andersonbarata
Marine/Ocean
- Feb 3, 2003
- 1
Hi, mates.
I'm trying to estimate the contact forces in a system composed by a drilling column in a marine riser. I'm using ABAQUS 6.2 under academical license. I'm not a novice user, but I've experienced a problem that has burnt my nerves. Let me share it. I've used beam elements B31H (to build the column and the riser) and tube support elements ITSCYL. The solver has given me a number of "negative eigenvalues" warnings, in spite of existing steady traction force across the whole length of the beam elements. The riser is properly constrained at its edges, it is under self weight, pulling force at the top edge and drag forces across its length. The drillpipe is pinned at the upmost edge and a spring at the downmost node replaces the stiffness of the continuation of the drillpipe; there is a pulling load at its upmost edge and self-weight. I see no obvious cause for the numerical problem. If I remove the coupling between the drillpipe and the riser, by means of MODEL CHANGE (to remove the ITSCYL elements), the analysis runs very fast and it produces valid results. This suggests that the problem may lie on the coupling or on the type of element I have used. Further details: in the first attempt and first iteration, the solver produces the warning of one negative eigenvalue. In the second attempt, it warns that 3620 negative eigenvalue exists. If someone can give me hints to overpass this problem, I'll be very grateful.
Anderson
I'm trying to estimate the contact forces in a system composed by a drilling column in a marine riser. I'm using ABAQUS 6.2 under academical license. I'm not a novice user, but I've experienced a problem that has burnt my nerves. Let me share it. I've used beam elements B31H (to build the column and the riser) and tube support elements ITSCYL. The solver has given me a number of "negative eigenvalues" warnings, in spite of existing steady traction force across the whole length of the beam elements. The riser is properly constrained at its edges, it is under self weight, pulling force at the top edge and drag forces across its length. The drillpipe is pinned at the upmost edge and a spring at the downmost node replaces the stiffness of the continuation of the drillpipe; there is a pulling load at its upmost edge and self-weight. I see no obvious cause for the numerical problem. If I remove the coupling between the drillpipe and the riser, by means of MODEL CHANGE (to remove the ITSCYL elements), the analysis runs very fast and it produces valid results. This suggests that the problem may lie on the coupling or on the type of element I have used. Further details: in the first attempt and first iteration, the solver produces the warning of one negative eigenvalue. In the second attempt, it warns that 3620 negative eigenvalue exists. If someone can give me hints to overpass this problem, I'll be very grateful.
Anderson