Why do we still have head gaskets? Learn me.

I was reading one of my old books and stumbled on a part about the Miller/Offenhauser race engine which dominated Indy from the 1920's to the 1960's.

Like a lot of high-end engines from the 20's and 30's (Alfa-Romeo, Delage, etc.), it didn't have any head gasket. The head and block was one solid block. It allowed it to run a lot of boost.

It got me thinking: with more and more turbo engines being built and with modern machining techniques, why do engines still have head gasket? I don't see a good reason why. It would reduce the number of parts, improve reliability, help cooling the head, etc...

One of you guys must know why we still have head gaskets on engines. 

Sure, if its designed so you can install the valves and pistons through the bottom of the bore it works.

I think it's because the potential advantages aren't needed in most engines and the manufacturing downsides (harder to assemble and do QA) outweigh them. Modern engines generally don't blow head gaskets and cool just fine. Head gasket/bolt weakness isn't the boost bottleneck. And eventually customers may catch on to the repairability downsides.

They could make sense in high-end supercars or race engines.

Because the world is still not ready to accept the true pinnacle of engine design:  

It would probably be better to just do factory o-ringing of the block and head if you need stupid amounts of boost.

It's just mass production.  There are variances in the block and head surfaces that are rendered inconsequential via the use of head gaskets.  They could be eliminated for sure but it would require much closer tolerances.  Likely not worth it on anything short of a full court F1 press.

In reply to jharry3 :

Yep.  And look at the required maintenance on a '20's / '30's Bugatti.  Basically engine out and full tear down for everything.  I read about that in Tom Perkin's book:

https://www.amazon.com/Classic-Supercharged-Sports-Thomas-Perkins/dp/0961226803

Of course when you are a high rolling venture capitalist it doesn't matter.  

Cost and ease of manufacturing.  It's going to be a lot simpler to cast and machine a block that doesn't have an integral cylinder head.

flatlander937 said:

It would probably be better to just do factory o-ringing of the block and head if you need stupid amounts of boost.

That's what the big boost diesel guys do with little to no issues. even top fuel dragsters, which make somewhere around 8,000 to 10,000 hp, still use a somewhat tradition style head gasket

and top fuel dragsters run enough cylinder pressure to compress the rods enough that they have to throw them out after a couple runs due to them becoming too short. So the amount of boost an engine with the head and block as one piece could handle would probably be beyond any actual useful range even with today's technology.

In reply to ¯\_(ツ)_/¯ :

If oil consumption and emissions and torque were not concerns, yes.  That said I’d love to have a 3rd gen RX7 with a 13b rew.  No LS swap for me.

In theory, we could use electrical systems without fuses, too. However serviceability and practicality prevail.

edizzle89 said:

flatlander937 said:

It would probably be better to just do factory o-ringing of the block and head if you need stupid amounts of boost.

That's what the big boost diesel guys do with little to no issues. even top fuel dragsters, which make somewhere around 8,000 to 10,000 hp, still use a somewhat tradition style head gasket

 

 

and top fuel dragsters run enough cylinder pressure to compress the rods enough that they have to throw them out after a couple runs do to them becoming too short. So the amount of boost an engine with the head and block as one piece could handle would probably be beyond any actual useful range even with today's technology.

I have "fire ringed" a couple of Cummins including my own, its amazing how much better the head seals with them. 

I want one of those discarded top fuel rods!

Edit: actually two, one for me, one for my New Year's game box.

F1 engines screw the liner into the head. The Offy and the Leyland O.500 are the other two headless four-stroke engines of note. For the Offenhauser it was a great advantage and extended its competitive lifespan a decade. For the O.500 it was the one thing that didn't go wrong.

A headless two-stroke makes even more sense but O-ring seals outside a copper fire gasket work well enough. 

chaparral said:

F1 engines screw the liner into the head. The Offy and the Leyland O.500 are the other two headless four-stroke engines of note. For the Offenhauser it was a great advantage and extended its competitive lifespan a decade. For the O.500 it was the one thing that didn't go wrong.

A headless two-stroke makes even more sense but O-ring seals outside a copper fire gasket work well enough. 

Husqvarna did air cooled 2 strokes w/no head gaskets for years. They recommended you lap the head to the liner (by hand) with valve grinding compound whenever you did a top end job. That went away with water cooling. 

I think I should ask my question in another way. 

