacts with oxide or dissolved oxygen to form steam
(water). As the water molecule is relatively large, it cannot diffuse out and
so remains and creates an internal gas pressure within the alloy, causing
porosity within the bulk and blisters at the surface, as well as grain
boundary cracking. At the surface, this grain boundary cracking results in
exfoliation of whole grains.
The less pure grades of copper often contain residual copper oxide and
is the reason why they should not be used in carat gold manufacture.
However, in this case, high purity electrolytic copper was used, implying
that oxide was unlikely, although relatively high dissolved oxygen levels
were found. It may be that this oxygen is present in the form of very
small copper oxide particles, which would tend to be found at grain
boundaries, and arise during continuous casting.
Thus two factors are responsible for the occurrence of these defects: a)
annealing in hydrogen-containing atmospheres at high temperatures and
b) high oxygen levels (possibly as fine oxide particles) in the wrought
alloy. These combine to allow the steam reaction to occur. Investigations
where carat golds with low (<10ppm) oxygen contents were processed by
the same route, including the same annealing schedules, did not result in
the formation of defects.
In the case of the brazed joints, examination of microsections at high
magnification revealed a great number of very fine particles, probably
copper oxide. This is consistent with analysis which measured 20-40 ppm
oxygen (by weight).

If the oxygen content is low or there is no hydrogen, then clearly, no
steam reaction can occur. If the annealing temperature is reduced, then
the diffusion of hydrogen into the alloy is markedly lowered.
Thus, there are three factors which, individually, should lead to
avoidance of this defect. The first - and preferred - is the use of a low
oxygen (oxide free) copper and silver for alloying (and clean, oxide-free
scrap, if recycling) and to ensure that the conditions for producing the
cast gold alloy result in a low oxygen content (preferably <10ppm). The
second is to avoid use of hydrogen-containing reducing atmospheres
where possible (use neutral atmospheres such as nitrogen or argon). If
this is not possible, then the third factor is to perform annealing at lower
temperatures for longer times. For 18ct (2N, 3N) alloys, the recommended
68
annealing temperature is 550°C, 30 mins. Lower annealing temperatures
6
also reduce the risk of excessive grain growth which can manifest itself as
‘orange peel’ surfaces in subsequent forming and bending.
Diffusion of gases in solid metals is well known. Importantly, the rate of
diffusion is very strongly temperature dependent. Among the precious metals,
the diffusion of oxygen and hydrogen in silver or of hydrogen in platinum and
palladium are well documented. Blistered silver sheet is the result of the ‘steam’
reaction of dissolved oxygen (or oxide) with hydrogen during annealing. Similar
behaviour in gold alloys has not been previously described in the literature.
However, several cases have been reported recently.
It seems that very small oxide particles, hardly detectable by examination of
microsections are especially critical for causing blistering. These micro particles
may be formed in continuous casting.
Another critical process is soldering (brazing) in a furnace in hydrogen-rich
atmosphere (which of course is favourable for the soldering procedure) where,
again, the temperature is relatively high and the diffusion rate of hydrogen
strongly increased.
Note: This case history is published by kind permission of Argor-
Heraeus S.A., CH-6850 Mendrisio, Switzerland.
69
PA R T C :
BASIC ASPECTS
O F T H E M E TA L L U R G Y O F
C A R AT G O L D S A N D
THE INVESTMENT
CASTING PROCESS
1 The following sections are intended to provide some additional, general
information on the metallurgy of carat golds and the investment casting
process as an aid to understanding the defect case histories described in
Part B. The intention is not to provide a substitute text book, but merely
to underpin the explanations given in Part B with some basic technology
involved as a guide for those with little formal technical training in
metallurgy or chemistry.
1 GENERAL PROPERTIES OF
M E TA L S
1.1 PHYSICAL PROPERTIES

