Inexpensive inert atmosphere UV curing source

C-Lab Home
A customer's product needed to be encapsulated in a very inert substance to be used on the human body.  We selected a UV cure encapsulant as the best solution, but then needed a way to cure the stuff for prototypes.  A little web searching showed us that the UV cure business can be an expensive club to join, with a minimal source costing on the order of $3,000.

This offended Doug's Scottishness, and the result was this project.



First we obtained a nearly airtight steel box with a hinged lid, of the sort used for holding electrical breakers outdoors.  At the same place (State Electric, a fine outfit) we obtained a 250 watt mercury vapor bulb and a ballast kit.  Here are the steps to a cheap UV source that will cure most substances on the market tack free very quickly.
  1.  Make the box reasonably airtight.
  2. Install the fitting for the argon or other inert gas near the bottom.
  3. Prepare the bulb -- cut off the outer envelope.
  4. Install the bulb and ballast.
  5. Install the reflectors.
  6. Proceed to use the device successfully.
1. Box dimensions: 10" x 12" x 5.5" high.  A somewhat larger box is fine, but takes longer to fill with whatever inert gas you'll use.  The box we obtained was airtight already, with welded seams and a gasketed lid.  If your box is not, use silastic to make it tight.  We like to use the high temperature red stuff you can get at the auto parts store even in relatively low temperature applications such as this one, as it is somewhat tougher than the run of the mill stuff you get at the hardware store for use in a caulking gun.

2.  We needed something that would allow the argon into the box, that would fit the tubing we have, not leak, and be strong enough to last through a little rough handling.  A quick look around the place located a .223 Remington shell which looked ideal for the job.  After necking it up a little in my 6mm PPC die, we cut the case head off.  Drilling a 3/8" hole near the bottom of the box gives a very nice, tight fit as the shell body is tapered.  We didn't have to get out a welder and mess up the paint, or even use any sealant.  All it took was a few gentle hammer taps to set the shell into the hole permanently.

3.  The streetlamp bulb we obtained has an outer glass envelope that blocks UV and is phosphor coated to whiten the light somewhat.  This has to go, we want direct exposure to the UV from the internal quartz capsule.  One could simply crush this in a vise, but we have a better way to handle glass here.  Put a diamond wheel in a dremel tool, with the whole thing mounted in a dremel drillpress.  Set the height appropriately so the cut will be near the mogul bulb base.  Now hold the bulb gently against the spinning wheel while rotating it carefully.  You will want to be careful not to tilt the bulb or do anything to let the line not be straight and joined at the end.  You could simply cut all the way through, but hey, that's work!  Here, we use 20 mil nichrome wire (from Alpha Aesar) in a loop, wrapped into the cut line.  Putting a few amps through this heats it up enough to shatter the glass in a perfectly clean cut.  If you're safety minded,  safety glasses are a good idea during all of this, as is sanding down the sharp edges while wearing work gloves -- the glass will sometimes cut the sandpaper.  You can imagine what would happen if you weren't wearing some tough gloves at that point.

4.  We made a little angle bracket to mount the mogul socket to the lid of the box.  Since the box is thick steel, we tapped 10-32 holes in it to mount the bracket, and needed no sealant on these.  We mounted the ballast to the bottom floor of the box.  We could have used silastic here, but got serious about it and tapped the ballast holes to 1/4-20 and used real bolts through the box into the tapped holes.  To be honest, silastic would have been a better idea, as the firm mounting provided by the bolts helps the box transmit the hum of the ballast transformer.  But we were impatient to try the thing, and didn't want to have to wait for another cure time.  No chance of leaving this thing on by accident!

5.  Aluminum foil makes a great cheap reflector for UV light.  Silastic in a thin layer holds it to the box surfaces nicely, as would contact cement, but remember, the stuff right behind the bulb is going to get pretty hot.  We cut pieces out to match each side of the box and glued them in.   It's not necessary to have it perfectly smooth, but don't crumple it up first either.

6. Use.  If you don't fill the box with an inert gas such as argon (nitrogen would probably work too) you will get an absolutely eye-watering amount of ozone from the shortwave UV the bulb emits while warming up.  We are talking well past "springtime fresh" here.  The effect is more like "Run, someone just released 30 cu ft of chlorine in the room".  Also, the surface layer of your UV cure substance won't cure tack-free, as the oxygen inhibits this.  It eventually does cure without using the inert gas, but it can take up to 24 more hours of sitting around, during which time it picks up dust.  Here, once we tried argon, borrowed from our MIG welder, we never looked back.  The little bit of extra cost is way worth it in practice.  The type of regulator you use with a welder is ideal for this, as we know the size of the box and the regulator is calibrated in cubic feet per hour.  A little math and you know how long to wait before starting up the lamp.  We put our parts to be cured into the box, start the argon flow, slow is better as you can use less.  The argon is heavier than air, so if you don't create a bunch of turbulence in the box, you can use just about exactly the box volume, and not need more.  This is why the fitting is at the bottom, so the displaced air just leaks out of the top.  During the run, we leave the gasketed top loose, and let the weight of the steel seal it "well enough".  Here we find that a 4-5 minute run, including lamp warmup time, works great.  Too much longer than this, and your part had better be fairly heat resistant.  Too much shorter and you get no cure.  The mercury in the bulb has to become vapor at medium pressure to put out the crucial 360 nm spectral line needed by the UV curing plastics.  Once there, it only takes a few seconds to actually do the curing.  I suppose if you were doing a production line sort of thing, this wouldn't be the most practical way.  In that case, use some of your money from selling the product and buy the professional model.

Total costs:

Box:  about $20 (actually, it was in our junk box for free)
.223 shell -- free to 30 cents, depending.
Bulb and ballast:  about $100 from State Electric.
Line cord:  In our junk box, call it $3 to buy an extension cord and cut off one end.
Aluminum foil: perhaps 10 cents worth
Silastic:  maybe a dollars worth of the expensive red stuff.
Time: less than one engineer-day, call it $200 worth.

Total cost:  $324.40 -- less than 1/10th of what it would normally cost to join this "club".
Here at C-Lab, that sort of fat "X" factor is what turns us on.  No wimpy 10% savings for this crowd!  Even if we add the cost of buying an argon tank and regulator at the local welding supply house, we only add about $120 dollars to this, and still have a fat factor, and a tool we know how to fix when it breaks.

Late breaking news:
We found that after some hours of operation, the shortwave UV from this source is eliminated due to "solarization "of the quartz lamp. Thus the argon is no longer needed after this period. There will be an update soon at: Coulter's Smithing, where we are using a newer version of this source for exposing other photosensitive things.

C-Lab Home