Professional Documents
Culture Documents
RF Dynam Pro Manual en
RF Dynam Pro Manual en
January 2020
Add-on Module
RF-DYNAM Pro
Natural Vibration Analysis, Response
Spectra, Time History, Equivalent Static
Forces
Program Description
Contents
Contents Page
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1 Add-on Module RF-DYNAM Pro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Using the Manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Opening the RF-DYNAM Pro Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2. Input Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1 Structure of Main Tab Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Mass Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.1 Mass Case Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.2 Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.3 Sum of Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.4 Global Settings for Mass Import in Details . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.5 Additional Nodal Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.6 Additional Line, Member and Surface Masses . . . . . . . . . . . . . . . . . . . . 16
2.3 Mass Combinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Natural Vibration Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.1 Number of Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4.2 Scaling of Mode Shapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4.3 Acting Masses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.4.4 Type of Mass Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.4.5 Eigenvalue Solver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4.6 FE Mesh Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.7 Stiffness Modifications as Initial Conditions . . . . . . . . . . . . . . . . . . . . . . 22
2.5 Response Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.5.1 Response Spectra According to Building Standards . . . . . . . . . . . . . . 24
2.5.2 User-Defined Response Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.5.3 Response Spectra Generated from Accelerograms . . . . . . . . . . . . . . . 25
2.6 Accelerograms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.6.1 Library of Earthquake Recordings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.6.2 User-Defined Acceleration-Time Diagrams . . . . . . . . . . . . . . . . . . . . . . . 28
2.7 Time Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.7.1 Transient Time Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.7.2 Periodic Excitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.7.3 Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.8 Dynamic Load Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.9 Dynamic Load Cases - Response Spectrum Analysis . . . . . . . . . . . . . . 33
2.9.1 Assign Supports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.9.2 Assign Response Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.9.3 Combination of Modal Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.9.4 Combination of Directional Components. . . . . . . . . . . . . . . . . . . . . . . . . 35
2.9.5 Export Result Combinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.9.6 Damping for CQC rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.9.7 Mode Shape Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.10 Dynamic Load Cases - Time History Analysis . . . . . . . . . . . . . . . . . . . . . . 37
2.10.1 Time History Analysis of Accelerograms . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.10.2 Time History Analysis of Time Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.10.3 Linear Solvers: Modal Analysis versus Implicit Newmark . . . . . . . . . . 40
1
Contents
2
Contents
3
1 Introduction 1
1 Introduction
1.1 Add-on Module RF-DYNAM Pro
Whether you are a beginner or already an expert user working with one of the previous RF-DYNAM
versions, everybody can quickly learn how to use the new add-on module. All the valuable hints
from our customers telling us about their everyday experience helped us to develop and improve
this add-on module. There are many new features available, which improves and enhances the
possibilities of a dynamic analysis.
The RF-DYNAM Pro module is split into four parts: The add-on module RF-DYNAM Pro - Natural
Vibrations is the basic module that performs natural vibration analyses for member, surface and
solid models. A multi-modal and multi-point response spectra analysis and a linear time history
analysis of the given structure can be performed with the module RF-DYNAM Pro - Forced Vibrations.
A nonlinear time history analysis is available in the add-on module RF-DYNAM Pro - Nonlinear Time
History. The add-on module RF-DYNAM Pro - Equivalent Loads covers the equivalent static force
analysis (based on the multi-modal response spectra analysis) in accordance to various building
standards.
We hope you enjoy working with RF-DYNAM Pro.
Your DLUBAL team
Natural Vibrations
RF-DYNAM Pro - Natural Vibrations determines the lowest eigenvalues of the model. The number
of the eigenvalues can be adjusted. Masses are directly imported from load cases or load combi-
nations (with the option to import the total mass or only the Z-component of loads). Additional
masses can be defined manually at nodes, lines, members or surfaces. Furthermore, you can
influence the stiffness matrix by importing axial forces or stiffness modifications of a load case or
a load combination.
The main features are listed below:
• Automatic consideration of masses from self-weight
• Direct import of masses from load cases or combinations
• Optional definition of additional masses (nodal, line, member, and surface masses as well as
mass moments of inertia)
• Combination of masses in different mass cases and mass combinations
• Pre-set combination factors according to EN 1998-1 CEN
• Masses can be neglected on arbitrary parts of the model
• Optional consideration of normal forces to modify the geometric stiffness matrix
• Stiffness modification (for example, you can import deactivated members or stiffnesses from
RF-CONCRETE)
• Consideration of failed supports or members possible as initial conditions
• Definition of several natural vibration cases possible (for example to analyse different masses
or stiffness modifications)
• Output of eigenvalue, angular frequency, natural frequency and period
• Determination of mode shapes and masses in FE mesh points
• Output of modal masses, effective modal masses, and modal mass factors
4
1 Introduction 1
• Visualization and animation of mode shapes
• Different scaling options for mode shapes
• Documentation of numerical and graphical results in the printout report
Four powerful eigenvalue solvers are available in RF-DYNAM Pro - Natural Vibrations:
• Root of the characteristic polynomial
• Lanczos method
• Subspace iteration
• ICG iteration method (Incomplete Conjugate Gradient)
The selection of the eigenvalue solver primarily depends on the size of the model.
After the calculation, the eigenvalues, natural frequencies and periods are listed. These result
tables are embedded in the main program RFEM. The mode shapes of the structure are tabulated
and can be displayed graphically or as an animation. All result tables and graphics are part of the
RFEM printout report. Moreover, exporting the tables to Excel is possible.
Forced Vibrations
RF-DYNAM Pro - Forced Vibrations is an extension of the add-on module RF-DYNAM Pro - Natural
Vibrations. Mechanical structures that are excited by transient or harmonic force-time or accel-
eration-time diagrams can be analysed using the modal analysis or the linear implicit Newmark
solver. Furthermore, a multi-modal and multi-point response spectra analysis can be performed.
The required spectra can be created according to the standards or individually. The add-on mod-
ule contains an extensive library of accelerograms from earthquake zones. They can be used to
generate response spectra.
The features of the time history analysis are listed below:
• Combination of user-defined time diagrams with load cases or load combinations (nodal,
member and surface loads as well as free and generated loads can be combined with functions
varying over time)
• Combination of several independent excitation functions possible
• Extensive library of earthquake recordings (accelerograms)
• Modal analysis based on the eigenvalue analysis or direct integration with the implicit New-
mark solver in the linear time history analysis available
• Modal analysis:
- Direct import of initial deformations from a load case possible
- Stiffness modifications from NVCs are taken into account
• Implicit Newmark solver:
- Independent of eigenvalue analysis
- The member type Dashpot provides the possiblity to model local linear viscous damping
elements with parallel connected linear spring elements (Kelvin-Voigt-Model).
• Structural damping using the Rayleigh damping coefficients or the Lehr's damping values
• Graphical result display in a time course monitor
• Export of results in user-defined time steps or as an envelope
5
1 Introduction 1
The features of the response spectra analysis are listed below:
• Response spectra of numerous standards (EN 1998-1 [1], DIN 4149 [2], IBC 2012 [3] etc.)
• Response spectra can be user-defined or generated from accelerograms
• Direction-relative response spectra approach
• Different response spectra can be assigned to different supports (multi-point option)
• Relevant mode shapes for the response spectra can be selected manually or automatically
(the 5% rule from EN 1998-1 [1] can be applied)
• Calculation is performed within RF-DYNAM Pro and is therefore linear
• Modifications of the stiffness matrix as defined in the Natural Vibration Cases are also used for
the calculation of internal forces and deformations
• Combination of the modal responses (SRSS rule or CQC rule) and combination of the results
from different excitation directions (SRSS or 100% / 30% rule)
The results from the time history analysis are displayed in a time course monitor. Here, you have
the possibility to superimpose different nodes or positions within one member. All results are
displayed as a function of time. The numeric values can be exported to MS Excel. In case of a time
history analysis, you can export results of a single time step or filter the most unfavourable results
of all time steps.
As the response spectra analysis is based on the eigenvalues and mode shapes and the internal
forces, deformation and further results are calculated internally within RF-DYNAM Pro, the calcula-
tion is purely linear, but the stiffness modifications as defined in the Natural Vibration Cases are
used for the whole analysis. Only result combinations are exported. A combination of the modal
responses and a combination of the results due to the components of the earthquake action are
done internally.
The input data of RF-DYNAM Pro and the exported load cases and result combinations are part of
the printout report.
6
1 Introduction 1
• Supported nonlinearities:
- Types of members (tension, compression members, cables)
- Member nonlinearities (failure, tearing or yielding under tension/compression, plastic hinge)
- Support nonlinearities (failure, friction, diagram, partial activity)
- Hinge nonlinearities (fixed options, friction, diagram, partial activity)
- Nonlinear material models
• The member type Dashpot provides the possibility to model local linear viscous damping
elements with parallel connected spring elements (Kelvin-Voigt-Model).
• Calculation independent of eigenvalue analysis
• Definition of structural damping via the Rayleigh damping coefficients
• Graphical result display in a time course monitor
• Export of results in user-defined time steps or as an envelope
The results from the nonlinear time history analysis are displayed in a time course monitor. All
results are displayed as a function of time. Results of single time steps or dynamic envelopes of all
time steps are available. All results can be exported to MS Excel.
The input data of RF-DYNAM Pro and the exported load cases and result combinations are part of
the printout report.
Equivalent Loads
RF-DYNAM Pro - Equivalent Loads is an extension of the RF-DYNAM Pro - Natural Vibrations add-on
module. With this module, seismic analyses can be performed with the multi-modal response
spectrum analysis. The required spectra can be created according to the standards or user-defined,
from which the equivalent static loads are generated.
The main features are listed below:
• Response spectra of numerous standards (EN 1998-1 [1], DIN 4149 [2], IBC 2012 [3] etc.)
• Input of user-defined response spectra
• Direction-relative response spectra approach
• Relevant mode shapes for the response spectra can be selected manually or automatically
(the 5% rule from EN 1998-1 [1] can be applied)
• Generated equivalent static loads are exported to load cases, separately for each mode and
direction
• The calculation of these load cases is performed in RFEM. Thus, a non-linear calculation can
be performed.
• Stiffness modifications that are applied in the Natural Vibration Cases are not automatically
applied in the Load Cases
• Combination of the modal responses (SRSS or CQC rule) and combination of the results from
different excitation directions (SRSS or 100% / 30% rule)
• Results envelopes with maximum and minimum results or signed result in accordance to a
dominant mode shape
Equivalent static loads are generated separately for each relevant eigenvalue and excitation direc-
tion. They are exported to static load cases, and a static analysis is performed in RFEM. Those load
cases are quadratically combined in result combinations, followed by the combination of different
excitation directions.
The input data of RF-DYNAM Pro and the exported load cases and result combinations are part of
the printout report.
7
1 Introduction 1
1.2 Using the Manual
Topics like installation, graphical user interface, evaluation of results, and printout are described in
detail in the manual of the main program RFEM. This manual, however, focuses on typical features
of the RF-DYNAM Pro add-on module.
The sequence and structure of the manual follows the input and results windows of the module.
In the text, the buttons are given in square brackets, for example [Edit]. At the same time, they
are pictured on the left. The expressions that appear in dialogue boxes, windows, and menus are
set in italics.
At the end of the manual, you find the index. If you do not find what you are looking for, you can
use the search option for our Knowledge Base to find a solution among the articles on dynamics.
The FAQ pages are also highly recommended. Furthermore, you can watch the recorded Webinars
and study the Verification Examples.
Menu
To open the add-on module, select
Add-on Modules → Dynamic → RF-DYNAM Pro
on the RFEM menu.
If RF-DYNAM Pro was opened before and is your current module, you can also select
Add-on Modules → Current Module.
8
1 Introduction 1
Navigator
Alternatively, you can open the add-on module in the Data navigator via
Add-on Modules → RF-DYNAM Pro.
When you right-click on the add-on module to open its shortcut menu, you can add RF-DYNAM
Pro to your Favourites (see Figure 1.2).
Panel
If RF-DYNAM Pro results are already available, you can also open the add-on module via the panel.
To display the panel, select the RF-DYNAM Pro case in the drop-down menu of the main program
RFEM and make the results visible with the [Show Results] button. If the panel is not shown, you
can activate it by clicking the [Panel] button.
Use the [RF-DYNAM Pro] button in the panel to re-open the add-on module.
9
2 Input Data 2
2 Input Data
When you start the add-on module, a new window opens. The window is organized in several
tabs and sub-tabs which you should go through from left to right when you enter data for the first
time. Not all tabs are shown right from the beginning, some tabs belong to special settings that
appear as soon as you have selected the corresponding check boxes.
The first window that appears when you open RF-DYNAM Pro is shown in Figure 2.1.
In the General tab, you decide which of the add-on modules of RF-DYNAM Pro you want to activate.
You can analyse natural frequencies and mode shapes of the structural system with RF-DYNAM Pro -
Natural Vibrations. With RF-DYNAM Pro - Forced Vibrations, you can perform a multi-modal response
spectrum analysis or a linear time history analysis. A time history analysis under consideration of
the defined nonlinearities in your model is possible with RF-DYNAM Pro - Nonlinear Time History.
The module RF-DYNAM Pro - Equivalent Loads offers you a multi-modal response spectrum analysis
with the export of equivalent static forces.
Add-on modules that were not purchased can be opened, but they run only as a demo version.
You can also activate a 90-day trial version of those add-on modules.
Natural Vibrations
This option is selected by default because a natural vibration analysis of the structure is required
in most cases. The tabs Mass Cases and Natural Vibration Cases belong to this option by default.
When you select the Mass combination check box, the corresponding tab appears (see Figure 2.1).
10
2 Input Data 2
Forced Vibrations
With this module, you can either perform a multi-modal response spectrum analysis or a linear
time history analysis. When you select both (Response spectra and Accelerations), you can generate
response spectra from an acceleration-time diagram. The Response Spectra tab appears when
Response spectra option is selected, the Accelerograms tab appears when Accelerations option is
selected. The Time Diagrams tab appears when the Time diagrams option is selected. The Dynamic
Load Cases tab is available for all three options of the RF-DYNAM Pro - Forced Vibrations module.
Equivalent Loads
This option allows the generation of equivalent static forces in accordance with various design
standards (EN 1998-1 [1], IBC2012 [3] and many others). The Response Spectra and the Dynamic
Load Cases tabs belong to the module RF-DYNAM Pro - Equivalent Loads.
11
2 Input Data 2
In the right part of the main window, you find the number of the selected case together with the
case description at the top (marked by a blue box in Figure 2.2). In the description box, you can
enter a case description manually or choose one from the drop-down list. Below you find the main
entering area for your input data which firstly opens in the General sub-tab. Some more sub-tabs
may appear depending on the selected check boxes.
In the General sub-tab, you can enter a comment. This space is marked by a green box in Figure 2.2.
Figure 2.2: Module Window Mass Cases with General tab – MC2 is selected and all four options to define
additional masses are selected to show the sub-tabs
The mass case number is set automatically and cannot be edited. If a mass case is deleted later,
the numbers do not change. You can enter a description manually, or you can choose one from
the drop-down list.
2.2.2 Masses
Select the From self-weight of structure check box when you want to include the self-weight inde-
pendent of any load case defined in RFEM. This option is shown in Figure 2.2.
You can import masses from a load case or load combination defined in RFEM by selecting the
From force components check box. Choose the relevant Load case or Load combination from the
12
2 Input Data 2
lists. Load types like forces and moments will be imported. Additionnally, the types Pipe content -
full and Pipe content - partial can be converted into masses.
The factor to modify loads of load cases, which can be set in the calculation parameters of RFEM
(Edit → Load Cases and Combinations), is not applied in RF-DYNAM Pro. To increase the masses
by a factor you have to employ Mass Combinations described in Section 2.3.
When you import a load case that contains self-weight of the model, make sure that the From
self-weight of structure check box in this tab is cleared. Otherwise you will double the self-weight
of the model.
In Section 2.2.4, global settings in RF-DYNAM Pro are described that influence the mass import
from Load Cases.
In addition, or as an alternative to the previously described options to import masses, it is possible
to define nodal, line, member, or surface masses. Depending on the check boxes you select,
additional sub-tabs will appear as shown in Figure 2.2. The settings within those sub-tabs are
detailed in Section 2.2.5 and Section 2.2.6.
The third option imports full loads regardless their direction. This is the only way to import masses
from horizontal earth pressure.
When different loads are acting on one FE-node, the load resultant is built first. It is then imported
to RF-DYNAM Pro.
The direction of gravity is defined in the direction of the global Z-axis when the positive orientation
is defined downwards. It is defined against the direction of the global Z-axis when the positive
orientation is defined upwards. The settings shown on the left can be found in the Edit → Model
Data → General Data menu of RFEM.
13
2 Input Data 2
Neglect Masses
It is possible to neglect masses from parts of the model. For this, select the Neglect masses option
as seen in Figure 2.4.
With the button, you open the Neglect Masses dialogue box (see Figure 2.5). There you can
define which masses of nodes, members, lines and surfaces are to be neglected.
