基于 WFT 信号的整车多方案仿真对比分析

Full Vehicle Simulation and Virtual Iteration

Otmar  Gattringer,2015

摘要：

本文基于 WFT 车轮力信号，采用 MSC.ADAMS 构建整车多体模型，对比四种整车仿真与虚拟迭代方案：直接 WFT 输入、车身固定 WFT 输入、四轮台架 + WFT 虚拟迭代、内部信号虚拟迭代。统一采用颠簸路载荷，对比轮力、悬架位移、球铰力、车身力等 70 通道信号与损伤结果。研究表明：直接 WFT 方法最简单但精度最差；四轮台架 + WFT 迭代便捷且精度合格；内部信号虚拟迭代精度最高，可通过载荷修正补偿模型误差。两种虚拟迭代方法可提取路面位移谱并跨车型迁移，为整车道路载荷仿真与疲劳分析提供优选方案。

Contents

· Overview

· Road load data

· MBS-model

· Simulation and VI

· Comparison of results

· Conclusion

Overview

4 different approaches of one full vehicle

A1Full vehicle simulation using WFT signals as input

A2 il,ef u tig WFT signals as

(“WFT with fixed body”)

A3 VI of full vehicle using WFT signals

(“VI using internal signals”)

· All simulations based on same road load data

· Comparison of results of the different approac

Road load data-application

Multilink suspension-front and rear axle

Wheel Force Transducer(WFT)

· KISTLER Roadyn

·3 Forces at 4 wheels 

·3 Torques at 4 wheels

Road load data

Measurements of rough road,

70 channels used by the 4 approaches:

· WFT signals (24 channels)

· Spring deflection (4 channels)

· Damper forces (4 channels)

· Ball joint forces

        - Front axle lower link (6 channels -X,Y,Z)

        - Front axle upper link (4 channels-X,Y)

        - Rear axle upper link (4 channels-X,Y)

· Accelerations

        - Wheel hub front and rear X,Y,Z(12 channels)

        - Body Z(4 channels)

· Steering rod forces axial (2 channels)

· Rear axle link forces axial (6 channels)

MBS-model

Full vehicle model (MSC.ADAMS/CAR)

· Front  suspension . Powertrain

· Front  stabilizer

· Steering

· Rear suspension

· Rear stabilizer

· Body

· Load

        - WFT signals  (free body or body fixed to ground)

        - 4 Poster with WFT signals

        - VI using internal signals

· Load at each wheel:

· (no driving/braking torque)

·A1  and A2:

- FX,FY,FZ,TX,TZ

(no driving/braking torque)

- Applied by GFORCE element

· A3 :

4 Poster with WFT signals    (5 channels at wheel center)

- vertical  displacement  (VI-4  poster)

- FX,FY,TX,TZ(measured       signals applied  additionally)

-Applied  by  GFORCE  and  MOTION elements

MBS-model

· A4 :

VI using internal signals (4 channels at wheel)

-vertical  displacement  at  wheel center

        - FX and TZ at wheel center

        - FY at tire contact point

        - Applied by GFORCE element at wheel center,GFORCE element at tire contact point and MOTION element

Full vehicle assembly

· Assembling all 8 subsystems

· Adjusting center of gravity of body so that measured static wheel loads are achieved

        · Adjusting preload of springs (measurement condition)

        - Front bumpstop clearance

        - Rear bumpstop clearance

Simulation and VI

A1: Full vehicle simulation using WFT signals

· Spring for stabilization (simple method):

        - X and Y direction at beginning (STEP function)

        - Z direction 10 N/mm

A1: Full vehicle simulation using WFT signals

· The system will be instable without springs:

A1: Full vehicle simulation using WFT signals

· The system will be stable with springs:

A2:Full vehicle simulation using WFT signals,body fixed to ground

· The system will be stable

A3:4 Poster of full vehicle with WFT signals

· VI of 4 poster

        - Load:vertical displacements

        - Desired:spring displacements and wheel center accelerations

        - 7 iterations automatically

· Applying additional measured WFT signals:

        - Vertical displacements of VI of 4 poster(7th drive)

        - WFT signals at wheel centers:FX,FY,TX and TZ

        - Simulation  of full  vehicle

A4:Vl of full vehicle using internal signals

· Load:16  channels

        - Vertical displacements DZ at wheel centers

        - FX and TZ at wheel centers

        - FY at tire contact points

· Desired:28  channels

        - Spring displacements and wheel center accelerations

        - Ball joint forces (X and Y direction),front and rear axle

        - Steering rod forces

        - Rear axle link forces

· 10 iterations automatically

· 11th drive:10th drive scaled manually by:

