mid nodes in buried piping bends

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,