Interesting Problem

[Matthew E. Herd]

Gentlemen,

I have observed the following phenomenon concerning my '76 Midget. I blew the head gasket in ?October? autocrossing (won that day, though!). The gasket burned through between cyl. 1 and 2. Turns out one of the guides needed replacing, the head needed a skim cut, and valve seals needed to be installed.

At any rate, I reinstalled the head (after it sat a good while) and fired it up a few weeks ago. I posted about having a problem with two head studs at the time of install. Since then I may have run the engine 6 or so times, driving perhaps 50 miles in total. Power definitely seemed to improve, and the smoking on startup went completely away (no more oil past the guides).

However, I previously sealed my cooling system because the pressure reservoir didn't fit with the custom airbox. However, the pressure that I saw recently seemed to build to the point where it'd come out of the seams, so I installed an inline filler neck/pressure cap. The fluid would bleed off a bit as it expanded, then reach a steady state (with some air in the system). However, I still had trouble with coolant leaks. I tightened every hose clamp, etc, and still the coolant would leak. Finally, while driving the car, it started to overheat. It had overheated the first time I took it out (highly unusual for this car), and it had no coolant in the system. At the time, I thought nothing of it and filled it up again. When it started to get hot the second time, I just figured too much coolant had bled out past the cap. However, as I drove it the remaining 7 miles home, it was leaking again (coolant was getting onto my windshield, which is how I knew it was leaking). It was also overheating again, within only 15 minutes and 7 miles or so.

While I still wasn't able to find the source of the leak (last week) I could hear hissing that sounded like it came from either the driver's side radiator hose (at the radiator) or the radiator endcap itself. I started to investigate today, and found that the coolant was halfway up the hose after having sat for a week. The car had overheated, but no coolant leaked from the hose or the radiator beyond a certain level. So, I opted to fill the system with water and let it idle (pressure cap off). I could see that the system took an enormous amount of water.

I ran the engine, and saw a bit of oil residue in the water (very slight). I assumed that this was from the filler neck, which was oiled from the manufacturer. However, another thing I noticed was that the water level dropped slightly as the car ran. Once the thermostat opened, the level may have dropped slightly faster. Finally, I heard a sound that was not unlike sheet metal flexing as wind hit it. Since I was inside a pole barn constructed primarily of corrugated sheet metal, I thought it was just the wind. However, then I heard it again (definitely from the engine) and immediately shut the engine off.

At this point, I decided to do a compression test, as I figured something was wrong. I worked on getting the alternator belt off (always a pain) and got all 4 plugs out. As I worked, I noticed that the coolant level dropped. I removed the valve cover and, sure enough, there was white foam all over the cover. I also checked the dipstick and that was foamy too. However, after I performed the compression test, I found that the readings were:

Cyl 1: 236
Cyl 2: 240
Cyl 3: 235
Cyl 4: 232

I plan on re-testing, in case coolant improved the seal somehow (if it's in the cylinders) once it has the opportunity to leak out. However, I suspect those numbers are genuine. I am trying not to skew the outcome of potential opinions, but I realized in retrospect that the static level of coolant is probably approximately at the level of the head/block interface.

As an aside, I noticed that when the vehicle started to overheat (or at least get much warmer than it's usual position, well left of "N", it never really overheated terminally) it would stumble on application of throttle (possibly a rich condition). This could simply be indicative of increased intake manifold temperature (and therefore the requirement for a leaner mix).

Let me know thoughts on this one. I just got done refurbishing the head after blowing the gasket, so the replacement was brand new, and everything looked to be in good order as I put it all together.

 

[RobSelina]

What's the pressure setting on that inline filler-neck/pressure cap you installed?

Just throwing some stuff out there, but how old is your water pump? Are there any restrictions on the bypass line for the pump? I'm assuming there's a problem with how the water pump regulates pressure in the system....

 

[Matthew E. Herd]

The pressure cap is 16 psi. When the system is open and the engine isn't running, the level was seen to drop.

The water pump is no less than 8 years old (the car sat 3 years before I got it, 4+ years ago). It seems to function.

BTW, I don't detect any external leaks at all.

 

[aeronca65t]

You have much higher compression readings than I do. My engine presently has 10:1 pistons and 0.115" skimmed off the head.
Hmmmm?

Maybe you have gotten water in the compression tester (I'm only guessing....I've never heard of this).

Otherwise, the foamy oil indicates water in the oil...a giveaway for a blown headgasket or related problems (such as warped head or cracked block).

