2021•Unpublished venueRequires access

Vibration Behavior of Powertrain Test Benches - Measurement, Analysis and Modelling

Alfons R. Wagner, Erwin Brosch

Open publisher page 2 citations

Abstract

Electric powertrains have unique characteristics, which imply new requirements and need additional testing methods within the powertrain test benches. This work presents a unique vibration model of a complete powertrain test bench based on both measurements and theoretical analysis of a real hardware system. The modelling method is a contribution towards understanding and analysing the new vibration behaviour of the electric powertrain on powertrain test benches, which refers to a shift towards higher frequencies. The procedure, which is developed in this work is based on a model of the physical vibration path of a driving machine within a test bench. The model is created and validated based on measurement data that are obtained from several tests then the overall model of the vibration of the complete test bench is constructed. Based on the entire mechanical system of the test bench, a simplified physical model of an individual machine within the test bench is first created in Matlab Simulink. Then, the individual models are combined together to form an overall system model including coupling parameters. The simplified model is constructed as an oscillatory multi- body mass spring damper system composed of several individual components. Special attention is taken when modelling the linking conditions in order to obtain realistic results from the overall model. The parameters existing in the model are determined either empirically from test data or theoretically with the help of the test bench documentation. In order to verify the individual model, several tests are carried out. For this purpose, high-resolution vibration measurement technology is applied to an existing test bench machine. In addition to that, vibration sensors are installed on the machine base and bed. Different speeds are approached and measurement data are recorded under steady state conditions in addition to dynamic tests. These tests are carried out under no load conditions. The test bench machine is also operated with and without explicit imbalance excitation and the relevant measurements are recorded synchronously at high frequency of 1 kHz. During verification procedure of the individual components, the model is compared with the measurement data and the deviations are iteratively eliminated by adapting the parameters used in the previous iteration. This procedure is repeated until the model matches the measurement data of the real test bench machine and the error is within a specified tolerance band. Finally, the measurements are repeated for the entire test bench and the overall model is validated. Based on a simplified physical model of a test bench machine the vibration behaviour of the entire test bench is modelled using measurement data of no-load tests at several speed with and without explicit imbalance excitation. Studying the vibration behaviour of powertrain test benches allows the testing of electric and electrified powertrains safely on existing test benches.

About this research paper

What this paper is about

Electric powertrains have unique characteristics, which imply new requirements and need additional testing methods within the powertrain test benches. This work presents a unique vibration model of a complete powertrain test bench based on both measurements and theoretical analysis of a real hardware system. The modelling method is a contribution towards understanding and analysing the new vibration behaviour of the electric powertrain on powertrain test benches, which refers to a shift towards higher frequencies. The procedure, which is developed in this work is based on a model of the physical vibration path of a driving machine within a test bench. The model is created and validated based on measurement data that are obtained from several tests then the overall model of the vibration of the complete test bench is constructed. Based on the entire mechanical system of the test bench, a simplified physical model of an individual machine within the test bench is first created in Matlab Simulink. Then, the individual models are combined together to form an overall system model including coupling parameters. The simplified model is constructed as an oscillatory multi- body mass spring damper system composed of several individual components. Special attention is taken when modelling the linking conditions in order to obtain realistic results from the overall model. The parameters existing in the model are determined either empirically from test data or theoretically with the help of the test bench documentation. In order to verify the individual model, several tests are carried out. For this purpose, high-resolution vibration measurement technology is applied to an existing test bench machine. In addition to that, vibration sensors are installed on the machine base and bed. Different speeds are approached and measurement data are recorded under steady state conditions in addition to dynamic tests. These tests are carried out under no load conditions. The test bench machine is also operated with and without explicit imbalance excitation and the relevant measurements are recorded synchronously at high frequency of 1 kHz. During verification procedure of the individual components, the model is compared with the measurement data and the deviations are iteratively eliminated by adapting the parameters used in the previous iteration. This procedure is repeated until the model matches the measurement data of the real test bench machine and the error is within a specified tolerance band. Finally, the measurements are repeated for the entire test bench and the overall model is validated. Based on a simplified physical model of a test bench machine the vibration behaviour of the entire test bench is modelled using measurement data of no-load tests at several speed with and without explicit imbalance excitation. Studying the vibration behaviour of powertrain test benches allows the testing of electric and electrified powertrains safely on existing test benches.

Why it matters

OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Electric powertrains have unique characteristics, which imply new requirements and need additional testing methods within the powertrain test benches. This work presents a unique vibration model of a complete powertrain test bench based on both measurements and theoretical analysis of a real hardware system. The modelling method is a contribution towards understanding and analysing the new vibration behaviour of the electric powertrain on powertrain test benches, which refers to a shift towards higher frequencies. The procedure, which is developed in this work is based on a model of the physical vibration path of a driving machine within a test bench. The model is created and validated based on measurement data that are obtained from several tests then the overall model of the vibration of the complete test bench is constructed. Based on the entire mechanical system of the test bench, a simplified physical model of an individual machine within the test bench is first created in Matlab Simulink. Then, the individual models are combined together to form an overall system model including coupling parameters. The simplified model is constructed as an oscillatory multi- body mass spring damper system composed of several individual components. Special attention is taken when modelling the linking conditions in order to obtain realistic results from the overall model. The parameters existing in the model are determined either empirically from test data or theoretically with the help of the test bench documentation. In order to verify the individual model, several tests are carried out. For this purpose, high-resolution vibration measurement technology is applied to an existing test bench machine. In addition to that, vibration sensors are installed on the machine base and bed. Different speeds are approached and measurement data are recorded under steady state conditions in addition to dynamic tests. These tests are carried out under no load conditions. The test bench machine is also operated with and without explicit imbalance excitation and the relevant measurements are recorded synchronously at high frequency of 1 kHz. During verification procedure of the individual components, the model is compared with the measurement data and the deviations are iteratively eliminated by adapting the parameters used in the previous iteration. This procedure is repeated until the model matches the measurement data of the real test bench machine and the error is within a specified tolerance band. Finally, the measurements are repeated for the entire test bench and the overall model is validated. Based on a simplified physical model of a test bench machine the vibration behaviour of the entire test bench is modelled using measurement data of no-load tests at several speed with and without explicit imbalance excitation. Studying the vibration behaviour of powertrain test benches allows the testing of electric and electrified powertrains safely on existing test benches.

Key concepts: Powertrain, Test bench, Vibration, MATLAB, Test data, Engineering, Computer science, Automotive engineering

Related papers

Back to paper searchBrowse research topicsOriginal source
Vibration Behavior of Powertrain Test Benches - Measurement, Analysis and Modelling — Research Paper | ScholarLens