na16tlee

Ok, I recently rebuilt my 1.6 with forged internals, cometic MLS head gasket, ARP head studs, and metal intake/exhaust gaskets. I'm having an issue with my coolant going into the expansion tank (overflow tank) and eventually completely filling the expansion tank and venting out. Coolant temp is OK, it runs in the range of 180-210 according to my gauges. If I drive the car easy, I don't have any problems, but when I start to run it hard is when all the coolant will go into the overflow and eventually fill up and vent out.

Tried several radiator caps (even bought an FM with a higher spring pressure) still no luck.

Verified head gasket not the issue, no combustion getting into cooling system, performed leak down test no bubbles from coolant. Had head re-worked and pressure checked before reassembly.

I have a water/oil cooled turbo, any chance excessive pressure may be getting into the cooling system from the turbo? Will it hurt the turbo if I dont run coolant lines to it for a short period of time just to verify?

Has anyone had an issue like this or have any suggestions? I've hit a wall with this one

Joe Perez

iTrader: (8)

Joe the Cynic asks if perhaps the thermostat itself is defective, installed backwards, clogged with a bunch of stray mouse fetuses that fell into the engine while you were assembling it, etc?


Virtually zero. The cooling jacket is separated from the rest of the center assembly by a wall of solid metal- no seals or gaskets in that path. If a crack / porosity / etc. existed, you'd probably be blowing coolant out the tailpipe and/or getting oil into the cooling system.

Nope. Some lazy ******** never even connect them at all.

na16tlee

Thanks for the feedback. I ordered a new TStat anyways since I can't remember the last time it was done, and a higher pressure radiator cap 22-24. The one I have now is supposed to be 18 psi but it is at least 2 yeas old now. I'll let ya know if I have any luck.

Any chance some improperly routed coolant lines could cause any weird issues too? I'm pretty confident they're ok but I re-routed some hoses to bypass the lines going to the throttle body. I know I need to do the re-route kit eventually but I wanted to get this issue figured out first.

fooger03

iTrader: (2)

How did you "verify" that the issue wasn't the head to block seal? Unless you put an AFR meter in your radiator, I'm still suspecting the head to block seal. Tell me more about getting your head "reworked" - what method did the machine shop use to clean/plane the bottom of the head? A compression / leak-down test may show the head sealed, but the force of WOT combustion is magnitudes greater than the force of simple compression. Did you strip any of the head studs out when tightening?

Joe Perez

iTrader: (8)

Interesting...

In the stock configuration, there are two small coolant lines connected to the lower/inner thermostat housing. One from the intake side of the engine (throttle body, etc) and the other going to the mixing manifold at the water pump inlet.

The one coming from the throttle body is not of great consequence. The one going to the mixing manifold is of high importance- it ensures that coolant is flowing past the thermostatic element even when the thermostat is closed.

Still, I'd expect the thermostat to open eventually...



I am also curious about this. In a perfect world, he'd put the car onto a dyno with a cold engine, remove the radiator cap, and then run the engine up to full load while looking for bubbles coming up out of the radiator.

In an imperfect world, he could do this same test with a cold engine while the car is simply parked in neutral with no load. (Bearing in mind that it's normal for the coolant level in the radiator neck to change very slightly when the RPMs change rapidly. Bubbles are what we seek.)

na16tlee

CNC machining on the block when i had the forged internals installed, and CNC on the head at a head shop. All of the head studs torqued just fine. I pulled the head a few weeks ago because that was my first thought was it was a head gasket issue. Head gasket looked just fine. While the head was removed that's when i had it pressure tested for any leaks or porous sections. No leaks all checked out ok. I removed all the head studs, cleaned all the bores and reinstalled and sealed the head studs again just to be sure. I have pressure tested the cooling system haven't had any leaks, no coolant getting into the cylinders even after pressure testing overnight.. i also used a tester that can detect combustion getting into the coolant by using a chemical that will change colors if it detects any combustion gasses, nothing found. I know this isn't a super accurate test but I feel pretty confident the head to block sealing is not the issue. I even had the car up on my lift and drove it in gear to about 7k rpms, I know its not the same amount of load as a dyno but i didn't have any air bubbles coming out of the radiator. Thanks again for the feedback so far

fooger03

iTrader: (2)

The problem is either load or speed based.

OP, take the thermostat out of the car and see if you still bubble. If the problem is fixed, then joe perez's recommendation of the coolant bypass hose between the water neck and the mixing manifold is probably correct; the convection process which transfers heat through the water in the head is too slow to open the thermostat for it to open in time. If this is the case, you should also see an accompanying spike on your temp gauge, assuming that you have a good temp gauge and not the OEM temp gauge which points to the middle for about a 10,000 degree range of heat. The air bubbles are caused by boiling water, and although a higher pressure radiator cap will help by increasing the boiling point of your water, it will not solve the problem, only reduce the symptoms.