Just out of curiosity, did you add cold water or coolant to the engine while it was real hot?

 

[Morris]

I am speculating wildly here, so please don't hold it against me if I say something stupid.Bbut in my experience, your cooling system really needs to "breath." If the only avenue for your coolant to expand is through a 16 psi cap, you have probably over pressurized your system such that it has blown through your head gasket.

Also, if you run without a catch can or surge pot of some kind, you will eventually overheat no matter what state the rest of your cooling system is in. Coolant will blow by the pressure cap and the system will have no means to recapture this coolant. Eventually, you will end up empty amd overheated.

 

[DrEntropy]

Yep, those comp. numbers are REALLY high. And the "chocolate mousse" is indicative of a head gasket failure.

 

[Matthew E. Herd]

well, I suspect, like Morris, that the system overpressurized somehow (blowing the gasket). While I don't believe that this SHOULD have happened, I think it did.

I was trying not to skew the results of the question, so I didn't give my diagnosis until now. Nial, what readings do you generally get? My CR was (prior to skimming the head last time ... perhaps .120" removed) 10.5:1. Now it should be slightly more. However, 14.7*10.5 = 154. This really makes no sense. I'll check the gauge to make sure it's not a 1 ... instead of a 2, but you're right, the numbers seem strangely high.

as to water, I did add some water at air temp (about 70 degrees) the second time it overheated from coolant loss. Morrs, as an aside, if you totally seal the system (assuming it doesn't eventually burst) it won't leak. I ran it this way for almost 3000 miles (until the head gasket blew the first time). It barely lost coolant at all, never taking more than 8 oz when I would refill it. Assuming the system is truly sealed, and everything can take the pressure, this is actually a fine way to do business!

I should also note that "overheating" was actually more like 30-45 degrees from the horizontal on the right hand side of teh gauge, as opposed to in the red and smelling coolant. It probably never exceeded 190-200 F.

 

[DrEntropy]

That shouldn't have pushed the gasket aside, then. And I'd expect compression with your description to be in the 150-ish range. Vexing to have to sit HERE while you and the car are over THERE, IYSWIM. /ubbthreads/images/graemlins/frown.gif

 

[Dave Russell]

Hi Matthew,
I really suspect that there are two problems, though possibly related.
1- The compression readings are a bit high.
"Cyl 1: 236
Cyl 2: 240
Cyl 3: 235
Cyl 4: 232"
High readings are only likely if the compression gage is faulty or there is liquid in the cylinders, as Nial said.

Quote: "However, 14.7*10.5 = 154."
This theory would only be correct if the intake valve closed at BDC. In reality, the intake valve closes some 30 to 70 degrees after BDC so the rated compression ratio is never acheived at low rpm. The compression pressure IS increased somewhat by mixture heating during compression which would partially compensate for the late intake valve closing, but not completely.

2- You are losing coolant. Combustion pressure is leaking into the cooling system. The previously mentioned head stud problem may be a factor.
Quote: "I posted about having a problem with two head studs at the time of install."

Quote: "well, I suspect, like Morris, that the system overpressurized somehow (blowing the gasket). While I don't believe that this SHOULD have happened, I think it did."

Quote: "That shouldn't have pushed the gasket aside, then. And I'd expect compression with your description to be in the 150-ish range. Vexing to have to sit HERE while you and the car are over THERE, IYSWIM."

As was said, coolant pressure would not disturb a "good" gasket seal.

I think your best bet at this time is to do a cylinder leak down test. This will tell for sure if combustion pressure is getting into the coolant & which cylinders are involved. It will also tell about the condition of rings, & valves. I can post a "DIY" about leak down testing if you wish.

Good luck,
D

 

[Matthew E. Herd]

No, that's ok Dave, I'm quite familiar with the process, I just haven't purchased the tool YET. I don't know for sure if there's coolant in the cylinder, and I'm reasonably sure there weren't compression leaks to the cooling system, but I can't be sure. I heard a slight hiss in cyls 2 and 4, but figured that was normal (it will leak out eventually). I didn't pay attention to the valve positions when doing the test, so it's possible they were closed on those cylinders (randomly) and not the other two.

Also, I'm quite familiar with the fact that the pressures shouldn't even be as high as atmosphere*CR psi, but was using that as a reality check.

 

[aeronca65t]

I think I get around 210 psi.

I recall that you have a "mild" cam in your's (right?). Generally, the hotter cams give a slight decrease in psi compared to a stock cam.

