The buried pipe modeller doesn't place any mid nodes on bends. If you go in and place mid nodes you may find significantly higher stresses, eg one model I ran with a bend angles of 80degrees and 10D radius gave about a 30% higher stress range in this manner. But I suspect this exaggerates the true stress due to the discrete nature of the soil restraint model..is it recommended to avoid/ignore mid nodes ?

I expect the answer to this one may involve the phrase 'use engineering judgement'. Anyhow heres the result of a little test on stress range for a 90degree bend:

1. use output from buried modeller directly:121MPa
2. add mid-nodes to above:142MPa
3. pre-divide bend into 4 tangent points before burial (don't add mid nodes):136MPa

So I guess I would tend to go with mid-nodes based on that outcome.

I think that option 3 would be what you need when you have a 10D radius.

It is an article in MEN june 2000 written by Dave that explain how.

Adding nodes after you buried the model it may not be correct.

Regards,