I have a temperature sensor in a -20C freezer. The probe is going into a bottle filled with propylene glycol solution. I want to seal the hole where the probe is inserted into the bottle.
What's a good sealant for this?
$\begingroup$
$\endgroup$
4
-
$\begingroup$ put the probe down through the neck of the bottle. $\endgroup$– Solar MikeMay 14, 2018 at 20:33
-
$\begingroup$ still need to seal it somehow though - i want to make sure the contents don't leak out if the bottle is tilted on it's side $\endgroup$– Misha MMay 14, 2018 at 20:38
-
$\begingroup$ Does the lid of the bottle need to be HDPE also? $\endgroup$– Jonathan R SwiftMay 14, 2018 at 20:53
-
$\begingroup$ That's what the container came with $\endgroup$– Misha MMay 14, 2018 at 20:54
Add a comment
|
6 Answers
$\begingroup$
$\endgroup$
1
Bonding to HDPE is very difficult. A few adhesives developed fairly recently do ok in some applications, but that kind of cooling will likely cause them to fail.
Instead, I'd recommend a mechanical seal. I would (and often have for this exact purpose) use a bulkhead fitting to seal against the wall and an instrument fitting, aka compression tube fitting to seal against the temperature sensor.
-
$\begingroup$ This one is good like a round rubber band you stretch through a hole or an ear plug. $\endgroup$– user4139Nov 8, 2018 at 21:55
$\begingroup$
$\endgroup$
6
Try this stuff. It's about as good as it gets off the shelf.
-
$\begingroup$ This adhesive will not bond to HDPE :-( $\endgroup$ Jun 4, 2018 at 22:49
-
$\begingroup$ Yes it will. This from the King Starboard site - "Products like 3M’s 5200 work well as a water sealing caulk but will not adhere King StarBoard® to itself or other materials in a permanent structural bond. It is preferable to mechanically fasten or weld King StarBoard®, but when an adhesive is necessary you can use a product called Lord 7542-AB, or 3M’s Scotch-Weld DP-8005, or Chem-Set™ 6105 Polyolefin Bonder." kingplastic.com/using-adhesive-with-king-starboard-2 $\endgroup$ Jun 5, 2018 at 1:34
-
$\begingroup$ 3M is pretty convinced it will, but DP8010 looks like it might be a better suited product, and 3M calls it a 'Key Product'. $\endgroup$ Jun 6, 2018 at 3:57
-
$\begingroup$ Try it, see how long the "bond" lasts :-) $\endgroup$ Jun 6, 2018 at 14:44
-
$\begingroup$ This family of adhesives works much better for flush surfaces, as you'll find in the product literature. It basically doesn't work to seal gaps around random misfit shapes. $\endgroup$ Jun 6, 2018 at 18:17
$\begingroup$
$\endgroup$
I suggest buying extra caps/stoppers and drill a close-fitting hole. If the caps (and your probe/wires can take it), perhaps use some hot-melt glue for a mechanical seal.
$\begingroup$
$\endgroup$
2
JB Water weld works very well for me in many applications like this. It sets under water and is not brittle. I built a lip for a water pump that eroded away. Please let us know what you went with?
-
$\begingroup$ Ended up going with a container with glass beads instead of a liquid. That eliminated the whole problem :) $\endgroup$– Misha MNov 7, 2018 at 17:23
-
$\begingroup$ @MishaM Can you put that as an answers on the process? $\endgroup$– user4139Nov 8, 2018 at 21:08
$\begingroup$
$\endgroup$
You may actually get your best bet by encasing the entire thing in shrink wrap (depending upon the size of the bottle).
Shrink wrap is typically made of LDPE. Due to the low density, the LDPE has many sites where it can still cross-link upon itself. Heating the wrap begins this reaction, and as the material cross-links, it shrinks to form HDPE. This shrinking action forms a good seal. In the area of the hole, additional heating could be applied until the LDPE and HDPE entered a plastic state, welding the two materials together.
Other methods, such as acrylics, work well, but this would insure similar chemical resistance inside and outside the freezer.
$\begingroup$
$\endgroup$
In the end, I decided to use a bottle filled with glass beads. This provided the necessary thermal stabilization and it has the added benefit of working below -50C.