Cavitation damage fuel pump - return top gas tank

[skywalker]

Hello people, let me first introduce myself. My name is Luuk, 40 years of age and I am from the Netherlands. I own a '73 super and a '67 split bus. The 1303S has a stock 1600 engine, the split a 2.2 type 4 with fuel injection. Both cars are for road use, not for the circuit or drag racing.

The split bus is consuming a lot of fuel pumps, something it should not do. In less then 2000 km (1250 miles) I recently mounted its fifth fuel pump. Asking friends to solve the problem got me pointed to this forum.

Of course I did some research why the fuel pumps brake down. The fuel pump that I use generates 130 litre/hour, max 6 bar. The fuel pressure regulator is set at 4 bar. After the first fuel pump broke down I replaced it by a new one. After that one broke down I replace the pump and the fuel filter. When the third one also broke down I searched a little harder and did I remove the -stock- filter in the fuel tank. It reduced the free outflow by approximately 150 litre/hour to 106 litre/hour: Without the tank filter I measured a free outflow of approximately 255 litre/hour. The last pump had some dirt in the pump itself (because of removing the filter in the tank). I cleaned the gas tank and replaced the fuel pump and fuel filter so now its running again and I want to keep it running this time.

Researching the cause I disassembled a fuel pump and I did found cavitation damage in the inlet of the fuel pump. Now I'm looking for the cause of this cavitation damage. So far I can only think of two reasons:

1) lower pressure in the supply line caused by obstruction of the standard tank filter and, by the last fuel pump, dirt in the fuel pump and heating up the fuel when driving long distance (fuel pumps brake down after longer trips). Lower pressure en the heat may result in gas bubbles which implode entering the fuel pump.

2) the return of the fuel line is in the top of my gas tank. When fuel is returning in the gas tank air bubbles are created. These air bubbles might become in my fuel pump and implode there because of the higher pressure. Heating up the fuel during longer trips promote gas bubbles.

At point two there are different meanings. Some say I should lower the return tot the – or make it in – the bottom of my gas tank, some say it is perfect as it is. No one has convinced me yet on what is the best, so I hope you can. Maybe I'm misjudging the whole situation and am I wrong in my thoughts, please share your thoughts with me. Specs of the engine and fuel system are below.


Engine specs:
case: VW / Porsche type 4, W

crankshaft, Porsche 914, 71mm
cylinders, Porsche 964 nikasil
bore, 100,00 mm
stroke, 71,00 mm
total engine displacement (four cylinder) 2230.53 cc
compression ratio, 9,22:1

camshaft, Web Cam 163/86b, 284/300°, 12.7mm
cam duration exhaust 300°
cam duration intake   284°
cam duration exhaust @ 0.050" lift 260°
cam duration intake @ 0.050" lift250°
cam lift 0,500"

heads, Porsche 914
valve diameter exhaust, 38 mm
valve diameter intake, 44 mm

exhaust, custom made 4-2-1 system

DTA Electronics E-48

throttle bodies, Jenvey, 2x2x45mm   
injectors, Bosch, 225cc/m
fuel pump 130 l/h max 6 bar
fuel pressure regulator set at 4 bar

170 HP / 237Nm at 4200RPM

Fuel system:


Note:
- gas tank vent is OK

[BeetleBug]

Welcome to The Lounge Luuk!

After looking at your drawing I would:

1) Use a pre-filter before the pump
2) Lower the return in order not to create any bubbles

Best rgs
BB

[Lt Charbon]

Hello,

If you want to reduce cavitation issues, you have to diminish the loss of pressure at the entry of the pump. In order to do that, you have to relocate the pump closer to the tank to diminish its lenght or replace the fuel line before the pump by a larger one. You can also diminish the flow of the pump or completely clean the tank and the fuel lines in order to prevent the dirt to stay in the line.

This is my point of view as a firefighter, maybe there are better idea...

[Martin S.]

