sensitivity of brake squealing concerning scattered …...sensitivity of brake squealing concerning...

Sensitivity of brake squealing concerning scattered component,

joint and bearing properties

Dr. Ronaldo F. Nunes, Daimler AG, Mercedes-Benz Cars and M.Sc. Christian Büttner (Altair Engineering GmbH)

22.11.2013

Page 2

Agenda

1 Simulation, Bench and Vehicle

2 NVH - Target

3 NVH – Squealing Joints Influence

4 Summary

Dr. Nunes, Büttner - WOST

Page 3

Introduction

Introduction

Dr. Nunes, Büttner - WOST

Page 4

Vehicle

Simulation

Bench

� Quality Control regarding Brake disc Eigenfrequency, lining material, etc.

Simulation, Bench and Vehicle(Interdisciplinary Strategy: Cross Correlation Simulation, Bench and Vehicle)

Source: EuroBrake 2012Dr. Nunes, Büttner - WOST

Page 5

Notice: Factor 1 means 0.8 to 1.5 mm tolerance

Operating Points

Material

(Lininig material,

Disc, Caliper, Axle,

Shim and so on)

Geometry

Uncertain Parameters: E-Modulus, Material Density and Geometry (for example Knuckle)

Dr. Nunes, Büttner - WOST

Page 6

Robustness Analysis Combined with Optimization Strategy (120 Designs – Complex Eigenvalue Analysis)

Baseline SystemNVH –Opt.

(Current Package)

NVH - Target : Focus to find out the best

solution taking into account the

influence of uncertain parameters

Dr. Nunes, Büttner - WOST

Page 7

NVH – Squealing Joints Influence

Dr. Nunes, Büttner - WOST

Page 8Dr. Nunes, Büttner - WOST

Backingplate

Lower ArmSteering Rod

Wishbone

Hub Unit

Pistons

Friction Material

Anchor (Caliper)

Wheel Bearing

Caliper

Tension Strut

Brake Disc

Knuckle

CAD Model – Car Brake and Suspension

Page 9

Mathematical Equations

� �M λ + D λ + C] {ω} = 0

� λ = ∝ ��ω

α = Real Part

� = Imaginary Part (Squeal Frequency)

� � � ∗ �� � � ∗ � ∝���

� �� ��� ∗ ���� � ��∗ ������

Stable / Unstable – Evaluation of complex Modes

� α < 0 System is stable

� α > 0 System not stable

Dr. Nunes, Büttner - WOST

Positive α indicates a system behavior with an increasing amplitude

of the brake oscillation

Background: Brake Squeal Analysis

Complex Eigenvalue

Analysis (CEA)

Page 10

• Without Friction: Independent Eigenmodes

• Convergence of Eigenfrequencies with increasing µ

� Mode Coupling at µcrit

� Unstable System (Squealing Propensity)

Dr. Nunes, Büttner - WOST

Background: Complex Eigenvalue Analysis

Independent

Eigenmodes

µcrit

Coupling

µcrit

Page 11

� Bearings, Ball Joints and Kinematic

Description

Joints Modeling

rigid

Kinematic CouplingRubber

Stiff Spring

Kinematic Coupl.

Dr. Nunes, Büttner - WOST

Page 12

Bench Test – Noise Problem

Bench• Noise problem identified

• Deterministic FE simulation was not

able to

find the critical frequency !!

� Sensitivity Analysis

Ball Joints, Bearings stiffness

Pad and Disc stiffness/Geometry

Brake PressureS

ou

nd

Pre

ssu

reL

eve

l

Frequency

Dr. Nunes, Büttner - WOST

Page 13

� Complexe Eigenvalue Analysis investigates Stationary System

� Frequency, amplitude and temperature depending behavior not considered

in single Calculation

Dr. Nunes, Büttner - WOST

Material Behavior in CEA

Frequency

Dyn

am

ic S

tiff

ne

ss

Page 14Dr. Nunes, Büttner - WOST

Workflow: Sensitivity Analysis

FEM Process

Stable?

� Calculation

Variation of Material

and Geometry

Post Processing

Statistical Evaluation of the

Analysis

Page 15

� Dynamic Measurements Available: Ball joints and Bearings

� Sensitivity Analysis was adopted in order to identified the influence of

Joint stiffness to the noise problem

Instability

Influence

Sensitivity Analysis: Joints Influence

Source: Christian BüttnerDr. Nunes, Büttner - WOST

Page 16

Bench and Sensitivity Analysis Validation

� Frequency identified

� Additional optimization to fix the problem Bench

FE

Sensitivity

Analysis

So

un

d P

ressu

reL

eve

l

Frequency

No

ise

In

dex

Bench

Dr. Nunes, Büttner - WOST

Page 17

Model Improvements

Dr. Nunes, Büttner - WOST

� Better Identification of main influences and

local minimum / maximum

� Better prediction of system behavior

Page 18

� Improvements in Modelling Brake System

� Better Identification of main influences

� Better prediction of system behavior

� Results fitting better with bench vehicles

� Using Sensitivity Analysis

� Identification of ideal parameter range

� Joint parameters haven‘t a strong influence

Conclusion

rigid

Kinematic CouplingRubber

Stiff Spring

Kinematic Coupling

Instability

Influence

Dr. Nunes, Büttner - WOST

Thank you for your attention !