Every technician that has any pwc experience has seen it. The engine whose rear piston has scored (seized) for no diagnosable reason (while the other piston(s) look perfect). subsequent teardown inspections out rule air leaks, fuel and coolant blocks, oil, ignition, etc. At Group K we believe that "nothing" happens for no apparent reason. However in the last decade of working on pwc engines, we have seen numerous seized rear pistons that we could not account for. This document makes an effort to account for an awful lot of the ones we've seen.
SOME IMPORTANT EXPERIENCES - Among Sea Doo race engine builders, richer jetting in the rear cylinder (to avert piston seizure) has been standard fare from the 580 cc days. We remember speaking with some of those technicians about the apparent need for "uneven jetting". They all told us that identical front and rear jetting in their Rotax's was a near guaranteed rear piston seizure. We argued that none of our 650 Waverunner's or Kawasaki SS twins had ever required uneven jetting. We felt certain that they had some other technical problem that they were overlooking. We continued to believe that, until the day we started testing our first Yamaha 701 Raider.
Our Sleeper kit development for the 701 Raider was relatively uneventful. However when we installed the Coffman pipe that drove rpms to 7100 (from the stock 6400), we began to experience a rash of scored rear pistons. We realized that we were probably facing the same gremlin that our Sea Doo pals had been living with for years. We were bound and determined that we would be the guys that would solve the problem. We believed that we would show that uneven jetting would not be needed. We then experimented with milder compression, retarded timing, richer overall jetting, dual cooling with additional flow for the rear cylinder...and plenty more. Three weeks, and a dozen seized rear pistons later, we finally had a strong running format that could run wide open for almost a full minute...before it seized piston number 13. We finally broke down and put one size larger main jet in the rear carb. With this richer jet, the machine still accelerated strong, and reached peak rpm quickly. We ran it full speed for over 15 miles...we couldn't make it seize (or foul a plug). In fact, to our surprise, later tear downs would show not even the slightest visual signs of over richness or fouling. Shortly there after, we repeated this same frustrating scenario with the (then new) Yamaha 1100 triples. Only the rear cylinder desired richer jetting.
Armed with this experience, we revisited our earlier model stand up and runabout dual carb engine kits. Remarkably, none showed the need for uneven jetting. Even more remarkable was that an identically modified 701 Raider motor, mounted in a stand up hull, did not require richer rear cylinder jetting. The stand up, with the Raider motor could run wide open all day long on "parallel" jetting. When that same motor, with it's parallel jetting, was then installed into a Raider hull...immediate rear piston seizure.
As all this went on, we realized we were creating far more questions than answers. For the meantime, we simply accepted our experiences at face value, and used parallel jetting when we could...uneven jetting when we had to.
During the beginning of our development on the 782 Laydown Rave motors, we immediately started to experience rear cylinder scoring (that was staved off by richer rear carb jetting). At that same time, we were lucky enough to be having a conversation about this problem with Ross Liberty of Factory Pipe Products. They were in the process of developing their pipe for the Laydown Rave in their newly completed dyno facility. This dyno, and it's instrumentation, can tell "everything" that's happening while the engine is under load. Ross mentioned that they too were chasing the "richer rear cylinder" gremlin. He said that they had eliminated every imaginable variable, and no matter what they did, the combustion chamber temperatures were never identical under full load. He said they suspected something related to the crankshaft twisting. He suspected that this twisting was affecting the "sameness" of ignition timing. We suspect that he is correct.
CRANKSHAFT TORSIONING "The Gremlin" - This suggestion, by Factory Pipe, is what we consider to be the true cause of the countless scored pistons we saw during our own tests. To understand it, one must first understand the difference between twisting and torsioning. Crankshaft twisting refers to a crank that has been rotationally jolted so hard that one of the press fit connections rotates out of index...and stays that way. When a crankshaft "twists", there is large and immediate loss in power that comes along with a noticeable vibration. When a crankshaft "torsions", there is a momentary rotational springing action...nothing comes out of index, there is no vibration, there is no noticeable power loss. A welded crank cannot "twist", but it can still "torsion". The total amount of "torsioning" that takes place depends on the length and rigidity of the crankshaft. The longer and less beefy...the worse the torsioning. What we now refer to it as "crankshaft torsioning" (not twisting)...works like this:
The front cylinder on all pwcs is the cylinder closest to the ignition rotor. The amount of torsioning that can take place between the ignition flywheel and the front crank pin is nearly non-existent. However, structurally speaking, the rear crank pin is much "farther away" from the ignition flywheel. When the pump (at speed) suddenly hooks up on some smooth water, the front cylinder and the ignition flywheel have enough rotational momentum to torsion the crankshaft over it's entire length. As this happens, the flywheel and front cylinder can actually get 2 to 3 degrees of rotation ahead of the rear cylinder. As this 2 to 3 degree torsioning rotation takes place, the front cylinder is still getting perfectly timed ignition sparks. However the rear cylinder is lagging slightly behind when it gets it's ignition spark. This means that the rear cylinder is firing 2 to 3 degrees more advanced than the front cylinder. Furthermore, this advanced timing is happening at the worst imaginable time...at high loads and high rpms. As any engine builder can tell you, running 3 degrees too much high rpm advance on a race engine is a guaranteed way to seize (or hole) a piston.
