[HOME]

ABOUT AIR LEAKS - The engine you save could be your own

It happens almost every working day. The poor guy on the other end of the phone is giving me every intimate detail about his modified engine in hopes that I'll be able to rattle out the exact cause of his most recent piston seizure. After he finishes his detailed description I always ask, "When was the last time you pressure checked the engine?" The caller usually tells me, "Oh it's always had about 180 psi compression." I explain that a compression test and a pressure (or leakdown) test are not at all the same thing. What he needs is an understanding of what air leaks are and how to check for them.

The theory of air leaks Unlike all 4 cycle engines, the two cycle engines in all personal watercrafts must have a completely air tight lower end. The normal vacuum in the 2 cycle engine's lower end will draw a mixture of fuel and air from the carburetor though the intake manifold. Once this fuel air mixture has been sealed off by the reed valve or a piston skirt, the resulting crankcase pressure moves the mixture through the transfer ports upward into the cylinder. A small "air leak" at a crank seal or an intake gasket will cause the usually perfect fuel/air mixture in the crankcase to become slightly lean...sometimes very lean. As we all know, a lean mixture can quickly result in a seized piston. Air leaks are, by far, the number one cause of personal watercraft piston seizures.

The pressure test Seal off the intake port and the exhaust port with pieces of rubber that are cut to the profiles of the gaskets. Then connect a pump with an in line gauge to the pulse line fitting on the crankcase. Pressurize the crankcases to 6 - 9 psi. Your in line gauge will show the loss of air pressure if you have a leak. The perfect engine will lose no pressure for about 3 minutes. If your gauge shows a leak, spraying the gaskets and seals with WD-40 will quickly expose the exact location of the air leak.

The realities of air leaks If you were to pressure test every personal watercraft engine at your local lake on a Sunday afternoon, you would find about 40% of them to have minor air leaks. Another ten percent will have major air leaks. Despite these potentially fatal leaks, most of these "leakers" will run for some time before they destroy a piston. These same air leaks that would be almost instantly fatal to a motocross engine, are easily tolerated by a watercraft engine...why? The answers are related to H.P. per cc and engine loads.

A modern 250cc motocrosser makes as much horsepower as a stock 650cc JetSki. That means the 250 is making about 250% more power per cubic centimeter than the jetski engine. Since the motocrosser is constantly being operated at close to maximum output and the rpms are being constantly maintained in a high range, the tolerance for an air leak of any kind is nil. A modern motocross engine with a significant air leak, is an engine that will soon produce a seized piston.

Most stock personal watercraft engines are incapable of turning much beyond 6500 rpm. Even at that limit, the amount of power and heat being produced is so relatively low that a small air leak would only cause a little bit of difficult starting and perhaps some uneven or temperamental carburetion. A large and significant air leak can still inflict a seizure on the stock watercraft engine. However the failure would only occur under unusually heavy use.

When this same watercraft engine is modified for greater torque and rpm's, the tolerance for an air leak of any kind is reduced greatly. When the engine is heavily modified for racing level output, the tolerance for any air leak is the same as the motocross 250...None. High output engines that have significant air leaks will usually run fine so long as the rpms are kept low. This accounts for why many freshly assembled engines operate well during the break in period, but seize the first time the throttle is hung wide open. Piston seizures related to air leaks are not just a result of an engine being run with an existing leak, but rather a result of the engine being run at close to maximum output with an existing leak.

Where to look When your engine seizes...don't disassemble it. If your want to cure your seizure problem once and for all, pressure testing should be the very first order of business. The most common and most deadly air leaks take place at the front crank seal and at the inlet manifold gaskets. The front crank seal is not only subjected to the fuels and pressures of the crankcase, but also the intense heat of the ignition compartment. It's very rare for a front crank seal to provide airtight sealing for more than two seasons of riding. Since most ignition covers are airtight themselves, a front crank seal leak can go on unnoticed for a long time on a stock output engine. If you find an oil residue in the ignition compartment when removing your ignition cover, there's a very good chance that your front seal is bad. Under any circumstance, it's good idea to remove the ignition cover or the ignition breather plug when pressure testing an engine.

Intake gaskets, particularly ones with no sealer on them, are always candidates for small air leaks. Since all the fuel/air mixture passes by these gaskets, any leak at all can cause a serious lean condition. If the leak is on the bottom side of the inlet gasket, it will often be necessary to remove the engine from the hull to se and accurately locate the leak.

Air leak fables - Air leaks can cause lower end bearing failures...wrong! Inside a two cycle engine there are two different lubrication needs. The top end desires a deluge of a light viscosity oil while the lower end desires a slow trickle of a heavier oil. When there is a serious lean condition in the engine, a piston seizure will occur long before any lower end bearings begin to suffer. As most victims of air leak seizures have noted during disassembly, there is usually plenty of oil still present in the lower end. The increased loads and increased temperatures of performance modifications mixed with a well worn lower end is what fails most lower end bearings.

Setting the carb slightly richer to accommodate the air leak can avoid a failure...wrong! Most air leaks draw a very minute amount of air at low engine speeds. In fact it's very common for air leaking engines to start and idle very nicely. However when the engine is brought up to full temperature at higher rpms, the castings will begin to expand and deflect slightly in different directions. At these higher temperatures the air intake of a leak can become many times greater. With this huge amount of additional air, even an "over rich" air leaking engine will experience a critical lean condition after only a few moments of full load operation. This accounts for boats that run trouble free for years at moderate speeds with a low skill level rider, yet seize in the first few minutes of being run wide open by an expert. No amount of richness can effectively avoid the failure.

Air leaks can be accurately diagnosed without pressure testing...wrong! Uneven or wavering idling is a common symptom of small air leaks on motorcycle engines. However in the case of watercraft engines, only huge air leaks will result in rising and falling idle speeds. The majority of the smaller, yet equally destructive, air leaks will have no effect on idling. Another common practice is the spraying of contact cleaner at potential leak areas while the engine is running. If the spray enters an air leak area, the idle speed will increase noticeably. This method is useful for large leaks in accessible places, however the majority of smaller hidden leaks can pass this test.

Freshly rebuilt engines are free of air leaks...wrong! Even engines that are rebuilt by the best technicians can air leak. About 20% of all freshly rebuilt engines will have small hidden air leaks. That's why most professional engine builders pressure test every engine before and after rebuilding. Installing fresh seals and gaskets is by no means a guarantee against air leaks.

Making the test gauge - The construction of a pressure test gauge is pretty simple. The photo shows a 0 -15 psi gauge, a standard air stem fitting, and a 1/4" hose barb all mounted in a brass "T" fitting. A hose from the barb fitting will connect to the pulse line fitting on the crankcase. Also shown are the rubber gaskets that will fit under the carb and the headpipe to seal off the engine. For those who would rather buy a finished kit, Group K sells an air leak test kit that has the rubber blockoff material for $30.00.

[Back to Home Page]