Multilinear Soil Springs - Multilinear Iteration Diverging

The SET MULTI 2 is just an alternative iteration schema when compared to the default one (SET MULTI 1 ) . Either way the software finds out the effective spring stiffness for every spring and if the change of effective spring stiffness between one iteration to the next is very small, the solution is considered to have converged. Before using SET MULTI 2, I would suggest that you ensure that you are not analyzing cases like seismic, wind or any other lateral loads in isolation without gravity loads. Doing so can unnecessarily increase the number of iterations and the solution may even fail to converge. By default the software iterates 10 times. If convergence is not achieved in 10 iterations, it generates warnings like the one you mentioned. One can increase the iteration limits to beyond 10 by using another command SET ITERLIM 15 which would set the limit to 15 iterations. It is advisable not to go beyond 15 iterations.

GOOD EVENING

i have to simulate with STAAD_Pro a casing pipe (burried pipeline) subjected to soil cover load, traffic load and lateral pressure. I have simulated pipe-soil interaction with multilinear soil pring which constitutive law have been defined according to ALA2005.

Pipe has been modelled with shell element and with proper mesh. To all nodes the following boundary condition have been assigned:

SUPPORTS
21 TO 120 FIXED BUT MX MY MZ KFX 314 KFY 1512 KFZ 87.33
MULTILINEAR SPRINGS
21 TO 120 SPRINGS -1.422 1512 -0.3 1512 0 0 0.3 50.41 1.422 50.41
MULTILINEAR SPRINGS
21 TO 120 SPRINGS -1.44 87.33 0.5 87.33 0 0 0.5 87.33 1.44 87.33
MULTILINEAR SPRINGS
21 TO 120 SPRINGS -1.44 314 0.5 314 0 0 0.5 50 1.44 314

I'd like to ask, if the procedure followed is correct. 

Perfoming a stati analysis, STAAD give the following warning message:   PLEASE CHECK RESULTS. Try rerunning with SET MULTI 2

What does it mean?

Thanks, regards

Currently in STAAD, at any given joint, only one multilinear spring can be assigned. It looks like, at each of the joints 21 to 120, your intention is to have 3 different multilinear springs: one for the X direction, one for Y and the third one for Z.

To assign different multi-linear springs for different d.o.fs at a specific joint, you can try the following procedure.

Define additional nodes very close to the actual nodes where the support is located. Connect these additional nodes to the existing nodes through rigid members. A rigid member can be a short beam segment and has a material with a high E and zero density. Place the spring support along the X direction at one of those new nodes, and along the Y direction at a second node, and along the Z direction at the third node. Each of these nodes can then be assigned its own multi-linear spring data.

Googmorning,

first of all i want to thank you for the quick answer. 

I agree with the procedure you describe before; therefore I have a lot of nodes in the f.e. model. The pipe has been modelled with shell elements with a regular mesh of 500mm, for a total lenght of 1000m. So, i have to define to much rigid link and in my opinion and this wil became complicated and long procedure.

I'd like to ask you, if in staad pro a such tipe of calculation (burried pipe, with vertical, horizzontal and lateral soil spring) has ever been modelled, and if you can indicate/provided some examples or pubblications. 

I'd like to ask you if it is correct to modell to interaction assigning the spring as indicated in the figure reported below:

Or considering only the bottom part of the pipe as follow:

In both models, vertical soil load, horizzontal soil load and vertical traffic load have benne applied.

The results are obtnained are differetent in term of stress.

Thank you, regards