        - DZ front left divided by 1.04|DZ front right divided by 1.08

        - TZ front left divided by 1.15|TZ front right divided by 1.25

Comparison of result-measured signals         

Results  of  different  an alyses

·Reference/desired   signals   show   satisfying   correlation   in   time   domain

         A1 Full vehicle simulation using WFT signals:

                -WFT    signals

        A2 Full vehicle simulation using WFT signals,body fixed to ground:

               -WFT    signals

        A34 Poster with WFT signals:

               -WFT    signals

               -spring displacements and wheel center accelerations

        A4 Vl using internal signals:

                -internal    signals

·Relative   damage   values   of   simulated   to    measured   signals,comparison between  the  4  approaches  (page  18-20)

·Comparison  of  body  forces   of  the  4   approaches  (page  21-25)

Wheel center forces:front left and rear left

· Common target range for relative damage values:[0.5,2](yellow area)

· Request are defined at the joint of spindle to knuckle (s mall deviation to applied signal)

·A1     -"WFT“,A2-"WFT   with  fixed  body"and  A3-“4 Poster with WFT"signals show excellent correlation logically

· The green bars of   A4-"Vl using internal signals"illustrate model inaccuracies

Vertical signals:spring displacements,damper forces and wheel center accelerations

· Blue bars of  A1-"WFT"and A2-"  WFT with fixed body"show that WFT forces are not optimal suitable for vertical load

· A3- “4 Poster with WFT signals"and A4 -"Vl using internal signals”show correlation to measurement with high quality for vertical direction

Ball joint forces (front axle),steering rod forces (front axle)and link forces (rear axle)

· Inadequate accuracy of A1-“WFT”,A2-    "WFT with fixed  body"and A3-“4  Poster with WFT signals"at some internal measurement signals (reflecting model inaccuracies)

· A4 -“VI using internal signals”shows satisfying results,except front link of rear axle is out of target area

Comparison of result-body forces          

Comparison of body forces (e.g.used for fatigue an alysis of body)

· No reference/measurement available

· Damage values computed relative to “VI using internal signals”

· Comparison for vertical-,longitudinal-and lateral forces (left side only)

            - Front axle (FA)spring/damper mount

            - Front axle (FA)subframe (front-,middle-and rear mount)

            - Front axle (FA)upper control arm front mount

            - Rear axle (RA)spring/damper mount

            - Rear axle (RA)subframe (front-and rear mount)

-Comparison of all body signals between simulation using WFT signals as input

            - Free body

            - Fixed body

Longitudinal body forces

· FX springs show very s mall amplitudes

· Upper control arm forces have s mall amplitudes in all 3 directions

·Longitudinal forces are s mall compared to A4-“VI using internal signals“

Lateral body forces

· Upper control arm forces have s mall amplitudes in all 3 directions

·A1-     "WFT"and A2-"WFT with fixed  body" have high deviation compared to A4 -“Vl using internal signals“

Vertical body forces

· Upper control arm forces have s mall amplitudes in all 3 directions,

specially in Z-direction

·A1   -"WFT"and  A2-"WFT with fixed body"have high deviation compared to A4 -“VI using internal signals“

Comparison of all body forces between free (A1) and  fixed  body(A2)consideration (simulation using WFT signals as input-no VI,forces at left side only)

· No information of engine mount forces with fixed body

· Accumulation of all s mall inaccuracies may lead to inexact evaluations (e.g.fatigue)

Conclusion   

· 3 approaches are common in practice,body fixed to ground most unusual

· A1 and A2-Full vehicle simulation using WFT signals is the simplest but most inaccurate  method  regarding  internal forces

· A3-  VI of a 4 poster with applying WFT signals additionally is a very

convenient approach and the accuracy is acceptable regarding internal forces

· A4 -VI using internal forces is a very accurate method assuming

measurement data is available.Model inaccuracies can be compensated partially by adjusting the load to get accurate internal signals

· The two approaches with VI (A3 and A4) can also be used for computing

vertical displacements at tire contact point including radial tire stiffness.These

displacements reflect the road surface (invariant signal)and can be

transferred to similar vehicles.The measured forces are not invariant and

cannot be transferred automatically.

· Accuracy of all approaches depends on model quality.The presented results cannot be transferred to other models generally.The vehicle is modeled

according ECS standard and is not specially trimmed for one

of the 3 approaches.

免责声明：

本页面/内容部分素材来源于互联网公 开 信 息，旨在传递更多信息，不代表本平台立场。

版权归原作者或机构所有，如涉及侵权，请通过平台联系我们，我们将在核实后第一时间处理。

本平台对转载内容的真实性、准确性不作任何保证，用户需自行判断并承担使用风险。