Problem with small models

[resammc]

I'm using v8.4 in unix environment (Ubuntu):

• Description:   Ubuntu 15.10
  Release:   15.10
  Codename:   wily
  

This morning, I also installed CentOS 7 to make sure that the problem is not from my operating system (or maybe a corrupted code in X Window), then downloaded a fresh version of nurbfeap from the website and did all the changes. Still the same problem.

• CentOS Linux release 7.2.1511 (Core)

The thing is, there is no change in the plot after making the suggested changes in plot routines (seems that the changes does not have any effect on the problem).

I will attach the modified plot routines (am I missing something here?).

Regarding the classical finite element method, it could not be an option for me since I need to use the nurbs shape functions.

[Prof. R.L. Taylor]

My fault.  There is one more change that needs to be done to define the hsize.  Add the attached file in ./program of main feap.

Then in meshck.f (in ./program directory) add the line:

!     Compute nodal spacing data for tolerance use

      call phsize(hr(np(43)),hr(np(44)),mr(np(33)))

at end just before the formats.

[resammc]

Dear Professor Taylor,

Thank you very much for your response. That solved the problem. Just one comment: apparently somewhere between the main feap and nurbfeap, hsize array is restarted because when I write out the hsize array in plot routines in nurbfeap (e.g. pmsh_lin.f ), it returns a zero array. Fortunately, calling phsize again within the nurbfeap module recalculates the hsize and then the plot routines work perfectly (currently, I've added the call to the beginning of pform.f, since I think it is one of the first routines which feap calls).

BTW, I would like to thank you once again for your patience during this debugging process.

[Prof. R.L. Taylor]

Yes, probably it is reset in ./program/pcontr.f

Look for the statements ' hsize (1) = 0.0d0' and delete the four values -- hopefully this is the last of the needed changes.

It is around line 1177 in the file (just before a call to formfe for isw=14 (which initializes all the history variables in elements)

[Prof. R.L. Taylor]

Oops, also delete the print statements after the call to formfe -- they should be in the new added module.

[resammc]

Perfect! Everything works smoothly now.

Many thanks.