Hello:

I have attached a text file where the model is written in Abaqus keyword language.

I am modeling proppant transport in 2D. Fluid, with proppants, is being injected into an already existing thin cohesive crack which is surrounded by porous reservoir. It is assumed that the entire reservoir has a constant depth of 30 m.

I am also assigning an initial value of 800 kPa to my entire reservoir (minus the fracture) as the pore pressure value. These are the stress values for my reservoir:

• sigma_xx = 12000 kPa

• sigma_yy = 8000 kPa

• sigma_xy = 15000 kPa

Also, for the inner nodes of my cohesive fracture (the nodes are located midway on two sides of the element COH2D4P), I am assigning an initial value of 820 kPa as the pore pressure value. This way, the fluid in the fracture will go from the fracture to the surrounding reservoir because of this positive pressure differential.

Now, the model is running fine. But, I am seeing some issues with it, mainly regarding the pore pressure values in my model.

First of all, during the entire duration of the simulation, the pore pressure value remains unchanged in the porous reservoir portion (800 kPa throughout the entire time). Nowhere, in my model, have I stated that pore pressure remain the same! I expect it change sometime as fluid flow happens.

And, the pore pressure inside the fracture, from the beginning to end of simulation, remain 0 kPa. I am not sure how that is possible. When I assign a specific pore pressure value to the mid-nodes inside the fracture, why does Abaqus not recognize that?

Second of all, I have noticed that when I assign initial pore pressure value of 800 kPa to the entire reservoir (not the fracture), the model runs just fine!

However, when I assign this 800 kPa pore pressure, initially, to just a part of the reservoir (say, some boundary nodes), the model does not converge and becomes unstable. Is it related somehow to any convergence criteria issue?

If you could help me with this, I would be grateful!

Thank you.