Kafuter thermal glues

9 posts / 0 new
Last post
Agro
Agro's picture
Online
Last seen: 2 min 19 sec ago
Joined: 05/14/2017 - 11:16
Posts: 6942
Location: Ślōnsk
Kafuter thermal glues

Yes, I know that Fujik is the most popular thermal glue on this forum. But seeing specs of various Kafuter glues (K-5202, K-5204K) I see that they have nearly double thermal conductivity (1.6W/m*K) while K-5204K is cheaper.
So…can I trust those specs? And aside from that, does anyone have any experience with them?

Enderman
Enderman's picture
Offline
Last seen: 2 hours 58 min ago
Joined: 11/03/2016 - 22:42
Posts: 4287
Location: Vancouver, Canada

Arctic silver thermal adhesive or a standard CPU thermal paste will have far higher thermal conductivity than that sad 1.6w/mk

Agro
Agro's picture
Online
Last seen: 2 min 19 sec ago
Joined: 05/14/2017 - 11:16
Posts: 6942
Location: Ślōnsk

I wanted something soft, so I didn’t like epoxy. But now, when checking Arctic conductivity (9W/m*K) I learned a new trick:
“you can mix about two parts of the adhesive with one part regular paste to achieve a compromise”
I guess that would conduct even slightly better. Though Alumina is 15 times as expensive (and after adding good paste it only gets worse), it still looks like a better option.

Enderman
Enderman's picture
Offline
Last seen: 2 hours 58 min ago
Joined: 11/03/2016 - 22:42
Posts: 4287
Location: Vancouver, Canada

If it needs to be soft then there are high quality thermal pads with conductivities of 10w/mk or more.
They also cost like a hundred dollars or so depending on how big.

kiriba-ru
kiriba-ru's picture
Offline
Last seen: 1 day 15 hours ago
Joined: 01/17/2016 - 02:34
Posts: 2130
Location: Russia

Agro wrote:

“you can mix about two parts of the adhesive with one part regular paste to achieve a compromise”

and get random thermal conductivity.
Agro
Agro's picture
Online
Last seen: 2 min 19 sec ago
Joined: 05/14/2017 - 11:16
Posts: 6942
Location: Ślōnsk

Enderman wrote:
If it needs to be soft then there are high quality thermal pads with conductivities of 10w/mk or more.
They also cost like a hundred dollars or so depending on how big.

I can’t use them for potting. But thanks for the suggestion anyway. Wink
kiriba-ru wrote:
Agro wrote:

“you can mix about two parts of the adhesive with one part regular paste to achieve a compromise”

and get random thermal conductivity.

Why random? Care to elaborate?
kiriba-ru
kiriba-ru's picture
Offline
Last seen: 1 day 15 hours ago
Joined: 01/17/2016 - 02:34
Posts: 2130
Location: Russia

It doesnt work as you think.
Cu have 360 thermal cond., aluminium alloy 200. Alloy which consist both are usualy under 100.

Enderman
Enderman's picture
Offline
Last seen: 2 hours 58 min ago
Joined: 11/03/2016 - 22:42
Posts: 4287
Location: Vancouver, Canada

kiriba-ru wrote:
It doesnt work as you think.
Cu have 360 thermal cond., aluminium alloy 200. Alloy which consist both are usualy under 100.

That’s because metallic alloys can have different crystal structures at different compositions, and those structures have different properties.

Unless the two pastes he mixes have some sort of chemical reaction the final thermal conductivity will be somewhere between both.
kiriba-ru
kiriba-ru's picture
Offline
Last seen: 1 day 15 hours ago
Joined: 01/17/2016 - 02:34
Posts: 2130
Location: Russia

Enderman wrote:
kiriba-ru wrote:
It doesnt work as you think.

That’s because metallic alloys can have different crystal structures at different compositions…

Unless the two pastes he mixes have some sort of chemical reaction the final thermal conductivity will be somewhere between both.


Disagree.
If you will build house with each even row from concrete bricks and each odd row from polyfoam brick, you wont get avarage thermal conductivity.
Each thermal grease/compound consist of two components – tiny particles of some metall and some liquid polymer. Main secret – size of this particles and density of the grease should have right proportions, that allow you apply very thick layer (close to particles size) of compound. If you just
Agro wrote:
you can mix about two parts of the adhesive with one part regular paste to achieve a compromise

you can easily meet situation when particles are far away from each other and cant influence on thermal conductivity.