TY - GEN
T1 - DoE-gestützter vorderachsentwurf auf einem virtuellen Rollenprüfstand unter Berücksichtigung experimentell ermittelter Lenkungsübertragungseigenschaften
AU - Münster, M.
AU - Klose, G.
AU - Wiest, L.
AU - Lehner, M.
AU - Zimmermann, M.
AU - Bosbach, R.
AU - Rixen, D. J.
N1 - Publisher Copyright:
© VDI Verlag GmbH · Düsseldorf 2016.
PY - 2016
Y1 - 2016
N2 - This paper presents a method for designing front axles based on design-of-experiment (DoE) studies using a virtual roller test rig. In addition to standard axle characteristics resulting from kinematic and compliance analyses, dynamic properties of both the steering and the front axle are considered in order to ensure desired steering behavior of the vehicle. The dynamics of the steering subsystem is available from experimental testing, whereas the front axle dynamics is calculated by multi-body simulation. Both subsystems are assembled in the frequency domain to predict vehicle performance. Such a hybrid approach combines the trustworthiness of experimental testing with the high efficiency of simulation models. The extent of the model is limited to the front axle including rolling wheels in order to enable large scale Doe-studies within reasonable time. Even the steering gear is not part of the model, as the steering subsystem is assembled from measurement data afterwards. In order to still perform realistic driving maneuvers, a virtual roller test rig for use with multi-body simulation was developed. After completion of the DoE all the results are used for the training of metamodels, which enable the application of statistical methods to identify solution spaces for all the design parameters of the front axle. By using metamodels, high-dimensional analyses of over one hundred design variables are feasible, without having to dispense with the accuracy of a detailed MBS-model.
AB - This paper presents a method for designing front axles based on design-of-experiment (DoE) studies using a virtual roller test rig. In addition to standard axle characteristics resulting from kinematic and compliance analyses, dynamic properties of both the steering and the front axle are considered in order to ensure desired steering behavior of the vehicle. The dynamics of the steering subsystem is available from experimental testing, whereas the front axle dynamics is calculated by multi-body simulation. Both subsystems are assembled in the frequency domain to predict vehicle performance. Such a hybrid approach combines the trustworthiness of experimental testing with the high efficiency of simulation models. The extent of the model is limited to the front axle including rolling wheels in order to enable large scale Doe-studies within reasonable time. Even the steering gear is not part of the model, as the steering subsystem is assembled from measurement data afterwards. In order to still perform realistic driving maneuvers, a virtual roller test rig for use with multi-body simulation was developed. After completion of the DoE all the results are used for the training of metamodels, which enable the application of statistical methods to identify solution spaces for all the design parameters of the front axle. By using metamodels, high-dimensional analyses of over one hundred design variables are feasible, without having to dispense with the accuracy of a detailed MBS-model.
UR - http://www.scopus.com/inward/record.url?scp=85106050988&partnerID=8YFLogxK
M3 - Konferenzbeitrag
AN - SCOPUS:85106050988
SN - 9783180922591
SN - 9783180922706
SN - 9783180922713
SN - 9783180922720
SN - 9783180922737
SN - 9783180922744
SN - 9783180922751
SN - 9783180922768
SN - 9783180922775
SN - 9783180922782
SN - 9783180922799
SN - 9783180922805
SN - 9783180922812
SN - 9783180922829
SN - 9783180922843
SN - 9783180922850
SN - 9783180922867
SN - 9783180922874
SN - 9783180922881
SN - 9783180922898
SN - 9783182343424
SN - 9783182343431
SN - 9783182343455
T3 - VDI Berichte
SP - 439
EP - 454
BT - VDI Berichte
PB - VDI Verlag GMBH
T2 - 18th Congress on Simulation and Testing in Vehicle Development, 2016
Y2 - 22 November 2016 through 23 November 2016
ER -