Alfa Romeo Forums banner

1 - 20 of 37 Posts

·
Registered
Joined
·
3,506 Posts
Discussion Starter #1
The '91L we laid up 1.5 yrs ago after it was rear-ended donated its engine and A/T combo to the other '91L automatic after its tranny started slipping atrociously just short of 160k miles. All had been well with the totaled car's engine, as far as we knew, and it had about 120+k miles on its engine/trans combination, so I swapped it as an assembly.

But as I try to start the system up, a quart or two of coolant goes missing every few miles of driving, and I'm getting a significant amount of bubbles (clear) in the water with which I've filled the coolant jacket. I didn't want to load it with antifreeze only to have to throw much of it away if something leaked or otherwise needed attention.

I'm suspicious of a head gasket leaking exhaust gases into the coolant. No oil is apparent in the cooling water, there is no sign of water in the oil, nor is anything strange coming out of the exhaust pipe. I'm thinking I should re-torque the heads. I probably should have done that while the engine was in transit to its new home, but I didn't because it had no known problems in its old home.

Suggestions? Alternate scenarios? Thanks....

Michael
 

·
Premium Member
Joined
·
2,811 Posts
MrT,
I can't address the problem, but I think your diagnosis is spot-on. When replacing the valves on the LS, I was shocked at how degraded the head gaskets were when the heads were removed. They literally crumbled in my hand. And that was on an engine that was running the day before.
 

·
Registered
Joined
·
272 Posts
water pump

When I had this problem, I was able to trace it to a bad waterpump. It was sucking air past the shaft seal at higher RPMs, but not leaking enough water past the seal to get noticed. A compression check ruled out the head gasket. Head gasket don't usually fail unless there is major overheating.
Alfahill
 

·
Premium Member
Joined
·
26,814 Posts
If coolant tank doesn't over pressurrize then head gasket is not likely leaking exhaust gases into coolong system.
 

·
Registered
Joined
·
1,765 Posts
But as I try to start the system up, a quart or two of coolant goes missing every few miles of driving, and I'm getting a significant amount of bubbles (clear) in the water with which I've filled the coolant jacket. I didn't want to load it with antifreeze only to have to throw much of it away if something leaked or otherwise needed attention.
You will have thought of this already I'm sure, but hopefully it is just a lot of air in the system that is taking time to fill with water. The pipe into the front of the expansion tank (from the radiator) may be blocked at the radiator outlet? I believe that pipe is there to allow the system to self-bleed.

Provided the system is not overheating, I'd say to just keep driving and topping-up. You could also remove the brass bolt on the thermostat housing until water flows out without bubbles. I can't remember if there is a bleed screw on radiator (I don't think there is) but removing that small tank-front pipe I mentioned should give same effect. Also heater core may have trapped air, so lots of possibilities.

-Alex
 

·
Registered
Joined
·
3,506 Posts
Discussion Starter #6 (Edited)
I've certainly thought of the trapped volume up in the heads. The bubbles come from the small diameter rubber tube which vents the radiator to the coolant reservoir. I've been topping up the car after every few-mile test outing. I go about 3-4 miles before the coolant tank low-level light comes on (straight water buoys up this float -- some have different experience, possibly because the floats are different?) and the tank takes a 1-2 qts (~liters) of water to top off. I see bubbles continuing to pass from the radiator to the coolant tank, at least when the engine is hot. It is possible that I have a leak in the system, but I don't see water dripping onto the ground under the car.

Fortunately, no water is visible in the oil. I'm thinking I need to re-torque the heads as a pre-emptive measure. I'm not certain whether to put some Barrs Leak (the plant fiber version) into the water prior to doing this. I also find that the water level drops overnight, but again I can't tell where the water is going. If I'd used glycol, I'd see the residue because it wouldn't evaporate like the water does. Maybe I should just consider it a debugging diagnostic investment.

Oh (addition) when I first fired up the engine for warmup, I had the reservoir cap off and saw some really disconcerting level changes, sometimes falling, sometime rising to push water out of the reservoir. I thought of this as possible indication of boiling inside the heads (making extra steam volume) from a stuck thermostat. I guess I should also plan on a compression check.

