Hinges between plates/shells

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Hinges between plates/shells

Post by bernd »

thschrader wrote: Mon Jul 12, 2021 2:07 pm
bernd wrote: Mon Jul 12, 2021 1:05 pm deformation is quite large (35 cm) ... May be some other could have a look as well ...
Load were far too high.
Did some modifications and comparison with my commercial tool.

concrete_plate_ts.FCStd

new_model.JPG
damn I knew something is wrong. Ahh it is Friday time to go into the Weekend ... :o :mrgreen:

OK the one from Thomas is much better ... We should take this one as our hinge example. Next step would be a Calculix input file. I have no idea how this would work with MPC ... http://www.feacluster.com/CalculiX/ccx_ ... de278.html Any Calxulix expert around?
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Hinges between plates/shells

Post by thschrader »

bernd wrote: Mon Jul 12, 2021 3:25 pm damn I knew something is wrong. Ahh it is Friday time to go into the Weekend ... :o :mrgreen:
I did a mistake too. Switched thickness of walls/floor. ;)
Here is a corrected one:
concrete_plates_test.FCStd
(138.21 KiB) Downloaded 49 times
Floor d=20 cm, walls d=16.
Now the deformation matches exactly with my commercial tool.
Its a long way to tipperary...
fandaL
Posts: 440
Joined: Thu Jul 24, 2014 8:29 am

Re: Hinges between plates/shells

Post by fandaL »

To be easier for hand modifications, I used very course mesh (two quad4 elements per width).
concrete_plates_test.FCStd
(12.16 KiB) Downloaded 53 times

Hand modifications of the inp file:
Nodes on the hinge edges are duplicated:
1, 3, 5, 7, 11, 17 duplicated as 40-45
Elements on the top face with replaced nodes:
element 45, nodes 3, 11 by 41, 44
element 46, nodes 1, 11 by 40, 44
element 53, nodes 7, 17 by 43, 45
element 56, nodes 5, 17 by 42, 45
Node translation DOF are bound by equations

Modified beginning of the inp file:

Code: Select all

*Node, NSET=Nall
1, 6000, 0, 2500
2, 6000, 0, 0
3, 6000, 2000, 2500
4, 6000, 2000, 0
5, 0, 0, 2500
6, 0, 0, 0
7, 0, 2000, 2500
8, 0, 2000, 0
9, 6000, 0, 833.3333333333
10, 6000, 0, 1666.666666667
11, 6000, 1000, 2500
12, 6000, 2000, 833.3333333333
13, 6000, 2000, 1666.666666667
14, 6000, 1000, 0
15, 0, 0, 833.3333333333
16, 0, 0, 1666.666666667
17, 0, 1000, 2500
18, 0, 2000, 833.3333333333
19, 0, 2000, 1666.666666667
20, 0, 1000, 0
21, 1000, 2000, 2500
22, 2000, 2000, 2500
23, 3000, 2000, 2500
24, 4000, 2000, 2500
25, 5000, 2000, 2500
26, 1000, 0, 2500
27, 2000, 0, 2500
28, 3000, 0, 2500
29, 4000, 0, 2500
30, 5000, 0, 2500
31, 6000, 994.5004138628, 831.4934419314
32, 6000, 996.8858840134, 1665.089340799
33, 0, 994.5004138628, 831.4934419314
34, 0, 996.8858840134, 1665.089340799
35, 4999.983071052, 999.9830710519, 2500
36, 996.0531366691, 996.0531366691, 2500
37, 1998.547335024, 998.547335024, 2500
38, 2999.640259853, 999.6402598526, 2500
39, 3999.92550759, 999.9255075899, 2500

** DUPLICATED NODES
** 1 -> 40, 3 -> 41, 5 -> 42, 7 -> 43, 11 -> 44, 17 -> 45
40, 6000, 0, 2500
41, 6000, 2000, 2500
42, 0, 0, 2500
43, 0, 2000, 2500
44, 6000, 1000, 2500
45, 0, 1000, 2500