Figure 2.5: Dialogue box Neglect Masses for masses on nodes, members, lines, and surfaces
You can enter a list of nodes (members, lines and surfaces) manually or pick them graphically by
using the button. The filter button [Select supports only] turns blue when activated; it is useful
for the selection of supports. If this filter is active, you can select the whole model, but only the
supported nodes (members, lines or surfaces) will be selected.
You can decide which directions of masses are to be neglected by selecting the relevant check
boxes for displacements uX , uY and uZ and rotations 𝜑X , 𝜑Y and 𝜑Z .
In the Sum of Masses section of the Mass Case tab (see Figure 2.2), the neglected masses are not
visible. There always the full masses are shown. The neglected masses on the single FE-nodes can
be checked in the result table 5.6 Masses on Mesh Points (see Section 4.1).
You can find an example of the neglecting masses feature in the Knowledge Base - 001222.
14
2 Input Data 2
2.2.5 Additional Nodal Masses
The Nodal Masses tab is shown in Figure 2.6. It is available when the Manually define additional
masses at nodes option has been selected in the General tab.
You can enter a list of nodes manually or use the function to select the relevant nodes in the
work window.
The masses m [kg] can be defined manually. They act in the directions defined in the Natural
Vibration Cases (see Section 2.4.3).
In addition, the Mass moment of inertia IX , IY , and IZ can be defined to model more complex mass
points (i.e. rotation of a machine can approximately be considered).
The buttons beneath the table provide common table functions as described in the RFEM manual,
Section 11.5.
The table entry can be stored in a library and can be opened whenever needed. The [Save] button
opens a dialogue box where you can enter a file name.
15
2 Input Data 2
2.2.6 Additional Line, Member and Surface Masses
In addition to nodal masses, you can define line, member or surface masses manually. The cor-
responding sub-tabs appear when the check boxes for additional masses at Lines, Members or
Surfaces have been selected in the Mass Cases tab (see Figure 2.2). The table to enter additional
line masses is shown in Figure 2.7.
As the tables to enter member and surface masses are very similar, they are not explicitly shown
here.
A list of lines (members or surfaces, respectively) can be entered manually or by using the
function which makes it possible to select the objects graphically in the work window.
Line and member masses are provided in [kg] per unit length. Surface masses are given in [kg] per
unit area. They can be defined manually and act in the directions defined in the Natural Vibration
Cases (see Section 2.4.3).
16
2 Input Data 2
2.3 Mass Combinations
Mass cases can be combined to mass combinations. This is done in analogy to load cases and load
combinations in RFEM as described in the RFEM manual in Section 5.5.1. The Mass Combinations
tab only exists when the corresponding check box has been selected in the General tab (see
Figure 2.1). When you open the Mass Combinations tab for the first time, a mass combination is
pre-set and the existing mass cases are listed with the combination factors set by default.
The module window with mass combination MCO1 containing self-weight and imposed load mass
cases is illustrated in Figure 2.8.
Figure 2.8: Module Window Mass Combinations with all existing mass cases selected for mass combination
You can add selected or all mass cases to a mass combination by using the buttons shown on
the left. By doing so, the mass cases move from the left to the right list. Combination factors are
pre-set automatically by RF-DYNAM Pro, but can be changed manually.
Similarly, you can remove single or all mass cases from a mass combination by using the buttons
shown on the left.
The combination factors are pre-set in accordance to EN 1990 [4] and EN 1998-1 [1]. Those factors
can be adjusted manually by entering a value or choosing a value from the drop-down list. As
regulated in [1] Section 3.2.4, additional masses beside the self-weight have to be considered to
calculate inertia effects.
∑ Gk,j “+” ∑ 𝜓E,i ⋅ Qk,i (2.1)
where Gk,j are the permanent loads and Qk,i any imposed load. 𝜓E,i are the combination factors
for the imposed loads defined as
𝜓E,i = 𝜑 ⋅ 𝜓2,i (2.2)
where 𝜓2,i are regulated in EN 1990 Table A.1.1 and 𝜑 are provided in EN 1998-1 Section 4.2.4.
The combination factors may be regulated differently in each National Annex of EN 1998-1 or
other international building standards.
17
2 Input Data 2
2.4 Natural Vibration Cases
The Natural Vibration Cases tab is the centrepiece of the RF-DYNAM Pro - Natural Vibrations
module. It is essential for the response spectrum analysis and the time history analysis based on
modal decomposition (modal analysis).
In this tab, you set how many eigenvalues you want to calculate, define which masses are used and
in which direction they act. You also set the eigenvalue solver, the type of mass matrix and define
how the mode shapes are scaled. This is also the place where you define stiffness modifications or
import axial forces as an initial condition.
The General tab of the Natural Vibration Case window is shown in Figure 2.9.
Figure 2.9: Module Window Natural Vibration Cases with General tab
For each natural vibration case NVC, you can set various calculation parameters as shown in
Figure 2.10.
The number of the natural vibration case is set automatically and cannot be edited. When a case
is deleted later, the numbers do not change. You can decide whether the specific NVC is to be
calculated or not by selecting or clearing its To Solve check box.
You can calculate each NVC separately by using the shortcut menu and apply Calculate. The colour
of a NVC is grey when no results of this natural vibration case are available. It turns green as soon
as the calculation is finished and results are available. For more information, see Section 3.3.
18
2 Input Data 2
Figure 2.10: Module Window Natural Vibration Cases with Calculation Parameters tab
19
2 Input Data 2
2.4.3 Acting Masses
For each natural vibration case (NVC), a Mass case or Mass combination can be imported. Select it
from the respective list in the Calculation Parameters sub-tab (see Figure 2.10).
You have to define the direction in which the masses are acting. The masses act in the global
translational X, Y, or Z-direction when you activate the corresponding check boxes, or rotationally
about the global X, Y, and Z-axes.
Those settings change your mass matrix: You will get different resulting mode shapes and frequen-
cies. To perform a planar calculation of your structure, consider only masses acting in one global
direction at a time. The planar simplification requires a regular structure. For a three-dimensional
analysis, consider masses acting in all global directions.
Diagonal matrix
When the type of the mass matrix M is chosen to be diagonal, the masses are lumped to the
FE-nodes. The entries in the matrix are the lumped masses in the translational directions X, Y, and
Z. Masses rotating about the X, Y or Z-axis are considered depending on the type of diagonal
matrix. There are three different types of diagonal mass matrices available:
- Diagonal matrix (translational DOFs)
Only masses in the translational directions X, Y, and Z are considered. The check boxes about axis
are not available. The diagonal mass matrix M is structured as follows:
M = diag (M1,X ,M1,Y ,M1,Z ,M2,X ,⋯ ,Mn,j ,⋯ ) (2.3)
where X, Y, or Z are the distances to the centre of total mass provided in the Mass Case tab described
in Section 2.2. The manually defined mass moments of inertia IX , IY and IZ as provided in the Nodal
Mass table (see Section 2.2.5) are also considered when acting in longitudinal direction.
- Diagonal matrix (translational and rotational DOFs) This type of mass matrix considers all
degrees of freedom. The masses in the translational directions X, Y, and Z and rotational directions
X, Y, and Z are taken into account. The diagonal mass matrix M on a specific FE-point with mass m
is structured as follows:
M = m ⋅ diag (1,1,1, (Y 2 + Z 2 ) + IX , (X 2 + Z 2 ) + IY , (X 2 + Y 2 ) + IZ ) (2.5)
where X, Y, or Z are the distances to the centre of total mass (provided in the Mass Case tab
described in Section 2.2), and IX , IY , or IZ are the mass moments of inertia (defined in the Nodal
Mass table described in Section 2.2.5).
20
2 Input Data 2
Consistent Mass Matrix
The consistent mass matrix is a full mass matrix of the finite elements. Thus, the masses are not
simply lumped to the FE-nodes. Shape functions are used instead for a more realistic distribution
of the masses within the FE-elements. With this mass matrix, non-diagonal entries in the matrix
are considered, which means that mass rotation in general is taken into account. The structure of
the consistent mass matrix is as follows (neglecting the shape functions here for simplicity):
1 0 0 0 0 0
⎡ ⎤
⎢0 1 0 0 0 0 ⎥
⎢0 0 1 0 0 0 ⎥
M=m⋅⎢ (2.6)
⎢0 0 0 Y2 + Z2 −XY −XZ ⎥ ⎥
⎢0 0 0 −XY X2 + Z2 −YZ ⎥
⎣0 0 0 −XZ −YZ X2 + Y2⎦
where the mass m and the distances X, Y, or Z refer to the centre of total mass given in the Mass
Case tab described in Section 2.2.
Unit Matrix
The unit matrix overwrites all the previously defined masses. A unit matrix is a consistent matrix
with diagonal members of 1 kg. By using the Unit matrix, the mass at each FE-point is set to 1 kg.
Translations and rotations of the masses are considered. This mathematical approach should only
be used for numerical analyses. More details about matrix types and especially the use of the unit
matrix are provided by Barth and Rustler [5].
Lanczos
The method by Lanczos is an iterative method to determine the p lowest eigenvalues and cor-
responding mode shapes of large models. In most cases, this algorithm allows to reach a quick
convergence. It is possible to calculate n-1 eigenvalues (n = degrees of freedom of the system).
For further details see Bathe [6].
Subspace Iteration
The subspace iteration is appropriate for large FE-models where you want to calculate only a few
eigenvalues. All required eigenvalues are determined in one step. The computer memory limits
the number of eigenvalues that can be calculated by this method. For further details see Bathe [6].
ICG Iteration
This method is also suitable for large models where only a few eigenvalues are required. But here
the eigenvalues are calculated successively. Therefore, the number of required eigenvalues is
proportional to the computing time. Theoretically, all eigenvalues of a model can be calculated.
21
2 Input Data 2
2.4.6 FE Mesh Settings
The button shown on the left links directly to the FE mesh settings of the main RFEM program. The
FE Mesh Settings dialogue box is also available under Calculation → FE Mesh Settings in RFEM.
Further details about the FE mesh and the parameters that can be adjusted are explained in the
RFEM manual, Section 7.2.
Members that are not integrated into surfaces are not divided into FE-nodes by default. When a
diagonal mass matrix is used, the mass of the member is lumped to the two end nodes. When
those end nodes are supported, you might not be able to calculate eigenvalues. In this case, you
need to activate the member division in the FE mesh settings.
The calculation parameters of a load case or load combination that are relevant for RF-DYNAM Pro
are shown on the left hand side. The check boxes Materials, Cross-sections, Members, and Surfaces
Calculation parameters activate the modifications made in the material, cross-section, member, and surface dialogue
of LC or CO in RFEM boxes in the main program RFEM (see RFEM manual, Sections 4.3, 4.13, 4.17, and 4.4). Further
details on the calculation parameters of load cases are given in the RFEM manual, Section 7.3.
Global stiffness modifications can be imported from RFEM independently of any LC or CO. Those
global modifications import the material partial factor 𝛾M , cross-section, member and surface
modifications that are changed in RFEM (see RFEM manual in Sections 4.3, 4.13, 4.17, and 4.4).
In RF-DYNAM Pro, those stiffness modifications are initial conditions. The calculation within RF-DY-
NAM Pro - Natural Vibrations is fully linear. You can activate the initial conditions as shown in
Figure 2.10.
When you want to consider axial forces only, make sure that no stiffness modifications are activated
in the calculation parameters of the imported LC or CO.
When you want to study deactivated members only, make sure you create a load case containing
no loads but the stiffness modifications you are interested in.
A figure describing the complexity of this feature is provided in the Knowledge Base - 001023.
22
2 Input Data 2
2.5 Response Spectra
A response spectrum is a plot of the maximum peak response to a specified input illustrated
usually versus the natural period of single degree-of-freedom (SDOF) oscillators. It is produced by
calculating the response to a specific input (i.e. average of several earthquake motions) for a family
of SDOF oscillators each having a different natural period but the same damping value. There
are computational advantages in using the response spectra method and it is a very common
method described in building standards. But you should be aware that it is only an approximate
method that calculates maximum internal forces of your system. For further details on the response
spectrum analysis, see for example Wilson [8] and Tedesco [9].
RF-DYNAM Pro offers the multi-modal and multi-point (only with RF-DYNAM Pro - Forced Vibrations)
response spectrum method. Ready to use response spectra curves are available in building stan-
dards, many of those are implemented in RF-DYNAM Pro. RF-DYNAM Pro can generate a response
spectrum from any given accelerogram (only available with RF-DYNAM Pro - Forced Vibrations).
The Response Spectra tab is only available when the Response spectra or the Equivalent loads options
have been selected in the General tab shown in Figure 2.1. This tab belongs either to the add-on
module RF-DYNAM Pro - Forced Vibrations or to the module RF-DYNAM Pro - Equivalent Loads.
In RF-DYNAM Pro, you have three options to enter response spectra: according to a building
standard, user-defined, or generated from an accelerogram. The Response Spectra tab is illustrated
in Figure 2.11.
Figure 2.11: Module Window Response Spectra with General tab – option According to Standard selected,
showing Code Parameters
23
2 Input Data 2
2.5.1 Response Spectra According to Building Standards
The parameters according to a building standard can be entered to generate a response spectrum.
The list of available building standards is provided in Figure 2.12.
Figure 2.12: List of implemented international and European building standards that regulate the earth-
quake design of structures and provide formulas for response spectra
Select a Standard from the drop-down list; the Code Parameters and their default values (see
Figure 2.11) change depending on the chosen standard. The parameters can be selected from
drop-down lists or entered manually depending on the type of parameter and depending on the
chosen building standard.
Many standards distinguish between a linear design spectrum employing the ductility factor q and
an elastic response spectrum employing a damping correction factor η. Note the drop-down menu
to choose between these types of spectra. Please note also the drop-down menu to distinguish
between horizontal and vertical response spectra definitions. The cells that are not editable
are determined by another parameter already set. For example in EN 1998-1 CEN [1], the ground
classes A to E determine the parameters TB , TC and TD ; those parameters cannot be edited manually.
When you change the ground type to Other, you can adjust the parameters TB , TC and TD manually
for the response spectra in accordance to EN 1998-1 [1].
The resulting response spectrum is illustrated in the graphic seen in Figure 2.11. The tool tip of
your mouse shows information about the displayed values. The values of the generated response
spectra are listed in the Table tab as shown in Figure 2.13.
The values can be exported to Excel. The time Step should be adjusted before exporting the data.
24
2 Input Data 2
2.5.2 User-Defined Response Spectra
Any kind of response spectrum can be defined by entering the period T and corresponding accel-
erations Sa [m/s2] in the Table shown in Figure 2.14.
Figure 2.14: User-defined entry of response spectrum by defining Period T [s] and Acceleration Sa [m/s2]
User-defined response spectra can be stored in a library and can be opened whenever needed.
The [Save] button opens a dialogue box to enter a file name.
You can export your user-defined response spectra to Excel or import response spectra from Excel
by means of the respective buttons.
Figure 2.15: Generated response spectrum from acceleration with graphics and tabular values
25
2 Input Data 2
The [Edit] button links to the Accelerogram tab described in Section 2.6 where you can define
acceleration-time diagrams or load accelerograms from a library of earthquake recordings. When
accelerograms are already defined, you can select the relevant entry from the drop-down menu.
You can adjust the parameters as seen on the left. The generated response spectrum is illustrated
in the graphic below, and the values of period T and acceleration Sa are listed in the table (see
Figure 2.15). Both will be updated as soon as you change one of the parameters.
The viscous Damping d [%] is the damping of the SDOF oscillator family for which the maximum
system responses are calculated. The larger the damping is set, the fewer accelerations will be
generated.
The Number of periods determines the number of steps between the minimum and maximum
time (period), and determines therewith the number of data points generated.
The Min time is the period of the first SDOF oscillator considered for the calculation. You see the
results in the first row in the table shown in Figure 2.15. The Max time is the period of the last SDOF
oscillator considered for the calculation. You see the results in the last row in the table shown in
Figure 2.15.
The response spectra generated might be different in each direction as the accelerograms might
be different in each direction. You can change the displayed response spectra in the graphic and
table by using the radio buttons shown on the left.
You can export the generated response spectra to MS Excel.
2.6 Accelerograms
Accelerograms are acceleration-time diagrams usually recorded from previous earthquakes. In
RF-DYNAM Pro accelerograms can be used to generate response spectra or to perform a time
history analysis. In both cases the system can be excited on all or some supports.
The Accelerogram tab is shown in Figure 2.16.
26
2 Input Data 2
The Accelerogram tab is only available when Accelerations have been selected in the General tab
shown in Figure 2.1. This tab belongs to the add-on modules RF-DYNAM Pro - Forced Vibrations
and RF-DYNAM Pro - Nonlinear Time History.
In RF-DYNAM Pro, you can load accelerograms from a library or enter them manually.
27
2 Input Data 2
2.6.2 User-Defined Acceleration-Time Diagrams
Any kind of acceleration-time diagram can be defined by entering the required values in the
Accelerogram table shown in Figure 2.18.
