Header overflow

[Roger King]

As Peter says, an expansion tank is there to soak up the expanding fluid as it gets hot, physics is physics.
John, I think you've got the problem the wrong way round. Evans expands like any other fluid - perhaps a bit more than normal coolant. It's you (and your choice of cap) that dictates whether the system is pressurized or not.

'Normal' ethylene glycol and water cooling system: as the coolant gets hot it expands, so fitting a sprung pressure cap has the effect of raising the pressure of the system. Raising the pressure has the effect of raising the boiling point of the coolant, therefore it will not boil over until a higher temp is reached - maybe 120-125°C with a 17lb cap. At this temp you will get pockets of steam forming and localised hotspots within the head castings etc. and effective cooling is compromised, leading to a vicious circle as everything gets hotter.
'Evans' propylene glycol (neat) system: the coolant will not boil until around 180°C, which the engine is extremely unlikely to reach. Therefore there is no need to pressurize the system with a sprung pressure cap, as you are not trying to increase the boiling point of the coolant. So you can run a single-seal cap with no spring, but with a suitable overflow tank to recover the coolant when it contracts in volume as it cools down again on switch-off.
With a sprung pressure-rated rad cap you will be creating pressure in the system whatever coolant you use. But with Evans you don't need the pressure, it's doing nothing except finding leaks and stressing the system.
That said, I run a 3lb cap in my 289 simply because it looks right on the McCord tank. I guess I could remove the spring seal from a standard cap, but haven't got around to doing that yet.
I run a straight-through (zero pressure) cap on the Mustang's 393 with a large fabricated overflow tank, as the engine bay looks nothing like stock.
Worth noting that the Ford SBF thermostat housing is a lousy design which has two retaining bolts, but would be far better with three. It is renowned for leaking, hence my preference for a home-made gasket using a rubberised gasket material with an aircraft adhesive-based Permatex sealant rather than RTV.
Also worth noting that coolant is flammable - do not let it leak on to a very hot exhaust system, especially in a fibreglass car!


edit - probably also worth mentioning that there is a better way of venting the overflow bottle than having a pressurised cap on it as Peter suggests. Most caps designed to work with a coolant recovery system go on the header tank (or radiator) as normal, but have two ring seals - one to seal the top of the rad opening, and another with the spring mount to open into the rad neck at set pressure to allow the expanded coolant into the rad neck (below the top seal) and thence go off to the recovery tank. In the centre of the spring plunger of the cap there is a one-way poppet valve which allows the coolant back into the header tank as it cools (and the closed system sucks it back in). Therefore there is no need for a second, pressurised cap on the recovery tank. So you need a cap that does three jobs, pretty common so not hard to find. My recovery tank in the Mustang has a billet threaded cap, with a simple pipe overflow at the top to allow air in and out, and the Cobra one is just a tank with an air vent hole at the top. Thinking about it, if you use a pressure (sprung) cap in the recovery tank it won't work properly as the air will not be able to move in and out of the tank freely to allow for expansion/contraction of the coolant. This part of the system should not be pressurised, whatever coolant you are using.

[IainS]

May also be worth noting that it is difficult to remove all air in the system in the standard Hawk set up such that the expansion tank is the only place you'll find air..... The heater needs to be as low as possible, the feed hose needs to sit below the top of the foot box....if you place a long sprit level on the expansion tank this becomes obvious. I raised my Hawk expansion tank 3 cm and have the heater as low as possible but I'm still not convinced I've removed all the air. Upshot is air expands significantly more than cooland and is the main source of pressure in the system...

Iain

[Karl]

I have a cap from company "Blau" in Austria on the expansion tank which allows you to vent the system. So when the engine is hot I press the button and you hear the air bubbles coming up. When cooled down you need to check the level and top it up. When you have done that three times all the air is out of the system and only the waterlevel in the expansion tank changes with hot and cold. Working since 6 years now.
Karl

[Roger King]

I have a cap from company "Blau" in Austria on the expansion tank which allows you to vent the system. So when the engine is hot I press the button and you hear the air bubbles coming up. When cooled down you need to check the level and top it up. When you have done that three times all the air is out of the system and only the waterlevel in the expansion tank changes with hot and cold. Working since 6 years now.
Karl

Karl, surely that will only work if the heater is lower than the header tank. Is your heater in the 'correct' position, and if so how hot does it get?

