Dear All
I am running a Plaxis 3D safety analysis for a shallow foundation with skirts for a subsea structure. I have the follow query and really appreciate if anybody can help
In Plaxis, the safety analysis can only start after a successful plastic analysis run.
My plastic analysis fails to reach the full load; by inspection it is believed due to the local soil failure around the foundation skirt - see attached screen shot (Figure 1); so I have slightly changed the tolerance from the default (0.01) to a slightly large number (0.02, up to 0.05) – this have made it a pass (See Figure 2).
then I run the safety analysis, at the beginning it gives below 1.0 and then it go back to 1.3 ( I stop the run, let’s say it become stable). see attached Figure 3. It is noted when i run the safety analysis, i used the default tolerance (0.01). the relative stress ratio after the safety run is shown in Figure 4.
Can I have you thought about this. can I take the stable value (say 1.3 ) as my safety factor?
Also, having read some early discussions, I will add a nil- step between the plastic and safety analysis - hopefully it will help sort some out of balance force out?
Noted all the below pictures attached are from a worst case. for other cases, the plastic run failed marginally (reach 98% full load) and I have just needed to increase the tolerance to 0.02. can you advise if there are any limit set for this.
Thanks all for your help - look forward for feedback
Figure 1 plastic analysis
Figure 2 plaxis analysis with updated tolerance 0.05
Figure 3 – Safety analysis with tolerance 0.01
Figure 4 – Safety analsysi with tolerane 0.01 (stress mobilisation ratio)
Dear Jin Li,
The larger the tolerated error, the more margin the calculation has for "wrong" stresses in order to still use equilibrium. So a larger tolerated error generally leads to a higher factor of safety. The drop in strength reduction factor at the beginning of the Safety phase is probably because the previous phase indeed used the 5% error margin to reach equilibrium and now the Safety phase only allows a 1% margin - which means that the Safety analysis first has to increase the strength (SumMsf < 1) to redistribute the stresses before it can reduce the strength and calculate the factor of Safety.
It's a very tricky situation, because this may indicate that the previous phase was in fact at the edge of failure and you prevented it by increasing the tolerated error. I wouldn't trust it and instead of increasing the tolerated error for instance increase the maximum unloading steps instead.
With kind regards,
Dennis Waterman
Thanks Dennis for your response - it has proved so helpful.
I did not use the above results, as I was not confident (though I have tried to relate it to local failure). In the end, I took an approach of optimising the loads, by which I did pass the plastic analysis with the default tolerance settings.
I did not mention in my previous post that I had doubled the maximum unloading steps from the default 5 to 10 in the above results - Really appreciate.
Kind regards Jin