On the Validation of Cross-Wind Calculation Models for Railway vehicles Carsten Bußmann, Bombardier Transportation (Deutschland GmbH), Hennigsdorf SIMPACK User Meeting, May 2011, Salzburg
Introduction - cross wind calculations according to RIL 807 - for full vehicle since 4-car EMU with Jakobs bogies - set up of 4-car SIMPACK- model - verification of model by comparing calculation results with - measured sway coefficient from - scale - verification problems with sway test - measured wheel unloading from - test runs - scale - verification sucessful - verified model serves as basis for cross wind calculations 2
Cross wind calculations according to RIL 807 vehicle for v max = 160 km/h, conventional, non tilting RIL 807, Class D EMU with Jakobs bogies full vehicle for calculation 3
CWC reference for class D vehicle acc. to RIL 807 Characteristic Wind Curves reference vehicle class D max. wind speed in m/s 35 aq= 0,0 m/s² reference vehicle RIL class D aq= 1,0 m/s² reference vehicle RIL class D 33 30 28 25 23 20 120 130 140 150 160 running speed km/h CWC for reference vehicle RIL 807, Class D 4
SIMPACK calculation model: Full vehicle MBS- Vehicle model including - four car bodies with articulation - two conventional end bogies - three Jakobs bogies 5
SIMPACK calculation model: Bogies MBS- Bogie models - two point air spring leveling + anti roll bar - nonlinear lateral bump stop characteristic - vertical secondary spring limitation by emergency springs and lift stop 6
Cross wind forces according to RIL 807 Chinese Hat wind scenario acc. to RIL 807 applied with time shift on different car bodies 7
Model adaption to test vehicle: static wheel loads model adapted to measured static wheel loads from test runs on straight track by manipulating mass and center of gravity data 8
Model adaption to test vehicle: sway coefficient sway coefficient 0.14 0.12 0.10 0.08 0.06 0.04 Measurement half and full 4-car unit (anti roll bar stiffness adapted for calculation, EBG +11,7 % / JBG - 9,2 % ) 4-car unit 1 T2 and M2 4-car unit 2 T3 and M3 DB-Measurement forward DB-Measurement reverse 4-car unit calculation 0.02 0.00 T2BG V T2 BG IV M2 BG IV M2 BG III BG V BG IV BG III MBS- Model adapted to sway test results by manipulation of anti roll bar stiffness 9
Calculation of CWC according to RIL 807 Characteristic Wind Curves 4-car unit 45 4 car unit aq= 0,0 m/s² max. wind speed in m/s 40 35 30 25 4 car unit aq= 1,0 m/s² aq= 0,0 m/s² reference vehicle RIL class D aq= 1,0 m/s² reference vehicle RIL class D 20 80 90 100 110 120 130 140 150 160 running speed in km/h CWC show values worse than expected homologation critical 10
Model verification: wheel unloading on the scale Wheel unloading Comparison Scale Measurements, Bogie 1, Wheel Set 1 Bogie 1 70 8000 Wheel load Q in kn 65 60 55 18110912 MOS 2.5 R11 18110916 MOS 2.5 R11 18110914 MOS 2.5 R12 18110918 MOS 2.5 R12 2110914 COT 2.3 R11 2110906 COT 2.3 R11 2110905 50 COT 2.3 R12 2110907 COT 2.3 R12 19110912 COT 2.2 R11 18110916 COT 2.2 R11 19110910 COT 2.2 R12 19110914 COT 2.2 R12 10021003 NUE 4.X R12 9021004 NUE 4.X R12 45-1.25-1.00-0.75-0.50-0.25 0.00 0.25 0.50 0.75 1.00 1.25 lateral acceleration aq in m/s² W h eel load in kg 7500 7000 6500 6000 5500 5000 0.00 0.10 0.20 0.30 0.40 0.50 lateral acceleration in m/s² R12 R11 R22 R22 sway test with measurement of wheel unloading, - calculated wheel unloading wrong with anti roll bar stiffness adapted to reproduce sway coefficient - calculated wheel unloading o.k. with nominal anti roll bar stiffness 11
