Radiator Core Size...cooling capacity?

[ardillolambo]

Hi There,
Due to space constraints I'm planning to use two cooling circuits with 2 small radiators out of a Yamaha 700 ATV. One circuit will be dedicated to the battery and the other will take care of both SDU's the DC/DC converter and the charger when charging. obviously one radiator per circuit. Looking at the Model S radiator core dimensions it have around 145 in3, the two Yamaha radiators combined will have around 170 in3. it is safe to say that the two radiator approach is more than enough to keep everything happy?

What you guys think?

[jrbe]

If you have more radiator area that's good. Just realize airflow in and keeping the hot air from recirculation back into the radiator are extremely important. Fans create high and low pressure areas that the hot air can unexpectedly just loop. On top of that, radiators aren't very airflow friendly, more of a brick than an open window.
Fam size and water pump capacity are very important as well. If you're racing that's one thing. Street use shouldn't be extremely demanding on cooling.

[arber333]

Hi There,
Due to space constraints I'm planning to use two cooling circuits with 2 small radiators out of a Yamaha 700 ATV. One circuit will be dedicated to the battery and the other will take care of both SDU's the DC/DC converter and the charger when charging. obviously one radiator per circuit. Looking at the Model S radiator core dimensions it have around 145 in3, the two Yamaha radiators combined will have around 170 in3. it is safe to say that the two radiator approach is more than enough to keep everything happy?

What you guys think?

I am using fiat seicento 900cc radiator similar to here https://www.rezervni-avtodeli.si/fiat-s ... niw-3299-0
In the past i used a dual fan server radiator https://www.overclockers.co.uk/barrow-d ... 01-ba.html and it was good enough to cool inverter during driving. However I had to add another such radiator because while driving stop and go in the city inverter would build up temperature and eventually trip out.

There is your problem. Motor will not be a problem so much. Inverter overtemperature will be reached before that. You can observe that and see if it will buildup with stop and go and then update your radiator design. In any case you need fans on both radiators since while charging car will be static. You need to dissipate temperature away from radiator to be effective.

[johu]

I've never had to use fans. Definitely not for driving and then as it turned out for charging also. While charging the water passes through so much passive metal that absorbs and radiates heat that there is not need for additional air flow.
Tested with the 3.3 kW Outlander charger on my Touran and the 10 kW Tesla charger on the Volvo conversion.

Addition: air cooled Polo with inverter charging also had no fan

Addition2: I use one 120mm fan that blows over the battery brick in Touran when I go on longer trips

[arber333]

I've never had to use fans. Definitely not for driving and then as it turned out for charging also. While charging the water passes through so much passive metal that absorbs and radiates heat that there is not need for additional air flow.
Tested with the 3.3 kW Outlander charger on my Touran and the 10 kW Tesla charger on the Volvo conversion.

Addition: air cooled Polo with inverter charging also had no fan

Addition2: I use one 120mm fan that blows over the battery brick in Touran when I go on longer trips

Well i must dissapoint then as i use Fiat radiator core in my Pug and without fan active Tesla gen2 charger would reduce power because of overtemp. Even 3K3 Outlander charger would reach uncomfortable temperature without active fan. So i simply add one 9in 80W fan directly to the core. It only consumes about 3A...

[MattsAwesomeStuff]

Yeah outside of race applications one radiator is probably sufficient. Otherwise, yes, more cubic inches than stock is going to be just fine.

Honestly... plumb one, put it into use, see what temperatures are, add a second one if needed. Leave some space for it but don't bother implementing it.

And yeah, what jrbe said, radiators have such dense fins on them that they're "sticky" to the air around them and convection doesn't happen much passively. You need a fan to create airflow and much more than single-digit percentage cooling out of them.

[P.S.Mangelsdorf]

On my LDU powered car, I have had many, many issues with cooling the drive unit on the street.

