Learning Forum

[manu87]

Hi,everybody. Inspired by the example VM278 in ANSYS help, I'm working on the problem that a beam is located within a cylinder. The beam is modeled by the Pipe288 element, while the cylinder is regarded as a rigid surface and meshed by the Targe170 element. A contact pair is established between the beam and cylinder, using Conta177& Targe170 elements. The boundary condition at the both ends are treated as a pin. In the first step, the gravity and a axial compress force is loaded. Then a rotational displacement is applied at the left end. 

However, it seems that the contact does not work, since the contact status and the penetration ara null. The beam deforms largely, and the radial displacement exceeds the extent of the cylinder.

Anybody would offer some help?

The APDL file is attached as follows.

finish

/clear

 

/PREP7

L=50 ! MODELING PARAMETERS

OD=0.5

WT=(0.5-0.4)/2

!*** SOLID MODEL ***

!

---

k,1,L,, 

K,2,

K,3,L,,-0.8/2

K,4,,,-0.8/2

L,1,2 ! GEOMETRY OF THE BEAM 

L,3,4

AROTAT,2,,,,,,1,2,360,

 

!*** MATERIAL PROPERTIES FOR STEEL IN NEWTONS AND METERS ***

!

---

MP,EX,1,2.1E11 ! YOUNG'S MODULUS

MP,PRXY,1,0.3 ! POISSON'S RATIO

MP,DENS,1,7800 ! DENSITY

MP,MU,1,0.35 ! COEFFICIENT OF FRICTION

 

!*** ELEMENT TYPES ***

!

---

ET,1,PIPE288 ! 3-D 3-NODE PIPE

ET,2,170 ! 3-D TARGET SEGMENT

ET,3,CONTA177 ! 3-D LINE-TO-SURFACE CONTACT

KEYOPT,3,2,1 ! PENALTY FUNCTION ALGORITHM

KEYOPT,3,3,2   ! TRACTION-BASED CONTACT TYPE

KEYOPT,3,7,1

KEYOPT,3,10,2

 

 

 

!*** SECTION DATA ***

!

---

! EXAMPLE SECTION:

SECTYPE,1,PIPE,,PIPE1 ! DEFINE PIPE SECTION

SECDATA,OD,WT,16 ! APPLY MODELING PARAMETERS TO PIPE SECTION

! *** REAL CONSTANTS ***

!

---

R,2,

RMODIF,2,3,-1e10 ! ABSOLUTE VALUE OF PENALTY STIFFNESS

 

 

 

! *** MESHING ***

!

---

TYPE,1   

MAT,1 

SECNUM,1

LESIZE,1,1,   

LMESH,1

 

*GET,dnodemax,node,,num,maxd

/pnum,node,1

 

 

asel,s,,,all

TYPE,2

REAL,2

amesh,all

 

 

 

! CONTACT/TARGET

TYPE,3 ! CONTACT TYPE

REAL,2

MAT,1

LSEL,S,,,1

ESLL

ESURF ! PLACE CONTACT ON PIPE ELEMENTS

esel,s,real,,2

ESEL,r,TYPE,,2

ESURF,,REVERSE  

 

 

/pnum,node,0

ALLSEL,ALL

FINISH

 

 

/SOLU

ANTYPE,4 

NLGEOM,1  

ACEL,,9.8

d,2,UX,,,,,UY,UZ

ddele,2,ux

f,2,fx,5e4

d,1,UX,,,,,UY,UZ 

KBC,0   

AUTOTS,1   

OUTRES,ALL,all 

TIME,2  

SOLVE

 

 

kbc,0   

*do,i,1/30,1,1/30

d,2,rotx,6.28*i

timint,on   

TIME,2+i

AUTOTS,1

solve  

*enddo

 

[John Doyle]

Try offsetting the beam slightly so the two are not perfectly concentric.

Also, use refined time increments with DELTIM.

[manu87]

Sorry for the delaying reply.

Thank you for your help. I tried your method, but it did not work.

Then I reread the help document, and the help shows that "CONTA177 is used to represent contact and sliding between 3-D target surfaces and a deformable line segment, defined by this element. The element is applicable to 3-D beam-to-surface, 3-D shell edge-to-surface, and 3-D beam-to-beam (or edge-to-edge) structural contact analyses. ". It seems that the case of 3-D beam-to-surface and 3-D beam-to-beam are distinguished. So would be there a probability that CONTA177 is just used in a non-closed surface, such as a half cylinder, or a half sphere?

[John Doyle]

Are you modeling the target as a rigid surface?

If so, I think your target element surface normals might be facing outward, so the contact elements are not seeing the target elements.

Can you try using beam to beam contact?  See See Section 5.3.1 of Contact Technology Guide.

 

[manu87]

Yes, the target surface points outwards defaultly, and it had been adjusted by the command ESURF,,REVERSE. Then the contact/target elements are seeing each other.

I'll try the second method you metioned. But, if any ideas you get about the CONTA177 element modeling, please contact me. Thank you.