error PMESH and PESURF

[cri]

Dear all,
I'm working on a large scale model of a huge structure (many nodes and elements) with FEAP 85. The first problem I got was the error '*ERROR* PMESH: More than 10000 files for PART'. Hence, I modified the 'setext.f' file by increasing the lenght of 'fext' character to 9 and modifying some lines as follows:

PREVIOUS VERSION:
fext(5:8 ) = '0000'
if(next.le.9) then
        write(fext(8:8 ),'(i1)') next
      elseif(next.le.99) then
        write(fext(7:8 ),'(i2)') next
      elseif(next.le.999) then
        write(fext(6:8 ),'(i3)') next
      elseif(next.le.9999) then
        write(fext(5:8 ),'(i4)') next


CURRENT VERSION
fext(5:9) = '00000'
if(next.le.10) then
  write(fext(9:9),'(i1)') next
elseif(next.le.99) then
  write(fext(8:9),'(i2)') next
elseif(next.le.999) then
  write(fext(7:9),'(i3)') next
elseif(next.le.9999) then
  write(fext(6:9),'(i4)') next
elseif(next.le.99999) then
  write(fext(5:9),'(i5)') next

Consequently, I also modified the last lines of 'ppart.f' to i5 as:
2000  format(/10x,'Part(',i5,') = ',a)
2001  format(/10x,'Part(',i5,'): Elm start =',i8,
     &       /21x,'Elm last  =',i8,
     &       /21x,'Nodes/el. =',i8)

and I increased the size of 'partname' and 'ptelm' in 'p_ptname.h'. I was then able to plot the model geometry without errors. Unfortunately, if I try to run an old input file (which worked before these changes) I get a new error, that is:

 '*ERROR* PESURF      PESURF  norm    tang    trac    disp    dofs    flux    mate    axis    plan    func    node    -Force  -Force  -B.C: Edge data fileSSMP_feb25error.txt.cbou0000 does not exist'

Can you indicate me if I have to modify other source files or if there is somenthing wrong in my modifications? Otherwise, is there a different procedure to avoid the first error ('*ERROR* PMESH...') I got?

Looking forward to hearing from you
Thanks in advance
Cristina

[Prof. S. Govindjee]

setext.f is called from a lot of different places and fext is declared in all of those calling units.  You will need to change the declared length of fext in all these locations.  Search the entire code for calls to setext and make the changes, then rebuild everything.

[Prof. R.L. Taylor]

What are you using to generate the mesh.  There should normally not be that many parts. 

[cri]

Dear Prof. Govindjee and dear Prof. Taylor,
thanks a lot for your replies.
I changed the declared length of fext in all the locations I found. However, I still got an error similar, but not equal, to the one I reported before ('*ERROR* PESURF..').
As regards the mesh, the model geometry is extremely irregular and, thus, I'm using an elevated number of blocks (almost one block for each finite element). For this reason, I modified the last lines of ppart.f file, as I described in my previous post.

Best regards

[Prof. R.L. Taylor]

Can you post the input file?

[cri]

Dear Prof. Taylor,
I really appreciate your interest and thank you for your precious support. However, we have introduced some changes and the code is currently working with this specific model. So, at the moment, there is no need to share the input file, as it is really complicated.

Thank you so much for your precious help
Best regards