Head Gasket Failure & Replacement Saga

[Bob Bidonde]

At 955 miles on a rebuilt Model "B" engine with a Synder's 6:1 cylinder head, its Best 573 Head Gasket failed between Cylinders 3 & 4 during a parade. At the time, the spark lever was advanced to about 15 degrees (30 degrees BFTDC). Running 87 Octane gasoline with GAV 1/8th turn open, TT10 Motorcraft Spark Plugs & 12V negative ground ignition.


Thinking about the cause of failure, I suspect: (1) Lugging the engine with too much spark advance during the parade; (2) Generally driving with too much spark advance (20 - 25 degrees ignition); (3) Loosening Cylinder Head Nut 8 to remove the distributor and not re-torque enough. The engine did not overheat since its rebuilding.



Replaced the failed gasket with a Best 509 Kevlar Gasket without sealant. Initially torqued head to 55 lbs-ft in steps of 30, 45 and 55. Overnight the gasket compressed without any engine run, so I re-torqued the head to 55 lbs-ft.



Timed the ignition at TDC #1 with the spark lever down 5 notches to regain full travel of the lever with the 6:1 head. With water as a coolant, I ran the car 13 miles, drained the coolant to add anti-freeze and re-torqued the head while it was hot to 60 lbs-ft. No leaks while engine is running.


More to follow as I put more miles on the car.

[norcalal]

Retorque head when engine is cold, and check two or three more times after it has gotten some miles on it and retorque again cold.

[alexiskai]

Rotten luck this year, Bob, sorry to hear you’re having more problems.

[Synchro909]

I NEVER put a head gasket on without sealer. Nuf said.

[Mike Peters]

Bob: Have you ever had your 6 to 1 Snyders head checked for flat? I personally have never had a Snyders head, but several club members have, and they tell me that the Snyders head straight out of the box can be as much as .012" out of flat at the head gasket mating surface, and had to be taken to a machine shop to be resurfaced. Installing the new head straight out of the box without checking has resulted in the same problems that you describe. Anyway, that's the word from the Rockford A's.

[alexiskai]

Sponsored Links (Register now to hide all advertisements)

Originally Posted by Synchro909

I NEVER put a head gasket on without sealer. Nuf said.

The 509G seals well dry, if the mating surface is clean and flat. AER did theirs that way for years.

[eagle]

I think the best way to keep the head gasket sealed is to check the torque of the head bolts/nuts often, for a year or so, then once a year after that. A head that is a bit out of spec for flatness especially needs to be kept tight.

[nkaminar]

I use the graphite Best gasket with the copper coat spray adhesive. That has worked the best for me.

Regarding flatness, you can smear some blueing on a sheet of flat glass and then rub it on the head. The high spots will pick up the blueing and the low spots will not. If the glass rocks then you are in real trouble and will have to have it decked.

If the block or head has small divots, you can fill these with JB Weld after a through cleaning with acetone or lacquer thinner. Dress the cured JB Weld flat with some fine wet and dry paper mounted on a flat object.

[alexiskai]

Originally Posted by nkaminar

If the block or head has small divots, you can fill these with JB Weld after a through cleaning with acetone or lacquer thinner.

I think it was Brent Terry who first clued me in that JB Weld ExtremeHeat is good for this purpose, and easier to work with because it’s a metallic putty rather than an epoxy.

[Bruce of MN]

Originally Posted by Mike Peters

Bob: Have you ever had your 6 to 1 Snyders head checked for flat? I personally have never had a Snyders head, but several club members have, and they tell me that the Snyders head straight out of the box can be as much as .012" out of flat at the head gasket mating surface, and had to be taken to a machine shop to be resurfaced. Installing the new head straight out of the box without checking has resulted in the same problems that you describe. Anyway, that's the word from the Rockford A's.

I must have gotten lucky. I took my 6:1 Snyders head to the machine shop to be checked and he said that it was surprisingly flat.

