V&V OF ANALYSIS RESULTS CREATED DURING DEVELOPMENT OF A MINE TRUCK AT ATLAS COPCO ROCK DRILLS AB. By Jari Hyvärinen/RCT

Transcription

1 V&V OF ANALYSIS RESULTS CREATED DURING DEVELOPMENT OF A MINE TRUCK AT ATLAS COPCO ROCK DRILLS AB By Jari Hyvärinen/RCT

2 CONTENT Atlas Copco Vehicles The Models Used During Development. The Methodology Used During Development. Measurements Included When Validating the Vehicle Behavior. Comparing the Numerical Results with the Measured Observations. Discussing the Limitations of the Assumptions used During Development. Alternatives for the Future Conclusions

3 ATLAS COPCO VEHICLES Mine Trucks Diesel Loaders

4 ATLAS COPCO VEHICLES Drilling Rigs Rock Reinforcement

5 LATELY DEVELOPED MINE TRUCK 65 ton Mine Truck Delivered Prototype Prototype 6 months after delivery

6 CONTENT Atlas Copco Vehicles The Models Used During Development. The Methodology Used During Development. Measurements Included When Validating the Vehicle Behavior. Comparing the Numerical Results with the Measured Observations. Discussing the Limitations of the Assumptions used During Development. Alternatives for the Future Conclusions

7 MODELS USED DURING DEVELOPMENT

8 ANALYSIS WORK PERFORMED DURING DEVELOPMENT OF THE 65 TON MINE TRUCK. Numerical and Experimental Analysis during MT65 development Dynamics Linear Static Analysis +Calculation shows that selected design was OK! Engine induced vibrations Vehicle Dynamics (needed) LC1 Sideway Towing LC2 LC.. +Tests in show that this is verified! Modal analysis Harmonic response analysis Not included in our analysis processes Tests showed that these results have underestimated the behavior in several regions.

9 TEST V&V What did we want to resolve with the test program: 1. Even for the design selected for the prototype the analysis showed weak regions. The stress levels were well above the acceptance levels. The questions to be answered were: Are these as severe as indicated? 2. How well does the current design specification ( also called design criteria) describe what is going on when the vehicle is used? 3. How important is it to consider driving induced dynamics when designing the vehicle?

10 CONTENT Atlas Copco Vehicles The Models Used During Development. The Methodology Used During Development. Measurements Included When Validating the Vehicle Behavior. Comparing the Numerical Results with the Measured Observations. Discussing the Limitations of the Assumptions used During Development. Alternatives for the Future Conclusions

11 MINE TRUCK ON TEST TRACK 65 ton Mine Truck Driving on Test Track

12 MEASUREMENTS INCLUDED WHEN VALIDATING THE VEHICLE BEHAVIOR. Measured data are evaluated against following load cases: Fatigue Life Evaluation when Driving the Vehicle with and without Load in the Dumpbox Verification of FE-model behavior Steering with Articulated Vehicle Sideway Towing Dumpbox (=DB) Push by Loader Speed Braking

13 FATIGUE LIFE EVALUATION In this presentation it will be shown that In 40 % of all measured locations, the model and upfront FE-analysis performed showed good agreement. In the remaining 60 % measured locations, the model and FE-analysis has strongly underestimated the stress levels not due to model limitations but due to too limited analysis procedures (only considering statics). Part 2, of this presentation will show that the reason for the underestimations are that vehicle dynamics is not considered during the FE-analysis according to the current design specification. Note: Adding/Removing static analysis load cases, and/or alter stress level criteria, or introduce new evaluation methods will not help to resolve this. We need to add more physics to improve this.

14 MT65 EVALUATION LOCATIONS OF STRAIN GAUGE'S AND ACCELEROMETER DURING THE TESTS.

15 STRAIN GAUGE LOCATIONS, 1

16 STRAIN GAUGE LOCATIONS, 2

17 ACCELEROMETER LOCATIONS, 3.X Load frame Waist Power Frame Above front axle, Power Frame Acc., 3.1, 3.2,3.3 (x,y,z) Rear axle, Power Frame Acc., 3.4, 3.5,3.6 (x,y,z) Upper Waist, Power Frame Acc., 3.7,3.8 (x,z)

