[catphish]

I'm beginning the assembly of my car. I'm struggling to find a good place to mount my header tank, which has lead me to realize that I don't have a very good understanding of coolant systems in general, and in particular how to get air out of the system.

In my car, the setup will be as follows:

• Tesla drive unit, at rear, coolant passes through twice, inverter and motor
• Battery pack, at rear immediately above drive unit
• Battery charger, at rear behind drive unit, approximately the same height
• Radiator, at front, roughly the same height as the batteries
• Pipes front-to-rear, one down each side of the car, well below any of the other components
• Header talk, front of car, next to but slightly above radiator

Please see artist's impression below (I'm not sure if the highest point is the radiator, or the batteries).

Will this setup suffice? Is air likely to get trapped at the rear in a system like this, and does it matter where the highest point is, or will the air be pumped around the system and leave via the header tank as intended? I've not seen much written on the forums about air in cooling systems, is it not much of a problem?

 

[remy_martian]

The "header" tank is at the top of the system - hence the name.

Which is why I suggested building it into the battery box lid in your car. A half inch space (or so...you do the math) over that area can be the equivalent of a coolant bottle.

You'll have to vacuum bleed the system in any case (see @windraver's CRX thread where he found out the hard way you're supposed to vacuum bleed).

In addition to air bubbles liking to go up, coolant also likes to go down. In the event of a pump failure, your fiddly bits stay immersed while you poop your pants pulling over due to temperature alarms going off. I have yet to see a surge tank or header be lower than anything else in a coolant system, but I have to leave something for Brian 😛

**Caution: this "top of the system" assumes there isn't an airhead driving the vehicle in your CAD drawing

ps I see you're going with ghetto wheel rims

 

[remy_martian]

One caution...if you do go with the lid coolant tank, you'll need to scavenge coolant from a few points to account for cornering, braking, acceleration, road camber, etc...

Pump seems to usually be fed first from the header tank in systems I've looked at.

 

[windraver]

I picked up an Airlift coolant refill tool off amazon:

550000 | UView Airlift™ Kit & Cooling System Replacement Parts

The 550000 UView Airlift™ Kit eliminates airlocks in vehicle cooling systems and refills in seconds. Find replacement parts and key product benefits here.

www.cpsproducts.com

word of warning that it requires a beefy air compressor to run. my air compressor wasn't keeping up well.

also important to note, the reservoir tank "leaks" vacuum so I basically clamped it to seal it off while I vacuumed the entire system.

 

[remy_martian]

Also brings up a point of pinching the homebrew surge tank off or it will collapse under vacuum because of its area if you incorporate it in the lid.

 

[catphish]

After much worrying and some advice from others, I have decided on a few things:
1) I will place a large at the rear of the car. This can store a bulk of water and the the pump can take water directly from the bottom of it.
2) I will place a filler tank above it, stashed in the back of the luggage boot with either single or dual pipes down to the main tank.
3) I will use OEM radiator at the front, and put a bleed valve next to it just in case, but I don't anticipate the pump having any trouble pushing the air out of it.

The hope is that air should get pushed around the system and gather in the top of the main tank, which can then be topped up via the header tank in the boot.

The other question this has raised for me is what to do about the batteries. I see three choices:
1) Put the batteries in the main cooling loop as I originally planned. This has the substantial disadvantage that heat from the motor may heat them excessively during discharge.
2) Don't cool the batteries at all. At charge currents of 0.3C and discharge averaging 1C (with peaks at 10C) cooling may not be necessary.
3) Run a second cooling system with a small radiator for the batteries.

The third option is clearly the superior one, but but will require a second (rear) radiator and a second set of tanks. Quite a lot of extra work.

 

[Ken S]

It's good to have not only a remote fill at the highest point in the system, but also one or more drains at the lowest point(s). If your lines on the radiator are at the bottom as the artist's rendition suggests, you will need to vacuum fill the system, and/or install a bleed valve to vent trapped air from the radiator. A simple screw with a rubber gasket will work. After filling, run the pump. Loosen the screw while the pump is running and wait until liquid seeps out, then tighten the screw. Go to mcmaster dot com and search "Sealing Blunt Screws for Sheet Metal" or "Sealing screws" for ideas.

 

[remy_martian]

Again, pumps do not push air out of a system - you keep wishing this will happen. If they did, anti-thump columns in water pipes (essentially a column of air to keep pipes from thumping when you close valves quickly) would run out of air.

