NVH CHALLENGES AND SOLUTIONS FOR MODERN AND ELECTRIFIED POWERTRAINS

NVH CHALLENGES AND SOLUTIONS FOR MODERN AND ELECTRIFIED POWERTRAINS AVL Vehicle and Powertrain NVH Stephan BRANDL AVL List GmbH (Headquarters)

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS CONTENT NVH for Conventional Powertrains Powertrain Development using Simulation and Measurement Engine Roughness Combustion Noise (Diesel Knocking) Turbocharger NVH Driveline NVH NVH in Electrification AVL s E-Driveline Development Approach NVH Frontloading for Power Electronics Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 2

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS All parts flexible COMBUSTION NOISE - ROUGHNESS Measured Cylinder Pressure Comprehensive Simulation Model Fully elastic; oil film bearings TVD Dynamic Stiffness & Damping EHD bearings Engine Mount Dynamic Stiffness Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 4

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS POWERTRAIN DEVELOPMENT - USING SIMULATION AND MEASUREMENT Powertrain NVH Assessment Installation on powertrain test bed Application of sensors Close to main bearing (excitation correlation) Engine surface (flexible part correlation) Data acquisition Data evaluation and comparison to simulation Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 5

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS COMBUSTION NOISE - ROUGHNESS Simulation Model Verification Sensor Position x, y, z Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 6

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS COMBUSTION NOISE - ROUGHNESS Root Cause Analysis Modal Contribution Analysis Transfer Path Analysis NVH Source Identification Cam cover Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 7

DIESEL COMBUSTION NOISE INTRODUCTION Why Do We Need Objective Sound Quality Criteria? Example: Overall Level vs. Sound Quality Criterion Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 9

Cylinder Pressure - bar Cylinder Pressure - bar CNI (Peak Peak) Cylinder Pressure - bar DIESEL COMBUSTION NOISE CNI (COMBUSTION NOISE INDEX) Theory Pressure Trace AVL CNI vs. CA AVL CNI vs. Torque AVL CNI Filter Time Masking Consideration Benefits Calculation based on cylinder pressure (possible in non-acoustic environments) Drawbacks Influence of engine structure not considered High effort to install cylinder pressure sensors Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 12

DIESEL COMBUSTION NOISE CKI (COMBUSTION KNOCKING INDEX) 3000 UPM 2000 UPM 1000 UPM Leerlauf Transfer Function Theory Band Pass and Envelope Calculation 3000 UPM 2000 UPM 1000 UPM Leerlauf Transfer Function 3000 UPM 2000 UPM 1000 UPM Leerlauf Transfer Function Consideration of Time and Frequency Masking CKI Result 0.4 0.4 0.2 0.2 Amplitude -0.0 Amplitude -0.0-0.2-0.2-0.4-0.4 0.10 0.15 0.20 0.25 0.00 0.05 0.10 0.15 0.20 0.00 0.25 0.05 0.10 0.15 0.20 0.25 Time - s Time - s Time - s Benefits Only microphone measurement Influence of engine structure considered High agreement between subjective assessment and CKI value Drawbacks Calculation based on airborne noise (for benchmarking) Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 13

DIESEL COMBUSTION NOISE DIESEL COMBUSTION NOISE ASSESSMENT Combustion Noise Assessment Cylinder Pressure Glow Plug Adapter Structure Borne Accelerometer Airborne Microphone Combustion Noise Benchmarking (engine test bed, vehicle interior) Combustion Noise Development Combustion Noise Monitoring AVL CNI Combustion Noise Index AVL CKI SBN Combustion Knocking Index AVL CKI ABN Combustion Knocking Index AVL Algorithms for Combustion Knocking Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 14

NEW MEASUREMENT PLATFORM expandable to 96 channels flexible configuration exchangeable modules Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 16

A COMPLETE SET OF DIESEL NVH ALGORITHMS AVL offers a complete set of algorithms for Diesel combustion noise assessment. These algorithms are suitable to cover the complete development process to ensure that NVH development targets will be successfully reached. AVL CNL Combustion Noise Level AVL CNI Combustion Noise Index AVL CKI Combustion Knocking Index Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 17