[Bob Bidonde]

I checked the head and block with a straight edge, feeler gauge and light, and found both to be flat within a few thousandths of an inch, good enough to use.

As for the torque, I estimate the clamping force as follows:
The attached chart estimates clamping forces vs torque. Looking at 7/16-20 Grade 5, 35 lbs-ft produces 4900 lbs clamp-up. Assuming the torque relationship to the clamp-up is linear, then 60 lbs-ft / 35 lbs-ft = 1.7, so 1.7 x 4900 lbs = 8330 lbs clamp-up.

Raising the torque to 65 lbs-ft increases the clamp-up by 770 lbs.

Katy! Good catch re the stud size. I fixed my estimate for 7/16-20 Studs.

[alexiskai]

Increasing torque doesn't do anything beyond the point needed to resist the combustion forces. It only increases the risk that the bolted joint fails. The failure point is likely to be the block itself. Those threads will shear off before the stud does.

[katy]

Looking at 5/16-24 Grade 5, 19 lbs-ft produces 3700 lbs clamp-up. Assuming the torque relationship to the clamp-up is linear, then 60 lbs-ft / 19 lbs-ft = 3.16, so 3.16 x 3700 lbs = 11,685 lbs clamp-up.

Your engine uses 5/16-24 head studs?

[Dave in MN]

Replaced the failed gasket with a Best 509 Kevlar Gasket without sealant. Initially torqued head to 55 lbs-ft in steps of 30, 45 and 55. Overnight the gasket compressed without any engine run, so I re-torqued the head to 55 lbs-ft.

Timed the ignition at TDC #1 with the spark lever down 5 notches to regain full travel of the lever with the 6:1 head. With water as a coolant, I ran the car 13 miles, drained the coolant to add anti-freeze and re-torqued the head while it was hot to 60 lbs-ft. No leaks while engine is running.


More to follow as I put more miles on the car.[/QUOTE]

Bob,
Based on my experience breaking engines in on a dyno, I believe there may be "more to follow."

The Kevlar/GraphTite (Best trademark) gaskets are prone to leak during the first couple heat/cool cycles. They often weep a bit of water until they fully seal. I find it takes an average of 3 heat/cool cycles for the gasket to bond to the surfaces. Changing to antifreeze with only one heat cycle may cause or allow the coolant to seep between the surface of the gasket and either the head or block. Once antifreeze gets between, the sealant imbedded in the gasket will not properly bond to the metal and then it leaks when put under load. I have proved this to be true multiple times with the dyno loading.

After observing a few failures when starting to use the Kevlar or Graphite gaskets, I now only run my new engines with clear water when on the dyno. I also ask my customers to use clear water for a couple hundred miles before changing to antifreeze. Since starting this practice, we have not had report of any failures. This procedure works for me.

Please keep this information in mind should you experience another failure.
Good Day!

[alexiskai]

Great advice Dave, I’ll do that on my next overhaul.

[Bob Bidonde]

I did a second head torquing today while the engine was hot. Some of the studs at the center of the head needed adjustment. I plan to do more re-torquing while the engine is hot until all of the studs stay at 60 lbs-ft.

[BRENT in 10-uh-C]

Bob, without performing a deck thickness test either with a sonic tester or by measuring the deck to pain rail with a micrometer, I would be leery of torqueing past 55#. The cast will actually begin pulling the threads and the area unsupported by the head upwards which can/will cause the gasket to leak. Torqueing head studs in a Model-A/B (-especially a Model-B) is definitely an area where if a little is good, then lots is better does not apply. Also, I would refrain from torqueing when hot and allow the entire package to return to ambient temperature.

I completely concur with Dave's comments in Post #14 above as those are basically my experiences too.

[1crosscut]

Like Dave I think coolant / water getting in between the gasket and the head / block shortly after putting on a new gasket can be the cause of a head gasket failure.

This is how I go about it.

