SPC Forces when running analysis with CGAP different then using RBE3 1

[brailos]

Hello everybody,

I have a small problem using CGAP elements. I use NX 7.5 and trying to output the reaction forces but when summing up the forces i have they are not the same with the Loads.

I am using CGAP elements in a linear analysis with SOL 101.
Solution parameters used:

AUTOMPC YES
BAILOUT -1
CDITER: 40
CDPCH YES
CDPRT NO
GRDPNT 0

For the CGAP properties i used :
Initial Gap Opening: 0.01
Axial Stiffness for closed Gap : 19000000
Axial Stiffness for open Gap : 19
Transverse stiffness when gap closed: 1900000.

All the Gap elements have been used to simulate shafts in lugs.


The reaction forces are much smaller the the initial loads.
I made a simple model with a simple lug and a shaft with Cbar and CGAP and the initial Load was 25000N and the reaction forces in the 2 constrained nodes where somewhere 9200N each node.
I made the exactly same thing with RBE3 elements instead and i got the correct reaction forces and a stress with 12% higher.

Not only that i need the correct SPC forces to prove the model is correct but it makes me doubt the stress values.

Any ideas? Please help me.

 

[brailos]

UPDATE!

Funniest thing happened, i kept trying changing parameters to see if anything affects the Reaction Forces, and i found out that by removing these parameters:

CDITER: 40 (this alone changes things)
CDPCH YES
CDPRT NO
GRDPNT 0

i get the correct reactions , plus the stress was at the exact same value.

Does anyone has an answer for this? , because i don;t understand how CDITER that defines the maximum number of iterations can affect the reaction force.

 

[Walterke]

my guess:

the required iterations to come up with the correct values is more then 40. So if you fill in 40 iterations in the variable, the calculation will stop too soon, giving you the wrong results.

NX 7.5
Teamcenter 8

 

[brailos]

I believed that too..let me explain better.
I have a big model with 4 shafts modeled with CGAP and CBAR, and today i realized the reactions are wrong.In that specific model CDITER was 40, but in the .f06 file all the subcases (12) are converging, it says that at the end of each subcase and they barely get to 20-22 iterations so 40 is actually to big not to low.
And because it took a lot of time to run that model , i made a smaller model with a singe lug and a single shaft.
And in the smaller model i declared CDITER to 10 , 20 and 40 and i get the same results. And the funny thing is that even if i declare a CDITER of 10 , in the fo6 file i see that the iterations goes until 12 , when it converges. So in conclusion it doesn't even take into consideration the maximum number of iterations because it passes it , which is kind of stupid thing.
If i remove the parameter CDITER i see the same text in the f06 file with the exact number of iterations ,converging at the end.
So i am in a kind of a fog here, because i realized not only that adding the CDITER parameter will change the reaction forces, but it does not even taking it into account passing the maximum number of iterations i declare.

Is this a model error?a local error?Or it is a bug in NX Nastran?

 

[AIUK]

When I need to simulate contact in a LINEAR (SOL 101) analysis, I normally make use of linear gaps (combination of SUPORT cards and MPC elements) calling the CDITER parameter that make possible an iteration process inside a linear analysis.
CGAP is a feature generally used in non-linear analysis, its stiffness will be update as the iterations of the non linear analysis procede. When you select the param cditer you are actually selecting the alternative approach that makes use of linear gaps.

 

[BlasMolero]

Dear Brailos,
A few suggestions here about CGAP elements for linear static SOL101 analysis:
• Avoid at all the use of PARAM,BAILOUT,-1, this is for debugging purposes only to detect FE modeling errores, not for working models.
• When a BCSET case control command exists, CGAP elements are treated as linear contact elements, if not the CGAP element will be considered as a spring.
• You can use CGAP elements in linear static analysis (SOL101) together with surface-to-surface contact defined (BCTSET bulk entries exist), or without (no BCTSET bulk entries exist). In the case where no surface-to-surface contact is defined, “n” on the BCSET case control command can point to a BCTPARM bulk entry which optionally defines PENN, PENT or PENTYP for the CGAP/linear contact elements, or to nothing if a BCTPARM bulk entry does not exist (an integer value for “n” is still required in this case).

