Keywords: processor farming, parallel computing, heat and moisture transfer, homogenization, stone masonry bridge.

This contribution presents a processor farming method in connection with a multiscale analysis. In the this method, each macro-scopic integration point or each finite element is connected with a certain meso-scopic problem represented by an appropriate representative volume element (RVE). The solution of a meso-scale problem then provides effective parameters needed on the macro-scale. Such an analysis is suitable for parallel computing because the meso-scale problems can be distributed among many processors. The processor farming method connected with a multi-scale analysis differs from classical domain decompositions. The macro-problem is assigned to the master processor while the solution at the meso-level is carried out on slave processors. At each time step the current temperature and moisture together with the increments of their gradients at a given macro-scopic integration point are passed to the slave processor (imposed onto the representative volume element), which, upon completing the small scale analysis, sends the homogenized data (effective conductivities, averaged storage terms and fluxes) back to the master processor. The application of the processor farming method to a real world masonry structure is illustrated by an analysis of Charles bridge in Prague. The three-dimensional numerical model simulates the coupled heat and moisture transfer of one half of arch No. 3. and it is a part of a complex hygro-thermo-mechanical analysis which has been developed to determine the influence of climatic loading on the current state of the bridge.

purchase the full-text of this paper (price £22)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £45 +P&P)