Rules of thumb for cooling

[Nigel R]

Gents

Do you have any rules of thumb regarding cooling holes for our electric setups?

I'm aware of the "out hole = 4 x in hole" one, but is there anything to relate the size of the in hole to the power of the electric system?

My context is a 350W (ish) electric conversion on a small (36" biplane (Precedent Bi Fly), I've experienced several thermal cutouts after a few consecutive flights which I can only pin on the ESC getting too hot (buried in the "under tank" area with little through flow). Obviously some mod to the cooling airflow is in order but I'd like to get it about right first time!

Thanks in advance

[Denis Watkins]

When I started electric flight, we fitted the ESC outside the model, with the wires routed inside where necessary. So the ESC cooling had more priority than a pretty model.

I had a BiFly for years before it warped due to storage, but the power plant, an OS25 IC, was out in the open?

Cannot your installation be up there amongst the bearers?

[Engine Doctor]

Had a similar problem with a Nigel Hawes  Tucano I built some years ago. Motor not a problem cooling with small air inlet in front of cowling but ESC kept getting too hot as tucked away beside battery The cure was a small air scoop mounted on the fuz just in front of the ESC . It doesn't look out of place and works well. The ESC still gets warm but cooler than before and keeps going even on the hottest days.I could have changed the set up,ESC,prop,motor etc for cooler running but the Tucano is small and wanted a hot/powerful set up for speed. The small scoop is in the prop wash so really drives the air in . Air exits in a hole at rear / underside of fuz where angle of fuz causes depression in pressure that draws air through the fuz. A small lip fitted in front of exit hopes can increase air draw and improve cooling.

Edited By Engine Doctor on 10/08/2017 09:19:55

[Chris Walby]

Plastic spoons...make nice air scoops be it either tea spoon or the larger versions. this jus leave you to make a suitable exit hole.

Just cut soon at highest point while lying flat, then mark suitable point on fuselage. Make hole in fuselage/make good covering & glue scoop in place. Paint to suit

I cannot take any credit, but they are cheap, work well and don't requite a lot of modifications + look way better that a straight hole!

[KiwiKid]

Yip spoon scoops work well and can also be used to cover up outlet holes to make them less noticeable.

Commercially moulded scoops can also look fairly unobtrusive.

And NACA scoops can be cut into canopies on non-scale models.

Addressing the OPs initial post, I don't think a 4X outlet hole is really necessary and you could get into some structural problems with some models. The aim is principally to create indoor/outdoor flow and 2X should do this adequately. The efficiency of cooling will be markedly affected by the outside air temperature, which is a variable, and I usually go with the biggest inlet that can be accommodated given the 2X exit hole needed. Then it's just a case of "suck it and see".

[Nigel R]

Spoons, brilliant idea.

Denis the motor is a mega inrunner, quite long, it takes up most of the engine bay, I think realistically I'm stuck with the under-tank location unless I do quite literally dangle the ESC out in the breeze.

[Brian Spearing]

Would a switch mode ESC produce less heat?

[Olly P]

Posted by Brian Spearing on 10/08/2017 14:28:18:

Would a switch mode ESC produce less heat?

Yes.

I'd look at if you can mount the ESC so the cooling fins/heat sink sit in open air. one of my long term projects has a very high wattage requirement and I plan the ESC to sit with the fins under the fuse belly, protetected by the U/C.

[Old Geezer]

Scale's one thing, sports flying's another. 20 feet up and 30 feet out, doing more than 10 or 15 mph neither you nor anyone else will see, let alone notice, or bother about your 'al fresco' ESC.

[Erfolg]

Is cooling to the ESC necessary?

IMO not always. I have gliders with no cooling to either the motor or the ESC. With a motor run of less than 30s, if things get warm, they cool during the next 5-10 minutes.

On the other had some models really need cooling. In some cases I cool both motor and the ESC. As has been suggested, in my TH CS, the ESC sits over a tunnel that is 1:1. In that case I also cool the motor. The Lipos, well they just have to tough it out in the body.

My trainer has the ESC sitting on the outside, as does the motor. However the LIPOs again tough it out in a cramped nose, with no cooling. The Lipos do not seem to like it.

In my opinion it is all about judgement and how much full power running you do, relative to the rating of each piece of kit.

[Nigel R]

"In my opinion it is all about judgement and how much full power running you do, relative to the rating of each piece of kit."

There's the thing. The model is quite draggy (as an aside, I'm trying to streamline the cowling and add a bubble canopy to cut drag a bit), and spends much of the flight at reasonably high throttle. Add in a breeze and you have to drive it quite hard to make headway.

As for ratings, the ESC is a 40A and it runs at around 25A full power, so not underrated by any means.

Perhaps a stick on heatsink might aid matters.

[Engine Doctor]

Posted by Nigel R on 11/08/2017 09:16:27:

"In my opinion it is all about judgement and how much full power running you do, relative to the rating of each piece of kit."

 

There's the thing. The model is quite draggy (as an aside, I'm trying to streamline the cowling and add a bubble canopy to cut drag a bit), and spends much of the flight at reasonably high throttle. Add in a breeze and you have to drive it quite hard to make headway.

