ShiningEngineer
Student
- Oct 11, 2023
- 27
Hello,
I have a question about the feasibility of using the conventional Abaqus GUI without scripting.
The following case:
I have already set up a homogenised Abaqus model in which hyperfoam material cards/data are assigned to the elements. I have a hyperelastic case with very large deformations in a compression. In reality, I have a lattice structure consisting of unit cells. By homogenising and assigning the real material properties on the basis of test data, Abaqus no longer has to simulate the complex filigree structures, but only the elements of the discretisation to which I have explicitly assigned the material properties. This also works so far in super correlation to experimental data.
Now my question,
Assuming I have 4 cases of ‘real’ lattice variants’, this results in 4 material maps that I have available.
Can I perform an optimisation for this in Abaqus? I can create an objective function, but how do I tell Abaqus to assign the created ‘materials’ one after the other during the optimisation and output the ‘best’ solution of my objective function? Is this possible without scripting or do you recommend that I implement this myself in Python? In itself it is quite simple, but I think Abaqus has not implemented such a function to control the ‘materials’.
My aim (see attached sketch) is to select areas 1 to 4 from the material TPU1 TPU2 TPU3 and TPU4 in order to fulfil a target function or come as close to it as possible. This could look like the following, for example. With a compression of 20'% or 30%, I would like to have a necessary reaction force of 8 kPA. A simple cube can be assumed in order to reduce the complexity as much as possible.
Regards
Info: When I talk about materials, I am talking about real materials in Abaqus with their own TEst data and Poisson ratio. I am using hyper foam material model.
I have a question about the feasibility of using the conventional Abaqus GUI without scripting.
The following case:
I have already set up a homogenised Abaqus model in which hyperfoam material cards/data are assigned to the elements. I have a hyperelastic case with very large deformations in a compression. In reality, I have a lattice structure consisting of unit cells. By homogenising and assigning the real material properties on the basis of test data, Abaqus no longer has to simulate the complex filigree structures, but only the elements of the discretisation to which I have explicitly assigned the material properties. This also works so far in super correlation to experimental data.
Now my question,
Assuming I have 4 cases of ‘real’ lattice variants’, this results in 4 material maps that I have available.
Can I perform an optimisation for this in Abaqus? I can create an objective function, but how do I tell Abaqus to assign the created ‘materials’ one after the other during the optimisation and output the ‘best’ solution of my objective function? Is this possible without scripting or do you recommend that I implement this myself in Python? In itself it is quite simple, but I think Abaqus has not implemented such a function to control the ‘materials’.
My aim (see attached sketch) is to select areas 1 to 4 from the material TPU1 TPU2 TPU3 and TPU4 in order to fulfil a target function or come as close to it as possible. This could look like the following, for example. With a compression of 20'% or 30%, I would like to have a necessary reaction force of 8 kPA. A simple cube can be assumed in order to reduce the complexity as much as possible.
Regards
Info: When I talk about materials, I am talking about real materials in Abaqus with their own TEst data and Poisson ratio. I am using hyper foam material model.
