Creative Cooling Ideas?

rg12

100 kW
Joined
Jul 26, 2014
Messages
1,584
I would like to run a few small packs harder than usual as a test (I usually limit my packs to no more than half of their claimed discharge rate).
I thought of a cooling idea by placing a material that doesn't conduct electricity but IS a good heat conductor flat on the nickels from both sides to speed up the heat disipation.

I once saw these active cooling pads connected to a power supply and never understood how it worked.
Also thought of a thin ceramic panel stuck to the nickels with thermal grease...
This is my general idea, to have the heat sucked from the nickels through a thin non electrical conductive that IS a good heat conductor plate.

Any ideas?
 
a few of the many:
https://endless-sphere.com/forums/search.php?keywords=cool*+batt*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
 
Your looking for a thermal pad or you could use thermal paste but would be messy. They are used in the PC industry to bridge heat to a heat sink or cooling block.
 
Thermoelectric Coolers (TECs) sound like what you remember. It's not a good solution because you need lots of power to run them and that will make your batteries generate more heat and defeat the purpose.

There are Thermal Interface Materials (TIMs) which I am planning on using but more to equalize cell temperatures than cool anything.

You could add a lot of thermal mass to act as a heat sink. That would reduce the rate of change in the pack but still not cooling. I may do a small amount of this with Aluminum sheet metal around the perimeter.

To actually cool a battery pack, you need airflow over the hot areas or liquid cooling. A less well known methods is using phase change wax, it takes an amount of energy to turn it from solid to liquid but it will hold it at a constant temperature until the entire mass of wax has melted.

In my opinion, it makes more sense to massively overbuild your battery pack so it barely heats up instead of spend a lot of money on a complex cooling system. You get increased range in the bargain so win-win.
 
Smoke said:
Thermoelectric Coolers (TECs) sound like what you remember. It's not a good solution because you need lots of power to run them and that will make your batteries generate more heat and defeat the purpose.

There are Thermal Interface Materials (TIMs) which I am planning on using but more to equalize cell temperatures than cool anything.

You could add a lot of thermal mass to act as a heat sink. That would reduce the rate of change in the pack but still not cooling. I may do a small amount of this with Aluminum sheet metal around the perimeter.

To actually cool a battery pack, you need airflow over the hot areas or liquid cooling. A less well known methods is using phase change wax, it takes an amount of energy to turn it from solid to liquid but it will hold it at a constant temperature until the entire mass of wax has melted.

In my opinion, it makes more sense to massively overbuild your battery pack so it barely heats up instead of spend a lot of money on a complex cooling system. You get increased range in the bargain so win-win.

I always size them right so they won't heat up but sometimes I have applications where the room is very limited so I'm looking for a solution for those.

Thought of the thermal silicone sheets used for ecu's and such to transfer heat to an alloy side wall that is the whole side of the battery (hopefully from both sides if the design allows it).
How well do you think this will work over having it inside a plexiglass box and heatshrinked?

Also, the thermal silicone suppliers offer different levels of conductivity that range from 1 to 8 and a sheet of 40*40*2mm cost $5 for conductivity of 1 (something) and about $40 for 8.
They recommend 2 but it sounds low and I ofcourse want the best but that is too damn expensive.
Any ideas?
 
If you are running them that hard, you would just about need a heat sink between each cell. I mean a sheet of aluminum sandwiched in the pack between each foil pouch. Then perhaps run water cooling in contact with each heat sink.

On the other hand, if you run shit that hard, its kind of a racing deal situation. One race and done, at least for another race. Might be easiest to just throw money at it if you plan to abuse cells by that much.
 
There are lots of dielectric potting compounds that conduct heat much better than air, and you have an aluminum case to act as a sink.
 
The thing about air is you can get convection or forced air cooling, that can be much more effective than the thermal conductivity of potting compound.

On the other hand, if your pack is covered in heat shrink with no air flow, that is when the air does nothing for you.

