XANSYS Forum Index
    Register    

FAQ    Search    Memberlist    Usergroups    SmartFeedSmartFeed    Profile    Log in
[APDL] Will it take same time for solid and shell elements ?
 
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies.    XANSYS Forum Index -> XANSYS
Author Message
debabrata.podder
User


Joined: 21 Feb 2013
Posts: 51
Location: NIT Meghalaya, India

PostPosted: Wed Apr 22, 2015 10:17 am  Reply with quote

Dear experts,

Just asking a simple question. I want to make a (Thermal/Structural) model where in one case, I want to take solid elements and in another case I want to take shell elements.
For the case of solid elements, only one element is considered along the thickness. Now if we consider shell elements instead of using solid ones, will it take same time to run the analysis or will it take less time ?

Regards-
_________________
Debabrata Podder(PhD)
Assistant Professor
NIT Meghalaya
India
Back to top
View user's profile Send private message
Martin Liddle
User


Joined: 15 Aug 2008
Posts: 1274
Location: Chesterfield, UK

PostPosted: Wed Apr 22, 2015 11:18 am  Reply with quote

On 22/04/2015 18:17, debabrata.podder wrote:
Quote:
Just asking a simple question. I want to make a (Thermal/Structural) model where in one case, I want to take solid elements and in another case I want to take shell elements.
For the case of solid elements, only one element is considered along the thickness. Now if we consider shell elements instead of using solid ones, will it take same time to run the analysis or will it take less time ?
I think it will be problem dependent. If the shells are thin then
aspect ratio considerations means that the brick mesh will need to be
finer than the shell mesh and hence the brick mesh will take longer to
solve. If the mesh density is similar then I would expect the run time
to be broadly similar for one element through the thickness,
--
Martin Liddle, Tynemouth Computer Services, Staveley, Chesterfield, S43
3TW, UK.
+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
_________________
Martin Liddle, Tynemouth Computer Services,
Chesterfield, UK.
Back to top
View user's profile Send private message Visit poster's website
zheng.kang
User


Joined: 04 Jun 2014
Posts: 24

PostPosted: Wed Apr 22, 2015 12:24 pm  Reply with quote

if everything else is the same, I would expect the shell element model
takes less time, because shell model will have way less elements.

On Wed, Apr 22, 2015 at 2:18 PM, Martin Liddle <xansys05@tynecomp.co.uk>
wrote:

Quote:
On 22/04/2015 18:17, debabrata.podder wrote:

Quote:
Just asking a simple question. I want to make a (Thermal/Structural)
model where in one case, I want to take solid elements and in another case
I want to take shell elements.
For the case of solid elements, only one element is considered along the
thickness. Now if we consider shell elements instead of using solid ones,
will it take same time to run the analysis or will it take less time ?

I think it will be problem dependent. If the shells are thin then aspect
ratio considerations means that the brick mesh will need to be finer than
the shell mesh and hence the brick mesh will take longer to solve. If the
mesh density is similar then I would expect the run time to be broadly
similar for one element through the thickness,
--
Martin Liddle, Tynemouth Computer Services, Staveley, Chesterfield, S43
3TW, UK.

+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+




--
Zheng Kang
Illinois Institute of Tech
Chicago, IL, USA
+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
Back to top
View user's profile Send private message
debabrata.podder
User


Joined: 21 Feb 2013
Posts: 51
Location: NIT Meghalaya, India

PostPosted: Wed Apr 22, 2015 12:58 pm  Reply with quote

Thanks Martin and Zheng for replying.
@ Zheng: Suppose the no. of elements are same for shell and solid elements. Then ??? I mean if along thickness only one element exist for solid elements model.

Regards-
_________________
Debabrata Podder(PhD)
Assistant Professor
NIT Meghalaya
India
Back to top
View user's profile Send private message
christopher.wright
User


Joined: 17 Jun 2009
Posts: 927

PostPosted: Wed Apr 22, 2015 8:00 pm  Reply with quote

On Apr 22, 2015, at 2:24 PM, Zheng Kang wrote:

Quote:
if everything else is the same, I would expect the shell element model
takes less time, because shell model will have way less elements.

Traditionally, the solution time depends on the total number of
degrees of freedom and the wavefront. Wavefront minimization has gone
the way of the floppy disk-suffice to say that if you're sloppy about
defining the mesh generation, you'll get a slower problem. Solid
elements typically have 3 DOFs per node which plte elements have 6
DOFs per node. With the same number of DOFs the solution times should
be equal, but you'll never have the same number of DOFs because
you'll probably need 4 solid elements in the through thickness
dimension to simulate (properly) the stress gradient in that
direction. Four solid elements at some point gives five nodes through
the thickness making 15 DOFa. A plate element will only have 6, so
you'd expect the plate element model to solve faster. Conventional
wisdom is that there's an additional penalty for higher order solids,
although maybe your mesh needn't be so small.


Christopher Wright P.E. |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
.......................................| John Sedgwick, Spotsylvania
1864)
http://www.skypoint.com/members/chrisw/



+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
Back to top
View user's profile Send private message
christopher.wright
User


Joined: 17 Jun 2009
Posts: 927

PostPosted: Wed Apr 22, 2015 8:02 pm  Reply with quote

On Apr 22, 2015, at 2:58 PM, debabrata.podder wrote:

Quote:
Then ??? I mean if along thickness only one element exist for solid
elements model.
This means that you're not able to capture through-thickness strain
variations with accuracy. It's a bad idea.


