Universiti Teknologi Malaysia Institutional Repository

Comparison of Weibull and normal probability distribution of flexural strength of dense and porous fired clay

Abubakar, M. and Mat Yajid, M. A. and Ahmad, N. (2016) Comparison of Weibull and normal probability distribution of flexural strength of dense and porous fired clay. Jurnal Teknologi, 78 (9). pp. 73-78. ISSN 0127-9696

[img]
Preview
PDF
792kB

Official URL: https://www.scopus.com/inward/record.uri?eid=2-s2....

Abstract

In this research, dense and porous fired clay were produced at a firing temperature of 1300°C. The flexural strength data of the dense and the porous fired clay were determined using three point bending test. Two-parameter Weibull and normal probability distributions were used to estimate the reliability of the flexural strength data of the dense and the porous fired clay. From the result, the Weibull probability distribution scale parameter for the dense (36.31MPa) and Porous (18.85MPa) fired clay are higher than the mean strength value for the dense (33.84MPa) and the porous (17.87MPa) of the normal distribution. Distributions of flaws in the dense and the porous fired clay have a significant effect on the Weibull and normal distribution parameters. The fractured surface of the dense fired clay shows a random distribution of cracks while that of the porous fired clay shows a distribution of pores in the morphology. The normal distribution considers failure at 50% of the flexural strength data while Weibull probability distribution is failure at 62.3% of the strength data. Therefore, two-parameter Weibull is the suitable tool to model failure strength data of the dense and porous fired clay.

Item Type:Article
Uncontrolled Keywords:Fired clay, Flexural strength, Normal distribution, Weibull distribution
Subjects:T Technology > TJ Mechanical engineering and machinery
Divisions:Mechanical Engineering
ID Code:74497
Deposited By: Widya Wahid
Deposited On:28 Nov 2017 07:42
Last Modified:28 Nov 2017 07:42

Repository Staff Only: item control page