Measuring the Resistance in a Rod to Indicate Cracking? 1

[chrisperry12345]

I am going to perform a test in which I will be bending various steel rods. Some of the rods will be exposed to conditions that make them much more brittle than others, and the test will consist of bending the straightened rods until the first signs of crack initiation. Comparison between the rods will be based on the maximum bend angle achieved prior to the first signs of cracking. My question is, would measuring the electrical resistance in the rod be a good way to indicate when cracking has started? In other words, let's say the rod has X ohms of resistance at the start of the test (straightened), would this number change drastically at the first signs of cracking? Theoretically it should right...

Chris

 

[1gibson]

Theoretically the cross sectional area will be reduced at the crack, and resistance will increase a very very small amount. I don't see how this is relevant to actually trying to measure it though. The temp increase from plastic deformation will have more of an affect on resistance than the first sign of cracks forming.

You'd have better luck I think searching for a coating that would crack only when the metal does, and set up a holiday test. Failure of the holiday test would alert you to the first sign of cracking. Too flexible and the coating will stretch, bridge gaps between cracks. Too rigid and the coating will crack before the metal. Still not an easy problem to solve, but it is more realistic and does have an electrical indication (which you may desire for whatever reason, automatically stop the bend maybe.)

 

[mfgenggear]

chris

why not do NDT (non destructive testing) after bending, to locate any cracking.
IE magnaflux , penetrant, buy a cheap penetrant kit, and do it your self.
or if a mag is available it takes five minutes once the setup is written.

Mfgenggear

 

[Metalhead97]

chrisperry12345,

The resistance changes are going to be determined by the constriction resistance associated with the crack, which goes rho/2a where a is the diameter of the constricted area. That calculation is for circular constricted geometries, but you might want to dig around for other geometries that are more 'real' for your situation. In reality, the constriction effects will not be significant. Think about it this way. People who do contact resistance measurements measure contact resistance changes in m-Ohms and that's is way more constricted than your situation! You would be better served with an eddy current type test.

MH

http://www.linkedin.com/pub/luke-autry/1b/510/566

 

[EdStainless]

are these smooth round rods? If so you might look at acoustic measurements. The modulus and resonant frequency will be very sensitive to minor cracks.

Though eddy current and penetrant examination would both be easy, fast, and sensitive.

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

 

[metengr]

My question is, would measuring the electrical resistance in the rod be a good way to indicate when cracking has started? In other words, let's say the rod has X ohms of resistance at the start of the test (straightened), would this number change drastically at the first signs of cracking?

No, it would be ineffective. Why, because as the rod deforms the resistivity will change due to cold working.

 

[IRstuff]

Copper is famous for cold working changing both resistance and ductility. We had to get a special copper that could tolerate repeated bending and flexing in a ribbon cable.

TTFN
faq731-376

 

[Metalhead97]

The plastic deformation effects from the bending can be separated out using a reference sample. For instance, you can run two simultaneous tests (sample plus known reference inspected by eddy current, ultrasound, etc.). You will get two V-out values at the exit leads, which are fed into the positive and negative terminals of an op-amp. The amplified signal will just be noise unless you can detect a crack, which you won't as my original response explains. In order to detect cracks you are talking micro-ohm type sensitivities or less.

MH

http://www.linkedin.com/pub/luke-autry/1b/510/566

 

[chrisperry12345]

Is there a way to determine when the rod first starts to crack without doing a destructive test? I know that mfgenggear suggested using a penetrant dye but I would rather not do this method due to the high volume of test samples. I'm bending the threaded rod and I would like to stop applying load at the first signs of cracking (I will not have a load cell setup either). Is there a way to determine this other than just visual?

 

[Metalhead97]

I think one of us mentioned using Eddy Current tests. Go here to get more info:

Link


MH

http://www.linkedin.com/pub/luke-autry/1b/510/566

 

[mfgenggear]

Metalhead

thank you for the link, very informative information.
Since I do not have personnel experience with Eddy Current,
my question which would be applicable to the OP is this.
what is the sensitivity of eddy current with finding very fine surfaces cracks.
I know with penetrant it will detect cracks easily.


