Charles Navarro

Long story short, engine blocks and cylinders end up with ovality and taper as they wear and are repeatedly heat cycled. We go by a limit of .002" of ovality and taper in any cylinder, aircooled or watercooled, before it must be reconditioned.

On a slightly separate topic, many years ago Alcoa sent us aluminum that was not stress-relieved. We began making cylinders and we could not turn a round part. As the material was machined, it would relieve itself, with the final part having runout and taper. We ended up having the aluminum cut up and cryogenically treated to stress relieve it.

I've considered adding a cryogenic treatment step to any block that is direct plated, like we do on most DFI blocks, but cost has been a mitigating factor, as this would quite a bit of cost and complicate logistics.

Prairiedawg

Very interesting, this may explain why southern cars are better? Would our exceptionally cold late January (-27*, I'm in the Chicago area also) cause problems long term? Anything left outside those two days were essentially cryogenically treated also.

EDIT: after reading about cryogenic treatment, it appears -27f is nothing compared to the hundreds of degrees below zero a cryo treatment consists of.

Charles Navarro

The tighter piston to cylinder clearances coupled with colder climates explain why we see more scoring issues from Canada and colder climates than we see in the south or out west.

The use of alusil is nothing new for Porsche, as A390 was used in the 3.0SC and the transaxle cars (944, 928, etc) without scoring. I believe the difference there is in the clearances being looser coupled with thicker oils with higher levels of ZDDP and less detergents.

I'm still hoping for a tribological solution to this problem.

Charles Navarro

It was bothering me, so I went back through all the technical material I've amassed and found this in the KS manual to reconditioning Al engines:

"Cast-in liners are subject to residual stresses that are difficult to control. Another problem is inhomogeneity as a result of casting tolerances and deformation. So it takes substantial development efforts to control multifactor-conditioned cylinder bore distortion problems with potential functional drawbacks, especially in long-term operation."

Basically, they expect the castings to move over time due to built in stresses.

Prairiedawg

Pardon my ignorance but what is the KS manual?

"Another problem is inhomogeneity as a result of casting tolerances and deformation."

Could tolerances in the alloys but still within manufacturer spec be the difference between a block that moves and one that doesn't, one that moves more in cold climates. Or they all move to a certain degree but some move more than others if put through similar stresses (heat cycles, climate, mileage etc)?

You stated earlier that the older 944's etc used alusil without issue due to looser tolerances, makes sense. Do the higher HP/liter engines require tighter cylinder bores? Would they be fine with looser bores, thicker oils? If so why would the manufactures go with tighter tolerances if it would it seems to me drive up the costs?

Sorry for all the questions but this is really interesting to me and now we're really seeing the meat on the bone of Bronz's misfortune.

edomund

Correct me if I'm wrong, but when LSPI occurs there should be some visual sign of the detonation like pitting or in extreme situations damage to engine parts. I didn't see this in this engine. As far as I can see LSPI was not a factor in the failure. It looks like it is similar to Baz's findings in these engines

black997er

As someone said earlier, if you watch Jake’s video you’ll see the part where he talks about wear starting to show on the rod bearings. This would be from the excess force generated by premature detonations.

Charles Navarro

I realized I forgot to mention probably the most important detail. The piston skirt has collapsed on cylinder 1 resulting in over half a mm of piston to cylinder clearance. That's why there is no wear in the middle of the piston skirt. While assisting with the RND program, I found quite a few forged pistons with collapsed skirts, however, these piston typically never showed any irregular wear, likely as they were pulled out of service before they could score due to another failure.

The second ring also ended up with a .036" total end gap where the other cylinders had a .030-.031" end gap, indicating accelerated ring wear. We didn't bother measuring the oil controls, but cylinder #1's were messed up, likely from the added stresses the rings were enduring while trying to guide the piston up and down the bore with all that extra clearance.

Charles Navarro

Tighter clearances reduce engine noise and to a point, reduce piston rock, which helps to stabilize the pistons and rings in the bore. By reducing ring flutter, the net effect is improved ring seal which can be directly attributed to lower emissions and oil consumption.

Remember, we've seen scoring in Cayenne's and Panameras as well for years, and colder climates historically are worse for scoring in these engines as well as the M96/M97.

