Component Based TPA Webinar

Component-based transfer path analysis
Your future-proof strategy for advancing NVH vehicle

development
Unrestricted © Siemens 2020 Siemens Digital Industries Software
Automotive OEMs have to reduce full vehicle testing to
handle wide variety of vehicles
IC
# of vehicle variants Virtual Vehicle Prototyping
PHEV
EV
• Increasing testing effort

• Prototype availability?
• Impact of modification?
Powertrain Body Component • …
Front-loading vehicle
How to ensure NVH performance while keeping development level component
time and cost under control? NVH testing
Unrestricted © Siemens 2020
Page 2 Siemens Digital Industries Software
Increasing complexity continuously challenges automotive
OEM’s and Suppliers
Typical OEM-Supplier Improved OEM-Supplier

cooperation cooperation
No correct Independent load

excitation data description
AOEM Increasing risk of integration

AOEM
issues of components
Implication Control NVH

Performance
Impact on time-to-market and possible

No detailed NVH Realistic NVH
design targets expensive modifications design targets
Supplier Supplier

How to keep control of the NVH Performance
at any stage of the development cycle?
Can we provide a method that

addresses all these challenges?
YES, WE CAN!
Component-based TPA
Classical Transfer Path Analysis
Source-transfer-receiver approach
Contribution at path or group (case vs. rpm or frequency) A
Engine 60.00
dB(A)
HVAC
Pa
10.00 sh_r:32:Z
Total
Measured
be_r:9:X
be_r:5009:X
sh_r:32:X
sh_r:5032:X
be_r:9:Z
be_r:5009:Z
sh_r:32:Z
sh_r:5032:Z
be_f:18:X
be_f:5018:X
be_r:9:Y
sh_f:30:X
subf:20:X
be_r:5009:Y
sh_f:5030:X
sh_r:32:Y
sh_r:5032:Y
subf:5020:X
be_f:18:Z
be_f:5018:Z
sh_f:30:Z
subf:20:Z
sh_f:5030:Z
subf:5020:Z
be_f:18:Y
be_f:5018:Y
sh_f:30:Y
subf:20:Y
sh_f:5030:Y
subf:5020:Y
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
50.00
RMS Sum
Spectrum: PRCM:0001:S
dB(A)
E-motor
Pa
Wiper-motor Spectrum: PRCM:0003:S
a Spectrum: PRCM:0006:S
-40.00
0.00 Hz 350.00
Transmission Tyres
EPS
Source (Fi,Qj) X Transfer (NTF) = Receiver (yk)

Source-Receiver interaction
Exchange of contact forces
Supplier AOEM
Predict Target Response 20

Harmonic Spectrum P3:T1:+Z Measured
Harmonic Spectrum P3:T1:+Z<Total Contact bench
10
Source` Source
Test bench 0
-10
CONTACT
dB
FORCES -20
g
-30
Test Bench Vehicle / Receiver -40
-50
-60
180.00
Phase
°
-180.00
Predicted target 20 50 100 150 200 250 300 350 400 450 500
Hz
deviates from
measured target!
Source-Receiver interaction
Exchange of Independent loads
Supplier AOEM
Predict Target Response

30
Harmonic Spectrum P3:T1:+Z Measured
20 Harmonic Spectrum P3:T1:+Z<Total Blocked forces
Source` Test Bench Source 10
INDEPENDENT 0
LOADS -10
dB
g
-20
-30
Test Bench Vehicle / Receiver -40
-50
-60
180.00
Phase
°
-180.00
20 50 100 150 200 250 300 350 400 450 500
Predicted target matches Hz
measured target!

