This article concludes here.
As a general comment on Harley-Davidson entering the high-performance world, none of these problems surprises me, as every performance parts manufacturer has gone through teething problems to some degree when initially venturing into the big-inch engine field. I don’t believe Harley-Davidson will be an exception. Attempting to rectify this situation with gasketing will not work if the problem proves to be a structural design or manufacturing hitch. A thicker fibrous gasket may alleviate the symptoms. Either way, I am a big fan of an O-ring around the oil passage as it intersects the head and cylinder. Perhaps the CVO 110 cylinder needs a redesign to allow for this simple but effective feature.
Here is another question: why don’t the TC 95 and TC 103 rebored cylinders suffer the same problem as the CVO 110? After all, not all three cylinder bores have enough room to use an O-ring around their oil return head gasket dowels. Yet only one of the three suffers an oil leak. I think it’s because the 110’s liner top portion is shifting. The H-D three-layer head gaskets need a very smooth surface to function correctly. MLS gaskets — the acronym stands for multilayer system — are comprised of three separate layers of stainless steel or aluminum. The Harley-Davidson MLS gasket’s two outer layers are embossed in 0.001"-thick Viton rubber-coated stainless steel. (Never use any sealer whatsoever with these gaskets. I can see where frustrated mechanics might try sealers as an assist.) The MLS gaskets are corrosion-resistant, reduce bore distortion, and withstand extreme cylinder pressures. This may well be the reason why MLS gaskets do not blow from the cylinder outward in the CVO 110, even with the extreme heat present in the rear cylinder. Unfortunately, they cannot be used with Viton rubber O-rings around the dowel pins, as they will compromise the gasket’s sealing effectiveness, although a cylinder and gasket redesign would handily take care of this irritation. The MLS gaskets need a smooth roughness average (RA) gasket surface to function effectively.
The earlier Harley-Davidson CVO 110 head gasket uses a Viton O-ring around the oil return dowel, but the proximity of dowel and cylinder liner lets the O-ring lay onto the shifting cylinder liner. A Viton O-ring, in conjunction with a thick fibrous gasket and fire ring, would probably resolve the blowing/seeping situation if the oil return dowel hole were centrally located on the alloy gasket surface, away from the disruptive cylinder liner. However, it cannot effectively do any sealing because it hangs onto the sleeve, close to the cylinder bore. This is where the speculation that the head gasket failures are exclusively the fault of cylinder liner movement evolves. The new MLS gasket does not use an O-ring and leads our causal diagnosis elsewhere. The MLS gasket has a ridge for dowel sealing, which it overlaps onto the liner. S&S Cycle, the stroker and big-bore king of the Harley-Davidson aftermarket, uses what appears to be an aluminum center covered by a Kevlar gasket with a fire ring on the inner diameter of its 4" bore gasket, with no problem of leakage or blowing that I am aware of. The H-D three-piece gasket is smooth on both sides, with a stainless steel center not bonded to the outer layers like the S&S gasket. The S&S 4" bore head gasket has a fire ring that can compress and fiber material to mold and seal more imperfections than the H-D MLS gaskets can.
At Heavy Duty Cycles, we have been substituting S&S 4" head gaskets for all of the H-D gasket series to date. Harley is hot on the case to correct this problem, so new developments can occur at any time. If the reader takes a look at a TC 88 head gasket, it is thick and fibrous to absorb variances, minor warping, and a rougher gasket surface. It also has a fire ring to seal compression in, not that compression leaks have been a problem with any of the Twin Cams, including the CVO 110. The TC 88 gasket also incorporates an O-ring around the two locating dowel pins on each head to add a further seal. This O-ring provides an excellent seal around the one dowel per head that does double duty as connecting and directing the oil return passage from head to mating cylinder. I mean, what is to understand here? In my opinion, the new MLS gasket is just too high-tech to correct other symptoms such as head warping, although it does seal the shifting top part of the cylinder liner. Lower the heat and many of the symptoms would disappear.
On my customer Chris’ CVO 110, the head gasket surface was not smooth, and it had crude machining marks. Tightening down the gasket to specifications may even be pushing the cylinder fins away from the top of the liner on the rear head. Again, here is the kicker: MLS gaskets require a surface finish of 50-micro-inch roughness average (50 RA) or finer for a proper gasket seal. Anything rougher will conflict with the gasket design. For aluminum heads, the numbers are even lower. The typical recommendation today for aluminum is a surface finish of 30 to 60 RA, with the preferred range being from 50 to 60 RA. For a comparative understanding, honing with #220 grit stones typically leaves a finish of just over 50 RA. Using #240 grit stones will usually leave a finish in the 38 to 45 RA range; #280 grit stones will generally produce a finish of 18 to 25 RA; and #400 grit stones lead to a finish of 10 to 15 RA. Honing with #600 grit stones can push the numbers down to 5 RA. A pane of window glass, by comparison, measures about 3 to 4 RA. Furthermore, it looks like the dealer mechanic tried to smooth some high spots on Chris’ rear head with a hand-held power grinder disc leaving an RA much too high for the MLS gasket to seal.
