tossin
Bioengineer
- Mar 3, 2011
- 2
Hello,
I am trying to model an idealized version of a fistula in ANSYS, which basically consists of an axisymmetric (axis = z) tube,the artery, with a branch jutting out from the side (symmetric across x and y-planes, but not axisymmetric), the vein. Size of the main tube is ~30mm, with inner radius ~ 2mm, outer ~ 3mm. All I want to see is the deformation/stress that occurs when I pressurize the inner surface of the object (up to 13000 Pa).
For now for simplicity, I have kept both the tube and the branch the same material (neo-Hookean, neoprene rubber) and fixed the three ends, though ideally, the tube and the branch would have different properties, and the three ends should only be fixed at one specific node/element.
The mesh itself has been imported from Hypermesh and built using second-order tetrahedral elements. In Hypermesh, the branch and tube were exported as Nastran .bdf files separately.
My problem is that the solution doesn't converge. ANSYS can simulate the tube and the branch separately, but not together (though if I change the object into a linear material like default steel, it is easily solved). I've tried greatly lowering the pressurization level and got convergence at 5 Pa, but not 50 Pa (over 3 seconds, time step = 1s).
Warnings I've received as messages or in the Solver Output:
- Solver Pivot Warnings encountered during solution as a result of ill-conditioned matrix possibly due to unreasonable material properties, under-constrained model, or contact-related issues.
- Element type 1/2 is SOLID187. It is not associated with fully incompressible hyperelastic materials. No suggestion is available and no resetting is needed. (I assume this is the tetrahedral element chosen in Hypermesh, but I don't think there is anything I can/should change here.)
- Material number 4 (used by element 65515 ) should normally have at least one MP or one TB type command associated with it. Output of energy by material may not be available. (I am not very clear on what this means.)
- The closed gap/penetration may be too large. Increase pinball if it is a true closed gap/penetration. Decrease pinball if it is a false one.
- The initial penetration/gap is relatively large. Bonded/no separation option may cause an accuracy issue. Switch to MPC algorithm or you may use the CNCHECK,ADJUST command to move the contact nodes towards the target surface.
- The closed gap may be too large. Increase pinball if it is a true closed gap. Decrease pinball if it is a false one.
The problem clearly seems to have to do with the contact/connections between the nodes of two objects, but I am unsure of how to proceed. Does anyone have any suggestions? Thank you for your time.
I am trying to model an idealized version of a fistula in ANSYS, which basically consists of an axisymmetric (axis = z) tube,the artery, with a branch jutting out from the side (symmetric across x and y-planes, but not axisymmetric), the vein. Size of the main tube is ~30mm, with inner radius ~ 2mm, outer ~ 3mm. All I want to see is the deformation/stress that occurs when I pressurize the inner surface of the object (up to 13000 Pa).
For now for simplicity, I have kept both the tube and the branch the same material (neo-Hookean, neoprene rubber) and fixed the three ends, though ideally, the tube and the branch would have different properties, and the three ends should only be fixed at one specific node/element.
The mesh itself has been imported from Hypermesh and built using second-order tetrahedral elements. In Hypermesh, the branch and tube were exported as Nastran .bdf files separately.
My problem is that the solution doesn't converge. ANSYS can simulate the tube and the branch separately, but not together (though if I change the object into a linear material like default steel, it is easily solved). I've tried greatly lowering the pressurization level and got convergence at 5 Pa, but not 50 Pa (over 3 seconds, time step = 1s).
Warnings I've received as messages or in the Solver Output:
- Solver Pivot Warnings encountered during solution as a result of ill-conditioned matrix possibly due to unreasonable material properties, under-constrained model, or contact-related issues.
- Element type 1/2 is SOLID187. It is not associated with fully incompressible hyperelastic materials. No suggestion is available and no resetting is needed. (I assume this is the tetrahedral element chosen in Hypermesh, but I don't think there is anything I can/should change here.)
- Material number 4 (used by element 65515 ) should normally have at least one MP or one TB type command associated with it. Output of energy by material may not be available. (I am not very clear on what this means.)
- The closed gap/penetration may be too large. Increase pinball if it is a true closed gap/penetration. Decrease pinball if it is a false one.
- The initial penetration/gap is relatively large. Bonded/no separation option may cause an accuracy issue. Switch to MPC algorithm or you may use the CNCHECK,ADJUST command to move the contact nodes towards the target surface.
- The closed gap may be too large. Increase pinball if it is a true closed gap. Decrease pinball if it is a false one.
The problem clearly seems to have to do with the contact/connections between the nodes of two objects, but I am unsure of how to proceed. Does anyone have any suggestions? Thank you for your time.