Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
PARAM SNORM in NASTRAN 2
- Thread starter Spirit
- Start date
![[surprise]](/data/assets/smilies/surprise.gif "[surprise] [surprise]")
? Thanks, bye.-
2
- #2
- Thread starter
- #3
OK, but it is strange that the SNORM parameter set to 20 also corrects an error in the calculation of shell structures modelled with QUAD elements with offset (the torsional stiffness was in fact underestimated in this situation)... I think they could have given more emphasis to this changes.
'Ability is 10% inspiration and 90% perspiration.'
'Ability is 10% inspiration and 90% perspiration.'
Spirit, i have forwarded your question to MSC and get following comments :
1. In Version 70.7 the default is 0.0. If param,snorm,0.0 is used for version 2001, it yields the same result as 70.7. However, 70.7 has a larger matrix to factor diagonal ratio of 2.8E+07. If you want and like the results of 70.7 but wants to run 2001, then I suggest using the parameter snorm set to 0.0.
2. The reason for using SNORM is that we are modelling smooth curved surfaces as facets.
This means that the drilling direction on one element couples with bending in adjacent elements on a curved plate. This causes the stiffness ratio problems. (which
is a bad thing) SNORM takes these elements and rotates the stiffnesses so that they use a common average normal. The drilling direction is then removed by AUTOSPC with no problems and no large stiffness ratios. In crude engineering terms this is like smoothing out the facets.
In a model where you have significant differences between 70.7 and 2001 it implies that you have significant loads being reacted through the plate normal rotation direction,
which is bad. Check your constraint forces for poor Autospc constraints and use the GROUNDCHECK case control request to verify your rigid body conditioning.
Switching SNORM to 0.0 will probably give more inaccurate results and worse rigid body behaviour so is not a good approach.
1. In Version 70.7 the default is 0.0. If param,snorm,0.0 is used for version 2001, it yields the same result as 70.7. However, 70.7 has a larger matrix to factor diagonal ratio of 2.8E+07. If you want and like the results of 70.7 but wants to run 2001, then I suggest using the parameter snorm set to 0.0.
2. The reason for using SNORM is that we are modelling smooth curved surfaces as facets.
This means that the drilling direction on one element couples with bending in adjacent elements on a curved plate. This causes the stiffness ratio problems. (which
is a bad thing) SNORM takes these elements and rotates the stiffnesses so that they use a common average normal. The drilling direction is then removed by AUTOSPC with no problems and no large stiffness ratios. In crude engineering terms this is like smoothing out the facets.
In a model where you have significant differences between 70.7 and 2001 it implies that you have significant loads being reacted through the plate normal rotation direction,
which is bad. Check your constraint forces for poor Autospc constraints and use the GROUNDCHECK case control request to verify your rigid body conditioning.
Switching SNORM to 0.0 will probably give more inaccurate results and worse rigid body behaviour so is not a good approach.
- Thread starter
- #5
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question
