Coolant Radiator Cap

[Sulley]

What do you think would be a proper psi for a Manta Mirage 350 given the system in the Manta? I'm running 22-25lb but I'm thinking that is too aggressive?

[Blueovalz]

I'm running a hybrid set-up in which the radiator has a high-pressure cap (much like you've noted above), but the actual overflow tank at the motor (and also at the highest point in the system) has a normal 15 PSI cap. My rationale was that I wanted the overflow cap to be the one that was the primary protection for the coolant system, and the radiator cap was simply a secondary that could also be cracked to release any air. I used a higher pressure cap at the radiator simply because I didn't want it to release water before the one on the overflow tank.

[Sulley]

My overflow tank is not the highest point in my system. It also does not have a pressurized cap. Does this sound like a problem?

[Blueovalz]

My overflow tank is not the highest point in my system. It also does not have a pressurized cap. Does this sound like a problem?

My assumption is that the cap is at the end of the hose, and the tank itself has no cap (which is how mine is set up). If there was no cap involved at all, it would not serve well as an "overflow" tank.

The tank itself does not need to be the highest point. I've had systems in which the tank was about 12" below the bleed-off (or radiator cap) for high pressure relief, and it was still able to draw water back up the hose (at cool-down) and into the coolant system with no issues. So as long as the location of the pressure relief is at the highest point, the level of the tank itself is less important (if at a reasonable height).

My understanding is limited to the fact that some means of collecting or separating the air in the coolant system is necessary. Designing a system in which a high point within the coolant flow path can accumulate that air is one means. I've done that with a very small collector that has a Presta valve screwed into the top of it. This allows air to be pushed out (by pressing on the valve) while the engine is fully warmed. The valve is not necessary, but it is very convenient in that no bleed screw is required. But the ideal system is to have the overflow tank hose (the hose that goes from the highest point in the coolant path, through the pressure cap, and into the tank) push air/or water cyclically into the reservoir. This allows air to more automatically be expelled from the coolant system on warm-up, and the drawing in of air-free coolant upon cool-down. Cycles of hot/cold incrementally remove air from the system.

I use a product called AirLift, that draws a vacuum on the drained coolant system, then uses that vacuum to draw uninterrupted coolant (usually from a bucket) back into the system, thus eliminating a lot of the air. This still leaves a small amount of air (as observed through the sound of water through the electric coolant pump), but I've never had an over-heating situation since using this product. With that said, I'm using the LS series motor in my Mirage, which does not have the benefit of the 2nd generation GM engines that can be filled without air in the system (for normal front engined cars that have the radiator cap as the highest point in a very short coolant circulatory system). Because our cars have an extended coolant circuit, depending on the configuration, it can be much harder to refill the coolant without getting all the air out, so bleed screws must be incorporated into all isolated high points in the system (e.g. top of our radiators, camel-hump bends in tubing or hoses, heads, etc).

In the attached photo, the coolant flows from a high points in the coolant system into the bottom of a very small cavity with a radiator cap on it. At specified pressure, the cap opens, expels coolant through a hose (hose is not connected in this photo) into the the bottom of the overflow tank. The tank is kept at atmospheric pressure by the hose coming out of the top, which dumps out the rear of the car (open air) if the overflow tank fills to capacity.