Stresses are not correct in Shell DKGQ element

[CBabakhanians]

I model one element shell model with Shell DKGQ element.
- The element has 4 nodes (1 2 3 4).
- The nodes 1 and 4 are fixed.
- A point load is applied to nodes 2 and 3 equal to 100 N.
- Shell thickness is equal to 0.02 m.
- Modulus of Elasticity is equal to 2.e011 N/m2.
- Poison's Ratio is equal to 0.3.

The model is analyzed and the results of the base reactions and node displacements are true.
- Base Reactions 1 and 4: 100 N
- Node Displacements 2 and 3: -0.000456 m; SAP2000: -0.000456 m;
- Element Node Stresses (Node 1): S11: 16.316 N-m/m; S22: 100 N-m/m; but in SAP2000 (S11:-200 N-m/m; S22:-60 N-m/m)
As seen, the results of Element Node Stresses are not correct.
I have tested several models, all results are true but the results of Element Node Stresses are not correct.

Can anyone help me???
Thank you!!

Here is the script:
#------------------------------------------------------------------------------------------
# Elastic Shell Moel
# Units, N,m

wipe;
model BasicBuilder -ndm 3 -ndf 6;

# Define Parameters
set S 1.0;
set L 1.0;

# Define Nodes

# X Y Z
node 1 0.0 0.0 0.0;
node 2 [expr 1.0*$S] 0.0 0.0;
node 3 [expr 1.0*$S] [expr 1.0*$L] 0.0;
node 4 0 [expr 1.0*$L] 0.0;

# Define Constraint

fix 1 1 1 1 1 1 1;
fix 4 1 1 1 1 1 1;


# Define Materials
set matTag 3;
set E 2.e011;
set v 0.3;


nDMaterial ElasticIsotropic $matTag $E $v;


nDMaterial PlateFiber 4 $matTag


# Define Sections
set secTag 2;


set Thick 0.02;

#section ElasticMembranePlateSection $secTag $E $v $Thick 1

# Tag MatTag Thickness
section PlateFiber $secTag 4 $Thick

# Tag iNode jNode kNode lNode secTag
element ShellDKGQ 1 1 2 3 4 $secTag


puts "End of Model Geometry"

# Define Recorders
recorder Node -file Base_OneElement.txt -node 1 4 -dof 3 reaction;

# Define Load

set LoadCorner 100; # N

pattern Plain 1 Linear {
# Tag Fx Fy Fz Mx My Mz
load 2 0.0 0.0 -$LoadCorner 0.0 0.0 0.0
load 3 0.0 0.0 -$LoadCorner 0.0 0.0 0.0
}
puts "End of Loading"

# Static Analysis
wipeAnalysis;
constraints Transformation;
numberer RCM;
system SparseGeneral;
test EnergyIncr 1e-7 25; # 0;
algorithm ModifiedNewton;
integrator LoadControl 0.01;
analysis Static;
analyze 100;


set dd [nodeReaction 1 3]
puts $dd
set bb [nodeReaction 4 3]
puts $bb
set aa [nodeDisp 2 3]
puts $aa
set b [eleResponse 1 stresses]
puts "Internal moments of element 12 in each nodes"
puts $b
puts "End of Static Analysis"

[yekose]

the reason could be that opensees gives the results on the gauss integration points, so that it may differ from the sap2000. You shall carry the stresses to the nodes by using some extrapolation methods and so.

[CBabakhanians]

Thank you for your reply.
would you please tell me a reference for extrapolating.
Thank you!

[yekose]

https://sites.google.com/site/kolukulas ... s-recovery

[CBabakhanians]

Hi
I hope this is the last question.
the results are for the natural frame. how can I transform the results to physical frame (x, y).
Thanks a lot.

[selimgunay]

Stresses will not be correct when you have a single element

[CBabakhanians]

Hi Dear Selimgunay,
I use SHELLDKGQ element for shallow foundation analysis but this element does not consider the shear deformation effects of "thick plates".

is there any shell element in OpenSees considering the shear deformation effects of thick plates?

Thank you!

[selimgunay]

There are 3D brick elements. You can check them

http://opensees.berkeley.edu/wiki/index ... ck_Element

[CBabakhanians]

many thanks!!

[CBabakhanians]

sorry!!
But I need shell forces to design the foundation.
the solid element forces are not appropriate for the design.

[selimgunay]

How about the regular shell element?

http://opensees.berkeley.edu/wiki/index ... ll_Element

[CBabakhanians]

Hi,
Unfortunately, the formulation of SHELLMITC4 is for tin plates and does not consider the effects of shear deformation.
I checked both MITC4 and DKGQ. DKGQ is very accurate for the tin plate but there are some errors when the thickness becomes considerable for example, in Mat foundations.

[Mojtabaai]

The ShellMITC4 element is applicable to the analysis of both thin and thick shells according to the reference:

https://www.researchgate.net/profile/Ed ... 000000.pdf

[havip]

Hi
I hope this is the last question.
the results are for the natural frame. how can I transform the results to physical frame (x, y).
Thanks a lot.

[vbokaeianOS]

Hi. I am going to model mat foundations in opensees, but I can't find the appropriate element to do so. Dear CBabakhanians, did you get your answer?