Metals are crystalline substances with some extraordinary properties.
The crystals (or grains) are formed by atoms lying in a regular array
(crystal lattice) which share a common ‘cloud’ of negative electrons.
This atomic construction is responsible for:
• optical reflectivity
• high electrical and thermal conductivity
• the unique working properties, ductility and work hardening
1.2 STRUCTURE
The structure of a metal crystal is, as in the case of any other crystalline
substance, determined by the shape of the unit cell. This is the smallest,
basic unit of a crystal, which when extended by the addition of many
other unit cells, builds up to produce the overall crystal.
The unit cells contain only a small number of atoms in a relatively
simple arrangement. The most common structures for metals are the
‘face-centred cubic’ (FCC) and ‘body centred cubic’ (BCC). These are
cubes with atoms on each cube corners and in the centre of the face
planes (FCC) or in the centre of the cube (BCC). Another common
structure is the hexagonal prism. Other less frequently occurring types of
structure can also exist with metals, but are not usually of concern with
gold alloys.
Any crystalline body tends to form a regular and typical macroscopic
‘crystal’ shape, well known from minerals and gemstones. Metals are a
little bit different. On solidification (see section 2.1), a number of crystals
start to grow in the melt. These embryonic crystals are called ‘crystallites‘.
The growing crystals (or ‘grains’) grow into and obstruct one another
from building up the regular external shape of crystals whereas the
regular inner atomic lattice structure is maintained. The result is a fine
(small) grained polycrystalline metal whose structure is only revealed by
examination of a polished microsection under a microscope.
The average size of the grains in a metal sample is called the ‘grain
size’ and is normally expressed in mm (or inch) or simply by a number
(the ASTM number). It is important as it influences the properties.
72
2
2 M E TA L L U R G I C A L E F F E C T S
2.1 MELTING AND SOLIDIFICATION
Melting of a solid piece of metal is a procedure which is not difficult to
understand. Increasing the temperature of a piece of solid matter causes
an increasing vibration of the atoms. At a characteristic temperature,
the vibrations are intense enough to overcome the atomic forces that
hold them in a regular crystalline array. The atoms start to move
relatively independently of each other, i.e. the material starts melting to
Figure 1. Schematic representation of a
become a liquid. dendrite
At this point, additional energy (heat) is needed to break the
inter-atomic binding forces. The temperature of the material remains
constant whilst the heating continues until all the solid has melted.
An example is melting ice. The temperature of a water-ice mixture
remains constant at 0°C as long as unmelted ice and water are present.
The same behaviour is shown by pure metals. For example, pure gold
melts at a constant 1063°C.
The melting behaviour of alloys [an alloy is a mixture of two or more
pure metals] is a little more complicated. In general, all alloys melt within
a melting range (with some specific exceptions). A more detailed
explanation is possible using phase diagrams but is beyond the scope of
this Handbook.
Solidification is the opposite of melting but more complex and is of
greater practical importance. The basis for the properties of a piece of
metal is laid at solidification.
The almost independent atoms of the liquid melt are not able to
instantly rearrange themselves into solid crystals. The growth of new
crystallites (the growing grains) starts from nuclei (‘seeds’). These nuclei
may be very small clusters of metal atoms or particles of impurities (e.g.
oxides or other insoluble metals such as grain refiners). The nuclei may
be initiated at preferred sites such as the roughness of the crucible or
mould wall.
The crystallites do not tend to grow uniformly in all three directions of
space until they meet adjacent growing grains, which would result in a
polyhedral array of equi-sized grains. In most cases, directional growth of
the crystallites along preferred crystallographic directions occurs.
Primarily, needle-shaped crystals are formed, which branch again and
again to produce the typical dendrite (see diagram, Fig. 1), a structure
which we find in jewellery alloys. The orientation of the various
dendrites in the melt vary, and they grow until they impinge on each
other.
In actual alloys, the composition of the crystallites changes during
solidification. The centre of the dendrites which is formed at the
beginning of solidification has a certain composition (defined on the
alloy phase diagram) and this gradually changes as solidification proceeds
to completion. This is known as ‘coring’. The average composition of the
dendrite is as one would anticipate from the overall alloy composition.
If solidification occurred very slowly, the different metal atoms would
73
2 migrate and tend to equalise differences in concentration from dendrite
centre to outside, a process which is called diffusion. Usually, solidification
and subsequent cooling occurs too fast to allow adequate diffusion to
equalise any differences in local dendrite composition.
Any type of crystal with defined composition range and atomic
structure is considered as a phase. A metal or alloy which is built of only
one kind of crystal is said to be single phased. An alloy can be single
phase but may also consist of two or more phases:
a) Two or more metals may form a homogeneous solid solution: only
one phase is formed which contains all atoms of the different
constituents. (A similar example is the dissolution of salt in water.
Salt water is only one phase which contains all constituents.).
This means that if the metals are completely soluble in one another,
the alloy is homogeneous and single phase.
b) Two and more metals may also form different phases e.g. phase
a and phase ß which contain either (almost) pure metal A or metal
B or both metals in different relations e.g. phase a: (high A + low B),
phase ß: (high B + low A). In this case, both metals have no or only
restricted solubility in one another. An alloy comprising a mixture
of phase a and phase ß is two phased or heterogeneous. Depending
on composition and temperature, some alloys can change from one
crystal phase to another.
In summary: The number and kinds of phases existing in a given alloying
system depend on composition, temperature and pressure
(for practical purposes, the influence of pressure can be neglected).
There are some special kinds of heterogeneous alloys. Only one will be
mentioned here. A ‘eutectic’ is a special defined composition in some
alloy systems where melting or solidification occurs at a specific defined
temperature rather than over a melting range. This eutectic temperature is
normally much lower than either melting point of the constituent pure
metals. A eutectic microstructure consists of a very fine grained mixture
of at least two phases with no dendrite grains. A eutectic is formed in a
number of alloy systems including silver-copper and gold-silver-copper
jewellery alloys.
A particular case of a phase in an alloy system is the formation of an
intermetallic compound, which resembles a real chemical compound
formed between two elements with different characteristics. The atomic
ratio of the constituent metals is fixed within a certain narrow range.
For practical purposes, most intermetallics are brittle. Thus formation of
intermetallic phases in an alloy may result in some embrittlement,
i.e. loss of ductility. Gold-copper alloys can form intermetallic phases,
and these can cause hardening of the alloy.
A final remark is necessary: The principles of alloying is one of the
most important subjects taught in metallurgy as it is basic to
understanding the resulting alloy microstructures and hence properties.
The basis for understanding phase relationships are phase diagrams.
As mentioned earlier, it is not possible to treat this subject in sufficient
depth here; an appropriate text book should be consulted.
74
2.2 DEFORMATION AND MECHANICAL PROPERTIES
2
A pure metal or an alloy usually shows two kinds of deformation
behaviour when a load (stress) is applied:
When a low load or force is applied (e.g. to a wire), a reversible
elongation is observed, i.e. after releasing the load from the metal,
it‘s original length is restored. The metal has been stressed in the ‘elastic’
range and the amount of deformation - or strain - is proportional to the
stress applied.
When the stress is further increased, a limit for such elastic
deformation is reached. Beyond this limit, part of the resulting
deformation remains after releasing the stress. The material is
permanently deformed and the deformation is said to be ‘plastic’.
The applied stress at the onset of plastic deformation is called the yield
strength.
When the stress is increased further, further plastic deformation occurs.
However, a maximum stress will be reached; beyond this value, the stress
needed to maintain continued deformation decreases until eventually the
metal sample breaks (fractures). The total plastic deformation at the
moment of fracture can be determined by joining the two parts of the
broken metal sample and measuring the increase in length compared
with the original length. The maximum stress is called the tensile strength,
the plastic increase in length at fracture is the ductility (or elongation in
the tensile test).
Yield strength, tensile strength and elongation (ductility) are basic
values for characterisation of the mechanical properties of a metal or
alloy. It should be noted that stress is defined as a load per unit area
applied to the original cross-section and the elongation is the increase in
length related to the original length (in per cent, %). In practice, it is not
always convenient or possible to perform tensile tests on a metal or alloy.
The use of hardness measurements can serve as a rough substitute.
The principal of the hardness test is simple: An indentor of a given
shape and high hardness (often a diamond) is contacted on the surface of
the material to be tested. A load is applied for a specified time and the
depth of penetration of the indentor is measured and from this a
hardness value calculated. There are a number of hardness test methods.
The most common are the Vickers and Brinell tests. The Vickers test is,
perhaps, the most commonly used for jewellery alloys.
Vickers hardness uses a diamond pyramid with a square base as the
indentor. The load applied to the test sample is selected to suit the alloy
and sample size. The diagonal of the resulting indentation is measured,
and a mathematical formula relates the indentation surface area and
applied load to give a value called the Vickers hardness, abbreviated to
HV. An example for a standard yellow carat gold is: 180 HV10; this is a
hardness value of 180 measured at an applied load of 10 Kg. Often, it is
just quoted as HV 180.
Influence of deformation on properties

