Into the Pump
By Kevin Cameron
Personal watercraft are propelled by a waterjet - a low-pressure water pump forcing water through a nozzle ring to form a rearward-directed propulsive jet. In order to force this jet rearward, the pump pushes itself - and the attached watercraft plus yourself - forward. This is the same process that propels airplanes, except that a propulsive jet of air is much lighter than one made of water - by a factor of 830 to one. This is why air propellers must be so large - in order to move the necessary large volume of air. In the case of jet engines, the mass of air accelerated is less, but jets make up for this by their very high exhaust velocity - up to 2500 feet per second. The heart of the waterjet pump is its impeller - the beautifully shaped propeller-like rotating element of the pump. Typical diameters for this impeller range from 138mm to 155mm, or about 5.4 to 6.1 inches. The impeller rotates inside the pump duct, the tips of its blades very nearly touching the inside of the duct in order to limit leakage loss. Because the rotation of the impeller imparts some rotation to the water flowing through it, slightly angled nonrotating vanes are placed in the flow to cancel this rotation. The job of the impeller is to convert the mechanical shaft power of the engine into the fast-moving jet of water that propels the craft. How fast? Usually jet velocity is made about double that of the craft. Because a high-end watercraft speed of 60 mph is 88 feet per second, you would expect the waterjet velocity to be about double this, or 176 fps. A rough relation between pump pressure and jet velocity is:
Pressure (in psi) = (jet velocity, squared)/145
For example, the above jet velocity of 176 fps translates to 213 psi. Just thinking about watercraft weight, power and speed reveals that a lot of energy is getting lost somewhere. A 150-hp engine should be able to move 600 pounds of watercraft and 200 pounds of rider a lot faster than a top speed of around 60 mph. Even powering a small car weighing over a ton, you'd expect a top speed of at least double that 60 mph. Where does all the power go? A lot goes into forcing the modified deep-V watercraft hull through the water - even when the boat is on plane, enough hull has to be down in the water to keep the pump intake submerged. More energy is wasted in friction between the impeller blade surfaces and the water through which they are slicing at speeds up to 160 feet per second. More loss occurs between the moving water and the inside surfaces of the pump ducting. More yet is carried away by the waterjet itself, because it is moving backward roughly twice as fast as the boat is moving forward.
You can think of a waterjet impeller as a screw whose rapid rotation gathers up water at its entry face, increases its pressure, and pushes it toward the restrictive nozzle ring. At the nozzle, pump pressure is converted into velocity in the waterjet. As you'd expect, this requires a difference in pressure between the forward and rearward faces of the impeller blades. Rotation of the impeller plus the wedge-like angle of the blades causes water to be accelerated rearward. On the forward faces of the blades, a much lower pressure exists. This is determined by how deep in the water the pump intake grate is located, the atmospheric pressure above the water, and any ram pressure or dynamic head generated by the forward motion of the boat. If the pressure at the face of the impeller falls low enough, water may not be able to enter the front of the pump as fast as the impeller is pushing it out the back. In an extreme case, this intake-side pressure can fall low enough that the water is pulled apart - cavitated - as it flows onto the impeller. Cavitation is a problem because (a) if the pump is partly pushing empty space instead of solid water, thrust is reduced, and (b) when such cavitated regions collapse as they move to higher-pressure regions of the pump, enough sudden pressure can be generated to damage metal surfaces. Light cavitation damage looks as if regions of the impeller vanes had been sand-blasted. In extreme cases, such damage looks as though the metal had been eaten away by an acid attack, with large areas of blade actually missing.
Cavitation bubbles are not a complete vacuum but are composed partly of gases that had been dissolved in the water and partly of water vapor (cold steam). During significant cavitation engine rpm will rise (because it is less work to pump bubbles than it is solid water), and the pump will Òsound funny.Ó Cavitation happens most easily at low boat speed, because the pump's intake grate is not moving rapidly enough through the water to generate intake pressure that can suppress cavitation.
Impellers are listed by manufacturers according to their nominal pitch. As applied to a conventional propeller, pitch is the distance the prop would advance through the water, without slip, in one revolution. Pitch is more complicated for impellers because it is not constant, but normally increases from leading edge to trailing edge. Thus, an impeller's pitch might be given as 14/19¡, meaning leading edge pitch is 14, and trailing edge pitch is a steeper 19. The reason for this increase is that cavitation is most likely to occur at the outer diameter (OD) of the leading edge. Reducing the pitch somewhat at the leading edge makes the impeller ÒbiteÓ better - that is, not cavitate. As water pressure increases along the length of the impeller, cavitation becomes less possible, and so the pitch can be increased toward the blades' trailing edges without risk.
Pitch is measured at the outer edge of the blade. Because a given blade element moves faster the farther it is located from the rotational axis, for constant effect blade pitch must increase as you move from the impeller OD inward toward the impeller hub.