Here's what I've learned:

• radiator tube size matters. Your radiator inlets should be the same size or bigger than the hose supplying them, or the coolant will pass through too quickly to properly cool. DO NOT step down the size of the hose to fit the radiator.

• pump power matters. I ended up with two pumps on the drive unit cooling loop, a main that pulls from the reservoir and pushes to the drive unit, and a secondary that pushes from the drive unit up to the radiator and reservoir.

• include a reservoir in the system. It is far easier than trying to get air pockets out of a closed loop system. Makes the filling process far easier.

• used electric water pumps can and do fail, usually at the worst times.

• having somewhere for the air that comes through the radiator to go is important. Like others mentioned, its easy for it to get stuck in an engine bay. I'm still working on getting the air out from underhood.

[arber333]

On my LDU powered car, I have had many, many issues with cooling the drive unit on the street.

Here's what I've learned:
....

• having somewhere for the air that comes through the radiator to go is important. Like others mentioned, its easy for it to get stuck in an engine bay. I'm still working on getting the air out from underhood.

See here the deair system i use .
I use a piece of central heating copper tube dia 18mm make a 8mm hole and solder a brass screw valve to it perpendicular. I use plummers paste to guarantee good joint with solder. Zinc i use is thick pure zinc wire no other additives. You can get everything in a hardware store.
The idea is that valve to be on the highest point of the tubing. Your header tank must be higher though to supply liquid consistently.
On the lowest point is the pump and a supply intake. I simply solder 10mm copper tube on 18mm copper tubing so my header tank will supply liquid to the lowest (practical) part and liquid will push air out through the top vent.
There is no problem in using the same header for cooling and heating system as long as it is closed on both sides.

[ardillolambo]

Guys,
I appreciated the info...I guess will go with the two radiators approach wich appears to be more cooling capacity than OEM. Wont put the fans, and see how it does, this car isn't intended to be a daily driver anyways.
Will follow P.S. Manglesdorf advice on using two pumps on the SDU's coolant circuit. For the pumps I'm using OEM ones.

In my case, the highest point will be the front SDU but I was planning to go around that with a bleeder screw as also shown by P.S. also swirl pot will be used.

[ardillolambo]

guys just want to share my design
The battery circuit will use 3/4" (19mm) tubing that matches the Yamaha Raptor 700 2019 radiator Inlet & Outlet ports
The SDU's circuit will have 1" (25mm) tubing. Parallel Tubing to the SDU's will be same length as much as possible. [img

[MattsAwesomeStuff]

or the coolant will pass through too quickly to properly cool.

I've heard this a bunch from normal car guys, and the logic has never made sense to me.

Coolant can't flow "too quickly to cool" through the radiator.

The radiator is removing a certain amount of heat per second. It does that regardless of how fast you're pumping the coolant through it.

If you slow down the coolant, yes, it will spend more time in the radiator, and the coolant that exits will be colder. But then the coolant that was in the engine (or whatever's making heat) will be spending more time in there too, and thus will be entering the radiator even hotter. Meaning, it'll be exiting the radiator even hotter.

You haven't gained anything. You're not making anything cooler either way. Except that, maybe if you insist on removing more heat from a given amount of coolant, since your delta-t is going to be much lower, you'll actually make it worse by slowing it down.

That's my reasoning anyway. I don't know that I'm right, but I haven't heard a convincing explanation of why that's wrong.

[evMacGyver]

Well i must dissapoint then as i use Fiat radiator core in my Pug and without fan active Tesla gen2 charger would reduce power because of overtemp. Even 3K3 Outlander charger would reach uncomfortable temperature without active fan.

Mitsubishi charger seems not to have high efficiency even having not done accurate measurements. For charging on below zero celsius, like -5'C, small ATV radiator without fan gets quite hot with it. This radiator was only a test, this heat is going to the batteries when cold using tesla valves.

North location needs even more planning I think, like I used hose insulator on lines where heating is needed during winter. Valves are needed to get heat either to batteries or radiator. And cabin heater being on the same circuit its getting more complex. Inverter and motor are always cooled while driving..