If that doesn't fix the problem, then either the head, the block, or both were improperly planed. It is absolutely possible that cooling system pressure testing (20psi) compression testing (150psi), leakdown testing (150psi), and small combustion events (200psi?) produced at idle, cruise, and light load are unable to escape from your cylinder into your cooling system, while WOT combustion events (600psi?) combined with localized heating produced by those events allows some of the combustion event to escape past the MLS and into the cooling system. I have seen a head gasket failure that only produced evidence under WOT before. I can think of few alternative failure modes which would allow air into the cooling system without also providing a free flowing "outlet" for your coolant to empty into the environment...

By chance, did you remove or alter the heater core routing?

hornetball

iTrader: (4)

Seems like your next test should be to run it with the turbo water cooling bypassed.

Just taking a step back, what is happening here is that your cooling system is being pressurized to a point beyond the radiator cap rating. Two things will cause this:

1. A source of high pressure is getting into your cooling passages. Usually pressure from the combustion chamber. There should be obvious secondary indications when this is happening.

2. Phase change of the coolant from liquid to gas (basically, boiling over).

The testing you've done so far tends to exculpate cause #1, which is a fairly common issue and was your (and ours) first obvious guess.

Moving on to cause #2, based upon your temperature readings, this doesn't seem to be happening within the engine block itself. The only other high temperature area that the coolant is exposed to is the CHRA. I haven't ever heard of boilover being caused by flashing in the CHRA, but I suppose that there is plenty of heat there at higher power settings if the coolant flow is inadequate. "Maybe" that's what's happening here. A bypassed run should prove or disprove it.

Anyway, rambling thoughts, hope they were helpful. Good luck and let us know how it goes.

Post edit -- agree with fooger, a run without a thermostat would be an excellent test too.

na16tlee

Well I have verified all the coolant lines are routed properly with another miata (without a turbo but close enough) Also verified heater core lines correct. I have 3 different cooling temp indicators, one for the megasquirt, one for the cluster, and one for my added auto meter gauge in the dash. According to the gauge in the cluster, it goes up to the halfway mark maybe just a touch above. (I know not accurate at all) My auto meter gauge reads 180-210 pretty consistently and the temp sensor used for the megasquirt is pretty close to identical. I do agree that the very high pressure when under boost could be the culprit and the tests I've done wont exactly rule that out, but before pulling the head again I guess I wanted to get an idea if maybe there was something I may have overlooked. I'll run the car without a thermostat this weekend and see if I get any different results, but if that and the radiator cap don't solve it I guess I need to look into the head to block sealing once again. Just out of curiosity is it possible that the head should be torqued differently or higher since I'm using head studs and MLS head gasket? The stock one was a composite gasket if I recall...just a thought...

hornetball

iTrader: (4)

Are you going to do a test with the turbo bypassed?

Also, there's a cooling passage to the intake manifold. This feeds the small lines on the intake side and you may (or may not) be using it to feed your turbo. Are you sure that this passage is open with your intake manifold gasket replacement?

Maybe you can draw out exactly how you've routed your cooling lines for us, including how coolant is fed and retrieved from the turbocharger?

Stock 1.6 head gasket is composite.

BTW, when I did my reroute, I got rid of all those little lines and OEM bandaids like the "cursed water plug." Brass or stainless pipe plugs are your friend. Definitely the way to go for long-term reliability.

na16tlee

Yea I will go ahead and bypass the turbo lines this weekend too. I have considered the passage in the intake manifold and I bought a steel gasket when I had the head off last time to try and rule that out. I may pull off the intake too and see if I find anything there. Ill try and draw something out when I get to a computer to show you the routing of my coolant lines.

crowder92

iTrader: (1)

Miata radiators place the radiator cap on the high pressure side of system. This reduces the effective pressure of the cap due to the pressure generated by the water pump. If the cap becomes weak it will have a tendency to leak, especially at higher rpms and power levels.

If the cap is new, is there any chance that the gasket on the radiator cap is not sealing properly or the core of the radiator is causing a flow restriction?

Heres some good info on cooling systems:
Stewart Components - High Performance Automotive Cooling

Joe Perez

iTrader: (8)

How would this even be possible?

The only thing that can create a pressure differential in a coolant loop is the flow of water across a restriction. Flow is created by the action of the water pump, so it must describe one boundary of the differential. The greatest restriction to flow is most likely the narrow orifice of the thermostat, so it most likely describes the other boundary.

Assuming that we define the "high pressure" side of the system as everything which is between the outlet of the water pump and the thermostat, and we define the "low pressure" side as everything between the thermostat and the inlet to the water pump, I'm not sure how you'd place the radiator on the "high pressure" side of the system.

