Problems about igafeap

[Brisk]

Dear FEAP users,

When I tested nurbs input files in ifeap of FEAP 8.5,  it tells that

*ERROR* PNUMS: Problem size does not permit a solution:
         Number of nodes         =         0
         Number of elements      =         0
         Number of material sets =         1

However, the same input files run correctly in nurbfeap of FEAP 8.4. So, are there something changes in igafeap 8.5? Or I used the wrong program?

[Prof. S. Govindjee]

There are some test files in the folder igafeap/inputs.  Do they work for you?

[Prof. R.L. Taylor]

I ran your files on my version of 8.5 and it runs perfectly?  Did you verify that when you run the first file, Iholep, with a solution added it works?  Even to plot the mesh???

[Brisk]

Dear Prof. Taylor and Prof. Govindjee,

Thanks for your advice.

I checked the Installation Manual and found that

$FEAPHOME8 5/igafeap/main contain special instructions for resolving the loader issue...

Following the instruction, I changed the makefile in $FEAPHOME8 5/igafeap/main and replaced $(ARIFEAP) \ on the loader lines by:

-L$(FEAPHOME8_5)/igafeap \
-Wl,-whole-archive -lifeap8_5 -Wl,-no-whole-archive \

and the problem solved finally.

[fethio]

Thanks for the solution, this solution worked for me as well.

[resammc]

Two matters which are somehow related to this topic:

• For FEAP86 on Ubuntu, I could only make igafeap running by using the following line in the makefile:
  

  -L$(FEAPHOME8_6)/iga \
          -Wl,-whole-archive -l:libifeap8_6i.a -Wl,-no-whole-archive \
  

  which is slightly different than the previous version. This might help someone.

• When I run the 'Ijointr' example in igafeap85 inputs, I get different behaviors in FEAP85 and FEAP86 for the same input file. I did not look deep into it, but, clearly, one of the codes is not showing the expected results.
  (the attached figures)

[Prof. R.L. Taylor]

There are bugs in the membrane -- if you change the Ijointi file to specify "finite' for the two membrane materials the problem runs.  We will work on the other element.