As stated on another thread, on my original-spec system the heater is higher than the header tank. It was on my Hawk, too. I fill my system by opening the top hose bleed screw and putting most of the coolant in the header tank. When I get a consistent level in that, I remove the intake hose for the heater at the bulkhead and back-fill the engine via the heater hose. Once coolant starts to come out of the top hose bleeder, I reconnect the heater hose and close the bleed screw. If I only fill the header tank to the top, the top hose is empty. And I'm not changing the design because it's exactly as original!

[Migge]

Heater? What is a heater?

[Roger King]

Heater? What is a heater?

Well, that should make life a lot easier!

[clive]

Heater? What is a heater?

Its a requirement for SVA/IVA. After that its never used.

[John H]

As Peter says, an expansion tank is there to soak up the expanding fluid as it gets hot, physics is physics.
John, I think you've got the problem the wrong way round. Evans expands like any other fluid - perhaps a bit more than normal coolant. It's you (and your choice of cap) that dictates whether the system is pressurized or not.

'Normal' ethylene glycol and water cooling system: as the coolant gets hot it expands, so fitting a sprung pressure cap has the effect of raising the pressure of the system. Raising the pressure has the effect of raising the boiling point of the coolant, therefore it will not boil over until a higher temp is reached - maybe 120-125°C with a 17lb cap. At this temp you will get pockets of steam forming and localised hotspots within the head castings etc. and effective cooling is compromised, leading to a vicious circle as everything gets hotter.
'Evans' propylene glycol (neat) system: the coolant will not boil until around 180°C, which the engine is extremely unlikely to reach. Therefore there is no need to pressurize the system with a sprung pressure cap, as you are not trying to increase the boiling point of the coolant. So you can run a single-seal cap with no spring, but with a suitable overflow tank to recover the coolant when it contracts in volume as it cools down again on switch-off.
With a sprung pressure-rated rad cap you will be creating pressure in the system whatever coolant you use. But with Evans you don't need the pressure, it's doing nothing except finding leaks and stressing the system.
That said, I run a 3lb cap in my 289 simply because it looks right on the McCord tank. I guess I could remove the spring seal from a standard cap, but haven't got around to doing that yet.
I run a straight-through (zero pressure) cap on the Mustang's 393 with a large fabricated overflow tank, as the engine bay looks nothing like stock.
Worth noting that the Ford SBF thermostat housing is a lousy design which has two retaining bolts, but would be far better with three. It is renowned for leaking, hence my preference for a home-made gasket using a rubberised gasket material with an aircraft adhesive-based Permatex sealant rather than RTV.
Also worth noting that coolant is flammable - do not let it leak on to a very hot exhaust system, especially in a fibreglass car!


edit - probably also worth mentioning that there is a better way of venting the overflow bottle than having a pressurised cap on it as Peter suggests. Most caps designed to work with a coolant recovery system go on the header tank (or radiator) as normal, but have two ring seals - one to seal the top of the rad opening, and another with the spring mount to open into the rad neck at set pressure to allow the expanded coolant into the rad neck (below the top seal) and thence go off to the recovery tank. In the centre of the spring plunger of the cap there is a one-way poppet valve which allows the coolant back into the header tank as it cools (and the closed system sucks it back in). Therefore there is no need for a second, pressurised cap on the recovery tank. So you need a cap that does three jobs, pretty common so not hard to find. My recovery tank in the Mustang has a billet threaded cap, with a simple pipe overflow at the top to allow air in and out, and the Cobra one is just a tank with an air vent hole at the top. Thinking about it, if you use a pressure (sprung) cap in the recovery tank it won't work properly as the air will not be able to move in and out of the tank freely to allow for expansion/contraction of the coolant. This part of the system should not be pressurised, whatever coolant you are using.

Thanks Roger, it makes sense that I am actually the cause of the stress! I filled the car with the front jacked up through the heater feed until it flowed out of the return, then through the header tank. From what Clive says and this I have overfilled and pressurised it! I will change the cap .
Thanks all...

[peterc]

I started off like Clive and didn't fill the header tank to the top. By filling to about 1.5" down from the rad cap flange the gap remaining allowed for the required expansion without loss of coolant. However I had problems getting rid of all the air even with the front of the car jacked up. It transpired that the top of the heater was higher than the top of the header tank flange.
This was checked out with a spirit level.
So the answer as related earlier is to have the header tank 100% full and the overflow bottle does the expansion bit. The advantage of course is also that the system is now a no loss system as the header doesn't vent to the ground.
I think that even with the advantage of the higher working temp of Evans you do need to check how high your heater is mounted relative to the header tank neck.
Peter C