User-defined accelerograms can be stored in a library and can be opened whenever needed. The
[Save] button opens a dialogue box to enter a file name.
You can export your user-defined accelerogram to Excel or import an acceleration-time diagram
from Excel.
The defined accelerogram is displayed in the graphic (see Figure 2.16). You can switch between
the X, Y, and Z-directions in the graphic.
28
2 Input Data 2
2.7 Time Diagrams
Time diagrams can be defined Transient, Periodic, or entered as Function. They excite the system
at a specific position. The load position is defined in static load cases (LC) where any type of load
can be entered. The static load cases (LC) are connected to the time diagrams (TD) in the Dynamic
Load Cases (see Section 2.10.2), and the multiplier k is used to determine the final value of the
excitation force.
The time diagram tab is only available when Time diagrams have been selected in the General tab
shown in Figure 2.1. This tab belongs to the add-on modules RF-DYNAM Pro - Forced Vibrations
and RF-DYNAM Pro - Nonlinear Time History.
Figure 2.19: Module Window Time Diagrams showing Transient time diagrams
The table entries can be stored in a library and can be opened whenever needed. Importing and
exporting from and to Excel is also possible.
29
2 Input Data 2
2.7.2 Periodic Excitation
To enter periodic functions, the Angular Frequency 𝜔 [rad/s], the Shift 𝜑 [rad], and the Multiplier k
have to be provided in tabular form. This is shown in Figure 2.20.
Figure 2.20: Module Window Time Diagrams showing Periodic time diagrams
Several harmonic functions can be overlain by filling several rows in the table. The periodic
functions are defined as follows:
f(t) = k1 ⋅ sin (𝜔1 ⋅ t + 𝜑1 ) + k2 ⋅ sin (𝜔2 ⋅ t + 𝜑2 ) + … (2.7)
The usual [Save], [Open], [Import] and [Export] functions are available.
2.7.3 Functions
You can enter the Function k(t) directly in the text box to define a time diagram. The parameter t is
reserved for the time. The resulting time diagram is displayed in the graphic as shown in Figure 2.21.
The generated tabulated values are listed on the right hand side.
Figure 2.21: Module Window Time Diagrams showing Function time diagrams
30
2 Input Data 2
The value Max t sets the maximum time for calculation; the time diagram k(t) = 0 for all t > tmax .
The Step size is used for the tabulated values in the table and for the automatic determination of
the time step for calculation (see Section 2.10.5).
All operators and functions that are available in RFEM can be also used in RF-DYNAM Pro. You can
use the parameters that you have defined in the main program RFEM. To avoid unit conversion, it
is recommended to employ dimensionless parameters in RF-DYNAM Pro. For more information
about the parametric input, consult Section 11.6 of the RFEM manual.
The tabulated values can be saved to the library or exported to Excel.
A complex example of how the time diagram functions can be used is shown in the DLUBAL webinar
Time History Analysis in RF-/DYNAM Pro - Walking and Running Across Pedestrian Bridge.
Figure 2.22: Module Window Dynamic Load Cases with General tab
Those options are only available when the specific add-on modules have been activated in the
General tab of RF-DYNAM Pro (see Figure 2.1).
31
2 Input Data 2
For the linear time history analysis, two solvers are available – the modal analysis which is based
on modal decomposition and requires a natural vibration analysis, and the implicit Newmark solver
which is a direct solver that does not require eigenvalues. For the nonlinear time history analysis,
two solvers are available as well – the implicit Newmark solver and an explicit solver. Details on the
time history solvers are provided in Section 2.10.3 and Section 2.10.4.
You have to assign a specific Natural vibration case (NVC) (discussed in Section 2.4) to the dynamic
load case for a response spectrum analysis, a modal analysis or an equivalent static force analysis.
The dynamic load case (DLC)) number is set automatically and cannot be edited. When a case is
deleted later, the numbers do not change. You can decide whether the specific DLC is to be solved
or not by selecting or clearing the To Solve option. You can also access this option in the shortcut
menu by right-clicking the corresponding DLC number.
You can calculate each dynamic load case separately by using its shortcut menu and applying
Calculate. The colour of a DLC is grey when no results of this dynamic load case are available yet.
It turns green as soon as the calculation has finished. For more information, see Section 3.3.
Load cases (LC) and/or result combinations (RC) are generated in each dynamic load case when
the check boxes in the To Generate section have been selected (see Response Spectra Analysis tab
in Figure 2.23, Time History Analysis tab in Figure 2.26 and Figure 2.27, and Equivalent Force Analysis
tab in Figure 2.32). Those load cases and result combinations are generated automatically and are
overwritten when the RF-DYNAM Pro calculation is performed again. The descriptions of the load
cases and result combinations clearly refer to the originating DLCs.
You can enter the number of the first generated load case or result combination. RF-DYNAM
Pro does not overwrite existing load cases or result combinations (static or generated from other
add-on modules). It chooses the first unused LC or RC number available. Load cases and result
combinations from RF-DYNAM Pro are deleted when the results of the corresponding dynamic
load case (DLC) are deleted.
It is possible to activate the export options when the calculation is finished. A re-calculation is
hereby not required.
In the following sections of this manual, the Response Spectrum Analysis, the Time History Analysis
and the Equivalent Static Force Analysis are discussed in detail.
32
2 Input Data 2
2.9 Dynamic Load Cases - Response Spectrum Analysis
A multi-modal and multi-point response spectra analysis is performed when the corresponding
radio button in Figure 2.22 is selected; this is only available in RF-DYNAM Pro - Forced Vibrations.
The definition of response spectra was discussed in Section 2.5. Mode shapes from the assigned
natural vibration case (NVC) can be selected for the analysis. The SRSS or the CQC rule is available for
modal combination. The calculation is done within the module RF-DYNAM Pro - Forced Vibrations
and is therefore completely linear. Stiffness modifications defined in the NVCs (Section 2.4.7) are
used for the analysis. The final results of the response spectrum analysis are exported to result
combinations (RC).
The sub-tabs when performing a response spectra analysis are General, Response Spectrum Analysis,
Damping, and Mode Shapes. The Response Spectrum Analysis tab is illustrated in Figure 2.23.
Figure 2.23: Module Window Dynamic Load Cases with Response Spectrum Analysis tab
33
2 Input Data 2
2.9.2 Assign Response Spectrum
A specific response spectrum can be assigned in each direction. At least one of the directions must
have been selected. The relevant Response spectrum can be chosen from the drop-down menu.
The Multiplier factor can be adjusted independently for each direction.
The excitation direction can be rotated in the XY-plane about the Z-axis. For example, when
your response spectrum shall excite the structure 45 ° rotated about the Z-axis, activate only the
X-direction with your response spectrum and enter 𝛼 = 45 °.
For each set of supports, the response spectrum in each specific direction must be activated
separately. This allows for different response spectra for each set of supports.
The combined results RSRSS result from the modal responses Rp from p modes of the structure.
In RF-DYNAM Pro, a modified form of the SRSS rule is available in order to calculate corresponding
results, i.e. corresponding internal forces. Compared to the standard form of SRSS rule, those cor-
responding results are usually much smaller and the signs are correct in relation to the controlling
force. The SRSS rule is applied as equivalent linear combination [11] as follows:
p
Ri
RSRSS = ∑ fi ⋅ Ri with fi = (2.9)
p
i=1
√∑ Rj2
j=1
In all other cases, the CQC rule must be applied. The CQC rule is defined as follows:
√
√
p p
RCQC = √∑ ∑ Ri 𝜀ij Rj (2.11)
⎷ i=1 j=1
where the correlation coefficient 𝜀 is
8 ⋅ √Di Dj (Di + rDj ) r 3/2 𝜔j
𝜀ij = with r = (2.12)
(1 − r 2 )2 + 4Di Dj r (1 + r 2 ) + 4 (Di2 + Dj2 ) r 2 𝜔i
The correlation coefficient 𝜀 simplifies when the viscose damping value D is equal for all modes to
the following:
8 ⋅ D 2 (1 + r)r 3/2
𝜀ij = (2.13)
(1 − r 2 )2 + 4D 2 r(1 + r)2
34
2 Input Data 2
The damping value Di that is required for the calculation of the correlation coefficient 𝜀ij is defined
in the Damping tab that appears as soon as you have selected the CQC rule. The settings for
damping are described below.
Analogously to the SRSS rule, the CQC rule can be applied as equivalent linear combination [11].
The CQC rule, still employing 𝜀ij as defined in Equation 2.12, is modified as follows:
p
p ∑ 𝜀ij Rj
j=1
RCQC = ∑ fi ⋅ Ri with fi = (2.14)
p p
i=1
√∑ ∑ Ri 𝜀ij Rj
i=1 j=1
Further information and mathematical derivations of the combination rules are provided in
[12–14].
Figure 2.24: Module Window Dynamic Load Cases with Damping tab
For the conversion from Rayleigh coefficients 𝛼 and 𝛽 to the Lehr's damping Di it is referred to
Section 2.10.9.
35
2 Input Data 2
2.9.7 Mode Shape Selection
In the Mode Shapes tab illustrated in Figure 2.25, the natural frequencies ω and f as well as the
periods T are listed with the corresponding accelerations Sa of the response spectrum and the
effective modal mass factors fme in the translational directions. The assigned response spectrum
is illustrated in the graphic. Corresponding values are displayed in red when you have selected a
row in the table as shown in Figure 2.25.
Figure 2.25: Module Window Dynamic Load Cases with Modes Shapes tab – first mode selected in table
When no calculation has been performed yet, use the button [Calculate Mode Shapes] to calculate
the frequencies internally without closing the module.
In the table, you can select modes that shall be used for the response spectra analysis. All modes
will be selected when the Select all option is activated. In this case, the selection cannot be changed
manually. By using the Deselect modes with option, you can deactivate modes with an effective
modal mass factor below a specific value. When both check boxes are clear, you can select the
modes manually in the To Generate column.
The sums of the effective modal mass factors fme are shown at the bottom of the table. According
to EN 1998-1 Section 4.3.3.3 [1], the effective modal mass factors of all modes taken into account
shall be at least 90%. When this cannot be achieved, all modes with a factor above 5% shall be
taken into account. Further details about the effective modal mass factors and how those are
calculated are provided in Section 4.1.
You can switch between the excitation directions X, Y and Z in the graphic. Note that also the
values Sa in the table update depending on this choice.
The values Sa in the table and the diagram reflect the multiplier factors applied to each direction
(see Response Spectrum Analysis tab in Figure 2.23). Note that – if you have rotated your assigned
response spectrum around the angle 𝛼 – the directions X and Y are local directions in the Mode
Shapes tab.
When you use the multi-point option and have defined several sets of supports in the Response
Spectrum Analysis tab (see Figure 2.23), the values of Sa in the table and the diagram in the Mode
Shapes tab change depending on the selected set of support in the drop-down menu.
36
2 Input Data 2
2.10 Dynamic Load Cases - Time History Analysis
A time history analysis is performed when either the Time history analysis of accelerograms or the
Time history analysis of time diagrams has been selected in the General sub-tab (see Figure 2.22).
The excitations and the results of a time history analysis are both data versus time.
The corresponding sub-tabs when performing a time history analysis are General, Time History
Analysis, Calculation Parameters, and Damping. The Time History Analysis tab is illustrated in Fig-
ure 2.26 and Figure 2.27. Note the difference in the user interface depending on the type of
excitation: Accelerograms excite the structure at the supports. The time history analysis of ac-
celerograms is discussed in Section 2.10.1. Time diagrams, however, can excite the structure
everywhere. The type of load and location of excitation is defined by load cases (LC) in RFEM. The
time history analysis of time diagrams is discussed in Section 2.10.2.
The time history analysis can be performed linearly, neglecting geometrically and structural non-lin-
earities with the add-on module RF-DYNAM Pro - Forced Vibrations. Two linear solvers are avail-
able – the Linear modal analysis and the Linear implicit Newmark analysis. They are described in
Section 2.10.3.
A nonlinear time history analysis is available with the add-on module RF-DYNAM Pro - Nonlinear
Time History. A large strain deformation analysis is performed under consideration of the defined
nonlinearities. Two nonlinear solvers are available – the Nonlinear implicit Newmark analysis and
the Explicit analysis. They are described in Section 2.10.4.
Time steps for the analysis are discussed in Section 2.10.5. The mass matrix settings are described
in Section 2.10.8 and the structural damping options in Section 2.10.9.
in which M is the mass matrix, D is the damping matrix and K is the static stiffness matrix. The
time dependent vectors u, u̇ and ü are the absolute displacements, velocities and accelerations.
üg is the ground acceleration which can be different in the different directions.
The definition of accelerograms was discussed in Section 2.6. In RF-DYNAM Pro, all supports can
be excited simultaneously with the same accelerogram. The multi-point option makes it possible
to excite single supports by different accelerograms (only available for linear solvers).
When nonlinearities play a role in the model, the time history of accelerograms might be favoured
compared to an response spectrum analysis.
The Time History Analysis sub-tab is illustrated in Figure 2.26.
37
2 Input Data 2
Figure 2.26: Module Window Dynamic Load Cases with Time History Analysis tab of Accelerograms
Assign Supports
The supports that are excited by the accelerogram have to be assigned. When you consider all
supports to be identically excited, then those supports can be nodal, line or surface supports. The
multi-point option is available for nodal supports only at the moment. Supports defined with a
spring constant cannot be used for a time history analysis of accelerograms at the moment. The
multi-point option is only available for a linear time history analysis.
You can create several sets of supports by clicking the button. You can switch between the
sets by using the drop-down menu. Use the button to delete a specific set. For each set, a
list of supported nodes must be provided. It can be entered manually or selected graphically by
using the function.
For background on the multi-point excitation, the reader is referred to the literature [8 and 10].
Assign Accelerogram
Choose the available accelerograms from the drop-down menu. You can apply different multiplier
factors in each direction.
The excitation direction can be rotated in the XY-plane about the Z-axis. For example, when your
accelerogram shall excite the structure 45 ° rotated about the Z-axis, activate only the X-direction
with your response spectrum chosen from the drop-down menu and enter 𝛼 = 45 °.
You can assign a specific accelerogram to each set of supports.
38
2 Input Data 2
2.10.2 Time History Analysis of Time Diagrams
With a time history analysis of time diagrams, any point of the system can be excited in any
direction.
M ü + D u̇ + K u = p(t) (2.16)
in which M is the mass matrix, D is the damping matrix and K is the static stiffness matrix. The
time dependent vectors u, u̇ and ü are the absolute displacements, velocities and accelerations.
p(t) is the external excitation.
The definition of time diagrams was discussed in Section 2.7. In the Dynamic Load Cases, those
time diagrams are combined with load cases (LC) to define the type, the location and direction of
the excitation.
The Time History Analysis sub-tab is illustrated in Figure 2.27.
Figure 2.27: Module Window Dynamic Load Cases with Time History Analysis tab of Time Diagrams
39
2 Input Data 2
2.10.3 Linear Solvers: Modal Analysis versus Implicit Newmark
In the General sub-tab (see Figure 2.22), you can choose between two linear time history solvers –
the Linear modal analysis and the Linear implicit Newmark solver. The linear analysis is geometrically
linear (small deformations) and ignores or replaces all defined non-linearities in the model, i.e.
cables are replaced by trusses and failure of a support is ignored. The linear solvers belong to the
add-on module RF-DYNAM Pro - Forced Vibrations.
The modal analysis uses a decoupled system based on the eigenvalues and mode shapes of the
structure, determined in the assigned natural vibration case (NVC). The multi-degree-of-freedom
(MDOF) system is transformed into many single-degree-of-freedom-systems (SDOF) (diagonalized
mass and stiffness matrix). A certain amount of eigenvalues is required to ensure accuracy. The
solution of the decoupled system is then found with an implicit (Newmark) solver. The mass matrix
settings and stiffness modifications are taken over from the NVC assigned. Initial deformations
can be applied. Once the eigenvalues are determined, the modal analysis provided in RF-DYNAM
Pro - Forced Vibrations is slightly faster than the linear implicit Newmark analysis.
The linear implicit Newmark solver is a direct time integration scheme that requires sufficient
small time steps to achieve accurate results. There is no natural vibration analysis required. Settings
for the mass matrix are defined in the Calculation Parameters of the DLCs. Stiffness modifications
and initial deformations are not available with this type of solver. For more theoretical background,
see e.g. [10]. The choice of time step is discussed in Section 2.10.5.
For the direct time history solver, dashpot elements can be used. This member type is available in
RFEM where you define the viscous damping coefficient c as well as the axial stiffness constant
(spring element) which is then acting parallel (Kelvin-Voigt-Model).
40
2 Input Data 2
This estimation provides a smaller time step than the stability limit in Equation 2.17. Note also
that many effects are not included in the time step estimation and that for accuracy reasons an
even smaller time step Δt may be required. RFEM uses a fixed time incrementation where the
fixed time step can be the initial stable time increment or a user-defined value.
Implicit schemes solve unknown values at time i + 1 based on the values at time i and i + 1. Thus,
nonlinear equations must be solved, iterations and convergence checks are required. The implicit
Newmark analysis is unconditionally stable, there is no upper stability bound in time step Δt.