Deformation (performed at room temperature) increases the tensile
strength and hardness. The material is said to be ‘work hardened’.
75
2 The ductility - or elongation (as measured with the tensile test) - is
consequently reduced. The workability is reduced; the material becomes
less ductile and will break if the deformation continues beyond a certain
level.
The amount of deformation (called ductility) a metal or alloy can
sustain is very dependent on the composition (and microstructure) of the
alloy. For example, nickel-white gold has only a very limited ductility,
whereas palladium-white gold can bear much more deformation, i.e. it is
more ductile. Pure gold can be deformed with almost no limitation (and
is the most ductile metal of all).
Influence of deformation on microstructure

The grains become elongated and thinner when the material is rolled or
forged or a wire is drawn to accommodate the overall shape change. With
a high degree of deformation, the grain structure is severely disrupted and
appears to disappear completely. The microsection shows a ‘fibrous’
structure.
The mechanism of deformation at the atomic level of a metal crystal or
grain is not easy to explain. The crystal lattice of a real metal crystal
contains a number of faults (‘dislocations’). At the beginning of
deformation, the faults move along certain crystal planes to enable the
planes of atoms to slide past each other, and this gives rise to an
elongation of the grains (and therefore an overall elongation of the sheet
or wire, etc.). Unfortunately, many more crystal faults are formed with
progressive deformation which collect and intersect with each other and
form ‘knots’, which, in turn, hinder movement of the faults, requiring
more stress to produce further deformation. Thus, the material gets
harder and the ductility decreases until the point is reached when no
more deformation is possible and increasing the applied force further will
break the item.
2.3 ANNEALING AND HEAT TREATMENT

2.3.1 SOFT ANNEALING
In most cases, annealing is carried out on work-hardened (deformed)
material to restore the original soft, ductile condition as the cold worked
material cannot be safely deformed any more. This is done by applying a
heat treatment called soft annealing which causes recrystallisation of the
deformed microstructure.
Increasing the temperature enables the atoms (and faults) in the
deformed crystal lattice to move. At first, the internal stress is removed
by this movement. This stress relief makes the item a little less brittle
(and, incidentally, prevents low carat gold alloys from suffering stress
corrosion cracking). When the temperature is increased further, a new
process starts, which is very similar to the formations of grains
(crystallites) at solidification. New crystallites are nucleated and grow into
the deformed material until all deformed material is consumed, resulting
in a microstructure of new equi-axed (i.e. uniform shape), undeformed
grains. The material has recrystallised.
76
The size of the resulting new recrystallised grains depends on several
2
factors:
a) The number of grain nuclei - the higher the degree of deformation,
the more nuclei are initiated. Therefore, only heavily worked
material should be soft annealed if a fine grain size is required.
Preferably, the degree of deformation should be between 50 and 70%
with normal carat golds. A critical deformation of about 15% is
needed to initiate recrystallisation and will result in a very coarse
(large) grain size.
b) Recrystallisation needs a certain minimum temperature to
commence. Below this level new grains are not formed. However,
exceeding this critical temperature to a significant degree causes the
grains to grow rapidly and any smaller grains are consumed by
larger ones which grow even larger. The result is a very coarse
grained structure, one consequence of which is a rough surface on
subsequent deformation such as deep drawing or bending - the so-
called ‘orange peel’ effect. Removing this by polishing is costly.
c) Recrystallisation needs a certain time to be completed.
The nucleation of new grains is very fast (in the range of minutes)
but some time is necessary for these grains to grow and consume all
deformed material. The influence of time on the growth of grains is
much more moderate than that of temperature. Therefore, it is
recommended to vary the time and keep the temperature constant
and to a minimum to effect control of the resulting mechanical
properties and grain size. However, if annealing by a gas torch, such
control is not easily attained.
The grain size obtained on soft annealing depends strongly on the
type of alloy as well as the amount of cold work. Pure metals (e.g.
fine gold) tend to result in a larger grain size as growth of new grains
is more rapid, whereas alloys shows a smaller but broader variation
in grain size. For example, pink or red golds usually have a
significant larger grain size than yellow golds. The grain size can also
be influenced by very minor additions of ‘grain refiners’ such as
ruthenium, iridium and cobalt. These enhance nucleation of new
grains, refining the resulting microstructure, if applied under correct
conditions (see also ‘grain refiners’). Additions of other metals such
as silicon, iron and chromium have a coarsening effect.
2.3.2 AGE HARDENING (HARDENING BY HEAT TREATMENT)

Certain carat gold alloys containing copper, in the range 9-18 carat,
undergo a hardening when slow cooled or heat treated.
This phenomenon can cause difficulties in jewellery fabrication but can
be utilised to deliberately harden these alloys and improve strength,
scratch and wear resistance of the finished jewellery. There are two
hardening mechanisms relevant to gold jewellery making: a) precipitation
hardening and b) hardening by ordering.
Precipitation hardening can be explained in a simplified way by
means of an example: The solubility of copper in silver is limited and
depends on temperature. Decreasing the temperature decreases the
77
2 solubility. If an alloy containing sufficient copper is slowly cooled down
(say from 700°C to room temperature), the surplus of copper precipitates
from the silver-rich matrix as a second phase - a copper-rich solid
solution - due to the decreased solubility for copper. If the alloy is rapidly
cooled by water quenching, such precipitation is suppressed and a
supersaturated (solid) solution with copper results. Subsequently, heat
treating at a temperature of about 200 to 400°C enables precipitation of
the copper-rich second phase in a very fine dispersed form which
significantly increases the hardness. (This explanation is simplified, the
real mechanism is more complicated).
This hardening process is not only restricted to silver-copper alloys but
is also of importance in yellow-pink gold alloys (consisting of gold-silver-
copper), especially 14ct alloys, where the precipitation of a copper-rich
phase influences not only the hardness but also the colour.
Hardening by ordering is a special mechanism occurring in gold-
copper alloys and in yellow golds where the copper content is relatively
high. In the gold-copper system, a homogeneous solid solution exists at
all concentrations above approximately 410°C. The gold and copper
atoms are randomly distributed in the alloy crystal lattice. However,
below this temperature, several ordered structures of intermetallic
compounds (AuCu, Au3Cu and AuCu3) will be formed. The copper and
gold atoms occupy fixed positions in the crystal lattice and the alloy is
harder and less ductile.
The ordering process can be suppressed by quenching the material
from a temperature above 500°C and the alloy retains its ductility and
low hardness. This state is preferred if the material is to be further
deformed. Heat treating this soft quenched material directly or after
further deformation in the range between 200 and 400°C will cause age
hardening by the ordering process. The material becomes harder and less
ductile. The magnitude of this effect depends strongly on the gold-copper
and copper-silver ratio. It is strongest in red and pink gold alloys in the
14-18 carat range. Hardness values of 300 HV can be easily obtained.
In some alloys, ordering cannot be suppressed completely by quenching,
making the material difficult to work.
Particular difficulties can occur with irreproducible cooling conditions
during investment casting. Subsequent annealing at about 700°C is
recommended to avoid cracks in critical red gold alloys.
In gold-copper-silver yellow alloys, the process is more complicated.
Prior to the ordering process, precipitation of a copper-rich phase from
the silver-rich solid solution occurs. The subsequent ordering process
mainly affects the copper rich phase.
2.4 LOW MELTING COMPONENTS

