PDA

View Full Version : Ansys hyperelastic + viscoelastic material trouble



rz_ca
2007-06-14, 17:57
Hi everyone,
I'm having trouble when using the combined hyperelastic material (neo-hookean) with viscoelastic (one term prony) in Ansys. The problem is a cone being pulled by uniform pressure at the top and fixed at the
bottom. Plane183 (8 node plane element with uniform reduced integration) is used with axisymmetric option. This element empolyed mixed u-p formation. The hyperelastic material is assumed to be incompressible (u=0.5) or nearly incompressible (u=0.4995). (actually if use u=0.4995, Ansys will not run at all, the element will not form at the first substep)

The model runs fine if only the neo-hookean hyperelastic
material (u=0.5) is used but will crash at about 30% of completion when viscoelastic material is combined. By looking at the hydrostatic pressure distribution in post-processing, there is obvious checkerboard pattern along the loading edge, so it seems that volumetric locking causes the analysis to stop.

My question is why this happens when viscoelastic material is combined but not happens when only hyperelastic material is used.

My guess is our version of Ansys is too limited (university intermediate version)?

So if anyone in this forum has higher version Ansys in his/her hand, can he/she kindly run this analysis for me and see if it's due to the version problem?


The attached input file ?scripts_e02r1pull? combines hyperelastic and viscoelastic material. The run will take only a couple of minutes. Thanks a lot.

rz


/BATCH
/FILNAME,e02r1_pull,0
/TITLE,e02r1

/PREP7

/com >> define element type and material
ET,1,PLANE183

KEYOPT,1,3,1
KEYOPT,1,6,1
KEYOPT,1,10,0

TB,HYPE,1,1,2,NEO
TBTEMP,0
TBDATA,,1,0,,,,

TB,PRONY,1,1,1,SHEAR
TBTEMP,0
TBDATA,1,0.4,14,,,,

/com TB,PRONY,1,1,1,bulk
/com TBTEMP,0
/com TBDATA,,0.4,2,,,,

/com >> define keypoint
K,1,0,0,,
K,2,0,1,,
K,3,1,0,,
K,4,0.1,1.012916581,,
K,5,0.117647058,0.993055471,,
K,6,0.12,0.992916581,,
K,7,0.6,0.992916581,,
K,8,0.1,1.6,,


/com >> generate quater circle
FLST,2,2,8
FITEM,2,0,0,0
FITEM,2,1,0,0
CIRCLE,P51X, , , ,90, ,

/com >> generate straight lines
LSTR, 9, 1
LSTR, 1, 10
LSTR, 8, 4
LSTR, 6, 7

/com >> generate small circle
LARC, 4, 6, 5

/com >> divide lines
LDIV, 1, 0.70

/com >> copy contact zone lines
LGEN,2,7, , ,0,-0.04, , ,0

/com >> shorten copied lines
LDIV, 8, 0.4539919309218844E-01
LDELE, 8

LSTR, 11, 14
LSTR, 10, 13


/com >> create area
AL,10,7,9,8

LSBL, 3, 9
AL,9,11,8,1,2

/com >> manual meshing the contact zone
/com >> transverse lines
LESIZE,8, , ,5, , , , ,1
LESIZE,7, , ,80, , , , ,1

/com >> mesh the area
MSHAPE,0,2D
MSHKEY,1
AMESH,1,1

SMRT,4
MSHKEY,0
AMESH,2,2

SAVE

/com >> apply surface load
DL,10, ,UX,0
DL,11, ,UX,0
FLST,2,68,1,ORDE,3
FITEM,2,1
FITEM,2,290
FITEM,2,-340
SF,P51X,PRES,-2.5

NSEL,S,node,,1418,1452
NSEL,a,node,,1372,1372
NSEL,a,node,,1373,1373

D,all,UY,0, ! APPLY DISPL
NSEL,ALL

/COM, CONTACT PAIR CREATION - START

/COM, CONTACT PAIR PROPERTIES - END

/sol
/com >>analysis parameters
ANTYPE,0
NLGEOM,1
NSUBST,100,10000,30
OUTRES,ALL,3
AUTOTS,1
NEQIT,80
RESCONTRL,DEFINE,ALL,3,10
TIME,1

/com >>message display control
/UIS,MSGPOP,3
/NERR,0,10000,


/com >>nonlinear diagnose
NLDIAG,MAXF,16
NLDIAG,NRRE,1
NLDIAG,EFLG,1
NLDIAG,CONT,1

/com >>run the job
SOLVE

/com 2nd step
/com nsubst,50
/com time, 2000
/com solve
FINISH
! /EXIT,ALL

Jorgen
2007-06-23, 19:55
I would be happy to run your simulation is you provide an input file for ABAQUS :wink:

Perhaps someone else can run it through ANSYS...

-Jorgen