Still, a reasonable small time increment is required to achieve accurate results. The time step
depends on the excitation, the frequency of the structure and the complexity of nonlinearities
in the structure. The implicit solver needs “time steps typically one or two orders of magnitude
larger than the stability limit of simple explicit schemes” [15]. There are no restrictions regarding
the mass matrix and Rayleigh damping when using the implicit Newmark solver.
The choice between the two nonlinear solvers depends mainly on the required time step but also
on the type of excitation. For both solvers, a time step convergence study must be done. For
further theoretical background, see [10 and 15–17].
For both nonlinear solvers, dashpots element can be used. This member type is available in
RFEM where you define the viscous damping coefficient c as well as the axial stiffness constant
(spring element) which is then acting parallel (Kelvin-Voigt-Model). The spring can be defined with
slippage and/or as diagram. Those nonlinear features of dashpot elements are only applicable to
the nonlinear solvers.
You can define the time step for calculation manually, or RF-DYNAM Pro determines it automati-
cally.
For the modal analysis and the linear and nonlinear implicit Newmark solvers, RF-DYNAM Pro
chooses the time steps for calculation automatically according to the assigned excitation. For
accelerograms, transient and function time diagrams, the automatic time step is determined using
the time step in between the data points, 𝛥t = Min{ti+1 − ti }/7. For periodic time diagrams, the
automatic time step is determined with 𝛥t = 𝜋/(10 𝜔) using the largest excitation frequency.
Note that this automatically determined time step is only in accordance to the excitation. Other
values, such as the frequency of the structure and non-linearities, are not taken into account, but
41
2 Input Data 2
they do influence the required time step. For the implicit solvers, it is therefore recommended to
define the time step for calculation manually.
The explicit solver determines the automatic time increment in accordance to the stability limit as
discussed in Equations 2.17 and 2.18. Usually, this time step is sufficient also in terms of accuracy.
Nevertheless, effects like e.g. rapid changes in stiffness are not included in the time step estimation.
For highly nonlinear systems, an even smaller time step Δt may be required for accuracy reasons.
A time step convergence study should be performed in any case. It is essential for nonlinear time
history analyses.
A result combination (RC) as a result envelope with maximum and minimum values of all saved
time steps is generated when you select the Create result combination option.
More information on the exported load cases and result combination is provided in Section 4.3.
42
2 Input Data 2
2.10.8 Calculation Parameters
The calculation parameters of time history DLCs are only available for direct solvers, such as the
linear and nonlinear implicit Newmark solver and the explicit solver. Here, you define the mass matrix
and the directions of acting masses. The Calculation Parameters sub-tab is shown in Figure 2.29.
Figure 2.29: Module Window Dynamic Load Cases with Calculation Parameters tab (direct solvers only)
In case of the Modal Analysis, the settings of acting masses, mass matrix and FE mesh settings are
taken over from the assigned NVC.
Acting Masses
You can import a specific Mass cases or Mass combination into each DLC. Those have been defined
as described in Section 2.2 and Section 2.3. Select the relevant mass case or combination from the
drop-down menu.
You have to define the direction in which the masses are acting. The masses act in the global trans-
lational X, Y, or Z-directions when you select the corresponding check boxes. They act rotationally
about the global X, Y, and Z-axes when the corresponding check boxes have been activated. Those
settings change your mass matrix.
When the explicit solver is used, masses have to act on each FE-node and in all directions.
FE Mesh Settings
The button links directly to the FE mesh settings of the main program RFEM. The FE Mesh
Settings dialogue box is also available in the Calculation → FE Mesh Settings menu of RFEM.
Further details on the FE mesh and the parameters that can be adjusted are described in the RFEM
manual, Section 7.2.
Members that are not integrated into surfaces are not divided into FE-nodes by default. When
a diagonal mass matrix is used, the mass of a member is lumped to the two end nodes of the
member. For a more refined distribution of the mass, you need to activate the member division in
the FE mesh settings.
43
2 Input Data 2
Initial Condition and Stationary State
The functions Initial condition and Stationary state are available. With the option Initial condition all
conditions from the selected load case or load combination at time step t = 0 s are imported and
then released shortly after. The function Stationary state considers all conditions over the entire
time.
In addition to the imported deformations and loads used to establish an equilibrium state at the
beginning of the calculation, all stiffness modifications made in the settings of the load cases and
load combinations are taken into account. These have already been explained in Section 2.4.7.
Furthermore the states of nonlinearities are imported in the nonlinear solvers.
It is possible to choose between load cases and load combinations. For the linear solver, only load
cases or load combinations can be used which are calculated according to a geometrically linear
analysis. Load cases and load combinations which are calculated according to a large deformation
or postcritical analysis can only be used for one of the two nonlinear solvers, however.
Figure 2.30: Module Window Dynamic Load Cases with Damping tab
using the factors 𝛼 and 𝛽. The Rayleigh damping coefficients are used for the linear and nonlinear
implicit Newmark and the explicit solver. The damping matrix C does not need to be a diagonal
matrix for the direct time history solvers. For more information about the Rayleigh damping, see
for example [18].
44
2 Input Data 2
The Lehr's damping values Di are used for the modal analysis. The damping matrix C needs to be a
diagonal matrix. The Lehr's damping values are defined for each single mode i as a factor between
the existing and the critical damping as follows:
ci
Di = (2.21)
2 ⋅ mi ⋅ 𝜔i
where ci are the entries in the diagonal damping matrix, mi the modal masses, and 𝜔i the angular
frequencies of the system.
The Rayleigh coefficients and the Lehr's damping are connected as follows:
1 𝛼
Di = ( + 𝛽 𝜔i ) (2.22)
2 𝜔i
This equation is visualised in Figure 2.31 using the Rayleigh coefficients 𝛼 = 0.2 and 𝛽 = 1 × 10−3 .
It gets clear that for one set of Rayleigh coefficients different values of Lehr's damping result for
each natural frequency of the structure.
RF-DYNAM Pro internally converts the Rayleigh damping coefficients 𝛼 and 𝛽 into the Lehr's
damping values Di if the modal analysis is selected. In this case, the solution is unique.
Figure 2.31: Relation between Rayleigh coefficients 𝛼 and 𝛽 and the Lehr's damping D using the examples:
1 – Total: 𝛼 = 0.2 and 𝛽 = 1 × 10−3
2 – 𝛽-Damping: 𝛼 = 0.0 and 𝛽 = 1 × 10−3
3 – 𝛼-Damping: 𝛼 = 0.2 and 𝛽 = 0.0
Illustrated are the Lehr's damping values resulting for each angular frequency 𝜔 of the structure.
For the implicit and explicit solver, the Rayleigh coefficients have to be specified. You can convert
Lehr's damping values Di to Rayleigh coefficients 𝛼 and 𝛽 by clicking the button: in a new
dialog box, define the most dominant frequencies of your model in combination with the required
Lehr's damping values.
The conversion from Lehr's damping values to Rayleigh coefficients is illustrated in the example
in Section 5.2.
45
2 Input Data 2
2.11 Dynamic Load Cases - Equivalent Load Analysis
The equivalent load analysis is performed when the corresponding radio button in Figure 2.22 is
selected; it belongs to the add-on module RF-DYNAM Pro - Equivalent Loads. The equivalent load
analysis in RF-DYNAM Pro is based on the multi-modal response spectra analysis.
The main differences to the Response Spectra Analysis in the add-on module RF-DYNAM Pro - Forced
Vibrations described in Section 2.9 are listed below:
• Load cases with equivalent loads are exported to RFEM separately for each mode and each
excitation direction.
• The calculation of load cases is done in the main program RFEM. Thus, nonlinearities are
considered by default, but stiffness modifications defined in the NVCs are not taken over.
• Accidental torsional actions can be considered automatically.
• Base shear forces can be easily evaluated separately for each mode.
• Result combinations are produced separately for each excitation direction (combined modal
responses with SRSS or CQC) and for the combination of results from different excitation
directions (SRSS, 100% / 30% (40%)).
• A signed result option is available that provides unique RCs using the signs of the dominant
mode shape.
• The results are reproducible step by step.
• All supports are excited identically (no multi-point option).
The corresponding sub-tabs when performing an equivalent load analysis are General, Equivalent
Force Analysis, and Mode Shapes. The Equivalent Force Analysis sub-tab is illustrated in Figure 2.32.
Figure 2.32: Module Window Dynamic Load Cases with Equivalent Force Analysis tab
46
2 Input Data 2
2.11.1 Assign Response Spectrum
In each direction, a specific Response spectrum can be assigned. At least one of the directions
must be selected. The response spectrum (RS) can to be chosen from the drop-down menu. It is
possible to adjust the multiplier Factor independently for each direction.
The excitation direction can be rotated in the XY-plane about the Z-axis. For example, when
the response spectrum shall excite the structure 45 ° rotated about the Z-axis, activate only the
X-direction with your response spectrum chosen from the drop-down menu and enter 𝛼 = 45 °.
The combined results RSRSS result from the modal responses Rp from p modes of the structure.
RF-DYNAM Pro includes a modified form of the SRSS rule in order to calculate corresponding results,
i.e. corresponding internal forces. Compared to the standard form of the SRSS rule, those corre-
sponding results are usually much smaller and the signs are correct in relation to the controlling
force. The SRSS rule is applied as equivalent linear combination [11] as follows:
p
Ri
RSRSS = ∑ fi ⋅ Ri where fi = (2.24)
p
i=1
√∑ Rj2
j=1
47
2 Input Data 2
In all other cases, the CQC rule must be applied. The CQC rule is defined as follows:
√
√
p p
RCQC = √∑ ∑ Ri 𝜀ij Rj (2.26)
⎷ i=1 j=1
where the correlation coefficient 𝜀 is
8 ⋅ √Di Dj (Di + rDj ) r 3/2 𝜔j
𝜀ij = where r= (2.27)
(1 − r 2 )2 + 4Di Dj r (1 + r 2 ) + 4 (Di2 + Dj2 ) r 2 𝜔i
If the viscose damping value D is equal for all modes, the correlation coefficient 𝜀 simplifies to:
8 ⋅ D 2 (1 + r)r 3/2
𝜀ij = (2.28)
(1 − r 2 )2 + 4D 2 r(1 + r)2
The damping value Di that is required for the calculation of the correlation coefficient 𝜀ij is defined
in the Damping tab. It appears as soon as you have selected the CQC rule. The settings for damping
are described below.
Analogously to the SRSS rule, the CQC rule can be applied as equivalent linear combination [11].
The CQC rule, still employing 𝜀ij as defined in Equation 2.27, is modified as follows:
p
p ∑ 𝜀ij Rj
j=1
RCQC = ∑ fi ⋅ Ri where fi = (2.29)
p p
i=1
√∑ ∑ Ri 𝜀ij Rj
i=1 j=1
You can find more information on the option Use equivalent linear combination and an example in
the Knowledge Base - 001098.
When the CQC rule is applied, the exported load cases contain information on the corresponding
angular frequency 𝜔 and the applied damping Di .
You can find more information on the CQC rule and an example in the Knowledge Base - 001263.
Independently of RF-DYNAM Pro, the CQC rule is available in the main program RFEM. This is
described in the Knowledge Base - 001259.
With the usual quadratic combinations like the SRSS and CQC rule, all maxima are positive and
all minima are negative in the produced result envelope. The signs of results are lost and the
results combinations are not unique. To overcome this side effect of quadratic combinations and
to produce a more realistic representation of internal forces, the so-called Signed Results option
is available in RF-DYNAM Pro. When the structure has a dominant eigenvalue, the signs of the
corresponding load case can be applied to the results in the result combination. This leads to
unique result combinations (RC). When you choose Automatic in the drop-down list, RF-DYNAM
Pro selects the dominant mode shape automatically by means of the largest effective modal mass
factor. You can also select any other eigenvalue manually.
You can find more information on the Signed Results option and an example in the Knowledge
Base - 001278.
Further information and mathematical derivations of the combination rules are provided in e.g.
[12–14].
48
2 Input Data 2
2.11.4 Combination of Directional Components
The internal forces resulting from different excitation directions can be combined quadratically
with the SRSS rule, or by using the 100% / 30% (40%) rule. The SRSS rule is applied as defined in
Equation 2.23, but now i = 1..p are the excitation directions X, Y, and Z. The SRSS rule for the direc-
tional combination can also be realised as equivalent linear combination applying Equation 2.24.
More information on the exported load cases and result combinations is provided in Section 4.4.
49
2 Input Data 2
2.12 Global Details Settings
The Details dialogue box can be accessed by pressing the [Details] button. You can also open it
via Settings → Details. In this dialogue box, you can set global parameters that are valid for the
overall dynamic calculation and independent of any defined case within the module.
The Details dialogue box is shown in Figure 2.33.
Figure 2.33: Details dialogue box with global parameters concerning RF-DYNAM Pro
Neglect Masses
Masses on nodes, lines, members, and surfaces can be neglected in the dynamic analysis. In
Section 2.2.4, details on the Neglect Masses feature are provided.
Instability Detection
The stability of a system can be analysed. For it, the default value in RF-DYNAM Pro is set to 0.01,
which means a very sensitive detection of instability and an early break-off limit. Please see RFEM
manual, Section 7.3.3 for further details.
50
2 Input Data 2
Equivalent Loads
As discussed in Section 2.11, load cases are exported containing the generated equivalent loads.
Loads can only be viewed in the graphic when less than 5,000 loads are produced. You can change
this value here, but note that the process of displaying more than 5,000 loads is slow.
The maximum number of loads that RFEM can produce and export to load cases is 99,999. As
equivalent loads are generated on each FE-node, you can only get a full set of equivalent loads
when your structure has less than 99,999 FE-nodes. When your structure has more FE-nodes, the
smallest equivalent loads are neglected, and only 99,999 loads will be exported. You can adjust
the maximum number of generated equivalent loads in the Details dialogue box.
Figure 2.34: Dialogue box Units and Decimal Places showing settings of RF-DYNAM Pro Input Data
Select the add-on module RF-DYNAM Pro in the list of modules. The Input Data tab is opened.
There you can choose the units from the drop-down menus and adjust the decimal places.
51
3 Calculation 3
3 Calculation
3.1 Check
Before starting the calculation, you can check the input data without closing the add-on module.
Click [Check] in the bottom part of the module.
The module shows the Input Data Verification dialogue box displaying any warning and error
messages. When no errors have been found, the message No consistency errors found is displayed.
When you click [OK & Calculate], the check is also performed. If an error is detected, the Input Data
Verification dialogue box is shown and you can decide whether you want to perform or cancel the
calculation.
The To Calculate dialogue box (Calculate → To Calculate) lists the add-on module cases as well
as load cases and load combinations. This is shown in Figure 3.3.
52
3 Calculation 3
You can filter the available cases with the drop-down menu shown on the left.
To transfer the selected RF-DYNAM Pro case to the list on the right, use the or button.
Click [OK] to start the calculation. All cases (NVCs and DLCs) will be calculated where no results are
available and which have been selected to be solved (To Solve check box).
When you have started the calculation, you can observe the analysis process in a separate dialogue
box shown in Figure 3.4.
53
3 Calculation 3
3.3 Availability of Results
The process to delete and calculate cases in RF-DYNAM Pro is differentiated: Cases like natural
vibration cases (NVCs) and dynamic load cases (DLCs) can be calculated separately, and also the
results can be deleted separately by using the shortcut menu (right-click the specific case). The
shortcut menu also offers to delete the case, set the case To Solve or Not to solve.
The colour of a case is grey when no results are available. The color of a case is green when results
are available; then also the exported LCs and RCs do exist. Exported RCs and LCs are tied to the
results, i.e. they will be deleted as soon as the results of a case are deleted.
The results of a specific case will be deleted, when
• you use the shortcut menu to delete the results,
• something is changed in the NVC or DLC within RF-DYNAM Pro,
• assigned mass cases (MC), mass combinations (MCO), natural vibration cases (NVC), response
spectra (RS), accelerograms (AC) or time diagrams (TD) change,
• assigned load cases LC or load combinations CO change.
54
4 Results 4
4 Results
The results of RF-DYNAM Pro are embedded in the main program RFEM. The general interpretation
of the results is described in Sections 8 and 9 of the RFEM manual.
Figure 4.1 shows the main program RFEM. The results of a Natural Vibration Case are displayed,
the first mode shape is illustrated, and the natural frequencies are listed in the table.
The Display navigator and the Results navigator are explained in Section 3.4.3 of the RFEM manual,
the Tables in Section 3.4.4 and the Control Panel in Section 3.4.6. In this RF-DYNAM Pro manual,
only the newly available results will be discussed.
You can access all relevant results tables of your dynamic analysis with the [Dynamic Analysis]
button.
Each of the results tables discussed in this chapter can be exported to MS Excel. The values within
the tables are saved as strings, so only the number of decimal places that are displayed will be
exported. If you need a higher accuracy, you have to adjust the decimal places in the Unit and
Decimal Places dialogue box accessible with the button.