Chris
Penetrant inspection is a very fast procedure.
clean with the approved solvent, remove all preservatives.
apply the penetrant
let it set for 1 minute
wipe off the penetrant with the required solvent
apply developer.
review for indications
dip in solvent or wipe clean with cloth or wipes
air blow
done
5 min's
This is a solid NDT process if cracks are there it will detect them.

all though if eddy current can easily detect very fine cracks
and that may be your cup of tea.
it would be less mess. :>)

Mfgenggear

 

[Metalhead97]

mfgenggear,

I cannot comment on the absolute flaw size that is detectable because it depends on the frequency, probe size, and technique used. If the probe is fixed you don't have to worry about any lift-off effects and the only changes (in terms of impedance) you will see on the scope are due to the formation of cracks.

MH

http://www.linkedin.com/pub/luke-autry/1b/510/566

 

[unclesyd]

The suggestion of using acoustic emissions is the absolute way to go. A little back ground information as to the sensitivity of the transducer. After mounting the transducers and in a spot that is in envelop of the next transducer we will take a m mechanical pencil with 2H lead, over extend the lead and press it against the material being checked. The cracking of pencil lead gives a very definitive signal.

Another possibility is they make a very small (?acoustical microphone?), very similar to a phonograph pickup head that maybe senestive enough to detect the beginning of crack.
With threads and the noise from the mechanical effort to bend the material may rule out this approach.

We have a manufacturing process where we grind different points on 1 mm diameter tool steel points or pins. Measurements of the ground area are made by passing the part through a ccil at very high speed. How this works I haven't clue, but it does. As we get points from different vendors commingling is a problem as each batch shows different electric charismatics. While working with a commingled lot of points the operator noticed he was getting a very high number measuring defects. I got requested to do an metallurgical analysis. About 95% of the rejected pins where from the same lot. Comparator measurements showed there was nothing wrong with measurements. Further analysis by optical and SEM revealed that the points had cracks that occurred during heat treating.
I'll try to get some information on how this works and if it might work for application. Our setup was extremely simple. The only instruments I recall were an oscilloscope, some type amplifier, and one unknown.

We also conducted a bend test on each lot of points using a with a very sensitive dial indicator. We were looking for the end point. Later on electronic measurements reveal a change in shape of the curve, may have been the onset of cracking. I don't think this was followed up.

 

[EdStainless]

I have seen AE sensors that actually use the material under test for the transducer.
You use a coil for a bias field and then some pickup coils.
The other one is a small piezoelectric pickup on one end of the rod.
There has been a lot work done on this. Search.

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

 

[unclesyd]

EdStainless, I.ve seen several references to the same thing. I've had no luck finding anything further, like a paper or thesis.

As i understand it thee are now handheld vibration monitors that can measure the Spike Energy. The old Spike Energy Meter was the first step in trying to analyze bearing failures. If you detected a high energy level on startup and pulled the bearing it took a stereo microscope around 100X to see the onset of the failure mechanism.
I would talk to some suppliers of hand held meters about this approach and rent one to do some testing.

Far Out:
I would like to try putting a small hole in a test piece and inserting a phonograph needle held in with epoxy. or a press fit, if I can make the hole good enough. Then insert the needle in a pickup head and condition the signal to visualize it on an oscilloscope. I've got some substandard Charpy sample of fairly brittle material where I could deepen the notch and figure some to break it with a little noise as possible, maybe over pins.
The problem right now is that there not any oscilloscopes laying around,even at two ham radio operators.

 

[gtaw]

Acoustic emissions will probably be the best means of detecting a crack as it initiates.

The formation and propagation of the crack tip will emit sound that can be picked up by the transducer and amplified by the machine for analysis. Multiple transducers will permit you to locate the crack with a degree of precision that is dependent on the equipment you use.


Best regards - Al

 

[gtaw]

Here's a web site that may be of interest.

http://www.ndt-ed.org/EducationResources/CommunityCollege/Other Methods/AE/AE_Intro.htm

Best regards - Al

 

[unclesyd]

Take a look at the first item on this page. it is a USB AE Node that hopefully will not break the bank or it maybe possible to rent/lease one. PAC used have very good technical support, it might still be there.

 

[EdStainless]

The only one that I have details on is for finding cracks in railroad wheels.
I think that Syd found what you need.

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