Using the Best brand graphite gasket I start the engine without any water/coolant and let it run at just above idle for about a minute.
While doing this I hold my hand on the head at the front right (passenger side) so I can feel the temperature as it starts to warm.
I find that that is the area of the head that is the first to warm up. Once I feel heat start to come up in that area I shut off the engine and wait to allow the heat to saturate through the head. Keep feeling around the head after it is shut off and you can get a good feel of how the heat moves through the head.
After it has had time for the heat to saturate through the head I torque it warm.
Notice I say warm... I don't run the engine until it gets hot.
Then after this I allow it to cool completely and torque it again cold.
Then I repeat the process at least a couple more times, more if necessary until the torque settles in at 55 lbs.

After I'm satisfied with it staying at 55 lbs. I fill the cooling system with water, place a dial probe type thermometer into the top of the radiator and cover the front of the radiator with a piece of cardboard.
Then I start the engine and watch the thermometer like a hawk and bring the temperature up to 190 degrees. Then I slide the cardboard down a bit as needed to keep the temperature at 190 for several minutes. After that I remove the cardboard and allow it to idle for a little longer to bring the temperature down a bit and then shut off the engine.
Then I torque it hot, allow it to cool completely and torque again cold.
I repeat at least once and sometimes more until I'm satisfied that the torque is staying rock solid at 55 lbs.

Not a fast process as it usually takes two or three days for the entire process but it has been working well for me.

Dave

[Bob Bidonde]

Here are some photos of the head gasket failure & replacement:

[Bob Bidonde]

It would be a significant aide to our hobby if someone would take a scrap cylinder block, insert some studs, torque them to failure, measure the force when failure occurs and describe what failed.

[alexiskai]

Originally Posted by Bob Bidonde

It would be a significant aide to our hobby if someone would take a scrap cylinder block, insert some studs, torque them to failure, measure the force when failure occurs and describe what failed.

I've seen Brent say in the past that over-torquing either results in sheared threads, if the threads are already corroded or marginal, or more likely it will distort the block, pulling that entire area of the block upward toward the head.

[johnneilson]

Bob,

I do some testing on the blocks before I do anything else, pressure test for cracks and then do a head stud thread test.
For the test, I use an ARP stud about 3" long and a wrist pin that the ID is close to the stud dia. Place the stud full depth into the block without forcing it, sometimes the threads need a cleaning. Place the WP over the stud and add washer and nut. Then I take the torque up to 75#. If the threads hold, all is good, usually if it is going to fail the stud just pulls out.
Once a block passes that test, it goes out for blanchard grind TOP and Bottom.

When the decks are being done, it usually shows the un supported sections pulled up away from the edges and the bores, and random movement around the valves.

It is interesting though and this is just an observation, that contributing factors into gasket failures come from the gasket compressibility, and the rigidity of the head.
I have noticed when someone runs two head gaskets the deck is deformed more than a single. In addition, we can run a solid copper gasket or no gasket and copper wires without failure. These in supercharged motors and NA high CR 13:1+
All of this using 55ft# torque. no more.

Best, John

[Ed in Maine]

I hate replacing head gaskets. It seems that I always get to do the job twice. The above comments are excellent and I will make note of them. There has been one comment about coolant seepage between the block and the graphite gasket that could prevent a good seal to be made if the coolant leaves a residue on the block surface. In the future I will use just water. Further, it may make sense to install the graphite gasket and torque the head to 55 ft-lbs and then let it sit for a week. Retorque as necessary to maintain 55 ft.-lbs but leave the engine dry for a period of time. If what you guys are saying, a waiting period before filling the coolant, should help the "bonding" process between the block and the gasket and the head. Ed

[Flathead]

I ran the engine briefly with NO water or coolant.

[Synchro909]

Originally Posted by alexiskai

the 509g seals well dry, if the mating surface is clean and flat. Aer did theirs that way for years.

aer???

[alexiskai]

Sponsored Links (Register now to hide all advertisements)

Originally Posted by Synchro909

aer???

