MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGA - Cylinder Test Results - Help!

Last weekend while driving in town, white smoke started coming from cockpit. Temp was steady at 220 degrees, the smoke was coming from three places, the front carb, the rear seam between the cylinder head and block, and the next-to-last rear mounting bolt for the manifolds. The BBS gurus suggest looking for a cracked head or a blown head gasket.

Tonight, I ran full compression tests:

Cylinder Cold Hot Hot/Wet
1 133 125 145
2 120 110 120
3 115 100 125
4 135 125 135

I then disconnected each cylinder in turn and recorded the idle speed (measured by tachometer):

Disconnected Cylinder RPM
1 800
2 950
3 800
4 800

Idle speed with all cylinders connected is 1000 RPM.

The engine ran great while testing until the temp hit about 180, at which point white smoke started coming from the front carb. The smoke didn't come from the other locations, but I suspect that it would if I heated the engine a little more.


To me, it looks like the gasket might be ruptured between cylinders 2 and 3. Cylinder 2 seems to holding the engine back for some reason as well.

Thoughts on this? Is it time to tear off the cylinder head?
Rich

Rich,
I would definitely tear off the cylinder head. The fact that your hot compression test is lower than your cold indicates to me that you have some problem. The problem that I would be afraid of is that if you continue to run with a blown head gasket you run the risk of needing a complete engine rebuild. A blown head gasket can cause the rod bearings to go, and running while blowing hot gases across a blown gasket area can wire draw the seating surface, which is even worse. A head gasket doesn't seem to be to inconvenient when compared to these.
mike parker

Rich,

Just to confuse you and to contradict Mike, your numbers are not great, but I wouldn't pull the head yet. I'd retorque the head (loosen each nut at least one full turn, one at a time, in the helical pattern shown in the repair manual, lube it, torque it to 50-55lbs). Then I'd adjust the valves. Then I'd check the compression again. If compression is still bad as you posted, you could live with it for awhile and keep watching it, or do a leakdown test to find out more.

If you're not losing anti-freeze you can keep driving it the way you are--although something did make the temp go to 220. If you are losing antifreeze, you've got to find out where. It may not be the head. Heater valves, heater matrixes, hoses, radiators, etc all leak. Putting a pressure tester on the system at about 10-15 psi for about 10 minutes will help find external leaks like that. If you've got a cracked head or bad gasket, the leakdown test may be necessary.

Feel free to call me at the shop if you'd like more advice.

--Carl
www.eclecticmotorworks.com
Carl Heideman

Hi Carl. I thought that adding oil to a bolt/stud being torqued gave an incorrect reading, i.e., too much torque...?

George

G Goeppner

Wouldn't a higher compression while Hot/Wet indicate a ring problem ?

Duncan
DA MacFarlane

Torque readings for studs or bolts are just a way to measure the tension on that bolt or stud. In this case we measure the rotational force needed to turn the nut. If there is any additional friction on the threads or the surface the nut is seated against then that friction will appear as rotational force and give a false reading as to the actual tension on the stud. Torquing a dry fastener will usually result in the tension being less than desired because of the friction induced in the rotational axis. To eliminate this source of error, many high stress fasteners such as connecting rod bolts in racing applications are now tightened using a bolt stretch measurment where the fastener is tightened until the bolt stretches a prescribed amount.
Bill Young

Bill is correct and so is George. Where the specification is Torque the the parts should be run with no lubrication. If lubrication is added, then there is a chance that the stud can get stressed beyond its yield strength causing the stud to never get to the required measured torque. As Bill stated, modern engines and race engines are specified to rovide a measured stretch just under the yield rating of the stud. Dry or oiled, when using that method, the stress remains the same.

Rich, I am a little confused as to how the manifold mounting bolt can be releasing the white smoke. Since you hd the engine out and back in, perhaps your manifold mounting bolts are not properly torqued and the other leaks are also really leaks from the manifold being disguised as carb and head leaks?

Chuck
Chuck Schaefer

When using the high strength bolts and tightening to a measured stretch, it is a one shot deal. You can't re-use the bolts/studs. This is method is also used on modern diesel head bolts.
John H

Carl is correct, but this sounds like a non-saveable gasket. I will disagree with Carl on the 50-55 torque though. MGA torque is 40, unless the head studs are centerdrilled or stamped "22" on the end; Then it is 45-50, as are MGB. Generally any of these will take +5, but I would be wary of 55 on an A.
All of you who are going on about lubrication have some but not all of what you need to know. It is a complex subject, and anyone who wants it can have my diatribe on head bolts, torque, and lubrication by sending an email request to me. The short version is that ALL threads and nut bearing faces must have some kind of lubrication or they will gall and be useless. Sometimes the "lube" is just natural oxides. It can be deliberate oxides, other chemical coatings like parkerising, oil, wax, anti-seize, or platings of various kinds. They do affect torque/stretch, but not in ways that are easily translatable simply as "dry" or "lubricated". All specifications for "dry" tightening assume certain material and coating conditions, which ARE NOT applicable to other combinations.
FRM
FR Millmore

