Need to drill 63,000 .020" dia. holes 6

[buckley8]

I am trying to find a machining process for making 63,000 very small (.020” dia) holes through .200” thick 6061 aluminum with .050” layer of epoxy (Scotch-Weld, Shore D hardness = 55) bonded to it. Hole spacing is .056” resulting in 318 holes per square inch. The holes do not have to be perfectly round, and the diameter tolerance is generous (.019-.024”)

This is for a rub strip over the rotor on a turbofan inlet for a test rig. The epoxy serves as a sacrificial abradable surface in case the rotor tips contact the nacelle and this is the only approved method/material for doing so. The holes are the focus of acoustic testing so they are not negotiable.

EDM can not be used since it requires electrically conductive material and epoxy is not.
Electron beam drilling (e.g. Acceleron Inc.) can not be used as the epoxy will off-gas and ruin the laser hardware. Anyway, this process is limited to .200” deep holes.
I fear that good-old-fashioned mechanical drilling is the only way this can be done and I figure that it will take 3 months or more to make this part (at 30 seconds per hole, 8 hours per day)= $$$$.

To make matters worse, the rub strip is not a flat plate but is a cylinder, 22” ID, 3” long.

I would also appreciate suggestions to "build-up" the aluminum/epoxy rub-strip with holes instead of drilling the holes from the rub-strip.


Jim Buckley
ASRC Aerospace
at NASA Glenn Research Center
Cleveland, OH

 

[SteveWag]

Why not woven aluminum mesh, 20 mesh, with 25, 27, or 30 thousandths, flattened (a little) and then coated. The square openings will round a little when coated and might look like round holes.

 

[IRstuff]

All very high tech solutions, but people seemed to blow off drilling way too quickly. Your time estimate for mechanical drilling is absurdly high. see:

http://www.pcbdriller.com/sz1000.htm

They quote 350 hpm, which works out to 21,000 holes per hour. Assuming that they're optimistic by a factor of 5, it would still only take about 16 hours to drill your device. Note that these guys are drilling 0.006" holes routinely, a 0.020" hole is probably gigantic to them.

TTFN

FAQ731-376

 

[buckley8]

I definitely like the idea of edm-ing the holes before the epoxy is applied. Of course, plugging the holes and then de-plugging them would not be much fun...
Are you suggesting an automated process for plugging the holes? How does the wire plug stay in the hole? Interference fit?

The abradable rub-strip is a safety requirement.

The two-part epoxy (3M p/n:EC-3524) is applied with a spatula or trowel and cures at room temp.

jistre: You are very close to understanding the rub-strip. The rub-strip is for a worst-case scenario. If the centrifugal force (from overspeed), excessive vibration, or temperature (or some other force) of the rotor blades is great enough to cause the blade to contact the inlet wall, the wall will be abraded, saving the expensive rotor blade and preventing fragmentation (at 10,000 rpm). It would only need to be used once. The epoxy is .050" thick. If the rotor goes through that, it has already come out of the hub and is on its way to doing some serious damage. You can't just have a lot of air clearance since it would seriously affect the efficiency of the compression of the fan. Clearance is ideally <.004". Therefore, a small excursion of the blade tip could cause it to scrape the rub strip.

Jim Buckley
ASRC Aerospace

 

[KENAT]

IRstuff, the different materials/material thickness might be the killer though.

PCB are a lot less than .2 thick and probably easier to drill than AL.

KENAT, probably the least qualified checker you'll ever meet...

 

[Compositepro]

Many hush kits for jets are made from carbon fabric prepreg and hard plastic pin blankets. After the epoxy is cured the pin blanet is pulled off leaving the holes. The pins are pushed through the prepreg with pneumatic hammers.

 

[ewh]

Drill the holes, aplly the epoxy and force air through before the epoxy dries, thus clearing the holes?

 

[jistre]

Ahh, it's the need for the tight air clearances that didn't click. Removing the epoxy and having a larger gap just won't work in that case, will it?

Then I still like the idea of using some sort of NC automation to get the holes drilled. Once it's set up and running, you should see a finished part in days, not months. Plus, if the pattern is regular, you can set up to drill by rows, which should make creating the tool path much less taxing. You just hit a hole, and then rotate through the proper angle and hit another. Jumping between rows should take just a lateral movement and a rotation to locate the first hole of that row, and you're set to repeat the rotate, plunge, rotate, plunge drill.

At 3 holes a minute, total drilling time would be about two weeks, and you should be able to get way more than 3 holes a minute.