Christopher Wright P.E. |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
.......................................| John Sedgwick, Spotsylvania
1864)
http://www.skypoint.com/members/chrisw/



+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
Back to top
View user's profile Send private message
debabrata.podder
User


Joined: 21 Feb 2013
Posts: 51
Location: NIT Meghalaya, India

PostPosted: Wed Apr 22, 2015 9:34 pm  Reply with quote

Thanks Christopher for your time. Actually in the high stress gradients I am using more no. of solid elements (SOLID 70/SOLID185) and in farfield zones I am using less (1 element along thickness). My plate thickness is very less also (3mm). So I was thinking if I use shell elements in the farfields, instead of using solid elements there, if it would have make any difference (reduce the solution time) !


Regards-
_________________
Debabrata Podder(PhD)
Assistant Professor
NIT Meghalaya
India
Back to top
View user's profile Send private message
christopher.wright
User


Joined: 17 Jun 2009
Posts: 927

PostPosted: Wed Apr 22, 2015 9:42 pm  Reply with quote

On Apr 22, 2015, at 11:34 PM, debabrata.podder wrote:

Quote:
So I was thinking if I use shell elements in the farfields, instead of using solid elements there, if it would have make any difference (reduce the solution time) !
As I said--check to see how many DOFs you're saving. You might try wavefront minimization as well if the commands still exist.

Christopher Wright P.E. |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
.......................................| John Sedgwick, Spotsylvania 1864)
http://www.skypoint.com/members/chrisw/



+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
Back to top
View user's profile Send private message
debabrata.podder
User


Joined: 21 Feb 2013
Posts: 51
Location: NIT Meghalaya, India

PostPosted: Wed Apr 22, 2015 10:56 pm  Reply with quote

Dear Christopher,
Can you give me some idea about the wavefront minimization technique ? How to use that? Is it possible to give an APDL example?

Regards-
_________________
Debabrata Podder(PhD)
Assistant Professor
NIT Meghalaya
India
Back to top
View user's profile Send private message
christopher.wright
User


Joined: 17 Jun 2009
Posts: 927

PostPosted: Wed Apr 22, 2015 11:25 pm  Reply with quote

On Apr 23, 2015, at 12:56 AM, debabrata.podder wrote:

Quote:
Can you give me some idea about the wavefront minimization technique ? How to use that? Is it possible to give an APDL example?
You need to read the manual梥ubject 'wavefront.'

Christopher Wright P.E. |"They couldn't hit an elephant at
chrisw@skypoint.com | this distance" (last words of Gen.
.......................................| John Sedgwick, Spotsylvania 1864)
http://www.skypoint.com/members/chrisw/



+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
Back to top
View user's profile Send private message
weizhuo.du
User


Joined: 16 Jun 2013
Posts: 13
Location: Shanghai

PostPosted: Thu Apr 23, 2015 2:09 am  Reply with quote

Hello Debabrata

If you want to save time by using shell and get some result just belong to
solid you can use shell-solid submodeling.

Weizhuo.du
PeraGlobal

-----邮件原件-----
发件人: Xansys [mailto:xansys-bounces@xansys.org] 代表 debabrata.podder
发送时间: 2015年4月23日 1:17
收件人: xansys@xansys.org
主题: [Xansys] [APDL] Will it take same time for solid and shell elements ?

Dear experts,

Just asking a simple question. I want to make a (Thermal/Structural) model
where in one case, I want to take solid elements and in another case I want
to take shell elements.
For the case of solid elements, only one element is considered along the
thickness. Now if we consider shell elements instead of using solid ones,
will it take same time to run the analysis or will it take less time ?

Regards-

------------------------
Debabrata Podder
Research Scholar
IIT Kharagpur
PIN NO.: 721302
India






+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

+-------------------------------------------------------------+
| XANSYS web - www.xansys.org/forum |
| The Online Community for users of ANSYS, Inc. Software |
| Hosted by PADT - www.padtinc.com |
| Send administrative requests to xansys-mod@tynecomp.co.uk |
+-------------------------------------------------------------+

Post generated using Mail2Forum (http://www.mail2forum.com)
_________________
Pip Du
Application Engineer
Pera Global
Back to top
View user's profile Send private message
patrick.tibbits
User


Joined: 27 Apr 2009
Posts: 38

PostPosted: Thu Apr 23, 2015 12:45 pm  Reply with quote

The SOLSH190 solid-shell elements are available: (8 nodes)(3 DOF/node) = 24 DOF. Compare to shells: (4 nodes)(6 DOF/node) = 24 DOF. SOLSH190s have been tested for sheetlike structures at one element through-thickness, where the structure is monolithic (single material). You can probably find documentation of the validation on Ansys.com.

The SOLSH190s do contact like solids, but solve as fast as shells.

Patrick Tibbits
Cobham
Back to top
View user's profile Send private message
Display posts from previous:   
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies.    XANSYS Forum Index -> XANSYS
All times are GMT - 7 Hours
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You cannot download files in this forum


Powered by phpBB © 2001, 2005 phpBB Group

sleek template created by Andrew Charron