MeanMini 's cooling system anomaly: coolant push to overflow

[wheming]

So i've been observing an occurance of coolant push to the overflow as i've increased the power level. I have yet to determine a cause and solution. I thought i had a bad radiator which was not drawing the coolant back in.

Here are some details that may be related. Water pump housing was smoothed and ported internally to remove casting flash and soften the edge/corner at the top, #4 cooling mod flows to heater core valve. 180 tstat. Was Nissens type brass/copper 2 row radiator, now Spectra Premium aluminum 1 row.

When i first noticed this i looked over the overflow system because once the engine cooled, the radiator level was low. I added hose clamps to the overflow hose because the spring clips seemed loose and i thought they might have allowed air in instead of vacuuming the coolant back in. I tried a different overflow reservoir. I tried a new radiator cap. Stant 10331 (lever vent) vented cap.
Recently i tried a stant 10330, lever vent unvented cap.

As an aside, I researched what the actual difference is with these caps and the return valve.
The vented caps have the valve hanging, and the seat upward when there is a quick rise in pressure, but normally they dangle and allow vapor to escape to the overflow. However i thought it was also allowing coolant to push out.
The unvented caps have the spring on the return valve. These will build system pressure and maintain it.
The lever venting feature on both isn't supposed to affect normal function, but provide safety release of pressure to remove cap.

I found the bottom tank on the Nissens radiator leaking slightly, and figured it was drawing in air when the system cooled so i changed the radiator. Still the level would not come back from the overflow.

Some recent and older threads talked about the possibility of coolant boiling in the turbo and being a cause. I am using a 50/50 mix of coolant right now.
I ordered a pressurized expansion tank, and i'd like to try installing it to see if that would help.
If there is some boiling occuring, that vapor needs to get out of the block but not occupy space in the radiator, and then when cooled that fluid needs to get back to the radiator. Having an expansion tank, with pressure cap, and maintaing a fluid level higher than the radiator cap should help. I would use the vented cap, or remove the valve from a cap for the radiator, the current low pressure overflow hose will be replaced and will attach to the bottom of the coolant reservoir, the reservoir has a pressure cap and an overflow line that will return to the overflow tank. The coolant will be removed entirely from the overflow tank. The pressurized reservoir will have about 30% level of coolant and 70% vapor space.
This is the reservoir i ordered:
Dorman 603-029 Coolant Reservoir https://www.amazon.com/dp/B00FY80P6M..._ciDmwbXP5RH7J

It looks like it comes with a 16psi cap. I looked at those coolant VW jetta ball shaped ones, but they appeared to be around 20-22psi.

I am trying to nip this in the bud so i can try increasing the power a bit. But right now i'm not sure if this situation is giving me some hot spots of vapor entrainment in the head as this coolant level gets lowered.

If anyone has further suggestions or ideas, i'd like to hear. To me it doesn't make sense why i would have to reinvent the wheel here. Something is going on, but i just can't find it. This should be a worthwhile improvement anyway.

[tryingbe]

Pictures of radiator setup, please.

[csxtra]

Wayne,

Are you using head studs or bolts? Are you getting any knock in your logs? I have seen that problem occur on multiple 8 valve cars when the power levels are high.

My theory is that it happens when the tune is right on the edge of safe (slightly lean or not enough octane/knock resistance) where it isn't knocking yet but there are some abnormal combustion events happening. The cars run fast, but the combustion pressures lift the head enough that combustion pressure gets into the cooling passages, blowing coolant out of the overflow.

That is why Warren Stramer added the extra head studs on the front of the block, to prevent this from happening.

My CSX did it consistently when I was running pump gas and Alky at 36psi of boost. The logs were clean, the car pulled hard (relatively), but I had to top off the radiator after every pass. I saw it happen this fall on another 8 valve running 116 octane race gas, but the tune was leaner than he usually runs because we were tweaking the fuel curves in the cal to lower base fuel pressure so he could run very high boost without putting too much strain on the fuel pump and decreasing flow.

It hasn't happened to my CSX recently because I have been too conservative lately with the boost level (aka slow )

I hope this helps!

Warren

[wheming]

Warren that is very interesting.
I have ARP headstuds, and a Cometic MLS gasket. The deck was surfaced for proper finish and so was the head. The gasket was modified by removing on of the steel layers to make a more appropriate thickness for the compression desired (iirc, we ended up around 8.2:1).
The last time out racing, i had tried 2 gals of 110 unleaded and i had a bit more in regular street gas than i wanted. I did see some brief timing retards. As frequent as my log samples are, sometimes i see 2 or 3 sample events showing some retard. For the most part my AFR's are richer than everyone else suggests i should run.
Earlier on, before the boost was increased much, i wouldn't see any timing retards through the run. And, coincidence or not, did not observe this coolant issue that i am aware of.

I don't really feel like i am on the edge, as i've tried to maintain a rich safety factor. The last couple days when trying to test a new radiator cap, then an unvented cap were on minimum boost (for Mini that is about 10, then boost creep through higher rpm) and it still pushed some coolant.

Do you think the headstuds need re-torqued? Have you found anything helpful, or just removing some timing & enriching?
I could go back through some logs, but so far i don't think i have hit more than about 24psi by the end of the run. There may be a run or two where i hit my overboost setting when i was tweaking in the g-valve (which at the time was 25-28psi).