Antique Engine Rebuilding (formerly of Illinois, now of Indiana).

[Y-Blockhead]

I have always followed Secrets of Speed Society recommendations and never have a issue.

[Bruce of MN]

One more vote for no antifreeze here. I followed the advice and the gasket is doing well with a 6:1 head,

https://www.fordbarn.com/forum/showt...te#post1145107

[cw253748]

I forgot where I read it but when I retorque my head gasket I always loosen the nut 1/8 of a turn and then torque to 55 lbs. That way you get a truer torque because it may take more than 55 lbs. of torque to over come the friction of the nut to head

[GPierce]

From the FORD GARAGE.COM

https://www.fordgarage.com/pages/bru...structions.htm


When re-torquing, it is usually not normal practice to back off nuts or bolts. However, on a Ford flathead, some of the cylinder head nuts can develop an increase or higher static friction as compared to the static friction of some of the other cylinder head nuts or a resistance to movement after they have been previously tightened or torqued. Or in other words, once tightened they sometimes tend to stick and the amount of torque required to get the nut to move is a little greater than the torque value that was intended to be applied.

When the torque is applied by the torque wrench in the standard way, the force must be great enough to break the nut free and move tighter. In the process, some of the torque is wasted overcoming the higher friction and the wrench may not apply the correct torque amount as set or viewed on the wrench. The result is the applied torque can be incorrect. And, if the increase in static friction is high enough, the nut may not break free at all at the torque amount set on the wrench.

In other words, the installer can be fooled into thinking that the nut is at the value set on the wrench whereas if the nut was not sticking it would turn tighter, as it should when the gasket has not completed its compression process. This possible sticking of the nuts can happen to a degree even with lubed threads after a little time goes by and a few hot/cold cycles. Talk about possible uneven clamp from nut to nut, leaky gaskets and premature gasket failure!! …. Anybody ought to be able to understand it.

If it could be guaranteed that the nuts would never stick, then backing them off would be unnecessary. However, to avoid the problem and to assure accurate re-torques and and even clamp from nut to nut, the nuts should be backed off, or cracked back slightly before being re-torqued and then retightened or re-torqued immediately.

Backing each nut off 1/8 turn does not loosen them in the true meaning of the word, loose. The procedure merely moves the nut only 45 degrees of rotation and just enough to overcome the stick that may or may not have occurred between the nut and stud threads and including the stud boss and nut face.

Since the torque readings are dependent upon the coefficients of friction present under the nut face and in the threads, there must be a smooth, uninterrupted rotation to have accuracy when the re-torque is applied. And last but not least, nuts have been observed that appeared to be tight at the final torque amount, when re-torqued and NOT backed off, only to turn as much as 2/3 to nearly a full turn farther when backed off and re-torqued at the same torque amount!

Apart from gasket surfaces being too far out of flat, insufficient final torque amounts, worn or low strength studs and nuts, etc.; many a leaking or seeping gasket and many a premature gasket failure on Model A’s are caused by the uneven clamp that is possible from nut to nut due to improper re-torques.

Larry Brumfield
July 2016

[katy]

Originally Posted by cw253748

I forgot where I read it but when I retorque my head gasket I always loosen the nut 1/8 of a turn and then torque to 55 lbs. That way you get a truer torque because it may take more than 55 lbs. of torque to over come the friction of the nut to head

True

[Cape Codder]

Mike Peters- Thanks for your post to this question. I do have a question about your statement below in part:
"but several club members have, and they tell me that the Snyder's head straight out of the box can be as much as .012" out of flat at the head gasket mating surface, and had to be taken to a machine shop to be resurfaced".

Do you know if any of the several members ever contacted Snyder's about the problem? I ask this because I feel that a head should be delivered FLAT or Snyder's should be responsible for the cost of the resurfacing!

Several years ago I had my motor rebuilt and at that time I decided to purchase a NEW Head from Snyder's. Upon receiving it I found that the head was 1/8" - 3/16" too short for the block, so I called Snyder's to inform them of the problem. I was told the best they could do was send me another head! Upon receiving the second head I found the same problem!

