does anodized layer impact heat transfer

[robincnn]

Type 2 and 3 Anodized layer is usually around 1.8 μm to 25 μm. Source wiki
Aluminum conductivity around 205W/(m K)
Hard anodized layer conductivity .7 to 1 W/(m K) here
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA191755&ved=0CB4QFjAA&usg=AFQjCNFX5PKDhRsYcsnUxFWfv5zW4IFUww

Would the anodized layer reduce the heat transfer between the LED and the heatsink?
Or the layer is too thin to have any issues with heat transfer

Thanks

 

[PICOGRAV]

You are asking a few different things in this post. Do you want to know if the anodizing will impact the thermal resistance where the COB is mounted or the total thermal resistance of the heatsink?

Hard anodizing would work the best and you should always sand it off where the COBs are mounted. There are two issues with hard anodizing, it is difficult to apply and you will be unable to color it.

 

[bicit]

Anodizing the heatsink increases the emissivity of the heatsink to the environment. However it would also increase the thermal resistance between the LED and the heatsink. So as pico said, you would want the portion where the LED was mounted to be as polished as practically possible.

Anodizing has very little (4-8%) effect on actively cooled heatsinks.

 

[robincnn]

Source wiki
moderate thickness 1.8 μm to 25 μm (0.00007" to 0.001")[9] are known as Type II in North America, as named by MIL-A-8625, while coatings thicker than 25 μm (0.001") are known as Type III, hardcoat, hard anodizing, or engineered anodizing

I was looking at this anodizing process
http://www.thefintels.com/aer/homealuminumanodizing.htm
Not sure if this 3 water and 1 acid with 12 volt gives a hard anodizing.

Thanks Picograv and Bicit

I will sand off the anodized layer and polishing it before fixing COB.

 

[robincnn]

You are asking a few different things in this post. Do you want to know if the anodizing will impact the thermal resistance where the COB is mounted or the total thermal resistance of the heatsink?

Hard anodizing would work the best and you should always sand it off where the COBs are mounted. There are two issues with hard anodizing, it is difficult to apply and you will be unable to color it.

Thermal resistance between heatsink and COB I understand that we can sand off the anodized layer so that Led sits on aluminum instead of anodized aluminum. Thanks

For second part, would anodizing impact total thermal resistance of active cooled heatsink. The anodized layer has lower conductivity so would the inner heatsink would be more hot? Or the anodized layer is too thin to have any impact on total thermal resistance of heatsink?
Or the cooling gained due to emissivity more than offsets the heat trapped by this low conductivity(insulating) thin anodized layer.

 

[alesh]

Thermal resistance between heatsink and COB I understand that we can sand off the anodized layer so that Led sits on aluminum instead of anodized aluminum. Thanks

For second part, would anodizing impact total thermal resistance of active cooled heatsink. The anodized layer has lower conductivity so would the inner heatsink would be more hot? Or the anodized layer is too thin to have any impact on total thermal resistance of heatsink?
Or the cooling gained due to emissivity more than offsets the heat trapped by this low conductivity(insulating) thin anodized layer.

Totally depends on how thick the layer is. It can be very beneficial for passive HS and it can also have negative effect on actively cooled HS.

 

[Rrog]

Not having an aluminum surface oxidizing would be great, as I avoid raw aluminum. The anodization is great for that. You can also anodize items in an electrolytic bath. Similar to the e-tank I clean cast iron in.

 

[PICOGRAV]

The surface on the COB is not polished, I would not polish the heatsink. Use a good thermal compound and sand to about 320 grit.

 

[bicit]

Totally depends on how thick the layer is. It can be very beneficial for passive HS and it can also have negative effect on actively cooled HS.

How could it negatively effect the performance of an activity cooled sink? Every thing I've read indicates minimal, but positive gains?

 

[bicit]

The surface on the COB is not polished, I would not polish the heatsink. Use a good thermal compound and sand to about 320 grit.

Cree cobs are actually flat. No laser etching like the vero line.

But @SupraSPL did some testing and found thatpolishing the heatsink and the quality of the thermal paste have very little impact on the cooling.

Arctic silver works great.

 

[alesh]

How could it negatively effect the performance of an activity cooled sink? Every thing I've read indicates minimal, but positive gains?

