need help with plastic finite element analysis

[rob768]

Hello all,

I need some help with a finite element problem i am facing. I want to calculate the plastic deformation of a steel pipe manifold (don't know the exact english phrase, but the pipe configuration looks like the letter Y), loaded by internal pressure. I want to model stress and make predictions with regards to failure behavior. However, I am totally new to non-linear calculations, and the manual give some examples but doesn't provide explanations as to the influence of the given paramaters. Thus, I have absolutely no feel of what I am doing. Can ayone help me out, or direct me to online answers or examples? I am using Femap 9.3.1 and NX-Nastran 5

tnx

 

[rb1957]

have you done a linear model already (and are looking to extend this into NL) ?

if not i think you should start there.

have you run an example problem ? if not this should be next !

have you contacted the FeMap help desk ? they're usually pretty helpfull.

which parameters concern you ? some are material constants, which you can only change by changing material, others are run parameters, and i suspect that most people will tell you is try different combinations till the problem works (and looks stable).

 

[rob768]

I do have a model i used for linear calculations, and no problem there. I did run example problems, i tried the paramaters used in those. The Femap helpdesk will be next.
I'll just experience for a while i guess.

 

[cbrn]

Hi,
one question just to avoid you dive in an "unknown world" unnecessarily:
why do you have to perform non-linear plastic analysis on a component like this? Are you undertaking the "Direct Route" of "Design-by-Analysis" of EN-13445 ?
If not, for pressurized components there are many other ways, the most used and recognized worldwide being ASME "Boiler and Pressure Vessel Code" VIII Div.2 (Mandatory Appendix 4, and perhaps also App.5).

Regards

 

[rob768]

It is a part of a failure examination in which a dredge pipe, used for offshore discharging, blew up. We located cause of failure as ductile overload and location of failure initiation, and this latter matches an area of high stress, as calculated in the linear model. I want to get an estimate at which point (pressure) failure may occur, as this is relevant with regards to the responsible party. If it is slightly more than normal operating pressure (of 8 Bar), it can not be blamed to the ship, but to the owner of the pipe. However there was mention of high pressure fluctuations and if the calculation indicates that failure only occurs at much more than normal pressure, the ship is to blame.

 

[cbrn]

Hi,
ah, OK, I understand. Now I can't help you further as I know nothing about Femap+Nastran. Just a tip: in order to demonstrate what you are describing, one way could be:
1- estimate the pressure load at which you want to check the net section failure
2- set up an analysis with multiple time-steps, in which the pressure load is ramped from 0 to the max at the last timestep
3- choose the number of timesteps so that you can "trap" pressure value with sufficient accuracy (for example, if you ramp from 0 to 10 bars in 10 timesteps, you will detect the critical pressure load with an uncertainty of 1 bar)
4- run the analysis; the last converged timestep will give you the max pressure load that can be held by the component; if all timesteps converge, then all your pressure range is sustainable (and the pipe is out of cause...)

Regards

 

[rob768]

thanks for your help. I am getting answers now, from some small test files I am experienting with. More important, I am Getting a feel of what i am doing, and the results match hand calculations and expected deformed shapes and all.
I am sure I am getting somewhere now.