Author Topic: general heat sink rules for fin dimensions  (Read 4936 times)

0 Members and 1 Guest are viewing this topic.

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
general heat sink rules for fin dimensions
« on: July 12, 2019, 05:49:26 am »
Is there a general rule for manufacturing heat sinks that can be followed, or restrictions?

The idea is to get copper or aluminum block and use the smallest endmill possible as long as possible which my little mill can handle in order to manufacture a heat sink.

From what I have seen in CPU coolers, you can basically go to playing card thickness on fins ?

The smallest endmill I have is 0.8mm. Assuming it does not break, and I have the will to sit there and make this very slowly, is there any reason to go to a higher mill size?

Also, is it optimal to have the same dimension for the spacing between the pin as the pin itself ?

Main concern is that I feel the endmill may break. I want to know if its the best to use 0.8mm or if I am actually hampering myself and taking additional risk (Which I will take even for small benefit, but IS there a benefit?)

« Last Edit: July 12, 2019, 05:51:15 am by coppercone2 »
 

Offline Black Phoenix

  • Super Contributor
  • ***
  • Posts: 1137
  • Country: hk
Re: general heat sink rules for fin dimensions
« Reply #1 on: July 12, 2019, 05:56:08 am »
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #2 on: July 12, 2019, 06:19:21 am »
Ok, do you have thermal simulations for the same heat sink with different spacings?

If I am doing forced air, I am thinking air condition manufacturers some how tuned this over the years. They use something like 1mm spacing, so all I would need to do is determine the flow required through those plates?

I wonder if they optimized it?

The document is a bit much, I feel like I am getting another degree to manufacture a single component.

though , this is good

« Last Edit: July 12, 2019, 06:22:19 am by coppercone2 »
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #3 on: July 12, 2019, 06:22:48 am »
why don't room air conditioners follow those guidelines though?

they use like 0.007 thick 0.07 spacing with heavy forced air.
https://www.powerelectronics.com/content/optimize-fin-spacing-how-close-too-close
« Last Edit: July 12, 2019, 06:32:38 am by coppercone2 »
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #4 on: July 12, 2019, 06:43:55 am »
https://myheatsinks.com/calculate/plate-fin-heat-sink-calculator/

maybe I can get a feel for it by doing 10 common heat sink sizes with a bunch of iterations in each and making a flip book with graphs
« Last Edit: July 12, 2019, 06:49:58 am by coppercone2 »
 

Offline Black Phoenix

  • Super Contributor
  • ***
  • Posts: 1137
  • Country: hk
Re: general heat sink rules for fin dimensions
« Reply #5 on: July 12, 2019, 06:55:38 am »
why don't room air conditioners follow those guidelines though?

they use like 0.007 thick 0.07 spacing with heavy forced air.
https://www.powerelectronics.com/content/optimize-fin-spacing-how-close-too-close

Room Air conditioners have a lot more surface to cool air in a small space. You want to cool a surface in direct contact with the heat source. The same applies to CPU coolers, specially the tower type, were you want the maximum cooling capability in the smallest space possible. Servers coolers use the same, but since the space between the top of the rack enclosure and the motherboard are small, they rely small dense fin coolers with fans able to move air faster.

Rule of thumb is the best thermal transfer metal (Cooper or Aluminium) with a good thermal interface and forced cooling via a small fan, directly applied to the cooler or an airflow path.
« Last Edit: July 12, 2019, 06:57:12 am by Black Phoenix »
 

Offline MosherIV

  • Super Contributor
  • ***
  • Posts: 1530
  • Country: gb
Re: general heat sink rules for fin dimensions
« Reply #6 on: July 12, 2019, 07:06:28 am »
Heatsinks (terrible word to desribe what they do) is about a material that transfers heat as fast/effieciently as possible
AND
Use as much surface area as possible to radiate the heat into the surroundings.

Commercial heatsinks are that shape because it is economical to make that shape in volume. Ie churn them out as fast as possible.

If you are making your own heatsink, you can investigate more complex shapes that take more time to machine.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #7 on: July 12, 2019, 07:06:46 am »
yes, it seems that in the calculator you can get better performance values at high flows at the sacrifice of low flow heat (so it requires cleaner conditions and is more sensitive to fan failures)
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #8 on: July 12, 2019, 07:07:48 am »
Heatsinks (terrible word to desribe what they do) is about a material that transfers heat as fast/effieciently as possible
AND
Use as much surface area as possible to radiate the heat into the surroundings.

Commercial heatsinks are that shape because it is economical to make that shape in volume. Ie churn them out as fast as possible.

If you are making your own heatsink, you can investigate more complex shapes that take more time to machine.

what complex shape?