) AVL Sound CNI - kpa Pressure - db(a) Sound Pressure - db(a) AVL CNI - kpa AVL Sound CKI Pressure - db(a) Sound Pressure - db(a) AVL Sound CNI - kpa Pressure - db(a) AVL CKI Sound Pressure - db(a) Sound Pressure AVL CKI - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) Sound Pressure - db(a) INDICOM & NVH SIGNAL PROCESSING TOOLS Various possibilities are given to display microphone, accelerometer, cylinder pressure signals as well as results of customized analysis algorithms. Both time and crank angle based data can be processed CYLINDER PRESSURE Object: Ford Panther GraphicDef.: AVL ENG ABN - A265008 3 rd Page: OCTAVE ABN max SPECTRA Torque 2000 rpm with different loads 100 90 70 50 MP1: Top 40 73.1 dba 90 30 83.0 dba 83.4 dba 20 31.5 100 315 1k 3.15k 10k Object: Ford Panther Octave Frequency - Hz GraphicDef.: AVL 70 ENG CN Diesel - A265008 Page: CNI_nru_lvergl AVL MP4: Left 100 CNI Load run ups 1000 2000 3000 4000 90 Rotational Speed - rpm 100 % Load Run Up 200 MP4: Left 110 70 150 50 90 100 40 76.7 dba 30 86.8 dba 85.6 dba 50 20 70 31.5 100 315 1k 3.15k 10k Octave Frequency - Hz 0 1000 2000 3000 4000 1000 2000 3000 Rotational 4000 Speed - rpm 5000 Rotational Speed - rpm 200 Object: Ford Panther GraphicDef.: AVL ENG ABN - A265008 Page: ABN nru Condition 110 100 100 40Nm Run Up MP1: Top Object: 0 Ford % Load Panther 1750 rpm GraphicDef.: 50 % Load AVL 1750 ENG ABN rpm - A265008 OVERALL Page: 100 ABN &% nru Load OCTAVE l100 1750 OL+Oct rpm LEVELS 100% load run up 100 110 90 100 70 90 110 50 100 MP2: MP1: Right Top 40 70 76.8 dba 90 30 86.7 dba 85.9 dba 20 31.5 1000 100 2000 315 1k3000 3.15k 10k 4000 Object: AVL Ford Scatterband Octave Rotational Panther Frequency Speed - Hz - rpm GraphicDef.: 70 AVL Target AVL Line ENG CN Diesel - A265008 Page: CKI_nru_lvergl Cylinder Pressure MP4: Left 1 110 Cylinder Pressure 2 1000 2000 3000 4000 Rotational Speed - rpm 100 1Nm 100 % Load Run Up Run Up 2002 AVL CKI Load run ups 240 90 150220 200 1001 70 1 50140 1000 2000 3000 4000 120 Rotational Speed - rpm 0100 1000 1000 2000 2000 3000 3000 4000 4000 5000 5000 Rotational Rotational Speed Speed - rpm- rpm 2002 100 % Load Run Up 1Nm Run Up Nm Run Up MP2: Right 20Nm 40 Nm Run Run Up Up 100110 40Nm Run Up 90 0 % Load Run Up 100 Motored Run Up 70 90 110 10050 MP3: MP2: Front Right 40 70 75.9 dba 90 30 86.5 dba 85.7 dba 20 31.5 1000 100 2000 315 1k3000 3.15k 10k 4000 Object: AVL Cylinder Octave Scatterband Ford Rotational Pressure Frequency Panther Speed 3 - Hz - rpm MP1: Top 70 GraphicDef.: AVL Cylinder Target Pressure Line AVL ENG 4ABN - A265008 MP2: Right 3D Page: SPECTRA ABN AVG nru All l100 0-1 Mics 3dkHz MP3: Front 100 1000 100% 2000 load run 3000up 4000 90 Rotational Speed - MP1: rpm Top Nm 200 2 1 4Run Nm 6Run Up 8Up 10 2 AVG All Mics 90 110 70 240 100 150 220 Test: 100 % Load Run Up Overall Level INJECTION Octave 500 SIGNALS Hz 50 200 90 3000 100 40 1 50 75.9 dba 30 1 86.0 dba 40 2000 85.2 dba 7020 50140 30 31.5 100 315 1k 3.15k 10k 120 Octave Frequency - Hz 20 1000 0 2000 3000 4000 1000 10 1000 Rotational 01000 2000 Speed 2000400 - rpm3000 0 4000 4000 05000 1000 5000 Rotational Rotational Frequency Speed Speed - rpm - Hz - rpm MP4: Left 0 % 200 2 Load 204 Nm Run 6Run Up8 Up 10 2 90 Rotational Speed - rpm 4000 MP3: Front Octave 1000 Hz Octave 2000 Hz Octave 4000 Hz 1000 2000 3000 Test: Rotational 100 MP4: % Speed Load Left Stephan 240BRANDL, Wolfgang SCHWARZ, Andreas LOECKER 240 4000 Vehicle and Powertrain NVH 24 November 240 2016 18 70 110 100 90 70 110 100 2 90240 220 200 3000 70 1 MP3: Front AVG All Mics 1 2000 1000 140 2000 3000 120 Rotational Speed Rotational Speed - rpm 4000 2 MP2: R 4Nm Run 6 100 1000 1000 0 2002000400 300 Rotational Frequenc Spe 2 0 % Load Ru

