Most Efficient Heat Dissipation System?

[Klem]

Here's a issue that has vexed me for a while...

Is the moving water that surrounds us when we dive more efficient as a heat dissipator than a big chunk of metal?

With high power LEDs we need to get rid of heat as quickly as we can. We glue metal bases to the emitters as heat dissipators, then plan a thermal path to surrounding material like the torch body and eventually to the outside water. This is especially important if we think the base is not sufficient enough as a heat sink.

Reading the 'Mineral Oil torch' thread the thermal conductive efficiency of typical metals we use (aluminium, steel, copper) are far more efficient than water. But don't these figures assume that water is static? If it is flowing over a torch does that not increase its efficiency, to perhaps even more efficient than metals?

Here's an applied scenario...Introducing water flow directly below the emitters via drilled holes, or machining fins into a Maglite head... as opposed to not drilling/machining and having that metal mass intact.

Thoughts?

 

[AnAppleSnail]

Here's an applied scenario...Introducing water flow directly below the emitters via drilled holes, or machining fins into a Maglite head... as opposed to not drilling/machining and having that metal mass intact.

Thoughts?

One way to model this is to assume that the water in your tubes is nearly the temperature of the surrounding water. That is, the thermal path from the LED junction to water is very short. Now, how valid is that assumption? It depends on how much power you're putting into the water. If water cycles quickly, it's a good assumption. If not, it won't give good data or improved heatsinking.

Generally, fins just make the light stay closer to ambient temperature. I don't think any LED light you can realistically use will require finning for water-cooling.

 

[Packhorse]

Contrary to a claim I made in another thread I do not have a degree in thermal dynamics. But I do have common sense. So take my advice with a grain of salt.

Sure water may not be as thermally conductive as aluminium but there is an awful lot of it and it is never really static.

Adding fins to a Maglite will increase the transfer of heat out of water but in water the transfer of heat is so great in water that it is pretty pointless.

An easy experiment is to heat up some aluminium ( say 1 kg) with a blow torch for 1 minute.
Now pick it up with your hand ( no not really)

Now half submerge the same piece of aluminium in a large body of water. and try and heat it up again with a blow torch. There is no way its going to get very hot at all.
Common sense tells us we need to need to extract the heat from the LED to the out side water as best as possible.

My guess is you could get a SST-90 on a star drive it at 9 amps while submerged in a large body of water and it will not blow.
Try the same thing out of water and I doubt it would last 10 seconds.

Now take a SST-90 mounted to a finned heat sink of appropriate size and let it sit in free air. The LED should last.

Now take the same LED and same heat sink and put it in a sealed plastic case and submerge it in water. The heat will have no where to go. It will heat up the air inside the container, the air is insulated from the water by the plastic. It will eventually over heat the LED.

The difference is thermal mass and thermal dissipation ( my terms)
The larger the mass the longer it takes to reach thermal over load.
The greater the surface area the more heat can be transferred to another medium be it water , air or aluminium.

 

[lucca brassi]

Actualy I have a few educational clases from thermometry on electrotehnical university .

Water in that case it is not important because in see it is almost ''unlimited'' quantity.
Important is ''conductive '' temperature path from source to sink.

there is transferr heat flow which want to equalize temperture in system from source to sink . When it equlized heat transfer stops.

source heatsink (heat transfer path) (sink)

Q/t (Q led energy)= m*cp*(Tled -T heatsink) (heatsink energy) = m*cp*(Theatsink-Twater) (water )

direct conduction is Q/t = K*S (T (hot)-T(cold)/d
http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/heatcond.html#c1


on short : points of ''heatsink energy''

- heatsink must be maximum heat conductive
(Cu , Ag , heatpipe)

- heatsink must have minimum mass ( we don't want to store heat )

- heatsink path must be short
(to long ribs of heatsink have no sense because water temperture dominant over leds so ribs don't work 100% and to big ribs give us unwanted mass)

- heatsink must have maximum surface ( for maglite lamp hollow cyliner
with bottom )

on short : points of ''heatsink energy''

- there must be enough mass (there is air problem we must have ventilation to get more mass)
(depend from Cp) to take transfered temperature - till equalization

-there must be enough surface
to take temperature from ribs of heatsink

-flow of sink(water .air )
ribs must work without ''cavitation'' (water ), turbulance (air) .because underpressure (vacuum) has low conductive factor. Ribs in most torches are verical on housing and that is OK because air flow go from lower point up (and there is no turbulance ) but if we have lamp in horizontal position and we move horizontaly fast with that torch ribs wouldn't work properly

all these points we best see at computer processor heatsinks

 

[badtziscool]

Here's a issue that has vexed me for a while...

Is the moving water that surrounds us when we dive more efficient as a heat dissipator than a big chunk of metal?

It's not more efficient as a heat dissipator but there's so much of it, it never gets heat soaked or heat saturated. In an ideal world, you should be able to take a mass of copper and use that as a heat sink, and when it gets saturated, instantaneously change out the mass of copper for another one that isn't heat saturated, and so on and so forth. Obviously that's not practical so you need another medium to carry away that stored heat/energy. That's where a fluid, like air or water, comes into play.

 

[Klem]

It looks like we are in agreeance then...moving water is a very different dissipator from still, and when diving it is always moving, and there is essentially no limit of it for heat sink, plus saturation is not an issue.

I'm thinking... If we lathe ribs into a Maglite, similar to commercial torches, this increases the surface area of the body to the surrounding water...creating a more efficient heat exchanger. If we also agree that the surrounding moving water is a fantastic heat sink, and the best place for this heat to be going, then using ribs is more beneficial than not using ribs.