Unless you have an electric boat or some type of ground source cooling for a power wall, eventually the heat is transferred in to the surrounding air, it's just difficult to get it out of the pack if it is sealed.
 
I would always want my packs to be as "weather resistant" as possible, so no airflow inside.

But then I also shoot for setting things up so that any detectable heat rise internally at all, is a problem symptom to be fixed, certainly not to be expected in normal usage.

Depends on the use case of course.
 
Heat in the battery is unavoidable. The question is how much power can you pull out of your cells before the temperature exceeds what you are comfortable with or what the cells will tolerate gracefully over its expected lifetime?
 
Smoke said:
Heat in the battery is unavoidable.

I guess it depends on what you call "heat", and what any particular builder is able or willing to do to avoid it.

If you mean "any temperature rise from ambient", then sure.

If you mean "a significant, easily noticeable rise from ambient" (something like 20-50F higher than ambient), then it's not necessarily "unavoidable". ;)

For instance, I use packs of low-enough internal resistance to not exhibit any noticeable temperature change from ambient, for my usage scenario, so they fall under the second "definition". I'm sure they do have some heat generation (easy enough to calculate, if not measure), but a "back of the hand test" inidcates as close as I can tell the same temperature as ambient, after a ride, regardless of starting SoC or length of ride. Of course this "test" is not really data, and I *should* at least try a generic IR thermometer pointed at them to reverify this. Also, I'm sure there is more heat in the core of the battery, but there's no detectable difference even after some time externally.

But they're also larger than most people want on their bikes. (20-35lbs depending on capacity and container, etc, and 1-2x the size of a common old steel 50cal ammocan).


I've tested a (small) 18650 pack that had such high Ri vs the current draws it was supposedly capable of that it heated massively beyond ambient under my usage...don't really want that. :/
 
Smoke said:
Heat in the battery is unavoidable.
Depends on the use case.

Usually only propulsion ones where light weight is critical actually "require" C-rates that high.

Like flying models, or eBikes used for racing, or "fun" levels of performance.

Others not so much, just keep adding Ah capacity and use low-resistance cells, until you cannot detect any rise in temperature from the outside of the cell while discharging.
 
With typical heat shrink insulation on a battery pack, what you can feel on the outside is not indicative of the temperature on the inside.

Also, there are not multiple definitions of heat. There just is not.

I am currently testing my cells so I can sort them and place them in my pack for the best thermal performance. They are in a fully loaded Opus charger doing 1A charge and discharge rate, inside a .30 call ammo can with the lid completely open (for some fire safety). That's a 0.2C rate, not really insulated at all and you can feel a noticeable rise in temperature on the body of the cells. I noticed the same thing at 0.1C rate also.

In the middle of a large pack, covered in shrink wrap, the thermal environment would be much worse. Maybe it wouldn't be bad enough to worry about at 0.2C but what packs are designed for 0.2C rates? I'm planning for about 0.7C and hoping I'll be able to keep it closer to 0.5C for most of my usage. That should be a lot more heat trapped in the battery. A less dense packing would help, and more parallel cells to reduce the C rate would help too but when you put insulation around a heat source, it doesn't take much to raise temperatures significantly.

I think most electronics are rated for 80°C but I would like my batteries to stay under 60°C if possible.
 
Sounds like cells may be a higher resistance than what would be ideal.

They are new, right?
 
Yes, new LG M50T 21700s. The Opus says around 4,800mAh and between about 60 and 120 mili-ohms.

I would prefer lower internal resistance but it's not too bad.

I'm going to do some statistics to see if any of the slots are giving results that differ too much from the other slots because I was expecting a bit more consistency.

Of course every battery is a compromise between capacity, internal resistance and discharge rate. I went for capacity and then went with 4p when 3p would have worked.

When you go with high discharge rate, you almost always compromise capacity so your C rate goes up even higher.