The very first failure I saw of scoring in the 9A1 was back in 2011 with a 2009 C4S from Kuwait, and it looked just like an M96/M97 failure with similar scoring on the bore and piston (which I actually posted photos on a 991 bore score thread yesterday).

edomund

Once again from my experience and from research when there is detonation you will see evidence of it on pistons etc not just bearing wear. In my 2 stroke racing engines when we run too much compression or get bad gas we see pitting on top of the piston. Also I've blown rotors sideways and destroyed apex seals in turbo rotary engines from detonation when a prototype ECU malfunction and added too much timing. Also in my turbo S2000 we would see LSPI sometimes on throttle tip in at low rpms and the datalog would pick up knock and you could audibly hear the detonation (sounds like popcorn popping). So what I'm saying is when there is LSPI you know it. I'm trying to eliminate variables that are not happening so we can figure out what we should focus our energy on. Also I'm not saying we shouldn't use a full saps oil as I'm using Motul 300V for various reasons.

Seems to me what the thread has taught us is and what Charles and Baz have said is warm the engine up before hammering it and use quality oil at lower intervals and stop worrying about all these other items. It's normal for blocks to have some movement and the only thing to keep in mind is the clearances are tight on this engine, so warm it up or you risk scoring. This is pretty straight forward and a practice we use in our race 2 strokes that run tight clearances and high compression, warm it up or you will stick a ring.

okie981

This is something I've wondered about, if LSPI is occurring to any degree whatsoever, will there always be an audible noise? I can see combustion not going as planned in a DFI engine with the different schemes used for multiple injections of fuel per combustion cycle at different engine RPMs and loads, and sometimes that disrupted plan for combustion causing abnormally high pressures and temperatures in the cylinder, but doesn't or wouldn't this always result in an audible sound with the naked ear?

Slightly off topic, but back in the days when I supported some guys with wrenching at kart races, we were at Daytona for the winter nationals (Christmas of 1984) and one of the guys was running a 2-stroke Dismore CKS engine. He leaned it out a bit each session and the last time, he never came back in after the session ended. Turned out he lost power on the back straight and had to get it hauled in. The top of his piston looked exactly like it had an acetylene blow torch held about 1/4" from the top of the piston right under where the spark plug was located. It burned a hole right through the top of the piston and you could see daylight through it. We inspected the rest of the engine, cleaned up the cylinder walls with some emery paper, put a new piston and rings in it and he used the engine for the rest of the weekend.

Balr14

Here's the rub. Most of us bought our 997.2 used. Even with complete service history, low mileage and a great PPI, there is no way to know if the previous owners followed these rules.

edomund

I'm talking about going forward. Obviously you can't control what someone else did before you, but that goes with any used product. Not sure what you're getting at. When you compare the 9A1 911 to other cars in its category it is VERY reliable, probably one of the most reliable comparatively. Look at how much run this thread has got, it's mostly because it is such a rare occurrence that people are intrigued. Granted we're talking about a high end/high strung sports car that's tuned closer to the limit than "normal" cars. I thought most people understand this when they buy a car like this. If you're looking for ultimate longevity buy a Toyota Corolla. If you're looking for a sports car that the manufacturer rung out most of the performance from and you want to have very good reliability as well buy a Porsche. If you want ultimate performance and have the money to not worry about it blowing up buy a McLaren or Ferrari. Look at all the forums for every high performance car, they all have their issues. A few 9A1 engines failing is not as you said poor engineering and it definitely doesn't make it an unreliable car. I believe saying so is a disservice to the knowledge base we try to create on these forums.

Anyways back to the point about what caused this rare failure.

bazhart

Lots of interesting responses - none of which seem to contradict what I posted a long time ago about the failures we have seen and measured - namely that the cylinders shrunk across the lower casting crossovers and because the clearances were so small from standard - they were too tight for piston clearance if the heat-up is too quick so the piston expands more than the cylinder bore and seizes.

I also note that we mentioned that the adjacent cylinders were slightly less shrunk and the other 4 not shrunk at all.

We have also only seen one side of the bore and one side of the piston whereas I showed that all ours were scored on both sides of the bore and piston which means that they seized not scored.

Seems to have been an awful lot of posting that didn't acknowledge what we originally posted and didn't really answer the questions about the bores and their measurements.

Its not (or shouldn't be) a competition to see who was right but so far I see nothing to contradict our much less involved and stretched out responses, analysis or advice.

We also analysed the piston profiles showing why the piston would pick up on the sides if the piston expanded too quickly compared to the bores. I sent all our analysis to Jake but not yet received any response from it relating to their findings.

Baz

Flat6 Innovations

Baz,
What you sent me to review was exactly what we’ve been seeing for 9 years. The pics could be confused with our own photos, because they are so much alike.