As pitch is increased, the impeller pushes water rearward farther per turn of the impeller. Naturally, it takes more engine torque to do this, so the more powerful the engine is made, the more the impeller's pitch may be increased to make that extra power pump more water. Clearly if the impeller pitch is too small, the engine will overrev at top speed, and vice versa.
Impeller manufacturers make the point that one maker's 138mm 14/19¡ impeller may give results quite different from an impeller for the same watercraft and with the same pitch numbers but made by another maker. Every maker publishes application charts, listing the suggested impeller for each watercraft model in three conditions: stock, limited mods and modified. Typically, an engine with limited modifications requires little or no impeller change, while a modified engine may pull another degree of front and rear pitch. Application charts are most easily found on the Internet but can be requested in printed form as well.
With the exception of turbocharged engines, watercraft powerplants lose power at higher altitudes because of decreased atmospheric density. To compensate for such power loss, it is normal to reduce impeller pitch slightly, thus allowing the engine to operate at its normal rpm rather than be pulled down by pitch that is too much for its reduced power. When you look at a variety of impellers, certain visual differences are obvious. The first of these is between straight and swept-back blade leading edges. Swept-back leading edges have a rounded shape more like that of a conventional marine propeller blade. The difference has to do with increasing the impeller's resistance to cavitation in marginal intake conditions. The rounded leading-edge shape - the so-called swirl impeller - results when those areas of the blade most likely to cavitate are cut away. Indeed, in the development of both propellers and impellers, one technique is to cavitate the unit deliberately and then cut away the parts of the blade on which cavitation effects can be seen. The result is an impeller better able to deal with rapid acceleration from low speed and with marginal pump intake conditions that can arise during vigorous maneuvering.
Operation in very shallow water can lead to junk going through the pump - and impeller damage. Nicks and dents in impeller blade leading edges provoke cavitation and thrust loss, so prop shops offer straightening services and/or welding to replace blade material that has been lost. Many prop shops also offer blade reshaping that might offer increased or differently optimized performance.
You have probably noticed variations in how much total impeller blade area there is. Impellers from Nujet generally have narrower blades than most, while the blades of other makers are often much wider. This is a compromise made between top speed capability and the impeller's resistance to cavitation at low speeds. The more blade area there is (this is sometimes called solidity), the smaller the pressure difference between pressure and nonpressure surfaces can be. This naturally tends to make cavitation less likely. But at the same time, increased blade area generates increased skin friction, which can limit top speed. It's hard to have it all - high top speed, good bite from lowest boat speeds, and high tolerance for rough water and vigorous maneuvering.
To give a better idea of the importance of skin friction, consider the difference between a 200-hp PWC and a 95-hp 250cc outboard hydroplane. The PWC may have a top speed close to 65 mph, but it also has a lot of wetted area - its hull and inside its pump ducts and in the form of impeller and vane surface. The hydro, by contrast, has risen much farther out of the water, until it is planing on only a very few square inches of hull surface. Almost nothing else is in the water save for the prop and its drive gears in the lower unit. Because of this reduction in wetted area, the outboard hydro can run nearly 100 mph on half as much power.
Therefore the user must decide what kind of performance is desired. If it is maximum possible top speed on flat water, the Nujet style of impeller may be your choice. If you spend more time in marginal intake conditions, you may want the increased cavitation resistance of the swirl-cut impeller of high solidity. If you are a racer, you may switch impellers according to conditions - the slower the turns on the course, the more important it becomes to compromise in the direction of cavitation resistance. In the end, it comes down to experiment, because although the impeller application charts provide a useful baseline, you the user are a unique application.
You can see from the ads in magazines that nozzle rings are offered in different diameters, usually given in metric measure - as "85mm," for example. For those who really want to get into the variables, changes in nozzle ring diameter can move the boat's speed of best performance up or down. Above it was noted that usually waterjet velocity is set to be roughly double the speed of the boat. If it were desired to increase efficiency at lower speeds, a larger nozzle ring might give better acceleration than a smaller one, while a slightly smaller one might eke out another mile per hour on top. Chuck Yeager, the Air Force pilot who was the first person to exceed the speed of sound, once played a Ònozzle ring trickÓ on his commanding officer. The new F-86H jet interceptor had just been released, and Yeager and his CO went up to test it - Yeager in the "old" F-model and his CO in the hot new H. When they opened up, the CO was stunned to see that Yeager's older airplane slowly pulled away and left him behind. Only later did he discover that Yeager had persuaded his crew chief to move the stops on his engine's variable exhaust nozzle, allowing it to become slightly smaller than normal (at some cost in increased turbine temperature). The resulting increase in exhaust velocity gave Yeager's aircraft more high-speed thrust, increasing its top speed.
PWC manufacturers naturally want their products to perform as well as possible, so stock pumps and their impellers are made to do a good job. But when you want something different to suit special conditions - whether that be engine modifications, a changed altitude or just a desire to experiment - the aftermarket is waiting to tempt you with many alternatives.