As you know, when airflow is forced to a HS, convection becomes the main mean of heat transfer to a point that radiation is almost negligible (2-7%@180LFM). Aluminum oxides have smaller thermal conductivity therefore it is only matter of airflow and how thick the anodized layer is before it starts restricting the flow of heat. How thick does the layer have to be at which airflow? That I don't know.
I've read this in an application note of a HS manufacturer but can't find it now. Hopefully I'll be able to do so to back up my words but I feel that it's pretty logical explanation.

 

[gk skunky]

Why not just mask off the mounting area before andozing? Then don't have to sand it off. I know it can be done my buddy runs a custom bike shop and they do all kinds of multi color andozing on wheels and stuff.

 

[bicit]

As you know, when airflow is forced to a HS, convection becomes the main mean of heat transfer to a point that radiation is almost negligible (2-7%@180LFM). Aluminum oxides have smaller thermal conductivity therefore it is only matter of airflow and how thick the anodized layer is before it starts restricting the flow of heat. How thick does the layer have to be at which airflow? That I don't know.
I've read this in an application note of a HS manufacturer but can't find it now. Hopefully I'll be able to do so to back up my words but I feel that it's pretty logical explanation.

Just postulating a bit, i imagine the heat load would have to be rather large and the oxidized layer pretty think before we would notice any real negative effects.

Could be wrong

 

[bicit]

Why not just mask off the mounting area before andozing? Then don't have to sand it off. I know it can be done my buddy runs a custom bike shop and they do all kinds of multi color andozing on wheels and stuff.

You could do that as well. Depends on who's doing the work.

Honestly though, anodizing aluminum is pretty straight forward. I plan on dabbling with DIY anodizing to save time/money since the setup fee's to do a single part are kind of expensive.

 

[robincnn]

Why not just mask off the mounting area before andozing?

Sounds good but acid is might cut through the mask if mask is not tough enough.

You could do that as well. Depends on who's doing the work.
Honestly though, anodizing aluminum is pretty straight forward. I plan on dabbling with DIY anodizing to save time/money since the setup fee's to do a single part are kind of expensive.

Are you doing this DIY similar to this method
http://www.thefintels.com/aer/homealuminumanodizing.htm

Not sure if the black dye after anodizing would form another thin insulating layer on heatsink.

 

[gk skunky]

Sounds good but acid is might cut through the mask if mask is not tough enough.


Are you doing this DIY similar to this method
http://www.thefintels.com/aer/homealuminumanodizing.htm

Not sure if the black dye after anodizing would form another thin insulating layer on heatsink.

There are many different masking options safe for andozing. I used polyester tape on the rear sets I did at his shop. And yeah I thought the same thing about the insulating possibility but sounds like it's actually the opposite and better. Idk haven't really got ti research it myself yet. @PICOGRAV may be able to touch on that more for us.

 

[bicit]

Sounds good but acid is might cut through the mask if mask is not tough enough.


Are you doing this DIY similar to this method
http://www.thefintels.com/aer/homealuminumanodizing.htm

Not sure if the black dye after anodizing would form another thin insulating layer on heatsink.

Probably something along those lines. Still researching it.

The color of the dye shouldn't matter really. Black would technically be the best if outright performance is the goal.

 

[PICOGRAV]

Blacks really the only pigment small enough to fit in the hard anodizing.

I used to do it for fun a while back. I wouldn't think your buddy would want to add contamination to his tank, something the tape or coating will leave behind.

Are you unable to source some 300-400 grit sandpaper?

Most engineers would design the cooling to just mount over the coating but we are going a little further in our scenario. Sometimes we need to think a little more practical and get things done

 

[gk skunky]

Blacks really the only pigment small enough to fit in the hard anodizing.

I used to do it for fun a while back. I wouldn't think your buddy would want to add contamination to his tank, something the tape or coating will leave behind.

Are you unable to source some 300-400 ount over the coating but we are going a little further in our sceario. Sometimes we need to think a little more practical and get things done

They do it all the time. No issues I've heard from him. Though Not like they use the same tank of dye over and over. They'd have to have hundreds of dip tanks. They have tons of shades and people building some these bikes want some specific stuff. The polyester tape is safe for both anodizing and powder coating. High temp and chemically resistive. Here. He uses the 3M stuff though.

http://www.viadon.com/Resources/Masking/Anodizing.aspx

 

[PICOGRAV]

Very interesting and if you have the time to spend on flash parts, go for it!

I've always been a little fan of the spray gun...