A method to develop a high-fidelity gearbox thermal model based on limited temperature measurements

Abstract

This paper presents a method to create a high-fidelity gearbox thermal model. A hybrid analytical-empirical model is presented, it is based on the thermal network approach and experiments are used to determine some unknown parameters of the model. A test campaign was led on a four stages helical gear unit and consisted of cooling measurements, no-load tests and loaded tests. These experiments were performed under different rotational speeds and loads. The results show that the developed thermal model provides an accurate prediction of the thermal behavior of different components of the gearbox (namely rolling element bearings and oil) by only using few thermal sensors. These results are corroborated by power losses measurements which demonstrate that the model can estimate the gear unit efficiency.

Keywords:

• Digital twin
• thermal network
• experiments

Acknowledgments

The authors gratefully acknowledge Association Nationale de la Recherche et de la Technologie (ANRT) (CIFRE No. 2020/1043). They also would like to express their gratitude to the CIRTrans (Consortium Industrie Recherche sur les Transmissions mécaniques) and especially to Andre Simonneau from TEXELIS and Loris Gimazane from ALSTOM.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

Table