Michael
 

·
Registered
Joined
·
720 Posts
If you have access to the equipment, a quick way to check what the bubbles are made of is to have someone with emission testing equiptment run their sensor over the expansion tank bubbles to see if they register as exhaust gasses. Probably much faster and easier than an compression test.

Good luck,

Jeff
Dallas, Texas
 

·
Premium Member
Joined
·
26,814 Posts
Oh (addition) when I first fired up the engine for warmup, I had the reservoir cap off and saw some really disconcerting level changes, sometimes falling, sometime rising to push water out of the reservoir. I thought of this as possible indication of boiling inside the heads (making extra steam volume) from a stuck thermostat. I guess I should also plan on a compression check.

Michael
You need to do a run in long enough to bleed system of air and understand that without 50/50 coolant mix the 164 engine will make hot enough for coffee water quicker than you wife's gas stove. Once you get thermostat open it is almost to late to prevent boil over in coolant tank since coolant tank gets hot water from rear of front head long before before radiator gets it through thermostat. That is by design so hopefully you can get heater to work before thermostat opens. Don't really blow a head gasket trying to save a dollar on coolant.
 

·
Registered
Joined
·
1,765 Posts
I go about 3-4 miles before the coolant tank low-level light comes on (straight water buoys up this float -- some have different experience, possibly because the floats are different?)
Ha, yes ;) And the floats will be different, as there are at least three suppliers - I was just lucky that time I think, my aged foam float was right on the limit of buoyancy.

I see bubbles continuing to pass from the radiator to the coolant tank, at least when the engine is hot. It is possible that I have a leak in the system, but I don't see water dripping onto the ground under the car.
The bleed pipe doing its job, then.

I'm not certain whether to put some Barrs Leak (the plant fiber version) into the water prior to doing this. I also find that the water level drops overnight, but again I can't tell where the water is going. If I'd used glycol, I'd see the residue because it wouldn't evaporate like the water does. Maybe I should just consider it a debugging diagnostic investment.
I had a bad experience with Bars Leaks (the black pellets version), so the one you're suggesting is probably fine, but my personal choice is the Wynns Premium Stop Leak in the purple bottle. It is a thick green solution that looks just like concentrated antifreeze, and it works really well. Solved a tricky metal-pipe-union coolant leak at the turbocharger on my Uno. Also on my last 164 after I replaced water pump/thermostat, I could not get those two hoses to the heads to seal. The Wynns sorted that out immediately.

So I'd probably add the Wynns as my pre-emptive strike, but also while there is air in the system, the level will continue to drop sharply as the engine cools, since the air expands by a great amount (as you know).

Oh (addition) when I first fired up the engine for warmup, I had the reservoir cap off and saw some really disconcerting level changes, sometimes falling, sometime rising to push water out of the reservoir. I thought of this as possible indication of boiling inside the heads (making extra steam volume) from a stuck thermostat. I guess I should also plan on a compression check.
Was the temperature gauge reading 'normally' or up-and-down like a yo-yo during this? If up-and-down, then yet again I put it down to air. The gauge, in my experience, doesn't read air temperature very well.

Usually the 164 system should be nicely self-bleeding, so I'm not sure what is going on with your system. I have seen cases on other cars where a minor leak in just the right place (the back of the thermostat housing on my 128, the suction side of the water pump) is enough to let air in, but not water out - but this is rare since most of the system is under pressure.

Just my thoughts and hopefully someone else will offer up a more concrete plan-of-action. EDIT: two of them already have, in the time I've written this ;)

-Alex
 

·
Registered
Joined
·
110 Posts
I agree with alexgs. I think you should watch carefully, but my car did EXACTLY what you describe. It was simply purging air after it sat for months undriven. Don't worry too much yet. Drive it for a while and see....
 

