Cracked headers, cylinder wash, how?

Based on my memory of reading the various responses on here for this in the last many, many years, you summed it up good
I would add that the exhaust reversion that can occur on the 355 may allow debri from the exhaust like broken up failed cat internals to enter the cylinders which over time would score the liners. Hopefully more knowledgeable folks will enlighten us.

 

I wonder if it's worth installing an air fuel ration gauge.

 

you would almost need 8 seperate thermo sensors to truly get a accurate reading of a lean / rich condition

 

Has nothing to do with air getting to O2 and fuel mixture. As you say it would effect 4 equally. Too many problems with that scenario to list.


Cyl damage is mostly from insulation debris in header getting in.

 

Interesting. It seems the insulation causes more problems than it's worth (seems to be part of what causes tube #2 to melt in the first place). I'm thinking pulling the headers and replacing the insulation with header wrap would be a worthwhile winter project.

 

From what I have seen, 1-4 manifold fails first, almost always. And it's usually cylinder number 2 tube that has burned through. At this point out comes the compression gauge. I go directly for cylinder 2. And more often then not, it's down on compression by a fair amount.

I would love to understand why the 1-4 bank manifold always seems to fail first, and why cylinder 2 seems to produce such high temperatures. Debris could be a reason for cylinder failure after the manifold has burned trough, but what is causing cylinder number 2 specifically to burn out. Yes the OE manifolds are crap, but is that the only reason why they fail? I don't think so.

 

Are the bends on 2 and five the tightest and also closest to the adjacent tube? That, coupled with the uneven insulation of the tube could create a localized hot spot that burns through first, especially if the mixture is off.

 

Yes..the bends are extremely tight and way too close together. Its a domino effect. The hole in cylinder 2 tube, I believe acts like a blow torch. Not only is the next cylinder heating the tube from the inside, but its now getting heated from the outside, far more then normal.

I believe these 2 holes cause cylinder 2 to run extremely lean causing the failure. But, what causes cylinder 2 to burn through? Is it just a lousy manifold design? I would think that's a big part of it, but is there something else.

Has anybody ever flow tested an intake and throttle body set up off the car on a bench, and monitored the flow of air to cylinder 2? Is the bend to tight at cylinder 2 not allowing exhaust to flow properly? It looks to me like the exhaust flow would almost be like hitting a wall, and creating a hot spot on the metal, and over time it just fails.

There are so many variables to this issue. I know Dave Helms has thoughts on it, but he doesn't post here anymore. Unfortunatly.

 

In mentioning Dave in this thread I will say that most recently in personal communication (4 months ago) he advised me to replace my headers with new OEM headers which I did. So, I am not so sure he prescribes to your above outlined theory of etiology.

It could be that he is confident of the redesign of the new OEM parts as having remedied an original problem as you described as well.

 

Fwiw, the below is all me throwing ideas out based on what I know about the car and not any measurements (as I don't have any).

I think the cats don't meltdown when this happens (at least right away) because the exhaust is pulling air in through the hole in the header, so it evens the mixture out at the O2 sensor (and hence the cat). You would get an SD light if the mixture went off and there were no hole letting oxygen in, but your O2 sensor should trip a CEL first - something like a 'fuel trim out of range' error - unless it was the cause of the engine going rich in the first place...

The headers do have two access ports for, presumably, checking the mixture, so you could just put 4 widebands in and measure cylinder pairs. You should see one pair go lean and the other rich at part throttle if you blew a hole in the header. At full throttle, one would go lean as the computer no longer uses feedback from the O2 sensors and just runs a fuel map, so it wouldn't adjust for the air getting into the stream and one cylinder pair (with the hole) would go lean while the other didn't.

It's interesting in the other thread that says the challenge cars didn't have the issue with headers. As the challenge cars would rarely be running in open loop mode (with feedback from the o2 sensors), could that mean it's due to whatever mixture Ferrari needed to run to pass emissions? I know they have to lean it out some for that reason, so the high egt's could be what kills the cylinder 2 header tube. A simple remap or piggyback setup to richen the system up some at part throttle could solve the issues provided it doesn't kill the cats (and you can still pass emissions).

