CRACK DETECTION 1

[CESSNA1]

HELLO ALL: I am familiar with dye penetrant and magnetic particle crack detection, but I am asking for input/references on other, newer methods such as ultrasonic acoustic emission. I am interested in equipment availabie, ease of use, ease of interpretation, relative costs, and the tough one is cost/benefit. The material is 2000 series aluminum which is painted all over and with machined pockets on one side.

Your inpur will be helpful

Regards

Dave

 

[unclesyd]

AE testing equipment is very expensive and requires one well versed in the art and technique to accomplish a meaningful test.
Our equipment cost over $75,000 dollars better than 20 years ago.

I think the best approach would be to contact one of the testing companies that do AE and let them look at the component and see if it is amenable to AE testing.

Here are the leader in AE testing and there are leads to several testing companies along with some very good information.

http://www.pacndt.com/

http://www.deci.com/index.html

I have heard that some AE equipment can be rented or leased, but this still requires a technician to run the test and interpret the data.

 

[TVP]

Sonoscan is also a provider of acoustic imaging equipment and services:

http://www.sonoscan.com/

 

[rb1957]

is there any reason you've gone from dye penetrant to AE, jumping over Eddy Current ?

 

[weeeds]

Another option to consider is Helium mass spectrometry under vacuum. We use it in our shop constantly but the application must be right.

 

[Cockroach]

What about restivity measurements? Does not restivity increase in the presence of a crack?

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[CESSNA1]

THANKS EVERYONE: rb1957, the game plane here is to identify small internal cracks before they become big external cracks and a ailure. Ken, are there companies that sell resisitivity equipment? If anyone else has info on pros/cons/methods for small, internal crack detection, please respond

Thanks again all you savvy people for your input

Regards
Dave

 

[Kwan]

Don't forget to consider probability of detecting a given size along with your cost etc. I assume you are working out inspection requirements based on DTA? Anyway, the size you can detect can add considerable time to your inspection.

 

[MintJulep]

Or you could do what one of my clients does.

Hit it with a stick.

One of the problems with any indirect inspection method is that they are sensitive to set-up conditions.

You want to detect small cracks before they become big ones. This implies that there is a maximum acceptable defect size, right?

For any indirect method, you can get an indication, remove the equipment, do nothing to the part under inspection, reconnect the equipment and get a slightly different indication.

So, at a minimum you need to factor in the uncertanty of the technique when establishing your limits.

 

[Kwan]

Yes Mint is right. it's called probability of detection. what percentage of operators detected a given size. It is also very very important to build calibration tools to ensure you have the machine settings correct for the given size you are looking for. And use the tool before each inspection.

The critical crack size should be based on residual strength. Now with some crack growth analysis you can determine the initial detectable size of your crack to provide a good inspection interval. So if you have sufficient crack growth time, then it's better to select a detectable crack size to provide the customer with a more economical inspection method (ie: a visual inspection does not cost as much as a bolt hole eddy current).

 

[ndeguy]

Cessna1, You do not mention ultrasonics - have you tried its capabilities on items with known defects representative of those you wish to detect? Many special composite probes are now available with enhanced properties. New proprietary UT sets are all-singing, all-dancing with so many enhanced functions. Definitely worth a try unless the item's geometry precludes its use.


ACFM is excellent for coated near-surface defect detection, very sensitive and no paint removal required, but its depth of penetration is limited - see for example:

http://www.tscinspectionsystems.com

Good luck

 

[rb1957]

going back a couple posts ... i'm assuming cessna1 is doing a QA inspection, rather than a repeatitive in-service inspection. so i doubt cessna1 is doing DTA to determine his detectable size.

I suspect that PoD (probability of detection) probably isn't going to help much, he's talking about some fairly exotic inspection methods, i doubt there is the data available.

how small a flaw are you looking for, and why ?

Ultrasonic and Eddy Current both will detect subsurface flaws. All of these methods, including hitting it with a stick (a MIL-SPEC calibrated stick, weilded by a MIL-SPEC defined guy, in a MIL-SPEC defined procedure ... you can see where this is going !), but how will you interpret the results (false +ve, false -ve) ?

just some random thoughts, trying to be +ve

 

[Fabrico]

Resistivity? I would be skeptical of elelectrical resistivity being a valid method unless the part is a) almost 2 parts, or b) you use many thousands of amps. On the other hand, thermal conductivity may be more revealing.

Internal cracks and defects are usually checked with X-ray.

 

[unclesyd]

The easiest way to find out if you have active cracking of any size is do an AE test if the component is amenable to testing.
AE is used for testing aluminum aircraft components. In fact one of the first big test by AE was the C5A in it's initial flight tests. It failed.

All the above mentioned NDT tests are great if you know that you have cracks in a particular area. Hunting cracks, especially non-surface breaking, blind is an almost insurmountable task if the component has any complexity.
I would talk with several of the AE people to see if AE is a viable approach.

 

[Kwan]

so let me make some more assumptions because RB is probably correct (he is quite sharp) about this being a QA requirement on part build. Part is going on a FAA certifed aircraft? Part is machined out of 7050 plate? Aircraft is damage tolerance certified? Thus, you need to ensure cracks no more than .05?

Typically you order the part ensuring that the plate has been ultrasonically inspected. (note: i've also had to call out min fracture toughness on plate for critical parts. We hand selected the material before part manufacture). Ultrasonic has a dead band on the surface and usually you order slightly thicker to machine off both side anyways. So now you have material that has been validated against a certain sized crack. Machining can, of course, cause cracks. Thus, call out pen inspect on the drawing. Is this not idustry standard?

Now if this inspection IS for repetative inspections and you call out resistivity methods, you have just hindered (want to use a stronger word but) your customers by forcing an inspection that can't be performed by the typical repair shop. This does not make much sense to me. Better to just redesign. Just my opinion. I'm in a bad mood and should refrain from writing but have been partaking of some good beer (hops are awesome).