What fuel pumps are you using? Mine was expensive so I would hate buying five! My fuel pump starts whining because it's cavitating especially when hot and I don't like that sound, but my mechanic assures me that it won't hurt the pump (!). He says the fuel is boiling and bubbling. He's got a plan to fix the plumbing to minimize the cavitation, but that's not a priority now. There was a small inline fuel filter on the system but it plugged easily so he changed the filter to a Porsche 356 inspired oil filter setup that should NEVER plug up. I have several thousand km on the setup and drive 45 minutes to work everyday and have had no problems with the fuel pump or engine at all. Here's a pic showing the fuel pump and filter. Return line is the small stock one thru the tunnel and supply is a 3/8 aluminum running along the bottom of the floor outside the heater channel.

[BeetleBug]

Return line is the small stock one thru the tunnel and supply is a 3/8 aluminum running along the bottom of the floor outside the heater channel.

Your return line should be bigger or at least the same size as the supply line.

Best rgs
BB

[Martin S.]

Haha, yes upgrading that is on my bucket list for sure. It may stop the cavitation. I'm having more fun just driving the car as is than fussing over the details. The weather has cooled off so HELLO TURBO!

[dragvw2180]

 Fuel pumps are made to push fuel not suck so the distance from the tank to pump is very important. I would install a larger fitting in your tank to increase the size of the feed line to the pump ( might even relocate the fitting in the tank  ) and keep the distance short as possible . I would also make sure your vent hose is big enough for the volume of fuel you are pumping . If excessive air is the culprit I would look at fuel/air seperator in this video,   <a href="http://www.youtube.com/v/Kqt6hOPDK3k" target="_blank">http://www.youtube.com/v/Kqt6hOPDK3k</a>    Hope a little of this helps solve your problem, Mike McCarthy

[skywalker]

@BeetleBug
- A pre filter is a good thought, I already have one. I did not install that filter because it reduces the free outflow to 140 litre/hour. So it holds up more than 100 litre an hour. I will order an high-flow 100-micron fuel filter soon (I'm considering other changes to the fuel system which I will reveal later).
- You are right about lowering the return line to the bottom will not create bubbles. But is it the reason of braking down the fuel pump?

@Lt Charbon
Correct. The supply line is 12mm and so is the connection at the entry of the pump. The supply line is about 20cm and the the pump is under the fuel tank, so I think that's not causing the problem.

What fuel pumps are you using?  Mine was expensive so I would hate buying five!

The first one was a Bosch, the second one a Kolbenschmidt Pierburg and the last three are from Electrix.

My fuel pump starts whining because it's cavitation especially when hot and I don't like that sound, but my mechanic assures me that it won't hurt the pump (!).

Interesting allegation... Besides the cavitation damage all of my broken fuel pumps have play, when I shake them I can feel and hear it (like the whole pump goes up and down inside the housing). I am assuming this is caused by the heat the cavitation is producing.

[skywalker]

Thanks Mike! That is an interesting video!
The stock fitting (6 mm) in the tank is already replaced by a larger (8mm) one. As mentioned in my previous post the supply line is short.

[dragvw2180]

 It is probably large enough but this is more like I was talking about, the AN 8 un it,     http://www.piersideparts.net/CB3288.html     plus a matching fittig in the pump . I believe the line I have to my pump is AN 10 .

[Martin S.]

I googled cavitation damage in fuel pump and mostly jet engine stuff came up. Does your fuel pump make the whining noise like mine does when hot? Is it well known that cavitation destroys fuel pumps quickly? Why is my pump ok with cavitation (at least for now!) and yours doesn't last?

[skywalker]

Nice product I can use Mike! Thanks! The 8mm fitting I have never pleased me, but for lack of better I use this one.

I googled cavitation damage in fuel pump and mostly jet engine stuff came up. Does your fuel pump make the whining noise like mine does when hot? Is it well known that cavitation destroys fuel pumps quickly? Why is my pump ok with cavitation (at least for now!) and yours doesn't last?

The cavitation is something is discovered after pump 4 broke down. I am not sure if it makes the noise, I never noticed, but I also never listen to it specific.  I've seen video's of fuel pumps that cavitate, so I know what kind of noise you mean. I am not sure the
cavitation is destroying the fuel pumps, it's something I assume. Like I said in my first post, maybe I am misjudging the whole situation and am I dead wrong.