Component Based Transfer Path Analysis
Predict full system level performance
Contribution analysis
Contribution at path or group (case vs. rpm or frequency) A
60.00
dB(A)
Pa
10.00 sh_r:32:Z
Total
Measured
be_r:9:X
be_r:5009:X
sh_r:32:X
sh_r:5032:X
be_r:9:Z
be_r:5009:Z
sh_r:32:Z
sh_r:5032:Z
be_f:18:X
be_f:5018:X
be_r:9:Y
be_r:5009:Y
sh_f:30:X
sh_f:5030:X
sh_r:32:Y
subf:20:X
sh_r:5032:Y
subf:5020:X
be_f:18:Z
be_f:5018:Z
sh_f:30:Z
sh_f:5030:Z
subf:20:Z
subf:5020:Z
be_f:18:Y
be_f:5018:Y
sh_f:30:Y
sh_f:5030:Y
subf:20:Y
subf:5020:Y
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
..
50.00
RMS Sum
dB(A)
Pa
-40.00
0.00 Hz 350.00
Pass-by Noise
75.00 75.00
F Left Side
F Left Side Virtual
B Right Side
B Right Side Virtual
dB(A)
dB(A)
Pa
Pa
65.00 65.00
-7.00 m 10.00
Sound synthesis
Step 1: Independent X Step2: Virtual Assembly = Step3:

loads (Fbl,Q) (FRF, K) Prediction(yk)
Component Based TPA
Independent Measurement Application Full Vehicle NVH

Source challenges Examples Synthesis
characterization
Theory of method From theory to Deal with the subsystem Combining all systems
measurements specific challenges & predict full vehicle
Blocked Forces as
Independent source Extensive mathematics Difference in Test benches NVH performance
description require accurate data & methodologies

Component Based TPA

characterization
Blocked Forces as

Component-based TPA
Independent load characterization
Structure-borne:
Blocked Forces
Free Velocities
Airborne:
Volume Velocities

Component-based TPA
Step1: Source characterization - Structure Borne
Three possible methodologies to obtain independent source description
1. Blocked Force 2. Free Velocity 3. In-Situ TPA
Source A Source A Source A Receiver B
4 5
𝑭2,𝐵𝑙𝑜𝑐𝑘𝑒𝑑 𝒂𝟐,𝑭𝒓𝒆𝒆 𝐴𝐵
F source F source F source 𝑯24
2 2 2 3
1 1 1 4
𝐴 −𝟏 𝒂 𝐴𝐵 +
𝑯22 𝟐,𝑭𝒓𝒆𝒆 = 𝑭2,𝐵𝑙𝑜𝑐𝑘𝑒𝑑 𝑭2,𝐵𝑙𝑜𝑐𝑘𝑒𝑑 = 𝑯24 𝒂𝟒
Rigid test rig → Most times not possible Source in free free conditions ISO 9611 Any receiver is valid ISO 20270
Source: Mondot, Petersson, Characterization of structure-borne

Source:Elliott, Moorhouse, Characterization of the structure borne
sound sources: The source descriptor and the coupling function
sound sources from measurements in-situ, 2008
1987
Contact force vs. Blocked force
Different measurement setup for local FRF
Classic TPA Blocked Forces TPA

Determination contact forces Determination blocked forces
Receiver B Source A Receiver B
4 5 4 5
1. Local FRF for 23 𝐴𝐵
𝑯24
3 𝑯𝐵34 1
load identification
4 4
Source A Receiver B Source A Receiver B

4 5 4 5
2. Operational 1 1
23 23
measurements F source F source
4 4
3. Force 𝐴𝐵 +
𝑭2,𝐵𝑙𝑜𝑐𝑘𝑒𝑑 = 𝑯24 𝒂𝟒
𝑭3,𝑪𝒐𝒏𝒕𝒂𝒄𝒕 = 𝑯𝐵34 + 𝒂𝟒
Identification

Structure borne application example
Steering system integration
Invariant
Source Mechanism Sub-Receiver Receiver
Source Synth. Model
Blocked
Steering Forces & Subframe Body
FEM/TEST FEM/TEST
System Impedances
FRF FRF
Mount Pos.
STEP 1: Source Characterization STEP 2: Full Vehicle Prediction
A 𝑝
𝐴 𝐴𝐶 −1 3
𝐹𝑏𝑙 = 𝐻𝑖1 ∗ 𝑎𝑖 2
1
𝐴 𝐹𝑟
1 𝐹𝑏𝑙 𝐴
𝐴𝐶 𝐹𝑏𝑙 A
𝐻𝑖1 Measured Coupled FRF B
x x
x bench 𝑎𝑖 𝐴𝐵
𝑝 = 𝐻32 ∗ 𝐹𝑟𝐴𝐵
In-Situ TPA© –Siemens
1. Unrestricted Apply Matrix
2020 Inversion: Blocked Forces = source characterization
Component-based TPA
Step2: Target prediction using Frequency based sub-structuring (FBS)
Source A Source A Free-Free or