All dynamic results tables can also be exported via File → Export. You can choose to export all
dynamic tables or only selected tables as shown in Figure 4.2. The results that are available for
each dynamic case are detailed in the following sections. Analogously you can export the data to
the CSV file format.
With the drop-down menu, you can switch between the available natural vibration cases (NVC) and
dynamic load cases (DLC). You can also switch between the cases using the and buttons.
The available results tables depend on that selection. This is clarified in the following sections.
The graphic in the work area of RFEM is also updated according to the selected NVC or DLC. All
results which are usually available in RFEM are accessible in the same way for the RF-DYNAM Pro
results. The interaction between tables and graphic works as common for RFEM.
55
4 Results 4
Figure 4.2: Dialogue box Export to Microsoft Excel - Settings to export all dynamic tables (NVCs and DLCs)
Figure 4.3: Table 5.1 Natural Frequencies with eigenvalues 𝜆 [1/s2 ], angular frequencies 𝜔 [rad/s], natural
frequencies f [Hz], and natural period T [s]
56
4 Results 4
The equation of motion of a multi-degree of freedom without damping is solved with the four
available eigenvalue solvers discussed in Section 2.4.5. Further theoretical details can be found
e.g. in BATHE [6] or TEDESCO [9]. The equation of motion is defined as
Mü + Ku = 0 (4.1)
where M is the mass matrix discussed in Section 2.4.4, K the stiffness matrix. u are the mode shapes
containing translational and rotational parts:
u = (uX ,uY ,uZ ,𝜑X ,𝜑Y ,𝜑Z ) T (4.2)
The eigenvalue 𝜆 [1/s2 ] is connected to the angular frequency 𝜔 [1/s] with 𝜆i = 𝜔i2 . The natural
frequency f [Hz] is then derived with f = 𝜔/2𝜋, and the natural period T [s] is the reciprocal of the
frequency obtained with T = 1/f.
For a multi-degree of freedom (MDOF) system, several eigenvalues 𝜆i and corresponding mode
shapes ui exist for each mode i.
Figure 4.4: Table 5.2 Mode Shapes by Node with standardised displacements uX , uY , and uZ and rotations
𝜑X , 𝜑Y , and 𝜑Z
Table 5.3 lists the values sorted by members, Table 5.4 sorted by surfaces, and Table 5.5 sorted by
mesh points. If there are no surfaces in the model, Table 5.4 does not exist.
The maximum values of u and 𝜑 are dependent on the selected scaling option discussed in Sec-
tion 2.4.2. The standardised displacements u are dimensionless, and the standardized rotations 𝜑
are given in [1/m]. Those units result from the scaling procedure.
57
4 Results 4
4.1.3 Masses in Mesh Points
The calculation of the masses in mesh points depends on the imported masses (see Section 2.2),
on the settings in the Details dialogue box (see Section 2.12), and on the finally assigned mass case
(MC) or mass combination (MCO (see Section 2.4.3). In Table 5.6, the masses in the translational
directions mX , mY , and mZ are provided. The sums of the masses are given at the bottom of the
table. Table 5.6 is shown in Figure 4.5. The coordinates of the mesh points in the global coordinate
system are listed as well.
Figure 4.5: Table 5.6 Masses in Mesh Points with translational directions mX , mY , and mZ for each mesh point
and sums of masses
Figure 4.6: Table 5.7 Effective Modal Mass Factors with modal masses Mi , effective modal masses me , and
effective modal mass factors fme related to global axes
where ui is the eigenvector of a single mode i as defined in Equation 4.2, and M is the mass
matrix discussed in Section 2.4.4. The modal mass Mi is independent of direction. It changes
depending on the scaling option chosen for the mode shapes (see Section 2.4.2); when the option
uTi Mui = 1 kg is chosen, all modal masses are Mi = 1 kg.
58
4 Results 4
The effective modal masses meff
ij
provide the masses that are accelerated in the j-direction, where
j = 1,2,3 for translations and j = 4,5,6 for rotations, separately for each mode i. Those masses
are independent of the scaling option for mode shapes, and directly related to the participation
factors 𝛤i,j .
1 T
𝛤i,j = u MTj (4.4)
Mi i
where Tj is the jth column in matrix T
1 0 0 0 (Z − Z0 ) −(Y − Y0 )
⎡ ⎤
⎢0 1 0 −(Z − Z0 ) 0 (X − X0 ) ⎥
⎢0 0 1 (Y − Y0 ) −(X − X0 ) 0 ⎥
T=⎢ ⎥ (4.5)
⎢0 0 0 1 0 0 ⎥
⎢0 0 0 0 1 0 ⎥
⎣0 0 0 0 0 1 ⎦
X, Y, and Z are the global coordinates of the considered FE-node provided in Table 5.6 (see Fig-
ure 4.5), X0 , Y0 and Z0 are the centres of total mass provided in the Mass Cases → General tab in
RF-DYNAM Pro (see Section 2.2). This matrix T exists for each FE-node. The definition of the partic-
ipation factors defining also the rotational degrees of freedom is detailed in [19], Section 15.7.5.
The participation factor 𝛤ij is dimensionless for translations and has the unit [m] for rotations.
The effective modal masses are defined with
mijeff = Mi ⋅ 𝛤ij2 (4.6)
where the effective masses for translations meX , meY , and meZ are provided in [kg] and for rotations
mu�X , mu�Y , and mu�Z in [kgm2].
The sums of the effective modal masses ∑ me are provided in Table 5.7 (see Figure 4.6). In the
translational directions, those sums are equal to the total sum of the structure Mtotal (see Table 5.6
in Figure 4.5) beside masses that are not activated, i.e. masses in fixed supports, and assuming
that all eigenvalues of the system are calculated.
The effective modal mass factors fme are required to decide whether or not a specific mode must
be taken into account for the response spectra analysis or the equivalent load analysis. EN 1998-1
[1] states in Section 4.3.3.3 that “the sum of the effective modal masses for the modes taken into
account” needs to be “at least 90% of the total mass of the structure” and that “all modes with
effective modal masses greater than 5% of the total mass” have to be taken into account .
The effective modal mass factors fme are defined as follows:
me
fme = (4.7)
∑ me
Further details on modal analysis can be found in the literature, e.g. Meskouris [20] or Tedesco [9].
59
4 Results 4
4.1.5 Export to Excel or CSV
You can export all results tables that have been discussed in this section to MS Excel or save them
in the CSV format using the File → Export option. The list of results tables is shown in Figure 4.7.
Figure 4.7: Results tables to export for natural vibration cases (NVC)
60
4 Results 4
4.2.1 Support Forces
The nodal support forces are provided in Table 5.11 which is shown in Figure 4.8. Similarly, the
line support forces are listed in Table 5.12. In those tables, the maximum and minimum support
forces are listed with the corresponding values.
Figure 4.8: Table 5.11 Nodes - Support forces with maximum and minimum results in directions X, Y, and Z
Figure 4.9: Table 5.13 Nodes - Deformations with maximum and minimum displacements and rotations
61
4 Results 4
Figure 4.10: Table 5.14 Nodes - Pseudo-Velocities with maximum and minimum velocities and angular
velocities
Figure 4.11: Table 5.15 Nodes - Pseudo-Accelerations with maximum and minimum accelerations and angular
accelerations
Pseudo velocities and accelerations are not included in the load cases or result combinations that
have been exported to the main program RFEM.
Figure 4.12: Table 5.16 Members - Internal Forces with maximum and minimum results N, Vy , VZ , MX , MY , MZ
62
4 Results 4
4.2.5 Surface Internal Forces
Table 5.17 provides the surface internal forces as shown in Figure 4.13. The maximum and mini-
mum values together with the coordinates of the FE-Nodes are provided in that table.
Figure 4.13: Table 5.17 Surfaces - Internal Forces with maximum and minimum results mx , my , mxy , vx , vy , nx ,
ny , and nxy
Figure 4.14: Table 5.18 Surfaces - Basic Stresses with maximum and minimum normal and shear stresses
63
4 Results 4
Figure 4.15: Dialogue box Edit Load Cases and Combinations with exported RCs from response spectrum
analysis, containing only result combinations of directional combination
The exported RCs cannot be modified, and the calculation parameters cannot be changed.
Figure 4.16: Results tables to export for dynamic load cases (DLC) with response spectrum analysis
The Export envelopes only option is selected by default. Only the results tables that belong to the
Dynamic Envelopes (i.e. X100% / Y30% / Z30%) will be exported. When you clear this check box,
the results tables that belong to the single directions will also be exported.
64
4 Results 4
using the drop-down menu shown on the left. The tabulated results and also the graphic in the
work window of RFEM will be updated according to your choice.
The results can also be displayed versus time in the Time Course Monitor which is accessible with
the button in the menu bar of the table or in the panel. You can activate the animation in
the main graphic by using the button in the menu bar of RFEM or in the panel. When you
have the Time Course Monitor open at the same time, the cursor moves through the plot while the
animation is running.
When you have selected the export of load cases and/or result combinations as discussed in Sec-
tion 2.10.7, the LCs and RCs are available in the drop-down list in the toolbar in RFEM shown on
the left. You have also access to the LCs in the Data navigator using Load Cases and Combina-
tions → Load Cases or similarly for the RCs using Result Combinations. Load cases and result
combinations are also accessible in the Edit Load Cases and Combinations dialogue box.
Load cases exported from a time history analysis are special: They do not contain loads, only the
results of the specific time step.
Figure 4.17: Table 5.18 Nodes - Support Forces with results in directions X, Y, and Z
Figure 4.18: Table 5.20 Nodes - Deformations with displacements uX , uY , uZ and rotations 𝜑X , 𝜑Y , 𝜑Z
65
4 Results 4
Figure 4.19: Table 5.21 Members - Internal Forces with normal forces N, shear forces Vy and VZ , as well as
moments MX , MY , and MZ
Figure 4.20: Table 5.22 Surfaces - Internal Forces with results mx , my , mxy , vx , vy , nx , ny , and nxy
The stresses of the surfaces are listed in Table 5.23 as shown in Figure 4.21.
Figure 4.21: Table 5.23 Surfaces - Basic Stresses with normal and shear stresses
66
4 Results 4
Figure 4.22: Table 5.25 Nodes - Accelerations with accelerations üX , üY , üZ and rotations 𝜑
¨X , 𝜑
¨Y , 𝜑
¨Z
Figure 4.23: Table 5.26 Nodes - Velocities with velocities u̇X , u̇Y , u̇Z and rotations 𝜑˙ X , 𝜑˙ Y , 𝜑˙ Z
Nodal accelerations and velocities are not included in the load cases or result combinations that
have been exported to the main program RFEM.
Figure 4.24: Time Course Monitor to display results versus time – here the displacements uX of three nodes
are displayed, showing the maximum and minimum values.
You can choose between Nodes, Members, Surfaces, and Support Nodes in the drop-down menu as
seen on the left. Depending on this choice, different results are available as shown in Figure 4.25.
67
4 Results 4
Figure 4.25: Available results for (a) nodes, (b) members, (c) surfaces, and (d) support nodes
For members and surfaces, you have to select the member or surface number from the list.
Depending on the choice of objects, either a list of finite element mesh points (see Figure 4.24) or
a list of distances on members (see Figure 4.26) is available. A multiple choice of listed points or
positions is possible by using the [Ctrl] or [Shift] key.
The legend in the graphic is set automatically; the axis labels and axis scales are also adjusted
depending on the selection. The minimum and maximum results with the corresponding time
steps are provided. You can use the mouse wheel to zoom into the graph. Values will be displayed
on the pointer when it is moved along the graph.
The work window of RFEM, the tables and the Time Course Monitor are interacting: When you select
the row of a specific member in a table or a node in the graphic, the settings in the Time Course
Monitor will be adjusted. You only have to choose the result that is to be displayed (Figure 4.25).
The root mean square (RMS) value of any results up to a specific time can be displayed if the check
box shown on the left is selected and a time is defined. The RMS value is calculated as follows:
1
RRMS = √ (R12 + R22 + … + Rn2 ) (4.8)
n
where Rn are the results of each time step with n considered time steps. The RMS value is illustrated
in the Time Course Monitor as shown in Figure 4.26.
Figure 4.26: Time Course Monitor with RMS value – here the moment MY is displayed versus time, showing
the RMS value up to 1.0 s.
The RMS value is only available when a single graph is plotted. If you overlay the results of several
nodes, the RMS check box is not available.
68
4 Results 4
With the [Print] button you can print the Time Course Monitor to the printout report. In the following
dialogue box, you can define the header of this graphic to be used in the report.
When you right-click on the Time Course Monitor, you have three options: (1) You can access the
Numeric Values in tabular view. The time steps are listed with all results displayed in the graphics.
An example is shown in Figure 4.27. The values as tabulated can be exported to Excel via the
button. (2) You can save the Time Course Monitor as Picture, and (3) you can access the Settings of
the graph. The Settings dialogue box of the Time Course Monitor is shown in Figure 4.28.
69
4 Results 4
Figure 4.29: Dialogue box Edit Load Cases and Combinations with exported load cases LC from a time history
analysis containing the results of single time steps
The exported LCs cannot be modified and the calculation parameters cannot be changed. Those
load cases do not contain loads, they contain the results of the specific time step. They cannot be
used in load combinations (CO), only in result combinations (RC).
Figure 4.30: Dialogue box Edit Load Cases and Combinations with exported result combination RC containing
the result envelope over all time steps from a time history analysis
The exported RCs cannot be modified and the calculation parameters cannot be changed.
70
4 Results 4
4.3.9 Export to Excel or CSV
You can export all results tables that have been discussed in this section to MS Excel or save them
in the CSV format using the File → Export option. The list of results tables is shown in Figure 4.31.
Figure 4.31: Results tables to export for dynamic load cases (DLC) with time history analysis
The Export envelopes only option is selected by default. Only the results tables that belong to the
Dynamic Envelopes will be exported. When you clear this check box, the results tables of all time
steps will be exported. Note that exporting can take considerably long when you export all time
steps.
71
4 Results 4
4.4 Dynamic Load Cases - Equivalent Load Analysis
The results tables that belong to equivalent load analysis cases are available when the correspond-
ing dynamic load case (DLC) is chosen in the drop-down menu. Tables 5.8 to 5.10 belong to this
type of dynamic load cases. The equivalent load analysis belongs to the module RF-DYNAM Pro -
Equivalent Loads. The input data required for an equivalent load analysis is discussed in Section 2.5
and Section 2.11.
You can switch among the tables showing the equivalent loads for all eigenvalues of the system,
or separately for each Mode Shape.
𝛤 Participation factors in X, Y, and Z-direction as defined in Equation 4.4 using a modal mass
of Mi = 1 kg
u Displacement values in X, Y, and Z-direction of the mode shape scaled so that
Mi = uTi Mui = 1 kg
Sa (T) Acceleration read from the response spectra diagram using the natural period T of the
considered eigenvalue
M Mass in the direction X, Y, and Z at the considered FE-node
With the formulas provided in Equation 4.9, the equivalent loads resulting from a response spec-
trum in X-direction are calculated. Accordingly, Equation 4.10 describes the loads resulting from a
response spectrum in Y-direction, and Equation 4.11 the loads resulting from a response spectrum
in Z-direction.
Those equivalent static forces exist at each FE-node within the structure, provided that the mass
M or the mode shape u is not 0 at this point.
The results tables for the equivalent load analysis are separated into the three excitation directions.
Table 5.8 gives the resulting equivalent loads for response spectra acting in X-direction, Table 5.9
those for response spectra acting in Y-direction, and Table 5.10 those for response spectra acting
in Z-direction. Table 5.9 is illustrated in Figure 4.32.
72
4 Results 4
When the equivalent loads are listed for only one eigenvalue of the system, the sum of equivalent
loads is listed at the bottom of Tables 5.8, 5.9, and 5.10. It is represented by the formulas
∑ FX,i = ∣∑(FX,i )∣ (4.12)
where i is the number of the considered eigenvalue. This sum of forces of each eigenvalue is the
base shear force.
When the equivalent loads are listed for All mode shapes of the system in one table, the overall
sum of the base shear forces from each eigenvalue is listed at the bottom of Tables 5.8, 5.9, and
5.10.
Table 5.9 is illustrated in Figure 4.32, listing the equivalent loads that belong to Mode shape 1 and
the base shear forces resulting from the excitation in Y-direction. Note that also the LC numbers
are provided to which the equivalent loads are exported.
Figure 4.32: Table 5.9 Equivalent Loads (Y-excitations) of mode shape 1 resulting from response spectrum in
Y-direction
where MZ is the torsional moment and FX and FY are the equivalent loads on each FE node as
defined in Equations 4.9, 4.10, and 4.11. The torsional moments MZ are considered in both positive
and negative directions.
73
4 Results 4
Figure 4.33: Dialogue box Edit Load Cases and Combinations with exported load cases LC from equivalent
load analysis showing two load cases for each mode and direction due to accidental torsion
The Angular frequency 𝜔 and Lehr's damping value D are also exported as LC properties from
RF-DYNAM Pro to enable the CQC modal combination. If the SRSS rule was set in RF-DYNAM Pro,
the exported damping value D = 0.
