Thin fins wins- Mtbr.com
Results 1 to 28 of 28

Thread: Thin fins wins

  1. #1

    Reputation:
    Join Date
    Jun 2009
    Posts
    82

    Thin fins wins

    How thin to make the heat-dissipating fins on our housings?

    To help answer that question, here are two housings, exactly identical except for the thickness of the fins. Here's the original "flatpack" housing, fins approximately 2.4mm thick.




    Back of envelope calculations told me that I should be able to make the fins much thinner without making a hoot of difference to the thermal performance. I hate carrying around extra weight, so re-machined the fins to half of original thickness, now 1.2mm thick.



    Here is the thermal performance of both housings, and they are identical within measurment error.


    So the calculations were right and, in fact, I could have made them half as thick again (0.6mm) and still not degraded the thermal performance. I'll post the basic calcs if anyone is interested, but the broad conclusion is that for our physically smallish housings, mechanical issues tend to limit the minimum thickness, rather than issues of being too thin to effectively conduct heat along the length of the fin. For example, if you make the fins much thinner than about 1mm, then they could easily get bent, and be likely to cut you.

    In my case I saved about 20gms by halving the fin thickness from 2.4 to 1.2mm, and if anything would have been easier to make, ripping the metal out from between the fins with a 4mm slot drill rather than a 3mmm one. I'll happily take the saved weight, but another approach would have been to keep the weight the same, and make the fins twice as long and half as thick, or twice as many at half the thickness, or whatever. However you look at it though, my original "thick" 2.4mm thick fins were just plain bad design, a bad use of metal. Bare housing is now around 65gms.

  2. #2
    mtbr member
    Reputation: yetibetty's Avatar
    Join Date
    Dec 2007
    Posts
    1,781
    I saved 20 grams too when I did mne.

    I like the new look of yours.

  3. #3
    Bandolero
    Reputation: notaknob's Avatar
    Join Date
    Apr 2004
    Posts
    2,437

    Nolla

    Quote Originally Posted by cdcdcd
    Back of envelope calculations told me that I should be able to make the fins much thinner without making a hoot of difference to the thermal performance.
    Why do you hate fat fins?



    Haven't they suffered enough?


    Love them, don't hate them.
    Slow-core. -.. .-. .. -. -.- .... --- -- . -... .-. . .--

  4. #4
    One Gear
    Reputation: .40AET's Avatar
    Join Date
    Jun 2007
    Posts
    1,077
    So could you use a smaller mill/bit and cut more fins into the same area and increase the cooling???

  5. #5
    mtbr member
    Reputation: zen bicycle's Avatar
    Join Date
    Mar 2007
    Posts
    463
    Fins only really help when you have air movement. It would be interesting to see that same test with air movement.

  6. #6

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    Fins only really help when you have air movement. It would be interesting to see that same test with air movement.
    No, not correct. Commercially available heatsinks for electronic devices are finned for good reason, yet have no forced convection. What actually happens is that the heated air rises, which creates movement of air past the fins.

    Fins are, of course, even more effective with forced air movement. This light, like almost all the lights we build, runs ony slightly warm when the bike is being ridden, so it is doubtful whether it is worth measuring the temp rise when moving as there just isn't a problem. However, I am completely certain that the cooling performance with air movement would also be unaffected by making thinner fins.

  7. #7

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    Here is another experimental result that may be of interest.

    The measurements were taken with the light horizontal, as it would be when mounted on a bike. With this particular housing design, that means the fins were also horizontal.

    I repeated the measurements with the light shining vertically upwards, so that all the fins were vertical, and found the temperature rise was about 15% lower. It is well known that aligning the fins vertically aids heat dissipation is still air, as the natural convection currents occur vertically due to the upwards movement of the heated air. If the light was designed for stationary use, it would therefore be worthwhile machining the fins in a vertical direction. With the fins horizontal as in this design, the cooling will be better (relative to vertical fins) when the bike is moving.

  8. #8
    mtbr member
    Reputation: zen bicycle's Avatar
    Join Date
    Mar 2007
    Posts
    463
    Sorry I should be more pointed in my statement. With the size finns we are seeing here it is mathmatically correct to say the make no difference. If there was air movement across these particular fins I was saying it would be interesting to see if the thinner fins were more effective at cooling than the thicker fins. Thermal loading and all that stuff.

