Analysis of road noise problem of a vehicle based on CAE technology

Analysis of road noise problem of a vehicle based on CAE technology

Analysis of road noise problem of a vehicle based on CAE technology

This paper addresses excessive noise at the driver’s right ear induced by road excitation during prototype testing of a domestic vehicle. By integrating experimental measurements with finite element technology, we used test data as excitation inputs for CAE-based transfer path analysis (TPA). Comparative validation with test results confirms the feasibility of CAE TPA. The root cause of the 53Hz peak identified in tests was successfully determined.

Keywords: Transfer Path Analysis (TPA), Road Excitation, Finite Element Method (FEM), Vibro-Acoustic Coupling

Introduction
With the rapid development of China’s automotive industry and increasing vehicle ownership, ride comfort has gained significant attention. Vibration and noise are critical factors influencing consumer purchasing decisions and serve as key metrics for vehicle quality assessment. To enhance market competitiveness, manufacturers prioritize NVH performance, often highlighting vibration and noise control as a selling point.

Road-induced vibration (predominantly <300Hz) is one of the three primary noise sources during driving, particularly pronounced on rough surfaces. Conventional solutions rely on experimental TPA to identify dominant noise paths for optimization. While frequency response functions (FRFs) and noise transfer functions (NTFs) are readily obtainable in classical TPA, accuracy is compromised by:

1. Load identification challenges:
   • Direct measurement: Demands high-precision sensors and ideal installation conditions.
   • Dynamic stiffness method: Limited by scarce accurate stiffness data for soft mounts.
   • Pseudo-inverse method: Ill-conditioned FRF matrices require ≥2× indicator points per load path, prolonging development cycles.

This study proposes a hybrid experimental-CAE TPA approach. Test-measured force spectra (60 km/h on asphalt) served as excitations for boundary element-based acoustic transfer vector analysis. CAE-calculated A-weighted SPL (1/3-octave, 20–200Hz) at the driver’s right ear was validated against test data. Subsequent CAE TPA evaluated vector contributions of each excitation path to pinpoint the root cause of road noise.

1. Analysis Theory

Road-tire interaction transmits vibration via the suspension to the body. Body panel vibrations generate noise that propagates through the cabin air, forming sound pressure at the driver’s right ear. Concurrently, air pressure constrains panel vibrations, creating vibro-acoustic coupling. Thus, analyzing sound pressure requires coupled structural-acoustic modal analysis, combining structural dynamics and fluid continuity equations:

Subscripts:

• s: Structure
• f: Fluid
• N, sN: Fluid pressure and structural shape function matrices

Structural modes (vacuum):

Fluid modes (rigid container):

Substituting Equations (2)–(12) into (1) yields the coupled system equation in modal coordinates:

2. Feasibility Verification

2.1 Test Data Acquisition

(1) Excitation Points:
Research confirms tire/road noise dominates at 50–60 km/h on dry surfaces, worsening on rough roads. Road excitation transfers to the body through two primary paths:

1. Suspension springs/dampers → body upper mounts
2. Brackets → body support points
   Fourteen excitation points (Fig. 1) were selected to evaluate driver’s right-ear noise.

(2) Test Conditions:

• Rough road surface, open surroundings
• Background noise >15 dB(A) below target
• Prototype coasting at 60 km/h

(3) Data Collection:

• Acceleration sensors at excitation points (Fig. 1)
• BSWA microphone at driver’s right ear
• Commercial software for data processing

2.2 CAE Simulation

• Frequency Range: 20–200 Hz (road excitation energy <100 Hz)
• Model: Trimmed Body (TB) excluding powertrain/exhaust/drivetrain/suspension/tires
• Preprocessing (HyperMesh):
  • 1,544,592 SHELL elements (10 mm mesh)
  • Connections: RBE2 (bolts), ACM/CWELD (welds), adhesives
  • Mass balancing: CONM2
  • Weight: 790.5 kg (Fig. 2)

• Acoustic Cavity: PSOLID elements (50 mm) for cabin air (Fig. 3)

• Solver: Nastran modal frequency response

2.3 Test-CAE Correlation

Driver’s right-ear sound pressure comparison (Fig. 4) confirms CAE accuracy, enabling path contribution analysis.

3. 53Hz Peak Noise Analysis

3.1 TPA Results

Dominant paths to 53Hz peak (Fig. 5):

1. Left-front damper (Y-direction)
2. Right-rear damper (Z-direction)
3. Right-rear trailing arm (Y-direction)
4. Upper-right linkage (X-direction)

3.2 Excitation Source Analysis

All four paths show force spectrum peaks near 53Hz (Fig. 6).

To isolate cause:

• Modified excitation: Flattened force spectra ±3Hz around 53Hz reduced cabin noise peak (Fig. 7).

• Stiffness modification: Increased Young’s modulus/thickness at excitation points showed negligible impact (Fig. 8).

Conclusion: The 53Hz peak originated from road excitation spectra, not structural stiffness.

4. Conclusions & Outlook

• Hybrid experimental-CAE TPA identified the 53Hz road noise source.
• Future work: Expand excitation databases for varied suspensions/roads to enable predictive NVH design during CAE phases, reducing development time/cost.

Source: Proceedings of the 2016 International Symposium on Automotive NVH Control

Contact:

Prof. Tian：WhatsApp：+86 15029941570 | Mailbox：540673737@qq.com