1. Source A: pre-load
𝐴 2 2
a. Blocked Force 𝑭2,𝑏𝑙 1 𝑯𝑨22 condition
F source
b. Inertance matrix 𝑯𝑨22 𝑭𝑨𝟐,𝒃𝒍
Receiver B Realistic
2. Receiver B: 5 boundary
a. Inertance matrix 𝑯𝐵33 3 𝑯𝐵33 condition
b. Noise Transfer Function 𝑯𝑩

𝟓𝟑
𝐴
3. Target prediction using blocked forces 𝑭2,𝑏𝑙 Source A Receiver B
a. Mount stiffness 𝑲 (not needed for rigid connections) 𝑲 5
2 3 𝑭𝑨𝑩
b. FBS 1 𝟑,𝒄
F source
−𝟏 −𝟏
𝑭𝑨𝟐,𝒃𝒍
𝑷𝑨𝑩
𝟓 = 𝑯𝑩 𝑨 𝑩
𝟓𝟑 ∗ 𝑯𝟐𝟐 + 𝑯𝟑𝟑 + 𝑲 ∗ 𝑯𝑨𝟐𝟐 ∗ 𝑭𝑨𝟐,𝒃𝒍
Structure borne application example
Steering system integration
Invariant
Source Mechanism Sub-Receiver Receiver
Source Synth. Model
Blocked
Steering Forces & Subframe Body
FEM/TEST FEM/TEST
System Impedances
FRF FRF
Mount Pos.
STEP 1: Source Characterization STEP 2: Full Vehicle Prediction

𝑝
A 2 3
1
𝐴 𝐴𝐶 −1
𝐹𝑏𝑙 = 𝐻𝑖1 ∗ 𝑎𝑖
𝐹𝑟
1 𝐴
𝐹𝑏𝑙 A
𝐴 𝐴
𝐴𝐶 𝐹𝑏𝑙 , 𝐻11 Calculate Coupled FRF B
𝐻𝑖1
𝐴 𝐵 𝐴 𝐴
𝐹𝑟𝐴𝐵 = 𝐻11 + 𝐻22 + 𝐾 −1 −1 ∗ 𝐻11 ∗ 𝐹𝑏𝑙
x x
x bench 𝑎𝑖 𝐵
𝑝 = 𝐻32 ∗ 𝐹𝑟𝐴𝐵
In-Situ TPA© –Siemens
1. Unrestricted Apply Matrix
2020 Inversion: Blocked Forces = source characterization 2. Component Based TPA – Apply FBS Substructuring for load prediction
Component Based TPA

characterization
Blocked Forces as

Component Based TPA: required tools
‘extensive mathematics require accurate data’
Accuracy Validation Completeness
• hard to reach location • reciprocity / linearity • FRF required at exact connection

• angle & position accuracy • directions errors / consistency center – assuming local rigidity
• repeatability / low noise / high power • driving point behavior • 6DOF FRF description of the
• large frequency range from 5Hz -10kHz • repeatability connection is required for completeness
Dedicated FRF shaker excitation - Use the Matrix heatmap to instantly verify Reduce the FRF’s to the connection center
engineered for TPA measurement large datasets using Virtual point transformation
Virtual Point Transformation (VPT)
Accurate FRFs at interface connection points
Challenge: VPT for correct blocked force estimation

80.00
Blocked force
▪ Transfer functions at difficult or in most even at connection
impossible to access positions Blocked force
dB
N
off connection
▪ High quality transfer functions at precise locations.
▪ Translational and rotational transfer functions (DOFs) -80.00
180.00
Phase
°
-180.00
0.00 Hz 1000.00
Solution:
VPT for correct assembly using FBS
▪ Geometrical Reduction / Virtual Point
Transformation VP
▪ Assumption: local rigidity in the connection
▪ Input: Geometry Information and FRFs