 

[Timelord]

buckley8,

You are wrong about the laser. Reread my post. If you use a PULSE laser of sufficient power you can ABLATE (not melt) the holes thru both the aluminum and the epoxy without heating up either. We use this technique to remove paint with a 1 in square focal spot and a 30 joule per pulse laser (at 200pps). We can remove paint one layer at a time without heating the substrate or the remaining paint. It will not ruin the epoxy. The blow back (off gas) can be handled by using a final optic with a long enough focal length and a special nozzle. You just need to locate a laser vendor who understands the problem and is willing to work with you.

Timelord

 

[rb1957]

"The two-part epoxy (3M p/n:EC-3524) is applied with a spatula or trowel and cures at room temp" ... and you think you're going to get a constant thickness of 0.05" ??

with a spatula inside a 21" dia tube ?

i think there has to be a better way; I mean, you are NASA.

 

[buckley8]

To rb1957:

The ID is machined after the epoxy cures. This has been done before several times. I didn't mention it since I didn't think it mattered how the epoxy thickness was arrived at. It's the holes that are new.

I just work for a contractor. NASA comes up with the far-out ideas. We then make them work. This is for aeronautical acoustic research (why aircraft engines are already so much quieter than they used to be and will eventually be virtually inaudible outside an airport).

I'm sure there's a better way. That's why I'm here!

Jim Buckley
ASRC Aerospace

 

[rb1957]

many ways to skin a cat ...

cats tend not to like any of them !

 

[hydtools]

I'd go to the water jet process. No heat affect, various materials, accuracy, programable.

http://www.flowcorp.com/waterjet-resources.cfm?id=268

http://www.flowcorp.com/uploadedFil...terjet_Brochures/Waterjet Cutting Systems.pdf

Ted

 

[vesselfab]

flow does make a 5-axis machine, but I do not know if they can hold that tight of a tolerance on those .020" dia holes

 

[minerk]

Water jets tend to enlarge the holes at the bottom of the cut. Would it be possible to make the holes, THEN machine the epoxy?

 

[hydtools]

minerk, according to Flow the hole is larger at the entrance not the exit.

vesselfab, rb1957 has stated no positional tolerance. The hole tolerance is .019 to .024 inch.

Use two aligned jets and cut holes half depth from each side if taper exceeds the tolerance range.

Drilling the holes with a drill bit will leave a burr that probably will have to be removed. Water jet will leave no burr.

Ted

 

[IRstuff]

Kenat,

That's why I downgraded by a factor of 5...


Actually, not, but the point was that even at a much slower step rate, it wouldn't take 3 months to drill. The full stroke retract time on drill is 18 ms. Assuming a drill rate of 20 in/min, the drill dwell time is 0.6 seconds.

Based on the manufacturer's velocity spec, it should take less than 10 milliseconds to move to the next drill position. The estimated cycle time is therefore less than 0.7 seconds per hole, resulting in a single piece drill time of about 25 hours, so 3 days, instead of my WAG of 2 days.

In any case, substantially shorter than 3 months, so definitely worth checking out.

TTFN

FAQ731-376

 

[sreid]

I'm all for the laser. We do this on a lathe (much smaller, fewer holes but still the same concept).

The part in the spindle rotates at constant RPM and the laser is fired based on hardware position matches in the controller. It is fired briefly for each hole. This has two advantages;

No time is lost from starting and stoping.

The holes are drilled a little bit with each laser pulse (minimizes heating and the ablated material chuff is small for each pulse).

Many modern machine tool controllers have the hardware match feature to drive the laser trigger input.

 

[NickE]

steps:

1) laser teh holes in teh Al.
2) Fill holes w/ wax (also wrap a matching mold on the ID, that creates a waxbutton .075" tall from each hole.)
3) Machine ID to get tight ID tolerance.
4) Melt the wax off and clean the part.


Hmmm?

(think casting, not machining) Cast the holes in the Epoxy.

 

[stodge45]

I wish I hadn't read this thread. I have not been able to sleep thinking of a solution. Anyway, try this: Purchase 63,000 sewing pins. The ones in my wifes sewing tin are 0.5mm dia. EDM drill the holes in the aluminium cylinder.
Insert the pins into the holes until the parallel portion protrudes no more than necessary. Apply the epoxy. Make a small slide hammer with a claw on the end that will just fit over the pin dia. Pull all of the pins out!
Stodge.

 

[MikeHalloran]

Sandwich the epoxy between two aluminum rings... yes, bonded to both.

Drill blind holes from the outside on a CNC machine.

Turn the inner ring to chips.



Mike Halloran
Pembroke Pines, FL, USA