Electrical Resistivity and Yielding in Steel 5

[mxracer17788]

Can electrical resistance be a good measure of the degree of yielding in a steel sample? If I have a steel plate and I measure it's baseline resistance and then subject that same plate to a cyclic bending test, would I get a measurable change in resistance? I know that certain metals will exhibit an increase in resistivity upon cold working, but does this phenomenon only occur for samples that plastically yield? If I did a cyclic fatigue test on a steel sample below the yield strength, would I observe an increase in resistance in the sample or would I have to enter the plastic range to observe any changes in resistance? On another note, is there another good method for measure the degree of yielding in a steel sample under cyclic loading?

Thanks,
Jeff

 

[SnTMan]

Kind of doubt it: thread330-327086

 

[mxracer17788]

Just heard from our mechanical engineer. Apparently we offer a refund on parts that have yielded, but as of now we have no way of doing non-destructive testing to see if yielding has occurred. It is a small steel plate 1" x 2" x 1/4", and the idea was brought up to use electrical resistance to check if the part had yielded. The problem is that, let's say the plate yielded in tension by an elongation of 1/8". Would this plate show a measurable increase in resistance when compared to a plate that has not yielded? SnTMan mentioned another previous thread asking a similiar question, and it was brought up that one could use acoustic emission testing to detect cracking. I'm assuming this test couldnt be used in my situation since we would not be monitoring the actual yielding event. At any rate, is there a good Non-destructive test that is capable of detecting yielding in a steel sample?

 

[IRstuff]

If it yields, presumably, it won't meet its drawing requirements, e.g., in your example, elongation is a dimensional change. It would not be that different than any acoustic test, i.e., a similar level of resolvability exists.

Seems to me your question is a bit off the mark; how did the customer decide it had yielded, so as to get the refund? If your customer can tell, without any fancy test equipment, why can't you?

TTFN
faq731-376

 

[mxracer17788]

Without delving too much into the proprietary technology, these pieces must be replaced after even a slight yielding has occurred. Is it possible to detect that yielding has occurred by NDT methods? If we know that on average, these parts have say X ohms of resistance, could we used changes in resistance to indicate yielding or is this method not sensitive enough. I'm trying to figure out a good NDT method that we can use to determine if parts have started to yield...

 

[mxracer17788]

Could ultrasonic testing be used to measure flaws in the material that could have occurred from yielding? To answer IRstuff, dimensional changes cannot be used because the piece might undergo cyclic yielding which could elongate the sample, and then subsequent return it to it's original shape. This is why I wanted to use resistance testing since this minor degree of cold working might cause an increase in resistivity but I'm unsure of it's resolving power.

 

[EdStainless]

I doubt that you could use resistance since the material has a fairly high resistance to start with. That would probably mask the change due to slight yielding.
If the part has yielded it is either the wrong size or not flat. You can;t hide yielding.
Even if resistance of UT could detect it I don't know how you would distinguish the effect of fatigue from yielding. Both cause similar microstructural damage.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[redpicker]

Perhaps your need to protect your proprietary technology is preventing us from understanding your problem.

But, if the pieces must be replaced after slight yielding has occurred, then there is some change in the pieces that has occurs with yielding that makes the pieces unsuitable for service. Why not use whatever that change is to determine if yielding has occurred?

Or, is there something I am missing?

In practice, the best check to determine if yielding has occurred that I am aware of is to measure strain. Either a permanent strain after load is released or a non-linearity between load and strain indicates yielding.

The major issue I have a trouble with is that you use "slight yielding" (which is not very definitive) to define when the parts must be replaced. It sounds as if you are looking for definitive test to check for a subjective result.

rp

 

[mfgenggear]

the only way seems to me is as explained by IRstuff or by having strain gages applied on the actual part. while in service.
& recorded.

