Reliability Analysis of an Axial Compressor Based on One-Dimensional Flow Modeling and Survival Signature

Miro, S, Willeke, T, Broggi, M, Seume, JR and Beer, M ORCID: 0000-0002-0611-0345 (2019) Reliability Analysis of an Axial Compressor Based on One-Dimensional Flow Modeling and Survival Signature. ASCE-ASME JOURNAL OF RISK AND UNCERTAINTY IN ENGINEERING SYSTEMS PART B-MECHANICAL ENGINEERING, 5 (3).

Text
Miro_et_al_2019.pdf - Author Accepted Manuscript
Download (542kB) | Preview

Official URL: http://dx.doi.org/10.1115/1.4043150

Abstract

<jats:p>This paper presents a procedure for the reliability analysis of a multistage axial compressor regarding blade-specific roughness effects, based on the survival signature approach. As a result, a time-dependent evolution of the system reliability is obtained along with a prioritization technique for monitoring and regeneration of the rough blade rows by capturing the most critical system components. For this purpose, a one-dimensional flow model is developed and utilized to evaluate the aerodynamic influences of the blade-specific roughness on the system performance parameters, namely the overall pressure ratio and the isentropic efficiency. In order to achieve transparency and high numerical efficiency for time-dependent analyses in practice, the physics-based compressor model is translated into an illustrative, function-based system model. This system model is established by conducting a Monte Carlo simulation along with a variance-based global sensitivity analysis, with the input variables being the row-specific blade roughness. Based on the system model, the roughness impact in different blade-rows is ranked by the relative importance (RI) index, and the corresponding time-dependent reliability of the compressor system in terms of pressure ratio and efficiency is estimated through its survival function. Furthermore, uncertainties in the roughness-induced failure rates of the components are modeled using imprecise probabilities. Consequently, bounds on the reliability function and the importance indices for the blade-surface roughness in each blade row are captured, which enhances the decision-making process for maintenance activities under uncertainty.</jats:p>

Item Type:	Article
Depositing User:	Symplectic Admin
Date Deposited:	09 Jul 2019 14:22
Last Modified:	19 Jan 2023 00:38
DOI:	10.1115/1.4043150
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3048188