DIESEL COMBUSTION NOISE SIMULATION BASED KNOCKING EVALUATION Source / Excitation Mechanism Transfer Response Pressure in combustion chamber Internal force path (transfer function) structural response CKI SBN CNI Accelerometer position External force path / radiation (transfer function) excitation source and mechanisms Microphone position radiated air borne noise CKI ABN CRUISE M crank angle resolved and mixture controlled combustion (MCC) simulation Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 20

Rotational Speed - rpm Sound Pressure - dba Time - s Sound Pressure - dba Time - s Sound Pressure - dba Rotational Speed - rpm Rotational Speed - rpm Sound Pressure - dba Rotational Speed - rpm Rotational Speed - rpm Sound Pressure - dba Rotational Speed - rpm Rotational Speed - rpm TURBOCHARGER NOISE QUALITY PARAMETERS FOR TC NOISE ASSESSMENT AND REFINEMENT Driver's ear lhs - Sound Pressure - db(a) Driver's ear rhs - Sound Pressure - db(a) 4500 90 4500 90 TURBO CHARGER NOISE PHENOMENA 3000 Tonal Noises 40 => directly related to the TC 2000 2000 20 20 Constant Tone Unbalance Whistle Noise Blade Passing Noise Ball Bearing Noise 4000 3500 3000 2500 2000 1500 4000 3500 2500 1500 Intake orifice 1500 2000 2500 3000 3500 4000 100 Frequency - Hz 4500 Firewall top lhs - Sound Pressure - db(a) 4500 4000 3500 3000 2500 1500 2000 2000 20 1000 10 20 20 0 100 200 300 400 500 0 700 0 900 1000 1100 1200 1300 1400 10000 11000 12000 13000 14000 15000 100 17000 100 19000 20000 4 Frequency - Hz Frequency - Hz 1500 10 1500 10 1500 2000 2500 3000 3500 4000 1500 2000 2500 3000 3500 4000 3 Frequency - Hz Frequency - Hz 0 1000 2000 3000 4000 5000 00 7000 00 9000 10000 11000 12000 Frequency - Hz Broadband Noises Flow Noise 90 70 50 40 30 4000 3500 3000 2500 2000 70 50 40 30 10 90 70 50 40 30 4000 3500 3000 2500 1500 1500 Turbo 2000 charger nearfield 2500 3000 3500 4000 Frequency - Hz 90 Firewall top rhs - Sound Pressure - db(a) 4500 4000 3500 3000 2500 9 8 7 6 5 70 50 40 30 20 Intake orifice lhs => non-synchronous, related to TC / intake system 70 50 30 10 Let Off Noise 90 70 50 40 30 110 100 90 70 50 40 30 Intake orifice Intake orifice lhs 4000 100 6 100 3500 90 5 90 3000 70 4 70 3 2500 50 2 50 2000 40 40 30 1 30 1500 0 5000 10000 15000 20000 Frequency - Hz 20 0 0 5000 10000 15000 20000 Frequency - Hz 20 Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 22

