Cavitation (aka Vapor Lock) Mysteries Revealed by Jack Letts

Summer is upon us and the classic pilot complaints and reports of fluctuating fuel flow have started. Last summer I wrote three long posts in BeechTalk about cavitation. Together those posts formed a primer on Cavitation and I think they offer a nice background to help understand what's actually going on inside the pump and fuel lines.

Cavitation is often referred to as Vapor Lock when discussing engine fuel systems, but I don't like the term because "Lock" is misleading; it implies that the system is locked and that no fuel is flowing. In fact, the system is working just at diminished capacity, it's not locked.

First, let me expound a little on why the fuel flow falls off with altitude. I've been in the "pump business" for 25 years so it's in my wheelhouse.

I think George Braly's Socratic teaching method is great. Give a little hint then let the student figure it out, in figuring it out the student gains a deeper understanding. For those of you who would just like to have it explained to you let me take a swing at it.

Pumps need atmospheric pressure in order to work. Pumps that draw a suction lift, like the engine driven pumps on our beloved Continentals that must draw the fuel up out of the low wing tanks in a Bonanza are even more dependent on atmospheric pressure.

Picture when you are drinking through a straw. Most people think they are "sucking" the beverage up into their mouth. In fact what is happening is that you are lowering the pressure in the straw and atmospheric pressure is pushing the tasty beverage up the straw. If you can picture somehow trying to drink through a straw in a vacuum, you can easily see that it wouldn't work. No differential pressure, between the surface of the liquid and the inside of the straw would mean no fluid movement.

At sea level, standard atmospheric pressure is 14.7 PSI, expressed another way that all us pilots know, 29.92 inches of mercury. If you want to keep thinking of a column of water like in your straw it would be about 33 feet.

As you climb, obviously you have less atmospheric pressure. By the time you get to 15,000' you are down to about 1/2 the sea level pressure. So what happens, why is the pump now putting out less flow at the same pressure? In a word CAVITATION.

Cavitation is a word that is used regularly but I find that it's poorly understood. To explain cavitation I like to use the swimming pool example. Imagine you are standing in the shallow end of a swimming pool and you take your hand and you move it as fast as you can just under the surface. You'll see a whole stream of bubbles appear behind your hand and they will immediately disappear as soon as you stop moving your hand. Your hand did not break the surface so the bubbles did not come from the atmosphere. Where did the bubbles go? they didn't float up and break the surface like air bubble, what are they? It's cavitation.

What you have done is put more energy into the water than it has affinity for itself. You've created little vacuum, or more accurately steam bubbles that slam back shut as soon as the energy that created them stops.

Now here's the next interesting observation. If you put on your scuba gear and dive down to 33' feet, you can't create those bubble that you easily create just a few inches below the surface. You are not strong enough to make the water cavitate at 33', no one is. (I picked 33' because it one additional atmosphere.) If you could find a swimming pool in a Himalayan village at 15,000' you would find it extremely easy to create cavitation bubbles with your hand because it would be only under 1/2 an atmosphere of pressure. We know from this thought experiment that it's harder to make a fluid cavitate at higher pressure, and it's easier to make them under lower pressure.

The next variable in the cavitation equation is the vapor pressure of the liquid. Water has a very low vapor pressure as compared to most liquids including gasoline because of the polar nature of the water molecules. Water is attracted to itself and resists cavitation. This is the same property that accounts for surface tension and capillary action. Gasoline does not have these properties and has a high vapor pressure meaning that it's easier to create cavitation bubbles. Gasoline's vapor pressure increases when it get hotter making it even easier to cavitate. Also gasoline vapor pressure can vary quite a lot depending on the batch and still be in "spec".

We went over the importance of atmospheric pressure and how it's essential for getting the liquid to the pump and how it's even more important when the pump is pulling a suction lift. Then I gave a little dissertation on cavitation and how it's affected by pressure and the property of the liquid called "vapor pressure".

As the plane climbs there is less and less atmospheric pressure available to push the fuel up to the pump. The pump is still creating low pressure in the suction line, but since there is less atmospheric pressure available the fuel can not get to the pump as easily. At some point pressure drops below the fuel's vapor pressure and cavitation bubbles begin to form inside the pump and in the piping on the suction side. It's like moving you hand through the water in that imaginary 15,000' Himalayan swimming pool. The same amount of energy that wouldn't form cavitation bubbles at sea level will easily form them at altitude.

The formation of the bubbles is exacerbated by the fact that fuel has a relatively high vapor pressure and will form cavitation bubbles easily. It's further exacerbated if the fuel is hot which raises the vapor pressure further .

The bubbles are formed in the pump and go into the pump from the suction piping, but they immediately collapse on the discharge side because of the higher pressure. What's happening is that the bubbles are occupying a volume that would normally be occupied by fuel. The volume of the bubbles displaces the fuel and results in a lower flow rate. Pretty simple actually.

The electric fuel pump is in a better position with regards to cavitation because it sits lower in the plane and has a flooded suction. When you turn on the electric fuel pump it restores some or all of the lost atmospheric pressure and keeps the pressure in the mechanical fuel pump and suction lines high enough to avoid the cavitation bubbles.