** Face elements
*Element, TYPE=S4, ELSET=Efaces
33, 31, 32, 10, 9
34, 32, 11, 1, 10
35, 32, 31, 12, 13
36, 11, 32, 13, 3
37, 2, 14, 31, 9
38, 12, 31, 14, 4
39, 33, 15, 16, 34
40, 34, 16, 5, 17
41, 34, 19, 18, 33
42, 17, 7, 19, 34
43, 6, 15, 33, 20
44, 18, 8, 20, 33
**45, 11, 3, 25, 35  REWRITTEN ELEMENTS
45, 44, 41, 25, 35
**46, 30, 1, 11, 35
46, 30, 40, 44, 35
47, 23, 22, 37, 38
48, 35, 25, 24, 39
49, 24, 23, 38, 39
50, 37, 27, 28, 38
51, 29, 30, 35, 39
52, 38, 28, 29, 39
**53, 17, 36, 21, 7
53, 45, 36, 21, 43
54, 37, 36, 26, 27
55, 21, 36, 37, 22
**56, 36, 17, 5, 26
56, 36, 45, 42, 26

** BINDING DOF OF DUPLICATED NODES
** NODES 1 AND 40
*EQUATION
2
1, 1, 1, 40, 1, -1
2
1, 2, 1, 40, 2, -1
2
1, 3, 1, 40, 3, -1
** NODES 3 AND 41
2
3, 1, 1, 41, 1, -1
2
3, 2, 1, 41, 2, -1
2
3, 3, 1, 41, 3, -1
** NODES 7 AND 43
2
7, 1, 1, 42, 1, -1
2
7, 2, 1, 42, 2, -1
2
7, 3, 1, 42, 3, -1
** NODES 11 AND 44
2
11, 1, 1, 44, 1, -1
2
11, 2, 1, 44, 2, -1
2
11, 3, 1, 44, 3, -1
** NODES 17 AND 45
2
17, 1, 1, 45, 1, -1
2
17, 2, 1, 45, 2, -1
2
17, 3, 1, 45, 3, -1   
inp_files.zip
(4.5 KiB) Downloaded 44 times
numbering_from_Salome.png
numbering_from_Salome.png (30.25 KiB) Viewed 1895 times
Original displacement magnitude is 3.09 mm
original.png
original.png (38.63 KiB) Viewed 1895 times
With hinges displacement magnitude is 7.8 mm
with_hinges.png
with_hinges.png (37.33 KiB) Viewed 1895 times

I tried also variant with *MPC cards. It triggered nonlinear calculation and gave me no results, spooles.out file saying: "matrix found to be singular". Just instead of *EQUATIONS, there were:

Code: Select all

*MPC
BEAM, 1, 40
*MPC
BEAM, 3, 41
*MPC
BEAM, 5, 42
*MPC
BEAM, 7, 43
*MPC
BEAM, 11, 44
*MPC
BEAM, 17, 45 
User avatar
johnwang
Veteran
Posts: 1345
Joined: Sun Jan 27, 2019 12:41 am

Re: Hinges between plates/shells

Post by johnwang »

Wrote a program to generate the mesh and then solved with Mystran. Use PLOAD2 on CQUAD4.

Why the max shows 'nm', maybe my MAT1 card is not right.

Code: Select all

 eletype = CQUAD4                                  
 Length =    6.000000    
 Height =    2.500000    
 Width =    2.000000    
 dxyz =   0.5000000    
 Nx =           12
 Ny =            4
 Nz =            5
 Plate = 20 cm walls 16 cm E=30000 MPa NU=0.17
 Loading 10 kN/m^2
Attachments
hinge_result.FCStd
(12.82 KiB) Downloaded 49 times
hinge.FCStd
(5.32 KiB) Downloaded 52 times
hinge.jpg
hinge.jpg (48.36 KiB) Viewed 1865 times
Last edited by johnwang on Tue Jul 13, 2021 10:36 am, edited 2 times in total.
hfc series CAE workbenches for FreeCAD (hfcNastran95, hfcMystran, hfcFrame3DD, hfcSU2 and more)
fandaL
Posts: 440
Joined: Thu Jul 24, 2014 8:29 am

