How disassemble epoxy / loctite glued bezel

[Walterk]

I need to remove the bezel of my flashlight Fenix TK40.
I want shorter focus length to a lens.
Immersing the head in boiling water didn't give slack after gentle force with vice and pliers.

How hot does a LED-head get, what temp are the electronics designed to withstand?
What temperature loosens epoxy-glue?
What temperature loosens loctite?


I know this question as been asked before, and I did several searches but didn't work.

Any helpfull suggestions welcome!

(And yes I know warrantys are lost when unscrewing as the electronics and glas are no longer gastight contained.)

 

[The Dane]

Loctite gets pliable at around 80 deg C and Epoxy around 100 deg C.
Try with a hairdryer and expect that it takes time for epoxy to soften up. I would say 3-5 min. and by heating with a hairdryer you're under the damaging temp for electronics (125-150 deg C) and out of harms way of water.

 

[wingnut86]

I actually put a glued Dorcy in the freezer for a couple of hours and it broke loose pretty easily with two pairs of rubber tipped channel locks. this method may not work with the Fenix though, could be, and probably is a better glue than that of the Dorcy...

 

[Walterk]

Thx, good info.

I might try hot water again, a bit longer submerged.
Temperature of air is harder to control I think.

Anyone else with experience, hints?

 

[Anglepoise]

I have successful used one of those small flame burners that run on propane.
Set the heat down and carefully and quickly, let the heat move over the area where the threads are. You want to localize the heat and although it sounds drastic, it has worked for me every time I have used it.

Good luck

 

[Justin Case]

I've put epoxied items in boiling water (100C) and the epoxy definitely gets soft and pliable at that temperature.

Assuming Loctite brand threadlocker, the temperature that you need to heat the thread area depends on whether you are dealing with blue, green, or red Loctite (red being the hardest to crack open).

See the datasheets for specific breakaway torques for various fasteners. In principle, you can open blue Loctite with just hand tools, no need to heat up the part.

The graphs below give you the relative strength reductions as a function of temperature.

150C is a very typical maximum junction temperature for ICs and LEDs. Examples include AMC7135, TPS61030 (used in the Badboy Nexgen boost driver), ST1S03 (used in the SOB buck driver), and Cree XR-E. All have a max junction temp of 150C. Both the TPS61030 and ST1S03 also list the max storage temp as 150C.


Blue Loctite

Green Loctite

Red Loctite

 

[tx101]

I just did my SF M6
it took three attempts
First attempt, left the head in a pan of simmering water for 5 minutes
that did not work. Then 10 minutes, then 20 minutes
After 20 minutes the loctite gave and I could unscrew
the bezel ring

 

[Justin Case]

I cooked my SF KL4 at 175F for about 10 min and it finally broke free. On the other hand, I'm still beating on a Blackhawk Gladius and a SureFire U2 (SSC P4 version). I've boiled and freezed the Gladius, cooked the U2 at 240F for 10 min, as well as used direct flame from a gas stove burner on both (probably too tentative with the flame). And then there was the KL4 that opened up just from pure torque.

 

[Walterk]

Really appreciate your 'wise-man's'voices on this matter ! Thank you.

Succeeded:
I have tried torch a few time again and it gave away.
The module for the led gave first, the glue was red. (I wasnt expecting that seam, after the second cool-fin.)
The bezel of the lens gave in later, the glue was blue.
For your information: all screwed connections have a rubber ring, and the glass has two itself.
The reflector and led-module are both glued in with red.
The housing didnt got too hot to hold, so that probably means the electronics havent been heated that much either.
The flashlight still functions anyway.

Update:
I succeeded in releasing the reflector from the cone.
After 3 minutes in boiling water it popped loose easily.
It has a snug fit, with 0.2mm glue around it, about 30% of the area was covered with red glue.

But! :
The coating of the aluminium reflector is diminished for the most part, it fell off or discoloured.
After cleaning it looks like shining copper.
However the effect on the beam is not as drastic as I expected it to be.
Somehow the gloss-factor seems more important than the color of the surface.




The reflector detached from the cone (no picture from)