Leafy

iTrader: (1)

Spin the water pump backwards.

crowder92

iTrader: (1)

Joe,

I agree with everything you said.
The thermostat is the primary restriction in a cooling system and for a good reason. What some people don't realize(those that remove the thermostat), is that the cooling system needs a small restriction on the outlet of the head.

When the water coming off the water pump comes into contact with the thermostat the restriction raises the pressure in the cooling passages and forces water into head cavities that may not get sufficient circulation otherwise. Increased block pressure also reduces water pump cavitation, keeps air in the system compressed and maintains the water to metal contact that is vital to prevent localized boiling and steam pockets.

You don't want so much restriction in the engine that it reduces coolant flow to the point of over heating. An ideal system would have a very little restriction to coolant flow to maximize heat exchange rates and very high block pressures to maximize the above mentioned qualities. Racing teams have done this by using what is referred to as a "pressurized system" which use an adjustable pressure relief valve and accumulator tank much like an accusump. These systems operate at 30+psi, and have allowed teams to downsize their cooling systems.

So that being said, when I first built my Porsche (mid-engined V8, front radiator) 13yrs ago, I installed the radiator cap on the top radiator hose just after the thermostat housing, which was the high point in the system. With this layout I was constantly puking coolant out the cap at high rpm and onto the water pump and alternator, which would in turn fling it all over the engine compartment. When I re-engineering the cooling system in 2005 I called C&R radiators and Meziere water pumps to consult on the system. C&R radiators informed me that locating the radiator cap on the upper radiator hose or inlet tank of the radiator lowers the operating pressure of the cap, and that I should only use cross-flow radiators for that reason (cap on the outlet side of the radiator). I ended up purchasing radiator cores and building my own end-tanks for the system. 26x19x3 radiator with an inline electric booster pump for the double pass radiator up front, and two "mini stock" 17x12x2 cross-flow radiators run in series in the rear quarter panels. All of the low temp sides of the radiators tie into a common expansion tank with a 28-32psi cap.

This system allows me to sit in traffic on 100-113* days we have here, and track the car with 800hp and not worry about puking coolant or overheating.

hornetball

iTrader: (4)

Before we get lost in semantics . . . let's boil this down to a clear concept.

There is 1 (ONE) controlled pressure point in your cooling system. It is the radiator cap. The pressure at this point will be atmospheric pressure + the pressure caused by the radiator cap spring.

Crowder's point is that you have a choice as to where to place the control point. If, for example, you place it in the radiator outlet tank by using a crossflow radiator, you raise the pressure of the entire system vis-a-vis placing it in the radiator inlet tank.

That said, I can say with 100% confidence that radiator cap placement in the OEM location is not OP's issue.

na16tlee

Yea I think I've come to the sad conclusion that it is indeed a sealing issue between the head and block. I've basically been trying to convince myself that it's something else or at least rule everything else out. I will be pulling the engine again once I get time and have someone else take a look at the head and block to see if one of the 2 are off. Thanks for all the replies and I will report back with the findings.

Thanks again

supercooper

I highly doubt the block will be warped, as its usually the aluminum head.. but, i guess you never know til you check it out

Pinky

iTrader: (1)

I'm just now reassembling a small block chevy at my shop that had the same symptoms as your car: the cooling system would over-pressurize and either blow off a hose or vent through the cap. It's a fairly radical small block; 11 to 1 with flat tops and double hump heads with 64CC chambers, and it was just allowing a tiny bit of compression and combustion gas to get past the head gasket. Barely noticeable bubbles in the radiator, and it *barely* read anything when we used a tailpipe sniffer to sample the gases at the radiator cap (if you have access to one it's a simple way to check for hydrocarbons in the cooling system- hold the sniffer ABOVE the coolant level, just at the top of the radiator neck, and wrap a rag around it to seal things up..) but a cooling system pressure gauge showed it slowly creeping up as the motor ran. (At 25 psi indicated on the gauge it blew off the lower hose again..)

Pulled it apart and didn't see any obvious signs of head gasket failure, although the original builder had used regular old composite type gaskets. We had the block and heads pressure tested and checked for straightness and then buttoned it all back up using Cometic MLS gaskets. With MLS gaskets it's a good idea to give them a coat or two of copper gasket spray, and ALWAYS retorque them a short while after firing up the motor for the first time. Just turn them all back a quarter turn and then retorque to spec in the correct order. MLS gaskets are awesome as long as everything is absolutely flat and clean, but a lot of folks aren't aware of the need to retorque them.

I'd be willing to guess that it's your head gasket; there just aren't that many places for it to happen otherwise. It might be so slight that a parts store combustion gas test kit doesn't register it.

j-po

Why would one want to retorque the head? A MLS gasket is not going to compress in use. Does Mazda call for a retorque with their oem MLS gasket?