Help!! Need an enclosure FAST...

[John Miles, KE5FX]

On Saturday, May 11, 2019 at 3:08:04 PM UTC-7, Rick C wrote:

Two, the emissivity effect you are talking about indicates the anodized
enclosure is actually better for getting rid of the heat.

Right, it doesn't conduct the heat, so much as it sheds it. These effects
are related to some extent.

-- john, KE5FX

 

[Rick C]

On Saturday, May 11, 2019 at 8:58:22 PM UTC-4, Lasse Langwadt Christensen wrote:

søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.56.05 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 6:22:32 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.11.44 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 5:30:22 PM UTC-4, Lasse Langwadt Christensen wrote:
lørdag den 11. maj 2019 kl. 21.34.09 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 8:39:29 AM UTC-4, Lasse Langwadt Christensen wrote:

and as I recently found out the thermal conductivity of the anodizing is absolutely horrible

Why would that matter? Anodizing is pretty thin. Nothing is a good thermal insulator if it is thin enough. What type of anodizing are you talking abougt? How bad is this supposed to be? I'm not familiar with the "horrible" spec.


the usually grey anodizing on extruded alu enclosures with heat sink fins

I didn't do the measurements but milling off the anodizing where a transistor was bolted to the inside made a big difference

I don't know what to tell you. Perhaps there was some other effect like the anodized surface not being very flat? Anodized layers are less than 1 mil up to 5 mils depending on type. It's pretty hard to get much insulating properties in 5 mils of thickness. How many watts? How large of a surface? What was the observed temperature change?


I didn't do the measurements, the surface is ~10mmx30mm, maybe 50watt

I did a quick calculation based on 30 W·m−1·K−1 3 cm^2 at 5 mil thickness (hard anodizing) would be about 70 W/K. That would seem to be less than one degree Kelvin from removing the anodized layer.


yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.


https://www.sciencedirect.com/science/article/pii/S0254058413004409

Did you actually read the paper? It is not clear to me just what they were doing. Without the paper you can't tell if what they were measuring is the same as the anodizing done for electronics.

Their results are 20:1 different at least from other published results. So it seems hard to believe without understanding what they measured and how it would apply to other uses. Then there is always the matter of unverified research. They seem to want $35 for the paper.

--

Rick C.

--- Get a 5,000 miles of free Supercharging
--- Tesla referral code - https://ts.la/richard11209

 

[Lasse Langwadt Christense]

søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:

On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.56.05 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 6:22:32 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.11.44 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 5:30:22 PM UTC-4, Lasse Langwadt Christensen wrote:
lørdag den 11. maj 2019 kl. 21.34.09 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 8:39:29 AM UTC-4, Lasse Langwadt Christensen wrote:

and as I recently found out the thermal conductivity of the anodizing is absolutely horrible

Why would that matter? Anodizing is pretty thin. Nothing is a good thermal insulator if it is thin enough. What type of anodizing are you talking abougt? How bad is this supposed to be? I'm not familiar with the "horrible" spec.


the usually grey anodizing on extruded alu enclosures with heat sink fins

I didn't do the measurements but milling off the anodizing where a transistor was bolted to the inside made a big difference

I don't know what to tell you. Perhaps there was some other effect like the anodized surface not being very flat? Anodized layers are less than 1 mil up to 5 mils depending on type. It's pretty hard to get much insulating properties in 5 mils of thickness. How many watts? How large of a surface? What was the observed temperature change?


I didn't do the measurements, the surface is ~10mmx30mm, maybe 50watt

I did a quick calculation based on 30 W·m−1·K−1 3 cm^2 at 5 mil thickness (hard anodizing) would be about 70 W/K. That would seem to be less than one degree Kelvin from removing the anodized layer.


yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.

https://www.sciencedirect.com/science/article/pii/S0254058413004409

 

[Rick C]

On Saturday, May 11, 2019 at 9:48:44 PM UTC-4, Clifford Heath wrote:

On 12/5/19 11:08 am, Rick C wrote:
On Saturday, May 11, 2019 at 8:58:22 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.


https://www.sciencedirect.com/science/article/pii/S0254058413004409

Did you actually read the paper? It is not clear to me just what they were doing. Without the paper you can't tell if what they were measuring is the same as the anodizing done for electronics.

Their results are 20:1 different at least from other published results. So it seems hard to believe without understanding what they measured and how it would apply to other uses. Then there is always the matter of unverified research. They seem to want $35 for the paper.


Anodising layer is very porous. You can see even from the abstract that
reduces thermal conductivity radically.

Radically??? Worse case, it would be approximately in proportion to the pore area, but when filled with heat sink grease, I expect somewhat less. Are you suggesting the anodizing is well over 50% pores? To explain a 20:1 difference it would need to be 95% pores. Sounds more like a net, holes tied together.