As for ratings, the ESC is a 40A and it runs at around 25A full power, so not underrated by any means.

Perhaps a stick on heatsink might aid matters.

 

Overheating will only be delayed by fitting a stick on heat sink unless you can increase air flow over the ESC.

[Erfolg]

On most of my questionable set ups, and occasionally on those where ducted cooling is used, after the initial flights, I do a finger tip test. Does anything feel warm or hot. It does seem habitual, i do this judgemental check quite regularly, particularly on those things I can get at easily.

I general most things turn out to be pretty cool. This could partly be down to conservative selection of ESC relative to current draw by the motor, I rarely run my motors at full throttle for long, and then tend to prop down for current draw. The same goes for the Lipo size and "C" rating.

In my case it is Lipos that seem to suffer the most, due to heating, even though I do try to be conservative.

Edited By Erfolg on 11/08/2017 10:27:23

Edited By Erfolg on 11/08/2017 10:27:53

[PatMc]

If running at full throttle is within the ESC's current range it creates less heat in the ESC than running at part throttle.

[Nigel R]

for more context, it's a fairly conservative setup all round:

lipo of 35C 4S 2200mAh; at full throttle of 25A that is about 12C, and the lipo is quite cool when removed after flight

the motor is a 200g (ish) mega inrunner (4/22/30) - like the ESC a max rating of 40A, also comes down very cool (but it is as noted out in the breeze)

[Erfolg]

So where is the problem?

From here, everything in the hen coupe seems well

[Nigel R]

The problem is really the lack of airflow to ESC, it simply sits and gradually warms up. At three consecutive flights it reaches cutout point...

The original question was more about ducting than the system, I guess.

[Frank Skilbeck]

On electric models it's really somewhere for the heated up air to go, if it can't get out of the model it just gets warmer and warmer, so it's not really like trying to cool an IC motor.

I did have one model where the inlet was bigger than the outlet and it worked just fine.

I also on another model put a temperature sensor on the ESC and the temperature went up more at part throttle than full throttle, but I can't be sure whether that was because the model was flying faster or the ESC generating less heat.

[Simon Chaddock]

It always seems odd to duct air inside when there is a lot of it traveling fast just outside the fuselage skin and when coupled with the fact that most of the heat is generated by the power MOSFETS which usually has a flat heat sink on them then why not mount the heat sink flush with the outer skin of the fuselage?.

Of course if the insulating shrink wrap around the ESC can be removed then the heat dissipation capacity is whole orders better.

These are the ESCs of an EDF that are mounted flush with the wing underside just ahead of the air intakes.

Really good 'drag free' cooling.

In principle an ESC produces nearly as much heat as a motor so its cooling has to be seriously considered.

[Erfolg]

Simon, we both know particularly on full size it is the opportunity to control the velocity and potentially to gain some thrust, using ducted radiators, to alleviate some of the ensuing drag increases. On our models it is generally about aesthetics. A bit of protection is also achieved.

Radiators sticking out vastly increase drag.

With our models, surface heat sinks will generally work, in principal in a similar way that the evaporative systems that never really worked on Full size.

My own thoughts are it is only those that are buried and well insulated, in high relative current applications that can be an issue. I have had one.

We have all seen some very powerful motorised Gliders, such as Pike Perfections,Hitch Hikers and other F5J,100s electric types, where everything is squeezed and cajoled into something that looks like 25mm dia (obviously a bit bigger). With not a hint of trouble.

However there are other extremes, as with my trainer.

As can be seen, an IC conversion from a 35 to a +400w set up. No cooling eventually fried the ESC, initially spotted by the heatshrink having shrunk to the extent it had mostly split, no longer covering the heat shrink. The biggest clue was that the set up stopped delivering full power. This meant I struggled to do a "A" award schedule. It (HK Blue Series) never stopped working, after all the abuse

This was modified to lets stick the thing outside, hanging in the breeze, to Hell with the aesthetics and aerodynamics (if it has that much effect anyway).

I do have a few other models, a profile and indoor model with the ESC stuck on outside.

Yet I think you really do have to Push the ESC hard, relative to its rating, for these measures to be necessary.

Edited By Erfolg on 11/08/2017 13:36:16

[John Emms 1]

Hi Nigel, there is little that I can add to what has been written so far.

Controller issues. Your 22/30/4 wants zero to around 6 degrees of timing advance, though the Foxy's tended to work best with the Megas at 22 degrees (the same as outrunners). If the timing or frequency is not at optimum, that could result in a build up of heat. I found optimum prop on the 22/30/4 to be around 11x5,5 running at maximum rating of 450W on 4S.

On the gliders, the top of the fin being open, with holes in the motor mount to allow cooling air through the motor, and to be drawn between the front of the fus, and the spinner always worked well.

On power type models, we always opened up all the formers in the fus, and generally cut a large hole underneath the rear of the fus for cooling, and that allows air in or out as required. Of course, air needs to be able to flow through the motor. For your ESC cooling, I would simply open up a hole in the bottom of the firewall that will allow air to flow through the fus, cooling both the ESC and the batteries.

