Difference between revisions of "Talk:Composite analysis"
| (5 intermediate revisions by 3 users not shown) | |||
| Line 4: | Line 4: | ||
Is this information in other wiki? | Is this information in other wiki? | ||
| + | |||
| + | ----- | ||
Well maybe you should kick it off with an example and others can tune in with improvements | Well maybe you should kick it off with an example and others can tune in with improvements | ||
/C | /C | ||
| + | |||
| + | ----- | ||
| + | |||
| + | Element size seems reasonable, if the results appears strange later on, it's just a matter of using code_aster to convert it to quadratic elements - but I'm no expert. | ||
| + | |||
| + | Well as you say; units matter, but I would prefer [m] in this case for some reason :) | ||
| + | |||
| + | /C | ||
| + | |||
| + | ----- | ||
| + | |||
| + | Thanks. It's nice to hear a voice with comments. | ||
| + | |||
| + | ---- | ||
| + | |||
| + | Hi, I suggest that you use quadratic/second order FE elements for your analysis, they usually give you the best accuracy. | ||
| + | |||
| + | If the shell in thin (span to thickness ration > 30) and relativelly flat Kirchoff plate element formulation should work fine (no transverse shear deformation) => elements called DKT in Code-Aster. | ||
| + | |||
| + | If the shell is thicker (span / thickness ration between 15 and 30) but curvature is negligible, the First order Shear Deformation Theory (FSDT) is recommended as it allows for transverse shear deformability => it is the DST FE formulation in Code-Aster (use ~0.8333 as shear correction factor for a better accuracy). | ||
| + | |||
| + | If the shell is thick and significantly curved, a real 3D shell formulation is required; in Code-Aster, these elements are called Coque3D. | ||
| + | |||
| + | To see an "old" example of a composite shell analysis, you can have a look at this archive: http://www.caelinux.com/downloads/examples/modal.tar.gz | ||
| + | |||
| + | Joël Cugnoni | ||
Latest revision as of 14:12, 31 March 2009
130 visitors and 0 contributions
Do you think that this is a good idea (comosite analysis wiki)?
Is this information in other wiki?
Well maybe you should kick it off with an example and others can tune in with improvements
/C
Element size seems reasonable, if the results appears strange later on, it's just a matter of using code_aster to convert it to quadratic elements - but I'm no expert.
Well as you say; units matter, but I would prefer [m] in this case for some reason :)
/C
Thanks. It's nice to hear a voice with comments.
Hi, I suggest that you use quadratic/second order FE elements for your analysis, they usually give you the best accuracy.
If the shell in thin (span to thickness ration > 30) and relativelly flat Kirchoff plate element formulation should work fine (no transverse shear deformation) => elements called DKT in Code-Aster.
If the shell is thicker (span / thickness ration between 15 and 30) but curvature is negligible, the First order Shear Deformation Theory (FSDT) is recommended as it allows for transverse shear deformability => it is the DST FE formulation in Code-Aster (use ~0.8333 as shear correction factor for a better accuracy).
If the shell is thick and significantly curved, a real 3D shell formulation is required; in Code-Aster, these elements are called Coque3D.
To see an "old" example of a composite shell analysis, you can have a look at this archive: http://www.caelinux.com/downloads/examples/modal.tar.gz
Joël Cugnoni