You can adjust the Calculation Parameters to your needs in the second sub-tab. Those parameters
are shown in Figure 4.34.
Figure 4.34: Calculation parameters of exported load cases from equivalent load analysis
74
4 Results 4
Changing the calculation parameters may be necessary when
• the structure has a dominant mode shape and the analysis shall consider the P-Δ effect,
• stiffness modifications were applied in the natural vibration cases (see Section 2.4.7) and the
LC shall use same modifications for consistency reasons (FAQ 002237 provides an example),
• non-linearities shall be deactivated in the load cases (equivalent loads are based on linearly
calculated eigenvalues and mode shapes, thus LCs should not consider non-linearities).
There might be a conflict of non-linearly calculated load cases and result combinations that are
applied in a dynamic analysis. Please consult the Section 5.5 of the RFEM manual to understand
the important difference between load combinations (CO) and result combinations (RC). Due to
the quadratic combination rules, only result combinations are possible in a dynamic analysis.
The equivalent loads can be displayed in the work window by pressing the [Show Loads] button.
If more than 5,000 loads are generated, the display is deactivated by default. But in the Details
dialogue box, the number of displayed equivalent loads can be changed (see Section 2.12).
Figure 4.35: Dialogue box Edit Load Cases and Combinations with exported result combinations from the
equivalent load analysis showing RC for modal combination in Z-direction
Dialogue box section Loading in Result Combination RC102 in Figure 4.35 lists the load cases (LC) that
contain the equivalent loads relevant in the considered excitation direction. Here, the accidental
torsional actions (described in Section 2.11.2) are activated, so two load cases for each mode and
direction exist. They are combined with the OR criterion before the SRSS or CQC rule is applied.
75
4 Results 4
The load cases are combined with the CQC rule according to the equivalent linear combination
(see Figure 4.36). For the Extreme value signs, the LC of the dominant mode shape is set in the
drop-down menu. Those calculation parameters are automatically generated by RF-DYNAM Pro
depending on the settings you have made in the add-on module. The calculation parameters of
the result combination (RC) are shown in Figure 4.36.
Figure 4.36: Calculation parameters of exported RC for modal combination with CQC rule as equivalent linear
combination applied and signs of LC corresponding to dominant mode shape set automatically
The calculation parameters of the result combinations can be changed in RFEM independently of
the settings in RF-DYNAM Pro.
In the second step of combination, the responses resulting from different excitation directions
are combined. One of the final result combinations (RC) is selected in Figure 4.37. Note the factors
1.00 and 0.30 that have been applied to account for the 100% / 30% rule.
Figure 4.37: Dialogue box Edit Load Cases and Combinations with exported result combinations from the
equivalent load analysis showing RC of directional combination with 100% / 30% rule
76
4 Results 4
4.4.5 Export to Excel or CSV
You can export all results tables that have been discussed in this section to MS Excel or save them
in the CSV format using the File → Export option. The list of results tables is shown in Figure 4.38.
Figure 4.38: Results tables to export for dynamic load cases (DLC) with equivalent load analysis
The Export envelopes only option is selected by default. Only the results tables containing the
equivalent loads of all mode shapes will be exported. When you clear this check box, the tables
that belong to each single mode shape will be exported as well.
Figure 4.39: Activation of automatic combination feature in General Data dialogue box of RFEM
77
4 Results 4
When you have activated the automatic combinations before you calculate any of the RF-DYNAM
Pro cases, the LCs and RCs imported from RF-DYNAM Pro will not be included in any of the Actions,
i.e. no action of the Earthquake category will exist. Consequently, the dynamic LCs and RCs will
not influence the load and result combinations that are created automatically.
Figure 4.40 shows the Action tab of the Edit Load Cases and Combination dialogue box. The LCs
imported from RF-DYNAM Pro - Forced Vibrations are not included in the Permanent action A1,
although the Permanent action category is allocated.
Figure 4.40: Actions tab of Edit Load Cases and Combinations dialogue box when automatic combinations
were activated before the RF-DYNAM Pro calculation – RF-DYNAM Pro results are available, but
imported LCs are not included in Existing Actions and no action of type Earthquake is available
When you switch on the Create combinations automatically option after you have calculated the
results in RF-DYNAM Pro and exported the LCs and RCs, make sure you keep the user-defined
result combinations as indicated in Figure 4.41.
Figure 4.41: RFEM warning when activating automatic combination feature – make sure to keep existing
RCs if RF-DYNAM Pro results are available
In this order, the Actions will take the LCs from RF-DYNAM Pro into account, and an action of the
Earthquake category will be generated. This is illustrated in Figure 4.42.
78
4 Results 4
Figure 4.42: Actions tab of Edit Load Cases and Combinations dialogue box when automatic combinations
were activated after the RF-DYNAM Pro calculation – Earthquake action exists with LCs imported
from RF-DYNAM Pro included
Nevertheless, this does not influence the automatic combinations, which means that no Action
Combinations, Load Combinations or Result Combinations will be created for the design situation
Earthquake. Load cases that have been imported from a time history analysis are included in load
combinations, though, but as they do not contain any loads (they contain only results of a specific
time step as described in Section 4.3), they do not have an influence on the results.
The result combination for the design situation Earthquake must be defined manually. Creating
Load Combinations is not sufficient, due to the quadratic combination rules that are requested
for a response spectrum analysis. For the time history analysis, Load Combinations are not suited
either because the exported LCs are only for single time steps and do not contain loads.
In EN 1990 [4] Section 6.4.3.4, Equation (6.12) provides the combination expression for the design
situation Earthquake as follows:
Ed = ∑ Gk,j “+” P “+” AEd “+” ∑ 𝛹2,i Qk,i (4.16)
j≥1
where AEd shall be the resulting RCs from RF-DYNAM Pro. The combination factors 𝛹2 are listed in
EN 1990 [4] Table A.1.1.
An example how this is done for a response spectrum analysis can be found in Section 5.1.
79
4 Results 4
4.6 Printout Report
You can create a printout report containing the dynamic results. Detailed information on the
printout report can be found in Chapter 10 of the RFEM manual.
When dynamic results are available, you can include the Input Data and the Natural Vibration Case
data in the Printout Report Selection dialogue box as shown in Figure 4.43 and Figure 4.44.
Figure 4.43: Dialogue box Printout Report Selection with Input Data available for RF-DYNAM Pro case
The Mode Shapes table that represent the selection of mode shapes and corresponding values read
from the response spectrum can be included by selecting Table 1.8.2.2 and/or 1.8.5.2. This table
was discussed in Section 2.9.7 and is relevant for a response spectrum analysis and an equivalent
load analysis.
Figure 4.44: Dialogue box Printout Report Selection with Natural Vibration Case data available
Each of the entries can be selected for all cases (NVCs or DLCs). When you clear the All check box,
you can select specific cases.
80
4 Results 4
All other RF-DYNAM Pro results are available in exported load cases or result combinations. The
printout report for those results can be adjusted as described in Chapter 10 of the RFEM manual.
Figure 4.45: Dialogue box Units and Decimal Places showing settings of RF-DYNAM Pro Results data
Select the add-on module RF-DYNAM Pro in the list of modules. In the Results tab, you can choose
the units from the drop-down menus and adjust the decimal places.
As the RF-DYNAM Pro results are embedded in the main program RFEM, most of the result values
can be adjusted in the RFEM list of units and decimal places as shown in Figure 4.46.
Figure 4.46: Dialogue box Units and Decimal Places showing settings for RFEM Results
81
5 Examples 5
5 Examples
An example how to perform a natural vibration analysis and a multi-modal response spectra
analysis using the module RF-DYNAM Pro - Equivalent Loads is shown in the webinar Natural
Frequencies and Equivalent Static Force Analysis with RFEM.
The simulation of walking and running across a pedestrian bridge demonstrating a complex time
history analysis using the module RF-DYNAM Pro - Forced Vibrations is presented in the webinar
Time History Analysis in RF-/DYNAM Pro - Walking and Running Across Pedestrian Bridge. It is
shown how time diagrams can be defined via functions employing parameters that are defined in
the main program RFEM. Also used are multiple load-time diagram sets.
In the webinar Nonlinear Time History Analysis – Machine-Induced Vibrations, induced vibrations
on a structure with nonlinear tension members are analysed with the nonlinear time history solvers.
More examples are available on our website that demonstrate the accuracy of the dynamic mod-
ules. The verification example of a Cantilever Beam (SDOF) with Periodic Excitation studies the
response of an SDOF system. There the results of RF-DYNAM Pro - Forced Vibrations are compared
with the analytical solution. The verification example of Equivalent Loads compares the equiva-
lent loads calculated by RF-DYNAM Pro - Equivalent Loads with the analytical solution. Considered
is a 5-DOF system with masses acting in X-direction.
In the following, Section 5.1 presents an example how an earthquake analysis is performed with
RF-DYNAM Pro - Equivalent Loads. The base shear forces in each floor are evaluated and the storey
drift is calculated. In Section 5.2, machine-induced vibrations are analysed with the module
RF-DYNAM Pro - Forced Vibrations. The case of resonance is studied and the structure is modified
to avoid resonance.
82
5 Examples 5
Three load cases are defined with self-weight and imposed loads. Those load cases are imported as
masses in RF-DYNAM Pro. LC1 defines the self-weight in +Z direction and 0.5 kN/m2. LC2 and LC3
contain 2.5 kN/m2 imposed load acting in the top and bottom level of the building, respectively.
In the General tab of RF-DYNAM Pro, the Mass combinations and the Response spectrum analysis
with generation of equivalent loads options are activated as shown in Figure 5.2.
Definition of Masses
Apart from the self-weight, additional masses have to be considered to calculate inertia effects.
This is regulated in EN 1998-1 [1] Section 3.2.4.
∑ Gk,j “+” ∑ 𝜓E,i ⋅ Qk,i (5.1)
where Gk,j are the permanent loads and Qk,i any imposed load. 𝜓E,i are the combination factors
for the imposed loads defined as
𝜓E,i = 𝜑 ⋅ 𝜓2,i (5.2)
where 𝜓2,i are combination factors regulated in EN 1990 [4] Table A.1.1. Assuming that the building
of this example is of category B (building with offices), 𝜓2 = 0.3. The values of 𝜑 recommended
in EN 1998-1 CEN [1] are 𝜑 = 1.0 for the top level of a building and 𝜑 = 0.5 for all other levels. To
account for the different values of 𝜑, the load cases with imposed loads have been split in LC2 and
LC3.
In RF-DYNAM Pro, Mass Cases are defined to import masses from the load cases of RFEM. The mass
cases MC1 to MC3 are shown in Figure 5.3 and Figure 5.4.
83
5 Examples 5
Figure 5.3: Mass case MC1 to import self-weight from LC1 (From self-weight of structure option not selected
as self-weight is activated in LC1)
Figure 5.4: Mass case MC2 to import imposed loads from LC2 acting in top floor
Figure 5.5: Mass combination MCO1 combining masses in compliance with EN 1998-1
84
5 Examples 5
5.1.2 Seismic Analysis in RF-DYNAM Pro
Figure 5.6: NVC1 defining mass matrix, number of eigenvalues, solver, and scaling of mode shapes
The natural frequencies f, the natural periods T, and the corresponding effective modal mass
factors fme are shown in Figure 5.7. In accordance with EN 1998-1 [1] Section 4.3.3.3, the sum of
the effective modal masses must be greater than 90% of the total seismic load. The first three
eigenvalues are relevant in the Y-direction, and only eigenvalue 28 is relevant in the X-direction.
Figure 5.7: Relevant natural frequencies f [Hz], natural periods T [s] and corresponding effective modal mass
factors fme [−]
The listed eigenvalues are considered in the multi-modal response spectrum analysis.
85
5 Examples 5
The dominant mode shape in Y-direction is illustrated in Figure 5.8, the dominant mode shape in
X-direction in Figure 5.9.
Figure 5.8: Dominant mode shape in Y-direction with natural frequency of 1.201 Hz
Figure 5.9: Dominant mode shape in X-direction with natural frequency of 5.28 Hz
86
5 Examples 5
Response Spectrum in Accordance with EN 1998-1 CEN
The horizontal design response spectrum in accordance with EN 1998-1 [1] Section 3.2.2.5 is
defined as:
2 T 2.5 2
0 ≤ T ≤ TB : Sd (T) = ag ⋅ S ⋅ [ + ⋅( − )] (5.3)
3 TB q 3
2.5
TB ≤ T ≤ TC : Sd (T) = ag ⋅ S ⋅ (5.4)
q
⎧ 2.5 TC
{ = ag ⋅ S ⋅ ⋅
TC ≤ T ≤ TD : Sd (T) ⎨ q T (5.5)
{
⎩ ≥ 𝛽 ⋅ a g
⎧ 2.5 TC TD
{ = ag ⋅ S ⋅ ⋅ 2
TD ≤ T : Sd (T) ⎨ q T (5.6)
{
⎩ ≥ 𝛽 ⋅ a g
where
T : Vibration period in [s]
ag : Design ground acceleration, ag = 𝛾1 ⋅ agR , in [m/s2 ]
agR : Reference ground acceleration in [m/s2 ]
𝛾1 : Importance factor
TB : Lower limit of the period of the constant spectral acceleration branch, in [s]
TC : Upper limit of the period of the constant spectral acceleration branch, in [s]
TD : Value defining the beginning of the constant displacement response range of
the spectrum, in [s]
S : Soil factor
Sd (T) : Acceleration values of the design spectrum in [m/s2 ]
q : Behaviour factor that describes the ductility
𝛽 : Lower bound factor for the horizontal design spectrum
In the example, the reference ground acceleration of agR = 1.0 and importance factor of 𝛾1 = 1.0
(importance category II) are assumed. The parameters S, TB , TC , and TD are chosen in accordance
with ground type C and spectra type 1. The behaviour factor q is kept with 1 on the safe side, and
the lower bound factor 𝛽 = 0.2 is kept as recommended in the standard.
The settings in RF-DYNAM Pro and the resulting horizontal and vertical response spectra are
illustrated in Figure 5.10 and Figure 5.11.
87
5 Examples 5
Figure 5.10: Horizontal design response spectrum in accordance with EN 1998-1 Section 3.2.2.5
Figure 5.11: Vertical design response spectrum in accordance with EN 1998-1 Section 3.2.2.5 – peak value
Sd (T) ≤ 2.5 m/s2 so that vertical direction can be neglected (see EN 1998-1 Section 4.3.3.5.2)
88
5 Examples 5
Response Spectra Analysis with Equivalent Loads
In the RF-DYNAM Pro - Equivalent Loads module, a multi-modal response spectrum analysis is
performed. Then equivalent loads are exported to RFEM separately for each eigenvalue and exci-
tation direction. The settings are done in the Dynamic Load Cases tab as illustrated in Figure 5.12.
Figure 5.12: Dynamic Load Cases tab with settings for Equivalent Force Analysis
The horizontal design response spectrum is applied in the X- and Y-directions. The Z-direction
does not need to be taken into account because the vertical spectrum is less than 2.5 m/s2 (see
EN 1998-1 [1] Section 4.3.3.5.2).
The option to Consider accidental torsional actions is selected to account for uncertainties in the
location of mass. This is described in EN 1998-1 [1] Section 4.3.2. The eccentricities to be applied
are calculated with ei = 0.05 ⋅ Li where Li is the length of the building.
The equivalent loads are exported to Load cases; the number of the first generated load case is set
to 4.
The combination of modal responses is done with the SRSS rule in this example. It can be used
when the eigenvalues are independent. This is the case when the criteria Tj ≤ 0.9 ⋅ Ti is fulfilled
with Tj ≤ Ti (EN 1998-1 [1] Section 4.3.3.3.2). All adjacent natural periods Ti are independent in
this example:
0.718 ≤ 0.9 ⋅ 0.833 = 0.750 (5.7)
0.514 ≤ 0.9 ⋅ 0.718 = 0.646 (5.8)
The SRSS rule is applied in form of the equivalent linear combination. The formula of this combina-
tion rule is provided in Section 2.11.3 in Equation 2.24.
The results of different excitation directions are to be combined according to the 100 / 30 % rule.
The directional combinations are regulated in EN 1998-1 [1] Section 4.3.3.5.1.
89
5 Examples 5
In the Mode Shapes sub-tab, the results of the assigned natural vibration case NVC1 are listed. Here
specific eigenvalues can be selected that are to be used for the multi-modal response spectrum
analysis. The response spectrum is shown in the graphic below the table. The corresponding
acceleration values of each eigenvalue are listed in the table and are marked in red in the graphic.
The selection of this example is shown in Figure 5.13.
The Load Cases and Result Combination are exported automatically. Due to the accidental torsion,
two load cases for each eigenvalue and excitation direction are exported – one with positive
torsional moments and one with negative torsional moments. The list of exported Load Cases is
shown in Figure 5.14.