There are 18650s that are rated for higher discharge than my 21700s but their capacity is about half. If I did a straight substitution, I would get half the capacity but the C rate would double, negating most of the benefit of a high discharge cell. If I doubled the number of cells to get the capacity back and match the C rate, the size has almost doubled and wouldn't fit in my case.

For me fit is critical because I don't want to have my battery or any of my electronics poking out of my bodywork. I could physically fit more but my cell layout is already more complex than I would like. What I have chosen fits, has high capacity and the nickel layout isn't too bad.

Once I settled on that, a plan for dealing with heat became the next priority.
 
Smoke said:
Yes, new LG M50T 21700s. The Opus says around 4,800mAh and between about 60 and 120 mili-ohms.

I would prefer lower internal resistance but it's not too bad.

I'm going to do some statistics to see if any of the slots are giving results that differ too much from the other slots because I was expecting a bit more consistency.

Of course every battery is a compromise between capacity, internal resistance and discharge rate. I went for capacity and then went with 4p when 3p would have worked.

When you go with high discharge rate, you almost always compromise capacity so your C rate goes up even higher.

There are 18650s that are rated for higher discharge than my 21700s but their capacity is about half. If I did a straight substitution, I would get half the capacity but the C rate would double, negating most of the benefit of a high discharge cell. If I doubled the number of cells to get the capacity back and match the C rate, the size has almost doubled and wouldn't fit in my case.

For me fit is critical because I don't want to have my battery or any of my electronics poking out of my bodywork. I could physically fit more but my cell layout is already more complex than I would like. What I have chosen fits, has high capacity and the nickel layout isn't too bad.

Once I settled on that, a plan for dealing with heat became the next priority.

Sounds like you are either buying fakes or low grade cells with that IR inconsistency.
Plus, 60 is pretty high, 120 is junk and the huge inconsistency says that something is terribly wrong there.
Plus, even if it wasn't like that, LG are usually junk, they sag like crazy and have the lowest life expectancy out of the major brands.
Samsung all the way!
 
Smoke said:
With typical heat shrink insulation on a battery pack, what you can feel on the outside is not indicative of the temperature on the inside.
My packs are not heatshrunk.

They are bare EIG NMC C020 cells, with (thin) aluminum heatspreaders on each cell across it's entire surface of one side, so what is on the side of the cells should be similar to, though lower than, the rest of the cell, and brass busbars from cell to cell on the top (under a flip-up acrylic lid to prevent shorting the bars but allow easy access if necessary) that also conduct heat from the cell tabs directly underneath them.

I'll see if I can find the pics I've posted and link them here (I'll edit the post when I do).


Also, there are not multiple definitions of heat. There just is not.
I'm not implying there are multiple definitions of heat, itself. But what is expected of a particular pack may be different from one application to the next. For me, any noticeable warmth generated by a pack would be "heat", and isn't acceptable--it means something is wrong with the pack itself (or it's sized too small), or the system it's powering, to create excessive current draw for what the pack itself is designed to handle.

From what I read here on ES and elsewhere, for many people using 18650-type packs, or 21700s, etc., that's not only often acceptable, it's expected. :(


I am currently testing my cells so I can sort them and place them in my pack for the best thermal performance. They are in a fully loaded Opus charger doing 1A charge and discharge rate, inside a .30 call ammo can with the lid completely open (for some fire safety). That's a 0.2C rate, not really insulated at all and you can feel a noticeable rise in temperature on the body of the cells. I noticed the same thing at 0.1C rate also.

That sounds like cells with what I would call a higher internal resistance. Not atypical for many 18650 types and some others, but is not the only type of cell out there.

The EIG cells I'm using are 20Ah, with 5C continous discharge capability, 10C burst for 10seconds, and the pack on the trike at present is 2p (1p gives the same "hand check" temperature results, though an actual instrumented measurement would show some difference). Total 14s2p pack resistance is estimated by the Cycle Analyst in the mid-30s milliohms (34-35). So that's about 5milliohm per cell.

