MODELING ELASTIC CONTLOVER BEAM WITH SHELL

[navidd1]

HI EVERYBODY
I AM MODELING A CONTLIVER BEAM WITH BOX SECTION USING SHELL S4R ELEMENT. INPUT CAN BE SEEN IN THE ATTACHEMENT. i DO NOT KNOW WHY THE RESULTS IS NOT SIMILAR TO THE ANALITICAL RESULTS AT THE YEILD POINT. I MEAN THE RELATION BETWEEN DELTA AND P MUST BE DELTA= PL^3/3EI BUT ITS NOT
(I(MM^4)=637632 & DELTA(YEILD TIME(0.64) *TOTAL APPLIED DISPLACEMENT)(20 MM)=12.8 & MODUL OF ELASTISITY(E)(N/MM^2)=200000 & L(MM)=1000 & P(N)FORMULATION=4897.01376 & P(N) FE RESULT=7392)

I WOULD BE GRATEFUL IF SOMEONE CAN HELP ME.

REGARDS
NAVID

 

[bassmanjax]

For starters, I calculated the 2nd moment of area to be 872640 mm^4.

What are you trying to determine with this analysis? For the beam you have specified, I calculated that you would need to apply a load of 8730N to cause yield by bending. This would require a displacement of 17mm by my reckoning.

Also, is there any reason why you are using shells to model the beam? It would be more efficient, not to mention quicker, to use beam elements.

Regards

Martin Stokes CEng MIMechE

 

[navidd1]

Thanks for the reply.
All your calculations are correct. I need to model a cyclic lateral displacement loading on a hollow column with a constant axial load so I need to use shell. I was modeling it and my results showed far less ultimate load capacity compared with experiment even in early steps when we have monotonic loading.
anyway as you said with 16.67 mm displacement(corresponding to 8730N) we will have yield but in the model yield happens in 12.8mm and with this displacement we will have 6720N load analytically and 6605N from the FE result(the difference is less than 2%) but the problem now is why yielding happens so early in FE model?

Regards
Navid