[P.S.Mangelsdorf]

I've heard this a bunch from normal car guys, and the logic has never made sense to me.

Coolant can't flow "too quickly to cool" through the radiator.

The radiator is removing a certain amount of heat per second. It does that regardless of how fast you're pumping the coolant through it.

If you slow down the coolant, yes, it will spend more time in the radiator, and the coolant that exits will be colder. But then the coolant that was in the engine (or whatever's making heat) will be spending more time in there too, and thus will be entering the radiator even hotter. Meaning, it'll be exiting the radiator even hotter.

You haven't gained anything. You're not making anything cooler either way. Except that, maybe if you insist on removing more heat from a given amount of coolant, since your delta-t is going to be much lower, you'll actually make it worse by slowing it down.

That's my reasoning anyway. I don't know that I'm right, but I haven't heard a convincing explanation of why that's wrong.

If you only change the radiator tube diameter, that does not effect the speed of coolant in the rest of the system. That speed is determined by the pump, which is not changing in this instance. When the radiator is a restriction, such as a smaller tube diameter, it speeds up the coolant IN the radiator, but the coolant slows back down upon reaching the larger diameter hose. It’s a localized velocity change due to a larger or smaller cross section of the tube.

[MattsAwesomeStuff]

When the radiator is a restriction, such as a smaller tube diameter, it speeds up the coolant IN the radiator, but the coolant slows back down upon reaching the larger diameter hose.

I'm following you but... ... So?

Coolant is still in the radiator, heat is still being radiated. No more or no less than otherwise.

I can understand perhaps if you simply restrict coolant flow by whatever method, (fold a hose nearly in half so barely any gets through) you won't circulate coolant fast enough. But the cause of that would be too much heat in the engine (or whatever), not "coolant moving too fast through the radiator". The problem is coolant, even if cold when it leaves the rad, moving so slowly that by the time it exits the engine it's way too hot and boiling. Not that it's moving too quickly.

[jrbe]

I think the coolant going too fast thing comes from high engine speed in a combustion engine causing cavitation on stamped impeller pumps. Same could happen from restrictions in the return line or with a pump that isn't sized to the rest of the system.
The cavitation comes from both ugly flow with the impeller blades and restrictions in the inlet causing vacuum which can cause the coolant to boil in those areas.
An added restriction on the pressure side of the pump could help "slow down" the system, but it's usually fighting a suction side restriction. It's tipping the scales to have the restriction on the pressure side to help eliminate cavitation and boiling from inlet restrictions.

Use a vacuum to fill the coolant system. Gets all the air out then sucks the coolant in. It's the struggle free method and well worth the tool cost. Airlift is one brand of many.

[ScythianNite]

Just wanted to jump in to say that I also had my eye on a similar setup to ardillolambo with the dual yamaha banshee radiators, but I think I might go with a larger sized cooler for the motor/inverter/charger size, specifically a civic radiator for the inverter loop and the banshee radiator for the battery since it keeps the same footprint of the original ICE radiator while allowing for more cooling capacity in case/when I take it out for fun drives or track days.

[arber333]

Usually most problems with cooling system are air in system related. Deair and circuit will function good.
Another problem is the wrong position of coolant pump. Pump should be at the lowest point in system with inlet positioned up to receive coolant directly into impeller. When i had it sideways sometimes liquid wouldnt reach impeller and pump went dry.
Another issue is if you have multiple headers and system not pressurized/sealed...coolant might travel between headers or do some weird things...

[jrbe]

guys just want to share my design
The battery circuit will use 3/4" (19mm) tubing that matches the Yamaha Raptor 700 2019 radiator Inlet & Outlet ports
The SDU's circuit will have 1" (25mm) tubing. Parallel Tubing to the SDU's will be same length as much as possible.