Figure 5.14: List of exported Load Cases containing equivalent loads separately for each eigenvalue and
excitation direction – due to accidental torsion two load cases are created for each eigenvalue
90
5 Examples 5
The list of exported Result Combinations is shown in Figure 5.15.
Figure 5.15: List of exported Result Combination for modal combination and directional combination
For the modal combination, load cases (LC) resulting from different eigenvalues but from the same
excitation direction are combined with the SRSS rule. The two LCs with positive and negative
torsional moments are grouped together (alternative combination). The combination scheme is
shown in Figure 5.16.
Figure 5.16: Modal combination scheme to combine modal response resulting from excitation in Y-direction
More information on the modal combination when accidental torsion is applied can be found in
the Knowledge Base - 001118.
For the directional combination, the 100% / 30% rule is applied in this example. This is simply
achieved with combination factors as shown in Figure 5.17.
Figure 5.17: Directional combination scheme to combine results from different excitation directions with
100% / 30% rule
RF-DYNAM Pro exports two RCs in this example because two excitation directions are taken into
account. The RC leading to the most unfavourable results is used for the seismic design situation.
91
5 Examples 5
Base Shear Force
There are several options in RF-DYNAM Pro and RFEM to evaluate the base shear force.
• List of equivalent loads in Tables 5.8 and 5.9
The equivalent loads are listed in the results tables 5.8, 5.9, and 5.10 together with the sum of
the equivalent loads. When Mode Shape 1 is selected, the equivalent loads FY are listed in Table
5.9. The sum is provided as 141.29 kN as shown in Figure 5.18. Note that the equivalent loads
are listed twice because accidental torsions have been activated. Thus, the loads are given for
the cases Torsion + and Torsion -. The real base shear force for Mode Shape 1 is FY = 70.65 kN.
Figure 5.18: Sum of equivalent loads ∑ FY resulting from Mode Shape 1, dominant in Y-direction – note
that the equivalent loads are listed twice for Torsion + and Torsion - cases
The sum of equivalent loads resulting from Mode Shape 28 is listed in Table 5.8 (see Figure 5.19).
Again the equivalent loads are listed twice. The real base shear force is FX = 96.83 kN.
Figure 5.19: Sum of equivalent loads ∑ FX resulting from Mode Shape 28, dominant in X-direction
Figure 5.20: Sum of loads ∑ FY for LC 4 resulting from Mode Shape 1, dominant in Y-direction
For the dominant eigenvalue in X-direction, which is exported to LC10 and LC11 respectively,
the base shear force is ∑ FX = 96.83 kN (see Figure 5.21).
92
5 Examples 5
Figure 5.21: Sum of loads ∑ FX for LC 10 resulting from Mode Shape 28, dominant in X-direction
Figure 5.22: Structural system with result beam placed outside of model, including all objects
In LC4, the shear forces Vy in the result beam provide the base shear forces resulting from Mode
shape 1 in Y-direction. In LC10, the shear forces Vz in the result beam provide the base shear
forces resulting from Mode shape 28 in X-direction. Both are displayed in Figure 5.23 and the
results are summarised in Table 5.1.
Figure 5.23: Base shear forces of result beam for LC4 (Vy resulting from mode shape 1 in Y-direction) and
LC10 (Vz resulting from mode shape 28 in X-direction)
93
5 Examples 5
The base shear forces of the result combinations (RC) can also be evaluated, which gives the
base shear forces of the combined results. RC5 contains the results of the SRSS rule in the
Y-direction, RC4 the combined results in the X-direction.
The shear forces in the result beam are shown in Figure 5.24.
Figure 5.24: Base shear forces showing combined dynamic results of result beam for RC5 (Vy in Y-direction)
and RC4 (Vz in X-direction)
The base shear forces of the single storeys are listed in Table 5.1.
Level Z = −8 m Level Z = −4 m ∑F
LC4: Mode Shape 1 - Y-Direction 39.76 kN 30.89 kN 70.65 kN
LC10: Mode Shape 28 - X-Direction 60.84 kN 35.99 kN 96.83 kN
RC5: SRSS - Y-Direction 39.86 kN 30.94 kN 70.80 kN
RC4: SRSS - X-Direction 60.84 kN 35.99 kN 96.83 kN
Table 5.1: Base shear forces of single storeys resulting from dominant eigenvalues and modal combination
94
5 Examples 5
Storey Drift
Maximal displacements can be used to evaluate the inter-storey drift. In accordance with EN
1998-1 [1] Section 4.4.2.2, the influence of the second-order effects (P-Δ effect) can be neglected
when the inter-storey drift sensitivity coefficient 𝜃 is small enough. It is defined as follows:
P ⋅d
𝜃 = tot r ≤ 0.1 (5.9)
Vtot ⋅ ℎ
where
𝜃 : Inter-storey drift sensitivity coefficient
Ptot : Total gravity load at and above the storey considered
dr : Design inter-storey drift, i.e. difference of the lateral displacements ds at the top
and bottom of the storey under consideration
Vtot : Total seismic storey shear
h : Inter-storey height
To evaluate the total gravity load at and above the storey considered, the masses on each FE-node
need to be analysed.
The masses on each FE-node together with the coordinates of those FE-nodes are listed in Table 5.6.
That table can be exported to Excel and filter options can be used to summarise the masses for
each storey. For the roof level Z = −8 m, the mass mZ=−8 m = 19,588.21 kg has to be considered.
The mass at level Z = −4 m consists of the bottom storey, which is mbottom = 17,927.17 kg,
and the mass at the roof level. Altogether the mass to be considered at the level Z = −4 m is
mZ=−4 m = 37,515.38 kg. For the level Z = 0 m, the sum of masses is mZ=0 m = 813.32 kg; those
are not considered as this is the level of support.
The storey shear forces were discussed in the last paragraph. The loads are summarised in Table 5.1.
The height of the storeys is h = 4 m.
The global deformations uY in RC5 (earthquake in Y-direction) are illustrated in Figure 5.25.
Figure 5.25: Global deformations uY resulting from earthquake excitation in Y-direction (RC5)
95
5 Examples 5
The design inter-storey drift for the top level is dr,top = 51.4 − 36.0 = 15.4 mm. For the bottom
level, dr,bottom = 36.0 mm. The inter-storey drift sensitivity coefficient for the top level can be
calculated with
19,588.21/100 ⋅ 15.4/1,000
𝜃top = = 0.02 (5.10)
39.86 ⋅ 4
and for the bottom level with
37,515.38/100 ⋅ 36.0/1,000
𝜃bottom = = 0.11. (5.11)
30.94 ⋅ 4
The limit of 0.1 is exceeded in the bottom level.
The global deformations uX in RC4 (earthquake in X-direction) are illustrated in Figure 5.26.
Figure 5.26: Global deformations uX resulting from earthquake excitation in X-direction (RC4)
The design inter-storey drift for the top level is dr,top = 3.2 − 1.8 = 1.4 mm. For the bottom level,
dr,bottom = 1.8 mm. The inter-storey drift sensitivity coefficient for the top level can be calculated
with
19,588.21/100 ⋅ 1.4/1,000
𝜃top = = 1.13 × 10−3 (5.12)
60.84 ⋅ 4
and for the bottom level with
37,515.38/100 ⋅ 1.8/1,000
𝜃bottom = = 4.69 × 10−3 . (5.13)
35.99 ⋅ 4
In the X-direction of the building, the limit of 0.1 is not exceeded.
In this example the inter-storey drift sensitivity coefficient 𝜃 according to Equation 5.9 in the
bottom level in Y-direction is slightly exceeded. In accordance with EN 1998-1 [1] Section 4.4.2.2 (3),
second-order effects can be approximately taken into account by multiplying the seismic action
effects with a factor of 1/(1 − 𝜃), provided that the sensitivity coefficient 𝜃 is smaller than 0.2.
96
5 Examples 5
Second-Order Effects
• Approximate consideration of P-Δ effects with factor 1/(1 - 𝜃)
To increase results in order to approximately consider P-Δ effects, Factor 1/(1 − 0.11) = 1.124
is used in the dynamic load case DLC2 as shown in Figure 5.27. Only the Y-direction is increased
by this factor.
Figure 5.28: Consideration of axial forces in natural vibration case to consider geometric stiffness matrix
with axial forces taken from CO2
Load combination CO2 combines the load cases with self-weight and imposed loads in accor-
dance with the Design Situation Earthquake. The combination factors 𝛹2 are in accordance
with EN 1990 [4] Table A.1.1. This CO2 is used for the modification of the stiffness matrix, as
this is the design combination which is relevant in an event of an earthquake.
97
5 Examples 5
The settings of the natural vibration case (NVC) influence the natural frequencies of the system.
The NVC has to be assigned to the dynamic load case (DLC) so that the modified frequencies
are used for the calculation of equivalent loads. The settings of DLC3 are shown in Figure 5.29.
Figure 5.29: Modified stiffness matrix assigned to DLC – the frequencies are slightly changed, as well as the
corresponding acceleration values Sa read from the response spectrum (cf Figure 5.13)
The equivalent loads are calculated with those modified frequencies and are then exported to
the main program RFEM. The stiffness modifications are not automatically applied in the load
cases. You need to consider the modified stiffness matrix also to calculate your final results like
deformations and internal forces. The settings as shown in Figure 5.30 have to set manually.
98
5 Examples 5
Figure 5.30: Stiffness modification of load cases exported from DLC3 – the geometric stiffness matrix has to
be taken into account for the calculation of those load cases
The combination factors 𝛹2 are listed in EN1990 [21] Table A.1.1. Note that those combination
factors have already been used in the mass combinations (MCO).
The combination for the seismic design situation has to be defined manually in RFEM. This can be
done using load cases (LC) or the priorly defined load combination CO2. Both options are shown
in Figure 5.31.
Figure 5.31: Seismic design situation in accordance with EN 1990: a) combination with load cases and b)
combination with load combination CO2 defined earlier using the same combination factors 𝛹2
99
5 Examples 5
The use of a load combination (CO) is recommended when nonlinearities exist in the model, or
when second-order effects have to be considered.
The resulting moment MY and the support forces on selected members of the structure are illus-
trated in Figure 5.32.
Figure 5.32: Results of RC10 for seismic design situation showing moments MY and support forces on
selected members
The internal forces are listed in Table 4.6 as shown in Figure 5.33 for member 6.
Figure 5.33: Table 4.6 Member - Internal Forces showing results of RC10 for seismic design situation for
member 6 at node 5
The modal combination rule was applied in form of the equivalent linear combination. The effect
of this can be seen on the corresponding internal forces – they are smaller compared to the
maximum values (displayed bold on the main diagonal) and have varying signs.
The second-order effects were taken into account (1) by an approximate method with the factor
1/(1 − 𝜃) (DLC2 in RF-DYNAM Pro) and (2) by a more exact method with the consideration of the
geometric stiffness matrix (DLC3). For both methods, the final seismic design combination was
built as described above. Figure 5.34 shows a comparison of those internal forces.
100
5 Examples 5
Figure 5.34: Results of seismic design situations considering P-Δ effect by a factor (RC20) and by the geo-
metric stiffness matrix (RC30) showing moments MY and support forces on selected members
The approximate method using the factor 1/(1 − 𝜃) leads to results on the safe side.
With those result combinations for the seismic design situation, the design of the single construc-
tion elements can be performed, for example with the add-on modules RF-STEEL, RF-STEEL EC3,
or RF-CONCRETE.
101
5 Examples 5
5.2 Machine-Induced Vibrations
In the second example, the dynamic excitation of two rotating machines is analysed by a linear
time history analysis in the add-on module RF-DYNAM Pro - Forced Vibrations.
The steel framework that supports the two machines is excited in resonance first, which means
that the frequency of the structure is the same as the frequency of the excitation. Resulting from
resonance, the systems experiences unacceptable large deformations and accelerations. Changing
the cross-sections and thus modifiying the natural frequency of the structure is compulsory.
Figure 5.35: Structural system with coloured cross-sections and materials and openings where machines
are located
In the openings in the lower and upper levels, two machines are located. The machines themselves
are not modelled – they are represented by their dead loads and the centrifugal forces that they
produce.
The self-weight of the steel frame in the +Z-direction and an additional surface load of 0.5 kN/m2
are defined in LC1. This load case will be imported as mass in RF-DYNAM Pro.
Each of the machines has a self-weight of m = 1.84 t. This dead load acts as line loads on the
adjacent members. It is defined in LC2 as illustrated in Figure 5.36.
102
5 Examples 5
Figure 5.36: Self-weight of machines, defined as line loads on members around openings
The mass of the rotor mR = 0.5 t rotates with 310 rpm with an eccentricity of e = 19 mm.
This movement is illustrated in Figure 5.37. The centrifugal force Fr acts perpendicularly to the
rotational movement.
Figure 5.37: Rotation of rotor mass mR within machine and resulting centrifugal force Fr
We will define time diagrams later in RF-DYNAM Pro when the excitation frequency 𝜔e is entered.
The value of the centrifugal force is defined with:
Fr = 𝜔e2 ⋅ e ⋅ mR = 10.01 kN (5.16)
103
5 Examples 5
The modelling in RFEM is simplified by defining only the horizontal and vertical centrifugal forces,
which is of sufficient accuracy. The horizontal component is delayed in time and defined with a
phase shift in the periodical time diagram. This will be considered when the time diagrams are
created in RF-DYNAM Pro.
Again, the centrifugal forces are defined as line loads along the openings. The vertical and hori-
zontal components are arranged in separate load cases. The loads in LC3 and LC4 determine the
size and the position of the time-invariant excitations. Those load cases are shown in Figure 5.38.
Figure 5.38: Centrifugal forces Fr defined as vertical and horizontal line loads along openings in two load
cases
Figure 5.39: General tab of RF-DYNAM Pro with Forced Vibrations module and Time Diagrams selected
104
5 Examples 5
Definition of Masses
In this example, only the self-weight of the steel frame and of the machines are of relevance. Thus,
the two relevant load cases LC1 and LC2 are imported as masses into RF-DYNAM Pro. Two mass
cases are defined: MC1 is shown in Figure 5.40, MC2 is defined analogously by importing LC2.
Figure 5.40: Mass case MC1 with self-weight imported from LC1 (option From self-weight of structure cleared
as self-weight has been considered in LC1)
The two mass cases of self-weight are combined in mass combination MCO1 using factors of 1.0.
This is shown in Figure 5.41.
105
5 Examples 5
Figure 5.42: Natural vibration case NVC1 employing a diagonal mass matrix and with masses acting in all
directions and about all axes
The relevant mode shape 6 in the X-direction with a natural frequency of f = 5.192 Hz is illustrated
Figure 5.43. The other eigenvalues are also listed together with the effective modal mass factors.
The dominant eigenvalues are highlighted.
Figure 5.43: Relevant mode shape 6 in X-direction illustrated in work window and all calculated eigenvalues
tabulated with effective modal mass factors
106
5 Examples 5
With its frequency of f = 5.192 Hz, the illustrated mode shape 6 is very close to the excitation
frequency of the machines which is fe = 5.167 Hz.
For time history analyses, the use of the consistent mass matrix is recommended in general. The
consistent matrix distributes the masses more equally over the members, which results in more
local mode shapes to be identified. In this example, however, a diagonal mass matrix is sufficient.
Time Diagrams
The centrifugal forces as defined in LC3 and LC4 shall act over time as harmonic excitations. There-
fore, the definition of time diagrams is required in RF-DYNAM Pro. The excitation frequency is
𝜔 = 32.463 rad/s (see Equation 5.15). For the vertical component, no phase shift is defined in
the periodic time diagram. The horizontal component has to be shifted to express the time delay,
however. This is done with a phase shift of 𝜑 = 𝜋/2. The factor k is set to 1 because the magnitude
of the load was defined in load cases LC3 and LC4.
The definition of the time diagrams is shown in Figure 5.44 and Figure 5.45.
Figure 5.44: Definition of periodic time diagram for vertical component of centrifugal force
Figure 5.45: Definition of periodic time diagram for horizontal component of centrifugal force
107
5 Examples 5
Time History Analysis
In the dynamic load case DLC1, the load cases LC3 and LC4 of RFEM are connected to the time
diagrams TD1 and TD2 that have been defined in RF-DYNAM Pro. The linear implicit Newmark
solver is chosen as the solving algorithm.
In the Time History Analysis tab, the time steps, the maximum time for calculation, and the export
options are defined as shown in Figure 5.46.
Figure 5.46: Settings for linear time history analysis of DLC1 with connected load cases and time diagrams,
time steps, and export options
The time step for calculation is set to 𝛥t = 1 ms to achieve the required accuracy of the results.
The relevant frequency of the structure and those of the excitation are about 5.2 Hz. The required
time step can be estimated with the following formula:
𝛥t = 1/(20 f) = 9.6 ms (5.17)
To ensure also exact velocities and acceleration, the time step is chosen even smaller for this
example.
In the Calculation Parameters sub-tab, the acting masses and the mass matrix options are set. This
is shown in Figure 5.47. The Diagonal matrix considering all degrees of freedom is selected.