I thought at most you can put 'scratches' or grooves into them, which would mean you need to make each fin individually and braze it (an option). Or I guess make weird machine tools that cut offset to the center axis (maybe a small ball tool)

can you give examples of stuff thats not mass produced  but good?

I have zero interest in mass production with this
 

Offline Black Phoenix

  • Super Contributor
  • ***
  • Posts: 1137
  • Country: hk
Re: general heat sink rules for fin dimensions
« Reply #9 on: July 12, 2019, 07:12:41 am »
One complex shape is the use of heatpipes to transfer the heat away from the source to a far away point, I've seen totally passive PC cases that are nothing more that 2 giant blocks of Aluminium to soak the heat and they are connected with various heatpipes to the base that makes contact with the heat source.



Basically using only natural convection.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #10 on: July 12, 2019, 07:58:42 am »
those are too expensive for me, I want to get entertainment from the mill

you can't make them without weird equipment unfortunately and the development kits are expensive

as much as I want to, I can't afford this
https://www.ebay.com/i/183203665283?

their like 10$ each if you buy singles

https://www.digikey.com/products/en/fans-thermal-management/thermal-heat-pipes-vapor-chambers/977?FV=1880003%2Cffe003d1&quantity=0&ColumnSort=1000011&page=1&k=heat+pipe&pageSize=500&pkeyword=heat+pipe

less for smaller ones. maybe
« Last Edit: July 12, 2019, 08:02:52 am by coppercone2 »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: general heat sink rules for fin dimensions
« Reply #11 on: July 12, 2019, 08:51:35 am »
The thickness and length of a fin can be derived from the heat spreading length of the material.

Thin fin radiators have frequent pipes, while thick fins are typical for longer heatsinks with heavy solid backplates.  Thicker fins or more frequent pipes are required for forced air.

Fin spacing similarly is driven by pressure drop.  The pressure drop of convection is very small so the fins need to be widely spaced.  There is much more time for thermal diffusion to occur, carrying heat into the thickness of the airflow, so this isn't a big deal.  Forced air can afford more pressure drop, and so shorter, denser fins are acceptable.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: general heat sink rules for fin dimensions
« Reply #12 on: July 12, 2019, 04:19:39 pm »
Coppercone2, what type of milling machine are you using?  For small end mills on the order of 1 mm you need a machine with tight ways and a fast spindle and have to watch the feed and speed carefully.  If fed too fast or if the end mill loads up with chips it'll break like glass.  Depending on the shape and spacing of the fins, it's probably better to use a slitting saw instead of an end mill.
 

Offline MosherIV

  • Super Contributor
  • ***
  • Posts: 1530
  • Country: gb
Re: general heat sink rules for fin dimensions
« Reply #13 on: July 12, 2019, 06:53:08 pm »
Most commercial heatsinks use extruded shapes. Meanung all straight lines and shapes.

How about machining concentric rings into a block.
I am no mathmetician or mechanical engineer but I would have thought you can get more surface area with circles than straight lines.

Not ecomonical to machine circles unless there is good reasons or necessity.

More thoughts rather than absolute facts.
 

Offline ejeffrey

  • Super Contributor
  • ***
  • Posts: 4033
  • Country: us
Re: general heat sink rules for fin dimensions
« Reply #14 on: July 12, 2019, 07:19:25 pm »
Straight lines are better for airflow than concentric rings.  And almost any heatsink whether forced air or natural convection ultimately relies on airflow to remove heat.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #15 on: July 12, 2019, 08:50:14 pm »
Coppercone2, what type of milling machine are you using?  For small end mills on the order of 1 mm you need a machine with tight ways and a fast spindle and have to watch the feed and speed carefully.  If fed too fast or if the end mill loads up with chips it'll break like glass.  Depending on the shape and spacing of the fins, it's probably better to use a slitting saw instead of an end mill.

just a little proxxon.

tried milling a steel wrench, some coins and some plastic with it, seems to work OK
 

Offline DBecker

  • Frequent Contributor
  • **
  • Posts: 326
  • Country: us
Re: general heat sink rules for fin dimensions
« Reply #16 on: July 12, 2019, 10:00:49 pm »
Almost all heatsink designs are driven by manufacturing cost.  They aren't picking optimal shapes, they are figuring out how to efficiently make good-enough shapes.

The optimal size and shape of fins/pins will be heavily influenced by the airflow and turbulence.  You can over-optimize to where a tuned design is not going to work in all conditions.  Long ago I had co-workers who had been on a exciting-but-uncomfortable airplane ride to diagnose their equipment failing at high altitude.  A highly tuned cooling design at sea level pretty much failed when the Reynolds number changed.

My experience with a high-power project taught me a few lessons. While an aluminum block with spring clips is very appealing, using even tiny copper heat spreaders under the devices dramatically improves the performance.  And that a round bored hole transfers surprisingly little heat to fluid flowing through, no matter what the flow rate.