Low melting components occasionally occur in jewellery gold alloys
either by additions made deliberately or by accidental impurities.
Two conditions have to be fulfilled for an addition or an impurity to
form a low melting component: (a) the addition has a low melting
temperature and a very limited solubility in the gold alloy, and (b) the
addition/impurity forms small amounts of low melting compounds of
78
limited solubility with the main alloying elements (gold, silver, copper).
2
Typical low melting additions are silicon and gallium. Silicon is much
more critical than gallium because of its small solubility in gold and its
insolubility in silver. Typical impurities are lead, bismuth, sulphur and
phosphorous. Lead can be introduced by: (a) gold alloy contaminated
with soft solder (lead-tin) or (b) by use of lead-containing (‘free
machining’) brass for alloying. Brass (copper-zinc) is usually used as a
means of introducing zinc into gold alloys to reduce zinc loss by
evaporation. [note: brass with less than 70% copper can contain lead;
with less than 62% copper, brass generally contains lead.]
Sulphur, as a sulphide, is introduced during investment casting by the
reaction of the melt with the investment or by the remelting of impure
metal (see Reaction with the Investment). Phosphorous (as phosphide) rarely
occurs, although it is sometimes used as a deoxidiser in silver.
Bismuth as an impurity in jewellery alloys may arise from use of soft
solders.
Low melting components tend to segregate at grain boundaries,
causing brittleness and fracture can occur with little applied force.
Grain size can influence this effect through the grain boundary surface
area available for such segregation; a fine grain size is preferable.
2.5 GRAIN REFINERS

Two kinds of grain refiners have to be considered: Those producing a fine
grained cast structure (such as iridium and ruthenium) and those effective
during soft annealing (e.g. cobalt). Only grain refiners used for casting are
considered here because they are more frequently the cause of defects.
Principles of working mechanism

At the commencement of solidification, the formation of crystallites
(grains) in the melt needs nuclei as starting points. The more nuclei that
are available, the more grains that can grow at the same time and the
smaller the resulting grain size of the dendritic structure. One method to
increase the number of nuclei is by adding high melting elements or
compounds to the melt. These should have no - or only a very small -
solubility in the solid alloy and have to be present as a fine dispersion
before the melt starts to solidify to be effective. For this reason, the best
effect is obtained if the compounds are homogeneously dissolved in the
melt and precipitate prior to the beginning of solidification.
The common grain refiners used are high melting metals of the
platinum group, with limited solubility in gold alloys, and some very
reactive elements which form intermetallic compounds or oxides or
nitrides as the effective nuclei.
79
3
Cono di ritiro 3 INVESTMENT CASTING
3.1 SHRINKAGE POROSITY

All metals and most alloys show a sharp decrease in volume on solidification
(solidification shrinkage) due to the closer packing of the atoms in the crystal
Lingotto
lattice. Such shrinkage effects can lead to defects in gold jewellery alloys.
At first, consider the casting of a simple ingot: The solidification of an
ingot starts from the mould wall, commencing from bottom to top.
The centre of the ingot at the top is usually the last part which solidifies.
Lingottiera
Due to the shrinkage, there is insufficient melt to fill the remaining space
Figure 2 Schematic: Shrinkage cavity on the top and a depression or funnel-shaped shrinkage cavity (pipe)
(primary pipe) on top of a cast ingot
remains, Fig. 2. Sometimes, a thin, deep cavity is formed (axial porosity,
secondary pipe). Sheet made from such an ingot can suffer delamination
and inclusions.
In investment casting, the situation becomes more complicated for
several reasons: The size of the casting is relatively small and the shape
complex; heat transfer is very dependent on the shape. A well-ordered
progression of solidification from outside to centre and bottom to top (as
in ingot casting) is not possible.
Early solidification of the gate (which connects the casting to the
central sprue or feeder) and other small cross-section parts can prevent the
supply of additional melt to compensate for shrinkage in the casting.
Solidification takes place in a very short time - a few seconds. Heat transfer
and growth of crystallites are strongly influenced by several factors.
Gold jewellery alloys (as with other alloys) start solidification by the
formation of dendrites (see section 2.1 ‘Melting and solidification’ above).
These dendritic grains impact and grow into each other, building a
sponge-like structure and increasing the resistance to further melt flow
needed to take account of solidification shrinkage. Therefore, the filling
of inter-dendritic spaces by additional supply of melt is hindered, leading
to interdendritic porosity with the typical dendritic shape.
Definite rules for avoiding shrinkage porosity are difficult to provide
but guidelines include:
• Avoid pronounced changes in cross-section within an item.
• Provide a gate of sufficient diameter to avoid its premature
solidification before the casting. Note: an extremely short and thick
gate can also cause shrinkage porosity.
• Where practicable, place the gate on the thickest part of the jewellery
item. If necessary, use two or more gates positioned to ensure feeding
of liquid melt to thick sectioned areas.
The tendency to shrinkage porosity depends also on the type of alloy.
Higher carat alloys seem to be less sensitive to this defect, especially if
they have a higher silver content. Casting and flask temperatures have a
significant influence, too. The relationships are complex and cannot be
readily explained here. Reference should be made to other publications
(see list at end). It is recommended that trials with varying temperatures
should be performed to find suitable working conditions.
It is important to note that improvements in casting quality by
80
optimising temperatures are only possible if items of similar size and
3
cross-section are used for building up the wax tree. An optimum set of
casting conditions cannot be achieved if the tree consists of a mixture of
filigree and heavy parts, for example.
3.2 GAS POROSITY

