Grunff wrote:
Andy Dingley wrote:

I was pondering this strange behaviour of UV light, and why it has to
be of a certain minimum wavelength to have any effect


Isn;t it just a photon energy threshold? Above a certain wavelength, the
photon energy isn't enough to kick the appropriate electron, below that
wavelength it is?

Not really. Most of the UV glues I have used have an absorption peak
matched to the 365nm mercury line (they are transparent again below
this). This will penetrate a thin layer of normal glass. Some also
have an absorption in the deep blue which helps them to cure in sunlight
("Glass Bond" is a DIY brand which IIRC will cure with UV or blue).
Dymax is one brand we use at work, Norland is another, though most of
their UV glues are designed and priced for use with precision optics.

"Disco" blacklights are 365nm, but there's not much of it, and what
there is is quite spread out, however that should get most of these
glues to go off. Germicidal/EPROM erase lamps (as pointed out
elsewhere) won't get through the glass and aren't guaranteed to set the
glue off anyway.

These glues can be extremely strong and can bond metal to glass as well
as glass to glass. A test we did at work was to bond a 20mm square 3mm
thick mirror along one edge to a bracket inside a dead PC case.
Dropping the case onto a solid concrete floor from ~1m rarely broke the
bond but frequently broke the glass (cleanly).

Both surfaces should be completely clean, especially grease free.

There is one other issue I discovered when attenpting to cure UV glue
with a ~370nm UV LED: If the intensity is too low you will get a
partial cure (tacky with no strength) which no length of exposure will
improve on.

--
Spamtrap in use
To email replace 127.0.0.1 with blueyonder dot co dot uk