Department of Civil & Environmental Engineering,
University of Bradford, Bradford, West Yorkshire, England. BD7 IDP
ABSTRACT
Many masonry structures resist applied loads and the effects of environmental change by complex interaction between the various structural elements and the supporting ground. Consequently, in order to obtain a realistic assessment of the in-service behaviour or strength of an existing masonry structure, it is generally necessary to use relatively sophisticated numerical methods of analysis. Any numerical or analytical model, however sophisticated, requires some form of constitutive model which contains a number of parameters (or phenomenological coefficients) to be found for each specific material.
Conventionally, the material parameters for a masonry constitutive models are determined directly from the results of tests on small assemblages or material samples in which it is usually assumed that the stress and strain fields in the test specimens are uniform. This assumption cannot be justified in the case of masonry which is an intrinsically inhomogeneous material. An alternative method of material parameter identification is proposed that better reflects the complex nature of masonry and the range of stress state types that exists in most masonry structures.
The method is based on the results obtained from tests on a series of larger specimens in which there is a variety of stress-state types. Assumed material parameters are initially used in a finite element simulation of the large-scale experiments. The differences between the results obtained experimentally and those predicted numerically are then minimized by adjusting the material parameters in the constitutive model using an advanced optimization technique.
1378.pdf