While we don't have any iron clad proof that these presumed effects of crankshaft torsioning are an absolute fact, we feel that have enough hands on experiences and supporting information to consider it "a very probable truth". Until some one with more insight and experience can come up with a more probable truth...we'll consider crankshaft torsioning to be a reality.
WHY IS THIS SUCH A PROBLEM "ALL OF A SUDDEN"? - Because late model pwcs have more power, more hull weight, and better turning abilities than ever before. All these features load the drive train harder, and increase the likelihood of crankshaft torsioning. Remember, the first machines to consistently experience this were the 580 Sea Doo's. While those old 580s may not be considered muscle boats, their engines generated lots of power per cc, the pumps hooked up great, and the hull could hold turns at full speed. The early runabouts from the other makers could do none of those things...hence they never loaded the crank hard enough to induce torsioning. In this same vane, high output stand-up boats are so lightweight, and so hard to keep hooked up, that crank torsioning could barely take place. That's why our old "parallel jetted" 701 Raider engine lived in the standup hull, yet seized the rear piston in the runabout hull. The weight and constant hook up of the larger Raider hull loaded the crank harder than any stand up hull ever could.
The apparent effects of crankshaft torsioning are likely among the reasons that Rotax made such a quantum leap in crankshaft "beefiness" for their 782 cc Laydown Rave engines.
OTHER SUPPORTING INFORMATION - Perhaps foremost in this area is the current trends among the boat makers themselves. The '96 ZXi 1100 ignition automatically retards the timing to the rear cylinder after engine temperatures increase beyond a specified level. The '96 Yamaha Blaster II ignition fires 2 degrees retarded at all times, and the rear combustion chamber has considerably less compression than the front. All the 1997 701 models, from Yamaha, also have this "staggered" compression arrangement.
We also suspect that crankshaft torsioning was a consideration for the new reed valve 950 Rotax engine. The rotary valve Rotax crankshafts are, by far, the longest of the 2 cylinder pwc cranks. This length is needed to accommodate the rotary valve diameter and related hardware. A 950 rotary valve twin would need even more length, if the necessary larger rotary valve disc were used. A 950 rotary valve engine with 135 hp, and the large GSX hull, would have certainly experienced unprecedented crank torsioning. The reed valve design would result in a much shorter and stiffer crank, not to mention reduced lower end total length and weight.
WHAT CAN YOU DO ABOUT TORSIONING - In a nut shell, anything that will reduce the combustion chamber temperature of the rear cylinder. For many engines, running slightly richer rear cylinder jetting is enough. Others, with a more serious temperature problem need the richer jetting "and" lower rear cylinder compression. Our testing, earlier this year, showed lower rear cylinder compression to offer no appreciable loss in overall power, along with a significant reduction in rear cylinder combustion chamber temperatures. As a result, many 1996 Group K engine sets have been prepared with slightly staggered compression ratios. (It bears noting that crank torsioning is a much bigger issue for modified pump gas engines than engines running on race gas. Our testing showed that many engines kits would overheat the rear cylinder on pump gas, yet have much more even temperatures on 110 octane race gas. However $4 - $5 a gallon gasoline is a very expensive solution.)
Of course, slightly retarding the ignition timing of the rear cylinder would seem to be the smartest solution. However the electronics of such a device are somewhat expensive and complex for an aftermarket approach. We are made to understand that MSD (the Texas ignition makers) manufactured a few Yamaha total loss racing ignitions, for one of the larger teams, that had an additional "plus or minus 6 degree" adjustment plate for the rear cylinder pickup. We suspect these plates were used to slightly retard the rear cylinder firing of some "tour" engines that were dancing on the edge of the reliability envelope.
In fourth coming year models, some boat makers may choose to drive the ignition off the back of the crankshaft. A design like this could be effective, but it would turn the rear area of the engine into "a very busy place" However pwc engine compartments, as a whole, are certain to become "busier places" anyway.
In time, we believe that all the boat makers will come up with their own special way of dealing this problem on their stock boats. With each new years machines making more power and more hook up, the phenomenon of crankshaft torsioning on high performance pwcs will have to be dealt with somehow.
By Harry Klemm - http://www.groupk.com/