  9. #9

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    With the size finns we are seeing here it is mathmatically correct to say they (the fins) make no difference.
    I presume you mean in still air? Are you saying that in still air, the temperature rise of my "flatpack" light (for example) would be the same, with or without the fins!!?? That seems extremely unlikely, given that the inclusion of the fins more than doubles the surface area, and why do the makers of heatsinks bother to put fins on them? Sorry, I'm not sure if I understand what you are saying here.


    If there was air movement across these particular fins I was saying it would be interesting to see if the thinner fins were more effective at cooling than the thicker fins. Thermal loading and all that stuff.
    I am not claiming that making the fins thinner improves the cooling per se. What I am saying is that you can save weight by making the fins thin, while maintaining essentially the same thermal performance compared to thick fins. Alternatively, you can keep the weight the same, and have more or longer thin fins, resulting in superior thermal performance for the same weight. To directly answer you question, with the bike moving, the temperature rise of my housing would be the same within measurement error with the original thick fins, or the re-machined thin fins. And it would run considerable hotter if the fins were removed altogether.

  10. #10
    I spelled Knievel wrong
    Reputation: StevelKnivel's Avatar
    Join Date
    Jun 2007
    Posts
    217
    You are correct cd^3. The fins only needs to be thick enough to efficiently spread the heat over the available surface area. Because your fins are pretty short height-wise, as long as heat gets spread along the base of the fin fairly well they do not need to be very thick, airflow or no airflow. It's only when fin height gets taller that thickness needs to increase for good heat conduction to the fin tip.

  11. #11
    mtbr member
    Reputation:
    Join Date
    Feb 2008
    Posts
    47
    I'm lost. I don't see how just the thickness of fins has any effect on cooling worth worrying about for bike lights? Aren't the fins just too short in most cases for this to matter that much.

    Commercial CPU heatsinks have paper thin fins so they can cram more of them in a smaller space and therefore increase surface area considerably as they use really tall fins.

    For our use in lights the limits are more to do with practicalityand structural concerns aren't they?

  12. #12
    mtbr member
    Reputation: scuppy's Avatar
    Join Date
    Nov 2007
    Posts
    94
    But the thinner fins means there is a bigger air gap between them (in the pics above). Surely that is a factor.

  13. #13
    mtbr member
    Reputation: znomit's Avatar
    Join Date
    Dec 2007
    Posts
    1,980


    Surely the base thickness is more important than the fin because it transports the heat to the base of every fin and to the front of the light which is where the wind is.

    Errr, yeah, no legit reason for quoting that picture.

  14. #14
    mtbr member
    Reputation: bumphumper's Avatar
    Join Date
    Jan 2009
    Posts
    136
    Hmmmmmmmmmmmmmm

  15. #15

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    We all seem to agree. For the shortish fins we are likely to use with our housings, we can save weight by making the fins as thin as possible, subject to structural and aesthetic considerations. Just cos I like numbers, I might post calculations showing just how thin the fins can be, as well as showing the effect of casing thickness which, as others have pointed out, is responsible for spreading the heat along the base of the fins.

    I would also like to thank notaknob for his excellent picture extolling the virtues of slim fins - pictures like this should be compulsory in all engineering texts. This is fine engineering at its best. Note there is no excess weight here whatsoever, no wastage, all components serve a purpose, and serve the purpose well I feel sure. A lovely example of functionality and aesthetics combined. I think I mentioned very early on that I have a tendency to rave on about good engineering ...
    Last edited by cdcdcd; 08-07-2009 at 06:03 PM.

  16. #16
    mtbr member
    Reputation: yetibetty's Avatar
    Join Date
    Dec 2007
    Posts
    1,781
    A lot depends on which direction the fins are ponting and for which situation you want to cool the light, for example when you are stopped or for a walking light where there will be little or no air flow the fins should be in a direction to allow the flow of rising cool air (convection).

    Most bike lghts don't really get hot when on the move due to airflow regardless of number or oriontation of fins so in my opinion it's more inpotant to have them orientated for those slow or stationary times and fins that point towards the wind for cooling whilst riding will have little effect whilst stationary and could infact trap air or hinder the convection cooling when stationary.

    One thing I have learnt through many experiments(and looking at ampifier heatsinks)is that pointed fins do a great job as heat will try and get to out side world air via the thinest part(the tip of the fin)but the heat has to get there some how(the thick base of the fin).
    In the end though it's all good news as less metal and less weight can't be bad.

