Closed grumpyinca closed 1 year ago
Changes were made to the Cycle_Life calculation (compression, extension & torsion).
These changes are a bit of a fudge (using a hard-coded smallnum) to avoid the problematic behavior. A more insightful change should be considered in a future release.
Double update: Branch 758-4.4 created, Compression, Extension and Torsion eqnset.js copied from branch 758 and checked in to branch 758-4.4 (I did not check in version.js nor messageOfTheDay.md, I only checked in the three eqnset.js files), branch 758-4.4 merged into the latest master, which is for new 4.4 development.
Branch 758 merged into master-4.3.1, closing
Specific input values can cause the Objective Value to become infinity. If that happens, Search is unable to resolve the situation.
This may (or may not) be associated with values of Cycle_Life much less than 1 (for example, 1e-36). Wild Guess: Because it happens without constraints on Cycle_Life, the root cause may be related to the validity penalty.
Steps to recreate:
Also, observe conflicting information:
Note: Cycle_Life is reported as infinity with values of Wire_Dia in the range of 0.0485 and 0.0525 Values of Wire_Dia > 0.053 produce a very small value of Cycle_Life Values of Wire_Dia <0.048 produce a very large value of Cycle_Life