Intercooler TURBOCHARGER NOISE QUALITY PARAMETERS FOR TC NOISE ASSESSMENT AND REFINEMENT MEASUREMENT OF INPUT DATA / DETECTION Measurement Setup Operating Conditions Vehicle Interior Artificial Head Engine Compartment Air filter housing Intake Orifice Near-field Cylinder Block Accelerometer TC Near-field Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 23

Time - s Time - s TURBOCHARGER NOISE QUALITY PARAMETERS FOR TC NOISE ASSESSMENT AND REFINEMENT TONAL NOISE CALCULATION EXAMPLE Blade Passing Noise Improvement during Development Process 16 Co-Driver ear lhs - Sound Pressure - db(a) Interior Noise 70 13 Co-Driver ear lhs - Sound Pressure - db(a) 70 15 12 14 50 11 50 13 40 10 40 12 30 9 30 11 20 8 20 10 10 7 10 9 8 Baseline 0 2000 4000 00 00 10000 12000 Frequency - Hz 0-10 6 5 with HFD 0 2000 4000 00 00 10000 12000 Frequency - Hz 0-10 The vehicle showed a high blade passing order in baseline condition. Implementing a high frequency damper on pressure side of the turbocharger significantly improved the noise. Blade Passing Noise Parameter 6.0 => 8.2 Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 24

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS TURBOCHARGER NOISE - SIMULATION Overview of Current Capabilities Rotor Floating Bushings Housing Development Measures Critical Speeds Rotor Displacements Speeds of Floating Bushings Bearing Forces, Oil Film Pressures Oil Flows, Oil Temperatures Dynamic Simulation Integrated Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 25

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS TURBOCHARGER NOISE - SIMULATION 20000 rpm Overview of Current Capabilities Harmonics Oil Pressure Distribution Rotor Orbital Path; Detecting Unstable Conditions Frequency First order frequency from unbalance Sub harmonics (can only be predicted when using EHD bearing definition) can be identified (important for NVH) 150 cycles last 50 cycles of total 150 cycles are plotted Typical displacement at the compressor nut Unstable behavior at 140000 rpm Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 26

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS NVH DRIVELINE DEVELOPMENT - SIMULATION Overview of Current Capabilities Fully Flexible Driveline Excite model incl. bending and torsion Engine mount characteristics Sound radiation Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 28

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS STEPWISE DEVELOPMENT APPROACH Engine Alone Engine & Transmission Entire Drive Line Torsional Approach Entire Drive Line Bending &Torsional Model Flexible Rear Sub-frame Flexible Sub-frame and Power Unit Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 30

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS NVH DRIVELINE DEVELOPMENT - MEASUREMENT In Vehicle Assessment Main measurement parameters Torsional vibration Acceleration Sound Pressure Level ECU data Dyno Assessment Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 31

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS NVH DRIVELINE DEVELOPMENT - MEASUREMENT Virtual testing Use simulation for component/subsystem optimization Connect existing simulation models to perform virtual system testing Real Testing Replace hardware components by dynos for subsystem or component testing E.g. Clonk testing on PT test bed Mixed testing Replace hardware components by simulation models Replace simulation models by hardware components Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 32

FULL ELECTRIC VEHICLES (EV) CONCEPTUAL APPROACHES BMW i3 Axle Drive RENAULT Zoe Single step final drive Single step final drive Differential Differential Single step final drive Differential Differential ZF Electric Drive Single step final drive Planetary gear set for speed reduction Co-axial layout with hollow shafts Planetary gear set for speed reduction Tesla Model S RWD and AWD Parallel-axial layout Differential Differential GM s Electric Drive Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 34

ELECTRICAL AND MECHANICAL NOISE OF FULL ELECTRIC DRIVELINE Shaft bearings Gears Shafts Torque ripple to driveline Electrical excitation, electromagnetic field Mechanical excitation, Multibody dynamic Sound radiation, from two sources, electrical excitation and driveline Driveline mount Structure borne noise to the vehicle structure Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 35