If you're using 2 pumps in your design consider splitting the system between the 2 drive units. Meaning one pump for each drive unit.

I do not see a coolant reservoir in the diagrams. That can help get the air out. Also having a reservoir right next to the suction of a pump means you have a visual of coolant level at the pump inlet and are getting bubble free coolant to the pump if it's designed well.

One important thing that doesn't show up on a diagram is giving some effort to not create air traps in the system. These are spots where say a 90° hose starts at a 45° angle from vertical then turns back down 90° leaving a bubble collector. If you have to leave these spots for some reason a bleed line tap could be added with a T fitting that goes back to the reservoir. You want the reservoir to be the air trap and want one at the reservoir so the bubbles make their way out. Add a T and bleed line in any spot like this. Search "coolant bleed tee" and similar will find you a ton of options.

Pressure differences matter here, you'll know if the bleed line is working properly if you see some coolant flowing back into the reservoir through the small bleed line(s). Most reservoirs have a spot to view this well. Some old and crusty reservoirs are so cloudy you can't see it.

Split systems are especially difficult to get the air out. 2 reservoirs might be worth the extra fitment hassle (one front, one rear) to stay air free.

If you're doing different size lines try to make the return a larger diameter than the feed. It's easy to push fluid fast through a small line. It's not easy to suck fluid through a small line.

[P.S.Mangelsdorf]

I'm following you but... ... So?

Coolant is still in the radiator, heat is still being radiated. No more or no less than otherwise.

I can understand perhaps if you simply restrict coolant flow by whatever method, (fold a hose nearly in half so barely any gets through) you won't circulate coolant fast enough. But the cause of that would be too much heat in the engine (or whatever), not "coolant moving too fast through the radiator". The problem is coolant, even if cold when it leaves the rad, moving so slowly that by the time it exits the engine it's way too hot and boiling. Not that it's moving too quickly.

But increasing radiator size does NOT change the speed of fluid in other parts of the system. It's not spending any more time in the engine, but it IS spending more time in the radiator.

Stepping back:
- It takes time for water to transfer heat.
- The more time water spends in the radiator, the closer to ambient air temperature it will become.
- If the water does not spend enough time in the radiator, it will not lose as much heat as it could have.

Let's take a system where the heat source (engine etc), hoses, and pumps all stay the same. Let's say the pump moves one cubic unit of fluid per second.

Let's imagine the radiator as a rigid tube of a fixed length. Radiator A's diameter is such that it holds 100 cubic units of fluid. Radiator B's diameter is such that it holds 200 cubic units of fluid.

Based on the pump's flow rate of one cubic unit per second, that means that a single unit of fluid spends 100 seconds in Radiator A and 200 seconds in Radiator B.

There's still one cubic unit of fluid entering and exiting the radiator every second (the flow rate) but the velocity of fluid within the radiator is lower (aka face velocity), meaning the same unit of fluid spends more time cooling and gets closer to ambient air temperature, WITHOUT slowing down the coolant in other parts of the system.

Said another way, by increasing radiator diameter, you're increasing the total volume of the system and that increase is in the radiator, which means a higher percentage of the system is being actively cooled.

[SuperV8]

I've heard this a bunch from normal car guys, and the logic has never made sense to me.

Coolant can't flow "too quickly to cool" through the radiator.

The radiator is removing a certain amount of heat per second. It does that regardless of how fast you're pumping the coolant through it.

If you slow down the coolant, yes, it will spend more time in the radiator, and the coolant that exits will be colder. But then the coolant that was in the engine (or whatever's making heat) will be spending more time in there too, and thus will be entering the radiator even hotter. Meaning, it'll be exiting the radiator even hotter.

You haven't gained anything. You're not making anything cooler either way. Except that, maybe if you insist on removing more heat from a given amount of coolant, since your delta-t is going to be much lower, you'll actually make it worse by slowing it down.

That's my reasoning anyway. I don't know that I'm right, but I haven't heard a convincing explanation of why that's wrong.

