3 Point bend test validation with FEA 1

[meher634]

Group,
I am trying to validate 3point bend test performed on a screw with FEA by comparing displacement values. From the test the Screw defelcted atmost of 1 mm, i tried to check the same with FEA but I am surprised that i dont get the same deflection, its deviating almost 35%. I performed FEA on couple of softwares and i get the same percentage error. I am just wondering if this error is in my mechanical testing or FEA. ( I made sure that the boundary condtions exaclty matched with those i did in testing, also the model is converged for results). Any suggestions?
Many Thanks

 

[SWComposites]

If you are trying to compare to cross-head travel of the test machine you are unlikely to ever get the FEA to match. The problem is that there is deflection in the test machine/fixture that you are not accounting for. For this reason, displacement measurements of a flexure test should ALWAYS be made with a displacement transducer directly measuring the deflection of the center of the specimen. Cross-head (screw) displacment is basically useless.
Other sources of potential error are:
- inaccurate material moduli; if this is a composite material then you probably have to account for different tensile and compression modului. For any material, if the material moduli varies with strain then you have to account for that also.
- large displacement behaviour; if the displacment is more than ~ the specimen thickness, then it is likely that you have to consider large (non-linear) displacments.
- change in geometry with deflection of the specimen; if round support bars are used in the test, as the specimen deflects the effective support span decreases.