when modeling a secant pile wall as a plate element in plaxis 3D , what is most appropriate, to model it as isotropic ? or Anistropic, and if it is anisotropic what will be E1/E2 ratio? How much it depends in the in the interlock between piles.

Dear Mohammed,

It highly depends on what you want to model and how you expect to secant pile wall to be loaded whether you want to model it isotropic or anisotropic. If you expect the wall to have one predominant loading it may not be worth figuring out the E1/E2 ratio, but if you expect a combination of loading it will be.

Be aware that by changing the E1/E2 ratio you will change both the EI1/EI2 and EA1/EA2 ratio and the interlocking between the piles probably does have a significant effect on the EI1/EI2 ratio, but not on the EA1/EA2 ratio of the wall. Hence, if axial forces are the most important you probably want to use a different E1/E2 ratio (closer to 1) then when bending is the most important loading.

It is possible to model the interlocks as rotational joints by using custom connections. However, be aware that it will be a lot of custom connections that will use serious calculation time and memory. Therefore it may be better to follow the suggestion to make a small model in PLAXIS 3D of let's say 3 piles of your secant pile wall, but then really model the piles as volume elements including the interlocking with rotational joints and put some loads on it to see how it behaves. Then make the model with a plate and calibrate E1 and E2 of the plate to capture the behaviour of your small model that has the real piles.

With kind regards,

Dennis Waterman

Dear Mohammed,

It highly depends on what you want to model and how you expect to secant pile wall to be loaded whether you want to model it isotropic or anisotropic. If you expect the wall to have one predominant loading it may not be worth figuring out the E1/E2 ratio, but if you expect a combination of loading it will be.

Be aware that by changing the E1/E2 ratio you will change both the EI1/EI2 and EA1/EA2 ratio and the interlocking between the piles probably does have a significant effect on the EI1/EI2 ratio, but not on the EA1/EA2 ratio of the wall. Hence, if axial forces are the most important you probably want to use a different E1/E2 ratio (closer to 1) then when bending is the most important loading.

It is possible to model the interlocks as rotational joints by using custom connections. However, be aware that it will be a lot of custom connections that will use serious calculation time and memory. Therefore it may be better to follow the suggestion to make a small model in PLAXIS 3D of let's say 3 piles of your secant pile wall, but then really model the piles as volume elements including the interlocking with rotational joints and put some loads on it to see how it behaves. Then make the model with a plate and calibrate E1 and E2 of the plate to capture the behaviour of your small model that has the real piles.

With kind regards,

Dennis Waterman