The pictures of the damage:
 

[spoolin70]

Have a read of this :

 http://fuelab.com/forums/topic/avoiding-cavitation/

They have a good Q&A forum too

Hope it helps
Darren

[hotrodsurplus]

A number of people have already revealed possible solutions to your problem but I'll sum them up and include my own observations. My experience is with power-steering assemblies, a Jr. Fuel dragster, and engineers that I've consulted for technical articles. The mechanical theories apply pretty much universally to all pump systems whether fuel or, as the Lieutenant indicated, water.

First, understand what cavitation is. Cavitation is the spontaneous creation of vapor (bubbles). Scientists call this the vapor point. The rest of us call it the boiling point. All three terms are roughly interchangeable. All fluids will vaporize/boil at very low temperatures if you give them the right environment.

Pressure influences a fluid's vapor point. Water is a good example. The atmospheric pressure at sea level prevents water from vaporizing until 100C/212F. The reduced atmospheric pressure at the top of Mt. Everest will let that same water boil at 68C/154F. And water will vaporize at very cold temperatures if you reduce the pressure sufficiently. The swirl trailing a boat propeller is vaporized water. That can occur at freezing temperatures which proves that vaporization/cavitation/boiling is possible in situations that seem highly unlikely.

The feed line is almost always the culprit in vaporization/cavitation issues. An insufficient feed line will reduce the pressure at the inlet side of the pump to well less than atmospheric pressure. Now consider that some components in gasoline can vaporize at temperatures as cool as 38C/100F. So cavitation almost always indicates a restriction in the feed line.

By the way, the 'whine' that you hear in a cavitating pump, whether fuel or power steering, is millions of vapor bubbles clapping shut when they reach the outlet side of the pump. While the mechanics are different the consequences of those bubbles clapping shut are very similar to the consequences of detonation. The pressure wave created by each bubble clapping shut is akin to a tiny hammer striking the pump body and gears. That's why the inside of a pump that suffers cavitation looks like the surface of a golf ball. The bubbles clapping shut also release a lot of energy in the form of heat. Heat will increase a fluid's tendency to vaporize/boil/cavitate. So cavitation creates a vicious cycle that ends in component failure.

Oh yeah, the 'swooshing' sound that accompanies the swirl in a propeller's wake is the sound of those bubbles collapsing. The groan you hear when you lock a power-steering system is the sound of those bubbles clapping shut. True, pumps do sort of whine when running; however, that's a mechanical noise that differs greatly from the groan or whine from cavitation/vaporization.

Here are tips to reduce/eliminate cavitation.

1. Use the largest possible diameter line to feed the pump and make that line as short as possible. Don't be afraid to use a line that's far bigger than the pump's inlet. The same goes for the fitting in the tank. It has to be large enough to flow what the pump requires. The line that feeds the pump should be the biggest and shortest line in the system. Think of the feed line as a tank in its own right. It is impossible to make a feed line too big or too short. The bigger and the shorter the feed line, the better the system will work.

2. DO NOT use any sharp elbow fittings (ell or 'street' or 'service' fittings) in the line from the tank to the pump. This applies to AN/JIC fittings as well. If you must use an angle fitting then choose one that distributes the bend over a large an area as possible (common tube-type fittings in AN/JIC land). And use fittings that have very large inside passages. Inexpensive fittings often have very small internal passages.

3. DO NOT install the filter in the line from the tank to the pump (I see by your diagram that you don't anyway). Instead, use a very large screen to protect the pump. But NEVER a filter between the tank and the pump.

4. Install the filter in the high-pressure line, preferably as close to the injectors as possible.

5. Make the high-pressure line the smallest line in the system. Pressure increases a fluid's flow capacity so even very small lines will flow far more fluid than you'd expect. I don't have the quote handy but several engineers and tuning specialists told me that a 5/16" or 8mm high-pressure feed line will support 400 to 600 horsepower in an injected application. An oversized high-pressure line can also aggravate a cavitation issue. A little bit of resistance in the high-pressure line is a good thing so long as the line can deliver sufficient fuel to the engine.