Normally I'm using them at about 100A or so bursts, for a few seconds at a time, which is 2.5c. Assuming I can math today (not guaranteed), and the CA's estimate is correct, that would be about 14w. (20A * 0.035ohms = 0.7v ; 0.7v * 20A = 14w). 3.5watts of heat within the pack during the acceleration from a stop to 20MPH.

Cruising is a lot lower current, around 20A depending on conditions, which is 0.5C. Assuming I can math today (not guaranteed), and the CA's estimate is correct, that would be about 14w. (20A * 0.035ohms = 0.7v ; 0.7v * 20A = 14w).

So yes, there is heat created, but it is not noticeable (by hand) even if running the pack from full to empty (around 30 miles or so of riding; I've only done this a couple of times).



Charging is done at 0.25c, 12A; the charger (Meanwell HLG-600H-54A) is mounted directly under the pack on the bottom (outside) of the wooden cargo/seatbox (the pack is on the top of the same wooden panel, inside the box).



I think most electronics are rated for 80°C but I would like my batteries to stay under 60°C if possible.
I would prefer to keep batteries under 40C, myself. ;)

Mine are typically close to ambient (which can be as high as 40-50C here; even higher if parked in the sun in a parking lot or riding in traffic, in late spring thru summer).

There was more heating when I was using just 1p, though the current was a little lower (a little more than 80A vs 100A peaks, and less than 20A continous). For instance, what I recall of some results a few years ago when testing in cold (for Phoenix, AZ) temperatures:

After sitting for 8-10 hours in the breakroom at work in high 60s F (warming the pack up to equalize to that), and then riding home to park in the shed at night. By morning I would see about 50F pre-ride temperatures at the very core, with the outer edges cooling to about 40F (ambient).

This had significant voltage sag, decreasing accleration; so I used an incandescent lamp under the seatbox for overnight storage to keep the pack warmer. This kept the pack in the 60s F for it's outer edges, and a few degrees warmer at the core, about the same as when I'd leave work. During the commute of around 2.5 miles, with the repeated hard acceleration and then 20MPH cruising between stops, there wasn't enough self-heating to register on the modified BBQ sensor used to test with at the time. (which is not as sensitive to small changes as typical thermometers as it has a much higher range).


As noted, though, the 18650 pack (13s4p) tested under warmer ambient conditions *did* heat quite noticeably, but despite the ratings claimed for it by the seller, it was not designed for anything like the loads I was putting on it. IIRC it was rated at 10Ah, so I was putting it thru up to 8C peaks, and almost 2C continuous. So I wasn't exactly shocked by the heat, even if I was disappointed. ;)
 
Yeah, 40C would be great but my ambient is about 30C as I speak, 10° rise is probably about what my cells at 0.2C in the charger do. If I add insulation on top of that (heat shrink) and pack them more densely, it's got no chance.

When my ambient gets to 40°, which happens, I'm doubly screwed because I'll probably want to get there quick to get out of the heat.

Having twice as many Ah and 1/10 the internal resistance or less is great but it must be big and heavy. By comparison, my battery pack is going to weigh between about 8.5-10 lbs, maybe 11 if I fill it with paraffin wax for phase change cooling.

By the way, I think you have 350 watts of heating at 100A.

I think I'm going to have about 35 watts at 10A but that's cruising, not a peak.

If I can find the right wax, I think about 2 lbs will let the battery heat to about 50C, and hold there for about 30 minutes, letting me do a 10 mile trip before the wax melts entirely.

If I can control my cruising speed but accelerate hard, I'll have 3 seconds peaks with about 300 watts of battery heating. Not horrible on short trips or infrequently but not great.

I still think it will work because I intend to use it mostly for trips of less than 5 miles (around 15 minutes) and my BMS has a thermistor so I'll be able to tell when my battery is getting too warm.

