Composite action in masonry walls under vertical in-plane loading : experimental results compared with predictions

A.T. Vermeltfoort, D.R.W. Martens

Research output: Contribution to journalArticleAcademicpeer-review

5 Citations (Scopus)
264 Downloads (Pure)

Abstract

In this article the results of five experimental test series on masonry walls supported by reinforced concrete beams or slabs are reported and compared to theoretical predictions of the load-bearing capacity. The experiments were performed on deep masonry beams built with respectively calcium silicate and clay brick. Investigated parameters were: position of the supports, concrete beam-masonry interface, concrete beam stiffness, type of loading and height of masonry wall and concrete beam. Based on literature, the method proposed by Davies and Ahmed (D&A) as well as the method according to Eurocode 6 were used to estimate the load bearing capacity of the tested masonry walls supported by concrete beams. The method of Davies and Ahmed allows for the determination of the stresses and stress resultants in the masonry. The analysis shows that near the support an inclined compressive force acts at the bed joint which means that a shear-compression stress state exists in the bed joint. Strength evaluation has been carried out using the Mann-Müller criterion which is adopted in Eurocode 6. Based on the test results, it may be concluded that both methods yield conservative values of the load-bearing capacity, as could be expected. Before cracking a linear elastic behavior was observed, while after cracking a strut-and-tie model may be applied. In order to develop more accurate design models, it is recommended to investigate the post-cracking behavior in more detail.
Original languageEnglish
Pages (from-to)449-462
Number of pages14
JournalCanadian Journal of Civil Engineering
Volume42
DOIs
Publication statusPublished - 2015

Fingerprint

Dive into the research topics of 'Composite action in masonry walls under vertical in-plane loading : experimental results compared with predictions'. Together they form a unique fingerprint.

Cite this