6. Make your return line with hose/tubing at least one size larger than your high-pressure feed line (two sizes if it fits). Remember, after the fuel leaves the regulator it enters an environment with slightly more than atmospheric pressure. Therefore, it will not flow as freely as it will in the high-pressure feed line. Any pressure in that return will also skew your tuning. Finally, understand that fuel can form bubbles as it leaves the regulator and enters the low-pressure atmosphere in the return line. A return line larger than the feed line will give the millions of tiny bubbles an opportunity to coalesce into larger bubbles before they reach the tank. So bigger is better, albeit not as important as the feed line.

7. Sump your fuel tank. This is very underestimated but there's a reason that the OEMs build tanks with complex baffles or sumps. If the fuel sloshes away from the pickup tube then the pump will run dry. Fuel soaks up the heat that a pump generates and when a pump runs dry it runs hot, even if only occasionally. So create a narrow depression in the bottom of the tank so the inlet is always submerged even during hard acceleration or cornering (aka, a sump). If you can't sump your tank then fit a swirl or surge tank in the feed line between the tank and pump. There is lots of information how to do that on the internet. It usually requires a lift pump between the fuel tank and the swirl tank but it's a very effective design.

Hope this information helps. It certainly helped me and a friend's team eliminate cavitation/vaporization/boiling issues in several applications.

[modnrod]

Just a couple of things to consider.
As Hotrod said above, excessive pressure will cause cavitation, heat will raise the fuel temp for more pressure, and a too-big pump will raise the fuel temp. I've run 300rwkW (not ACVW, but WOW that would be fun!  ) with only a 200L/hr pump at around 3Bar (45psi, close enough). I did however use an in-tank pump, but still only 8mm lines.
I would maybe reduce the line pressure then retune, and also possibly check your pump isn't too big?

[hotrodsurplus]

Just a couple of things to consider.
As Hotrod said above, excessive pressure will cause cavitation, heat will raise the fuel temp for more pressure, and a too-big pump will raise the fuel temp. I've run 300rwkW (not ACVW, but WOW that would be fun!  ) with only a 200L/hr pump at around 3Bar (45psi, close enough). I did however use an in-tank pump, but still only 8mm lines.
I would maybe reduce the line pressure then retune, and also possibly check your pump isn't too big?

Actually I have to clarify a few things. I never mentioned problems associated with excessive pressure--if I accidentally did then I didn't intend to. The killer is negative pressure.

True, pressure and heat are functionally interchangeable; however, the engine and injectors' needs dictate the operating pressure and the pressure a fuel system usually encounters usually isn't sufficient to cause problems. And an excessively large pump won't cause problems problem so long as the rest of the system can meet its flow capacity. An excessively large pump will generally alter the operating pressure at the injectors more than anything else. So if you can't reduce the operating pressure at the injectors to a given value then it's time to consider increasing the line capacity or decreasing the pump capacity. 

The pumps usually used for fuel injection applications can generally operate fine even when the high-pressure line is shut off. If the pump can't displace any air/fluid then it cannot ingest any more to pump. Closing off the pressure side will make them run hot but that's not a very common circumstance.

[skywalker]

Thanks hotrodsurplus, that is helpful!

I know what cavitation is, it controls my life at the moment.
Here is a great video explaining cavitation causes and effects.

You confirmed my thoughts I pointed out in my first post at option 1).
My conclusion: the stock filter damaged my first three pumps. I removed it after fuel pump 3 was damaged.
The dirt in the last fuel pump killed that one, I cleaned my tank after that. (After I removed the stock filter from the tank I checked if fuel was coming out clear, and that seems to be OK at that moment.)



To make my fuel system better I ordered the -8AN CB fuel tank adapter tot get rid of my 8mm fitting as in above picture. It is the smallest part in my supply line and as I mentioned before I never was thrilled about it.

In my split bus the return line is very short, nevertheless I will replace it with a lager hose, It's a small and easy thing to do.


 
I am going to measure the fuel temperature at specific interfalls, just to be sure heating up my fuel is no problem.