A leakdown test is the way to go. There's also a tool for checking leakdown on the cooling system.
You need a calibrated coolant temp gauge.....those stock ones are horrid.
I bought a mechanical temp gauge in AutoZone for about $20.
For an autocross car, I would "gut" the thermostat (remove the spring and valve) and just leave the brass ring to act as a blanking plate.

 

[RobSelina]

[ QUOTE ]
I think I get around 210 psi.


[/ QUOTE ]

that's almost 1:14.3 compression. You sure about that?

 

[Matthew E. Herd]

Trouble is, it's also a street car. I figure I'll do a leakdown test first (gotta buy that tool), and perhaps also test cooling system pressure (I can borrow this one). I figure I will have to pull the head, and probably check that the block isn't warped, but I rather doubt it. We'll see. Now to order the toy ...

As to pressures, I'm not sure why they're so high, as they defy logic. However, if I'm not the only one with higher than theoretical pressures, I don't feel so bad. /ubbthreads/images/graemlins/smile.gif At least I can brag about my amazing consistency! Total spread is like 8 psi!

 

[Dave Russell]

It's pretty easy to make a leakdown tester. A small pressure regulator, a zero to 80 psi gage, few fittings, quick disconnect coupler, sparkplug hole adapter & an .040" orifice. This can be a hole drilled through one of the fittings which is soldered closed. See attached pic.
D

 

[Bugeye58]

Dave using a higher pressure gauge, and 100 PSI as the reference pressure, makes the math easier! /ubbthreads/images/graemlins/grin.gif
Jeff

 

[Dave Russell]

You're right, it would. I think 80 psi is a quasi-standard. Just trying to stay close to other gages so comparison would be easier. If you were to use 100 psi, the restrictor would need to be about .036" diameter. Most drill sets don't go below .040". I do have a set of 60 to 80 number drills, but they are not very common. If you are so inclined, the gage face can be repainted & marked backward to read actual leakage %.
D

 

[Matthew E. Herd]

I already ordered one, as I don't really have the time to make one (and it wasn't terribly expensive at $94.00 for a Longacre Racing item). Also, I can use it on a semi-regular basis, so it's not like it's a one-time use item.

 

[Bugeye58]

[ QUOTE ]
If you were to use 100 psi, the restrictor would need to be about .036" diameter. D

[/ QUOTE ]
Dave, you're absolutely right. I too, have number drills down to #80, as well as reamers as small as .032", but, I got lazy and bought my leakdown tester.
Jeff

 

[DrEntropy]

Got the drill bits. But my reamer collection is ~obviously~ inadequate now...

Must get MORE TOOLS! Thanks fer th' "heads-up" Jeff! /ubbthreads/images/graemlins/jester.gif

...and when SWMBO asks why I -NEED- reams down to tirty-five thousandths, I'll tell her: "Well, Jeff has 'em!" --- and send her YOUR direction. /ubbthreads/images/graemlins/devilgrin.gif

 

[Matthew E. Herd]

Update to the problem ...

I removed the head after doing a cooling system pressure test (sure enough, the sound of trickling could be heard in the engine). I examined the gasket and found that it did not fully compress around one of the studs (one which failed during installation). I also found that two more of the studs had actually also galled the threads, and were junk.

Upon reinstallation, I made sure to lubricate the threads, although the Haynes manual does not specify that this is required. I then torqued them to 40 ft-lbs (the spec is 46, but I didn't have the manual in front of me at the time). I let the car sit for a week, as I didn't have any more time last weekend. Yesterday, I went to tighten the studs to their final torque (46 ft-lbs). I had used the original studs as replacements for the failed (new) studs. Not one of the old studs exhibited any problems, and did not fail under tightening. I had not yet taken the matter up with Moss, who supplied these parts, but will tomorrow.

At any rate, while tightening the studs, two more of them suddenly became loose, with the torque dropping off sharply. I removed one, and it was very difficult to turn out of the block. I examined it, and there was evidence of necking just above the thread root on the threads which would go into the block. However, the threads were undamaged. I could not get the replacement stud to thread in. Currently, one stud is suspect, and the other is removed, but won't thread in. I imagine that there is some swarf in the hole, and I don't have air power available at the location of the vehicle (it's not in my 'shop' at the moment). Obviously, I can't inspect the hole because the head is installed. I want some opinions on the appropriate course. I figure I'll probably have to pull the head again (3rd time). I don't have taps or thread inserts. I have an engine guy who is nearby, and could tow it to either him or my usual mechanic if necessary.