It's not clear what you mean by the wall curving outward. Secondly, there are several external failure echanisms for MSE. Suggest that you think about whether the 2d situation is conservative vis-a-vis driving and resisting forces for a 3d failure surface. Most MSE failures that I am aware of involve bearing capacity/settlement rather than deep failures.

J
 

what I meant by 'curving outward' is that the wall face is on the convex side of a 'L' shape. The wall is built on a 6H:1V slope and the overall stability is quite critical. Before looking into the 3D approach, I want to have a feel on how much the 3d effect may have on the slope stability. Thanks a lot for your input.

I would ignore it from a benefit side.  The MSE is flexible, will not act the same a a concrete wall would in this situation.  You may get some benefit out of the reinforcing that is running back from the perpendicular face of the wall, but it will be limited to just that area.  It is probably not worth changing the design for just that area.  It could lead to things getting messed up in the field.

Also be aware of the grid to grid interfaces at the corner.  It is my experience that this does not provide the same strength as the grid with backfill on the top and bottom of it.  If it is a sharp bend, consider shifting the grid elevations between the two sides.

TDAA, Thanks a lot for your input. Could you please explain why the flexibility of the wall will affect the extent of 3D effect?

As to the geogrid/geogrid interface shear strength, we are aware of that. Our construction specification requires 0.5 ft of separation between the grids.

rocks09,

For 3D effects, I would expect that the surface area of the resisting surfaces in the immediate vicinity of the outward projecting corner would be greater than for the 2D "inifnite strip foundation" case.

It would be similar, to my thinking, to stresses near the corner of a flexible footing.

J
 

As jdonville indicated the area in the "immediate vicinity" would be greater.  Now the question becomes how large is that area?  That would be dependent on the load transfer in the wall.  The more flexible the wall, the less impact. The degree the effect will have on design / cost reductions also has a lot to do with the wall.  Are we talking a 10 foot, 70 foot, multi-tiered, etc.?

Hi Rock09,
I located Taiwan, we ever meet same question as yours. We solve the problem by Plaxis 3D program to realize the different of 2D and 3D effect. Our result indicate the 2D is more conservative than 3D. So we adopt 2D result for designing. But we found another interesting question inside. When we design MSEW, only consider the strength of geogrid at machine direction(MD) and ignore the strength of geogrid at cross machine direction(CD). Some casees show the failure at CD due to difference settlement occure. I think that you might consider the strength of XCD, not only MD strength.

just a generalized thought here: from my observations, the corners are problematic not from bearing standpoint but due to the difficulty in adequately getting the sharp corner "tied" together during construction due to the orientation of one face and grid versues the adjoining face/grid. also depending on the scenario, differential fill induced settlement could be problem due to the corner condition versus the adjacent full face and mass fill conditions.