Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Best way to apply edge pressure in 2D plane strain? Referring to OS-V: 0085

Reference is made to OS-V: 0085 which uses nodal pressure to compute a pressure vessel (modelled as quarter segment).

I'm using different parameters to study a thick-walled cylinder. However, results show a kind of "hourglassing" effect with nodal pressure. It is not real hourglassing because it is only localised to the inner radius, elements are already fully integrated (second order) and the effect does not disappear with mesh refinement. See the figure below:

After checking the reference manual, it seems to be possible to try "PLOAD2" as alternative to radial nodal pressure. The figure below shows that PLOAD2 pressure applies to the element edges instead of the nodes:

But now I'm stuck, because these loads are not included in the analysis. Did I miss something? Could CQPSTN elements be incompatible? Or is there a better way to define edge pressure on a 2D model? See the model in attachment.

test_pload2.hm

Find more posts tagged with

English

OptiStruct

v2022.1

Accepted answers

Adriano_Koga

Thanks for the response!

PLOADE1 seems only available for LGDISP nonlinear, I cannot find it for the current static/linear model (see screenshot):

Is there an equivalent option for the linear/static case? I already tried PLOAD1; same result as PLOAD2 (no error, no deformation).

However PLOAD does at least give the below error message, maybe possible to solve?

*** See next message about line 143 from file: C:/temp/hw_session/b.fem
"PLOAD 1 1.0+7 10 11 0"
*** ERROR # 1000 *** in the input data:
Incorrect data in field # 6.

My experience in Optistruct is limited, perhaps I missed an easy way to solve this?

look at the 2 attached models, for linear usage of PLOADE1 and PLOADSF

Pressure_Vessel_LS_PRESSURE_LOAD.fem

Pressure_Vessel_LS_PRESSURE_LOAD_PLOADE1.fem

All comments

Adriano_Koga

I haven't tested but you probably need to use this one PLOADE1

Wesley_Peijnenburg2

Thanks for the response!

PLOADE1 seems only available for LGDISP nonlinear, I cannot find it for the current static/linear model (see screenshot):

Is there an equivalent option for the linear/static case? I already tried PLOAD1; same result as PLOAD2 (no error, no deformation).

However PLOAD does at least give the below error message, maybe possible to solve?

*** See next message about line 143 from file: C:/temp/hw_session/b.fem
"PLOAD 1 1.0+7 10 11 0"
*** ERROR # 1000 *** in the input data:
Incorrect data in field # 6.

My experience in Optistruct is limited, perhaps I missed an easy way to solve this?

Adriano_Koga

Thanks for the response!

PLOADE1 seems only available for LGDISP nonlinear, I cannot find it for the current static/linear model (see screenshot):

Is there an equivalent option for the linear/static case? I already tried PLOAD1; same result as PLOAD2 (no error, no deformation).

However PLOAD does at least give the below error message, maybe possible to solve?

*** See next message about line 143 from file: C:/temp/hw_session/b.fem
"PLOAD 1 1.0+7 10 11 0"
*** ERROR # 1000 *** in the input data:
Incorrect data in field # 6.

My experience in Optistruct is limited, perhaps I missed an easy way to solve this?

look at the 2 attached models, for linear usage of PLOADE1 and PLOADSF

Pressure_Vessel_LS_PRESSURE_LOAD.fem

Pressure_Vessel_LS_PRESSURE_LOAD_PLOADE1.fem

Wesley_Peijnenburg2

look at the 2 attached models, for linear usage of PLOADE1 and PLOADSF

The PLOADSF fem file example is exactly what I need... Hopefully my final question: how did you do it?

PLOADSF for a new pressure BC is not available as below screenshot shows:

However this was solved using the answer here:

https://community.altair.com/community?id=kb_article_view&sys_kb_id=ea88ba531b0c8554507ca6442a4bcb95

PLOADSF can be accessed with the search option (Ctrl+f). So one obstacle is removed.

Now the documentation on PLOADSF mentions it can be defined using SURF or SET entry. It seems that you used a set? I tried following your footsteps with the set definition, but this is where I am now stuck.

This is the set in your model (which I want to reproduce), see dots on inner radius:

However when I try to reproduce (using model>contacts>surfaces) and select the edges I get something different, see blue cross marks:

Strangely, I see no difference in the context menu of both (except for ID and color).

What is causing this difference in set definition? Sorry if I'm missing something obvious and many thanks for your time and effort. Strongly appreciated!

Community-Pasted-Attachment-1673344419995.png

Adriano_Koga

The PLOADSF fem file example is exactly what I need... Hopefully my final question: how did you do it?

PLOADSF for a new pressure BC is not available as below screenshot shows:

However this was solved using the answer here:

https://community.altair.com/community?id=kb_article_view&sys_kb_id=ea88ba531b0c8554507ca6442a4bcb95

PLOADSF can be accessed with the search option (Ctrl+f). So one obstacle is removed.

Now the documentation on PLOADSF mentions it can be defined using SURF or SET entry. It seems that you used a set? I tried following your footsteps with the set definition, but this is where I am now stuck.

This is the set in your model (which I want to reproduce), see dots on inner radius:

However when I try to reproduce (using model>contacts>surfaces) and select the edges I get something different, see blue cross marks:

Strangely, I see no difference in the context menu of both (except for ID and color).

What is causing this difference in set definition? Sorry if I'm missing something obvious and many thanks for your time and effort. Strongly appreciated!

I've create a set of type SURF (Ctrl+F >> SURF) and selected the 'edge'.

at least in 2022.2 it is working just fine.

Wesley_Peijnenburg2

I've create a set of type SURF (Ctrl+F >> SURF) and selected the 'edge'.

at least in 2022.2 it is working just fine.

The problem is now completely solved. Below screenshot shows the stress distribution is very uniform along the inner radius with PLOADSF:

Thanks again!