18 DRIVING 20 MINUTES ON THE TEST TRACK The next few pages give examples of different types of sensor data that were stored during the tests meas mat 30 Test Track Drive Speed Monitoring km/h Vehicle SPEED ( max/min: 36.6/-0.0 ) time[s] meas mat Speed km/h km/h GPS-SPEED ( max/min: 35.0/0.0 ) time[s] meas mat 50 ustrain 0-50 Flat Track Drive Speed Monitoring QB ( max/min: 99.4/ ) time[s] meas mat ustrain QB ( max/min: 108.3/-91.9 ) time[s]

19 FATIGUE LIFE EVALUATION On the next pages measured stress at strain gauge locations are presented. Reported How have these stresses been calculated? Answer: See following example Evaluations of stress during driving (example point 2.1, test 32 ): We know that for a plate (cut edge) to have a fatigue life of 10,000 h, subjected to 400,000 cycles, the stress level need to be below 150 MPa ( for a BSK Whöler curve with slope of 3). In test 32 we calculated the cubic mean of the strain in a A-frame dog-bone (2.1) to 31 µs and this appears 9412 cycles during 0.32 hours. Assuming that this is one typical work cycle we can now normalize this to the strain it would correspond to if measurements were made during 10,000 hours of work. Application of above gives: Eps(mean) = 31*(10,000*9412/400,000*0.32)^1/3 = 280 µs or about 60 MPa (for steel), hence below 150 MPa and OK!

20 FATIGUE LIFE EVALUATION Let s look at the results! The results will be given in a table. Each sensor result have also been evaluated individually.

21 FATIGUE LIFE EVALUATION Accelerometer data (expressed in G ) This table illustrates the multiaxis type loading that appear when driving. This in contrast to the 1D load used during an analysis. No Dumpbox Load Full Dumpbox Point Test 06 Test 09 Test 33 Test 35 Part PF-x Specified Load in FE-analysis PF-y PF-z 3.0 G LF-x LF-y LF-z 3.0 G Test track, 20 minutes On flat ground, 20 minutes The table shows cubic mean acceleration levels.

22 FATIGUE LIFE EVALUATION Qubic Mean Stress in MPa Empty Dumpbox Full Dumpbox FE-calculations Point Test 06 Test 07 Test 09 Test 32 Test 33 Test 34 Test 35 LC15 Part PF-bone PF (vert) (73) LF LF-cent PF-cab LF-rear (48) D-box_c (DB Load scaled to 57 ton load) PF-left D-box_r Dynamic amplification (40) D-box_f LF-front D-box_s Hydro static pressure

23 FATIGUE EVALUATION + It is assumed that a work cycle is 1 Hour. + Dumpbox is assumed to be full 60 % of the time and 40 % empty. + We use BSK Whöler curves as material data ( C=90, value base on that stress appear at thermally cut edges.) + Palmgren-Mainer cumulative damage approach is used for fatigue analysis.

24 ANALYSIS MODEL, SIDEWAY LOADING Problem regions

25 MT65 ANALYSIS MODEL Below Cabin (Scale=1.96), 2.3 % prob. of failure BSK 1. 1,20E+01 1,00E+01 8,00E+00 6,00E+00 4,00E+00 2,00E+00 Empty Dumpbox Cumulative Fatigue Damage, Empty on test track + Loaded on flat ground: Fatigue Life Approx.= 83 Hours 2. Empty on test track + Loaded on test track: Fatigue Life Approx.= 85 Hours (195) 3. Empty on flat ground + Loaded on flat ground: Fatigue Life Approx.= 96 Hours ( ) 50 % prob. of failure value!!!! 0,00E+00 0,00E+00 1,00E+08 2,00E+08 3,00E+08 4,00E+08 5,00E+08-2,00E+00 1,00E+01 9,00E+00 8,00E+00 7,00E+00 6,00E+00 5,00E+00 4,00E+00 3,00E+00 2,00E+00 Full Dumpbox Cumulative Fatigue Damage, 2.9f 1,00E+00 0,00E+00 0,00E+00 1,00E+08 2,00E+08 3,00E+08 4,00E+08 5,00E+08