You know, torque is something that nobody ever agrees on.
Rich, after a quick read here, I agree with most all posts. All torque values are based on stretching the bolt or stud to just within the limits of elastic deformation, and as FRM states, lubrication of some sort is necessary to prevent both galling and improperly torqued fasteners.
Carl is correct that the head gasket may just need re-torqueing. I forgot about this, as it has been a long time since I have worked with older head gaskets. Old head gaskets were not really able to seal completely on one torque. The normal procedure on copper and flat metal head gaskets is to torque, heat up the engine and retorque. Some people feel that step torque sequences ( 40/45/50 in 3 steps, etc.) is the way to go, and I do this on all modern engines, with modern style gaskets and with the high stress style fasteners described above.
Additionally, there are posts in the archives about countersinking the thread area on the block to prevent improperly tightened gaskets. The idea is to remove the chance that the stud will pull the threads up from the surface of the head and impair the sealing ability of the head gasket.
As for me, I think it may pay to get FRM's little lesson.
mike parker

You cannot measure bolt stretch with head studs as you can only get to one end of the stud. You can do this with rod bolts.

The factory workshop manual, under general data, states "Torque Wrench Settings, Cylinder Head Nuts:" 50 lb ft. At the shop, we've found 55 is often helpful when there is weaping of water between cylinders 2+3.

Third, I agree with the comments about lube/no lube with NEW hardware. We always lube installed nuts because they are usually corroded. You can't measure torque accurately with corroded hardware. We always get questions about this in our seminars, etc., but it's the method I learned when I worked for John Twist (and he still uses) and I've had no trouble in 20+ years.

I hope this clarifies things,
Carl
Carl Heideman

Carl;
My MGA books have all gone missing, but it was about that time when I learned about the marked studs, which were an upgrade. My Magnette book shows 40 for both 1500 & 1622 . My early B book gives 45-50. Does your A book say anything about the marked studs? It's not just corroded studs/nuts, but any used ones, as use wears off the black oxide or maybe parkerising they came with. I am hesitant to advocate increasing torque, since so many people seem to report breaking studs. For my part, I have been doing this since the early 60s; I lube the threads and bearing faces with Lubriplate and spec torque, or anti-seize and low spec torque. I re-torque on first run, after a couple of days and several heat-cool cycles, after 500 miles or a couple of months, 1000 miles/1 year, and about once a year thereafter. Time and heat-cool cycles is more important than mileage. I have never stripped/broken a stud or had a gasket failure.
FRM
FR Millmore

Hi gents,

To add to the confusion, I did not pull the cylinder head off at all during the rebuild. The engine remained intact. However, I did pull off the manifolds, and I did not torque the nuts on the studs, (perhaps incorrectly) believing that the torque on the intake and exhaust manifolds wasn't all that important.

My amateur restorer theory is that the engine was running lean, and thus hot, which caused the head gasket to blow...

Rich

Rich,

I don't think that you have a cracked head and I would be very reluctant to "tear the head off" until I was sure what was going on. Your compression figures aren't great but close to within 10% of each other so whilst the engine has probably done a fair mileage since a rebuild is still servicable if you don't do huge annual mileages. I don't think that your head gasket is leaking across cylinders otherwise two adjacent cyls would have an identical reading and on the assumption that the coolant is remaining in the rad it isn't leaking anywhere else either. B series engines have a habit of being uneven on idle where removing plug leads has variable effect on idle speed

Now FRM is of course correct and advises against overtorquing the studs, be warned as they will break like a carrot. Check them by all means but don't exceed 40 ft/lbs. Check your manifold bolts as well to ensure that they are fully tightened, from memory torque is 12/15ft/lbs but check to be sure. Now 220 degrees is far too hot but if the pressure cap is doing its job properly the system should just about be OK but is your gauge accurate ? This could be a simple case of the thermostat needing replacement or the rad needing a good flush out or even recored. It's worth reverse flushing the whole system to ensure that its not sludged up before pulling what seems like a servicable engine apart all to no avail.
Iain MacKintosh

This thread was discussed between 13/09/2005 and 19/09/2005

MG MGA index

This thread is from the archive. The Live MG MGA BBS is active now.