I'll get a few logs loaded up in a g-drive folder if anyone wants to peak through.
This occurance, and your experienced observations now have me a bit concerned about running at Cecil without somehow addressing it.

[fixit]

i'd expect the volume to increase ~ 1pint when the engine is hot.. so the overflow should be ~1/3 full when cold.

If the thermostat is removed, and the engine is left idling and reaches temperature .. if then its driven hard --- then let to cool down -- when the engine is cool is the coolant full ? (if not could be a leaking head gasket)

I've had a bad head gasket not leak under an overnight pressure test and would idle fine for an hour.. once driving exhaust would displace coolant ....

have you checked the coolant for exhaust ?

[wheming]

Pictures of radiator setup, please.

I don't know how much you'll be able to see but here it is...

Stock type aluminum replacement. Installed in stock location and stock radiator hoses. Stock fan (might have been from newer van though).
180 deg tstat.
4th cyl cooling mod is hard to see, but hose barb 90 deg elbow point down to about 4 o'clock and hose routes over trans and to the far right connection on heater valve.

Overflow hose clamped at radiator and overflow tank spring clamp was tight.

My plan was to install the pressurized reservoir up around where the windshield spritzer is. Maybe i would remove it, or find a smaller one.

[wheming]

[GLHS60]

Wayne, I personally think checking / re torquing the head is important. When checking I like to start down about 50 ft/lbs and work up at 10 ft/lb increments to whatever your final value is. Dave Zelkowsky recommended to final torque the four corner bolts / studs 5 ft/lbs less that all the others and I follow his advice. Sometimes you get movement sometimes not but its a good practice in my opinion.

Rob Walsh tried a very unpopular experiment concerning coolant "evaporation". He ran straight antifreeze in his Turbo engine and his evaporation problem evaporated. Maybe he will notice and reply, although most everyone totally bashed his idea I think it might have some merit and will be trying it myself.

Thanks
Randy

[wheming]

I just came in from a test ride. My coolant level in the overflow is showing higher than when i left, but i want to see what happens after it cools.

I checked the head torque on the ARP studs. Checked at 85 lb-ft. There were about 3 that barely moved. They moved in the sense that it was only enough to know it moved, i can't even describe how slight, but nonetheless it was enough to know it moved.

I let the car warm up with the cap loose, so i know all the coolant had expanded. Then i mental marked the overflow. (Just above half between Min/Max). When i returned just now. I am right at the top of the word Max.

I heard something today, and would want to verify if there are two different ARP stud kits for our application. Did James Reeves have a similar issue and found he had the wrong headstuds? I'll need to check that out.

[GLHNSLHT2]

The studs are the same, it's just the nuts that are different.

[thedon809]

You could put a block tester on it to see if you are getting combustion gasses in the cooling system.

[wheming]

The studs are the same, it's just the nuts that are different.

Is there any way to identify which ones are these?

[wheming]

You could put a block tester on it to see if you are getting combustion gasses in the cooling system.

Do you mean the same as a leakdown tester? I almost tried that today.

[tryingbe]

Was the coolant overflow tank checked for crack? Especially near the nipple area? Was the hose ever changed?

[GLHS60]

Harry, I don't think Wayne would have any crack stashed in his overflow tank !!

Thanks
Randy

[wheming]

Was the coolant overflow tank checked for crack? Especially near the nipple area? Was the hose ever changed?

Yes, and yes.
However, i may try and pressure test it all. As i did not see an issue visually.

[guyd_15]

Cooling system pressure test is a good first step, I've used combustion gas detector in the past. Worked well for me, although the results usually weren't good news.

Here's a link to an inexpensive kit.

http://www.summitracing.com/parts/li...FYcTHwodyF4J0Q

Guy

[tryingbe]

Harry, I don't think Wayne would have any crack stashed in his overflow tank !!

Thanks
Randy

I've learn to never assume when you can check/ask for verification.

When i first noticed this i looked over the overflow system because once the engine cooled, the radiator level was low. I added hose clamps to the overflow hose because the spring clips seemed loose and i thought they might have allowed air in instead of vacuuming the coolant back in. I tried a different overflow reservoir. I tried a new radiator cap. Stant 10331 (lever vent) vented cap.
Recently i tried a stant 10330, lever vent unvented cap.

What are all the radiator cap you have try? Part number? I know you had stant 10331 and stant 10330, but what else?

[Vigo]

Normally, all the pressure in the cooling system comes from the expansion of the coolant due to heat. Having said that, what I would do is warm up the car completely with the system open so that no pressure builds up in the system. Then, put a cooling system pressure tester on the car, using it only as a gauge, and then go do a full boost run down the road with it hooked up and see if the pressure in the system rose substantially when you come back. If so, that pressure would be coming from a combustion leak into the cooling system.

You can also do a 'block test', but I am not sure how recently the gases have to have leaked into the cooling system for the fluid in the tester to change color.

[wheming]

What are all the radiator cap you have try? Part number? I know you had stant 10331 and stant 10330, but what else?

Those numbers are for stant "lever-vent" caps. Also had the no lever stant 10231. That was a vented cap (return valve hangs down).