The two truck models in this demo were prepared using several Nastran DMAP alters and Matlab functions available with TAMKIT. All Nastran DMAPs and Matlab functions described in the following five steps are available in TAMKIT. A summary of the process is highlighted below and detailed descriptions of each step follow. (All Nastran input decks are provided in the demo folder for reference.)
Select target modes for each FEM configuration—Calculates full FEM modes of each FEM configuration and writes mode shapes to OP4 using write_modes alter. Select full FEM target modes for each configuration. (Nastran decks truck_wload fem_modes.dat and truck_fem_modes.dat)
Select candidate set DOF for each FEM configuration—Determine the candidate set DOF for each FEM configuration using fast_iter_rke alter which writes candidate set of DOF to the punch file. Matlab function wrt_irke_aset.m is used to convert each punch file to Nastran ASET cards. (Nastran decks truck_wload_fem_irke.dat and truck_fem_irke.dat)
Merge candidate set DOF for each FEM configuration—Create a single candidate set for both FEM configurations that is the union of the two individual candidate sets, then intersected with each FEM’s DOF. (Nastran bulk files truck_wload_union_irke.aset and truck_union_irke.aset)
Validate candidate set DOF TAM—Check the accuracy of each FEM configuration’s candidate set TAM compared to the full FEM modes using the ortho alter. Matlab functions read_tam.m and rpt_tam.m are used to generate reports that document the accuracy of the candidate set TAMs for each configuration. (Nastran decks truck_wload_irke_tam.dat and truck_irke_tam.dat)
Extract matrices for Genetic Algorithm—Write out the characteristic matrices of the candidate set TAM for each FEM configuration. The output from these Nastran runs will be directly imported into the Genetic Algorithm for accelerometer placement. (Nastran decks truck_wload_irke_tam_out.dat and truck_irke_tam_out.dat)