    Cd, what is the thickness of the actual part that that your LED's are mounted to?

    EDIT:Sorry, I should have read the whole thread!!!
    Last edited by yetibetty; 08-07-2009 at 06:20 PM.

  17. #17

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    A lot depends on which direction the fins are pointing and for which situation you want to cool the light, for example when you are stopped or for a walking light where there will be little or no air flow the fins should be in a direction to allow the flow of rising cool air (convection).
    Yes. For example with my particular light the measured difference between having the fins horizontal or vertical is 15%, not insignificant.


    Cd, what is the thickness of the actual part that that your LED's are mounted to?
    I cal lthis part the "firewall", and it is 5mm thick. The firewall needs to be thicker than any other part of the casing, especially just where the led(s) are mounted. Even so, my 5mm is overkill, but the area of the firewall is so small that there is very little weight to be saved by making it thinner.


    Most bike lghts don't really get hot when on the move due to airflow regardless of number or oriontation of fins so in my opinion it's more inpotant to have them orientated for those slow or stationary times and fins that point towards the wind for cooling whilst riding will have little effect whilst stationary and could infact trap air or hinder the convection cooling when stationary.
    So what are saying here, is that my fins are pointing the wrong way!! Arguably so, but I'm saving the best for last, for my housing was also designed for optimum heat transfer to the bars, and the advantage I may gain here could make arguments of fin direction insignificant. We'll see.

  18. #18
    mtbr member
    Reputation: yetibetty's Avatar
    Join Date
    Dec 2007
    Posts
    1,781
    No, not saying the wrong way, depends on the intended use of the light.

    It's about time that "part" had a name, so firewall it is and as long as it is the thickest part of the housing that's all that matters as heat will then follow a path through progressively thinner parts for an easy route out.

    I made a new housing today and have settled on 1.5mm fins with 1.5mm between them & 5mm tall yet managed to fit more of them in the same space...... well one more

    As you say strengh and lack of sharpness are also inportant as is the fact that they were easier to make.
    Last edited by yetibetty; 08-08-2009 at 11:42 AM.

  19. #19
    mtbr member
    Reputation: troutie-mtb's Avatar
    Join Date
    Sep 2007
    Posts
    3,062
    I have been banging on about using the bars for the heatsink for 2 years now and it is the best way ever
    what about no fins and just bars



    this concept will keep cool the daftest lights and driver size is the bigest bugbear

    then making the mounts from ali and direct contact to the bars is the next best way
    as the insane quad MCE


    this light will overheat in no time when not on the bars but maintains a nice cool temp for ever even when the bike is parked in the garage
    doesnt even make a hand warmer possible.


    Then there was the one Yetty should remember well as he helped in the concept



    it was made whilst waiting for the MCE to arrive and in its life got butchered to take all sorts of leds and worked as you can imagine pretty well with the driver inside the stem and the switch in the steerer tube cap like Znomits

    OK no good for on the helmet

  20. #20
    mtbr member
    Reputation: znomit's Avatar
    Join Date
    Dec 2007
    Posts
    1,980
    Quote Originally Posted by cdcdcd
    So what are saying here, is that my fins are pointing the wrong way!!
    I hear this a lot, though I have yet to see a trail where I need 1000lm when riding at 5kph.

    One more Fin...


  21. #21

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    We all agree that for our typical housings the fins can be "quite thin" before causing any significant degradation of thermal performance.

    Just for fun, I ask the question "how thin"? Back-of-envelope calculation follows for my "flatpack" dual MC-E housing, which is roughly typical.

    There are 32 fins, each roughly 40mm long by 5mm high by 1.2mm thick. Let's say 1mm thick for simplicity.

    The electrical power input for 2 MC-Es is 2.8A x 3.2V x 2 = 18W (approx)

    In rough terms, about half of that 18W will enter the base of the fins, the other half will be ejected out the front as light and heat, and dissipated from the surface area of the rear,and the area of the casing between the fins. So, very approximately we have 9W flowing into the base of the fins. As there are 32fins, the thermal loading per fin is therefore 9/32 = 0.28W per fin.

    However, each fin dissipates this power along it's length, so the thermal loading is 0.28W at the base, and zero at the tip. The average thermal loading per fin is therefore about 0.14W/fin.