Re: Hinges between plates/shells

Post by fandaL »

johnwang wrote: Tue Jul 13, 2021 4:21 am Why the max shows 'nm', maybe my MAT1 card is not right.
Hinge1CQUAD4.dat.txt contains only the load by 10 Pa, if the units are m, N, Pa

Code: Select all

$-------SID-----P-------EID1----THRU----EID2----7-------8-------9-------+-------
PLOAD2  350     10.00   26      THRU    85
User avatar
johnwang
Veteran
Posts: 1345
Joined: Sun Jan 27, 2019 12:41 am

Re: Hinges between plates/shells

Post by johnwang »

fandaL wrote: Tue Jul 13, 2021 6:46 am Hinge1CQUAD4.dat.txt contains only the load by 10 Pa, if the units are m, N, Pa
Corrected to:

Code: Select all

$-------SID-----P-------EID1----THRU----EID2----7-------8-------9-------+-------
PLOAD2  350     -10000.026      THRU    85 
Going down and NU=0.17
It is 3.55 um now. Still small.
Last edited by johnwang on Tue Jul 13, 2021 11:27 am, edited 1 time in total.
hfc series CAE workbenches for FreeCAD (hfcNastran95, hfcMystran, hfcFrame3DD, hfcSU2 and more)
fandaL
Posts: 440
Joined: Thu Jul 24, 2014 8:29 am

Re: Hinges between plates/shells

Post by fandaL »

johnwang wrote: Tue Jul 13, 2021 9:32 am Corrected to:

Code: Select all

$-------SID-----P-------EID1----THRU----EID2----7-------8-------9-------+-------
PLOAD2  350     -10000.026      THRU    85 
Going down and NU=0.17
It is 3.55 um now. Still small.
The pressure is applied on elements 26-85. If I count well, that is 60 elements. Looking at the picture, top face contains 12x4 = 48 elements. In fact, some elements on top are not loaded and some side elements are loaded and the direction of the load is upwards. If you will correct this, displacement magnitude will not change much. In MSC.Nastran I get total deformation 0.00353 m.

Couldn't be the reason in reading units by FreeCAD?
User avatar
johnwang
Veteran
Posts: 1345
Joined: Sun Jan 27, 2019 12:41 am

Re: Hinges between plates/shells

Post by johnwang »

fandaL wrote: Tue Jul 13, 2021 11:01 am The pressure is applied on elements 26-85.
I messed up node with element, should be 21 THRU 68. Result is 3.62 um.
But Mystran's F06 shows T3max(ABS)=3.620582E-03. Close to thschrader's result 3.65 mm.
Strange!
F06.jpg
F06.jpg (23.06 KiB) Viewed 1732 times
Hinge1CQUAD4.dat.txt
(19.33 KiB) Downloaded 52 times
hfc series CAE workbenches for FreeCAD (hfcNastran95, hfcMystran, hfcFrame3DD, hfcSU2 and more)
user1234
Veteran
Posts: 3333
Joined: Mon Jul 11, 2016 5:08 pm

Re: Hinges between plates/shells

Post by user1234 »

bernd wrote: Sun Jul 11, 2021 10:56 pmlast FEM object we added was this one ... https://github.com/FreeCAD/FreeCAD/comp ... 282db7c8c6
I inspected that already. The main problem for me will be git itself and write (understand) the inp file. I hope i can make some results until to the end of the weekend.

Greetings
user1234
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Hinges between plates/shells

Post by bernd »

user1234 wrote: Tue Jul 13, 2021 1:58 pm
bernd wrote: Sun Jul 11, 2021 10:56 pmlast FEM object we added was this one ... https://github.com/FreeCAD/FreeCAD/comp ... 282db7c8c6
I inspected that already. The main problem for me will be git itself and write (understand) the inp file. I hope i can make some results until to the end of the weekend.
Source_code_management
Post Reply