Can your mathcad find a solution to the example CSTR ?

[ptc-4735478]

I think I may be the only one who cannot have a converged solution in the last solve block. Maybe it is because I do not have the latest version. The only changes that I made to Harvey's file is the mfin and tout index. the solution does not converge if I deviate by even 10% of the initial condition. Just want to check with the community that the file is indeed working fine. Thanks

[AndyWesterman]

Hi Henry,

I'm not familiar with the chemical processes that you are modelling, but my question is - should this give an exact answer, or are you looking to get the best fit to the data input?

If you only need the best fit then try changing the Find to Minerr.

regards

Andy

[Werner_E]

If the question is to test in a newer Mathcad version - In MC15 M020, the current release, the solve block does not converge, too.

If I set CTOL:=0.002 (default is 0.001) I get a solution.

[AndyWesterman]

Using the Minerr path , the residuals calculate to distinctly non-zero values, so Find must (nearly) always fail if this is the "best " answer that Mathcad can derive.

.

[ptc-4735478]

This is not my worksheet. It belongs to Harvey and he has been kind enough to help me in the document section. Since I cannot post attachment in document section, I have to post it in discussion. He said I have to pick a new inlet temperature, I think now I understand the worksheet better now.

Besides, I think I found what is the problem too. It is Qm, Qm cannot be picked abituarily only at certain Qm the system would converge. Look at the previous red S shape graph, if Qm is picked such that the slope would change sign while mathcad do its iteration, then mathcad fail to converge. Since the S shape curve can be shifted left and right by tin, changing tin is what is needed if I want to keep Qm to be zero.