The way I see it, making a modern DOHC engine with the head and block as a single piece instead of two separate pieces means:

1 - The only machining that would be harder would be machining the valve seats. But if they could accomplish that in the 20's, I really don't see this as a problem. But you also get rid of two very large machined surfaces. The rest of the machining is exactly the same.

2 -  Having the head and block as a single unit makes it much stiffer and stronger. You can use that to reduce weight or increase boost depending on your goals.

3 - You remove an edge in the combustion chamber which reduces detonation. 

4 - You remove a failure point. While modern engines have less head-gasket problems, there are still problems *cough Subaru cough*. In fact, I would argue that head gasket problems are still the number one problem on engines in my experience.

5 - You remove an assembly operation and a bunch of parts that can all break (head gasket and bolts/studs)

The only real problem that I can see in this is the casting of the block/head. If someone has some experience in casting, I'd like his opinion.

Edit:

And as far as maintenance, the only thing it would change is that you wouldn't be able to replace the valves with the engine in the car.

I'm not sure how relevant this is any more but if you have separate head and block castings you can put together different engines like legos.  Like the LSX.  Consolidate your engineering dollars.

My Ducati bevel drive engines don't have head gaskets. Neither do air cooled VWs.

The engines in the 1913 Moyer and 1908 Ford Model S I'm working on right now don't have them either. 

You can change the valves with the engine in the car, just pull the valve plug out, remove the valve keepers and pull the valve out. That being said, emissions compliance on a T-head engine might be a bit difficult for manufacturers. 

This is assuming you have all aluminum or all steel blocks and heads.

mixing steel and aluminum means you have to have a gasket to deal with the dissimilar metals and their different expansion rates.

fanfoy said:

I think I should ask my question in another way. 

The way I see it, making a modern DOHC engine with the head and block as a single piece instead of two separate pieces means:

1 - The only machining that would be harder would be machining the valve seats. But if they could accomplish that in the 20's, I really don't see this as a problem. But you also get rid of two very large machined surfaces. The rest of the machining is exactly the same.

2 -  Having the head and block as a single unit makes it much stiffer and stronger. You can use that to reduce weight or increase boost depending on your goals.

3 - You remove an edge in the combustion chamber which reduces detonation. 

4 - You remove a failure point. While modern engines have less head-gasket problems, there are still problems *cough Subaru cough*. In fact, I would argue that head gasket problems are still the number one problem on engines in my experience.

5 - You remove an assembly operation and a bunch of parts that can all break (head gasket and bolts/studs)

The only real problem that I can see in this is the casting of the block/head. If someone has some experience in casting, I'd like his opinion.

Edit:

And as far as maintenance, the only thing it would change is that you wouldn't be able to replace the valves with the engine in the car.

If a manufacturer were to incorporate this combo cylinder and head they would have to make the cylinder assembly separate from the block casting like many motorcycle engines for ease of manufacturing. So you still end up with the same amount of machining, but your gasket just moves somewhere else.

fanfoy said:

I think I should ask my question in another way. 

The way I see it, making a modern DOHC engine with the head and block as a single piece instead of two separate pieces means:

1 - The only machining that would be harder would be machining the valve seats. But if they could accomplish that in the 20's, I really don't see this as a problem. But you also get rid of two very large machined surfaces. The rest of the machining is exactly the same.

Playing devil's advocate...

I don't know if this is true or not.  There's been a lot of R&D into combustion chamber design in the last hundred years, and it's likely that the machining done on the head today is pretty different than what they did in the 20 and 30s, for both emissions and performance reasons.  If nothing else, gas engines in the 20s weren't direct injection.  I'm not a machinist, I don't know whether that could all be done through the length of the bore.

Installing the valves through the cylinder bore is going to impose some constraints on how long the valve stems can be and what angle you can install them at.  Does that matter?  I don't know.  Again, the constraints of a modern economy car engine are pretty different than an 80 year old race engine.

2 -  Having the head and block as a single unit makes it much stiffer and stronger. You can use that to reduce weight or increase boost depending on your goals.

Well, modern engines are all aluminum, right?  How many of them are sleeved?  Can you install sleeves through the bottom?

The only engine I've taken apart to this level is from a Miata, and those pistons won't go in the bottom because the main bearing housings block the edges of the cylinder.  To fix that you'd either need to increase the cylinder bore spacing or decrease the width of the main bearings, both of which have significant implications.

3 - You remove an edge in the combustion chamber which reduces detonation. 