In Chris’ second H-D dealership rebuild, it’s a coin toss to figure out whether the disc grinding, which left an approximate 70-80 RA gasket surface finish, or extreme heat blew the gasket first. Both conditions will do it. It’s a question of which came first, the grinding or the gasket?
Smoothness has become a major issue because the difference in thermal expansion rates between an aluminum head and cast-iron cylinder liner creates an amount of sideways shearing force and scrubbing action on the head gasket. If the surface finish is too rough (more than about 60 RA), the metal will bite into the gasket and pull it sideways as the head expands and contracts. The cumulative effect over time can cause a delaminating effect in the gasket, literally tearing it apart, causing it to leak and fail. Some aftermarket gaskets can handle a rougher gasket surface finish on these types of engines, but it all depends on the design of the gasket. Even so, smoother is definitely better on these engines. The Harley-Davidson MLS laminated, stainless steel gasket outer is extremely durable in smooth surface conditions. The multiple layers of stainless steel have two outer layers with a 0.001" thin coating of Viton rubber. The design prevents the gasket from losing torque over time. The composition also reduces the amount of torque that’s required on the head bolts to seal the gasket, which in turn reduces cylinder bore distortion for better combustion sealing and reduces internal compression ring blowby. I cannot fault the Motor Company for trying to cure this heating mess with progressively better, superlative gaskets. Even though the gaskets are failing, they are leading-edge technology.
The Causes
During engine warm-up, the H-D EFI runs in open-loop mode at a 12.0:1 air/fuel ratio. The default idle air/fuel ratio is 12.0:1 for new engines, no matter what the engine temperature is, even if you disconnect the exhaust O2 sensors. Once the ECU determines when the engine has heated sufficiently, the EFI goes into closed-loop mode at hot idle. The engine will hot idle at a lean, stoichiometric 14.7:1. Normal idle speed is 1000 plus or minus 50 rpm. Cold idle will be the higher figure. In high-temperature mode, the engine drops to 950 rpm and the ECU goes to open-loop mode, richening the fuel mixture to 12:1 in an attempt to cool the engine.
Once the rider detects the rpm change, it’s time to ride so the air moving over the cooling fins can do its job cooling the engine. I never let my bikes sit at idle. I get on them and ride conservatively until they’re warmed up, letting the air work its cooling magic. There is no airflow when stopped at idle. If the engine idles up to 1200 rpm, usually because of the rider feathering or blipping the throttle, the ECU retreats back to closed-loop mode at 14.7:1, which is too lean for an air-cooled engine.
An EFI throttle should be left untouched by human hand, as the smallest amount of throttle will lean the engine out as long as the ECM is operating in closed-loop, high-heat mode. EFI functions quite well untouched at idle. The interfering human touch only makes a dire situation, that’s attempting to cope as best as possible within EPA limitations, much worse. If the bike accelerates past 1200 rpm at idle, the engine will stumble under high heat. The continuation of the lean 14.7:1 air/fuel ratio condition can drive engine temperatures even higher, especially in the air-shrouded rear cylinder. On super-hot days, stop-and-go traffic, or some combination of both, the ECM will sense a potential overheating situation in progress and put the engine into parade-duty mode. The ECM will direct the EFI out of closed-loop mode, thus lowering the air/fuel ratio to an approximate 12.0:1 rich condition. This richer air/fuel ratio attempts to cool an air-cooled engine that is not getting relief from air circulation. Cylinder head temperatures can go nuts in this type of situation, double sometimes the acceptable 225 degrees Fahrenheit and my preferable 210 to 215 degrees. In fact, the stoichiometric air/fuel ratio of 14.7:1 goes up to 14.8-16.0:1 on the rear cylinder due to insufficient cooling from airflow, leading to the detonation. The intense heat causes the 0.625" of cylinder liner not tied in by spiny lock construction to expand. The rear head gasket surface begins to warp in a plane between the head intake and exhaust ports. A three-piece MLS gasket that depends so much on a smooth gasket surface cannot protect the rear oil return dowel, in line with this warped plane. And Harley-Davidson keeps inventing new, leading-edge gaskets to treat the symptom not the cause as per usual with the CVO 110. Of course, lots of fiberglass, especially lowers on the FLHT models, only exacerbates a bad situation.
Donny Petersen
Tattoo Tony’s Heavy Duty Cycles
Toronto, Canada
www.HeavyDutyCycles.com 
Techline-April09-H-D 110" CVO - Part IV: The cylinder liners, etc. (concludes) 
Linear Mode
