Categories
Technical Literature

Working wonders with APDL Math – Ep 02 : Thermal Harmonic Analysis

Summary: What did we learn this time?

What we have seen through this example is the usage of the linear solver commands (*LSENGINE,*LSFAC, *LSBAC) for obtaining frequency responses, which is a working tool for system analysis across many disciplines. We also learned how to dump the results into either ANSYS binary results (.RTH) or ASCII files, and I recommend doing both.

But there are some options for you to discover. In particular when tasks become computationally demanding, it is possible to have fine control over the solver:

  • For example, in case where independent excitations are present, there is a possibility to generate multiple RHS vectors, do a single factorization, and only repeat the computationally inexpensive back substitution (much alike the KUSE command). This is particularly useful when trying to determine which heat path is the most prevalent, or when searching for the optimal locations of actuators (always a hot debate).
  • Additionally, the recently introduced commands *LSDUMP/*LSRESTORE allows to dump the factorized linear solver system to a binary file, for later usage (to be honest I stumbled onto those while writing this post so I haven’t used them so far). So, once a large system has been factorized, it may be put aside and re-used, which is very convenient when all load cases are not known beforehand.

And finally, returning to the conclusion of my first post: the number of APDL Math commands is now 21 in lieu of 18, so keep looking for new possibilities!

References

[1] Brewer, Bryan, McCure and Pearson, “Thermal Effects in Dimensional Metrology”, Lawrence Radiation Laboratory, Livermore, California, 1965

[2] M. Casado, “Water stable isotopic composition on the east Antarctic plateau: measurements at low temperature of the vapour composition, utilisation as an atmospheric tracer and implication for paleoclimate studies,” Ph.D. thesis, Paris Saclay (2016).

This post has been previously published as a white Paper on the ANSYS website in 2019. This is the original version.