UMAT - How to compute Jacobian matrix?

[idaapm]

I'm implementing a 3-dimensional progressive damage model for FRC using Fortran. The current version of the UMAT accounts for damage initiation (based on Hashin’s criteria) and damage evolution (based on equivalent displacements). The input data to the UMAT subroutine includes the strain state, the estimated constitutive matrix based on the given material properties array PROPS and the computed data defined in the solution-dependent variable array STATEV. The output variables passed back to ABAQUS/Standard from the UMAT subroutine includes the trial stresses, the trial constitutive matrix, the tangent constitutive matrix (Jacobian matrix) and the updated values of the solution-dependent variable array.

Questions:
1. At the moment I'm having some issues regarding the computation of the Jacobian matrix (DDSDDE). What is the best way to compute the Jacobian matrix?

2. What solution-dependent variables should I used within the UMAT?

3. I’ve included the user material option UNSYMM to account for viscous regularization, however it's not working with this option. Should I change anything in order to be able to run my UMAT successfully? Would it be better to implement my own code for this? Should the option UNSYMM work likewise?

4. At the moment, the code only accounts for the elastic regime. If I want to go beyond this, what changes will I have to take in consideration?