The you are a simply nastran input without the need to define BCTPARM entry, this way CONTACT WILL BE LIMITED TO 20 ITERATIONS MAXIMUM, if not convergence is achived then the solution stops:

INIT MASTER(S)
NASTRAN SYSTEM(442)=-1, SYSTEM(319)=1
ID cgap,Femap
SOL SESTATIC
TIME 10000
CEND
TITLE = NX Nastran Static Analysis Set
ECHO = NONE
DISPLACEMENT(PLOT) = ALL
SPCFORCE(PLOT) = ALL
OLOAD(PLOT) = ALL
FORCE(PLOT,CORNER) = ALL
STRESS(PLOT,CORNER) = ALL
BCRESULTS(TRACTION,FORCE,SEPDIS,PLOT) = ALL
SPC = 1
BCSET = 108
LOAD = 1
BEGIN BULK
$ ***************************************************************************
$ Written by : Femap with NX Nastran
$ Version : 10.3.1
$ Translator : NX Nastran
$ From Model : D:\MODELOS\test\cgap.modfem
$ Date : Thu Mar 08 13:06:35 2012
$ Output To : D:\Scratch
$ ***************************************************************************
$
PARAM,POST,-1
PARAM,OGEOM,NO
PARAM,AUTOSPC,YES
PARAM,K6ROT,100.
PARAM,GRDPNT,0
CORD2C 1 0 0. 0. 0. 0. 0. 1.+FEMAPC1
+FEMAPC1 1. 0. 1.
CORD2S 2 0 0. 0. 0. 0. 0. 1.+FEMAPC2
+FEMAPC2 1. 0. 1.
$ Femap with NX Nastran Load Set 1
PLOAD4 1 4 -1.
PLOAD4 1 2 -1.
$ Femap with NX Nastran Constraint Set 1
SPC1 1 123456 3
SPC1 1 123456 4
SPC1 1 123456 7
SPC1 1 123456 8
$ Femap with NX Nastran Property 1 : Mid-Surface t=5.
PSHELL 1 1 5. 1 1 0.
$ Femap with NX Nastran Property 2 : Mid-Surface t=5.
PSHELL 2 1 5. 1 1 0.
$ Femap with NX Nastran Property 3 : GAP Property
PGAP 3 0. 0.1000000. 0. 0. 0. 0.+
+
$ Femap with NX Nastran Material 1 : Acero (MPa)
MAT1 1 205000. .3 7.85-9 0. 0. +
+ 260. 260.
GRID 1 0 0. -15. 2.5 0
GRID 2 0 0. 15. 2.5 0
GRID 3 0 -50. 15. 2.5 0
GRID 4 0 -50. -15. 2.5 0
GRID 5 0 50. 15. 7.5 0
GRID 6 0 50. -15. 7.5 0
GRID 7 0 100. -15. 7.5 0
GRID 8 0 100. 15. 7.5 0
GRID 9 0 50. 15. 2.5 0
GRID 10 0 50. -15. 2.5 0
GRID 11 0 0. -15. 7.5 0
GRID 12 0 0. 15. 7.5 0
CQUAD4 1 1 1 2 3 4
CQUAD4 2 2 5 6 7 8
CQUAD4 3 1 9 2 1 10
CQUAD4 4 2 11 6 5 12
CGAP 5 3 5 9 1. 0. 0.
CGAP 6 3 6 10 1. 0. 0.
CGAP 7 3 11 1 1. 0. 0.
CGAP 8 3 12 2 1. 0. 0.
ENDDATA d50e6f3a

If you need to increase the numbr of contact iterations, then the contact control parameters on the BCTPARM bulk entry (optional) can help you adjust the contact algorithm when you are having problems getting a solution to converge and complete, or when the contact results are not as expected.

You can now optionally define multiple BCTSET/BCTPARM bulk entry sets, each set with unique contact set IDs (CSID), and then combine them with a single BCTADD bulk entry. The multiple BCTSET/BCTPARM bulk entry sets are created to adjust certain contact parameters locally. Contact parameters can also be adjusted globally with a BCTPARM bulk entry having the same CSID as the BCSET case control command.

For instance, if I want to increase the number of iterations to 50 with the NX NASTRAN solver then I can enter the following command in the bulk data:

BCTPARM,108,MAXS,50

Best regards,
Blas.

~~~~~~~~~~~~~~~~~~~~~~
Blas Molero Hidalgo
Ingeniero Industrial
Director

IBERISA
48011 BILBAO (SPAIN)
WEB:

http://www.iberisa.com

Blog de FEMAP & NX Nastran:

http://iberisa.wordpress.com/