Cooling or heating is all about mass flow rate, which is the relationship between power kW, flow LPS & delta/T (difference in temperature)
power = flow x dT
dT = power/flow
flow = power/dT
For a given power - the lower the delta/t the more flow you need.
For increasing power and fixed dT you need more flow. (more power through a smaller radiator means you need more flow through that radiator)

[MattsAwesomeStuff]

There's still one cubic unit of fluid entering and exiting the radiator every second (the flow rate) but the velocity of fluid within the radiator is lower (aka face velocity), meaning the same unit of fluid spends more time cooling and gets closer to ambient air temperature, WITHOUT slowing down the coolant in other parts of the system.

I think I followed you and agree with you all the way through your whole example. I just don't see how it is related to your claim about reducing the coolant tube size or the rate at which coolant flows.

Unlike in your example, the size of the radiator hasn't changed, so there's still the same amount of cooling in the system.

Said another way, by increasing radiator diameter, you're increasing the total volume of the system and that increase is in the radiator, which means a higher percentage of the system is being actively cooled.

Wait, by "increasing radiator diameter" are you talking about putting in a larger radiator? I interpreted that to mean the sizes of the tubing and such getting necked down. It has nothing to do with the rate of flow through the system, it's just the rate at which a given radiator size can pull heat out. Doesn't matter how fast the coolant is moving through that radiator, fast or slow. I.E. If you put on a pump that was 300% the size or flow rate, you don't change anything about how much heat is removed from the system... you'd only change how much temperature *difference* you'd have between start and end of the radiator.

If what you mean is just that a bigger radiator has more cooling then, yes, obviously.

[P.S.Mangelsdorf]

Wait, by "increasing radiator diameter" are you talking about putting in a larger radiator? I interpreted that to mean the sizes of the tubing and such getting necked down.

I'm talking about a radiator with a larger tube size. The radiator itself may be the same physical outside dimensions.

[MattsAwesomeStuff]

I'm talking about a radiator with a larger tube size. The radiator itself may be the same physical outside dimensions.

Nope, then I lost you again.

I mis-read your example. I thought you had radiator A that was twice as long as radiator B, which is how radiator B holds 2x as much fluid. But also has 2x the surface area, so that made sense to me. What you actually said is that they're the same size, (kinda, one's just fatter, which, isn't going to lose heat any quicker).

I still contend that a radiator of a given size has the ability to remove heat from a system at a constant rate (aside from slight changes in Delta-T, which actually favor keeping it a bit hotter). It doesn't matter how much time the coolant spends in there, the radiator is still pulling heat out at that same rate. Fast, slow, doesn't matter.

I do get that if fluid spends 2x as long in a radiator it will get cooler before it leaves, except that if to do that you're adding more coolant to the radiator, the same radiator takes 2x as long to cool the coolant down so the net result is nothing changed.

...

I'm not saying I'm right, I'm just saying the explanations don't make sense and haven't convinced me. It's not your job to convince me, but if you've got a different way of explaining it maybe that'd click for me.

[P.S.Mangelsdorf]

I'm not saying I'm right, I'm just saying the explanations don't make sense and haven't convinced me. It's not your job to convince me, but if you've got a different way of explaining it maybe that'd click for me.

I appreciate that. Apologies that my explanations have not been great, thermodynamics and fluids were not my best classes in college. (now questions about sound? Those I can explain to anyone)

I still contend that a radiator of a given size has the ability to remove heat from a system at a constant rate (aside from slight changes in Delta-T, which actually favor keeping it a bit hotter). It doesn't matter how much time the coolant spends in there, the radiator is still pulling heat out at that same rate. Fast, slow, doesn't matter.

I think this is the error. A radiator does not have a fixed rate in that way. A radiator is simply a device that increases the surface area of a fluid exposed to a lower temperature. The rate of heat transferred is a property of the fluid, not of the radiator.