2LT cylinder head cross cut, the reason it cracks and fails

[GTSSportCoupe]

Interesting to know the 3B thickness; thanks.

It's almost like Toyota scaled the head based on displacement.

2.4litres/3.4litres = 0.7 (ie. 2LTE is 0.7 displacement of 3B)

0.7 * 8.5mm = 5.95mm

5.95mm is pretty close to the 5.7mm that Fantom measured.

Scaling doesn't work when talking the same pressure and temperature requirements though...FAIL Toyota.

As depressing as this is, and after reading hundreds of threads about people's bad luck with the 2LTE head, I'm going to keep running mine. I love my LJ78, and I love the way my engine runs after the improvements I made. Lots of power; runs cool. Truck is comfy and capable.

 

[Fantom]

The design might be questionable or it might have a reason, what we should do is get an engineer that knows thermal and pressure effects on steel and see if the wall thickness, compression pressure, combustion/coolant temperature, and area all add up to a failure.
Now John, from Ateb is a person who has immense knowledge about diesels, racing and modifying them, he's been running 2LTE modified engines for years with upwards of 20PSI and no issues. He's rebuilt more engines than I've seen and he insists the issue lies in cooling and improper monitoring/setup of most 2LT engines. I respect his opinion (because he knows more than I) but I am still going to find someone who can do some accurate math on this wall thickness.
Yes, many heads, including the 3B and countless others have a cooling chamber between the valves, and a dimple between the two ports, but regardless of the size of the engine, 2.4, 3.4, 4.2, they all usually have similar compression ratios (20:1-22:1) and have almost the same compression pressures and coolant/combustion temperatures. So in essence, what happens in a combustion chamber of a 2.2L "L" series engine (500psi, 21.5:1 ratio, 1000-1200*F combustion, 215*F coolant) is not much different than what happens in a 4.2L 12HT with very similar numbers.
So if the wall thickness of the head, in that same spot is thicker, then it would be a logical conclusion that's the culprit.
I can't wait to cut another head apart...

 

[Wheelingnoob]

My Isuzu 4BD1T (DI) also has that dimple as dose the 4BD2T (IDI). The 2T has issues with cracked heads but the 1T do not. Not sure if its the IDI contributing to the cracked heads on those or not.

I think my 1T head has coolant passages between the valves since I have small frost plugs where they drilled through. You are not getting my head though

 

[Fantom]

You're no fun, it's for a good cause!

 

[Wheelingnoob]

You're no fun, it's for a good cause!

Hah no chance. Im sure i cod find you a 2T head.....

 

[Fantom]

I'll pass on the fish

 

[gerg]

I think it has to be an IDI thing as the pre cup chamber sheds incredible amounts of heat into the head its not funny. That will add alot of localized stress. Its pretty much the only reason that makes them so ineffecient. Most folks dont realize that IDI actually get better air fuel mixing than DI, but the net heat loss is so great it losses right there. Its the reason IDI can run higher rpms than DI due to better mixing. They also run lower peak cyliunder pressures as well. Newer common rail multi injection stuff asside though cus that stuff is pretty advanced. That is why the swirl intake is so nice on your isuzu head and the 1hdt Jereme as it gives way better air fuel mix with cylinder swirl giving you a noticably higher rpm peak torque than the 4bt, that and actually having a normal intake runner. Also give you better BSFC naturally. The cummins intakes are weird and dont breath at all at rpms above 2000ish. Its an incorporated part to so its very hard to modify flow. That is why you always see cummins intakes consist of a pipe that drops right down into the block and disapears. The isuzu is a race engine compared to it.

 

[Jan-78FJ40]

I have a 1HDT head with a tiny crack here in San Diego that I'd happily donate for cutting it up. Shipping would kill the deal though, I guess.
cheers,
J

 

[ozwallaby]

I have a 1HDT head with a tiny crack here in San Diego that I'd happily donate for cutting it up. Shipping would kill the deal though, I guess.
cheers,
J

Just take it in to any machine shop and get them to slice it on a metal band saw or similar..... I feel sure it would be relatively cheap.. Or get a mate who works at one to slice for free ... Maybe there should be a thread dedicated to sliced heads....now there's an idea

 

[kim]

I could supply a 2H head. arnott just did one though.

 

[Dougal]

Yes, many heads, including the 3B and countless others have a cooling chamber between the valves, and a dimple between the two ports, but regardless of the size of the engine, 2.4, 3.4, 4.2, they all usually have similar compression ratios (20:1-22:1) and have almost the same compression pressures and coolant/combustion temperatures. So in essence, what happens in a combustion chamber of a 2.2L "L" series engine (500psi, 21.5:1 ratio, 1000-1200*F combustion, 215*F coolant) is not much different than what happens in a 4.2L 12HT with very similar numbers.
So if the wall thickness of the head, in that same spot is thicker, then it would be a logical conclusion that's the culprit.
I can't wait to cut another head apart...

The pressures are very different between an ~18:1 Di engine (12HT etc) and a ~21:1 idi engine. So there's an increase in stress.
The precombustion cup is the difference in heat.

The geometry of these in conjunction with the heat flow and mechanical stress make analysis very difficult and expensive. You're talking at least a $10k job requiring far more expensive computer software than I've got access to. At least these days hardware isn't expensive.

It is however right in line with the problems universities like to feed their engineering students. It would make an excellent fractures assignment and then the results would be free. How good the results will be will depend on how good the student is.

 

[Fantom]

Yes, the compression in DI engines is smaller, but the 2H as well as most IDI engines sit over 20:1.

Hmm, I could visit BCIT and give them the head to play with if they are interested. I'll look into that when I get some time.

