A boundary element method (BEM) for large-scale acoustic analysis is accelerated efficiently and precisely with Graphics Processing Units (GPUs). Based on Burton-Miller boundary integral equation, an implementation scheme that can be handled efficiently in GPU is derived and applied to accelerate conventional BEM. Data caching techniques in GPU are introduced to improve efficiency of the prototype algorithm. A double-single precision algorithm implemented with single-precision floating-point numbers is employed to reduce numerical errors. It shows that the improved algorithm sustained a highest GPU efficiency of 89.8% for large-scale problems, and its accuracy was almost the same as that with double-precision numerals directly while costing only 1/28 in time and half in GPU memory consumption of the latter. The largest problem size up to 3 million unknowns was solved rapidly on a desktop PC (8GB RAM, NVIDIA GeForce 660 Ti) by the method. Its performance was better than the fast BEM algorithms in both time and memory consumption.

A multiscale algorithm with element/mesh momentum transfer of parameters and variable information exchange in coupling transition zone between discrete element method (DEM) and finite element method (FEM) is deduced.Multiscale computation of time and space with coupling of DEM and FEM is established.The algorithm and the model are applied to damage of a pre-tensioned aluminum plate under laser irradiation.Compared with results in FEM model,multiscale model of space and multiscale model of time and space,the multiscale model of time and space by coupling DEM and FEM is shown accurate and highly efficient.Damage of a pre-tensioned aluminum plate under laser irradiation is simulated in multiscale model of time and space from macro-and meso-scale.Simulated results are coincident with experimental resuts.

Recovery by equilibrium in patches (REP) technique is adopted for the non-linear transient heat conduction adaptive FEM analysis of casting system, which takes account of thermal conductivity of the interface. In order to test the present program, ANSYS and superconvergent patch recovery (SPR) methods are also employed for analysing the same problem respectively. The results of numerical analysis are compared with the experiment results, which show that REP method is cost effective.

A method of burnup calulation is introduced in calculation of reactor physic in this paper. Burnup calculation is completed by interpolation table of macroscopic cross section vs burnup, interpolated method and solving diffusion equation. The calculated result compared with shearon Harri s's result is satisfied. The method has characteristic of use friendly, less calculated time and occuracy demand. Now it is used in the practical calculation of reactor physic. A effective tool is provided for engineering design.