Yumoto 2006

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

SAE TECHNICAL
PAPER SERIES 2006-01-0693

Integrated Brake Disc Design System


Hiroyuki Yumoto and Toshikazu Okamura
KIRIU Corporation

Reprinted From: Brake Technology 2006


(SP-2017)

2006 SAE World Congress


Detroit, Michigan
April 3-6, 2006

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

The Engineering Meetings Board has approved this paper for publication. It has successfully completed
SAE's peer review process under the supervision of the session organizer. This process requires a
minimum of three (3) reviews by industry experts.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.

For permission and licensing requests contact:

SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: permissions@sae.org
Tel: 724-772-4028
Fax: 724-776-3036

For multiple print copies contact:

SAE Customer Service


Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-0790
Email: CustomerService@sae.org

ISSN 0148-7191
Copyright  2006 SAE International

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract to Secretary, Engineering Meetings Board, SAE.

Printed in USA
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

2006-01-0693

Integrated Brake Disc Design System


Hiroyuki Yumoto and Toshikazu Okamura
KIRIU Corporation

Copyright © 2006 SAE International

ABSTRACT therefore plays a significant role in the design process


for automotive components.
We have constructed an original brake disc design
system by standardizing, automating, and speeding up Table 1 Required performance, evaluated factor and
each design process. This system consists of two steps. design methodology
In the first step, a designer, without professional
knowledge or skills regarding CAE, can easily carry out Requirement Evaluated factor Design methodology
FEA by manipulating pull-down menus, parametric Basic x Stress x Classical
design, and automated modeling and meshing. Further, strength mechanics, partly
our new computer program automatically identifies each Thermal x Temperature x Thermal-transfer
eigenmode. It takes less than two hours for a complete strength x Thermal stress analysis
FEA of a new design. In the second step, the developed x Thermal-stress and
postprocessor makes it easier and faster to compare the -deflection analysis
simulated results of many design alternatives and Judder x Thermal x Thermal-stress and
determine the optimum solution. deflection -deflection analysis
Squeal x Resonant x Eigenvalue
Through the accumulation of vast numbers of FEA frequency analysis
cases and tests for validation, we have obtained the x Coupling
knowledge of the effects of disc configurations,
dimensions, and material properties on resonant T. Valvano et al. [1] utilized a PC-based computer
frequencies and thermal deflections. As a result, we can program to calculate the necessary thermal parameters
now release CAD data of a new brake disc in only two and apply the results as input to a finite-element-based
days from scratch and provide the highest design quality thermal-stress analysis for a brake disc. The advantage
by making the best of classical FEA methodology. of using a PC-based tool is its fast run-time and friendly
user-interface with its many customized features. In
INTRODUCTION various development processes for other automotive
systems, such as engines and transmissions, design
The fundamental function of a brake system is to support systems applying CAE have also been
decelerate or stop a vehicle. However, sometimes developed in response to the accelerated requirements
drivers and passengers may suffer some discomfort due in regards to development lead-time reduction and
to squeal or judder during braking. It is necessary to quality improvement [2-4].
reduce noise, vibration, and harshness (NVH) problems
and respond to the increasing customer needs Considering this background, we constructed a brake
concerning braking performance and weight reduction. disc design system that can be easily operated by not
In addition, shortening the lead-time for a new product only a CAE specialist but also an ordinary design
development has become ever more important. engineer, provide informative simulation results, and
Enhancing development competitiveness has therefore support decision making regarding new disc designs.
become the most significant issue for a brake- We have established the brake disc design system, by
component supplier. standardizing, automating, and speeding up each
process of design and CAE in order to ease the time-
Obtaining an optimum solution for a new brake disc consuming studies on basic strength, thermal stress and
needs time-consuming studies to satisfy various deflection, and resonance frequencies. We are
requirements, such as basic strength, thermal stress and improving the simulation accuracy and high-speed
deflection, and NVH performance. Nowadays, finite design process with CAE by making the best of classical
element analysis (FEA) is mainly used for these studies FEA methodology. In the following sections, we describe
as shown in Table 1. Computer-aided engineering (CAE) the developed system, the key factors for its success,
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