 

[Hulsty]

The geometry of these in conjunction with the heat flow and mechanical stress make analysis very difficult and expensive. You're talking at least a $10k job requiring far more expensive computer software than I've got access to. At least these days hardware isn't expensive.

It is however right in line with the problems universities like to feed their engineering students. It would make an excellent fractures assignment and then the results would be free. How good the results will be will depend on how good the student is.

If anyone has a CAD model I might be able to knock something up. I occasionally use ANSYS to do static structural and Transient thermal analysis of the Rubber things I make. Looking in the Anlysis systems I can also set up a Transient Structural analysis, temperature and pressure can be used as conditions. Analysis wise a simple model would be best to start with, could take ages to setup and run, I have not done that style before.

I could start with a flat steel plate and many assumptions, just need the time to do it

 

[Fantom]

I could easily do a solidworks 3D model of it. But again, time. Right now I'm beyond busy, but over the next couple months I will start looking at it.
I'd be able to measure it up and have it modeled within very tight tolerances, wouldn't be accurate enough to build a brand new head (without a CMM ) but it should do for this test.

 

[Hulsty]

I dont need a full model, just the smallest representative "slice" Most things I do have some form of symmetry so I do quarter models or a infinitely thin "slice" when it is completely symmetrical in the area I am looking at, otherwise run times get stupidly long and it takes a long time to find errors.

 

[Dougal]

One cylinder worth (the cracked one) will be enough for analysis. But it'll be an absolute bastard of a modelling job with all those internal complex surfaces.

 

[Fantom]

Actually it's not bad. Model the main outside shape, model the intake, exhaust and valve ports, then shell the entire thing to the wall thickness and it's done. Takes more time to measure the head properly than to do it in the software....well almost
I would just model one cylinder chamber, no need to do more.

 

[Ebike toter]

Hi all.. I'm interested if anyone has the coolant flow pattern for the 2L-TE head in the top cooling chamber at the valves. I'm looking for a clear definition of all the coolant passages from the block thru the head. I'm trying to design a work around by optomizing coolant flow in addition to adapting standard components used in oil & gas steam / cooling systems to

1. automatically vent steam vapor from coolant pockets ( vent valve)
2. automatically relieve coolant /steam by pressure (PSV valve)
3. intercommunicate coolant passages at the head mating surface areas to better distribute coolant flow..

Any ideas??? Comments???

 

[GTSSportCoupe]

Hi all.. I'm interested if anyone has the coolant flow pattern for the 2L-TE head in the top cooling chamber at the valves. I'm looking for a clear definition of all the coolant passages from the block thru the head. I'm trying to design a work around by optomizing coolant flow in addition to adapting standard components used in oil & gas steam / cooling systems to

1. automatically vent steam vapor from coolant pockets ( vent valve)
2. automatically relieve coolant /steam by pressure (PSV valve)
3. intercommunicate coolant passages at the head mating surface areas to better distribute coolant flow..

Any ideas??? Comments???

I don't think anyone has a schematic diagram per say. But the coolant flows from the block up into the head, and then out the front of the head to the radiator. I've looked into these passages of the head, and basically the coolant flows above/beside the valve areas and relies on eddy currents to cool the areas lower down and closest to the pre-cups (that are hottest). When these heads crack, the rear most cylinder gets it worst, and usually cylinder one (front) doesn't even crack. It's gets progressively worse from the front of the motor to the back. This means the back of the motor is the hottest.

Now something you should know, is the pictures on the first page of this thread are the pre 1989 rocker head engine that is totally different then what is on the 2LTE. That earlier engine actually doesn't even have the major cracking problems of the 2LTE. See pictures below of the first generation 2LTE head cross cut. Now note that this is the early part number. Toyota changed something to the later part number that largely stopped the cracking issue. I'm guessing they filled in the little divit as all it did was boil in there. The boiling creates a vapour barrier that prevents the heat from transfering away from the metal. This super heats the metal in some sections, causing fatigue and eventual cracking.

You need to PREVENT the steam in the first place, not just vent it away. The rad cap takes care of the venting. Ways to prevent the steam are run a waterless coolant that has a much higher boiling temperature (which I do and it works excellent - 385F boiling point). A very successful shop in Vancouver (that worked on these motors for a good 10 years before the US started getting them) adapted a 1KZTE much larger thermostat to the 2LTE housing using a custom spacer. They claimed this was THE FIX for the 2LTE. It allowed faster coolant travel through the engine/head, thus less time for the coolant to over heat and turn to steam. I have to agree that the factory T-stat is much too small. I'd say if you want to do something special for the 2LTE that you could likely profit from, why don't you make a new custom T-Stat housing that bolts to the factory cylinder head. I would not use the 1KZTE t-stat, but look a the 3RZFE t-stat which is also huge and easier/cheaper to get.

You also need to understand that Toyota made the 2LTE from 1990 through to 2005. The only versions of this motor that have head cracking problems are the 1990-1994 versions or so. After that Toyota produced at least another three part numbers of this cylinder head which DO NOT have the same head cracking issues. So as admirable as your efforts are, there are already off the shelf solutions to the head crack issue.

 

[Dougal]

I would not say only the 90-94 versions crack heads. Here in NZ we have all of them as used japanese imports. The L and KZ series engines keep engine rebuilders alive.

The issue is that IDI engines put more heat into the head so you've got more stress from thermal cycling. Each failure point takes a long time and a lot of miles to show. Revise the design and start the process again.

But that also means that if you have a cracked head, just install a new one and you'll be good for many more years. The new one will last longer than the last.