There's also Capristo's theory that the setup of the valved exhaust is causing the problem as when the valve is open, it pulls the exhaust back through the primary cats super heating them and causing the whole system to overload, but that doesn't make sense given the observation that challenge cars don't have the issue (as they run in bypass most of the time even with the valve installed).

I also wonder if some of the issue could be due to the bypass valve not fully closing, causing some feedback through the bypass circuit that the system picks up at the o2 sensor and mistakenly adjusts the mixture for. You'd need to spend a day (or 3) with wideband o2 sensors on all 4 ports and egt monitoring measuring the car in all kinds of different scenarios to get a feel for how different things affect the mixture and egt. Perhaps the issue is that the SD light triggers at too high a temperature and doesn't catch the issues earlier?

 

I havn't taken apart the updated set of manifolds. So I can't comment on the new design, or what could have been changed. I know they are much cheaper then they used to be which is really nice. Obviously Ferrari has recognized there is an issue, which carries all the way up to the 430, and possably the 458.

The manifold is definatly a weak point, there is no arguing that fact. But I can't help but think there is another issue, which results in big failure if the burn through is not caught early enough.

 

The motor hasn't been made where all cylinders breathe the same and run at the same temps. In race motors we try very hard to achieve that and it requires a lot of gymnastics to get close. Remember the old big block Group 7 cars with the uneven height intake stacks? Just trying to get the cylinders to breathe the same. And it wasn't unusual to run a bunch of different pill sizes on individual cylinders to try and get even exhaust temps.

 

Reversion happens on intake and exhaust. Well known, well accepted. I have proven in legal cases that debris from a failed cat can get back all the way to the intake manifold from reversion.

Further when exhaust flow and extraction is disrupted by a failure of one pipe in a header its flow design characteristics just went out the window. Even a small crack can destroy extraction.

It doesn't go rich. It would be a CEL if it did.

In 96/97 we were recording all nature of parameters and sending them to Ferrari at their request to see "A" Why the header failures and "B" How they can be proven by the computer for warranty purposes and it was shown over and over there was not necessarily one single parameter out of spec on cars with failed headers.


They also knew damn well that cylinder wear could become an issue. They never came out and said it but we were asked in many cases to measure and report cylinder wear on cars with multiple header failures. This was all back when the cars were still in production. It is all very old news.

 

No choice with cam actuated valves. If you don't start opening the valves early, you end up with insufficient timing (duration).

Secondly, more cam overlap means more lower midrange.

Thirdly, exhaust gasses are not reinjested, the exhaust gasses are at a low pressure and are perfectly happy to pull fresh intake into the cylinder even before the piston starts down.

Only in the sub 2300 RPM range do exhaust gasses get back into the cyclinder as a positive wave.

 

Very good info to keep in that black hole in my mind, appreciate you sharing that.

Honestly for me I dont see these headers as a problem item. They seem to last as long or longer then most high performance/ race style headers I have used. 2-3 years for headers with "hot" motor and I was a very happy camper. I know due to cost we would like them to last a lifetime but these are not honda civics

 

Interesting discussion. Has anyone placed a thermocouple on the individual suction manifolds and looked for slight increase or decrease in temperatures relative to airflow? Am thinking specifically cyl. 2. issue. Does the air velocity increase or decrease relative to other cylinders. Engines are thermodynamically based. Decrease in intake air velocity and/or temp might splain' some things...

 

Why bother? We know what happens. We know how to fix it.

 

Well, If you look at the airbox design vs. say the 360, you can see that it's doesn't have as much taper, which could lead to velocity/volume issues between cylinders. You'd think that would affect an end cylinder more, though. Maybe the balance pipe that was added back to the 360 (348 had it as well) makes a difference?

 

My thought was that if it was slightly greater airflow, disrupting it preliminarily in the airbox might slow it specifically for that cylinder, thus saving any new headers that are put on, or those very early in their life. (of course you're right.. a new set of headers made of higher quality stainless would solve the problem)

That's kind of what I was wondering as well... Interesting.

 

You're never going to get all 8 exactly the same. If you insist on running things on the ragged edge of melting **** you need to use better ****.

 

No one else has some info other than Brian. At least we still got Brian giving us some details.

 

When the damage occors to the n° 2, what kind of damage is seen ?