My understanding is that auto line constraint allows you to connect a line element to a shell element node. It does this by internally providing a 'rigid' link to the next adjacent nodes. I've emailed them about this and they said that the exact formulation is proprietary but that is basically what it does. If you look at suggested modeling for software that doesn't have this feature you would see that you have to do something equivalent to this manually - Etabs (and SAP) is attempting to do it for you. A good way to get a sense of what it is doing is to look in the Risa help manual, there they have a good diagram about doing this manually (which is required in Risa) - now imagine that Etabs has internalized that. You can turn it off if you have a case where it isn't doing what you need.
I've found that you do need to be somewhat careful with this, as you change your mesh size you will see a big change in behavior at some point as that internal rigid link gets shorter and shorter. Overall though it seems to work fine.
bookowski, so you were saying is when there are two shell elements where the mesh lines between the two elements does not coincides with each other and the line constraint is activated then it will generate a rigid link between mesh lines, and if the auto line constraint is deactivated then the only rigid link is only to the location where the mesh lines coincides with each other. Please let me know if my understanding is correct
Check out this pdf from CSI, it's better than I can describe.
If it's all shells then I still try to mesh following the usual rules rather than trust the line constraint.
The only place where you need to use it is when you are connecting line elements to shell element, the most common example that I know of being link beams between shear walls (coupled walls). The line constraint allows you to directly connect a frame element to the wall elements without any further work.
and check this out as well, between these two it should be clear. If you are looking at meshing shells to shells the CSI paper is more relevant, if you are doing beams to shells this should help more. This is from the Risa help manual. If Etabs didn't have the line constraint feature you would have to do what Risa is showing here - Etabs is doing this internally - the specifics of how they don't tell you, but it seems to generally work.
