Initial Experience for the Verification of Vehicle Models ... · Initial Experience for the...

Initial Experience for the Verification of Vehicle Models according to EN14363: 2016 Annex T

University of Applied Sciences BielefeldInstitute of System Dynamics and MechatronicsWorkgroup Railway Vehicles

2nd SIMULIA European Multibody Simulation UGM7th November 2017

Prof. Dr.-Ing. Rolf NaumannChristian Mallwitz, B.Eng.Sönke Lück, M.Sc.

Stephan Behringer, M.Sc.

DB Systemtechnik GmbHI.T-IVP 23 (1)Minden

2Verification of Vehicle Models According to EN14363

Prof. Dr.-Ing. Rolf Naumann, Christian Mallwitz, B.Eng.

7th November 2017 · Braunschweig2nd SIMULIA European MBS UGM

1. Introduction

2. Overview EN 14363 Annex T

3. MBS-Model of locomotive BR120

4. Validation Process

5. First Conclusion and Future Work

Content of presentation




1. Introduction

Simulation can often be used in the context of the homologation testing process

• Reduction of the number of test variants• Reduction of testing times,• Increasing the probability of successful proof,• Increasing the security of critical questions.

• Reduction of costs:

If the modelling process and the use of the simulation takes place in the design phase for other questions (crosswind, proof of derailment safety, comfort calculations, strength requirements, etc.), a reduction of the costs is to be expected.

If the modelling process and simulation is only for the proof of the testing a cost reduction is not necessarily to be expected

The biggest challenge is the validation of the simulation models




1. Introduction

• New validation method for simulation models for railway vehicles developed within the EU projectDynoTRAIN (2013)

• Models used in numerical simulations must be validated by comparison with on-track test results

Aspects of the EN14363 Annex T:

• Use of experimental data from the normal measuring method with measuring wheelsets

• Definition of the

• data to be compared

• minimum number of evaluation points

• data filtering

in accordance with the evaluation of the measured data.

• Calculation of the mean values and standard deviation of the differences between simulation and measurement

• Assessment by comparison with defined validation limits

• At least 3 sections for each of the 4 test sections (12 sections in total) should be examined• Length of sections of curves and straight track line must be at least equal to the measured

sections




1. Introduction

Validation should include the following:

• Sufficiently detailed modelling of relevant design data with regard to vehicle dynamics

• Test results for the modeled vehicle that can be used for model validation, including

• Time courses of measured data in digital form

• Tests and data should include a representative range of track conditions, curves, cant deficiency, speed and contact conditions

• Sufficiently exact track data of the original test track

If there is no complete data set available an estimation of the error influence must be drawn up and, if necessary, the scope of application is to be restricted!

When using numerical simulations for a vehicle under different conditions (eg. empty, laden, air spring operation, emergency spring operation, etc.), individual models must be validated for each condition.

Two validation methods are available




2. Overview EN 14363 Annex T

• EN 14363-2016 Annex T simulation of on-track tests

• Validation against static, quasi-static and dynamic tests

Method 1: (“previous conventional method”)

• Comparison of selected tests with simulation results

• Assessment by expert

Method 2: ("mathematical comparison")

• Simulation of defined configurations (tests) and evaluation with mathematical methods (max values, statistics)

• Comparison with tests

• No assessment by expert necessary




2. Overview EN 14363 Annex TSource: Polach, O.: Validierung der Simulationsmodelle für die fahrtechnische Zulassung [3]

Definition Validation: It has to be proven that the simulation model simulates the examined physical tests with sufficient accuracy.

Problematic: what does " sufficient accuracy " mean? How is this judged? Which methods are available?

Selection of a representative number of test configurations to test the entire scope of the simulation model.

Method 1: required and recommended tests defined (own decision)

Method 2: Test and simulations are precisely defined

Simulation of the test configurationsMethod 1: Recommended inputs and boundary conditions

Method 2: (sufficiently) precisely defined

Assessment of the comparison between measurement and simulation results

Method 1: Limits values for deviations defined for a small amount of values, otherwise not specified.

Method 2: Exactly describes the assessment method and limit values




2. Overview EN 14363 Annex T Method 2

Quantities to be assessed (Table T2):

Source [1]




2. Overview EN 14363 Annex T Method 2zu beurteilende Größen (Tabelle T2) Fortsetzung:

Source [1]





Evaluation of the relevant values for simulation SV

Measurement

sec. 1 sec. 2

Simulations

sec. 1 sec. 2

Evaluation of the relevant values for measurement MV

Source [3]





Multiplication with sign of MV

Berechnung Mittelwert und Standard-abweichung

Normalization with the validation limits

Calculation of differences DV

Source [3]





Validation result:

Standard deviation of all required valuesMean value of all required values

Successful validation if all values are less than 1!

Remark / Criticism of method 2:

• The dynamic behavior of the vehicle such as natural frequencies are not tested• Thus, an important property of the simulation model is not validated!