and some correlation data between CAE simulation PREPROCESS AND MAIN PROCESS
results and experiments.
In the main menu window, which has three frames, we
TOOLS AND FEATURES OF THE DEVELOPED can start the modeling of a new disc, browse
DESIGN SYSTEM accumulated simulation results, and manage simulation
files as well. We define a new project in the first frame or
OUTLINE OF THE SYSTEM select an existing one from a project list, and register a
new design or select a registered one in the second
Figure 1 shows the flow diagram of CAE processes used frame. The simulation result files of the selected design
in the developed system. This system consists of two are listed in the third frame.
parts. The first part simulates the individual performance
of each design. A design engineer, without professional Table 2 lists the operations classified into manual and
knowledge or skills with CAE, can easily carry out FEA. automated ones, which are operated by our new system,
The second part is a decision-support system that in the preprocess and main process.
compares the simulation results of several alternative
designs. Table 2 Manual and automated operations in preprocess
and main process
CAE specialists, and not ordinary design engineers,
have traditionally been in charge of CAD and CAE Manual Automated Human
Step
operations. This is because special knowledge and operation operation verification
complicated operations, such as modeling, meshing, Parametric x Input x 3D-data Basic-design
and boundary conditioning, are necessary to carry out design dimension generation integrity
CAE simulation. With our new system, a meshed model x Select x 2D drawing
is created and the analysis is processed automatically Material x Mass
by inputting dimensions, vehicle specifications, and calculation
braking schedules in each operational window, and Basic- x Command x Meshed- Meshed-
clicking the execution button. model execution model model
creation creation integrity
We have developed PC-based design-support tools for Eigenvalue x Command x Full-model Results after
preprocessing and postprocessing with the analysis execution generation postprocess
MSC.Nastran™ solver. In addition, we have x Calculation
programmed application tools ourselves on the Thermal- x Input x Boundary Analysis-
Windows® platform with Microsoft® Visual Studio® and analysis condition conditioning model
structured query language (SQL). In its early stage of modeling integrity
the development, this system won the 2002 Windows Thermal x Command x Calculation Results after
Digital Engineering System Award for Encouragement analysis execution postprocess
from Nikkei Business Publications, Inc., a leading
publisher of business and technical information This system has the flexibility needed to verify the
magazines in Japan [5]. results of each operation and correct them manually if
necessary. This feature assures the extendibility for a

Fig. 1 Brake disc design system block diagram


Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

new disc configuration. This also prevents the design each part. Using the pull-down menu at the top of the
process in the developed system from being a black box right frame, we can select registered standard types with
for ordinary design engineers. Using the system, design various connecting parts between the hat and friction
engineers can get opportunities to obtain knowledge on plate. In addition to this sectional view, other views for a
the brake system behavior and FEA operational skills. fixing plate and ventilation passages with vanes are
available. All the dimensions of a brake disc can
Parametric Design therefore be input parametrically with this tool.

