For running abaqus, we should determine time period and then increment. I don't know how to set increment. Thank you very much for your help.
The option influences your Newton Raphon solver. You have a solution at t and you want the solution at t + dt. To obtain a stable solution, sometimes, the increment have to be very small.
If you chose a changing increment, Abaqus will increase or reduce your increment to find a solution. E.g. if you run a linear problem the increment will be increased to the highest number possible, because a stable solution is easy to find. However, if you have a non-linear problem Abaqus will adapt the increment to reach a stable solution.
The fixed step width allows it to define the increment by yourself.
Most of the times the changing increment option works well and should be used.
The option influences your Newton Raphon solver. You have a solution at t and you want the solution at t + dt. To obtain a stable solution, sometimes, the increment have to be very small.
If you chose a changing increment, Abaqus will increase or reduce your increment to find a solution. E.g. if you run a linear problem the increment will be increased to the highest number possible, because a stable solution is easy to find. However, if you have a non-linear problem Abaqus will adapt the increment to reach a stable solution.
The fixed step width allows it to define the increment by yourself.
Most of the times the changing increment option works well and should be used.
thank you. I'm talking about dynamics problem.
in a dynamic problem how i determine time increment?
which one i can select for correct answer in my solution : 'unlimited' option or adjusting value.
whats difference between 'global' and 'element-by-element' ?
thank you very much.
I think you talk about the explicit solver of Abaqus. An explicit time integration schema need small time increments to be stable. The highest eigenfrequency of your model gives you the step width. You find the formula in the Abaqus manual as well as in literature. Global and element by element is the method to estimate this frequency.
If you really are talking about Explicit dynamics, in fact the issue it is the Courant criterion.
http://en.wikipedia.org/wiki/Courant%E2%80%93Friedrichs%E2%80%93Lewy_condition
Advice: try to use fixed size elements, avoid smaller ones.
Read about so called mass scaling (ABAQUS manual)
what does it mean by the error "The increment required is less than the minimum specified." ?
Hi Soma, it means your minimum specified increment is not small enough for abaqus to get a converged solution. Simply you need to reduce the minimum increment value.
It depends on your solver.In explicit solver Curant number automatically determined the step time but you can change it.In implicit solver, it depends on the natural frequency of the problem or body.This number is so important in high-frequency problems.For the error,in the implicit solver, you can set minimum maximum and initial time step.Be careful about this numbers, the implicit solver is so costly solver and take a long time to solve the problem.Also,if can solve your problem in a quasi-static situation.I strongly suggest to you to read carefully the getting start doc.
How can find correct Increment Number?
or generally how fill increment tab for steps?
thanks for your helps
Oh....almost 4 month ago
Any way, in Explicit you cannot do more. But in implicit solver , it depends on your loading frequency. 1/20 of your shortest loading period time (According to ANSYS Docs) is a good estimate. Again, I insist, this is very tricky and needs a lot of experiences in each field.
In my experience, the time increment usually depends on your scheme: implicit or explicit. For explicit scheme, this value must be small enough. While for implicit scheme, you can switch the time increment to ensure the convergence.
Sincerely yours,
Xuan Nam Do
See the link below:
http://en.wikipedia.org/wiki/Courant%E2%80%93Friedrichs%E2%80%93Lewy_condition
- Badges
- Science topic