Consider an upside down U bent pipe with air in it. The pump will pump fluid happily over the U but air will remain in the upper cross section at the top of the U because fluid can move from the upward flow to the downward flow in the lower cross section of the pipe, not disturbing or displacing the air space. Now put your IGBT at the top of that U. It has no liquid in contact and will fry. This is what @windraver observed in his build.

In every cooling system I've seen, fluid is returned to the surge tank...at the very top of the system - where your "filler" is. That surge tank has an air space in it to accommodate expansion. You don't have that in your drawing.

 

[Ken S]

Mr. Martian is mostly right. If you have a trap with no path for the air to bleed out, there will be an air pocket. A bleed valve needs to be located where is can vent the undesired air out. In order expel air from the system you have displace the air and you have to open the bleed valve(s). The upside down U is a common feature of many septic systems. (Though in that application, the air gap serves a function.) Look closely at the fittings near the U, and you'll probably find a bleed screw.

Mr. Martian is correct on another point - you need to accommodate thermal expansion of the coolant. The expansion tank is good for that.

Most of us played in the bathtub as children. I expect most of us have bled the air out of a brake system. Many work with hydraulics, and some do it for a living. But every project has its unique challenges. You'll get lots of advice from plenty of people who have not seen the details of your implementation (present company included). Take whatever ideas you find most helpful... Just get the air out.

 

[catphish]

Again, pumps do not push air out of a system - you keep wishing this will happen.

This can only be partially correct. If a pump can pump water then it must be able to push air ahead of the water. But clearly only as much as is necessary for the water to pass.

Consider an upside down U bent pipe with air in it. The pump will pump fluid happily over the U but air will remain in the upper cross section at the top of the U because fluid can move from the upward flow to the downward flow in the lower cross section of the pipe, not disturbing or displacing the air space. Now put your IGBT at the top of that U.

I absolutely agree, this is clearly a problem where water travels up and back down, as I'm sure it will in my build.

In every cooling system I've seen, fluid is returned to the surge tank...at the very top of the system - where your "filler" is. That surge tank has an air space in it to accommodate expansion. You don't have that in your drawing.

I've essentially reached the conclusion that the best option is to use a header tank, at the highest point possible, with the pump drawing water from the bottom, and being returned to the same tank and the end of the loop. This gives somewhere to fill and air to escape. The only problem this doesn't solve is that elsewhere in the system, air may get trapped at high points. I will attempt to add bleed valves in appropriate places to alleviate this, or may consider vacuum filling.

Thanks again for the pointers.

 

[GrayRaceCat]

I picked up an Airlift coolant refill tool off amazon:

550000 | UView Airlift™ Kit & Cooling System Replacement Parts

The 550000 UView Airlift™ Kit eliminates airlocks in vehicle cooling systems and refills in seconds. Find replacement parts and key product benefits here.

www.cpsproducts.com

word of warning that it requires a beefy air compressor to run. my air compressor wasn't keeping up well.

also important to note, the reservoir tank "leaks" vacuum so I basically clamped it to seal it off while I vacuumed the entire system.

I came here to say exactly this. I used one at work for the last 15 years. I started with the UView Airlift II (plastic) but wore it out. I replaced it with the UView Airlift (Brass) that windraver shows above. I recommend getting a clear bucket with a lid to hold the fresh coolant, you will need to see the level so you don't draw in any air. https://www.amazon.com/Chemical-Guy...gallon+clear+bucket+&+lid,aps,212&sr=8-1&th=1 Vacuum filling is the best way I've found to fill a cooling system.

 

[GrayRaceCat]

Also brings up a point of pinching the homebrew surge tank off or it will collapse under vacuum because of its area if you incorporate it in the lid.

Yup! 24+ inches of vacuum will find the weak points in your design/construction fast.

 

[rob_kr]

I know these are frowned upon, but I've had good luck with them... My cooling circuits are kind-of like yours with some very high points and a radiator way down there, there's batteries in the front (the entire engine compartment up to the hood) and in the back (under the trunk). To make it all work without crawling all over all of the time, I've run my lines to where I have a few along-the-way high points. Those are way lower than the absolute top of the system, but high enough compared to what's right around them to make bubbles want to gather there. Then I put these things on each of the high-points. Each one vented its fair bit of air when I first started things up, and now when I press them manually they all produce water right away so the air is "out"...
https://www.grainger.com/product/WATTS-Automatic-Air-Vent-Valve-150-46A964

 

[remy_martian]

Yeah..that doesn't work for the air pockets in the inverter or a HEX.