SEDL / STP / STP1203-EB / STP14899S

Statistical Analysis of Profilometric Sampling for Roughness Parameters

Drury, WJ
Graduate Research Assistant and Professor, Georgia Institute of Technology, School of Materials Engineering, Atlanta, GA

Gokhale, AM
Graduate Research Assistant and Professor, Georgia Institute of Technology, School of Materials Engineering, Atlanta, GA

Pages: 13    Published: Jan 1993

Download this paper for $25 PDF (308K)          View License Agreement
Abstract

In quantitative fractography, roughness parameters RL and RS are often used as important descriptors of fracture surface geometry. These parameters are usually estimated by analyzing fracture profiles acquired by sampling the surface of interest with metallographic sections perpendicular to the global plane of fracture. Because of the limited sampling, there is an inherent statistical error to the measured RL and RS values. This work addresses the determination of this sampling error using basic statistical concepts. In addition, a simple parameter is introduced that facilitates estimation of the amount of profile data required to achieve a specified precision to the roughness parameters. Experimental results of sampling error and effort are provided from ‘typical’ steel (AISI 4340) fracture surfaces.