 

[mxracer17788]

So UT is capable of detecting yielding? The problem here is that let's say we test 100 of our parts to define a baseline ultrasonic profile. If we take a part at random and introduce a subtle yield, let's say a minor elongation, could we take a UT profile of this sample and tell that it has yielded when compared with our baseline signal? I'm unsure of UT's resolving power. Also you mentioned that fatigue and yielding can produce similar microstructural damage, but perhaps both conditions are serious enough to warrant a refund since fatigue damage is definitely not a good thing.

 

[IRstuff]

Resistance is futile...

"cyclic yielding which could elongate the sample, and then subsequent return it to it's original shape."
Really? How?

I think the only way to determine whether something has yielded is to manipulate the piece in the same fashion that caused the yield to begin with. Unless the piece has the same stiffness as before shipment, it's likely to have yielded.

TTFN
faq731-376

 

[mxracer17788]

UPDATE: I was not aware that we attentively monitor each of these parts at the time of manufacture. I have reviewed the drawings and these parts have a fairly tight dimensional tolerance. As a result, I am going to take everyone's advice here and have gauge blocks made to create a "go, no go" scenario for measuring these parts in the field. Thanks everyone for the help.

Jeff

 

[Guest102023]

No.

While permanent strain is known to increase resistivity (due to increased dislocation and vacancy densities), this is next to impossible to determine by any practical method other than dimensional measurement. You probably don't even know what degree of strain was present to begin with, so you don't even have a reference. In a real-world steel I can't demonstrate 'slight' yielding metallographically, so I don't know how measuring resistivity would show it.

Yielding implies a permanent change in dimension; dimensional measurement is your best hope.

 

[ManOfSteel8357]

Since it seems like the general consensus from everyone is to use dimensional analysis to determine if the part yields, is there a quick and easy way to take accurate density readings of the part before and after yielding? Since we inspect these parts before they go into service, it wouldnt be too much of an issue to record the density first. My thinking is that say the steel plate permanently elongates by 0.20", it should technically have a lower density when compared to before yielding since it's volume has increased. My question is, is there a way to accurately measure the volume of part up to the thousandth in accuracy? (or a machine that can peform this function. I'm thinking a fluid displacement method since these plates are machined into irregular shapes). If I were able to measure the volume to that accuracy, would permanent yielding really show a change in density that I could measure?

 

[ManOfSteel8357]

By the way, me and mxracer work in the same lab and we really need to find a solution to this problem! All the help has been much appreciated.

 

[Maui]

Plastic deformation in steel components is often modeled as a constant volume process. I doubt that you would find a detectable change in density. But if there is a measureable change, a relatively simple way for you to quantify it would be by using an acoustic method. Utrasonic waves passing through an isotropic linear elastic medium have a longitudinal wave velocity that only depends on the density of the material and the modulus of elasticity according to the relationship V = [E/rho]^0.5 where E is the modulus and rho is the density. This is obviously an oversimplified set of assumptions, and for a more accurate and in-depth analysis you should read this:

http://solidmechanics.org/text/Chapter4_4/Chapter4_4.htm

You can use a standard ultraonic inspection apparatus and piezoelectric transducer to measure the wave speed in your parts before they are placed in service, and after they are returned to you. If there is a change in density, this would probably be the most practical way for you to determine it.

Maui

http://www.EngineeringMetallurgy.com

 

[ManOfSteel8357]

Well if plastic deformation is modeled as a constant volume process in steel, than this might not be a good method. The part is designed to yield in the middle, so I'm wondering if we could take ultrasonic readings just in the center of the part and record these values, and then test the part after it has plastically deformed. I'm wondering what the sensitivity of this test method is...

 

[Maui]

One of the reasons that I recommended it is that it is an extremely sensitive and repeatable method.

Maui

http://www.EngineeringMetallurgy.com

 

[Compositepro]

Using strain gauges is the usual way to measure strain. To measure plastic yield strain, compare before and after measurements between gauge marks (scratches) on the specimen.
If you mean that you want to be able to measure whether parts coming back from the field have yielded, the only reason that would be of interest is that they are now dimensionally out of tolerance.