I understand your point 7 completely and will give it a thought.

I would maybe reduce the line pressure then retune, and also possibly check your pump isn't too big?

Thanks for your thought modnrod. Reducing the pressure will not reduce the volume of returning (or circulating) fuel. The fuel pump will pump 130 litre/hour any way and the engine will consume the same amount of fuel. This fuel pump is the smallest I can use.The Kolbenschmidt Pierburg fuel pump did 160 litre/hour and the Bosch fuel pump even more.


For now I only have one question left. What is best for entering the fuel back into the gas tank: leave it as it is or should I lower it to the bottom of my gas tank?

[dragvw2180]

 I have not done it myself yet but will be trying to install fuel cell foam into a stock gas tank . If I do and it works I suspect that the foam will act as an air separator and anti slosh baffle for the fuel and returning fuel in my tank. Mike McCarthy

[Fiatdude]

For now I only have one question left. What is best for entering the fuel back into the gas tank: leave it as it is or should I lower it to the bottom of my gas tank?

Return input is not a real issue -- BUT the lower it is in your tank the less foaming of your fuel occurs -- I always weld a bung to the tank at the best/easiest place to put the return and then weld a shot section of pipe to the fitting that screws into it so that the end of the pipe is near the bottom of the tank -- -- One point to remember is that IF you have to break open the fuel lines and both return and feed are below the fuel level, fuel will coming out very fast from both sides of the connection

[hotrodsurplus]

Thanks hotrodsurplus, that is helpful!

Hey, you're welcome. Actually that video explains cavitation quite elegantly. It gets really good from about 11:00 onward. Water does indeed have intense damage potential. Water under extreme pressure explains how water-jet cutting works and it's very effective on many metals.

To make my fuel system better I ordered the -8AN CB fuel tank adapter tot get rid of my 8mm fitting as in above picture. It is the smallest part in my supply line and as I mentioned before I never was thrilled about it.

Okay, eliminating the stock fuel pickup had to help but I have my doubts as to whether that CB fitting will solve the issue. Remember, the chain is only as strong as its weakest link and the bung in the bottom of the tank could have too small of an ID to work properly. After all, JIC/AN fittings are sized in /16 increments so a -8 would be 8/16 or 1/2". That's quite large.

Now what size (diameter) is the high-pressure line from the pump to the regulator? If it's gigantic like -8 then that might further aggravate the problem. If left to its own devices the pump might have the capacity to outflow its own inlet diameter. If that happened that would cause cavitation regardless of inlet-hose size. I distinctly remember one manufacturer's engineer telling me that the high-pressure line didn't have to be very large and for some reason I think he told me that it's a good idea that the high-pressure line be considerably smaller than the inlet line. There has to be some sort of critical balance at some point--sort of like the critical balance between intake and exhaust ports. Disrupting that balance will cause the engine to run improperly. But this is just speculation.

In my split bus the return line is very short, nevertheless I will replace it with a lager hose, It's a small and easy thing to do.

I don't think length is as critical as diameter. And at that i don't think diameter is all that important so long as it's larger than the high-pressure line from the pump to the regulator.

Return input is not a real issue -- BUT the lower it is in your tank the less foaming of your fuel occurs -- I always weld a bung to the tank at the best/easiest place to put the return and then weld a shot section of pipe to the fitting that screws into it so that the end of the pipe is near the bottom of the tank

I think that's pretty sage advice. Even though a light froth from splashy fuel probably doesn't cause cavitation it might push things over the edge if cavitation is near. At the very least it's a good idea to minimize the splash.

From the examples I've seen, the OEMs minimize splash by running the return line to within about a half an inch of the tank bottom. That way the fuel flows like water from a garden hose when it's about an inch away from the sidewalk. It won't break up and splash that way. In fact some of the in-tank pump mounts suspend the pump from the side of the return line.

One point to remember is that IF you have to break open the fuel lines and both return and feed are below the fuel level, fuel will coming out very fast from both sides of the connection

So true. And that's reason enough to consider using an in-tank pump-mounting kit like the ones Tanks Inc and others offer. You can make your own, by the way. It's really quite easy.