At a glance, disc configuration seems to be simple. If we select a gray-cast-iron grade from the pull-down
However, it cannot be defined as a single standard menu, disc material properties necessary for FEA, such
configuration, because its details have a diversity to as density, Young’s modulus, and thermal conductivity,
meet various customer requirements through controlling are linked to the created 3D model data. We can monitor
eigenmode alignment and thermal deflection. We the model's conformity to the weight target by using a
analyzed a lot of disc configurations, and found that the mass-calculation utility of our modeling tool in this step,
connecting part between the hat and the friction plate without using a commercial 3D-CAD system. This
causes most of the diversity. We therefore defined basic modeling tool has another utility for model data export,
standard configurations of a disc by classifying and and 3D-CAD data for a newly designed disc can be
combining the following components: easily generated. At present, our system can export
model data only to a certain commercial CAD system,
x Disc(s) for friction plate but we are going to expand this function to other CAD
x Ventilation passages with vanes and their circular systems.
periodicity
x Connecting part between the hat and friction plate in Basic Meshed-Model Generation
the shapes defined with arcs and straight lines
x Cylinder portion of hat Most commercial CAE tools have an automated mesher,
x Another disc for fixing plate with holes which mainly uses tetrahedron elements that makes
automated meshing easier. However, they are
Figure 2 shows a window for model creation by sometimes inferior to hexahedron elements in view of
parametric design tools. The left frame shows a simulation accuracy for a model that has fewer degrees
sectional view of standardized disc configurations. They of freedom to reduce calculation time. Moreover in some
are composed of various disc types, such as solid and cases, a tetrahedron element causes poorer
ventilated, drum-in-hat, and inner-hat (also known as reproducibility of meshed models that have the same
reverse-hat, inside-out, or out-board-flow) discs. basic configuration with small difference in several
Dimensions of a disc are listed in the right frame and we dimensions. Therefore, we automated the meshing
can input each dimension into designated cells in the list. operation with hexahedron elements to achieve higher
calculation accuracy and shorter processing time. We
analyzed the operational procedures of several CAE
operators, and standardized the meshing process with
the highest accuracy, least operations, and best
productivity.

If we execute the meshed model generation, the


modeling and meshing tool reads the input parametric
data for a 3D model, and automatically generates a
basic meshed-model of a symmetrical section with a pie-
slice shape, directly from the parametric model data. In
other words, our system does not mesh a 3D solid
model of a disc, but creates a 3D meshed-model from
scratch. We can verify the meshed model on the monitor
screen, and correct it manually if necessary.

Eigenvalue Analysis

If we click the execution button, a full meshed-model is


Fig. 2 Window for parametric design generated from the selected basic meshed-model of a
pie-slice shape and necessary material properties linked
When we finish inputting dimensions into the right frame, in the parametric design process are applied to the
the disc shape in the left frame is transformed meshed model. The generated full meshed-model is
proportionally to the actual size on a prescribed scale. then transferred to a solver, and eigenvalue analysis is
We can grasp the actual shape at first glance and easily automatically processed.
judge the preliminary validity of the design, such as
thickness, stiffness, strength, and their balance between
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

Thermal Analysis

Thermal analysis is divided into two steps in the same


way as described in the references [1, 6, and 7]. The
first step is a thermal-transfer analysis, where the
temperature of the meshed model is calculated in
accordance with a prescribed schedule with repetitive
braking and cooling. The second step is a thermal-stress
analysis, where the thermal stress and deflection of the
model are simulated under the calculated temperature
distribution at a prescribed moment. If we input
simulation conditions and click the execution button, the
following steps are sequentially completed.

First, we input the fundamental specifications for the


vehicle and brake system, such as gross vehicle weight
and tire dynamic radius. Secondly, we input the
conditioning parameters for a braking schedule, such as
Fig. 3 Window for results of eigenvalue analysis
initial vehicle speed, initial temperature, deceleration,
and brake-application interval. Finally, if we click the
Furthermore, without a moment's delay (which might
execution button, the system automatically recalculates
disturb the thinking process of the design engineer), the
and applies boundary conditions to the basic meshed-
shape of any selected eigenmode is shown visually like
model of a pie-slice shape, and a thermal-transfer
the animation images in the right frame of Figure 3. The
analysis of the model is processed. The calculated
mode shape can be determined with the help of this
temperature distribution and other boundary conditions
animation, even if a disc model has an eigenmode that
are then applied to the basic meshed-model. The
is difficult to be identified automatically.
meshed model with the temperature distribution is finally
transferred to a solver, and a thermal-stress analysis of
Results of Thermal Analysis
the model is processed.
Figure 4 shows a window for displaying the results of the
POSTPROCESS FOR INDIVIDUAL ANALYSIS
thermal-transfer analysis. If we click the execution button,
the postprocessor reads the thermal-analysis results and
We have developed postprocess tools that automatically
provides various visual expressions. The horizontal axis
abstract the simulated results and express them visually.
of the graph in the lower right frame indicates time
We selected specific elements and nodes that are
measured in seconds, and the vertical axis indicates
necessary for the succeeding evaluation of the
temperature measured in degrees Celsius. This is a
performance of a simulated design. The postprocessor
typical simulation result for the JASO fade test code.
identifies the specified elements and nodes, and
The temperature change over time at a prescribed
abstracts the calculated values at the identified nodes
position, for example, on the friction surface of a disc
from the simulated results. The postprocessed data are
along the brake effective radius, can be evaluated.
saved in CSV format and referred to later in the
decision-support system for comparing several design
alternatives. These data are also accumulated in our
knowledge base, which is utilized for designing another
brake disc and for studying the effects and sensitivity of
various disc parameters, such as dimensions, gray-cast-
iron material properties, and their variations.