I can't tell anything from the abstract about the thermal conductivity of anodizing on heat sinks because I don't know anything about how they created the anodizing in the study or how they measured the thermal conductivity. Can you?

--

Rick C.

--+ Get a 5,000 miles of free Supercharging
--+ Tesla referral code - https://ts.la/richard11209

 

[Clifford Heath]

On 12/5/19 11:08 am, Rick C wrote:

On Saturday, May 11, 2019 at 8:58:22 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.56.05 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 6:22:32 PM UTC-4, Lasse Langwadt Christensen wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.


https://www.sciencedirect.com/science/article/pii/S0254058413004409

Did you actually read the paper? It is not clear to me just what they were doing. Without the paper you can't tell if what they were measuring is the same as the anodizing done for electronics.

Their results are 20:1 different at least from other published results. So it seems hard to believe without understanding what they measured and how it would apply to other uses. Then there is always the matter of unverified research. They seem to want $35 for the paper.

Anodising layer is very porous. You can see even from the abstract that
reduces thermal conductivity radically.

 

[Jasen Betts]

On 2019-05-11, Lasse Langwadt Christensen <langwadt@fonz.dk> wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1
and thought no problem, but then I looked up numbers for
anodizing and found multiple papers where they had measured
it as more like 1 W·m−1·K−1

But it's so thin! does this number really matter?

--
When I tried casting out nines I made a hash of it.

 

[Clifford Heath]

On 13/5/19 6:16 pm, Jasen Betts wrote:

On 2019-05-11, Lasse Langwadt Christensen <langwadt@fonz.dk> wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1
and thought no problem, but then I looked up numbers for
anodizing and found multiple papers where they had measured
it as more like 1 W·m−1·K−1

But it's so thin! does this number really matter?

Anodising is very porous, it seems. That's how it takes dye.

 

Clifford Heath <no.spam@please.net> wrote in news:CycCE.12432$%
55.10902@fx43.iad:

On 13/5/19 6:16 pm, Jasen Betts wrote:
On 2019-05-11, Lasse Langwadt Christensen <langwadt@fonz.dk
wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1
and thought no problem, but then I looked up numbers for
anodizing and found multiple papers where they had measured
it as more like 1 W·m−1·K−1

But it's so thin! does this number really matter?

Anodising is very porous, it seems. That's how it takes dye.

Bare Aluminum is porous. Anodizing is not so much as a hard
anodized Aluminum surface is non conductive. If it were 'porous' as
you state, it would fail at a very low voltage. It does not, so the
surface treatment must be pretty complete.

The final step in anodizing Aluminum is a water finish to close up
the 'pores' on the oxidized layer. That is also when they add the
tinting. After that prossess the pores are sealed up by the water
reacting with the oxide layer.

Chem etch and chem film is NOT anodizing.

 

[George Herold]

On Saturday, May 11, 2019 at 8:58:22 PM UTC-4, Lasse Langwadt Christensen wrote:

søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.56.05 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 6:22:32 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.11.44 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 5:30:22 PM UTC-4, Lasse Langwadt Christensen wrote:
lørdag den 11. maj 2019 kl. 21.34.09 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 8:39:29 AM UTC-4, Lasse Langwadt Christensen wrote:

and as I recently found out the thermal conductivity of the anodizing is absolutely horrible

Why would that matter? Anodizing is pretty thin. Nothing is a good thermal insulator if it is thin enough. What type of anodizing are you talking abougt? How bad is this supposed to be? I'm not familiar with the "horrible" spec.


the usually grey anodizing on extruded alu enclosures with heat sink fins

I didn't do the measurements but milling off the anodizing where a transistor was bolted to the inside made a big difference

I don't know what to tell you. Perhaps there was some other effect like the anodized surface not being very flat? Anodized layers are less than 1 mil up to 5 mils depending on type. It's pretty hard to get much insulating properties in 5 mils of thickness. How many watts? How large of a surface? What was the observed temperature change?


I didn't do the measurements, the surface is ~10mmx30mm, maybe 50watt

I did a quick calculation based on 30 W·m−1·K−1 3 cm^2 at 5 mil thickness (hard anodizing) would be about 70 W/K. That would seem to be less than one degree Kelvin from removing the anodized layer.


yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.


https://www.sciencedirect.com/science/article/pii/S0254058413004409

Huh, Interesting thanks. 5 mil for the anodized layer seems like a lot.

(OK 5 mil is 'common'
https://www.defelsko.com/resources/anodizing-thickness-measurement-on-aluminum)

George H.

 

[Carl]

"Clifford Heath" wrote in message news:CycCE.12432$%55.10902@fx43.iad...

On 13/5/19 6:16 pm, Jasen Betts wrote:
On 2019-05-11, Lasse Langwadt Christensen <langwadt@fonz.dk> wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1
and thought no problem, but then I looked up numbers for
anodizing and found multiple papers where they had measured
it as more like 1 W·m−1·K−1

But it's so thin! does this number really matter?

