I'm trying to reconcile the stress results shown in AutoPIPE with what I would expect the code stresses to be based on the shown values.
The AutoPIPE code used set to ASME B31.4
Please see the screen shot below with reference to the highlighted line, the results of interest shown are:- Hoop stress (Sh)= 180.41 N/mm2- Max Long (SL)= 317.05 N/mm2- Combined (Se)= 362.84 N/mm2- Shear stress (Ss)= 1.01 N/mm2
I'm currently doing a design to PD8010 so need to check for the equivalent Von Mises stress based on:
Se = (Sh^2 + SL^2 - Sh.SL + 3Ss^2)^0.5
Therefore using the numbers above - I calculate the Se to be 275.4 N/mm2, however as you can see the Combined stress reported by AutoPIPE (which I would have expected to be the VM stress as well) is much higher at 362.84 N/mm2.
Can you please help with clarifying how AutoPIPE calculates the Combined stress?
In Tools>Model Options>Results have you set the "Total Stress" to O (Octahedral)? This results in the calculation of Octahedral stress i.e. Von Mises as required by PD8010.
Thanks very much for your response.
As per your advice I've adjusted the "Total Stress" option to "O" and re-checked the results and have noticed that now the 'Combined stress' has come down to 314.59 N/mm2 (previously 362.84 N/mm2), which is great - however this is still not as low as the manual calculation for the Von Mises/Octahedral stress which should be around 275 N/mm2.
Any assistance to clarify why there is still this difference would be greatly appreciated.
Thanks in advance.
Please note that Von Mises stress is calculated for at least 3 locations, neutral axis, maximum bending and minimum or negative bending. Only the maximum of these is reported.
Thanks for responding.
Is it possible for me to use the reported (Hoop), (Shear) and (Max Long) stress component values to manually calculate the Maximum Von Mises Stress that matches the value shown in the (Combined) column?
The VM stress that I calculate using the component values are lower that the (Combined) value calculated by AutoPIPE - so are you saying that the reported stress component values are not the maximum values for ALL locations (i.e. NA, bending, -bending) so can't be used to manually determine the same (Combined) stress?