Now there is more going on here. For example, I don't understand why the electric fuel pump doesn't always restore full fuel flow. According to some, the electric pump has to be built so that it performs at the top of its spec or it won't fully solve the problem. I'd think the electric pump would easily restore the pressure and quell the cavitation. I also don't fully understand some of the fluctuating flows and pressures that some pilots are reporting. It may be that the cavitation bubbles are getting big enough to cause the pump to momentarily lose prime.

As I have been thinking about it there are a couple more things are going on that might be helpful if explained a little further. In the earlier sections, I went over the fact that it's actually the atmospheric pressure that pushes the fuel up to the pump. As you go up in altitude you have less atmospheric pressure sort of holding the fuel together and at some point, the low pressure in the suction line causes the fuel to form bubbles, that is what we call cavitation. It would be accurate to say that the fuel is boiling, but I don't like to use that term when explaining this because boiling implies heat. We live in a sea level atmospheric world and we generally forget that a fluid will boil at any temperature, the boiling point is established by the fluid's properties and pressure.

When fluid is moving through a pipe it creates turbulence. We all know, because we are pilots, that turbulence creates drag. For the purposes of understanding this, you can just think of it as drag, and the drag is further lowering the pressure in the suction pipe and at the inlet to the pump.

Just like on an airplane any obstructions create further drag. In this context the obstructions are things like elbows, valves, strainers, or fittings that the fuel passes through. All these things are lowering the pressure in the suction line and bringing the fuel closer to its vapor pressure with the resultant bubbles and loss of flow.

As a pump expert, when I look at the suction pipe arrangement on a Bonanza, it's a total mess. What you want is a short straight pipe with no obstructions. What we have is a long pipe with a bunch of obstructions; the coarse fuel strainer that is actually in the tank, the selector valve, the gascolator and the screen, the electric fuel pump, and a whole assortment of elbows and fittings thrown in for good measure. All these things are working against us and adding to the cavitation problem.

Some obstructions are worse than others. Obstructions that are close to the pump are much worse than obstructions that are far from the pump. For example a 90° elbow right at the inlet of the pump would be a disaster.

Why? you ask. Whenever the fluid passes an obstruction a low pressure area is created just downstream. Think Bernoulli. If the fluid is close to its vapor pressure the cavitation bubbles will form at the low-pressure spot just past the obstruction. If this low pressure spot is far from the pump, depending on all the factors surrounding pressure and vapor pressure, the bubbles are likely to close back up before reaching the pump. If the low pressure area is right in front of the pump, the bubbles are likely to remain intact and enter the pump resulting in the loss of flow for the reasons we've already discussed.

Obviously, our Bonanzas are certified aircraft and you just can't start changing the suction piping around for kicks. Most of the problems inherent in the suction setup are baked in the cake. However, there is some variability in the hoses and fittings between the firewall and the pump.

Here are a couple of areas to examine:

Make sure the hose that leads from the firewall to the pump is in perfect condition. If it's a Teflon hose it's probably Ok and you can easily inspect it. If it's an old hose lurking under a fire sleeve it may be slightly crushed or degraded. This is the most important part of the fuel suction setup since it's close to the pump and needs to be tip-top.

Try to keep the fuel hose away from anything hot. Heat lowers the vapor pressure of the fuel and takes it that much closer to cavitation. Even if I had a Teflon hose, I'd consider putting a fire sleeve over it to help insulate it. Also, check and make sure the cooling air duct for the pump is in place and in good shape. The object of the duct is to help cool the pump, which fights off cavitation tendencies.

Try to eliminate all the fittings between the firewall penetration and the fuel pump. Try to have just one hose from the firewall penetration to the pump with no fittings. Have the hose make long radius bends. If there is a 45° or God forbid a 90° elbow actually attached to the pump find a way to get rid of it. You want the gas to flow into the pump with as long a straight run as you can manage. I don't know what's possible here, however, my Bonanza is NA so I'm not trying to pump 32 GPH to the engine at 15,000'. My pump is probably cavitating when I'm at high altitude, but I can still get 12 -13 GPH to the engine so I'm not worried about it. As a result, I haven't really looked at the suction piping. But the TN folks may need to pay special attention.

All of the problems with cavitation are on the suction side of the pump and in the pump itself. Once the fluid is through the pump and the pump is "pushing" it, cavitation is no longer a factor. Newtonian liquids are incompressible so you can push on them as hard as you want, you just can't pull them.

The mechanical fuel pump used in our Continental aircraft engines is a Gear Pump and it falls in the family of pumps called Positive Displacement pumps. Positive displacement pumps all share a common trait, they create a cavity then crush the cavity.

This gives all positive displacement pumps two common qualities. First, their flow rate is set by the RPM. The volume of the cavities is built into the pump meaning that it will pump a certain amount per revolution. So, in order to change the flow rate, you must change the RPM. Second, there is no theoretical limit to how much pressure they can generate. Since liquids are incompressible, the pressure generated when you crush the cavity is only limited by the strength of the materials, and horsepower available. That's why there is always some pressure relief mechanism on a Positive Displacement pump. And, that why we have a fuel return line.

