MGA Research Corporation Real Time Simulation Testing Gerald Roesser David Nagle Thomas Hutter MGA Research Corporation 1
MGA Research Corporation PRESENTERS Gerald Roesser BSEE MGA Associate since 2001 Group Leader Simulation Engineer Dave Nagle BSME MGA Associate since 2006 Simulation Test Engineer Thomas Hutter BSEE MGA Associate since 2000 Simulation Engineer 2
MGA Research Corporation Independent test services Began in New York, 1977 ISO/IEC 17025:2005 Accredited Specialize in Safety Regulations (FMVSS, ECE, SAE, ASTM, MIL_STD, etc.), Vibration, Noise, Life Cycle, Environmental, Equipment, etc. Industries served include: Automotive, Military, Aerospace, Rail, Other Transportation, etc. 3 Burlington, Wisconsin Troy, Michigan Greer, South Carolina Akron, New York Manassas, Virginia
MGA Research Corporation MGA and Simulation Testing 10 years of experience with Simulation Testing 13 Multi-Axis Simulation Tables 150 plus servo-hydraulic loops for multi-axis durability testing Hundreds of Simulation Tests and Components validated per year 10 Simulation Engineers Simulation Test supplier to all major OEMS and the majority of their suppliers 4
WEBINAR OVERVIEW Introduction: Vehicle Durability Testing Simulation and MAST testing overview Road Load Data Acquisition and Analysis Fixture Design Drive File Development Question and Answer 5
INTRODUCTION 6
INTRODUCTION Vehicle Durability Testing Simulation Testing is used in all transportation industries Using actual road data can help predict product performance more accurately. Design Goals for Quality can be met with less development time Multi-Axis Simulation Table (MAST) will be used as an example of simulation testing 7
INTRODUCTION Vehicle Durability Assessment Methods Test Driving, Actual Highway Miles Highly accurate prediction Time consuming Not Practical for new vehicles Proving Ground Track Loop Accurate life predication Highly Correlated Expensive Time and labor intensive Available late in the design process 8
INTRODUCTION Vehicle Durability Assessment Methods Component Block Cycle Testing Good for suppliers Great tool for A to B comparison Not representative of real world Not adequate for complex systems Real-Time Road Simulation Saves development time Accurately correlate to road data and predicts product life 9
Real Time Simulation Testing 10
Simulation Testing Overview Simulation Test Flow Chart 11
Simulation Testing Overview Determine end use of product or vehicle Perform a study to determine what type of road surfaces the vehicle will travel on Create Testing Goals Generate a Durability Schedule based on life expectancy and 95 th percentile usage Select a Test Rig that will be able to apply the collected road inputs 12
Simulation Testing Overview Surface # Time (s) Repeats Test Time (h) 1 2 3 4 5 6 100 528 66 42 1525 42 1456 42 547 57 900 57 Total Test Length 14.7 0.8 17.8 17.0 8.7 14.3 73.1 Example of a Generic Durability Schedule 13
Road Load Data Acquisition (RLDA) Data Types Collected During RLDA Wheel Force Transducers Wheel end to body displacement Chassis, body, or component tri-axial accelerometers Wheel-end acceleration Strain Gages Vehicle Speed GPS Environmental Conditions 14
Road Load Data Acquisition (RLDA) Optimize Data for Test Rig Use Durability Schedule to calculate occurrences of different road surfaces Check data and prepare for analysis and reduction 15
Road Load Data Acquisition (RLDA) Data Normalization Collected data must be normalized so that relative comparisons can be made between different road surfaces, events, and recordings made from the test rig itself. There must be a method in place so that accurate correlations can be derived. In order to compare surfaces, correlate to the real world, or find the worst case vibration scenario, the fatigue damage must be calculated. 16
Road Load Data Acquisition (RLDA) Normalize data for easy comparisons Conduct Relative Fatigue Damage Analysis Create Fatigue Damage Spectrum Establish relationship between actual road miles and laboratory miles using damage figures 17
Road Load Data Acquisition (RLDA) 18 Damage is a unit-less number that normalizes complex data for comparison Data can be reduced to shorten time Damage comparisons are performed to ensure data still correlates Damage Analysis saves a significant amount of testing time
Simulation Test Rigs Upon editing and optimizing the data, it is time to select an appropriate test rig for your application. There several types of real-time simulation rigs and many ways to apply these inputs to your product. Real-Time simulation can mean applying force inputs, acceleration, displacement or even strain to your product on as many axes as collected in the field. 19
Simulation Test Rigs 20
MAST TABLE Multi-Axis Simulation Table (MAST) Made up of 6 hydraulic actuators connected to a test table Capable of moving in 6 degrees of freedom. X, Y, Z, roll, pitch, and yaw Frequency band of 2 to 50 Hz. Each actuator is instrumented with an LVDT to measure displacement. Samples are mounted to the table top 21
Test Fixture Design 22
Test Fixture Design The fixture must not a have a natural frequency in the frequency band of the test data. The fixture must also be light weight so as to not add too much moving mass to the test rig. CAD data is used to generate the fixture to ensure that the sample is mounted exactly as it would be on the vehicle and no extraneous stresses are induced. Test fixture itself can be run through a computer simulation to determine where improvements can be made. 23
Drive File Development 24
Drive File Development The drive file is the corrected data set that is generated to accurately match the desired data. A closed loop servo-hydraulic system will not immediately be able to match the desired data on command Data must be corrected for the dynamics of the system. PID tuning is effective for fixed frequency, fixed amplitude testing It is very difficult to get the command and feedback on a multi-axis system with crosstalk and complex dynamics at play to converge. Advanced software and controls in modern hydraulic controllers make the process easier. (RPC Pro) 25
Drive File Development Accelerometers are mounted to a test fixture or the sample just as they were when the data was collected. (Normally 3 tri-axial Accels) A random (white noise) input is commanded to each of the drive channels as the accelerometer response is recorded. This response yields a transfer function which can be used to match the acceleration response with the desired acceleration data collected from the road surfaces. Transfer functions and inverse transfer functions are generated. 26
Drive File Development Drive File Iterations System Inverse Example 27
Drive File Development Play-out of this drive file on the test rig. Calculate Error Adjust Gain on Drive File Repeat until data converges as pictured on the right 28
Drive File Development 29
Drive File Development The drive file iteration process is complete when all time history response files have been adequately matched to the desired data RMS value of each channel should be at least within 5% Maximum and minimum acceleration peaks should be within 10% Overlay the desired spectrum and response spectrum 30
Drive File Development 31
32
QUESTION AND ANSWER 33
QUESTION AND ANSWER 34 Thank you for participating! MGA Contacts: Group Leader Gerald Roesser (Gerald.Roesser@mgaresearch.com) (248) 259-7331 Simulation Engineer Dave Nagle (David.Nagle@mgaresearch.com) (248) 670-4838 Simulation Engineer Thomas Hutter (Thomas.Hutter@mgaresearch.com) (586) 491-6161