Search Results

The usage of multi-body dynamics tools for the prediction of vehicle system/sub-system loads, has significantly reduced the need to measure vehicle loads at proving grounds. The success of these tools is limited by the quality of the digital representations being used to simulate the physical test roads. The development of these digital roads is not a trivial task due to the large quantity of data and processing required. In the end, the files must be manageable in size, have a globally common format, and be simulation-friendly. The authors present a methodology for the development of high quality 3-dimensional (3-D) digital proving ground profiles. These profiles will be used in conjunction with a multi-body dynamics software package (ADAMS) and the FTire™ model. The authors present a case study below.

Measured vehicle loads have traditionally been used as the basis for development of component, subsystem and vehicle level durability tests. The use of measured loads posed challenges due to the availability of representative hardware, scheduling, and other factors. In addition, stress was placed on existing procedures and methods by aggressive product development timing, variety in tuning and equipment packages, and higher levels of design optimization. To meet these challenges, General Motors developed new processes and technical competencies which enabled the direct substitution of analytically synthesized loads for measured data. This process of Virtual Road Load Data Acquisition (vRLDA) enabled (a) conformance to shortened product development cycles, (b) greater consistency between design targets and validation requirements, and (c) more comprehensive data.

Automotive engineering development processes are growing more dependent on the use of multi-body dynamic (MBD) models for generating vehicle loads that at one time could only be measured using physical hardware. A certain technique combines these two approaches using a minimal set of physical measurements to excite a vehicle MBD model for predicting loads at various vehicle interfaces. This approach eliminates the use of a tire model, often the roadblock in MBD-based loads prediction simulations. However, for various reasons, the direct application of loads to a model can lead to problems with the simulation. Alternatively, the model can be artificially constrained but this also has its disadvantages. The purpose of this paper is to present a loads prediction technique that relaxes the use of artificial boundary conditions for applications involving the input of measurements to an MBD model.