For any everyday example of a Positive Displacement pump (PD) think of a pressure washer. When you let go of the trigger, the pump is still pumping. The flow is just going through the pressure relief valve and is piped back to the suction. If there were no pressure relief valve the pump would continue to build pressure until something gave; the engine might stall because it just couldn't make any more power, or something might break or slip like a shaft or a belt. The pressure would try to go to infinity and it would only stop when if finds the weakest part of the machine.

Our pumps are always pumping the same amount of fuel at a given rpm (assuming no cavitation). I don't know what it is, but let's just say it's 40 GPH at 2500 RPM. So, if you are flying along at 2500 RPM burning 13 GPH, that means that 27 GPH is going back to the tank. Push the throttle or mixture in so now you're burning 25 GPH, now 15 GPH is going back to the tank via the return line. The total amount of pressure is set by the pressure relief valve, not the pump. The pump could make lots more pressure as long as the fuel is getting to the pump.

And there we are, back to cavitation. The problem is that the fuel isn't getting to the pump, it's boiling and creating bubbles. The bubbles in the pump are occupying space that would be occupied by fuel if it weren't cavitating. Result, the flow drops off.

This leads me to my last point. I consider myself to be somewhat of a pump expert, but I'm not a Continental fuel injection expert. I have a good working knowledge of how the system works, but I find myself occasionally puzzled.

The most recent example of my puzzlement is in a recent post that says "My A&P has also talked to TAT and I understand that there is a larger orifice that can be installed to help with this, somewhere in the engine." I can't think of any reason an orifice would need to be installed anywhere in the system. If the orifice is on the suction side of the pump, then it's in the perfect position to create lots of cavitation bubbles in exactly the wrong place. If anyone can explain what this orifice does I'd be interested in hearing about it

There you have it, my musings as a "pump expert" regarding cavitation, which we pilots have always referred to as "Vapor Lock". I hope this narrative has helped improve your understanding of our fuel system and the causes of "CAVITATION".

Just to add a bit more content to this narrative, I thought I'd post this video showing FF fluctuations in a NA Bonanza at altitude with a high OAT.

I shot this video last summer at 12,500' with an OAT of 75 degrees. While 75 degrees may not sound hot, it's unusually hot for that altitude. Boost pump was off.

So what's going on here, why is it doing that? If you read the previous parts of this narrative  you've probably figured out that its cavitation, but why the bouncing?

It's actually fairly easy to predict cavitation, pumps generally have a curve that shows the pressure where the pump will start to cavitate. You have to make corrections depending on the vapor pressure of the liquid, but it's all straightforward engineering work. What's hard to predict is exactly what it will "look like". Will it be a straight drop off or will it pulsate? That depends on the exact piping and pressures.

Say you are working with a bigger water pump that's electric driven, something where you can actually hear the water flowing through the pump. It will sound like you are pumping gravel.

As an experiment, say you want to force the pump to cavitate so you can see what happens. You can do this easily by simply closing the suction valve partway. Pretty quickly you will start to hear the gravel sound. As you continue to close it, you will find places where the discharge pressure fluctuates wildly and the needle is bouncing all over the place. What's happening is that vortices are setting up in the pump. The vortices act like a little temporary plug that drops the pressure even further. When the flow rate/pressure drops it changes the conditions that set up the vortices and it breaks up. Then the conditions are back to where they were and the vortices set up again. So it's the vortices setting up and dissipating that's causing the fluctuation. The fluctuations can be from several times a second to several seconds or even longer.

When you see a fluctuating discharge pressure gauge (that's what the Beech factory gauge is after all), you have an excellent indication of cavitation or a problem on the suction side of the pump. So we suspect cavitation because of the temp and the altitude, the bouncing FF confirms it. Also, turn the boost pump on and re-lean and it's gone, case closed.

Now the real question is "is it hurting anything"? After pondering this for a while, I don't think so. As I pointed out in an earlier post, the energy level in our little 1/2 gallon a minute pumps is quite low. I don't think the cavitation bubbles that can be so destructive in larger high HP pumps are actually damaging anything in our little pumps. I'm sure the spray pattern at the injectors is oscillating, but apparently, it's averaging out because the engine is smooth.

Of course, you can stop the bouncing needle right now by turning on the boost pump, but that has its own risk.

On most NA planes you just have a single speed pump and the single speed is high. If you turn the pump on and re-lean at 12-13,000' you must remember to turn it back OFF when you descend. Otherwise, if you go full rich when you perform GUMPS or initiate a Go-Around, it may be so rich that the engine quits. This could be a very bad if you go full rich for landing and it quits while you are at a low altitude.

I've decided to just let it bounce unless it actually starts to affect the engine, which I have never seen.

The TN guys have another set of issues as they try to keep 35-36 GPH flow rate all the way to 18,000'. They must use the boost pump on low then high to keep the cavitation at bay. Even with the boost pump on high they often can't keep the high fuel flow at altitude.