    Now to calculate what we really want to know, being the temperature difference between the base and tip of the fin. Approximate formula is :

    dT=(PL)/(CA) where

    dT is the temperature difference in DegC
    P is the average power flowing along the fin (W)
    L is the length of the fin (m)
    A is the cross sectional area of the fin (m^2)
    C is the thermal conductivity of the aluminium fin material (SI units) ~180

    dT = (0.14x0.005)/(180x0.04*0.001)

    dT ~ 0.10 DegC

    However, that is the temperature difference all the way from the base to the tip. The average temperature difference between the housing and the fins is only half of that, a completely negligible 0.05 DegC.

    Hmmm. So what does that actually mean? Well, if the 1mm thick fins were made 20 times thinner, then the average temperature difference between casing and fins would be 20x0.05 = 2.0 DegC, which is still negligible compared to the other temperature differentials in the thermal path, such as the 28 DegC from semiconductor junction to led solderpoint, or the 40 DegC or so from casing to ambient.

    Heck, 20 times less than 1mm is 0.05mm!!! That's an amazing result IMHO. For my "typical" housing with 2 MCE leds and 32fins, the fins could be thinner than a sheet of paper at only 0.05mm (0.002"), and still the degradation in thermal performance would be negligible.

    So, at least from the point of view of heat flow in the "radially outward" direction from the base of the fin to the tip, the fins can be as thin as you like, even down to an absurd 0.05mm, without making a hoot of difference to thermal performance. Interesting, eh.

    A similar calculation in the axial (=front-back) direction would be interesting to see if fin thickness is necessary to transport heat along the base of the fin, in the axial direction. Or, to look at this another way, calculate how thick the housing needs to be to transport heat in the axial direction.

    When I get a free moment I'll do it.
    Last edited by cdcdcd; 08-12-2009 at 03:54 PM.

  22. #22
    mtbr member
    Reputation: piesoup's Avatar
    Join Date
    Feb 2009
    Posts
    455
    Man you have a lot of time on your hands!! Interesting as I am building a 55w light and need all the heaksinking I can get!

  23. #23
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    453
    cdcdcd, don't get TOO thin! A 0.05 mm fin will not last long in MTB environment. Or maybe you could start using razor blades

    There is another option, to alternate thick and tall and thin and shorter fins so that the thick and tall protect the thin and short.

    Like this:
    OooOooO

    l

  24. #24

    Reputation:
    Join Date
    Jun 2009
    Posts
    82
    cdcdcd, don't get TOO thin! A 0.05 mm fin will not last long in MTB environment. Or maybe you could start using razor blades

    You may misunderstand me. There is no benefit in making fins that thin - obviously a 0.1mm thick fin is mechanically ridiculous, and once the fins get down to around 0.5mm thick they weigh half of nothing anyway, so there is absolutely no reason to make the fins ridiculously thin.

    Mainly, I just thought the calculation was interesting. What is shows is that we can make the fins as thin (and therefore lightweight) as we like based on mechanical considerations only.

  25. #25
    mtbr member
    Reputation:
    Join Date
    Sep 2007
    Posts
    453
    Quote Originally Posted by cdcdcd
    Mainly, I just thought the calculation was interesting. What is shows is that we can make the fins as thin (and therefore lightweight) as we like based on mechanical considerations only.
    This calculation is already done in metal - check the "bonded fin" heatsinks, where the fins are made of thick foil.

  26. #26
    Bandolero
    Reputation: notaknob's Avatar
    Join Date
    Apr 2004
    Posts
    2,437

    My Name is Jack

    Sorry to threadjack again... but I came across another thin fin. It'll be the last time and I profusely apologize for attempting to derail this technical thread with such irrelevance.



    It really was Zemike's fault with his diagram of 0oo0.. that brought this on.
    Slow-core. -.. .-. .. -. -.- .... --- -- . -... .-. . .--

  27. #27
    mtbr member
    Reputation: zen bicycle's Avatar
    Join Date
    Mar 2007
    Posts
    463
    keep interupting please

  28. #28
    mtbr member
    Reputation: piesoup's Avatar
    Join Date
    Feb 2009
    Posts
    455
    Yip, feel free to barge in on the thread at any time.

Members who have read this thread: 0

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2020 VerticalScope Inc. All rights reserved.