Gases which are dissolved in the melt, or introduced with the flow of
liquid metal or formed by chemical reaction, can produce a special kind
of microporosity after solidification.
In its most typical appearance, the micropores are almost spherical (see
Case 1). However, many other irregular or dendritic shapes are also
possible (see Case 2). For this reason, a clear identification of the nature
and origin of the porosity is not always possible.
In gold alloys, the most common source of gas porosity is the
formation of sulphur dioxide by reaction of the melt with the mould
investment. The presence of copper oxide in the melt can also cause gas
porosity as it is believed that the oxide can decompose on solidification,
releasing oxygen under some conditions. On the other hand, porosity
generated by entrapped air or protective atmosphere is believed possible
although not proven.
To avoid gas porosity, two rules should be observed:
• Avoid conditions causing reaction of melt with investment

- The reaction of investment with the melt is covered later in
section 3.5.
• Use only clean material for preparing the melt.
- Use of too high a proportion of used/scrap material or recycling
of ‘dirty’ casting scrap (e.g. central sprues), insufficiently
cleaned, for remelting is potentially dangerous.
3.3 INFLUENCE OF INVESTMENT ON CASTING QUALITY

The quality of investment castings is strongly influenced by the
properties and behaviour of the investment from which the mould is
made. Properties which have to be considered include:
• setting behaviour of the slurry
• strength in the ‘green’ and the fired state
• gas permeability
• thermal conductivity
• chemical stability, i.e. the reaction of investment compounds with
the molten metal.
For a better understanding, some remarks are made on the composition
of investment powder; the investing procedure and firing process.
Composition of investment powder: Generally, the investment powder
consists of fine grained mineral powder as the main component (‘body’)
and a binder. The binder gives the strength and can be added as a solid
directly to the dry mineral powder (as it is the case with gypsum-bonded
investment) or as a liquid together with the water in preparing the slurry
as is the case with phosphate-bonded investment. In gold jewellery
casting, gypsum-bonded investment is normally used.
81
3 The main components are:
• silica (silicon dioxide) in two crystal modifications (quartz and
cristobalite) as the refractory body,
• gypsum (Plaster of Paris) - calcium sulphate - as the binder,
• some additions e.g. detergents, for improving wetting behaviour of
the slurry with the wax, and agents to modify setting time, thermal
expansion and strength.
Whereas silica is relatively stable and does not lead to problems in casting
of gold alloys, gypsum can cause many problems, depending on the alloy
to be cast, flask and melt temperatures and working conditions.
Note: Gypsum-bonded investment is not recommended for casting
palladium-white golds. Even with nickel-white golds, the use of gypsum-
bonded investment can be critical. For good reasons, it is a material best
suited for the casting of coloured carat gold alloys with relatively low
melting ranges. Alternatively, phosphate-bonded investment can be used
for casting jewellery alloys with a higher melting range (and/or greater
reactivity). In this investment, phosphate compounds as used as the
binder (in combination with magnesia) in place of gypsum; an organic
silicate can also be used as the binder, producing another type of
investment with similar properties.
The advantages and disadvantages of both types of investment are:
• Gypsum-bonded investment is convenient to handle and relatively
cheap. However, the instability of calcium sulphate is the cause for
many casting defects, as is discussed later.
• Phosphate- (and silicate-) bonded investment is not so easy to handle
and more expensive. Its thermal and chemical stability in jewellery
casting is excellent, but the subsequent removal of phosphate-bonded
investment from the casting can be difficult.
Because of its dominant use in jewellery casting, only the behaviour of
gypsum-bonded investment is considered in the following, especially
with relation to casting defects.
3.4 THE INVESTING PROCESS (GYPSUM BONDED

INVESTMENT)
3.4.1 SETTING TIME
The investing process starts with addition of water to the investment
powder which must be mixed thoroughly. A reaction occurs, with the
partially dehydrated gypsum taking up water and forming the stable
hydrated state. At the end of the reaction, the investment slurry sets to a
solid mass of ‘green’ investment of sufficient strength.
The time between the first contact with water and solidification of the
investment is called the setting time. The setting time can be easily
determined by observing the time to ‘gloss off’, i.e. when the setting
process is almost complete, the glossy surface of the setting investment
(visible on the top of the flask) becomes dull and matte. The setting time
depends on:
- the brand/grade of the investment
- the mixing ratio, that is the powder to water ratio (e.g. 100:37
by weight)
82
- the temperature of the slurry
3
- the water quality (deionised, distilled or tap water which has a
variable hardness)
A typical value is 12 minutes; however, the actual values should be
checked in practice. Where the observed setting time differs significantly
from the correct (manufacturers recommended) value, difficulties can be
anticipated and defects are likely to occur on casting.
A prolonged setting time can be caused by:
- old (bad) batches of investment powder
- powder stored in a humid environment
- too much water added
- water and /or powder unusually cold
- changes in water quality
Result: investment is too weak. Cracks will occur, especially on edges of
heavier parts and with centrifugal casting, and the cast item exhibits fins
(Case 6). Melt can also penetrate into the surface of the weak
investment, resulting in the cast item showing a sandy surface (Case 6).
Alternatively, the slurry can separate and release the excess of water,
leading to the cast item showing water marks (Case 7).
A shortened setting time can be caused by:
- too little water (slurry is viscous)
- prolonged application of the vacuum in closed vacuum
investing equipment (water evaporates more readily)
- a high water temperature (e.g. in unusually hot weather)
- a change in water quality
Result: investment tends to crack during handling and fine details of
cast items are not well reproduced.
3.4.2 WORKING CYCLE

