IMPROVING THE IMPACT RESISTANCE OF MASONRY PARAPETS

There are over 60000 masonry bridge parapets in the UK. Whereas steel and concrete parapets are well covered by design standards masonry parapets are not. Initially LS-DYNA was used to model vehicle impacts on various forms of masonry parapet. The results were encouraging, with good qualitative agreement between tests and analysis. The same finite element modelling strategy was subsequently used to help develop a guide for assessment (County Surveyors’ Society Guidance Note, 1995) and a new British Standard (BS 6779 pt 4, 1999). The findings of the initial study demonstrated that unreinforced masonry could perform reasonably well in preventing vehicle penetration and controlling the rebound of impacting vehicles. However the work highlighted two areas that required additional study: (i) in critical situations requiring improved impact resistance, advice on alternative strengthening methods or new build options was needed; (ii) the discrete analysis approach adopted, using LS-DYNA tied interface type 9, required use of unrealistic failure parameters in order to prevent a premature (brittle) failure. The current project is addressing these two issues by the use of both physical tests and finite element analysis. The test work involves a range of small and medium scale tests that provide data for the numerical work. These tests address the issue of dynamic enhancement of the shear and tensile strength of masonry. In addition, the effects of fracture energy and dilatancy under high strain rates are being considered. Test methods have also been developed to evaluate the interaction between the masonry and reinforcement under varying strain rates. Full-scale tests are also being used to both provide analytical data in the early stages of the project and to validate the use of LS-DYNA as a predictive tool in the latter stages of the project. The fullscale test walls range from approximately 9m to 20m in length. In the numerical models the existing LS-DYNA interfaces are being modified to incorporate the effects of fracture energy and dilatancy. Early results are promising, allowing realistic material properties to be used and seeming to explain the apparent high strain rate sensitivity of the measured data. Strategies for modelling reinforcement in a masonry wall are also being developed. Related work where LS-DYNA is being used to model masonry arch bridges is also summarised within this paper.