LEM vs FEM

[massagung]

hi all,

as we know LEM (limit Equilibrium Method) as one of ways to define SF on Slope Stability Analysis (Bishop, Janbu, Fellenius, Spencer, Sarma), and currently we also know FEM (Finite Element Method to define SF on Slope Stability Analyis, as my limited of experience about both of this method, here's the question :
1. could anyone to explain what are the fundamental differences between this both method?
2. how about the difference of SF if we do analysis with LEM and FEM? is the difference of SF still reliable (around 5%)?
3. what is your prefer method in design LEM and FEM? and why?

thanks in advanced,

M.A.

 

[ntschwanz]

1. Too much to say to try to cover here but there are two main concepts you should consider: LEM is based on plasticity and you are solving directly for a SF for a trial slip surface and then you need to check other trial surfaces to find the slip surface with the lowest SF; FEM is indirectly solving for SF by the non-convergence of the model through the strength reduction process where this leads to large strain and the contours of this strain identifies the slip surface (no trial slip surfaces to assume and no searching). There are well respected individuals who suggest that you should always use the FEM while there are others that say the opposite.

2. See the paper by Griffiths and Lane on "Slope stability analysis by finite elements", Geotechnique 49, no. 3. I'm more confident in getting comparible results where the models do not contain structural elements than when they do. I say this because the introduction of structural elements can lead to differing failure mechanisms.

3. LEM tools have been developed for practicing engineers and you can find these that make setting up and running a model quick and relatively easy (perhaps too easy where non-geotech engineers can do this - but that's another subject). The FEM software that I'm familiar with doesn't have all the bells and whistle's as the LEM software and it takes me longer to create and run the model I want. I use the LEM for normal slope stability problems and move to the FEM method for complex or SSI type problems or where I want to verify uncertain results from a LEM analysis. Even with the FEM models it's wise to check with a LEM model for similar failure mechanisms. This is similar to checking one LEM analysis using another LEM software. In other words, I use both but the FEM is not routine for me for slope stability.

 

[massagung]

thanks a lot ntschwanz for your input, i will learn more to get clearly concept about this.,


M.A.