Failure Shear Strain analogous to Elongation to Failure

[BobbyCY]

Hi everyone who reads this...

I’m trying to justify a design which is under significant and dominant shear loading. The load-case is an “ultimate” case, such that the design is allowed to yield (so we’re in the plastic region instead of below 0.2% Yield/proof) as much as it likes as long as it doesn’t fail/rupture.

My question is whether there is a similar shear-related property to the tensile “% elongation to failure” that I could use to set a limit to my plastic shear deformation? Is there anyway to estimate such a property based on uni-axial "elongation" data? Or is there already a standard limit that I'm missing for this type of plastic deformation?

Thanks in advance,
Bob

 

[metengr]

I think your best bet from a design standpoint would be to use shear yield strength based on this expression 0.57*tensile yield strength (0.2%offset). In other words, if applied shear stresses are below 0.57*tensile yield strength, you will not have plastic deformation under the application of shear stresses. By the way, elongation is a quantity that can be expressed under tension or shear loading.

 

[metengr]

Sorry, I just realized you do allow some plastic deformation in shear. In this case, the same concept would apply expect use a shear stress limit between 0.57x tensile yield strength and some percentage of the ultimate tensile strength of the material of interest. Take steel, you could use a limit of 0.5x ultimate tensile strength (UTS) to decide on a maximum limit of shear deformation . Of course, you could reduce the limit to 0.35XUTS but still above 0.57*TYS.

 

[BobbyCY]

Thanks for the reply!

Yes, I had considered that, however I was advised that something like the UTS or Yield Strength should be compared to elastic-equivalent stress results (i.e. the stress from analysis using a constant elastic stress-strain relationship).

The analysis I'm conducting is elastic-plastic with a non-constant stress-strain relationship (Ramberg-Osgood), and I believed that in such analyses I should be using a strain-based failure criteria instead of stress-based. Or should this not really matter, as the UTS will occur at the "elongation to failure" strain?

Thanks,
Bob

 

[metengr]

Bob;
Take a look at this paper.

http://www.google.com/url?sa=t&rct=..._V_e9E0IapcaeUjxg&sig2=EPc54B3aFBu0K4Gv1gO45w

 

[TVP]

If you are conducting elastic-plastic analysis, you need true stress-strain data for the material, not engineering stress-strain. The UTS is not meaningful if this load case does not produce uniaxial tensile stresses, uniform elongation to necking, etc.

 

[CoryPad]

If the loading is predominantly shear, then the failure criterion will not be based on shear strain since that does not develop cracks.

 

[metengr]

The more I think about this a torsion test on a tubular specimen where torque and angle of twist data are converted into a true shear stress and true shear strain curve would be most applicable for the above elastic-plastic analysis design. In fact, this method would accommodate the evaluation of large shear strains versus relying on tension test data where there is greater oppurtunity for evaluating ductile behavior in shear.