i experienced problem in definition of variable shell thickness by Nodal thickness feature in quadratic element type. It seems not working properly, below example cases (link).
I made some testing some months ago. Posted files contain second order shells. There was not too much enthusiasm about it. I guess it’s difficult to model and write the cards by hand.
I also considered the possibility of building the weld root with this increased thickness at the base. That is very simple as it only requires a few nodes alignement definitions.
** Generated by Mecway 23
*NODE
1,0,0.046875,0.156
2,0,0.0625,0.051
3,0,0.046875,0.111
4,0,0.0625,0.081
5,0,0.046875,0.126
6,0,0.046875,0
7,0,0.046875,0.246
8,0,0.046875,0.201
9,0,0.046875,0.216
10,0,0.0625,0.156
11,0,0.0546875,0.186
12,0,0.0625,0.186
13,0,0.046875,0.0015
14,0,0.0546875,0.216
15,0,0.046875,0.141
16,0,0.0625,0.216
17,0,0.046875,0.231
18,0,0.046875,0.051
19,0,0.0625,0.246
20,0,0.0546875,0
21,0,0.046875,0.171
22,0,0.046875,0.006
23,0,0.046875,0.021
24,0,0.0546875,0.036
25,0,0.046875,0.186
26,0,0.046875,0.036
27,0,0.0546875,0.006
28,0,0.046875,0.081
29,0,0.0546875,0.003
30,0,0.0625,0.006
31,0,0.046875,0.096
32,0,0.046875,0.066
33,0,0.0625,0.003
34,0,0.0546875,0.156
35,0,0.0546875,0.126
36,0,0.0625,0
37,0,0.0625,0.111
38,0,0.0625,0.141
39,0,0.0625,0.171
40,0,0.0625,0.201
41,0,0.0546875,0.246
42,0,0.0625,0.231
43,0,0.0625,0.036
44,0,0.0546875,0.066
45,0,0.0625,0.0015
46,0,0.0625,0.066
47,0,0.0546875,0.096
48,0,0.0625,0.096
49,0,0.046875,0.003
50,0,0.046875,0.0045
51,0,0.0625,0.126
52,0,0.0625,0.0045
53,0,0.0625,0.021
*ELEMENT,TYPE=S8
1,49,33,30,22,29,52,27,50
2,22,30,43,26,27,53,24,23
3,26,43,46,32,24,2,44,18
4,32,46,48,31,44,4,47,28
5,31,48,51,5,47,37,35,3
6,6,36,33,49,20,45,29,13
7,5,51,10,1,35,38,34,15
8,1,10,12,25,34,39,11,21
9,25,12,16,9,11,40,14,8
10,9,16,19,7,14,42,41,17
*NSET,NSET=Team1
50
52
*NSET,NSET=Team4
6
20
36
*NSET,NSET=Team2
29
33
49
*NSET,NSET=Team3
13
45
*NSET,NSET=N-SHELL1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
*ELSET,ELSET=SHELL1
1
2
3
4
5
6
7
8
9
10
*SURFACE,NAME=Master
*SURFACE,NAME=Slave
6,S3
*MATERIAL,NAME=MAT1
*ELASTIC,TYPE=ISOTROPIC
210000000000,0
*DENSITY
7850
*BOUNDARY
6,2,,0
6,1,,0
6,3,,0
6,5,,0
20,3,,0
20,1,,0
20,5,,0
36,3,,0
36,1,,0
36,5,,0
*SHELL SECTION,ELSET=SHELL1,MATERIAL=MAT1,OFFSET=0,NODAL THICKNESS
0.005
*NODAL THICKNESS
N-SHELL1,0.005
Team1, 0.005
Team2, 0.005
Team3, 0.009
Team4, 0.013
*STEP
*STATIC
*CLOAD
19,1,333.3333333333
41,1,333.3333333333
7,1,333.3333333333
*NODE FILE,GLOBAL=YES
U,RF
*EL FILE
S,NOE
*END STEP
thanks for looking up my model and given another example. In the past i have several tests of nodal thickness of beam element, both of linear and quadratic element, straight and curved in geometry. All of them working properly without any issues.
it seems nodal thickness of shell element did not take overwrites as described in documentation, need to defining twice for the whole nodes. I also tested for linear shell element, but still have the same problem.
can it be confirmed to work properly on multiple mesh also? i’m not succeeding here.
*Shell section, Elset=Shell_part-2, Material=S235, Offset=0
10
*NODAL THICKNESS
Node_Set-4,10
Node_Set-3,15
Node_Set-1,17.5
Node_Set-2,12.5
It works properly.
You are missing the NODAL THICKNESS parameter on the *shell section.
“It indicates that the thickness for ALL nodes in the element set are defined with an extra *NODAL THICKNESS card and that any thicknesses defined on the SECTION card are irrelevant”.
NOTE: irrelevant doesn’t mean it is not necessary in this case. If omited, will throw an error so I guess this card is not properly descrived in the documentation.
*SHELL SECTION,ELSET=Eall,MATERIAL=EL,OFFSET=0,NODAL THICKNESS
0.01
*NODAL THICKNESS
Team1, 0.0820
Team2, 0.0103
Team3, 0.0820
…
…
1 Like
got it, this actually i miss since the beginning. Thank you
*Shell section, Elset=Shell_part-1, Material=S235, Offset=0, Nodal Thickness
5
*NODAL THICKNESS
Node_Set-3,5.0
Node_Set-1,10.0
Node_Set-2,7.5
unforttunately, maybe some serious book such from Eriksson and Lignell (2003) link is needed to be cited.