BTW, are you planning to use an end mill on bulk copper?  That will be... challenging.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #17 on: July 12, 2019, 10:12:50 pm »
i was thinking tellurium copper
 

Offline ahbushnell

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: us
Re: general heat sink rules for fin dimensions
« Reply #18 on: July 13, 2019, 12:10:06 am »
The problem with milling fins on heatsinks is the fins start to vibrate when they get long and skinny. 
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #19 on: July 13, 2019, 12:10:39 am »
I thought about that and I thought I might be able to fill them with wax or something in between cuts

maybe hot glue, not sure what kind of strength you need
 

Offline ahbushnell

  • Frequent Contributor
  • **
  • Posts: 751
  • Country: us
Re: general heat sink rules for fin dimensions
« Reply #20 on: July 13, 2019, 02:25:51 am »
I thought about that and I thought I might be able to fill them with wax or something in between cuts

maybe hot glue, not sure what kind of strength you need

I know for custom heatsinks there are vendors that use ganged saw blades to cut them all at once.

3D printed metal would work.  But that would be expensive. 

3D print plastic and make mold and then cast the aluminum.  Good home project.  :)

Just make the fins a little thicker and machine them.  Do it slowly. 

Good luck

 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #21 on: July 13, 2019, 02:28:22 am »
hehe vaccum silver casting would do the trick

I think it offers very good resolution.

a big to-220 heatsink is 51gm, so you get something like 19cm3,

19cm3 of silver is 200gm, 106$

Now, how do you get the black anodized coating on silver (obviously a problem) to increase its performance like with Aluminum?


Given the price of a LT3080 is 6$ in single unit quantities, the heat sink price is slightly less absurd. Even more worth it for a LT3083 high grade part (11$+) :-DD


add the absurd cost of a stainless screw and thermal compound and it makes more sense

could aluminum plating the silver and anodizing the aluminum and dyeing it do the trick?
« Last Edit: July 13, 2019, 02:41:33 am by coppercone2 »
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5569
  • Country: us
Re: general heat sink rules for fin dimensions
« Reply #22 on: July 13, 2019, 03:41:51 am »
The common thread running through many of the answers here is that there is no single rule of thumb or design equation.  Optimum depends on the particular problem being solved.

For you, it seems that the problem be solved is how to put your mill to use.  So the answer may revolve more around what can it do, what is good enough for cooling purposes, and perhaps what clever mounting technique can you mill in that lets you optimize around your willingness to invest design and machine time in a unique design.

That line of thinking lets you ignore ganged saws which is really the only good way to cut fins, even if cutting is the right answer.

If getting a usable heat sink is the priority your best bet is probably to find or buy some extruded finned aluminum and then mill as appropriate for mounting purposes.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: general heat sink rules for fin dimensions
« Reply #23 on: July 14, 2019, 06:44:10 pm »
Silver naturally combines with sulfur to make silver sulfide that is quite black, and commonly called a tarnish.  You'll just have to accelerate the process.  Black and white film and papers use various finely divided silver halides to produce quite black coatings.  I don't know how hard these coatings are, but silver tarnish is hard to remove.  I do know that aluminum oxide (anodized finish) is quite hard and that's why it's used.

I've got a few copper server processor heat sinks that were scrapped because of machining faults eg., broken drill bits or taps embedded in them.  Pure copper is hell to machine.  BTW, I had never heard of tellurium copper.  Thanks for the tip - it seems like it's much more machinable than ordinary copper.

In addition to the heat sinks, I have some spot welder tongs that I'd like to make black to improve heat dissipation.  Copper tarnishes all by its lonesome but I've never tried to do any sort of coating myself.  I understand there's a process using "Liver of sulphur" (Anything like eye of newt or horn of toad?) that is a conglomeration of potassium-sufur compounds.
 

Online coppercone2Topic starter

  • Super Contributor
  • ***
  • Posts: 11336
  • Country: us
  • $
Re: general heat sink rules for fin dimensions
« Reply #24 on: July 14, 2019, 07:25:40 pm »
careful its still dangerous, at least on a lathe, just less so (in terms of getting machine tools stuck in the ceiling or your head). Much nicer though (hence the ultra machinable), but not as nice as normal metals

I wonder if the tarnish improves the thermal conductivity as does black dye anodizing (5% or so)

I don't think the oxide needs to be black, more like you need a ultra thin film to lock in a black dye (but I guess a naturally black oxide would be better? I don't really understand the dynamics.

I kind of wonder if undistributed, would the ultra-black soot of an air acetylene flame out perform all these things. Maybe vacuum deposit the thinnest layer of super glue on it afterwards to protect it a bit?
« Last Edit: July 14, 2019, 07:29:09 pm by coppercone2 »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf