Laust Pedersen

Last Sunday I was testing my car (after adjusting water injection flow), which consisted of a few power runs with and without brakes applied. As usual my Accuboost was adjusted to 25psi, but all of a sudden I saw a huge white cloud in the rear view mirror and the engine temperature began to rise, so I turned off the freeway. There was so much smoke, that when I stopped at an intersection with a slight tail wind, I could not see 3 feet ahead.
After the cloud cleared, I limped ahead until the temperature got too hot, stopped and opened the hood to cool down. Fortunately I was very close to Danno’s home and 20 minutes later was able to limp there to call AAA to tow me home.

There were no ugly sounds from the engine, except that it was only running on 3 cylinders only. This obviously had the sure signs of a blown headgasket (HG), but how could that be, when I only 2 months ago replaced it with an MLS gasket (GURU’s ultimate HG), which in my engineering analysis is way better than the standard and the WF HG. We (Danno and I) were also very careful with proper installation of the MLS HG and before “the event” there were no oil or water leaks, the oil stayed clean, its consumption was very low and the 850 rpm idle vacuum was 18-19 inHg. All signs of a successful mini-rebuilt (con-rod bearings and a rod were replaced too).

After taking the cylinder head off, I found that the problem was a cracked # 4 cylinder on the passenger side from top to almost bottom. This is the side where the piston-cylinder forces are the highest (previously calculated to about 1500 lb). The crack has the signs of being a fatigue fracture rather than a one time stress fracture (brittle break). The MLS HG was shiny at the location of the crack, indicating that the (cracked?) cylinder wall had moved a significant amount. The other cylinders showed no such signs and no visible signs of cracks.

My car was just dyno’ed at 325 Wft-lb torque at 25 psi, which I can’t help trying almost every time I drive the car. With port water injection I am sure there is more torque while driving due to the increased airflow relative to the 20 mph fan at the dyno-shop.

So why did only cylinder # 4 crack? It is closest to the flywheel and therefore does not see any of the softening torsional flex of the crankshaft. It also was the troubled cylinder, when I blew the WF headgasket 2 months ago.
I am afraid that frequent use of 325 Wft-lb torque simply may be the limit of the block rather than a weakness in my particular engine.

What is the highest documented torque that has been run for extended periods on a standard block?

Laust

PS Pictures will follow once I get my camera back from the dyno-shop, where I forgot it.

Jeremy Himsel

Laust,

I suspect that you ran lean and the o-ringed head and MLS gasket were tougher then the cylinder walls. I have a block here that had the same thing happen (except it was # 2) to it running 18 psi but was out of tune and ran way lean. I think the extra heat generated by spinning the 26/6 so high was a contributing factor. There are a ton of guys running way more torque then that and haven't had issues.

J Chen

Hi Laust.
Have you ever tested egt readings for each
cylinder ? I suspect that cylinder 4 could be
running a little leaner hence a higher temp.
which could lead to metal fatique. I've read
that listers are claiming that the manifold is
the culprit to uneven air distribution among
cylinders but I have another therory. It could
be that the design of the fuel rail that's causing
uneven fuel distribution among the cylinders &
not uneven air distribution.

Ski

I blew my head gasket on #4 and when we blew it on the track car(low ocatane) it was number 4.

Laust I do have a question though; after talking with Neil at PD about a couple of weeks ago, he told me he had NOT run the MLS gasket with an o-ringed head. Not that you can't do it, he said you just had to make sure that the dimple and the o-ring lined up perfectly where the o-ring was in the middle of the dimple - hence the reason we did not run it on the track car but how did you check the alignment of the dimple compared to the o-ring in the head? TIA

fast951

Laust, first sorry your are having engine problems, it's no fun. Based on the MLS supplier they do not recommend an O-ringed head... Running the K26 at 25psi will make it VERY inefficient and you are generating plenty of heated intake temps.. Did you get any detonation?

As far as running 325 rwft-lb TQ on a stock engine, that's not a problem as long as the mixture and timing are fine. I run my race car stock engine (WFR HG & O-ring head) at 20+psi on the track, under load. The car is making over 400rwhp/tq. I have been doing this for years...

Chris White

Hmmm….I have your block’s twin sister! I split the #4 cylinder due to too much spark timing. The ignition timing was accidentally advance by 15 degrees and then a full power run. split the cylinder but left the widefire head gasket intact.
Cracking a cylinder is much more indicative of timing or detonation issues rather than too much torque. Any hick up in a water injected motor (if it has been tuned to run with the water on) can cause catastrophic failure in a short time. The increase in intake charge with running your turbo at higher than 20 psi can be pretty impressive.

All that being said – the failure was not due to too much torque – there is still a bunch of ‘headroom’ left above your out put.

Chris White

Mike S

Wow....25psi on the stock K26. I was seeing and EGT of 1550F at 18psi on my stock turbo. I can't imaging how hot it would have been at 25psi.

Laust Pedersen

No, I didn’t run lean. The AFR was in the 11.5 - 12.0 range when the car was dyno’ed about 2 weeks ago. Danno started out conservatively and gave me too much fuel in the beginning, causing it to soot (black spots) in the back. The combustion chambers and inside of the headers showed no sign of having run too lean and my air-fuel meter (narrow band sensor) always shows rich while under power.

Also the rules are different for water injected (WI) cars. Running lean is only dangerous due to heat, but enough water can be added before quenching the combustion to adequately cool it. While not having tried it, I am sure there is enough cooling to run below stoichiometric.

Detonation (pinging, knocking) comes in various intensities from light (barely audible) to heavy (sounds like breaking metal) and I would claim that detonation that cannot be heard with a trained ear does no harm, at the same time acknowledging that detonation drown in noise above maybe 5000 rpm and that the racetrack is not the place for listening to detonation. I will also claim that occasional for a second or two does no harm, unless the cylinder and/or power transmitting components are already stressed to their limit. Detonation increases (and shortens) the peak combustion pressure.

No I did not have detonation, apart from occasional fractions of a second (and my prior HG “event”) and the combustion chambers shows no sign of it.

To once more test the limits of the effectiveness of WI, I at one time filled it up with 87 octane gas and had no pinging at 25 psi. However it did ping strongly for a fraction of a second when it for the first time was priming the water lines. On a daily basis I am running 89 octane and 91 octane on the racetrack.


Jeremy & Fast951: The extra heat generated by the K26/6 is easily cooled by the water, see below.

J Chen: No I have not done EGT measurements and maybe I for documentation purposes should do that in the future. However the 3 (non washed) combustion chambers looked fairly uniform with respect to deposits and showed no signs of excessive heat.
The only temperature sensor I have is the coolant sensor, which can be trusted to go up when the temperature increases and down when it decreases and seems to be repeatable, but here is the interesting part; when I am driving close to the water trigger point (about 14 psi boost) the gauge goes above half and when I push it above the trigger point, even into 25 psi it goes below!! half.
Well, physics works it takes a lot of calories to evaporate water and fortunately the amount of water needed to keep the temperature at reasonable levels is less than what is needed to quench (extinguish) the combustion.
As to your fuel rail theory: The damper is there to minimize pulsations -> standing waves -> local pressure peaks and valleys -> uneven fuel delivery, so I don’t subscribe to that theory for the standard setup, but most based on my general respect for German engineering.

Ski & Fast951: My O-rings are just inside the dimple in the MLS gasket, but has left a uniform indentation in the HG. These O-rings (also to my surprise) are not closed, but have an end gap of maybe .002” - .005”, so they cannot seal all by themselves, but they help in fixing the HG, which probably is not needed for a MLS HG. That the manufacturer is not recommending O-rings is different from them outright discouraging them. In my evaluation, O-rings are an advantage on std. and WF HG’s and do no harm on an MLS HG.
However the MLS HG may have a disadvantage in not fixing the cylinder, by not providing a locking indentation for the cylinder top (fixing = preventing the cylinder from rocking due to the high piston-cylinder side forces).

Chris: The torque is a good indicator of the forces in the combustion chamber needed to crack the cylinder and I prefer that there is enough headroom for brief unintended events such as detonation and excessive heat.
No doubt that missing water for some time in an engine relying on its cooling effects is a killer.

Fast951: Frankly I am impressed with your car, which strangely is contrary to my experience. I would be even more impressed if you have both old and new dyno charts showing that your numbers have not dropped off with time.

Mike: The port water injection keep the heat generated by inefficient turbo range, adiabatic compression of intake air and added power (with associated heat) at very tolerable levels.


