transport velocity in soil

[Shuaihao-Zhang]

Remove the tensile instability in cohesive soil using transport velocity.

[Xiangyu-Hu]

I am wondering why we need this new class. Could we use a generalization-extension approach to add this new feature?

[Shuaihao-Zhang]

I am wondering why we need this new class. Could we use a generalization-extension approach to add this new feature?

Yes. It's better to use a generalized approach.
A simple approach is to just add a judgment statement in the original TransportVelocityCorrection.
When a particle is identified as a surface particle, corrections are applied only in the tangential direction. However, it is uncertain whether this approach will affect other cases.

[Xiangyu-Hu]

No. Using a switch is not a generalization-extension approach. We need first find a type in the original formulation such that it can be generalized, then change it to a template parameter (as a templated class) and reimplement the original class by realizing the template parameter with a small function object. For extension, we write the new feature as a new function object which substitute the template parameter in the generalized templated class. In this way, the choose new feature is don in compiling time not runtime.

[Shuaihao-Zhang]

No. Using a switch is not a generalization-extension approach. We need first find a type in the original formulation such that it can be generalized, then change it to a template parameter (as a templated class) and reimplement the original class by realizing the template parameter with a small function object. For extension, we write the new feature as a new function object which substitute the template parameter in the generalized templated class. In this way, the choose new feature is don in compiling time not runtime.

I will think about this.