How to create voids PURPOSELY to test for NDT

[pkiskool]

Hey, I'm doing some NDT research here at my job and my boss has asked me to create artificial voids within the carbon panels using prepregs. I've tried two methods thus far which both didn't work:

1)I've tried curing small sections of a prepreg sheet by sandwiching it between two aluminum sheets with cut-out sections, and heating it in the radiation light source (so that the cut-out section will be cured, but the other parts won't since the UV will reflect off the aluminum sheets). I successfully cured the small sections of the prepreg sheet, then I laid this sheet up in the middle of a regular prepreg layup and cured the entire thing (with vacuum) in the oven hoping that would create a void somehow: but no, every layers sort of cured together, I couldn't even see the pre-cured layer.

2)This time I took 2 pieces of a small cured prepreg parts and stacked them together. I then sealed all four edges with a sealant tape and inserted this "pocket" of cured prepreg in between my layup of 14 prepreg plies, then cured the entire thing. I was so sure that this would create a void inside the panel, if not at least a piece of tape in there for my NDT device to pick up, but no, everything sort of melted together again and there is absolutely no void, no tape, na-da.

Anyone know a effective way of creating an artificial voids?
My NDT device is an infrared camera, FYI.
Panel must be made of prepregs as well.

Thank you all.

 

[KirbyWan]

I would skip the prepreg all together. Just take two layers of non-perforated FEP, seal the edges together and put them in your laminate. (note: I've never done this, just spitballing here what might work) Are you doing a high temperature layup?

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[pkiskool]

So you're suggesting inserting a pocket made of non-perforated FEP?
We get up to about 250 F or so.

 

[RPstress]

Just progressively reducing pressure if it's autoclave cured or vacuum if it's out-of-autoclave vacuum only should give progressively worse cures. Not consolidating the stack during layup may also have/contribute to the desired effect. You can sometimes get porosity when it's an autoclave cure by maintaining full vacuum when the cure cycle recommends reducing vacuum (can let volatiles present come out of solution while resin is good and liquid).

What's the materials's spec.?

 

[Compositepro]

How are you determining that you do not have voids? The only reliable method that does not need some other type of calibration is to cut and polish the specimen and examine with a microscope.

What you are attempting has been tried by many people before you. It is not easy to do with any degree of control. There are many different forms of porosity as well as degrees. The best approach is to get samples of rejected parts. You could also try fiberglass prepreg so that you can actually see the voids you are trying to detect by some other method.

 

[pkiskool]

It's an out-of-autoclave cure, using a technology called the Pharos(TM) oven. Basically a closed aluminum oven with heat lamps emitting UV rays.
Going to try what Kirby suggested with the two layers of non-perforated FEP.
Thanks ya'll

 

[berkshire]

try sticking a dab of mold release compound, "Moldwiz" "Kantstick" or similar between two plies. Then curing it.
B.E.

 

[Demon3]

You can add PTFE particles. They will behave just like voids because they won't bond to the epoxy. Unlike air bubbles, you can be sure what size and shape particles you are adding and how much of them. Air bubbles or foaming agent will give holes that change size depending upon solubility and pressure.

You can buy PTFE particles from DuPont or cheaper from Shamrock.

Chris DeArmitt PhD FRSC CChem

http://www.phantomplastics.com

Consultant to the plastics industry

 

[KirbyWan]

pkistool,

It looks like most of the suggestions (from people who know more about composites then I do) are to get porosity. My suggestion was to just get a delaminated area. The mold release might work as well. How are you testing for voids? sonic tap testing (i.e. tap hammer) or an ultrasonic detector? With the right equipment you should be able to tell which layer is disbonded.

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[pkiskool]

Thanks all for suggestions.
We are aiming for the delamination to be exact, not so much the porosity.
But who knows, with application of enough mold release, it could cause a delam along with the porosity.

And Kirby, our focused testing method currently is using the Infrared camera. There would be light source behind the specimen and we would look at the camera from the front.
Hopefully it is accurate and sophisticate enough to detect a small delaminations.

 

[pkiskool]

By the way Kirby, my name on this forum is 'pkisKool' with a K, not 'pkisTool' with a T.
LOL

 

[KirbyWan]

pkiskool,

You're kool, not a tool. Sorry for the error, I'm just a fool.

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[PanelGuy]

You could try the tip of a rubber glove. They always make a nice void and show up with Ultrasound. Atleast you could cut shapes and have them well defined.

The mold release deal should work well.

It might be hard to get measurable voids in a small sample, easy to evacuate. You may need to flex it a little to get your intranal structure to "stop touching". Maybe a thin sheet of Teflon plastic, again, it would be able to be well defined.

What about sliding a needle between and intentionally injecting an air pocket in the sandwich before cure...like making bismarks?

My thoughts anyway. Good luck!

 

[Sparweb]

I just completed a test where a void was deliberately placed in the lay-up. We used a strip of plumber's teflon tape. A piece of mold release film might work, too.


Steven Fahey, CET

 

[GregoireMC]

If you want to do a "heat" picture of the composite, inserting a strip of polymer (Teflon...) will disturb the heat flow differently than delamination with no foreign object of the "real world", so may be the release agent is better (if it's working of course).

 

[KirbyWan]

SparWeb,

Did you use just a single layer? I would think even teflon would have some bonding with the resin on both sides which is why I was thinking two layers would be needed. I do understand that for thermal through transmission that may not make a difference, but for sonic or ultrasonic I think it might.

-Kirby

Kirby Wilkerson

Remember, first define the problem, then solve it.

 

[Sparweb]

The teflon tape was suggested to me by an engineer at Transport Canada, who was referring to various books as he was explaining the topic. Sorry, I can't remember which book it was - something from Boeing? Niu? Only one layer of Teflon was disucssed.

The teflon was plainly obvious in the fiberglass part, as you'd expect. It was easily detectable by coin-tapping, too. We didn't bother getting out an ultrasonic scope because the plane requires very little ultrasonic anyway. The load test procedure was so elaborate that I was too busy to spend any more time thinking about the flaw inclusion than that.


Steven Fahey, CET

 

[Radagast]

Hello,

We successfully produced voids in our test specimens by stamping out 1/8, 1/4 and 1/2 inch circles (defects) of Grafoil (flexible graphite) and placed them at various depths within the laminate. We used both prepreg fabric, unidirectional tape and a combination of both for our tests, cured using our normal production run. It worked like a champ.

Mike

 

[pkiskool]

Sorry for getting back a little late on here.

I tried both Teflon tape and Release tape (wrapped a cured prepreg sheet).
The two panels came out, I couldn't see the defect using my IR camera, so we cut the panel to check out the cross-section.
So void/delam.

Frustration is amounting.

Post from Radagast - regarding Grafoil, would be my next attempt, but going to step aside with this for a bit.

 

[pkiskool]

Just wondering Radagast, did it show up as a void/delam in your detection? or just a foreign object?
By putting in a piece of steel or plastic would do the same trick IF it's just picking up a foreign object, no?