26 MT65 ANALYSIS MODEL LC15 Area 2.5 (Scale=2.31), 2.3 % prob. of failure BSK 1. 7,00E-02 Empty dumpbox Cumulative Fatigue Damage, 2.5 6,00E-02 5,00E-02 4,00E-02 3,00E-02 2,00E-02 1,00E-02 0,00E+00 0,00E+00 2,00E+07 4,00E+07 6,00E+07 8,00E+07 1,00E+08 Full dumpbox Cumulative Fatigue Damage, 2.5f 1. Empty on test track + Loaded on flat ground: Fatigue Life Approx.= Hours 2. Empty on test track + Loaded on test track: Fatigue Life Approx.= Hours (35200) 3. Empty on flat ground + Loaded on flat ground: Fatigue Life Approx.= Hours ( ) 50 % prob. of failure value!!!! 9,00E-02 8,00E-02 7,00E-02 6,00E-02 5,00E-02 4,00E-02 3,00E-02 2,00E-02 1,00E-02 0,00E+00-1,00E-02 0,00E+00 5,00E+07 1,00E+08 1,50E+08

27 MT65 FATIGUE LIFE A.F.O SPEED Load Frame Fatigue Life a.f.o Drive Speed Power Frame Fatigue Life a.f.o Drive Speed hours, 72 ton (9200, 57) hours, 72 ton (500, 57) hours, 72 ton (1550, 57)

28 Fatigue Life Based on Usage in Mine Operations

29 FATIGUE LIFE BASED ON OPERATIONS IN IRELAND Measurements between 10/4 2/5 Estimated Fatigue Life Base on Daily Usage Life 2.13 Life 2.9 Life 2.7 Life 2.5 Poly. (Life 2.5) Note: At the later parts of the test period results are close to the levels found during the Kvarntorp tests.

30 What Causing the Difference in Calculated and Measured Behavior?

31 POWER FRAME ACCELERATION, , DURING 20 MINUTES DRIVING Several Dynamic amplifications due to excitation of structural modes or tones The PSD Sideway Acceleration Spectrum Power Frame 3.2 Power Frame Y Power Frame 3.3 Power Frame Z Acc.location ( ) Normalized Power Spectral Density, PSD in (g) 2 /Hz 10-4 Normalized Power Spectral Density, PSD in (g) 2 /Hz f(hz) The PSD Vertical Acceleration Spectrum Hz f(hz) 50 Hz

32 POWER FRAME DYNAMIC AMPLIFICATION SIDEWAY BENDING of POWER FRAME

33 POWER FRAME DYNAMIC AMPLIFICATION AROUND 50 HZ (DUMPBOX UNSELECTED)

34 Dynamic amplification MT65, ANALYSIS MODEL, Experimental response PSD at 2.19 (dumpbox side) Normalized Power Spectral Density, PSD in (Unit) 2 /Hz QB f(hz) This is the 20 Hz excitation mode. Normalized Power Spectral Density, PSD in (Unit) 2 /Hz 3.1 Power Frame Power Frame X f(hz) Deformation amplitude contours when the mode is excited. Stress field induced when the mode is excited.

35 Dynamic amplification MT65, ANALYSIS MODEL, Experimental response PSD at 2.19 (dumpbox side) Normalized Power Spectral Density, PSD in (Unit) 2 /Hz QB f(hz) This is the 40 Hz excitation mode. Stress field induced when the mode is excited. Normalized Power Spectral Density, PSD in (Unit) 2 /Hz Power Frame 3.1 Power Frame X f(hz) The scale stress field of the 42 Hz dumpbox mode shows that a 14 mm vibration amplitude generates 130 MPa stress level in the center of the plate.

36 CONCLUSIONS What can we do to improve the Mine Trucks today? Used available spectrums as basis when evaluating design changes to remove/limit dynamic amplification problems experienced. This enable shifting dynamic properties in the right directing and at the same time decreasing or maintain weight. With the goal to achieve acceptable fatigue life.

37 VISION ON VEHICLE DEVELOPMENT AT ATLAS COPCO Quasi-Static Analysis This iterative effort Decrease will with deeper Insight Created by Dynamic Analysis!!!! Engine Induced Vibrations Modal Analysis Response Analysis Driving & Operations Induced Vibrations Modal Analysis PSD Spectrum Analysis FE-modelling MT65 Develop Loaders and Trucks CFD Fluid dynamics for cooling managment 73 ton vehicle Prototype Field Evaluation Load Spectrum Statistics Prototype Analysis/Evaluation Engine Induced Vibrations Test Track Lab. Testing Modal Analysis, Material & Component Testing PSD Spectrum Excitation (Time Compressed Investigation of Vehicle/Component Fatigue Life)

38