The working cycle comprises:
- Mixing
- Degassing
- Pouring into the flask
- Degassing the flask
- Vibrating the flask
The total time which can be used for performing this cycle is called the
working time. The working cycle should use almost the entire setting
time, less about one to two minutes.
Working time = Setting time minus one or two minutes
• If the cycle is completed a long time before setting occurs and the
flask is left alone without any movement, water will separate from
the slurry and collect on the surface of the waxes, causing typical
water marks (Case 7) on casting. The danger of the occurrence of
water marks increases with increasing water content.
• Where the working time is equal to or longer than the setting time,
cracks can be caused.
• During the setting process, the flask should be still and not be
83
3 touched or moved.
• Degassing should be extended until the foam collapses. (Where the
investment contains a significant detergent content, the foam does not
disappear completely. Experience is required to find the correct time.)
• Pouring the slurry into the flask has to be done as gently as possible,
not only to prevent breaking of fragile waxes but also to avoid uptake
of an excess of air. In any case, additional evacuation of the filled
flask is necessary to remove air from the setting slurry.
• Evacuation and subsequent vibration is essential to remove air bubbles
from the wax surfaces. If this procedure is not performed properly,
small spheres (‘pearls’) are visible on the surface of castings. Air
bubbles form voids in the investment and the voids are filled with
melt during casting.) This defect is named after its origin: air bubbles.
The procedure described above is typical for simple investing equipment
consisting of a mixer and an evaporation and vibration table. With the
more sophisticated closed automatic equipment, the danger of getting air
bubbles is minimised because the introduction of air is avoided as the
mixing and pouring is done in a vacuum. However, another problem may
occur: prolonged evacuation at a relatively high vacuum results in a
greater quantity of water being evaporated, with the result that the slurry
becomes too ‘sticky’ and the setting time is reduced. The investment can
crack. The water to powder ratio should be increased to compensate for
this enhanced loss (follow manufacturer’s recommendation).
Altogether, the available working time should be split in a relatively
long mixing period, with sufficient de-aeration and a moderate vibration
time. The splitting ratio is very dependent on the equipment and needs
some experience to optimise.
3.4.3 BURN-OUT CYCLE

Dewaxing
Dewaxing is also an important step to prepare the flask for a sound
casting. The wax can be removed simply by heating the inverted flask to
about 100°C. A great portion of the wax will flow out through the central
sprue. However, a portion remains in the cavities. Where the investment
is already relatively dry (by standing the flask for a long time between
investing and dewaxing), the wax tends to be absorbed into the
investment. The presence of water vapour, however, inhibits the uptake
of wax by the investment and ‘pushes’ the liquid wax out. The
subsequent burn-out process is considerably enhanced. A dry flask should
be soaked with water again before dewaxing.
A more efficient method for removing wax is steam dewaxing which
should be used wherever possible. The advantage of steam dewaxing is
not only reduced wax retention in the flask; it also decreases
environmental pollution (from combustion of the residual wax) and
protects the burn-out furnace.
Removing water
This should be completed at about 150°C (a temperature higher than
100°C is necessary to remove the free water) as hydrated gypsum loses a
part of its water in this temperature range.
84
Allow the flask to dry out for a sufficient time in the temperature range
3
150° - 250°C (furnace temperature), otherwise the investment will get a
rough surface from violently boiling water.
Transformation of cristobalite
Cristobalite (a form of silica) transforms to quartz at about 400°C, with a
consequent volume change. It is advised that this temperature should be
passed through gently to avoid cracking.
Decomposition of calcium sulphate
Pure calcium sulphate decomposes at about 1200°C. However, in the
presence of silica, the decomposition temperature is lowered towards about
780°C. Therefore, this temperature sets the highest temperature possible in
the burn-out cycle. For safety, the maximum furnace temperature should
not exceed 730 to 750°C. Other factors such as residual carbon can lower
the decomposition temperature even further (see below).
Removing residual wax
Wax absorbed by the investment tends to resist relatively high
temperatures until it is completely removed by combustion with oxygen.
Air circulation within the narrow channels of a flask is inhibited,
so adequate time needs to be given for wax residues to be fully
combusted (oxidised). Normally, the maximum temperature in the
burn-out cycle should be about 730°C, maintained for at least 2 hours
(actual temperature within the flask, see later). Where gem stones are
being cast ‘in situ’, such a high temperature cannot be applied without
damaging the stones. Any carbon or wax residues may lead to defects.
An indicator for an incomplete burn-out process is the presence of a
greyish zone in the investment around cast items when the investment
is broken up after casting.
Additional to a sufficiently high temperature, an excess of oxygen has
to be maintained. In a well insulated electrical furnace, overfilled with
flasks to the top, or in a gas furnace, with reducing flame gases allowed to
enter the muffle, this condition might not be fulfilled.
Low thermal conductivity of investment
Investment is a thermal insulator comparable with good insulating bricks,
etc. For this reason, the actual temperature within the flask lags
considerably behind the furnace temperature during the heating
programme. To ensure a uniform temperature within the flask at
maximum temperature and, later, at casting temperature, a minimum
time of 2 hours at set temperature should be maintained (i.e. the time
elapsing after the furnace has reached the set temperature). The time
necessary to equalise the temperature depends strongly on flask size,
furnace construction, filling of the furnace and position of the flask
within the furnace. Filling the furnace to the top should be avoided.
The flask should be positioned well away from the heating elements
and door openings to avoid overheating or too low a temperature,
respectively.
Conclusions
Heating flasks through the temperature range up to 400°C has to be done
very slowly; preferably, a holding time in the range 150° to 250°C should
be included. In principle, further heating up to the maximum
85
3 temperature can be done as fast as possible, consistent with the
temperature within the flask following the rising furnace temperature.
A maximum temperature of 730 - 750°C must not be exceeded.
For complete removal of carbon residues, a holding time of 2 hours at
maximum temperature should be included. After cooling down to the
working (casting) temperature -in most cases about 600°C - an
‘equalising’ time of two hours should be provided, too.
3.5 BEHAVIOUR OF GYPSUM-BONDED INVESTMENT

DURING CASTING
The most reactive component of the investment is gypsum (calcium
sulphate). As mentioned earlier, pure calcium sulphate decomposes at
about 1200°C to sulphur dioxide gas and calcium oxide, i.e.:
CaSO4 ➞ SO2 + CaO

Calcium sulphate ➞ sulphur dioxide (gas) + calcium oxide
The presence of silica in the investment lowers the decomposition

temperature substantially. Residual carbon from incomplete burn-out of
wax residues also acts to enhance the decomposition of gypsum,
effectively lowering the decomposition temperature significantly to the
critical value of about 800°C. Even worse, the base metal alloying
elements in jewellery alloys and their oxides can further enhance this
decomposition of gypsum. Their detrimental effect increases from copper
to zinc and nickel. Palladium as an alloying element is also critical.
Of course, the relative influence of an element depends on its
concentration in the alloy.
Another very critical factor is the casting atmosphere: a reducing
atmosphere enhances the instability of gypsum, too. From this reason,
casting in a closed chamber filled with forming gas (a mixture of nitrogen
and hydrogen) should be avoided. Covering open casting equipment
(crucible and flask) with forming gas may be advantageous if not too
much gas is supplied but difficulties can be expected.
Vacuum acts in the same way as a ‘reducing atmosphere’ and should be
avoided. In closed vacuum casting equipment, the pressure has to be
increased to atmospheric pressure (or even higher) with a neutral gas at
the moment of casting. The recommended atmosphere for casting is a
neutral gas such as nitrogen (preferable) or argon.
Not only gaseous compounds can be formed in the decomposition
reaction. Especially with reducing conditions, compounds between
sulphur, silver and base metals may be formed, leading to undesirable
inclusions, typically sulphides, in the casting, for example:
CaSO4 + 4Cu ➞ CuS + CaO + 3CuO

