Display principal stresses

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
bernd
Posts: 8340
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland

Re: Display principal stresses

Postby bernd » Wed Sep 26, 2018 9:22 pm

Very value input Stefan. It is the way to go IMHO. The graphical output will even look better, similar to the picture in the first post.
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Tue Oct 16, 2018 11:39 am

Very interesting, I will need it sometime...

kudos. ;)
UR_
Posts: 983
Joined: Tue Jan 03, 2017 8:42 pm

Re: Display principal stresses

Postby UR_ » Tue Oct 16, 2018 7:14 pm

AFAIK, Paraview offers three methods for reducing glyph density

ParaView 5.6.0-RC2 64-bit-000079.png
ParaView 5.6.0-RC2 64-bit-000079.png (6.77 KiB) Viewed 373 times

Mode 2 and 3 are only dropping glyphs, there is no interpolation or other maths ;)

BTW, because of ordered mesh numbering, necessary for optimized solving, there is no major difference in visualization between mode 2 and mode 3.
Perhaps not worth the effort.
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Sat Dec 22, 2018 8:56 am

Hello,

I have a doubt regarding the stresses and their posterior calculation like principal stresses....

Taking a row vector from calculix .frd and .dat files...

Are the indexes of stress values sorted in the same manner?

I check that in the .dat file the stress for each element integration point is indexing in the following manner:

Code: Select all

stresses (elem, integ.pnt.,sxx,syy,szz,sxy,sxz,syz)
So if we include the values in the stress tensor, it is in the following way:

S=[sxx sxy sxz
sxy syy syz
sxz syz szz]

Using numbers as index...

S=[1 4 5
4 2 6
5 6 3]

However, the .frd file shows that:

Code: Select all

 -5  SXX         1    4    1    1
 -5  SYY         1    4    2    2
 -5  SZZ         1    4    3    3
 -5  SXY         1    4    1    2
 -5  SYZ         1    4    2    3
 -5  SZX         1    4    3    1
This means that the stress tensor values for nodes is:

S=[1 4 6
4 2 5
6 5 3]?

To sum up, are sxz and syz components exchanged?

Best regards,
fandaL
Posts: 340
Joined: Thu Jul 24, 2014 8:29 am

Re: Display principal stresses

Postby fandaL » Sat Dec 22, 2018 9:27 pm

EkaitzEsteban wrote:
Sat Dec 22, 2018 8:56 am
To sum up, are sxz and syz components exchanged?
I think you are right that components are written in a different order.
I don't exactly recall where it was discussed, if somewhere on this forum or on the ccx mailing list.
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Sat Dec 22, 2018 9:37 pm

fandaL wrote:
Sat Dec 22, 2018 9:27 pm
EkaitzEsteban wrote:
Sat Dec 22, 2018 8:56 am
To sum up, are sxz and syz components exchanged?
I think you are right that components are written in a different order.
I don't exactly recall where it was discussed, if somewhere on this forum or on the ccx mailing list.
Thank you FandaL!
User avatar
HarryvL
Posts: 1043
Joined: Sat Jan 06, 2018 7:38 pm

Re: Display principal stresses

Postby HarryvL » Sat Dec 22, 2018 9:44 pm

It was discussed in this topic, see a few posts back:

https://forum.freecadweb.org/viewtopic. ... 10#p240408
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Mon Dec 24, 2018 11:59 am

fandaL wrote:
Sat Dec 22, 2018 9:27 pm
EkaitzEsteban wrote:
Sat Dec 22, 2018 8:56 am
To sum up, are sxz and syz components exchanged?
I think you are right that components are written in a different order.
I don't exactly recall where it was discussed, if somewhere on this forum or on the ccx mailing list.
Response from cxx mailing group https://groups.yahoo.com/neo/groups/cal ... ages/15199, I copy and paste it.

I don't know about .dat but I'll confirm that the .frd has the YZ component as the 5th component. I just ran a model with known shear stress and it appeared in the 5th position. It shouldn't matter though because you should use the ICIND1 and ICIND2 values to identify the component rather than the order.


Be careful when postprocessing .frd and .dat files...
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Mon Dec 24, 2018 12:00 pm

HarryvL wrote:
Sat Dec 22, 2018 9:44 pm
It was discussed in this topic, see a few posts back:

https://forum.freecadweb.org/viewtopic. ... 10#p240408
Thumbs up!
EkaitzEsteban
Posts: 100
Joined: Wed Sep 12, 2018 1:31 pm

Re: Display principal stresses

Postby EkaitzEsteban » Mon Dec 24, 2018 1:17 pm

Hello,

My last confirmation regarding stresses. I make a simple test: inital conditions pure stress (500 MPa - SXY direction) in element 51.

Code: Select all

#Element information
El=[46, 9, 3, 8, 6
    47, 5, 3, 6, 4
    48, 3, 4, 8, 6
    49, 7, 5, 6, 4
    50, 5, 3, 4, 2
    51, 3, 2, 1, 4]
# From FRD file:
NodeS=[[b]-5.53205E-01  7.20325E+00  7.61790E+01 -4.43743E+00 -2.79101E+02 -1.59312E+01[/b]
        1.45933E+01  1.11598E+01  8.20929E+01  9.95657E+00 -2.91087E+01 -2.00360E+01
        5.04025E+00 -9.63229E-01  1.54117E+00  1.13816E+00 -1.38002E+00 -3.10907E+00
       -2.67077E+00 -1.59571E-01 -1.99464E-02  2.53921E-01  5.64170E-02 -8.97747E-02
       -1.78328E-01 -1.56922E+00  1.64284E+00  2.19851E+00  8.63545E+01  4.65185E+00
       -7.29664E+00 -5.57991E+00 -4.10465E+01 -4.97829E+00  1.45544E+01  1.00180E+01
       -2.52012E+01  4.81615E+00 -7.70584E+00 -5.69080E+00  6.90009E+00  1.55453E+01;
        5.44095E-01 -1.24779E+00 -4.05538E+01 -1.07905E+00  1.00185E+01  9.87834E-01
        1.33538E+01  7.97857E-01  9.97321E-02 -1.26961E+00 -2.82085E-01  4.48874E-01]
# From DAT file:
ElementS=[1.335383E+01  7.978568E-01  9.973210E-02 -1.269606E+00  4.488735E-01 -2.820850E-01
         -5.073507E+00 -2.464022E+01 -7.537246E+01 -1.206424E+01  2.255105E+01  3.128036E+01
         -1.226564E+01 -3.293428E+00 -8.120725E+01 -8.885034E-01  1.526795E+00  2.031913E+01
         -2.520123E+01  4.816147E+00 -7.705843E+00 -5.690797E+00  1.554534E+01  6.900095E+00
          2.973975E+01  1.511639E+01  8.800681E+01  2.435058E+01 -2.414085E+01  2.208832E+02
         [b]-5.532049E-01  7.203245E+00  7.617900E+01 -4.437433E+00 -1.593121E+01 -2.791007E+02][/b]
I check the values in NODE 1 from element 51:

# Stress FRD file
-0.5532 7.2032 76.1790 -4.4374 -279.1010 -15.9312
# Stress DAT file
-0.5532 7.2032 76.1790 -4.4374 -15.9312 -279.1007

Sxy and Sxz are in different order.

Best regards,
Ekaitz.