Results of Eigenvalue Analysis

Figure 3 shows a window for displaying the results of the


eigenvalue analysis. When the execution button is
clicked, the postprocessor automatically abstracts only
the necessary simulated results and identifies each
eigenmode from the eigenfrequency and the calculated
displacements at predetermined nodes. The
postprocessor subsequently lists the identified results.
The upper half of the left frame of Figure 3 contains the
list of the identified eigenmodes and frequencies. The
lower part of the left frame contains indexes for Fig. 4 Window for results of thermal-transfer analysis
evaluating the coupling of in-plane and out-of-plane
modes within a disc.
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

Figure 5 shows the window for displaying the results of and calculated mass of design alternatives. Even if no
the thermal-stress analysis. The postprocessor solution completely meets all the requirements, we can
automatically calculates and shows the indexes for select the second best or think of another design
evaluating thermal-performance requirements in the alternative from those reports.
lower left frame, such as the maximum stress at the
friction surface and its ratio to the acceptable strength ACHIEVEMENTS OF THE DEVELOPED SYSTEM
for the selected material. The color contouring
expressions of temperature, stress, and deflection help REDUCTION IN CAE LEAD TIME
in evaluating the thermal performance of the simulated
design. Figure 6 plots the chronological change in the average
time for an FEA case and the volume of analyses for
every half year from 1999 to 2004. An FEA case here
means one eigenvalue or thermal analysis, from model
creation to a visual report for each analysis. The
horizontal axis indicates the passage of time measured
in six-month periods, the left vertical axis indicates the
average analysis time in hours shown by dots, and the
right vertical axis indicates the number of analysis cases
shown by bar chart.

Fig. 5 Window for results of thermal-stress and -


deflection analysis

DECISION-SUPPORT SYSTEM

The decision-support system is a reporting tool for


helping to determine an optimum design. By selecting
design cases for comparing performances from the list
of simulation result files in the main menu window, the
decision-support tool reads the postprocessed data of Fig. 6 Change in average time and volume of FEA case
each design, and then shows a comparison of the
evaluation indexes for required performance of selected We started to apply the first phase of the developed
design alternatives. As a result, it is easy to determine system from fiscal year 2001, the second phase from
the optimum design that satisfies the customer 2002, and have added support tools one after another
requirements. since then. The system has reduced the time for a CAE
analysis of a new disc from two days to less than two
Our eigenvalue-analysis report shows the lists and hours, and a design engineer can carry out simulations
graphs of frequencies for the prescribed eigenmodes of more than five different designs a day. The lead-time
and compares various design alternatives. It is therefore reduction before 2001 was partly due to the operational
easy to compare the simulated design cases from the experience of engineers and the progress in computer
viewpoint of potentials of coupling of adjacent in-plane hardware. We can now release CAD data of a new
and out-of-plane modes and resonance between the brake disc with a standard configuration in just two days
disc and other neighboring components, such as from scratch. This is a significant reduction in
calipers and pads. The thermal-analysis reports, on the development time, one fifth of three years ago. This
other hand, show the graphs that give a full comparison system has been constructed following our guiding
of temperature, safety allowance for thermal strength, principle of "ikki-seisan" in the KIRIU production system,
and thermal deflection among design alternatives. which means manufacturing products within each site
and each production line to eliminate “muda” (i.e.,
In the development process of a new brake disc, various "waste" in Japanese) completely.
customer requirements need to be considered in
combination with each other. We can determine the best
solution for satisfying multiple requirements by
examining the eigenvalue and thermal analysis reports
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