Calcium sulphate + copper ➞ copper sulphide + calcium oxide + copper oxide
Thus, decomposition of gypsum can occur at temperatures close to or at

the normal solidification temperatures of the alloys being cast. Hence, the
need to keep both flask and melt temperatures to a minimum consistent
86
with good mould filling. Temperature and heat transfer play an important
3
role. The higher the casting and/or flask temperature, the more likely is
the reaction of the melt with the investment. However, temperature alone
is not the only decisive factor. The weight of an item to be cast is
important, too. The quantity of heat transferred to the investment locally
depends on it and determines the temperature reached at the adjacent
surface of the investment. For example, it is possible to cast thin-walled,
nickel-white gold items satisfactorily in standard investment whereas
heavy cross-sectioned parts can cause serious problems.
A number of kinds of defect can be produced by the instability of
gypsum. The most obvious defect is spherical gas porosity at and just
beneath the surface (Case 1). The sulphur dioxide formed by the reaction
cannot escape and stays in the liquid metal, thus forming pores in the
cast metal. In most instances, the casting looks sound in the as-cast state,
but pores are exposed during subsequent polishing. Sometimes, the
porosity caused or favoured by sulphur dioxide shows a typical dendritic
structure, similar to shrinkage porosity (see Case 2) and so it is difficult to
determine it’s cause from shape alone.
Another critical case appears when sulphide inclusions are formed.
Sulphide inclusions are bluish-grey particles, only detectable in
microsections using a microscope with high magnification. In many
cases, the inclusions are randomly distributed within the structure as
roughly spherical particles, causing little harm. Sometimes however,
sulphides accumulate on grain boundaries and, in this instance,
severe embrittlement occurs. The jewellery item can break with the
application of very little force.
Another detrimental effect of sulphide inclusions is seen when
remelting contaminated scrap material as part of a new metal charge
since gas porosity can result. The remelting of used/scrap material - or
adding such used material to fresh alloy - can be a source of many
problems. Both sulphide-containing material as well as oxidised alloy can
produce gas porosity. The worst case is where scrap material (e.g. sprues)
contaminated with remnants of old investment are remelted: inclusions
and gas porosity are likely to result.
Some rules for avoiding gas porosity and other defects resulting
from reaction with gypsum bonded investment
• Keep the casting and flask temperatures as low as possible.
- This is especially important for the casting of heavy items. This
means also that casting filigree and heavy items on the same tree
should be avoided.
• Never cast in a reducing atmosphere (e.g. forming gas).
- A neutral atmosphere (e.g. nitrogen) is preferred.
• Phosphate-bonded investment is preferred for casting high melting
alloys with reactive alloying additions (e.g. palladium white gold, ‘990’
gold-titanium).
- Nickel-white gold is on the limit for being cast in gypsum-bonded
investment and only light/filigree items might be cast with success.
• Take extreme care if remelting used or scrap material.
- It must be thoroughly cleaned of remnants of old investment, oxides,
etc. and not contain sulphide inclusions.
87
FURTHER READING
There are relatively few good books on gold jewellery manufacturing that
approach the subject from the technology and underpinning science
rather than from a craft standpoint. Most are only available in English
language editions and some are out of print, unfortunately.
Much useful and up-to-date information on gold jewellery
manufacturing and materials technology and best practice has been
published in the gold jewellery technology journal, Gold Technology,
published by World Gold Council in several language editions, and in the
proceedings of the Santa Fe Symposia on jewellery technology, published
by Met-Chem Research Inc. Other useful conference proceedings and
books are published by the International Precious Metals Institute, USA,
and occasional technical articles appear in some of the trade press such as
American Jewelry Manufacturer, published by MJSA, USA.
Back issues of Gold Technology are available from local offices of World
Gold Council. World Gold Council also publishes a growing number of
jewellery technology publications that complement this Handbook and
which the reader will find invaluable. Increasingly, some of these are
becoming available translated into other languages as well as English
(contact your local World Gold Council office for information). At the
time of printing, these include:-
1. Technical Manual for Gold Jewellery - A practical guide to gold
jewellery manufacturing technology, published 1997
2. Investment Casting - A technical advisory manual for goldsmiths,
published 1995.
3. The Assaying and Refining of Gold - A guide for the gold jewellery
producer, published 1997.
4. Finishing Handbook - to be published 1998.
Information on these publications can be obtained from World Gold
Council local offices worldwide or direct to:- World Gold Council,
Industrial Division, 1st Floor, King’s House, 10 Haymarket, London SW1Y
4BP, UK. Tel: +44 171 930 5171; Fax: +44 171 839 4314; E-mail:
[email protected]. Internet: http://www.gold.org.
Some relevent articles are listed below.
Defects
1.“Examples of typical defects in jewelry casting”, Dieter Ott, Proc. Santa
Fe Symposium, 1989, 297-310.
2. “Defects in jewelry - a new version of an old problem”, Dieter Ott,
Proc. Santa Fe Symposium, 1991, 171-197.
3. “Analysis of common casting defects”, Dieter Ott, Gold Technology,
no.13, July 1994, 2-15.
4. “Control of defects in casting”, Dieter Ott, Gold Technology, no.17,
October 1995, 26-35.
5. “Stress corrosion cracking in white golds”, D.P.Argarwal, G.Raykhtsaum
and Joann DeRoner, Proc. Santa Fe Symposium, 1990, 257-262.
88
6. “Stress corrosion cracking failure of wrought gold jewelry alloys”, G.
Normandeau, Proc. Santa Fe Symposium, 1991, 323-352.
7. “Tarnish and corrosion of silver and gold alloys”, L.Gal-Or, Proc. Santa
Fe Symposium, 1990, 19-36.
8. “An overview of tarnish in dental and carat gold alloys”, Mark
Grimwade, Proc. Santa Fe Symposium, 1993, to be published.
9. “Residual stresses: their causes and how they can be minimised” (on
stress corrosion cracking), Mark Grimwade, Gold Technology, no.8,
November 1992, 9-13.
10. “Environmental effects on gold alloys and their resistance to
tarnishing and corrosion”, Mark Grimwade, Gold Technology, no. 9, May
1993, 13-17.
11. See also: “Basic metallurgy for goldsmiths - why you should know
something about metallurgy & - melting, alloying and casting”, Mark
Grimwade, Gold Technology, no.2, June 1990, 2-4 & 5-10.
Investment Casting
12. “Gold jewellery investment casting”, Theo Groenewald, Gold Bulletin,
13(2), 1980, 80-81.
13. “Investment casting of gold jewellery”, Dieter Ott & Christoph J.Raub,
Gold Bulletin, 18(2), 1985, 58-68.
14. “Investment casting of gold jewellery - porosity in casting”, Dieter Ott
and Christoph J.Raub, Gold Bulletin, 18(3), 1985, 98-108.
15. “Investment casting of gold jewellery - surface properties”, Dieter Ott
and Christoph J.Raub, Gold Bulletin, 18(4), 1985, 140-143.
16. “Investment casting of gold jewellery - temperature changes occurring
in metal and investment”, Dieter Ott and Christoph J.Raub, Gold Bulletin,
19(1), 1986, 2-7.
17. “Investment casting of gold jewellery - factors affecting the filling of
moulds”, Dieter Ott and Christoph J.Raub, Gold Bulletin, 19(2), 1986, 34-
39.
18. “Casting as a total system”, Larry Diamond, Proc. Santa Fe
Symposium, 1989, 235-240.
19. “Temperature gradient casting”, Larry Diamond, Proc. Santa Fe
Symposium, 1991, 225-239.
20. “Casting - gas pressure effects”, Dieter Ott and Christoph J.Raub,
Gold Technology, no.7, July 1992, 10-17.
21. “Casting - surface properties”, Dieter Ott and Christoph J.Raub,
Gold Technology, no.7, July 1992, 28-31.
22. “Casting - porosity, causes and prevention”, Dieter Ott, Christoph
J.Raub and W.S.Rapson, Gold Technology, no.7, July 1992, 18-27.
23. “Porosity in investment casting”, Dieter Ott, Gold Technology, no. 11,
November 1993, 15-20.
24. “Wax elimination, burnout and the mould’s effect on porosity in
casting”, Eddie Bell, Gold Technology, no.11, November 1993, 21-27.
25. “Shrinkage porosity in investment casting: a consideration of the
factors affecting its formation”, Dieter Ott, Gold Technology, no. 13, July
1994, 16-21.
89
26. “Chaos in casting: an approach to shrinkage porosity”, Dieter Ott,
Proc. Santa Fe Symposium, 1996, 383-406 and Gold Bulletin, 30(1), 1997,
13-19.
27. “The effect of investment and metal casting temperatures on the
quality of castings”, Greg Normandeau, Proc. Santa Fe Symposium, 1990,
209-256.
Mechanical Working
28. “Cold working and annealing of carat gold jewelry alloys”, Alex
Langford, Proc. Santa Fe Symposium, 1990, 349-374.
29. “Basic metallurgy for goldsmiths - working and annealing”, Mark
Grimwade, Gold Technology, no.2, June 1990, 17-22.
90
ACKNOWLEDGEMENTS
The co-operation of many jewellery manufacturers in providing defective
jewellery for investigation is gratefully appreciated; they are too
numerous to name individually. Without such assistance and willingness
to expose their problems, this Handbook would not be possible.
The author especially thanks the technical committee of the Santa Fe
Symposium and World Gold Council for their support and
encouragement to undertake the defects project and to publish the
results. Thanks are also given to Eddie Bell and Valerio Faccenda for their
contributions and especially to Christopher Corti for his many useful
suggestions and the editing of the manuscript. Lastly, he would like to
thank his colleagues at FEM for their support and practical assistance.
91

