Can't figure out why coolant won't draw back into cooling system

[Howard]

So my set up is a 427W with a QF carburetor. The chassis has got about 25k miles on the clock. The engine is a Dart block from Fortes, and was installed about 5k miles ago. Recently (1000 miles or so ago) I've started to notice that the coolant does not draw back into the cooling system from the overflow tank. About 1000 miles ago I replaced the radiator (from FFR) as well as adding a solid lower radiator hose from Breeze. Those were the only changes.

My cooling set up is a stainless steel SPF surge tank with the radiator cap on it. The overflow tube comes out of the side of it and goes to the large Canton overflow tank mounted on the X brace.

I recently did about a 650 mile trip and noticed that the overflow tank ended up completely filling up and ended up overflowing by the end of the 3 day trip. The coolant temps were never an issue and the engine has never overheated - max temp on the gauge was 195* in 98* weather and stop and go traffic for 30 mins. Once the car is moving above 40 mph, it drops to 180 or so.

I'll try and explain my observations and what I've done. As best I can tell, the system appears to be staying pressurized, which would explain why the fluid is not being sucked back in when it cools down. So I've let it get fully warmed up, and sure enough fluid moves into the overflow tank. However once it cools down, the PVC hoses are still obviously pressurized. I can't squeeze them much at all.

Once after it was cooled down, I've removed the radiator cap, and about 1/2 a gallon of coolant comes gushing out (it's actually mostly water as i"m running 80% water with water wetter).

So after the next heat cycle was complete (filled it all back up etc) and the engine was cold, I opened the petcock at the top of the drivers side of the radiator, and got a gush of air then it stopped and no fluid came out. Once that stopped, I released the radiator cap, and then opened the petcock again, more air came out, and coolant drained from the surge tank into the radiator. Probably about 1/2 the volume of the surge tank drained into the radiator.

As best I can tell it appears that air is getting pumped into the system which would lead me to think of a blown head gasket. I'm using MLS steel head gaskets. However there are no other symptoms of a blown HG. No white smoke from the pipes, no overheating etc. Oil looks good, although since I'm using water, I'm not sure if it would be as obvious - I've had the chocolate milkshake experience before and would prefer not to repeat it!

So to check for a blown HG I rented a block tester kit, and the fluid never changed color to indicate the presence of hydrocarbons. The upper corrugated radiator hose got hot, so I'm pretty sure the thermostat was open.

Next I got a cooling system pressure testing kit - although in retrospect since the system was holding pressure, it seems obvious now that there would no leaks, and sure enough there weren't any. There have been no coolant leaks at any time.

So my conclusion (or question I guess) is could there be a leak in the head gasket that is causing air to get pumped into the cooling system without any other symptoms?

Any suggestions are appreciated!

Thanks

[CraigS]

First, I'd love to see some pics of your installation. Especially important are how the rad cap and the hose are set on the surge tank. W/o pics my first 2 guesses would be an improper rad cap (not set up for a return under suction system) and a hose to the Canton tank that collapses under vacuum.

[GoDadGo]

First, I'd love to see some pics of your installation. Especially important are how the rad cap and the hose are set on the surge tank. W/o pics my first 2 guesses would be an improper rad cap (not set up for a return under suction system) and a hose to the Canton tank that collapses under vacuum.

Without pictures, it is tough to help, but I'd check the overflow hose and cap first as CraigS suggested.

I had an issue many, many years ago (1980 Camaro Z/28) where the hose going to the overflow tank was so long that it had a little kink in it...It let fluid flow to the tank, but not back...That itty-bitty kink in the hose created a sort of one-way check valve...It actually caused a vacuum in the cooling system as the engine cycled, so much so, that the upper and lower radiator hoses would collapse and the engine cooled when it was shut off...The engine would boil over even though the engine was far below the boiling point...The cool thing is I purchased that car for a stupidly cheap price because it was obviously an issue that the dealer couldn't figure out.

We all know that when we increase the boiling point by having pressure in the system but reduce that pressure and the boiling point occurs far below 212-F / 100-C.

[GFX2043mtu]

I agree you could be having a system back pressure issue. Radiator caps hold what ever pressure they are rated for plus or minus a tolerance usually plus/minus .1 bar. But the key here is that they have a little valve that opens if negative pressure goes above usually -2 to -3 psi (.1 bar). If you pull a radiator cap, that little flat washer thing in the middle of the rubber seal will pull away and light spring force will be felt (it should pull away ~1/16-1/8”). That is the negative pressure valve that opens when vacuum is to high. This is key as the system can push to a surge tank at a much higher pressure then it can pull it. If the overflow tank line is restricted in any way the coolant will be pushed to the tank when the engine warms but won’t be able to be pulled back. Vise versa if the over flow has a radiator cap the pressure ratings are added which is a safety issue and the surge tank pressure cap can cause the radiator cap to leak when surging fluid tries to go to the overflow. Now if the system is restricted when you open the cap air will be pulled into the system causing an issue. Some pictures of the system would be awesome.

