Local Yielding

[ninjaneer777]

This is a general question about stress analysis. Let's say you're reviewing stress results and there is very nearly some yielding in a local area. The area is not one that looks too concerning to you, e.g. not on a weld, not near a sharp corner, etc., the stress magnitude is not close to the material's ultimate strength, the part is not in a fatigue situation and no one will die from part failure.

I know that everything always depends on the situation, material, function and many other factors, but I just wan't to get some input from others. Again, sorry if it is too general.

What concepts, like creep, would you consider for situations like this? Could it be just a little strain hardening and nothing to worry about, or would you be concerned that over time it would lead to failure?

Thanks!

 

[corus]

In my understanding, creep is related to temperature and should be considered separately in relation to specific design codes. Similarly I'd refer to design codes for assessing stresses above yield. Normal structural design codes don't consider local yielding as they're based upon simple hand calcs. A pressure vessel design code does go into more detail and would consider localised yielding as that which contributes to failure through fatigue and not overall collapse. In your assessment, therefore, try and refer to recognised design codes.

Tata

 

[rb1957]

if the location is not near anything special, why is it high ? is it an FE result or a hand calc ??

 

[ninjaneer777]

"if the location is not near anything special, why is it high ? is it an FE result or a hand calc ?? "

The question isn't from from any particular situation, it just popped into my head and I was wondering what I would do if that came up. At first I thought that it would always be unacceptable to have stresses near/at yield, if your using a factor of safety greater than one, but then I reconsidered that there may be situations where the area you're concerned about is fine even though some other area looks high.

Here's a hypothetical. A thin plate supported at both ends, with some type of mass in the middle of the plate. The assembly is dropped landing on the wall supports. The middle of the plate experiences high stress but the plate to wall connection is very well designed and suffers acceptably low stress. Now, I know an analysis of this would be wrought with many assumptions and after all is just a simulation so you're stresses could be much different in actuality, but it's just the best I could come up with right now, so just take that as the general idea: a situation where high stress is happening in an area that does not look like it will lead to failure.

Thanks to anyone who replies despite it being conceptual and not specific.

 

[rb1957]

that's quite what i meant ! ... you have high stresses 'cause something's making them that way. in your example it is the inertia load applied could well be overstressin the plate (which is designed by the support reactions).

by you can accept high stresses if you understand them and can rationalise them away. i'll accept high compression stresses (> ftu) on a solid (= thick) flange in bending where they are limited to a small depth of the flange, arguing that localised load redistribution (= plastic bending) will keep things together.

 

[ninjaneer777]

rb1957, thank you for your reply. That is the kind of answer I was looking for; a situation in which you use an educated judgement to say that the stress in that area is high, but because of x or y I don't think it will be a problem. I probably should have phrased my opening question better. Lesson learned.

Can anyone else think of a situation where you've made an educated judgment to deem a high stress area acceptable? Thanks.

 

[Twoballcane]

“Here's a hypothetical. A thin plate supported at both ends, with some type of mass in the middle of the plate.”

Dynamic Case (Velocity Shock): Some yielding is good because this will create some damping saving the mass, but the plate will have to be replaced.

Dynamic Case (Vibration): Yielding is not good and will cycle to failure.

Static Case: If you’re going to sell this at Walmart, yielding does not matter. Plan to use this on the NASA shuttle, yielding is not good. Since us engineers want to design and build quality products, we should keep stress under the Sy point.

If the material has reached its Sy point, it is just a hop, skip, and a jump away from Su, so why risk it? When we do analysis, we do it in an ideal world where we assume simple case of static load or dynamic load where in reality other degrees of freedom will occur making it a worse case then we thought. When the product fails, try explaining it to your boss/customer that you knew it yielded but decided to continue with the design anyway.



Tobalcane
"If you avoid failure, you also avoid success."

 

[chris9]

If it is not a fatigue case then from a design point of view it is best to keep stress values below yeild. However, if you ever put your component on test you will probably find that it fails when the average stress across the cross sectional area has exceeded yield and this will be at a load that is well above the load required to reach yeild stress at a stress concentration. That is assuming it is steel, plastics are an entirely different ball game.

Chris

http://www.value-design-consulting.co.uk