For a general rule, with glow models, I always sealed them as much as possible, and with an electric model I open up the inside of the fus to cooling air as much as possible - and yes, that still feels strange! Using this strategy, we (as a test and display team of 5) never had an issue with overheating.

Hope that helps,

John

[Erfolg]

John

I certainly agree with the general thrust, of what you do and advice.

I am coming around to one of your implied concepts, that is to get some air moving around the Lipos. Particularly where the Lipos are pretty much squeezed in. The drag increase will be next to nothing in practical terms, as the air bleed in will be a very slow moving body of air in the fuz, a largish hole towards the rear, will effectively have no drag consequence.

I have more issues with short live Lipos than anything else. It will always be the same models. The ratings will always be conservative. I am pretty sure it is the lack of effective cooling. Like the ESC, it does not seem to be much that is ever needed (perhaps other than DF).

So more often than not, I have started to look to get some air moving past the Lipos, if the cabin area has little free volume. I am a long way from all of my models using this concept. Having moved from gliders to sport electric models, I am far more aware that it is the long motor runs at the limit that can be an issue.

[John Emms 1]

HI Erfolg, sorry, I don't have a real name.

You got me thinking. I or we got used to the idea of really opening the inside of the fus up to air freely moving through, which was a real shock after the idea that everything is packed inside and wrapped in foam! The big fear was things breaking loose, so batteries always have been stuck to a light ply plate with self adhesive Velcro, and then have a Velcro strap to make sure the batteries stay in place. I said we never had a problem - but I remember one:

The LN Model Accord 2200 2,2m was a freestyle or 3D model designed for 40cc. Libor of LN then produced a lighter version intended for 10S electric - I bought two as demonstrators. In hindsight, Libor had allowed for cooling air around the cowl that contained the speed controller, but had not allowed for cooling air through the fus that contained the batteries. We had the best 3700 cells available at the time (25C), and peak power was conservative at around 65A. At the end of every flight I was aware that the batteries were warmer than I was comfortable with, and became very slightly puffed. The same power system with the very same batteries was used in my Rascal 110 conversion. The batteries were in a ply box or tube that was in the tank bay, and slid in under the cowl. The battery box had openings at the end, and cooling air could escape through a hole cut in one of the side windows. The batteries were hardly warm at the end of the flight, and I was never concerned about them - other than knowing they had slightly puffed during use in the Accord.

I had always thought that people suffering short battery life were being sold cheap batteries, using them at high currents, and not balancing regularly, but I think you could well be right, that good levels of battery cooling are a large part of long battery life. Yes, I also agree with you that a small amount of air that is allowed to flow freely through the fus does make a huge difference.

Regards,

John

[Nigel R]

Some good points all, many thanks. John thanks for the specifics on the mega motor. Must admit power wise it is about right on the 10x5, freshly charged that is giving about 450W down to around 400W at empty, and I'm getting a 6min flight in with about 25% left in the battery.

Today I flew with the ESC hanging outside in the breeze - 4 consecutive flights, no hint of a problem.

The battery is coming out just warm, so I think that is OK as is. The old fuel line hole serves that compartment quite well.

I think the answer is simple at this point, I will open up the former at the rear of the wing saddle, and open a hole in the fuselage at the same point. And then sort a small scoop to get some air into the ESC compartment (or a hole in the firewall if I can get access with a drill).

[Erfolg]

I guess that many of us have wrestled with the issue of Lipo life and cost that John Emms has raised.

As with so much a lot has changed and possibly will change significantly in the future.

On returning to model making, the switch to brushless motors, was starting, in earnest. The advent of Lipos presented far more viable options for the ordinary modeller such as myself. My first Lipos were I believe called Energens?. These were followed by Overlander products including the Supreme version. They were OK, not brilliant, unfortunately the lives of these units was very variable. Probably the first so called budget brand was Loonmax (again, or something like that), these Lipos were certainly as good and arguably better than the earlier Lipos. Since then i have used loads of different brands, Rhino, Turnigy.

All I can really say, there does not appear to be one brand that appears to be consistently good, because so far most if not all the brands have had a shorter life than expected. I now have a internal resistance meter, that loads the cell, to provide a reading. I do not treat the output as definitive, providing an indicator on a numeric basis on the condition of each of the cells. Often right from the word go, the resistance are often different by an amount that is noticeable. With usage, it becomes clearer that a cell within the Lipo has such high resistance, it is time to bin it. I do have a spread sheet which i have entered the initial values and the values when it becomes apparent when the pack seems to be failing.

But at the end of the day, no definitive answers, although a suspicion, that even a Lipo pack becoming warm, is not the best of news. Although I think this seems to be born out, by it seems more often than not, it is the middle cell on a 3s exhibiting a notably higher resistance than the outer cells. Which may be implying that a warm outer wrapper, indicates that the central cell is much warmer, probably down to less effective cooling. It is just a contention that it is heat that degrades the pack life. I am sure that pulling heavy currents is also a factor. I certainly do not have enough results, nor undertaken with sufficient controls, to say anything beyond I suspect.

Edited By Erfolg on 15/08/2017 16:08:47