Add coupling fields used for CFD-DEM

[davidscn]

Adds

• the drag force
• explicit and implicit momentum fields
• writing the pressure gradient

[davidscn]

This change is critical, but necessary. In previous versions, gradP was a scalar for some reason. Maybe we should introduce a normal gradient separately.

[MakisH]

According to our versioning scheme, this would be a breaking change / major version jump. But I also see the point that this is wrongly defined (or with a very restrictive assumption).

I also stumbled upon this in another context, and we should indeed make a separate normal gradient. We also have other gradients that need reconsideration or renaming (AlphaGradient, TemperatureGradient, VelocityGradient).

I would suggest that, as a first step, we rename this to PressureGradientVector (or a better name), so that it can be merged as-is and released in a v1.4.0, and then we revamp the fields in a v2 of the adapter.

[MakisH]

Many nice new features, thanks! Remember to also update the documentation.

I have not looked into details yet. Is there any test case already available?

Note that one of the major updates we were planning with @vidulejs for v2 was the generic readers/writers (#138, #279, #295, #346), and this PR gives us more examples to consider.

[MakisH]

According to our versioning scheme, this would be a breaking change / major version jump. But I also see the point that this is wrongly defined (or with a very restrictive assumption).

I also stumbled upon this in another context, and we should indeed make a separate normal gradient. We also have other gradients that need reconsideration or renaming (AlphaGradient, TemperatureGradient, VelocityGradient).

I would suggest that, as a first step, we rename this to PressureGradientVector (or a better name), so that it can be merged as-is and released in a v1.4.0, and then we revamp the fields in a v2 of the adapter.