Also note that I am systematically following my “max reliable performance with min mods” goal and am trying to figure out if this set-back is specific or generic. In essence I am quite sure I have increased the HP/TQ while keeping good control of the fuel and ignition and with this thread I am trying to find out how pervasive cylinder cracking is.
At this point I believe that it is worthwhile to come up with some reinforcement of cylinder and have numerous ideas on how to do that relatively easily.


Thanks for your comments, sympathy and sharing your experiences.

Laust

fast951

I wish I can dyno the car in the same configuration as when I dynoed it last. Due to a close encounter with the wall at Road Atlanta many items had to be changed, the intercooler and piping to name a few.. I plan on running it on the dyno so we shall see.

adrial

You dont need to search very far to find that the block is not the issue. 325 ft lbs of torque to the wheels is nothing compared to what other people have done with a stock block....

How does a stock HG fix the cylinders in place any better than an MLS gasket?

turbite

Sorry to hear about your troubles...

Its possible that water got in that cylinder and hydrolocked the engine.

I've heard of this happening via leaking head gaskets, and I suppose it could also have happened if your water injection nozzle stuck open...

J Chen

Hi Laust,
How about getting a custom deck plate installed
& weld it about 0.5 cm below the top of the
piston sleeve. I've always wondered about
the reliability of the open deck design of the
block under high hp operation especially in
racing application.

UK952

Purely a stab in the dark but could it be related to the high pressures?
You are putting a mixture into a cylinder at 25psi, this is then squashed further (complete with some hard to compess water). This repeated high pressures has caused some expansion and contraction in the surrounding metal eventually leading to fatigue and cracks?
Tony

porshhhh951

Uk that's actually what I am kinda of thinking.

There are ton's of people making more than 325rwtq and having no problems what so ever. So it didn't crack because of the tq. Question is what was it.

Laust Pedersen

Well, I finally got some pictures taken.

The first picture shows the crack in cylinder #4 at the power side (passenger or exhaust direction)

The next picture shows the MLS HG upside down and it can be seen that the cylinder wall movement has polished the mating area on the gasket.

The last picture shows cylinders # 3 and #2 with an unrelated problem, namely some deep scoring in # 2 at 6-8 O-clock which may have been caused when I in June replaced a slightly bent con-rod and may not have noticed that the piston also was distorted as a compensation. At that time the scoring was only at around 6 O-clock. The marks in cylinder # 3 are merely optical and cannot be felt at all.

In essence the block is dead and since the car had a few other “issues” (PO accident bigger than I thought) I have already found a substitute, which appears to be a success story, which I will show in another thread.


Fast951, I look forward to your next dyno chart.

Adrial, The std and WF HG are softer (compressible) than the MLS and the cylinder top leaves an indentation in the HG, which in turn creates a ridge that to some extent holds the cylinder. The MLS HG is quite slippery especially if copper spray is used (as we did) and too hard to provide a holding indentation.

Turbite, All the water nozzles are “stuck open”, when the water injection is on. They have an upper flow limit defined by the nozzle size and water pressure. If it was hydro lock the most likely result would be a seriously bent connecting rod and the piston-cylinder forces would also be in the opposite direction.

J Chen, I am thinking of a solution along those lines, but without welding.

UK952, You got me thinking along those lines, but slightly different:

If you take a look at the piston tops in picture #3, maybe you can see that some of the thermal coating is flaking off (I tested it with an ohm meter), but only at the passenger side. Either HPC has an adhesion problem with their coatings or there is more to the story. Maybe the boost threshold for the water injection is too high and too much water is added, resulting in a thermal shock treatment of that part of the piston with a resulting build up of stress (from thermal gradients) high enough to chip off the coating.

The same argument could be used on the cylinder wall, namely that the abrupt water cooling of the hot inner cylinder surface induces stress (from thermal gradients), which added to the cylinder-piston force exceeds the yield strength of the aluminum.
Based on torque curves I have with and without water at 15 psi boost, I was definitely using too much water at that level.

I have in my past experienced that abrupt cooling (large thermal gradients) can do serious damage. This happened when spraying cool liquid (carb cleaner) into the intake of a fast running engine, resulting in numerous cracked ring-lands and broken rings.

A question to all of you with experience in cracked cylinders: Was the crack on the power (exhaust or passenger) side?