Analysis of Various Parameters Responsible for Uncertainty in Measurements

Abstract

newline Measurement is the fundamental tool to control the quality parameters in any newlineengineering system. So obtaining the accurate results is very vital in every measurement newlinesystem. Since measurement of mechanical properties by using various material testing newlinemethods always possess some uncertainties therefore, it is in many cases expressed that newlinemeasurement result is finished just when joined by a quantitative explanation of its newlineuncertainty. Accurate measurement of mmechanical properties like compressive strength, newlinedynamic Young s modulus of elasticity, impact resistance, are required for effective newlinestructural analysis. Measurement of these properties shows a significant variation because newlineof several influencing factors which need to be taken into consideration for the evaluation newlineof uncertainty. In this research measurement uncertainty analysis for both destructive newlinetesting (Charpy impact test and Compression test,) and nondestructive testing (Ultrasonic newlinepulse velocity test and Rebound hammer test) is carried out. The various uncertainty newlineparameters for all the tests are identified and classified in two major categories as type A newlinestandard uncertainty and type B standard uncertainty. Type A standard uncertainty is newlinemainly due to repeatability of experimental observations in particular material testing newlinemethod. Type B standard uncertainty consist the effect of machine testing speed, newlinecalibration error, Dimensional variation in test specimen, supporting measurement newlineinstruments like vernier caliper, temperature, moisture content , operator bias etc. The newlinecombined effect of both type A and type B standard uncertainties with related coverage newlinefactor is presented in the form of expanded uncertainty. Monte Carlo simulation method newlineis used to study the propagation of uncertainties in all the tests and obtained results are newlinecompared with evaluated uncertainty values.