TAGGED: -2d-materials-and-metals, lsdyna, membrane-elements
-
-
November 20, 2023 at 2:08 pm
r.bombardieri
SubscriberI am attempting to evaluate the modal response of a pressurized membrane-like structure.
I have created a simple example consisting of a square membrane, simply supported along the edges, and subjected to uniform pressure. Here are some additional details:
I am utilizing a fully integrated Belytschko-Tsay membrane (ELFORM=9) that is designed to be hourglass-free.
I am using a linear elastic material (*MAT_ELASTIC).
The pressure load is applied using the *LOAD_SHELL_SET_ID keyword.
Currently, I have opted for the explicit solver, applying the pressure linearly.
I am constraining the translational degrees of freedom (DOF) of the edges only.
I am using version R13.1.1-13-g5480eca294 mpp_dp for Windows.
Upon running the simulation, I immediately observe what I thin kto be significant hourglassing in the membrane deformation and am unsure about the cause (please see the attached picture).
How can I suppress this behavior? What is causing it?
Best regards,
Rocco
-
November 20, 2023 at 5:41 pm
r.bombardieri
SubscriberInterestingly enough, reducing the time step using keyword *CONTROL_TIMESTEP delivered a very good result. Also the implicit solver was able to deliver the same robust results using the following settings:
*CONTROL_ACCURACY
$# osu inn pidosu iacc
1 4 0 0
*CONTROL_IMPLICIT_AUTO
$# iauto iteopt itewin dtmin dtmax dtexp kfail kcycle
1 30 10 0.0 1.0E-4 0.0 0 0
*CONTROL_IMPLICIT_DYNAMICS
$# imass gamma beta tdybir tdydth tdybur irate alpha
1 0.6 0.38 0.01.00000E281.00000E28 0 0.0
*CONTROL_IMPLICIT_GENERAL
$# imflag dt0 imform nsbs igs cnstn form zero_v
1 1.0E-4 2 1 1 0 0 0
*CONTROL_IMPLICIT_SOLUTION
$# nsolvr ilimit maxref dctol ectol rctol lstol abstol
12 15 15 0.001 0.011.00000E10 0.91.0000E-20
$# dnorm diverg istif nlprint nlnorm d3itctl cpchk
1 1 1 1 0 1 0
$# arcctl arcdir arclen arcmth arcdmp arcpsi arcalf arctim
0 0 0.0 1 2 0 0 0
$# lsmtd lsdir irad srad awgt sred
5 2 0.0 0.0 0.0 0.0 -
November 20, 2023 at 5:55 pm
r.bombardieri
SubscriberInterestingly enough, reducing the time step by keyword *CONTROL_TIMESTEP help achieve a good solution. Same soluction could be obtained using the implicit solver with the following settings.
*CONTROL_ACCURACY
$# osu inn pidosu iacc
1 4 0 0
*CONTROL_IMPLICIT_AUTO
$# iauto iteopt itewin dtmin dtmax dtexp kfail kcycle
1 30 10 0.0 1.0E-4 0.0 0 0
*CONTROL_IMPLICIT_DYNAMICS
$# imass gamma beta tdybir tdydth tdybur irate alpha
1 0.6 0.38 0.01.00000E281.00000E28 0 0.0
*CONTROL_IMPLICIT_GENERAL
$# imflag dt0 imform nsbs igs cnstn form zero_v
1 1.0E-4 2 1 1 0 0 0
*CONTROL_IMPLICIT_SOLUTION
$# nsolvr ilimit maxref dctol ectol rctol lstol abstol
12 15 15 0.001 0.011.00000E10 0.91.0000E-20
$# dnorm diverg istif nlprint nlnorm d3itctl cpchk
1 1 1 1 0 1 0
$# arcctl arcdir arclen arcmth arcdmp arcpsi arcalf arctim
0 0 0.0 1 2 0 0 0
$# lsmtd lsdir irad srad awgt sred
5 2 0.0 0.0 0.0 0.0
-
- The topic ‘Modal response of membrane subjected to uniform pressure’ is closed to new replies.
-
2773
-
960
-
830
-
599
-
591
© 2025 Copyright ANSYS, Inc. All rights reserved.