Standard-issue:
http://www.tanksinc.com/index.cfm/page/ptype=product/product_id=253/prd253.htm

http://www.tanksinc.com/index.cfm/page/ptype=product/product_id=227/category_id=61/home_id=-1/mode=prod/prd227.htm

And one with its own fuel trough:
http://www.tanksinc.com/index.cfm/page/ptype=product/product_id=84

Oh yeah, see how those kits hang the pump from the return line? And see how the return line points to within a fraction of an inch of the tank bottom? That's what I was talking about.

[hotrodsurplus]

I just saw the full-size photo of that return line. I think that's perfectly adequate--it's quite large diameter in fact. I don't think the length is an issue either.

In my split bus the return line is very short, nevertheless I will replace it with a lager hose, It's a small and easy thing to do.

[skywalker]

Now what size (diameter) is the high-pressure line from the pump to the regulator? If it's gigantic like -8 then that might further aggravate the problem.

Al fuel lines after the pump are 8mm and all connections are -6 (AN6).

I just saw the full-size photo of that return line. I think that's perfectly adequate--it's quite large diameter in fact. I don't think the length is an issue either.

Return line is the same as the pressure line, 8mm and connections are -6.

[skywalker]

Here's a picture showing the fuel lines and fuel rails quite well:

[hotrodsurplus]

Al fuel lines after the pump are 8mm and all connections are -6 (AN6).

I just saw the full-size photo of that return line. I think Return line is the same as the pressure line, 8mm and connections are -6.

Based on what I know about existing fuel systems the 8mm line is sufficient to flow well more for the power your engine makes AND is sufficiently small to maintain balance with the -8 line. So I would say that you're safe there.

This is just ammunition for the conversation you should have with a fuel-system specialist but I bet that 8mm is adequate for the return line too. Remember, with length comes resistance and that's a very short line. If this was a conventional system where the tank is far from the regulator then the 8mm might not be adequate.

Again, this is just my opinion. An engineer might see it totally differently.  

[skywalker]

...but I bet that 8mm is adequate for the return line too. Remember, with length comes resistance and that's a very short line. If this was a conventional system where the tank is far from the regulator then the 8mm might not be adequate.

I agree, but it is very easy for me to replace it by a larger line. The lower the resistance the less gas due to lower speed. All bits will help.

[hotrodsurplus]

I agree, but it is very easy for me to replace it by a larger line. The lower the resistance the less gas due to lower speed. All bits will help.

True. Very true. Might as well eliminate all of the variables so long as it's cost effective. A return line design inside the tank that mimics the OEM fuel line would likely help too if only by reducing or eliminating the splash. 

I suspect that the flat tank bottom and lack of baffles lets the fuel slosh which starves the pickup.

[Fiatdude]

Just keep the tank full-ish -- can't slosh if there is no where for it to slosh to

Or go to one of those tall small tanks --

a short story from a long time ago -- someone who will remain nameless -- half way across the tank drilled and welded in 3/4" tubing -- a lot of them -- about 1/4" apart -- connected them all and ran R12 thru them to chill the gas (back when that was the big rage) and created the world greatest baffle lol

[skywalker]

Time flies when having fun! 

To end this story, I think it is all fine now. I drove 6.400 km this year with the same full pump. Longest trip was from the Nurburgring to home in one run (370 km / 3h45).

What I did last winter:
- I replaced the 8mm tank outlet for a AN8 tank adapter as mentioned by dragvw2180
- I extended the return in the tank to the bottom of the thank.

The 8mm outlet seems to be the problem I had causing the pumps to break down, as being to narrow.

[BeetleBug]

Time flies when having fun! 

To end this story, I think it is all fine now. I drove 6.400 km this year with the same full pump. Longest trip was from the Nurburgring to home in one run (370 km / 3h45).

What I did last winter:
- I replaced the 8mm tank outlet for a AN8 tank adapter as mentioned by dragvw2180
- I extended the return in the tank to the bottom of the thank.

The 8mm outlet seems to be the problem I had causing the pumps to break down, as being to narrow.

Thanks for sharing. Very nice bus!