Hi,

I am working on the stress analysis of a 6" oil flowline. The total length of the line is nearly 7km. I am having convergence issues with this model.

Modeling methodology used:
1. For Soil Stiffness

Performed a small stress buried model of 6” line using the actual design data. The small model had 25 nodes, with each node spaced at 5D distance. The modeling for the buried pipeline is carried out, using the buried modeler of Caesar II, with assumed soil depth of a little more than half the diameter of the pipeline, 100mm for the 6" line with the typical soil design parameters. After taking the buried conversion, the Y2 stiffness as seen in the converted sample model corresponding to a mid-node number (to make it free from the end conditions)is used in the actual pipeline model as a normal rest support stiffness (placed at a span length of 5D) with a friction factor of 0.5.

2. The whole model is split into two CAESAR models (Since the number of restrains are high). Please note that there are considerable elevation differences in this model, and the line has got 25 crossings (it is buried in these areas). Temperatures are Max:100C and Min:5C). There are nearly 3500 supports in each model.

Issue: When I run these models, it is not converging. I tried changing the friction stiffness tolerance values whose results are mentioned below.

1. About 700 supports are non-converging (with default friction stiffness tolerance values)

2. When I reduced the value to 1.75E004 it reduced to 300 supports.

3. When I reduced the value to 1E003, it reduced to 20 supports.

I tried to level the elevation slopes at these points but then it shows non-convergence at some other support locations.

My doubts are:

1. What is the exact procedure to solve non- convergence issue for surface laid lines in CAESAR (I cant give a hold down or change the friction factor in these supports, like we do in piping networks)

2. Can I reduce the number of non-converging nodes by changing the friction stiffness to a minimum value? If so what is the minimum value of the friction tolerance that will give me a realistic result for the stress values.

3. Please comment about the methodology that we are using to model surface laid lines in Caesar.

Advance thanks for all the responses and support.

Cheers.