4 - You remove a failure point. While modern engines have less head-gasket problems, there are still problems *cough Subaru cough*. In fact, I would argue that head gasket problems are still the number one problem on engines in my experience.

Well, you're not going to build a boxer engine this way, so I don't think that solves the Subaru problem. :)

There's also the fact that on many engines, the head gasket acts as a kind of fuse.  You overheat the engine, you blow the head gasket, the car spews steam everywhere, and most people aren't *quite* stupid enough to keep driving it after that happens.  If you get rid of the head gasket then that probably means they won't stop until the pistons seize.

5 - You remove an assembly operation and a bunch of parts that can all break (head gasket and bolts/studs)

While that's true, you also remove the ability to fix certain types of failures by replacing the head.  This means that a number of warranty repairs are going to be change from "replace head" to "replace motor", which is going to significantly increase costs for those repairs.

My Porsche 964 doesn’t have headgaskets. Mine are dry, but some leak oil at that interface when cold.

As long as headgaskets can make it past the factory warranty period, none of this will  change. 

Car manufaturers aren’t interested nor inclined to be to make extremely long-lasting engines.

I would think when a manufacturer heard that people are keeping their cars for 300,000 miles, there’s gonna be some discussions to be had.

codrus said:

fanfoy said:

I think I should ask my question in another way. 

The way I see it, making a modern DOHC engine with the head and block as a single piece instead of two separate pieces means:

1 - The only machining that would be harder would be machining the valve seats. But if they could accomplish that in the 20's, I really don't see this as a problem. But you also get rid of two very large machined surfaces. The rest of the machining is exactly the same.

Playing devil's advocate...

I don't know if this is true or not.  There's been a lot of R&D into combustion chamber design in the last hundred years, and it's likely that the machining done on the head today is pretty different than what they did in the 20 and 30s, for both emissions and performance reasons.  If nothing else, gas engines in the 20s weren't direct injection.  I'm not a machinist, I don't know whether that could all be done through the length of the bore.

Installing the valves through the cylinder bore is going to impose some constraints on how long the valve stems can be and what angle you can install them at.  Does that matter?  I don't know.  Again, the constraints of a modern economy car engine are pretty different than an 80 year old race engine.

2 -  Having the head and block as a single unit makes it much stiffer and stronger. You can use that to reduce weight or increase boost depending on your goals.

Well, modern engines are all aluminum, right?  How many of them are sleeved?  Can you install sleeves through the bottom?

The only engine I've taken apart to this level is from a Miata, and those pistons won't go in the bottom because the main bearing housings block the edges of the cylinder.  To fix that you'd either need to increase the cylinder bore spacing or decrease the width of the main bearings, both of which have significant implications.

 

3 - You remove an edge in the combustion chamber which reduces detonation. 

4 - You remove a failure point. While modern engines have less head-gasket problems, there are still problems *cough Subaru cough*. In fact, I would argue that head gasket problems are still the number one problem on engines in my experience.

Well, you're not going to build a boxer engine this way, so I don't think that solves the Subaru problem. :)

There's also the fact that on many engines, the head gasket acts as a kind of fuse.  You overheat the engine, you blow the head gasket, the car spews steam everywhere, and most people aren't *quite* stupid enough to keep driving it after that happens.  If you get rid of the head gasket then that probably means they won't stop until the pistons seize.

5 - You remove an assembly operation and a bunch of parts that can all break (head gasket and bolts/studs)

While that's true, you also remove the ability to fix certain types of failures by replacing the head.  This means that a number of warranty repairs are going to be change from "replace head" to "replace motor", which is going to significantly increase costs for those repairs.

 

I'm too lazy to address the other stuff this morning, but as for your last sentence, it's a value game for the consumer. X repair becomes more expensive, but X repair becomes less likely. 

This is why I write instead of code or do math for a living. :)

The biggest reason I see is modern packaging limits. You can not build a block with spacing that wide to allow a bottom end rebuilding due to the length.  If the pistons have to come out the bottom then the main bearing have to go out wider, this means block grows and now you deal with heavy crankshafts and flexing too as there wider spaced between main bearings or you skip mains. A lot of the old motor built this way had only 3 main bearings and big heavy , flexible cranks.

The other option is to have the cylinder bolt to the crank cases like a motorcycle however then your left with high forces on that flange....watch some top fuel Harley's and see how well that works...