In the end I returned BOTH heads to Snyder's and was given a full refund including postage that I had paid to get them. I have always wondered if the problem had ever been resolved?

I understand the problems vendors are having BUT if no one tells them of problems they will just keep on buying the same product without any corrective action!

Lastly, I must say I still order parts from Snyder's because even with the disappointment in the above product they are a Great company!

Originally Posted by Mike Peters

Bob: Have you ever had your 6 to 1 Snyders head checked for flat? I personally have never had a Snyders head, but several club members have, and they tell me that the Snyders head straight out of the box can be as much as .012" out of flat at the head gasket mating surface, and had to be taken to a machine shop to be resurfaced. Installing the new head straight out of the box without checking has resulted in the same problems that you describe. Anyway, that's the word from the Rockford A's.

[alexiskai]

Originally Posted by Cape Codder

Several years ago I had my motor rebuilt and at that time I decided to purchase a NEW Head from Snyder's. Upon receiving it I found that the head was 1/8" - 3/16" too short for the block, so I called Snyder's to inform them of the problem. I was told the best they could do was send me another head! Upon receiving the second head I found the same problem!

In the end I returned BOTH heads to Snyder's and was given a full refund including postage that I had paid to get them. I have always wondered if the problem had ever been resolved?

This is a "known bug" with the Snyder's 5.5:1 and 6:1 heads. What I've been told is that the heads were designed by making a mold from a stock head and then altering the combustion chamber. When you make a mold from a cast iron object and then use that mold to duplicate the object, the copy will be very slightly smaller than the original because the iron shrinks as it cools. So all the Snyder's heads are very slightly shorter on the outside than the stock heads and the block. This is not a problem the new Burtz heads have, since they were designed by laser-scanning an original head into a computer.

[Fhane]

Makes me wonder who is doing the machine work on the Snyder heads. Most installers would not check the dimensions on a head The would assume the head is good. If I remember correctly, all of the new heads we installed in the last 8 years or so had to be resurfaced to make true. Some of the newer customer heads we replaced were out of whack.

[Steves46]

Since I recently blew a head gasket using a Snyder’s Head I am taking it have the flatness checked. I will say I was running it for 2 years before it blew so maybe mine will be good. Will share the results.

[Y-Blockhead]

Originally Posted by Y-Blockhead

I have always followed Secrets of Speed Society recommendations and never have a issue.

Altho I have never had a problem not using a sealant, it has been pointed out to me by Vince Falter (Ford Garage http://www.fordgarage.com) that BEST recommends using a sealant with their GraphTite gaskets so I will be following their recommendation from now on. Best to go with manufacturer.
[IMG]https://www.fordbarn.com/forum/attachment.php?

[/url]

[alexiskai]

Originally Posted by Y-Blockhead

BEST recommends using a sealant with their GraphTite gaskets so I will be following their recommendation from now on.

I suspect the difference in recommendations is due to the fact that the rebuilders on this forum have control over the mating surfaces. They can ensure that the head and block are flat, with the proper finish, so no sealant is needed. Best’s recommendations need to account for the hobbyist with warped surfaces who just blew a head gasket and wants to get back on the road.

[BrianH]

Interesting info about backing off the head nuts before retorquing.

My Snyder head 6.1 had to have right at .006 taken off to get flat. I also had my block surfaced flat so hopefully I won't have a head gasket issue any time soon.

[Steves46]

Originally Posted by Steves46

Since I recently blew a head gasket using a Snyder’s Head I am taking it have the flatness checked. I will say I was running it for 2 years before it blew so maybe mine will be good. Will share the results.

The flatness was good so it's all clean an ready installation. However I had to order a replacement stud and darn it all it only makes one turn by hand and nothing more. Sure hate to try and force it in so one of my local Model A buddies has a bunch in his stash and will try one of those.