·
Registered
Joined
·
3,506 Posts
Discussion Starter #11
OK. Plans revised. I posted the question/quandry here just so I could get better ideas. I'll invest in (or designate already purchased supplies) putting glycol into the coolant system. I was considering doing the standard mild acid (citric acid or something like that in the standard packages) flush on the system, but I'll forego that now.

It _could_ be something nasty like the fact that the engine didn't have coolant in it for many months and maybe the head gaskets had soaked up water. But the glycol wouldn't have evaporated. Anyway, I'll re-torque the front head, at least. The engine has been operated on the road for something like 1/2 hr so far, with the only symptom being coolant loss.

Michael
 

·
Registered
Joined
·
272 Posts
Don't waste your time as adding Glycol won't make a difference. Plain water transfers heat better. Do the easy diagnosis first: See if the coolant is being lost from the tank at high RPM. If you rev the engine at 3K RPM, you will see if the bubbles are pushing liquid out from the overflow cap. If so, you either have a bad head gasket or a bad water pump. Compression test will tell you which. Try the simple stuff first. I still think you have a water pump with a dry seal that is sucking air at high RPM> -Alfahill
 

·
Premium Member
Joined
·
15,537 Posts
I wouldn't use just plain water in the cooling system, as the systems are designed to take advantage of the water/antifreeze mixture characteristics.

Things to be considered which are important are that the boiling temperature is raised significantly and the freezing temp is lowered when using modern antifreeze mixtures, and, you miss out on the cooling system anticorrosion additives which are in the mixture. In fact, many newer cars will overheat when just plain water is used. Friend had an Oldsmobile which, after a waterpump change, kept overheating until they drained out the plain water, and replaced it with the antifreeze mixture. Dealer confirmed that the radiator was smaller to take advantage of the antifreeze characteristics, and that just water would not work.
 

·
Registered
Joined
·
3,506 Posts
Discussion Starter #14
I appreciate the niceties of engineering, and that certain non-intuitive things can happen. I don't understand exactly why plain water shouldn't work well, but I _can_ see that if the thermostat were to be set too high, water wouldn't give good results. Knowing of this one counter-example that I understand, I can believe there might be others that I don't understand. I had only planned on using water for the leak check portion of the testing, but I've not understood fully what I'm seeing and that has delayed me.

I will put the glycol in tonight or tomorrow and give it a run. Local boiling would really fog up my crystal ball. But I've ruled out the simple faults now, and maybe there are no others hiding. I'll see soon, and it won't cost me much and I'll not be fretting over it for days, so that is all to the good.

Thanks for the great input from _everyone_.

Michael
 

·
Premium Member
Joined
·
9,742 Posts
I appreciate the niceties of engineering, and that certain non-intuitive things can happen. I don't understand exactly why plain water shouldn't work well, but I _can_ see that if the thermostat were to be set too high, water wouldn't give good results. Knowing of this one counter-example that I understand, I can believe there might be others that I don't understand. I had only planned on using water for the leak check portion of the testing, but I've not understood fully what I'm seeing and that has delayed me.

I will put the glycol in tonight or tomorrow and give it a run. Local boiling would really fog up my crystal ball. But I've ruled out the simple faults now, and maybe there are no others hiding. I'll see soon, and it won't cost me much and I'll not be fretting over it for days, so that is all to the good.

Thanks for the great input from _everyone_.

Michael
I would use 50/50 mix. Here is why I think this is a better way to go than just water although I will contradict myself later...
1. lubrication of thermostat and water pump, regardless if a short run.
2. if you have air in the system you want maximum cooling with ethelyene glycol. If there is AIR trapped in the block somewhere you are creating a hot spot which can cause more damage than a $13 bottle of Anti-Freeze. Even if you use 2,3,6 bottles before you figure out the problem you still saving money.
3. Corrosion!

