Skip to Main content Skip to Navigation

Diagnostic de fuites internes dans une pompe à pistons axiaux

Abstract : The work presented in this thesis is intended for the fault diagnosis of axial piston pumps. In this context, we propose different methods able to diagnose internal leakages in an axial piston pump, in particular those caused by one or several worn pistons. We begin this manuscript by recalling the fundamental concepts related to hydrostatics and explaining the working principle of axial piston pumps. After that, we discuss the modeling and simulation of the behavior of this type of pump. This step is essential to understand the variation of characteristic variables of the pump (pressure and flow rate). Next, we compare the simulated pressure signals with those acquired on an experimental test bench. This is intended to demonstrate the robustness of the used modeling and to highlight cases where the modeling differs from reality. Finally, we propose three diagnostic methods based on different approaches. On the one hand, the first two methods aim to identify the faulty piston when there is a piston leak. These are model-based methods that use the estimation of interest parameters to make the diagnosis decision. On the other hand, the last method is data oriented. It uses the collected data for various health conditions and under different operating conditions (speed and load) in order to discriminate several classes. Each class corresponds to one health condition. This method offers the possibility to distinguish the most robust indicator that allows piston leak diagnosis, regardless of the operating condition.
Complete list of metadatas

Cited literature [123 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Wednesday, June 24, 2020 - 12:14:12 PM
Last modification on : Friday, October 23, 2020 - 5:04:44 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02879843, version 1



Djihed Bensaad. Diagnostic de fuites internes dans une pompe à pistons axiaux. Génie mécanique [physics.class-ph]. Université de Lyon, 2019. Français. ⟨NNT : 2019LYSES020⟩. ⟨tel-02879843⟩



Record views


Files downloads