Handbook of Casting Defects

Uploaded by

Document Information

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

Handbook of Casting Defects

Uploaded by

Copyright:

Available Formats

HANDBOOK ON CASTING

AND OTHER DEFECTS

Publication Date: November 1997

Published by World Gold Council, Industrial Division,

Produced by Dieter Ott, FEM, Schwäbisch Gmünd, Germany

Part A: Survey of Defect Types 9

Part B: Case Histories: Description of Defects 15

2 Ingot Casting and Continuous Casting 45

3 Defects caused by Alloy Composition 49

4 Corrosion, Tarnishing, Discolouration 57

Part C: Basic Aspects of the Metallurgy of Carat Golds

Investment Casting, Remelting of A

Pores and Remelting of B

Investment Casting, Polluted crucible, A

Defects caused by investment or investing process

Investing Wrong investing C

Casting, alloy Low melting Yellow gold with A

Brittle Casting in reducing Reaction with Jewellery alloys B

Heat treatment, Age hardening in Red and pink C

Investment Shrinkage General, 14ct alloys A

Investment casting is a very useful technique in medium- to large-scale

Description of the Defect

Recommendation for Avoidance

Description of the Defect

Recommendation for Avoidance

Description of the Defect

Extended Explanation and Further Reading

Recommendation for Avoidance

Description of the Defect

and vacuum assisted method. An important detail is the ‘gating’ of the

Such a pronounced dendritic porous structure drastically reduces the

More about gating

Recommendation for Avoidance

Recommendation for Avoidance

Check burn-out cycle

Recommendation for Avoidance

Recommendation for Avoidance

In most cases, the defect is restricted to a small surface zone, as in this

Recommendation for Avoidance

CASE 11: PORES AND INCLUSIONS ON THE SURFACE - CAUSED BY

Description of the Defect

Recommendation for Avoidance

Recommendation for Avoidance

Description of the Defect

Recommendation for Avoidance

2.1 CRACKS AND MATERIAL SEPARATION

Recommendation for Avoidance

Recommendation for Avoidance

3.1 LOW MELTING IMPURITIES

Description of the Defect

Fig. 17.10 200 x

Recommendation for Avoidance

Description of the Defect

Recommendation for Avoidance

Description of the Defect

RECOMMENDATION FOR AVOIDANCE

Key Words Fig. 21.1

Recommendation for Avoidance

Description of the Defect

Fig. 23.1 Fractured 14ct gold ring Description of the Defect

Recommendation for Avoidance

Fig. 24.1 Broken prong on a white gold

Recommendation for Avoidance

Extended Explanation and Further Reading