An added bonus would be, this ought to lessen the severity of heat heading axially towards the electronics and batteries (if the battery compartment is not separated by water as in a can light).

Advantages of using water more efficiently in the design;
1. Cooler insides...for all the obvious reasons.
2. More amps, brighter torches.
3. The metal mass immediately behind the emitter can be smaller...for a more compact torch.

Disadvantages
1. The design may not work in air (diving only).
2. Need access to a lathe, plus the time and effort.


I'm also considering the pros and cons of separating the emitter/lens section of the torch from the electronics/battery compartment by a few milimeters of space. Once underwater this space would act in two ways; as a heat dissipator, and an insulator between the two compartments. Obviously there would be a few structural joining bolts, with one being hollow for cabling.

The gap in the sketch is a little exagerrated but you get my drift.


Or, using a ready-made 'radiator', like a thin piece of rectangular tubing. A few bolts through the middle to pull/hold the head and back sections onto the radiator, with a hollow bolt for wiring. Water flows through the middle, and the two side walls provide structural integrity.

Again, this photo has an exagerrated gap but you get my meaning. Saw some recently on EBay for hobbyists. All different sizes with gaps as narrow as 1cm and as wide as 7cm. (Checked just now...not there anymore).

 

[Packhorse]

I'm thinking... If we lathe ribs into a Maglite, similar to commercial torches, this increases the surface area of the body to the surrounding water...creating a more efficient heat exchanger. .....

I dont see the need.
The efficiency of the heat exchange from a Maglite to water is extremely high as it is.
By machining enough fins into the Mag light your maglite you may increase the surface area by 100%. This may or may not increase its heat dissipation by 100% but its probably close. But to what advantage?
I run a 35 watt LED Maglite and I can assure you that in water in never feels warmer than the ambient water temperature. I also ran W200 light head at 35 watt with the same result and that has far less surface area and mass than a maglite. My gut feeling is you could run 10 times this power and still not have an issue.

Cooling fins are great for dissipating heat in air but in water they are not needed unless we are talking huge amounts of heat that you will never see in a dive light.

But fins do have a few down sides. They can snag on lines for one, and if you machine them into a maglite it exposes un anodised aluminium for another.

 

[wquiles]

I dont see the need.
The efficiency of the heat exchange from a Maglite to water is extremely high as it is.
By machining enough fins into the Mag light your maglite you may increase the surface area by 100%. This may or may not increase its heat dissipation by 100% but its probably close. But to what advantage?
I run a 35 watt LED Maglite and I can assure you that in water in never feels warmer than the ambient water temperature. I also ran W200 light head at 35 watt with the same result and that has far less surface area and mass than a maglite. My gut feeling is you could run 10 times this power and still not have an issue.

Cooling fins are great for dissipating heat in air but in water they are not needed unless we are talking huge amounts of heat that you will never see in a dive light.

But fins do have a few down sides. They can snag on lines for one, and if you machine them into a maglite it exposes un anodised aluminium for another.

+1

I feel the same way. For a diving head I am making for a customer I (like you) felt fins were not necessary and could get snagged into things. But I wanted to provide some grip, so instead of a completely smooth surface I machined some shallow grooves.

 

[Klem]

Yes, granted you don't feel radial heat coming through the sides of an un-ribbed Maglite underwater, but that doesn't mean heat is not heading axially into the back section.

That said, If 35watts works with you, without a discernable reduction in lumens efficiency, Li-ion battery life and capacity (through overheating) then that is heartening news. That's given your torch has batteries behind the emitter and is not just an empty wand with cable heading to a separate canister. If batteries are out of the picture by design, and the emitter is cool enough to operate at max efficiency then your reply for that situation is sound.

I take your point about machining off the anodising... although washing in fresh water afterwards should cure the effects you allude to.

Catching the torch on cabling, speargun line is a fair comment...hadn't thought of that. Although after consideration, weighing up the probability of this happening, against the utility of having fins...Can't see myself rejecting the idea completely because of this.

Bear in mind I am trying to think laterally about emitter cooling, and while the fins concept is one idea, separating the sections with a radiator gap allows for designs for people don't have ready access to a lathe; so can't machine nice tight-fitting chunks of heat sink.

 

[lucca brassi]

I think so : simple enough ?

 

[Klem]

Excellent diagram Lucca.

Yes, I see what you mean. Your idea is a solid interface of metal with 2 radiator holes drilled at right angles. Enough surface area and you thermally isolate the front section from the battery section.

Your 'cross' design ensures an even spread of radiator area over the whole diameter. The center of the torch can (and does) radiate heat, probably moreso than the edges... and your solution covers that nicely.

Like 'Pack' said, it's probably overkill but I figure it doesn't hurt to consider the water surrounding our torches from every perspective. We spend a lot of effort keeping water out of our torches, and all I'm saying is consider working with it, rather than exclusively against it.

 

[AnAppleSnail]

bably moreso than the edges... and your solution covers that nicely.

Like 'Pack' said, it's probably overkill but I figure it doesn't hurt to consider the water surrounding our torches from every perspective. We spend a lot of effort keeping water out of our torches, and all I'm saying is consider working with it, rather than exclusively against it.

You can cut metal costs. You really only need metal in the thermal path between the LED and the water. For most purposes, polymers are getting pretty sturdy, they just aren't conductive yet. With this you'd have a G3-like light, with a polymer handle and metal head...

 

[Packhorse]

Keeping some warmth in the battery section can lead to extended run times since batteries dont like being too cold ( or hot).

 

[Klem]

Granted, but depending on the 'C' rate and power draw batteries often get there all by themselves (internal resistance).