Virtual Point Transformation (VPT)
Application examples
TRANSMISSION T/M BRACKET
Transmission
g/N
dB
VPT for Measured
FBS
improving FBS 180

90
Phase
0
°
-90
-180
50.00 500.00 1000.00 1500.00 2000.00
Hz
Source:” A Frequency-Based Substructuring approach to engine mount bracket performance assessment for gear noise”, ISMA Siemens,Toyota Motor Europe 2020
TIRE/WHEEL SUSPENSION
Road Noise
VPT for spindle
forces/moments
and FBS
Source:” Structure-Borne Prediction on a Tire-Suspension Assembly Using Experimental Invariant Spindle Forces”, SAE 2019 DOI: 10.4271/2019-01-1541
COMPRESSOR CAR BODY
HVAC
Force reduction
for improved
prediction

Component Based TPA

characterization
Blocked Forces as

Hitachi AMS – Steering systems
Estimating in-vehicle noise of EPS using test bench
Development Component Based TPA methodology for NVH assessment
Test bench condition Vehicle condition
Blocked Force comparison In-vehicle Noise prediction

• More realistic design targets for 20
0
F Vehicle
20
0
F Vehicle
20
0
F Vehicle
60
55 F Measured : In-Vehicle
60
55 F Measured : In-Vehicle
F Bench F Bench F Bench F Noise Prediction : with Bench Blocked Force F Noise Prediction : with Bench Blocked Force
components dB
-10
-20
-30
-10
-20
-30
-10
-20
-30
50
45
50
45
dB
dB
N
N
-40 -40 -40 40 40
-50 -50 -50 35 35
-60 -60 -60
30 30
• Appropriate evaluation method at

-80 -80 -80 25 25
100 400 600 800 1000 1200 1400 1600 2000 100 400 600 800 1000 1200 1400 1600 2000 100 400 600 800 1000 1200 1400 1600 2000
Hz Hz Hz 20 20
dB(A)
dB(A)
Frequency Frequency Frequency
Pa
15
Pa
15
10 10
20 20 20
component prototype phase F Vehicle F Vehicle F Vehicle 5 5

0 F Bench 0 F Bench 0 F Bench 0 0
-10 -10 -10
-20 -20 -20 -5 -5
dB
dB
dB
-30 -30 -30
N
N
-10 -10
-40 -40 -40
-50 -50 -50 -15 -15
• Predict in-vehicle noise/vibration

-60 -60 -60 -20 -20
-80 -80 -80 -25 -25
100 400 600 800 1000 1200 1400 1600 2000 100 400 600 800 1000 1200 1400 1600 2000 100 400 600 800 1000 1200 1400 1600 2000 -30 -30
Hz Hz Hz 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 2000 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 2000
Frequency Frequency Frequency Hz Hz
Frequency Frequency
using component based TPA

method ▪ Test bench identified blocked forces correlate with vehicle blocked forces
▪ Predicted cabin noise using test bench blocked forces well up to 1500Hz
Aoi Nakanome Hitachi AMS, 2019 Simcenter Conference Amsterdam

Automotive OEM – Wiper System
Derive full vehicle structure-borne noise prediction from test rig
Identification of blocked forces Full vehicle assembly from Prediction of target

on Mockup component level testing response
Mount characterization Transferability + FBS prediction
Mount stiffness
dB
g
Measured
N/m
dB
Predicted
Measured 1
Measured 2
C1/C1 180
C1/C2
Phase
C2/C1
0
°
180 C2/C2
-180
Phase
0
°
20.00 100.00 200.00 300.00 400.00 500.00 600.00
-180 Hz
20.00 500.00 600.00
Hz
Substructuring
Wiper + Vehicle FRF → Performance evaluation
→ using ISO/CD 21955 for prediction
→ Identification of path contribution
g/N
dB
→ Input for target setting
→ Source in operation: wiper system at 35 cpm.
Measured FRF Target / Connection 2
FBS FRF Target / Connection 2
→ What if analyses:
→ Blocked forces calculated using in-situ TPA 180
Phase
0
- Sensitivity to mount stiffness
°
on dedicated test rig – matrix inversion. -180

20.00 100.00 200.00 300.00 400.00 500.00 600.00
Hz - Receiver modification.
-…
Ortega, Bianciardi, Corbeels, Ullmann, Desmet, MOUNT CHARACTERIZATION ANALYSIS IN THE CONTEXT OF FBS
FOR COMPONENT-BASED TPA ON A WIPER SYSTEM, , FA 2020, Lyon, Siemens, KULeuven, BMW
Automotive OEM – Road Noise
Wheel center blocked forces
Identification of blocked forces Full vehicle assembly from Prediction of target response
Vertical (Z)
component level testing
dB(A)
dB(A)
Pa
Pa
dB
N
▬ Measured Autopower
Toe (Mz) ▬ Comp-TPA: FBS assembly
50 Hz
Hz 250
Hz Total
Nm
dB
dB(A)
Pa
Hz
Hz
→ Invariant forces: receiver → Independent characterization of → Performance evaluation
independent source and receiver → Identification of path contribution
→ Transferable between receiving → Virtual vehicle assembly → Input for target setting
structures

Automotive Supplier – Tire manufacturer
Road noise – Wheel center blocked forces
On board validation of target response on

Identification of blocked forces on test rig
test rig
dB
X
N
Hz
Blocked Forces and Moments
dB
g
dB
Y
N
Measured
Hz Predicted
Hz
dB
Z
N
Hz
AutoPower RIG:TARGET_1:+Z
Nm
AutoPower RIG:TARGET_1:+Z<RIG:TARGET_1:Z
MX
dB
AutoPower RIG:TARGET_1:+Z<TIRE:RIM_CR:RX
Hz
dB
g
AutoPower RIG:TARGET_1:+Z<TIRE:RIM_CR:RZ
AutoPower RIG:TARGET_1:+Z<TIRE:RIM_CR:X
Nm
MZ
dB
AutoPower RIG:TARGET_1:+Z<TIRE:RIM_CR:Y
Hz
AutoPower RIG:TARGET_1:+Z<TIRE:RIM_CR:Z
Hz
→ Source in operation: 20 / 40 / 60 / 80 / 100 kph.
→ On-board validation
→ Identification of path contribution
→ Blocked forces calculated using in-situ TPA: matrix inversion
→ Input for realistc target setting & prediction
using multiple integral shakers for FRF a
→ Independent
→ Blocked forces measured on rigid test rig using force cell (usable
only up to
Unrestricted 300 Hz)
© Siemens 2020
Component-based TPA
Electromotor blocked forces – Assembly vibration prediction
Identification of blocked System assembly from
Prediction of target response
forces component level testing
Order60, X-direction
FBS vs Measured ▬ Measured Order

Order80, Y-direction
▬ Comp-TPA: FBS assembly
▪ Invariant forces: receiver independent ▪ Independent characterization of ▪ Performance evaluation

▪ Transferable between receiving source and receiver ▪ Identification of path contribution
structures ▪ Virtual vehicle assembly ▪ Input for target setting
▪ High frequency up to 4.5kHz
Unrestricted © Siemens 2020 Bianciardi, Corbeels, Choukri, HIGH FREQUENCY SOURCE CHARACTERIZATION OF AN E-MOTOR USING
COMPONENT BASED TPA, SIA 2020, Siemens
Component Based TPA

characterization
Blocked Forces as

Component based TPA for full vehicle NVH assessment
Model Based Development for NVH – Concept
Combining
Digital Twin Test & Simulation Compare Contribution Analysis
Model data
Vehicle Component Scenarios

Architectures
Pass-by Noise Synthesis
Library 75.00 75.00
Methods F
F
B
B
Left Side
Left Side Virtual
Right Side
Right Side Virtual
Forces
dB(A)
dB(A)
Pa
Pa
FRF
65.00 65.00
-7.00 m 10.00
Sound Synthesis Result

Stiffness
NVH
Target synthesis NVH Driving Simulator Evaluation
Requirements
Solver

Simcenter Testlab NVH Synthesis
Model Based Development for NVH
Contribution
Define the Create Populate
Perform Analysis &
vehicle Validated Assemblies /
Calculation Audio Replay
architecture Components Configurations & Reporting
Supported by Model (Lifecycle) Data Management

How can Simcenter help you ?
Implementation roadmap for component-based TPA
Combining test
and simulation
End-to-end
Technology test system
transfer Simcenter Solution
• Single source provider for test
and simulation models that can
Testing Solutions be used in a NVH Synthesis
context
• Integrated solution for transfer path
Engineering and analysis with shakers, sensors, • Process implementation to predict
NVH performance during the
Consulting hardware and software
entire development process –
• The reference solution for airborne concept, detailed engineering,
• Worldwide technology and structureborne source
deployment & technology full-vehicle
characterization, substructuring and
transfer services NVH Synthesis.
• Dedicated approach for each • Embedded in test-based NVH
application engineering solution covering TPA,
structural dynamics, acoustics and
rotating machinery
Thank you
Unrestricted © Siemens 2020 Siemens Digital Industries Software

Component Based TPA Webinar

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Component Based TPA Webinar

Uploaded by

Copyright:

Available Formats

Component-based transfer path analysis

Your future-proof strategy for advancing NVH vehicle

• Increasing testing effort

Typical OEM-Supplier Improved OEM-Supplier

No correct Independent load

AOEM Increasing risk of integration

Implication Control NVH

Impact on time-to-market and possible

Unrestricted © Siemens 2020

Can we provide a method that

Contribution at path or group (case vs. rpm or frequency) A

Wiper-motor Spectrum: PRCM:0003:S

Source (Fi,Qj) X Transfer (NTF) = Receiver (yk)

Predict Target Response 20

Test Bench Vehicle / Receiver -40

Predict Target Response

Source` Test Bench Source 10

Predicted target matches Hz

Unrestricted © Siemens 2020

Step 1: Independent X Step2: Virtual Assembly = Step3:

Independent Measurement Application Full Vehicle NVH

Unrestricted © Siemens 2020

Independent Measurement Application Full Vehicle NVH

Unrestricted © Siemens 2020

Unrestricted © Siemens 2020

Three possible methodologies to obtain independent source description

1. Blocked Force 2. Free Velocity 3. In-Situ TPA

Source A Source A Source A Receiver B

Source: Mondot, Petersson, Characterization of structure-borne

Classic TPA Blocked Forces TPA

Source A Receiver B Source A Receiver B

Unrestricted © Siemens 2020

STEP 1: Source Characterization STEP 2: Full Vehicle Prediction

Source A Source A Free-Free or

a. Inertance matrix 𝑯𝐵33 3 𝑯𝐵33 condition

b. Noise Transfer Function 𝑯𝑩

STEP 1: Source Characterization STEP 2: Full Vehicle Prediction

Independent Measurement Application Full Vehicle NVH

Unrestricted © Siemens 2020

Accuracy Validation Completeness

• hard to reach location • reciprocity / linearity • FRF required at exact connection

Challenge: VPT for correct blocked force estimation

Unrestricted © Siemens 2020

improving FBS 180

COMPRESSOR CAR BODY

Unrestricted © Siemens 2020

Independent Measurement Application Full Vehicle NVH

Unrestricted © Siemens 2020

Development Component Based TPA methodology for NVH assessment

Test bench condition Vehicle condition

Blocked Force comparison In-vehicle Noise prediction

• Appropriate evaluation method at

component prototype phase F Vehicle F Vehicle F Vehicle 5 5

• Predict in-vehicle noise/vibration

using component based TPA

Aoi Nakanome Hitachi AMS, 2019 Simcenter Conference Amsterdam

Identification of blocked forces Full vehicle assembly from Prediction of target

on dedicated test rig – matrix inversion. -180

Unrestricted © Siemens 2020

On board validation of target response on

Blocked Forces and Moments

FBS vs Measured ▬ Measured Order