Research — Numerical Modelling
Our research in this area is strongly linked to our project work because we are so often asked to solve open-ended problems, or problems for which there is no established method of solution. A good example of a research project was the simulation of hydro-mechanical coupled processes to determine the effect of glacial advance and retreat on the performance of a deep geological repository for radioactive waste. Our modellers are continually required to push the limits of software and hardware. We exercise control over the models we run by understanding the limitations and strengths of the source code, and by maintaining the ability to adjust, debug, and improve such codes, in house. We maximize the benefit gained from a given code by developing necessary pre- and post-processing algorithms in house, typically as objects within mView. In the area of hardware, we maintain a stable of powerful computers whose capabilities are exploited to the best ability through batch processing, and parallelization.