Source [3]




3.1 BR120 - Overview

History [6]:

- First mass-produced 3-phase current locomotive for DB

- First prototypes were produced in 1979

- Testing and proving finished in 1984

- Until 1989 DB received 60 BR120 from different suppliers

- Universal usage (passenger transport, heavy freight trains)

- Service perfomance in 2010: 17.000 km/month

Technical Data [7]:

- Length over buffer: 19.2 m

- Mass: 84 t

- Vmax: 200 km/h

- Tractive force at starting: 340 kN




3.2 BR120 - MBS-Model - Car Body

Wheelset 1,2Wheelset 3,4

Bogie 2 Bogie 1

Car Body




3.3 BR120 - MBS-Model - Bogie

Primary Suspension:

- Springs

- Damper

- Wheelbars

- Lift Lock / Fixed Stop

Secondary Suspension:

- Springs

- Damper

- Fixed Stop / Lateral Buffer

Traction Link:




3.4 BR120 - MBS-Model - Wheelsets

Wheelsets:

- Measured wheel profiles

- Measured rail profiles

- Left & right profile for every validation exercise

(averaged profiles)

- Rail track ballast force element

- Representing roadbed properties




4.1 Validation Exercise - Overview

Start: Uffenheim (95.3 km)

End: Ansbach (51.3 km)

Validation Exercises:

Exercise 1:

- Section of interest: 85.8 - 84.5 km

- Section with constant radius: 85.5 - 84.8 km

- Curve radius: 581 m

- Velocity: 110 km/h

- Test zone: 3

Source: Google Maps

3




4.2 Validation Exercise - Measured Data for Validation

BR120:

- Lateral & vertical wheel forces of bogie 1

- Lateral and vertical accelerations of car

body on top of bogie 1

Track:

- Track layout (.trm)

- Track irregularities (.tre)

- Averaged rail profiles (.prr)




4.3 Validation Exercise - Process & Results

Lts

Pre-processing / Simulation

Yjk Qjk ÿ*i z̈*i

Yjk Qjk ÿ*i z̈*i

Simulation Results

Test Run Results

Post-processing / Validation

Filter Low pass, Band pass, Sld mean

EvaluationMedian

0.15% / 99.85% PercentileRoot-Mean-Square

Defining Evaluation Section (Lts)

1 2 3

21 3




4.3 Validation Exercise - Process & Results

Difference SV, MV and

Transformation�� = (�� −��)

��

��

Mean Value and

Standard Deviation

Normalization, Final Verification

��,��23��,��

≤ 1;��,�.��.

��,��≤ 1

0

2

4

6

8

10

Section 1Section 2Section 3 Mean S. Dev.

Diffe

rence [

kN

]

Yqst

Right Wheel

Left Wheel

Mean

S. Dev.

1.84 0.56




5 Conclusion

- Model building process

� Technical drawings

� Data sheets

� First verification with stationary tests (no part of method 2)

o 150m curve

o Bogie rotation test

o Stationary vertical wheel forces

- First verification test according to method 2

� 581m curve

- The preparation of measuring data is complex (synchronization, resampling), an automatization of this process would be useful to increase efficiency

- First results are showing good coherence, adaption of model is still necessary

- Analysis of Different Validation Exercises from DynoTrain-Project

- Method 2 of EN14363 is suitable for automatization, if measured data is provided pre-processed

- The quality of verification, for using the model in framework of approval, can only be judged at the end of our project




Thank you for your attention!

Prof. Dr.-Ing. Rolf Naumann

Christian Mallwitz, B.Eng.

University of Applied Sciences Bielefeld

Institute of System Dynamics and Mechatronics

Workgroup Railway Vehicle

E-Mail: [email protected]

Interaktion 1

33619 Bielefeld

Germany




[1] EN14363:2016 Bahnanwendungen – Versuche und Simulationen für die Zulassung der fahrtechnischen Eigenschaften von Eisenbahnfahrzeugen – Fahrverhalten und stationäre Versuche; Deutsche Fassung EN 14363:2016

[2] EN14363:2013 Bahnanwendungen – Zulassungsprozess bezüglich der fahrtechnischen Eigenschaften von Eisenbahnfahrzeugen – Prüfung des Fahrverhaltens und stationäre Versuche; Deutsche Fassung prEn14363:2013

[3] Polach, O.: Validierung der Simulationsmodelle für die fahrtechnische Zulassung, aus: Fahrzeuge Rolling Stck, ZEVRail 140 (2016) 6-7 Juni-Juli, S.222 – 226.

[4] Mongiardini, M.; Malcom, H. Ray; Anghileri, M.: Development of a Software for the Comparison of CurvesDuring the Verification and Validation of Numerical Models, Italy, 7th European LS-Dyna Conference, 14th – 15th May, 2009, Salzburg.

[5] Polach, O. et. al.: Validation of simulation models in the context of railway vehicle acceptance, aus: Proceedings of the Institution of Mechanical Engineers / Part F, Journal of rail and rapid transit, Band 229, 6, S. 729 – 754, London, 2015.

[6] Kratochwille, R.: Description class BR120, DB-Systemtechnik GmbH, 2010.

[7] DB-Systemtechnik GmbH: Hauptdaten der Lok BR 120, interner Bericht.

Literature

Initial Experience for the Verification of Vehicle Models ... · Initial Experience for the...

Documents

Transcript of Initial Experience for the Verification of Vehicle Models ... · Initial Experience for the...