One trip I'll do frequently is about 3 miles and I'll probably stop and accelerate about 12 times. That 36 seconds of peak power should use up about 20% of the pack's thermal capacity and the remaining 8 minutes and 42 seconds at cruising speed should use up another 30% or so. 5 miles should be no problem even if I increase the cruise speed a bit and try not to use the entire 1,800 watts of power from every stop. My controller has three power modes so maybe I can limit acceleration that way when I'm trying to control pack temperature on longer trips.
 
Smoke said:
Yeah, 40C would be great but my ambient is about 30C as I speak, 10° rise is probably about what my cells at 0.2C in the charger do. If I add insulation on top of that (heat shrink) and pack them more densely, it's got no chance.

When my ambient gets to 40°, which happens, I'm doubly screwed because I'll probably want to get there quick to get out of the heat.
It's not even June (end of spring) yet, already over 35C and it's about to be over 40C here next few days, and probably just keep going up from there. So I sympathize. ;)



Having twice as many Ah and 1/10 the internal resistance or less is great but it must be big and heavy.
It is larger and heavier than the average ebike pack, but my trike, with me on it, is around 500lbs, and about 11 feet long, built to carry St Bernards and/or loads of dogfood, carts of groceries, and trailers of whatever, so it doesn't really matter. Drop in the bucket kind of thing. ;)

It is something like 20lbs, in a old steel 50cal ammocan, for the 14s1p 52v 20Ah version. It sits in the can "facing" down, ends toward the can ends, sides toward the can sides, "back" facing the lid. So it's not really any bigger than a typical ebike pack that's also 20Ah. (I have an A123 20Ah pack that's about the same size, but several pounds heavier, also in a 50cal ammocan; I've never characterized it so dunno it's temperature changes/etc).

The 40Ah 2p version, bare, is about 35lbs, and is a little smaller than the ammocan side/end dimensions, and almost twice as "tall" as the ammocan from bottom to lid (just a double-height stack of cells). That's the size presently on the trike.

You can see the cells over in Jimbob01's thread in the for-sale section, and some other EIG threads around the forum, if you're curious.


By the way, I think you have 350 watts of heating at 100A.
You may be right; I'm too tired to work it out right now--I tried a few times and got different answers each time. :(

But if so, I wonder why it doesnt' appear to heat teh pack, because it ought to turn it into a right good oven. Even if it's only a few seconds at a time, I get that kind of power draw at least a couple of dozen times on just a 2.5 mile ride to work (or home).

And it would mean that the heating at 20A cruising would be 70 watts, which is still quite a lot, and I would expect to notice that even by hand on longer rides (the pack is enclosed inside the wooden seatbox, no airflow, and there is styrofoam insulation between the wood and the battery to help keep the road heat and direct sun heat out of the box as long as possible)
 
rg12 said:
I would like to run a few small packs harder than usual as a test (I usually limit my packs to no more than half of their claimed discharge rate).
[...]
Any ideas?

iced salt water bath?

Alternately, not pushing equipment until overheating is a problem?
 
I think your 20A number is right at 14 watts heating.

1/2 the heating with 4x the mass equals 1/8 the temperature rise. Instead of a my 30° rise, yours is probably 4° over the same amount of time.

With the size of your trike, a 40 lb battery does sound appropriate. If it would fit and the nickel would work out, I would have probable bought more cells to go 6p and reduce the internal resistance by 1/3.

I just got what I could fit.
 
Smoke said:
1/2 the heating with 4x the mass equals 1/8 the temperature rise. Instead of a my 30° rise, yours is probably 4° over the same amount of time.

W/Ah is the figure you want to kno. W/Ah.



Balmorhea said:

iced salt water bath?

I was thinking alcohol. Lol. So many alcohols in this world. So many.
 
good subject and right on time.
it is getting warmer and warmer weather.
our batteries hate heat.
keeping our packs in cool environment while not riding .
I have a problem at my work place.
They disabled /management/ AC cooling .
and temp. gets to 24C at the end of my working day.
I tried putting cooling pads over my battery , blowing fan, etc.
any other ideas?
 
Back
Top