Electrical Noise Mechanical Noise 40 20 0-20 -40-100 40 20 0 Isa Isb Isc Tem_IM4 0.3 0.32 0.34 0.36 0.38 0.4 Time (s) ELECTRICAL AND MECHANICAL NOISE OF FULL ELECTRIC DRIVELINE PWM Harmonics FEM magnetic field Radial, tangential & bearing forces mapped to FEM Torque and bearing forces to EXCITE, calculate eccentricity and come back Electrical + Mechanical Coupled simulation excitations in frequency domain (forced response analysis) Ripple torque Tool: Multi-body Dynamic Driveline, whine, etc. Export excitation in frequency domain Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 36

ELECTRICAL AND MECHANICAL NOISE OF FULL ELECTRIC DRIVELINE Electromagnetic excitation Electromagnetic excitation + Mechanical Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 37

E-COMPONENT AUDIBLE NOISE PREDICTION POWER ELECTRONIC Target: Noise distribution prediction of emitted noise from the source (single components L, C, silicon) via the distributor (PCB (printed circuit board), mechanic, housing) to vehicle interior without hardware. Benefit: Simulation driven assessment and optimization already in an early phase of development Support component selection and schematic diagram development Improved positioning strategies and structural design of PCB and housing Assessment of noise contribution to vehicle interior estimation of annoyance Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 39

E-COMPONENT AUDIBLE NOISE PREDICTION POWER ELECTRONIC Approach: Noise source prediction by simulation in an early phase, based on schematic diagrams technical component information Noise transfer prediction via simulation of structural vibrations and radiated noise Noise source analysis at start of development Noise transfer during development I/U/f Physical model Schematic based audible noise detection CAE based audible noise simulation Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 40

VALIDATION OF PCB STRUCTURAL MODEL Detailed 3D FEM Model (NASTRAN) PCB without excitation Eigenfrequencies PCB with excitation (x, y) position dependence Free-Free Conditions Constrained (mounted on bed plate): Non-isotropic material FR-4 Mass 27.695 g Copper (Cu) RBE2 s CBAR s SPC s Tin (SnPb) Results of FEM modal analysis and forced response analysis are compared in terms of eigenfrequencies and transfer functions measured with laser vibrometer. Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 41

rows 25mm EXPERIMENTAL SYSTEM IDENTIFICATION OF ELECTROACOUSTIC TEST BOARD MODELING OF TEST BOARD EXCITATION POSITIONS 20mm columns f00 Excitation by small hammer f22 f44 Animation of Excitation Positions Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 42

FREE-FREE EIGENFREQUENCY ANALYSIS 377 Hz 567 Hz 906 Hz 924 Hz 1144 Hz 1582 Hz 1822 Hz 1855 Hz 2541 Hz 2734 Hz Measurement 21 Hz 3076 Hz Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 43

TRANSFER FUNCTION COMPARISON FULLY ISOLATED PCB System Response of fully Isolated System V1.0 (laser vibrometer) FEM free-free calculation sol111 Fully isolated by elastic bands f11 f22 f33 F-range: 0-5kHz (coherence (evaluation window): 270Hz 3kHz) Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 44

TRANSFER FUNCTION COMPARISON PCB MOUNTED ON BED PLATE System Response with mounted plate ( Bed Plate ) V1.1 (laser vibrometer) FEM constrained calculation sol111 Mounted on Bed Plate Frequency range: 0-5kHz (coherence: 270Hz 3kHz) f11 f22 f33 Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 45

DEVELOPMENT APPROACHES FOR MODERN POWERTRAINS SUMMARY NVH for Conventional Powertrains Complete powertrain models can be used in early development phase to predict complex NVH phenomena (e.g. roughness) Improved NVH simulation and parameters for components (e.g. Turbocharger NVH) Simulation delivers valuable imput in NVH driveline development NVH in Electrification Electromagnetic and mechanical (esp. gear contact) excitation need to be considered New NVH frontloading approaches for Power Electronics are under investigation Stephan BRANDL, Wolfgang SCHWARZ, Andreas LOECKER Vehicle and Powertrain NVH 24 November 2016 46

THANK YOU www.avl.com