Cannot evaluate this accurately at one or more of the specified values

[ptc-5205519]

Goodness this is the worksheet that keeps on giving....errors, that is.

I've worked through most of the bugs after converting this worksheet to the new Mathcad format and thought FINALLY, I've got it up and running as well as it worked in 15. Unfortunately, no.

The problem is that it will only work when ncutoff:=0. As soon as ncutoff:= 1 or 2 the program crashes and I get a "Cannot evaluate this accurately at one or more of the specified values, because the values are outside of the range of accuracy of the function" error in the Morphological Time Loop routine (pg.37). Now, the functions that depend on ncutoff directly (as far as I can tell) include:

1. (H.o a.o L H.ero) on pg. 7 of the doc

2. Morphological Time Loop

-Q_p (pg. 44)

-Q.t (pg. 45)

The purpose of the Q.t loop is to identify which Q value to use for the further calculations, based on which fissure model is used, ncutoff, and a few other variables. I'm thinking the problem is that when ncutoff:=1 or 2 it uses a different Q value (namely Q.a). As I am not all that familiar with Mathcad I am having a hard time identifying a solution to the problem. It baffles me that this program runs in the old version but not Prime. Any help would be appreciated!

Note: fissure_model must be set to 5 in order for the program to use the ncutoff variable.

[Werner_E]

This worksheet of yours really is a tough one. Seems to be like an endless job.

Haven't looked at your new one but just thought to point you to the solution of the last question of yours here http://communities.ptc.com/message/214723#214723

I also outlined how to trace the error even if we unfortunately don't have the trace command in Prime.

There was/still is a problem with posting to the forum as of a big spam attack and so postings were withhold for some days and still yet are delayed because of "moderation".

[jkucan]

Interesting worksheet. The issue is reported by the call to "mod(accnts, hdiv)" function inside the definition of "Evolve". I am looking into the details what would be triggering this specific message. I'll let you know what I find out.

To find out the root of the error:

1) Recalculate worksheet (this is only necessary once after you opened the worksheet).

2) On Calculation tab, click on |< (red VCR buttons) to trace to first error in the chain.

Hope this helps,

--Jakov

[jkucan]

mod returns this error when the relative difference in magnitude of its arguments exceeds precision. Usually this means that the first argument is very large. For example mod(10^40,2).