Difference between theorticall and experimental results 1

[meher634]

Hello,
I performed simple test on a plate(steel) subjected to bending (both experimental and FEA-static linear) but i found out that the results from the experimental study and FEA was quite different.The deflections observed in the experimental study are very high compared to the FEA study,Do I need to perform non-linear analysis since the experimental study deflections are very high ,or is it something to do with the material proerties in FEA study.The stresses observed in FEA are well below yield values and the deflections are too small.Can anyone suggest me regarding this?
Thank you all in advance.

 

[Cmfg]

meher634:

There are several things that could be happening.

- Boundary conditions in your FEM model. Some times boundaries cannot be assumed either as fully clamped nor as simply supported.

- Finite Element used. Solids or shell elements?
Shear locking is very common in thin plates modeled with low order solid elements.
Solutions: - Shell elements
- Higher order elements
- Elements with enhanced modes
- Hourglass elements

- Properties used: E, poisson. Improbable since the material is steel. Variation could be only about 10%.

CMFG

 

[DBruceG]

A couple of other things might be happening as well. How many elements did you use? If you bisect(or whatever) the mesh, do you get the same answer?

DBruceG

 

[prex]

Linear theory for plates is only valid when the maximum deflection is a fraction of plate thickness. Beyond that high non linearities occur because of the importance of in plane stresses.
However, if you make an analysis with geometric non linearity turned on, you'll get lower deflections in the calculation, so your result will be father apart from experiment.
You should check in the first place the boundary conditions, as suggested by Cmfg.
If you want to go more in depth in the discussion here, you should post all the relevant data of your model and experimental setup.

prex

http://www.xcalcs.com

Online tools for structural design

See

http://www.xcalcs.com/docs/symbolinfo.htm

if you want to use symbols on these fora

 

[TheTick]

As one of my professors put it (paraphased)...

"The difference between actual and theoretical results is the differerence between what really happens and what you failed to consider."

"Great ideas need landing gear as well as wings."--C. D. Jackson

http://www.EsoxRepublic.com

 

[meher634]

CMFG/Bruce/Prex
Thank you for the feedback,It was a good suggestion by CMFG that shear locking is common with thin plates modelled with low order elements,I see that the problem was rectified by changing the type of elements.
Thank you all.