Pill/reflector coming unscrewed with heat

[Duster1671]

I've been messing around with a custom build using a Lumens Factory KL4 head and H17Fx driver. I've built this up with an XP-L HI and a 219b (w/max output programmed down) and I'm having a nagging problem with the brass pill and aluminum reflector coming unscrewed from one another when the head heats up.

The Lumens Factory head consists of a brass pill with male threads that screws into the back side of the aluminum reflector. This assembly then drops into the KL4 head and is retained with the lens, o-ring, and bezel ring. When I run the light on turbo for a minute and then remove the head from the body, there is invariably a rattling sound when I shake the head. Disassembly reveals that the brass pill and aluminum reflector are no longer screwed tightly together. These threads being tight is what puts pressure on the MCPCB for a good thermal connection, and in one case I cooked an LED when the reflector loosened up.

I'm guessing this has to do with differences in thermal expansion between brass and aluminum, but I don't know for sure. What would be the best way to keep this from happening? Ideas so far...

1. Loc-tite - I could use loc-tite between the reflector and pill, but I'm not sure if 242 (blue) would be good enough considering the heat
2. Silicone - another thought is silicone RTV on the threads. Maybe better under heat, but also not sure about this...
3. Staking - perhaps I could screw the reflector and pill together and then stake them by hitting across the seam with a cold chisel. Again, not sure if this will work and how difficult it would be to get the components apart in the future if needed.

Any ideas or suggestions would be appreciated!

 

[scalpel_ninja]

I would start with a weaker Loctite than blue, such as purple. Purple is meant to hold adjustable parts in place, such as tension screws, control knobs, etc.

To maintain good metal to metal contact, I wouldn't try to apply it wet and thread in because this would spread the solution over 360 degrees of the thread. If space allows, perhaps place a dot of the purple on one spot, place the part in a Ziploc bag (anaerobic curing) and then install the part once it's set.

If purple is too weak, there's always blue to try.

 

[Duster1671]

That's an interesting idea about pre-curing the loctite. I don't think it would create a bond to the part installed after curing, but it might gum up the threads enough to prevent loosening.

Another thought was to put a drop of two-part epoxy on the seam after assembly. This would probably keep things together and it could be flicked off with a screwdriver if I need to do some rework.

 

[LEDphile]

Aluminum has about 30% more thermal expansion than brass. I'd forget about trying to prevent things from moving - the thermal expansion will either result in high stress in the parts, or simply break things loose. Instead, I'd try to build some compliance into the joint so that the expansion and contraction can occur but won't cause any issues. Something like a wave washer or a split-lock washer to provide some spring action, for example.

 

[fulee9999]

you can also try fixing the reflector to the pill with JB weld if you don't need to dissasemble it anytime soon

 

[sween1911]

I've had the same thing with the generic P60 drop-in kits. Screwing and unscrewing would make the pill back out of the reflector. I've just tried to crank them down tightly. That might not always be an option.

If you want to use Loctite, get your hands on some Purple. It's good for aluminum and brass and offers flexibility.

 

[Duster1671]

I think unscrewing the head from the body is a factor here. When I tighten the head, the end of the body jams into the solder blobs that bridge between the driver and the pill. Loosening the head puts some torque on the pill, potentially acting to loosen it from the reflector. It's too bad the pill and reflector are not made with left-handed threads.

I tried a small bead of two-part epoxy between the pill and reflector and that didn't work at all. The epoxy sheared right at the seam.

I agree with LEDphile that we're probably talking about large forces here, as a result of the thermal expansion. I'm just not sure how one would build compliance into this particular type of joint.

 

[fulee9999]

I tried a small bead of two-part epoxy between the pill and reflector and that didn't work at all. The epoxy sheared right at the seam.

oh wow, what epoxy did you use? did you leave it to cure for a sufficient time?

 

[orbital]

+

Unscrewing the head your light / taking it apart, is unscrewing your drop-in.
Heat alone isn't going to disassemble a drop-in.