108
5 Examples 5
Figure 5.47: Settings of DLC1 in Calculation Parameters tab for acting masses and mass matrix
In the Damping sub-tab, the structural damping is defined. The settings for damping used in this
example are shown in Figure 5.48.
Figure 5.48: Settings of DLC1 in Damping tab for Rayleigh coefficients 𝛼 and 𝛽
The linear implicit Newmark solver requires Rayleigh damping coefficients. To convert available
Lehr's damping values to Rayleigh coefficients, the following relation can be used:
𝛼 𝛽 𝜔i
Di = + (5.18)
2 𝜔i 2
In this example, the most important eigenvalues in the X- and Y-directions shall be damped with
D = 0.01. The relevant eigenvalues (1 and 6) have angular frequencies of 𝜔1 = 7.623 rad/s and
𝜔6 = 32.620 rad/s. In the Lehr's to Rayleigh damping calculator which can be accessed via the
button, the angular frequencies and the damping values are defined as seen in Figure 5.49.
After [OK], the corresponding Rayleigh coefficients 𝛼 = 1.235839E − 01 and 𝛽 = 4.9692E − 04
are set in the Damping tab (see Figure 5.48).
109
5 Examples 5
Figure 5.50: Displacement u resulting from time history analysis with dynamic envelope in work window and
displacement uX at node 73 displayed versus time in Time Course Monitor showing resonance
110
5 Examples 5
Adjustment of Structural System
To avoid that kind of resonance case as seen in Figure 5.50, the cross-sections of the beams and
columns are modified as illustrated in Figure 5.51. This is done to achieve a considerable shift of
the dominant natural frequency.
Figure 5.51: Structural system with modified cross-section of beams and columns
The changed frequencies of the structure are listed in Figure 5.52. The dominant eigenvalues are
highlighted.
Figure 5.52: Eigenvalues of modified structural system showing dominant mode shape 12 in X-direction
with frequency f12 = 4.00 Hz
The dominant frequency is now about f = 4.00 Hz and, therefore, shifted by 1.2 Hz compared to
the critical frequency.
Additionnally, the Rayleigh damping coefficients need to be adjusted so that the dominant frequen-
cies are damped with D = 0.01. With the modified frequencies, the coefficients are determined
as 𝛼 = 7.82 × 10−02 and 𝛽 = 6.71 × 10−04 .
The other settings are kept as before. Then the time history analysis is performed again.
111
5 Examples 5
As a result, the displacements are much smaller. In the Time Course Monitor, the effect of damping
can be well seen compared to the case of resonance. There the root mean square value (RMS),
which is calculated from the displacement values from t = 0 s to t = 10 s, is shown as well.
Figure 5.53: Displacement u of altered structural system with dynamic envelope in work window and
displacement uX at node 73 displayed versus time in Time Course Monitor and RMS value of
6.9 mm showing no resonance
This example demonstrates how to analyse machine-induced vibrations by RFEM and the add-on
module RF-DYNAM Pro. It shows how important it is to analyse the frequencies of the structure in
order to eliminate resonance cases.
112
A Literature A
Literature
[1] EN 1998-1: Design of structures for earthquake resistance Part 1: General rules, seismic actions
and rules for buildings. CEN, Brussels, 2004.
[2] DIN 4149: Buildings in German earthquake areas - Design loads, analysis and structural design
of buildings. Beuth Verlag GmbH, Berlin, 2005.
[3] IBC2012: International Building Code. International Code Council, 2011.
[4] EN 1990: Basis of structural design. CEN, Brussels, 2002.
[5] Christian Barth and Walter Rustler. Finite Elemente in der Baustatik-Praxis. Beuth, Berlin, 2.
edition, 2013.
[6] Klaus-Jürgen Bathe. Finite Element Procedures. Prentice Hall, 1996.
[7] Hans-Günter Natke. Baudynamik. B. G. Teubner, Stuttgart, 1989.
[8] Edward L. Wilson. Three-Dimensional Static and Dynamic Analysis of Structures. Computer
and Structures, Inc., Berkeley, California, USA, 3rd edition, 2002.
[9] Joseph W. Tedesco, William G. McDougal and C. Allen Ross. Structural Dynamics : Theory and
Applications. Addison-Wesley, 1st edition, 1999.
[10] A.K. Chopra. Dynamics of Structures - Theory and Applications to Earthquake Engineering. Pren-
tice Hall, 2001.
[11] C. Katz. Anmerkung zur Überlagerung von Antwortspektren. D-A-CH Mitteilungsblatt, 2009.
[12] A.K. Gupta. Response Spectrum Method in Seismic Analysis and Design of Structures. CRC Press,
1992.
[13] E.L. Wilson, A. Der Kiureghian and E.P. Bayo. A replacement for the SRSS method in seismic
analysis., 1981.
[14] A. Der Kiureghian. A response spectrum method for random vibration analysis of MDF system.,
1981.
[15] Abaqus. Abaqus 6.14 - Theory Guide. Simulia., 2014.
[16] Abaqus. Lecture 1 - Overview of ABAQUS/Explicit. Abaqus Inc., 2001.
[17] Ivan Němec and Vladimír Kolář. Finite Element Analysis of Structures - Principles and Praxis.
Shaker Verlag, Aachen, 2010.
[18] U.Stelzmann, C.Groth and G.Müller. FEM für Praktiker - Band 2: Strukturdynamik. Expert Ver-
lag, 2008.
[19] ANSYS Inc.. Theory Reference for the Mechanical APDL and Mechanical Applications, Release
15.0., 2013.
[20] K. Meskouris, K.-G. Hinzen, C. Butenweg and M.Mistler. Bauwerke und Erdbeben. Vieweg und
Teubner, Berlin, 3. edition, 2011.
[21] EN 1990: Grundlagen der Tragwerksplanung. Beuth Verlag GmbH, Berlin, 2010.
[22] Program Description RFEM 5. DLUBAL GmbH, 2013.
[23] C. Petersen. Dynamik der Baukonstruktionen. Vieweg & Sohn, 1996.
[24] Manual SAP2000. CSI Analysis Reference Manual - For SAP2000, ETABS, and SAFE2007., 2007.
[25] V. Červenka, L. Jendele and J. Červenka. ATENA Program Documentation Part 1 - Theory. Čer-
venka Consulting s.r.o., 2014.
[26] OENORM B4015:2007: Design loads in buildings - Accidental actions: Seismic actions. Austrian
Standards Institute, 2007.
[27] NBC2010: National Building Code of Canada. Canadian Commission on Building and Fire
Codes, National Research Council of Canada, Thirteenth edition, 2010.
113
B Index B
Index
A EN 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17, 82
Acceleration . . . . . . . . . . . . . . . . . . . . . . 26, 37, 66 EN 1998-1 . . . . . . . . . . . . . . . . . . 17, 36, 47, 82, 87
Acceleration of Gravity . . . . . . . . . . . . . . . . . . . 13 Equivalent Linear Combination. .34, 35, 47, 49
Acceleration of Response Spectra . . . . . . . . . 36 Equivalent Load . . . . . . . . . . . . . . . . . 7, 46, 51, 72
Accelerogram . . . . . . . . . . . . . . . . . . . . . 25, 26, 37 Equivalent Load Analysis . 11, 23, 46, 72, 82, 89
Accidental Torsional Actions . . . . . . . 47, 73, 75 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82, 102
Acting Masses . . . . . . . . . . . . . . . . . . . 20, 43, 108 Excitation Direction . . . . . . . . . . . . 34, 38, 39, 47
Angular Frequency . . . . . . . . . . . . . 30, 40, 41, 56 Explicit Solver . . . . . . . . . . . . . . . . . . . . . . . . 32, 40
Assign Accelerogram . . . . . . . . . . . . . . . . . . . . . 38 Export . . . . . . . . 32, 35, 42, 49, 63, 69, 70, 73, 75
Assign Response Spectrum . . . . . . . . . . . . 34, 47
Axial Forces . . . . . . . . . . . . . . . . . . . . . . . . . . 22, 97 F
Axial Strain for Cables and Membranes . . . . 50 Failing Member . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Failing Supports . . . . . . . . . . . . . . . . . . . . . . . . . 22
B FE Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22, 43
Base Shear Force. . . . . . . . . . . . . . . . . . . . . . . . .92 Forced Vibrations . 5, 11, 26, 29, 33, 40, 60, 102
Building Standards . . . . . . . . . . . . . . . . . . . 24, 87 Frequency . . . . . . . . . . . . . 10, 19, 40, 41, 56, 105
Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
C
Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . .11, 52 G
Calculation Parameters . . . . . . . . . . . . . . . 18, 43 Global Parameter Settings . . . . . . . . . . . . . . . . 50
Cancel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Global Stiffness Modification. . . . . . . . . . . . . .22
Centre of Mass . . . . . . . . . . . . . . . . . . . . . . . 12, 13
Check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11, 52 H
Combination Factors . . . . . . . . . . . 12, 17, 83, 99 Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Combination Rules . . . . . . . . . . . . . 34, 35, 47, 49
Complete Quadratic Combination (CQC)34, 47 I
Consistent Mass Matrix. . . . . . . . . . . . . . . .21, 43 ICG Iteration. . . . . . . . . . . . . . . . . . . . . . . . . . . . .21
Implicit Newmark Solver . . . . . . . . . . 32, 40, 108
D Initial Condition . . . . . . . . . . . . . . . . . . . . . . 22, 44
Damping . . . . . . . . . . . . . . . . . . . . . . . . . 35, 44, 49 Initial Deformation . . . . . . . . . . . . . . . . . . . . . . . 42
Dashpot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40, 41 Input Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Deactivated Member . . . . . . . . . . . . . . . . . . . . . 22 Instability Detection . . . . . . . . . . . . . . . . . . . . . 50
Decimal Places . . . . . . . . . . . . . . . . . . . . 11, 51, 81
Demo Version . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 L
Details . . . . . . . . . . . . . . . . . . . . . . . . 11, 12, 13, 50 Lanczos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Diagonal Mass Matrix . . . . . . . . . . . . . . . . . 20, 43 Lehr's Damping . . . . . . . . . . . . . . . . . . . 35, 44, 49
Direction of Gravity . . . . . . . . . . . . . . . . . . . . . . 50 Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Direction of Mass . . . . . . . . . . . . . . . . . . . . . 20, 43 Line Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13, 16
Directional Combination . . . . . . . . . . . . . . 35, 49 Linear Solvers . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Display Response Spectra . . . . . . . . . . 23, 36, 50 Linear Time History Analysis . . . . . . . . . . . . . . 11
Dynamic Load Case . . . . . . . . . . . . . . . . . . . 11, 31 Load Case . . . . . . . . . . . 12, 32, 39, 42, 49, 69, 73
Loading - Time Diagram Set . . . . . . . . . . . . . . 39
E
Earthquake Analysis. . . . . . . . . . . . . . . . . . . . . .82 M
Earthquake Recordings . . . . . . . . . . . . . . . 26, 27 Machine-Induced Vibrations . . . . . . . . . . . . . 102
Effective Modal Mass . . . . . . . . . . . . . . . . . . . . . 58 Mass . . . . . . . . . . . . . . . . . . . . . 12, 14, 58, 83, 105
Effective Modal Mass Factor . . . 19, 36, 58, 105 Mass Case . . . . . . . . . . . . . 12, 17, 20, 43, 83, 105
Eigenvalue . . . . . . . . . . . . . . . . . . . . . . . 19, 40, 56 Mass Case Type . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Eigenvalue Solver . . . . . . . . . . . . . . . . . . . . . . . . 21 Mass Combination . . . . . . . . . . . . . 17, 20, 43, 83
114
B Index B
Mass Conversion Type . . . . . . . . . . . . . 12, 13, 50 Result Combination32, 35, 42, 49, 63, 70, 75, 99
Mass Import . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Mass Matrix . . . . . . . . . . . . . . . . . . 15, 20, 43, 108 Root Mean Square . . . . . . . . . . . . . . . . . . . . . . . 68
Mass Moments of Inertia . . . . . . . . . . . . . . . . . 15 Root of the characteristic polynomial . . . . . . 21
Masses in Mesh Points . . . . . . . . . . . . . . . . . . . 58 Rotation . . . . . . . . . . . . . . . . 20, 34, 38, 43, 57, 58
Maximum Time . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Member Internal Force . . . . . . . . . . . . . . . . 62, 65 S
Member Mass . . . . . . . . . . . . . . . . . . . . . . . . 13, 16 Scaling of mode shapes . . . . . . . . . . . . . . . . . . 19
Modal Analysis . . . . . . . . . . . . . . . . . . . . . . . 32, 40 Second-Order Effect . . . . . . . . . . . . . . . . . . 22, 97
Modal Combination . . . . . . . . . . . . . . . . . . 34, 47 Self-Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Modal Damping . . . . . . . . . . . . . . . . . . . . . . 35, 49 SDOF Oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Modal Mass . . . . . . . . . . . . . . . . . . . . . . . . . . 19, 58 Set of Supports . . . . . . . . . . . . . . . . . . . . . . . 33, 38
Modal Responses . . . . . . . . . . . . . . . . . . . . . 34, 47 Signed Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Mode Shape Selection . . . . . . . . . . . . . 36, 49, 90 Solvers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21, 40
Mode Shapes . . . . . . . . 10, 19, 36, 46, 49, 56, 57 Square Root of the Sum of the Squares (SRSS)
Multi-Modal Response Spectrum Analysis. .23, 34, 35, 47, 49
36, 46, 89 Standardized Displacement . . . . . . . . . . . . . . 57
Multi-Point Excitation. . . . . . . . . . . . . .23, 33, 38 Standardized Rotation . . . . . . . . . . . . . . . . . . . 57
Start Calculation . . . . . . . . . . . . . . . . . . . . . . . . . 52
N Start RF-DYNAM Pro. . . . . . . . . . . . . . . . . . . . . . .8
Natural Frequency . . . . . . . . . . . . . . . . . . . . . . . 56 Stationary State . . . . . . . . . . . . . . . . . . . . . . . . . 44
Natural Period . . . . . . . . . . . . . . . . . . . . . . . . 36, 56 Stiffness Modification . . . . . . . . . . . . . . . . . 22, 97
Natural Vibration Analysis 4, 10, 18, 56, 85, 105 Stiffness RF-CONCRETE . . . . . . . . . . . . . . . . . . . 22
Natural Vibration Case . . . . . . 10, 18, 20, 32, 56 Storey Drift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Navigator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Structural Damping . . . . . . . . . . . 35, 44, 49, 109
Neglect Mass . . . . . . . . . . . . . . . . . . . . . 13, 14, 50 Subspace Iteration . . . . . . . . . . . . . . . . . . . . . . . 21
Nodal Acceleration . . . . . . . . . . . . . . . . . . . 61, 66 Sum of Masses. . . . . . . . . . . . . . . . . . . . . . . . . . .13
Nodal Deformation . . . . . . . . . . . . . . . . . . . 61, 65 Support Force . . . . . . . . . . . . . . . . . . . . . . . . 61, 65
Nodal Mass . . . . . . . . . . . . . . . . . . . . . . . . . . 13, 15 Surface Basic Strains . . . . . . . . . . . . . . . . . . . . . 63
Nodal Velocities . . . . . . . . . . . . . . . . . . . . . . 61, 66 Surface Basic Stresses . . . . . . . . . . . . . . . . . 63, 66
Nonlinear Solvers . . . . . . . . . . . . . . . . . . . . . . . . 40 Surface Internal Force . . . . . . . . . . . . . . . . . 63, 66
Nonlinear Time History Analysis . . . . . 6, 11, 40 Surface Mass . . . . . . . . . . . . . . . . . . . . . . . . . 13, 16
O T
Open RF-DYNAM Pro . . . . . . . . . . . . . . . . . . . . . . 8 Time Course Monitor . . . . . . . . . . . . . . . . . . . . . 67
Time Diagram. . . . . . . . . . . . . . . . . . . .29, 39, 107
P Time History Analysis . . 29, 32, 37, 64, 102, 108
P-Delta Effect . . . . . . . . . . . . . . . . . . . . . . . . 22, 97 Time Steps . . . . . . . . . . . . . . . . . . . . . . . . . . 41, 108
Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47, 73
Periodic Excitation. . . . . . . . . . . . . . . . . . .30, 107 Torsional Moment . . . . . . . . . . . . . . . . . 47, 73, 75
Printout Report. . . . . . . . . . . . . . . . . . . . . . .69, 80 Transient Excitation . . . . . . . . . . . . . . . . . . . . . . 29
Program Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Trial Version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Project Navigator . . . . . . . . . . . . . . . . . . . . . . . . . 9
U
R Unit Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Rayleigh Coefficients . . . . . . . . . . . . . . . . 44, 109 Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11, 51, 81
Resonance. . . . . . . . . . . . . . . . . . . . . . . . .102, 110
Response Spectrum . . . . . . . . . . . . 23, 25, 60, 87 V
Response Spectrum Analysis 11, 33, 46, 60, 72, Velocities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
89 Viscous damping element . . . . . . . . . . . . . 40, 41
Result Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
115