Now I have a tech section on my site(upgrading it soon)that explains the cooling system. Quoted from my site:

Water has a higher specific heat than an ethylene glycol or propylene glycol coolant mix. Therefore, it provides the best heat transfer performance in a cooling system. If a cooling system is marginal, that is, it only overheats on the hottest of days, then running with water as a coolant in the summer and an ethylene glycol or propylene glycol coolant solution during the rest of the year will probably solve the problem. Commercial coolant solutions provide cooling, anti-freeze protection, corrosion inhibitors to protect the metals in the cooling system, and a lubricant for the water pump. When running water as a coolant for maximum heat transfer, a product that provides a corrosion inhibitor and water pump lubricant should be added to the water.

In terms of the relative heat transfer performance of ethylene glycol versus propylene glycol coolant bases, they are pretty much equal when mixes according to the manufacturers’ recommendations, usually a 50/50 water to glycol mix. Ethylene glycol coolant solutions provide slightly higher heat transfer performance over propylene glycol solutions at low coolant flow rates.


Jason
 

·
Registered
Joined
·
1,233 Posts
How hot was the engine? You might also want to check whether the radiator and external coolant hoses are also hot. If they are not as hot, then it is possible that there is a circulation problem (weak water pump or stuck thermostat?).

It is not a problem to use plain water to test the system temporary but keep in mind that water boils at 212F but a pressurized 50/50 antifreeze mix system will withstand above 250F+ without boiling. The heat distribution of the engine is not that even and there are spots hotter than others. The combustion chambers easily see a few hundred F. With the coolant (or water) at the other side of the chamber, the cylinder wall has a temperature gradient dropping from several hundred degrees to the temperature of the coolant through the thickness of the cylinder wall. If the coolant is not circulating well (i.e. stagnant) , the temperature will certainly rise rapidly. When the coolant temperature rise to the coolant boiling point, the water content (after it has absorbed enough latent heat :)) will start to change phase into gaseous form (i.e. steam). Now, Michael, any possibility that the bubble that you were seeing was actually steam?
 

·
Premium Member
Joined
·
15,537 Posts
Therein lies the problem, as was stated, water in a standard pressurized cooling system (about 15 psi above ambient) will boil at 220-230F, depending on impurities, while a proper antifreeze/water mix of 50-60% antifreeze will not boil before about 250-265F at the same 15 psi. Besides the other problems of not having anticorrosion additives and lubricants, you would have to be absolutely sure that your engine coolant temp anywhere in the system will not rise above about 220F if you use plain water for anything long term, else you can get steam. Not good.

Since modern cooling systems are designed and sized for the above characteristics, ie, they often are set to run hotter than in the past, mostly because the engine is more efficient at higher temps, giving a gain in fuel economy, and also because the cooling system can be reduced in size, therefore saving a little weight, also gaining fuel economy.
 

·
Registered
Joined
·
272 Posts
I have tested the 164 cooling systems for short periods with water and it does not boil over. I am not suggesting that you run your car year round without antifreeze, but water is good enough for testing. I don't now whether the water will boil if the system is not holding pressure.
Besides the water pump and head gasket, a blockage of water flow could cause boiling leading to bubbles (if the bubbles are indeed steam and not air or combustion gas). Could mice have gotten into the cooling system while the engine was in storage and try to build a nest? I have seen mice build a nest in the top tank of a spider radiator.
 

·
Premium Member
Joined
·
15,537 Posts
Take my word for it, if there is no pressure, ie, no coolant cap, the water may boil as the engine can easily reach 212F (boiling point of water) in the 164, locally maybe even higher.

If you retorque the front head, you should really take the time to retorque the rear as well, as that one is usually not as tight as the front because of exactly of what you said you would do, ie, skip the rear because it is a little more difficult to access.

With my 91S, I found exactly that situation, the rear was not torqued up anywhere near as much as the front, and the front was low already. Retorquing both heads eliminated the minor oil contamination and slight bubbling I had found.

Doesn't mean that is your problem, but it's a good start, and something which should be done in any case, regardless. Good luck.
 

·
Registered
Joined
·
503 Posts
word of caution - i had a bad coolant cap on my LS, replaced, cooling system worked perfectly but the presure blew the heater core, and thus i had a head job to do......i am sure the heater core was on its way out but i had no idea it was going.....
 
1 - 20 of 37 Posts
Top