[Norm B]

The system staying pressurized after cool down leads me to believe you have a blown head gasket, cracked head or block. How is your oil? Any discolouration?

Norm

[Howard]

Thanks for the replies - I'll try and shoot some video over the next couple of days and upload it to youtube. My impression is that there is air being forced into the cooling system creating the pressure which is making me think blown head gasket. The block is a Dart block, so hopefully it's not cracked! The oil looks fine, but I think I may try draining it and see what it looks like. The system is primarily water so not the typical coolant/water mix. Certainly less messy to deal with.

[Howard]

Here are some pics. I've had an issue before (years ago) with coolant not drawing back into the system that turned out to be a tiny hole that had worn from rubbing in the tube going to the overflow tank - it would let air into the hose, so coolant couldn't be hydraulically pulled up the tube. That was an easy fix. I've removed the hose and checked for leaks and didn't find any. It's also pretty firm rubber and doesn't appear to be collapsing.

[Norm B]

No water in the oil is a good sign. I doubt your block is cracked. Most likely the system didn’t fully purge when you changed the rad and hose leaving an air pocket in one of the heads. This would create a hot spot that most likely cooked the head gasket.
I would try running the engine with the rad cap off and watch for bubbles at the fill port. Be careful doing this because a blown head gasket can cause a geyser out the fill port at any time. A sure sign of a head or head gasket failure is a gurgling sound followed by a large geyser of coolant a couple of seconds after shutdown when testing with the cap removed.

Good Luck

Norm

[GoDadGo]

Could you have some crap in the bottom of the overflow tank since the hoses don't seem to be compromised?
Basically, is it possible that the tank has junk in it that may be clogging the inlet when the coolant is being drawn back into the engine as it cools?

[Howard]

Could you have some crap in the bottom of the overflow tank since the hoses don't seem to be compromised?
Basically, is it possible that the tank has junk in it that may be clogging the inlet when the coolant is being drawn back into the engine as it cools?

I thought of that as well - I removed the overflow tank and tubing from the car, and thoroughly examined it - tank looked fine.

[rich grsc]

Well, right off the bat, you have a degas tank. You should not run a degas tank 100% full of coolant. It works by having an air space in it, when the coolant heats and expands, it compress the air space. No coolant should be forced out of the system, you don't run a recovery tank. I leave about 1 1/2" of space in the tank, and never have trouble.

[Hoooper]

Yeah, everything I have seen with a surge tank holds pressure for days after it is run if there arent any leaks. The tank needs to start out with an air space to work properly otherwise its going to push fluid out at temp but it shouldnt be sucking anything back in, it should be piped to drain that relief line to the ground. Many surge tank vehicles dont even have a drain they just have a cap that releases coolant all over the tank if the cap pressure is exceeded like the coyote surge tank below. You can also see how much air space is in the tank in this picture.

[Mag]

Well, right off the bat, you have a degas tank. You should not run a degas tank 100% full of coolant. It works by having an air space in it, when the coolant heats and expands, it compress the air space. No coolant should be forced out of the system, you don't run a recovery tank. I leave about 1 1/2" of space in the tank, and never have trouble.

Absolutely right. Didn't know the correct term for it (degas tank) we always called it a burp tank. My buddy had one on a 65 390 Galaxy. You don't run an overflow tank with a burp tank. With the engine cold, he would set the level to half full and never had an issue.

[NAZ]

Lots of good well thought out ideas above.

As Rich said, you leave an air space in the degas tank to account for expansion. The amount of expansion can be calculated (that's how you determine the size of a recovery tank) but it's just as easy to let the degas tank puke and once the system is cool the air space left in the tank will show you the cold full level.

You stated that the system is holding pressure and if that's the case, the recovery tank will never return coolant to the degas tank. I'm trying to decide how a leaky head gasket (especially when the dye check eliminated that) would leak like a check valve -- pressurize the cooling system and then not allow that pressure to return through the leak path to a cylinder or cylinders. I've never come across that before and being familiar with Windsor head gaskets, having a hard time imagining that happening. The persistent pressurization is puzzling and hard to believe since I've never seen this and can't imagine how that happens. If not for that I would have tried a known good cap to see if the problem continues and I'd likely still try that.

This is an interesting one so let us know what you find when you finally crack the code.

[bobl]

So, to address the elephant in the room. Why did you replace the radiator to begin with? Could this problem be related? You indicated that it started about the same time. Maybe you had the degas previously at the right level, hence no problem, but after replacing the radiator it was filled too much? Did the engine overheat prior to replacing the radiator?

[CobraboyDR]

NAZ is correct. On FE's (or, really, any engine) with puke tanks, fill it to the brim, get the engine heated to operating temperature, let it burble out, shut the engine down, cool it off, THEN put on the cap.