Actually, it is bizarre, i completely agree with Mohammad Zaid check your boundary condition and also try it with Abaqus.

Perhaps you are getting a correct answer.

Soils are much happier in horizontal layers than in steep slopes, and they tend to move down hill over time especially in moist environment, or with lots of freeze thaw cycles. I have been watching some fence posts with barbed wire fence placed vertically into a steep slope along a roadway. In the last twenty five years the post have gone from vertical, to nearly horizontal as the soils creep down the slope.

A similar process occurs at sloped bridge abutments covered with concrete slope pavement. The soil is not confined vertically, or horizontally on any slope, so it creeps down hill.

Michael Clark

Thanks for the answer. After many hours of work I am started to believe that this could be the correct behaviour. At sides of the soil domain there is only dashpots so there is no rigidity at the sides of the soil domain. During the dynamic analysis soil tends to make lateral deformation at the sides. You agree with that right?

Yes it deforms both laterally, and vertically. A slope will slump at the top. and bulge at the bottom. Look at the U-TUBE Videos of landslides, which is the extreme limit of soil movement. Look at Google Earth, in some of the major cities

where bridges cross over the Freeways ( Interstate Highways ). It was Very common for bridges built in the 1960's, thru 1990's to have concrete slope paving

that bulges , with cracks, at the bottom, leaving a void under the slope paving at the top. Colorado Bridges along I-25, and I-70 are good examples as there is a lot of freeze-thaw cycles in Colorado Winters that expand and contract soils with moisture in them.

Currently, Newer Bridges, are being built with vertical abutments on three sides, which introduces a whole new set of problems related to settlement of approach slabs, especially if the approach slabs are too short. The State of Louisiana is now building bridges with 40 foot long approach slabs, that are made "too high" so that in a few years, they will settle down to a correct approach elevation.

When Modeling, try making your grid squares parallel to the slope. The compressive forces parallel to the slope will make the squares shorter, at the bottom, and longer at the top, and turn them into parallelograms. The area of each grid square will remain the same, but the shape of each grid square will change. Shorter, and fatter at the bottom, but longer and more slender at the top.

Thus, the bottom bulges up, while top slumps down.

Mike Clark

Golden Colorado

Thank you for detailed answer sir.