Explanation of nonlinear wheel unloading on the scale Wheel unloading Comparison Scale Measurements, Bogie 1, Wheel Set 1 70 65 Wheel load Q in kn 60 55 18110912 MOS 2.5 R11 18110916 MOS 2.5 R11 18110914 MOS 2.5 R12 18110918 MOS 2.5 R12 2110914 COT 2.3 R11 2110906 COT 2.3 R11 2110905 COT 2.3 R12 2110907 COT 2.3 R12 19110912 50 COT 2.2 R11 18110916 COT 2.2 R11 19110910 COT 2.2 R12 19110914 COT 2.2 R12 10021003 NUE 4.X R12 9021004 NUE 4.X R12 45-175 -125-75 -25 25 75 125 175 superelevation in mm roll angle 0.0 (0 mm) roll angle 1.5 (39.3mm) Application of cant leads to contact point jump between 40 and 65 mm due to scale measurement procedure. This suggests a nonlinear wheel unloading which does not occur during real test runs roll angle 2.5 (65.5mm) 12
Model verification: measured wheel unloading Wheel load changes Measurement test run, Bogie 1, Wheelset 1 Wheel load changes Measurment test run, Bogie 1, Wheelset 2 100 100 90 90 vertical wheel load Q in kn 80 70 60 50 40 30 20-1.5-1 -0.5 0 0.5 1 1.5 uncompansated lateral acceleration aq in m/s² Q11 curve right Q11 curve left Q12 curve right Q12 curve left Q11 CR regression test run Q11 CL regression test run Q12 CR regression test run Q12 CL regression test run Q11 CR - Sigma Q11 CL - Sigma vertical wheel load Q in kn 80 70 60 50 40 30 20-1.5-1 -0.5 0 0.5 1 1.5 uncompensated lateral acceleration aq in m/s² Q21 curve right Q21 curve left Q22 curve right Q22 curve left Q21 CR regression test run Q21 CL regression test run Q22 CR regression test run Q22 CL regression test run Q21 CR - Sigma Q21 CL - Sigma Q11 CR + Sigma Q11 CL + Sigma Q21 CR + Sigma Q21 CL + Sigma Q12 CR - Sigma Q12 CL - Sigma Q22 CR - Sigma Q22 CL - Sigma Q12 CR + Sigma Q12 CL + Sigma Q22 CR + Sigma Q22 CL + Sigma scale measurement Vers.: 9021004 Q12 scale measurement Vers.: 10021003 Q12 scale measurement Vers.: 9021004 Q22 scale measurement Vers.: 10021003 Q22 Wheel unloading from test runs coincides with scale results 13
Comparison of measured and calculated wheel unloading Wheel unloading in 400m-curve Comparison of test ride measurements and calculation Wheel unloading in 400m-curve Comparison of test ride measurements and calculation 110 110 100 100 vertical wheel load in kn 90 80 70 60 50 40 30 0 0.2 0.4 0.6 0.8 1 1.2 1.4 lateral acceleration aq in m/s² Q1i curve right Q1o curve left calculation 400m-curve right Q11 calculation 400m-curve left Q11 Q1o curve right Q1i curve left calculation 400m-curve right Q12 calculation 400m-curve left Q12 Radaufstandskraft q in kn 90 80 70 60 50 40 30 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Querbeschleunigung aq in m/s² Q2i curve right Q2o curve left calculation 400m-curve right Q21 calculation 400m-curve left Q21 Q2o curve right Q2i curve left calculation 400m-curve right Q22 calculation 400m-curve left Q22 Q1o regression test run (400m<R<600m) Q1i regression test run (400m<R<600m) Q2o regression test run (400m<R<600m) Q2i regression test run (400m<R<600m) Q1o - sigma Q1o + sigma Q2o - sigma Q2o + sigma Q1i - sigma Q1i + sigma Q2i - sigma Q2i + sigma scale measurement Vers.: 9021004 4T-NUE Q12 scale measurement Vers.: 10021003 4T-NUE Q12 scale measurement Vers.: 9021004 4T-NUE Q22 scale measurement Vers.: 10021003 4T-NUE Q22 Good agreement of wheel unloading results from calculations and test 14
Recalculation of CWC with verified model Characteristic Wind Curves 4-car unit "Wheel Unloading" 45 4 car unit wheel unload., aq= 0.0 m/s² 4 car unit wheel unload., aq= 1.0 m/s² max. wind speed in m/s 40 35 30 25 aq= 0,0 m/s² reference vehicle RIL class D aq= 1,0 m/s² reference vehicle RIL class D 20 80 90 100 110 120 130 140 150 160 running speed in km/h CWC for verified model show higher cross wind stability homologation no longer critical 15