How to calculate cooling properties via mass?

Flashlight Modding and DIY Parts
1

Friend of mine said he "may" be interested in milling a custom copper heat sink for me to bolt on my X80 just for testing purposes. The idea is to surround the head of the x80 with this finned copper chunk that would make full contact with the fin grooves of the x80 and transmit the heat to the new surrounding copper chunk. Would be almost round but still allowing for the original button and vent and should be an easy clamp on. The X80 step down is temp controlled so adding more heat sink may work.

My questions are:

1: Would this just be a waist of time?

2: Is there some sort of calculation that could give us an idea how much heat would be dissipated by the weight of the new copper so that the size of it could be calculated before milling to give an idea how run times would increase?

3. Would a good chunk of copper substantially increase run times on turbo and turbo max, or would the increase be minimal?

I have two x80's and don't mind one of them being used for testing. The whole idea of this is to have fun and see what can be achieved and to possibly make the addition aesthetically pleasing as well. It's not a serious project, just something to play with in hoping for good results. We also discussed having a temp readout in the copper head but not sure how to go about that yet.

2

The calculations aren’t easy.
There is conduction through the stock fins, resistance from the anodizing, resistance from the contact area, conduction through the new fins and convection from the new fins.
If you’ve taken heat transfer and thermodynamics in university you should be able to solve the problem.
Otherwise another option is to do a simulation in solidworks, but that would require modeling the entire flashlight and additional heatsink.
So really there is no easy way to find out how much of an impact adding that extra heatsink would be.

3

Thanks Enderman, one of the first things I was wondering about was how much resistance anodizing would have on the affect of total heat transfer. Lots of things to think about. Would be disappointing having a really nice heat sink only to find out it affected run times by only 10 to 20%.

4

Interesting link
I have calculated such things long time ago during my study, Unfortunately that knowledge is not up to date anymore :wink:

https://www.google.nl/url?sa=t&source=web&rct=j&url=https://www.digikey.com/Site/Global/Layouts/DownloadPdf.ashx%3FpdfUrl%3DF51974C9A6D544F1A7D8F119514B67FF&ved=2ahUKEwiJ7M3p897eAhXMsqQKHdcuBWM4ChAWMAd6BAgGEAE&usg=AOvVaw0BR0YtakNDKisWXX1HtYu5

5

Wow good link thanks Yokiamy

6

It honestly probably won’t be a lot because the stock fins would restrict the heat transferred to the outer heatsink by quite a bit.
A significant improvement could be had if you remove the stock fins and anodizing by lathing the head to leave a smooth aluminum cylindrical contact surface and then attached the copper heatsink to that, but at that point you might as well machine a whole new head out of copper.

7

1. We need to take into account thermal mass first. Let’s be mean and say the Acebeam X80 consumes about 300W of power, so 300J/s. Copper has a thermal mass of 385J/kg*k, which equates to 385W/kg*C.

If you make a 6cm diameter x 1cm thickness copper disc, you are going to get a 250g chunk of copper.

If you make it 4cm thick instead, that will give you a 1kg chunk of copper.

With an ambient temp of 25C, a max head temp of 65C, 250g copper will heat up to 65C in about 15 seconds, giving it 13 seconds of runtime just based on mass.

With 1 kg of copper, you would be getting about 51 seconds of additional turbo runtime.

Just based on mass alone, you’d be getting 51 seconds of absolute minimum additional max turbo.

2. With that 0,001884m2 of copper, a heat transfer coefficient of 13,1W/m2*k in passive cooling, you’d be getting an additional 1 second to that runtime, so 52 seconds.

So, at minimum, with no finning, you would be getting 52 seconds of turbo runtime of top of what the light is capable of. That is without taking into consideration time of course. The energy would be dissipated into the air, so you would likely get more in practice.

There’s a good reason why CPU cooler employ active cooling, along with lots of finning.

You’d be limited in many other factors though, like the TIM, the finning of the light, etc.

I’m just doing some math here :slight_smile:

I may be wrong in some places though.

8

@Enderman, it does make more sense to mill a completely new head out of copper then. Too many variables the other way around.

@BlueSwordM, lets say one really gets about a minute more on turbo max, that still doesn't sound very exciting really because it's still only 2 minutes instead of one minute for all that effort. BTW, I used an online super computer to decipher your math lol. There needs to be some type of major break through in physics to get around the cooling problem as all the effort still doesn't result in reasonable run times. On the X80, running plumbing around the head to keep it cooler defeats the purpose of the compact design. Not as simple as I thought overall and glad I started this thread before diving deep into it all.

9

Yep. That’s the problem with passive cooling inside of a flashlight.

BTW, 1 min of extra runtime is the minimum, without finning. With finning, I’m willing to bet that a copper head would help with passive cooling by a lot.

10

Ah, you got me interested again. I will see if I can bribe my friend to build me a copper head then. I built a deck for him last year to help him out so he's somewhat motivated to do this project lol. I still want to find a way somehow to put a nice little temp gauge in the head as well. He may be a little stumped with the vent valve as well that the X80 has in the head.

11

LoL sorry LMFAO! :-D

12

haha, trouble with that light is you need a truck to carry it I don't want to end up there with the x80 just to get a few extra minutes of run time.

13

DBCustom said Richard’s modded TR-J20 can sustain about 39k lumens until the batteries drain. The stock TR-J20 was measured by someone here at 1,479g. The light also has insane amount of finning for maximum surface area to radiate heat.

14

Thanks for that link! After reading it, the next problem comes along..... great heat sinking, lots of fins and running on max rapidly depletes the cells.

15

:-D