Design Quality Improvement KEY FACTORS FOR SUCCESS

This system has brought about many intangible Figure 7 explains the continuous-improvement cycle of
achievements as follows: the developed brake disc design system. We have
focused on a brake disc and standardized the disc
1. The automated meshing with hexahedron elements configurations, FE modeling, and postprocessing
and boundary conditioning assures the through our accumulated experience with many CAE
reproducibility of the meshed model and the analyses and experimental results. Furthermore, we
accuracy of analysis results. We can compare all the have automated and speeded up each CAE process,
simulated results evenly with disregard to such as modeling, eigenmode identification, and final
differences among designers. evaluation of thermal-strength analysis. Figure 7 also
2. In eigenmode analysis, we can obtain a reliable reveals the following four key factors that are essential in
estimation of high-frequency mode, up to 15 kHz, by successfully establishing this system.
using a many-degree-of-freedom disc model of
50,000 hexahedron elements within reasonable x Focus on and expertise in brake disc design
processing time. x Simplicity of brake disc configuration
3. Knowledge of the effects of disc configurations, x Gradual system construction
dimensions, and material properties on resonant x Continuous improvement
frequencies and thermal deflections can be obtained
through the accumulation of vast numbers of FEAs Some of them are explained in detail in the following
and tests for validation. This makes it easier to sections.
deliberate on design alternatives. In addition, the
design system with decision support helps in
SIMPLICITY OF BRAKE DISC CONFIGURATION
selecting an optimum solution from ten design
alternatives within two days.
We defined multiple standard configurations of brake
4. An optimum solution for meeting tradeoff
discs. However, a brake disc basically consists of simple
requirements, such as squeal, judder, thermal
bodies such as a cylinder, discs and ventilation
strength, and weight reduction, can be determined at
passages. A circularly symmetric brake disc model
the early stage of the development by collating
therefore can be easily divided into simple elements.
reports of simultaneous analyses on eigenvalues
This simplicity of the disc shape makes it easier to
and thermal stress and deflection.
construct various disc configurations with a combination
of these simple elements and apply an automated
The developed design system has provided many discs
meshing algorithm systematically. Due to this simplicity,
that satisfy various requirements of most of Japanese
the coordinates of nodes in a meshed model are used
OEMs, some of them directly and others via Tier 1s.
for identifying specific nodes and elements in the
preprocess and postprocess tools.

Fig. 7 Kaizen cycles for design quality and productivity


Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

GRADUAL SYSTEM CONSTRUCTION requirements for discs and drums in a brake system as
well as in a full vehicle under various operational
Node identification by using the coordinates described conditions. We have an integrated process for discs and
above generalizes the data transfer between CAE drums from research and development to manufacturing.
processes. This generality has exempted us from We can therefore exceed OEM or Tier 1 in terms of the
applying a complicated algorithm for numbering a great depth of our expertise focusing on brake discs and
many nodes and elements. The sequence of the CAE drums.
operational processes does not always agree with their
priority of the effectiveness of productivity improvement Through various opportunities like the SAE Annual
in each process. However, our node-identification Brake Colloquium, the brake community shares the
system makes it possible to develop system tools one accumulated knowledge on the correlation of brake
after another with disregard to the sequence of the CAE system behaviors between real-field, proving-ground,
operational processes. As a matter of fact, we chassis-dynamometer, brake-dynamometer, and test-
constructed the postprocessing tool for eigenvalue bench results. Based on this shared knowledge and
analysis first and the preprocessing tool for parametric individual information for each targeted vehicle project
model creation later. provided by OEM or Tier 1, the reliability of our
developed products is assured by CAE-simulated results,
Furthermore, we can apply old and new design systems, which are validated by test results on a brake
namely, our developed system and a commercial dynamometer and a test bench, as shown in Figure 9.
process, alternatively. For example, we can design a
new disc through the developed processes exclusively,
or we can create a meshed model with a commercial
FEA tool and then evaluate the simulation results with
the developed system.

MAKING BEST USE OF CONVENTIONAL CAE

LIMITED SCOPE BUT PROFOUND EXPERTISE

Original equipment manufacturers (OEMs), Tier 1, and,


Tier 2 of the supply chain have different roles and
scopes, as shown in Figure 8.

Fig. 9 Correlation between CAE & experiment

CONVENTIONAL CAE MIGHT EXCEED STATE-OF-


THE-ART CAE

The developed design system has helped us to obtain


Fig. 8 Hierarchy of supply chain and correlation between the knowledge of the effects of disc configurations,
test results by facilities dimensions, and material properties on resonant
frequencies and thermal deflections, through the
A. Jerhamre et al. [8] studied the cooling efficiency of accumulation of vast numbers of results of FE analyses
ventilated brake discs through complete vehicle and validation tests. Based on our accumulated CAE
simulations and measurements. We cannot exceed experience, original eigenmode-identification algorithm
OEM or a system supplier in terms of the range of and thermal-stress evaluation indexes were developed
activities and knowledge like that, since we design and and loaded into the postprocess tools. We have made
supply only discs and drums for a brake system and our the best of classical FEA methodology by allotting all of
simulations and measurements are mainly based on the degrees of freedom, which are computationally
dynamometer testing. However, we understand the available, only on a brake disc model. Furthermore, we
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

can obtain more accurate CAE results for our discs, properties that were determined for each gray-cast-iron
because we use the material properties of the castings grade but not modified them in each analysis on different
that we cast mainly in our in-house foundry and measure discs.
some of the properties by ourselves. Consequently, we
can provide the highest design quality at the early stage Table 4 Material and foundry of discs for Figs. 10 and 11
of new vehicle development without need for a state-of-
the-art complex eigenvalue analysis [9-11] or nonlinear Material FC190 JIS FC250 SAE G3000 Total
analysis [1]. JP 16 9
Foundry 27
US 2
EIGENVALUE ANALYSIS

T. S. Shi et al. [10] provided a comparison among real


modal analysis, transient analysis, and complex
eigenvalue analysis. Table 3 compares conventional
eigenvalue analysis and the state-of-the-art complex
eigenvalue analysis. The complex analysis can analyze
the behavior of multibodies and determine the stability of
a brake system under operational conditions. On the
other hand, we focus on the vibrational behavior of a
brake disc, and we can obtain an optimum solution
without complex eigenvalue analysis. By processing a
lot of FEAs on a disc design with the developed design
system, we assure the simulation results with practical
accuracy and speed.

Figures 10 and 11 are our typical eigenvalue analysis


results compared with experimental resonant
frequencies, which were mean values of 5 to 10 samples
for each disc design. Most of the discs are front
ventilated and the others are rear drum-in discs. Their
outer diameters ranged from 280 to 350 mm. Table 4
lists the casting material and the location of foundries, Fig. 10 Resonant frequency correlation between CAE
where "JP" indicates in-house foundries in Japan and and experiment for nodal-diameter bending mode
"US" indicates outsourcing to a certain foundry in the
USA. The lines in Figures 10 and 11 indicate plus or
minus 3% zone. These figures confirm that conventional
real-eigenvalue analysis provides excellent accuracy of
calculated frequencies. We have applied the material

Table 3 Comparison between conventional (real) and complex eigenvalue analysis

Conventional eigenvalue by our system Complex eigenvalue


Simulated condition Disc without neighboring components under Multibodies of caliper, pad, disc, etc.
free-free constraint without damping including contact element and damping
Calculus Real number Complex (real and imaginary) number
Stability Coupling Frequency interval between in-plane and Real part of eigenvalue in addition to
determinant out-of-plane modes frequency intervals
Resonance Frequency interval between disc and other
neighboring components at several modes
FEA Degree of 50,000 hexahedron elements only for full For example in reference [11], 26,000
accuracy for freedom (DOF) disc model elements for all bodies including caliper
disc and pads
Limitation Quite accurate in nodal-diameter-bending Less accurate in calculated
and in-plane modes, but less correlated with eigenfrequency due to limited allocation
experiments in hat-related mode for of DOF to disc model
assembled discs
Verification Experimental modal analysis
CAE Model creation A few minutes Several days or weeks due to difficulty
lead-time Processing Less than 2 hours from scratch through FEA in automated modeling and meshing
time result report and larger DOF for the same number of
elements as conventional analysis
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

an alternative determinant for thermal resistance of a


brake disc.

Figure 12 plots the correlation between simulated


stress/strength ratio and experimental number of brake
applications when more than one crack went through
either the inboard or outboard friction plates in a high-
speed braking test. We use a constant thermal-
distribution ratio between the disc and pad with
disregard to lining materials, and determine the thermal
strength of a new disc by collating the simulated
stress/strength ratio with specified criteria for each
friction material. The upper and lower lines for each
friction material indicate the plus or minus 10% zone.
This figure, including the results of 25 ventilated front
discs made of three gray-cast-iron grades, confirms that
linear thermal analysis can be used as a criterion for
estimating the thermal strength of a simulated disc.

Fig. 11 Resonant frequency correlation between CAE


and experiment for in-plane mode

BASIC STRENGTH

The stress-strain curve for steel is typically nonlinear.


However, gray cast iron is a fragile material and has an
almost linear stress-strain curve. Its tensile strength,
yield point, elastic limit, and proportional limit are
practically the same at room temperature. Yielding or
plastic deformation of a brake disc therefore need not be
taken into consideration below 200°C.

THERMAL ANALYSIS

In the thermal analysis, we have to deal with nonlinear


properties and behaviors. There are a variety of
temperature-dependent nonlinear properties as follows: Fig. 12 Correlation between simulated stress/strength
ratio and experimental number of brake applications
x Tensile strength
LIMITATIONS AND FUTURE STUDY
x Young's modulus
x Specific heat capacity
Focusing on a brake disc, we have simulated and
x Thermal-expansion coefficient
verified its behavior as a single-body FEA model. We
x Thermal-conductivity coefficient have obtained CAE simulation results with practical
accuracy through the approximation of nonlinear
In preliminary simulations, however, we can approximate properties and contact conditions with neighboring
the nonlinear properties above to be linear or constant. components. However, a brake disc is assembled and
operated in a vehicle. Furthermore, the hardware and
Tensile strength nonlinearly depends on temperature. software for CAE are making rapid progress. For the
With repeated brake applications, the maximum study of brake disc cooling in particular, sophisticated
temperature of the disc friction surface becomes very simulation with computational fluid dynamics (CFD) has
high, and even cast iron behaves as a plastic material. J. become practical by OEMs [8, 13-15]. We will employ
Brecht et al. [7] revealed the relationships among these advanced tools in consideration of the speed and
temperature at the friction surface, compressive stress, economy of a new product development by gradually
material yielding, residual tensile stress, and crack widening our limited scope as a component supplier.
initiation. Y. Jimbo et al. [12] also explained the same
mechanism that compressive plastic deformation during Benchmarking is important in designing new products. T.
braking causes thermal crack formation. We therefore Chatterley et al. [16], studied trends in the configurations
calculate tangential stress at the disc friction surface and materials of European brake discs. One of the
with linear analysis, and adopt a maximum compressive authors also described the trend in geometric accuracy
stress calculated under high-temperature conditions as
Downloaded from SAE International by Stony Brook Univ, Sunday, August 12, 2018

according to benchmarked data of brake discs from Technical Report, No. 21 (2003), pp.180-185 (in
Japan, the USA, and Europe [17]. We are continuously Japanese)
accumulating benchmarking data and studying the 5. NIKKEI DIGITAL ENGINEERING, No. 60
following trends: (December 2002), pp.96-107 (in Japanese)
6. A. Fukano et al., Development of Disc-Brake Design
x Basic configurations such as inner or ordinary hat, Method Using Computer Simulation of Heat
and straight or spiral vanes; Phenomena, SAE Paper 860634
x Disc size and mass versus vehicle specifications 7. J. Brecht et al., Influence of Material Selection on
such as maximum speed and gross vehicle weight. Stresses in Ventilated Brake Discs, SAE Paper
980595
We are taking these trends into consideration and 8. A. Jerhamre et al., Numerical Study of Brake Disc
incorporating new configurations into our system. As a Cooling Accounting for Both Aerodynamic Drag
result, we will continue providing brake discs with Force and Cooling Efficiency, SAE Paper 2001-01-
competitive performance in a short development lead- 0948
time by utilizing the new design system described here. 9. H. Ouyang et al., On Automotive Disc Brake Squeal
- Part II: Simulation and Analysis, SAE Paper 2003-
CONCLUSION 01-0684
10. T. S. Shi et al., Advances in Complex Eigenvalue
We have constructed an original brake disc design Analysis for Brake Noise, SAE Paper 2001-01-1603
system, by standardizing, automating, and speeding up 11. A. Bajer et al., The Influence of Friction-Induced
each CAE process. The main achievements of this
Damping and Nonlinear Effects on Brake Squeal
system are summarized as follows.
Analysis, SAE Paper 2004-01-2794
12. Y. Jimbo et al., Development of High Thermal
1. An optimum and robust design can be selected from
Conductivity Cast Iron for Brake Disk Rotors, SAE
many alternatives quickly and easily.
Paper 900002
2. The lead time to find an optimum solution is reduced
13. S. Sakamoto et al., Development of High Cooling
to two days, and the design quality is improved at
Performance A New Brake Rotor, TOYOTA
the early stage of development, through continuous
Technical Review, Vol. 44, No. 1 (May 1994), pp.90-
improvement of CAE processes.
95 (in Japanese)
3. Knowledge of the effects of disc configurations,
14. T. Shimazu et al., Development of A New Brake
dimensions, and material properties on resonant
Rotor with Improved Cooling Capacity, JSAE
frequencies and thermal deflections can be obtained
Review, Vol. 25, No. 3 (July 1994), pp.128-133 (in
through the accumulation of vast numbers of FEAs.
Japanese)
4. A design methodology was established to satisfy the
15. C. Repmann, Investigation of the Thermal Behaviour
requirements for a brake system, such as squeal,
of Vented Disc Brakes by Means of CFD, Advanced
judder, and thermal strength, by improving the
Brake Technology, 2003, pp.51-75
simulation accuracy of eigenvalue and thermal
16. T. Chatterley et al., Cast Iron Brake Discs - Current
analyses with conventional CAE methodology.
Position, Performance and Future Trends in Europe,
SAE Paper 1999-01-0141
We will continuously improve this system by automating
17. T. Okamura et al., High-Precision Brake Discs to
the final decision making, applying sophisticated CAE
Reduce Judder, SAE Paper 2005-01-3924
methodology like nonlinear analysis, and constructing a
design system for brake drums.
CONTACT
REFERENCES
Hiroyuki YUMOTO, KIRIU Corporation
2, Omata-Minami, Ashikaga-shi, Japan 326-0142
1. T. Valvano et al., An Analytical Method to Predict
E-mail: h-yumoto@kiriu.co.jp
Thermal Distortion of a Brake Rotor, SAE Paper
URL http://www.kiriu.co.jp/
2000-01-0445
http://www.kiriu-usa.com/
2. Y. Takahashi, et al., Establishment of an Intake Port
Design Support Tool Using CFD, Honda R&D
Technical Review, Vol. 15, No. 1 (April 2003),
pp.103-108 (in Japanese)
3. T. Uchida, et al., Practical Examples and Future
Directions for CAE-Based Development for
Automobile Transmissions by Designers, Honda
R&D Technical Review, Vol. 15, No. 1 (April 2003),
pp.109-114 (in Japanese)
4. H. Yokohata et al., Development of "PT-CS": CFD
Analysis System for Design Engineers, Mazda

You might also like