Topic: IntegrationPoint vs. GaussPoint

When looking in the IntegrationRule class I see that we have the member
GaussPoint **gaussPointArray
but shouldn't this member rather be
IntegrationPoint **integrationPointArray

Furthermore, is it really necessary to have a GaussPoint class, why not just stick to IntegrationPoint? It seems like the term Gauss point is used interchangeably with integration point, which I can understand since Gauss integration is almost always used.

However, I currently use a mixed integration rule adopting a Gauss integration in two directions combined with a Lobatto integration in the third (for a layered shell element). I, therefore, don't have a GaussPoint or a LobattoPoint (which doesn't exist in the code) but rather a "HybridPoint". I don't want to introduce such a class and is currently treating my point as a GaussPoint, but I don't see a need for differentiating between integration point "types" - they only differ in location and weight.

Or am I missing something here?

Re: IntegrationPoint vs. GaussPoint

When you use typedef, e.g.
typedef GaussPoint IntegrationPoint;
all you are doing is creating a synonym. They can be used interchangeably in the code. GaussPoint is not meant to imply GuassIntegrationRule.

What this comes down to is; "Should we rename GaussPoint to IntegrationPoint ?"
That would definitely make things more consistent. I'm not to bothered either way.

Re: IntegrationPoint vs. GaussPoint

So to answer your question, yes I think it should be renamed for the sake of consistency. The fact that there is a typedef GaussPoint IntegrationPoint; is to me the statement "let's treat them the same".

Well I'm not to bothered either with the matter. I only brought it up as I noticed the inconsistent naming and was wondering if there was any thought behind it. Maybe I will get around to change it later.

Re: IntegrationPoint vs. GaussPoint

If we decide to change this, we should start by using IntegrationPoint everywhere, then rename the class itself and remove the typedef.