Anodising is very porous, it seems. That's how it takes dye.

The oxide layer formed during anodizing is porous and kind of "fluffy" and
damages easily but takes dye well. At this point you can wipe most of the
oxide off with your hand. The final step in anodizing with or without dye
is to simmer in boiling water, usually with nickel acetate and maybe other
proprietary goodies added. This changes the crystal structure of the
aluminum oxide, collapses the pores in the oxide layer which traps the dye
making the color durable, and the densification greatly improves the
abrasion resistance of the oxide layer.

--
Regards,
Carl Ijames

 

[Phil Hobbs]

On 5/13/19 11:53 AM, Carl wrote:

"Clifford Heath"  wrote in message news:CycCE.12432$%55.10902@fx43.iad...

On 13/5/19 6:16 pm, Jasen Betts wrote:
On 2019-05-11, Lasse Langwadt Christensen <langwadt@fonz.dk> wrote:

yes I also looked up aluminium oxide and saw 30 W·m−1·K−1
and thought no problem, but then I looked up numbers for
anodizing and found multiple papers where they had measured
it as more like 1 W·m−1·K−1

But it's so thin! does this number really matter?

Anodising is very porous, it seems. That's how it takes dye.

The oxide layer formed during anodizing is porous and kind of "fluffy"
and damages easily but takes dye well.  At this point you can wipe most
of the oxide off with your hand.  The final step in anodizing with or
without dye is to simmer in boiling water, usually with nickel acetate
and maybe other proprietary goodies added.  This changes the crystal
structure of the aluminum oxide, collapses the pores in the oxide layer
which traps the dye making the color durable, and the densification
greatly improves the abrasion resistance of the oxide layer.

I believe that the porous film is sapphire (Al2O3). It has to be porous
so that ionic current can still pass through it, or the oxide stops
growing. I don't think further chemical processing touches the
sapphire, just the underlying metal and the pores.

The original process just used the hot water to form aluminum hydroxide
to fill the pores. That works OK for some things, but as you say, you
can fill them with other stuff, e.g. fluorpolymers for Tufram/Polylube.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Rick C]

On Monday, May 13, 2019 at 9:22:01 AM UTC-4, George Herold wrote:

On Saturday, May 11, 2019 at 8:58:22 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 02.07.02 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 7:13:25 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.56.05 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 6:22:32 PM UTC-4, Lasse Langwadt Christensen wrote:
søndag den 12. maj 2019 kl. 00.11.44 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 5:30:22 PM UTC-4, Lasse Langwadt Christensen wrote:
lørdag den 11. maj 2019 kl. 21.34.09 UTC+2 skrev Rick C:
On Saturday, May 11, 2019 at 8:39:29 AM UTC-4, Lasse Langwadt Christensen wrote:

and as I recently found out the thermal conductivity of the anodizing is absolutely horrible

Why would that matter? Anodizing is pretty thin. Nothing is a good thermal insulator if it is thin enough. What type of anodizing are you talking abougt? How bad is this supposed to be? I'm not familiar with the "horrible" spec.


the usually grey anodizing on extruded alu enclosures with heat sink fins

I didn't do the measurements but milling off the anodizing where a transistor was bolted to the inside made a big difference

I don't know what to tell you. Perhaps there was some other effect like the anodized surface not being very flat? Anodized layers are less than 1 mil up to 5 mils depending on type. It's pretty hard to get much insulating properties in 5 mils of thickness. How many watts? How large of a surface? What was the observed temperature change?


I didn't do the measurements, the surface is ~10mmx30mm, maybe 50watt

I did a quick calculation based on 30 W·m−1·K−1 3 cm^2 at 5 mil thickness (hard anodizing) would be about 70 W/K. That would seem to be less than one degree Kelvin from removing the anodized layer.


yes I also looked up aluminium oxide and saw 30 W·m−1·K−1 and thought no problem, but then I looked up numbers for anodizing and found multiple papers where they had measured it as more like 1 W·m−1·K−1

That seems pretty unrealistic. Are you perhaps not getting the right units, with the thickness already having been taken into account or something? That's what I would expect from a paper analyzing thermal conductivity of anodizing.


https://www.sciencedirect.com/science/article/pii/S0254058413004409

Huh, Interesting thanks. 5 mil for the anodized layer seems like a lot.

(OK 5 mil is 'common'
https://www.defelsko.com/resources/anodizing-thickness-measurement-on-aluminum)

George H.

Depends on what you ask for. There are multiple types of anodizing. Chromate is typically 1 mil or less. Sulfuric acid is used to make anodizing up to 5 mil... well, I think I read it was sulfuric acid. Wikipedia has an ok page and lots of anodizers also have pages. The thicker sulfuric acid anodizing is called "hard" anodizing.

--

Rick C.

-+- Get a 5,000 miles of free Supercharging
-+- Tesla referral code - https://ts.la/richard11209