Engine/Transmission (1998.5 - 2002) Cavitation is NOT your VP-44's friend!

[Bob V]

Another good video Gary! This has been a good discussion and it has led me to do a little testing of my own, however... I haven't mastered the technology required to get a video out of the shop and onto the internet! That's going to be my next project!!



I totally agree with CIWYM, something is happening at 15psi. That's the bottom line!



What I tried to do is separate the relief valve from the cavitation issue. My thinking is that if I could take take the relief valve out of the equation, I could more accurately evaluate the cavitation.



Here is what I did:



My set up today did not include gages. Tomorrow's set up will include a pressure gage on the output and a vacuum gage on the input (very important, as I will explain later).



1. Duplicated Gary's set-up w/o gages. Fuel on the floor below the pump, cranked down on the output and sure enough get the sudden drop in output and pump speeds up. Cloudy fuel in the line.



2. Removed the input fitting and inserted the clear tubing into the pump so that it sealed inside the pump, past the bleed port for the relief valve. Repeated the test, cranked down on the output and sure enough got the sudden pressure drop. Only this time fuel is dumping out of the input, indicating that the pressure drop was due to the relief valve opening.



This seems to indicate that the sudden pressure drop is caused by the relief valve opening and the cloudy fuel is from the aggitation in the pump.



3. Now to take the relief valve out of the equation, I cut a short (~3/8" long) piece of clear tubing and wrapped it with electrical tape. My goal was to press this into the input and effectively seal up the bleed port for the relief valve. If it didn't seal, at least restrict the flow enough to reduce the effect of the relief valve. Put the short piece of tubing in place and put the fitting back into the inlet. Fired up the pump, cranked down on the output, no sudden pressure drop, cranked some more, pump working harder and harder. Finally stopped the flow all together with the pump still working hard. Never did experience the sudden pressure drop.



My guess is that tomorrow when I put the gage on the output, I will see that with the relief valve sealed off, I will get higher than 15 psi without the drop in pressure.



This test leads me to believe that what Gary demonstrated in his first video is really the relief valve opening at 15 psi.



So Greg... how do you explain the second video? INQUIRING MINDS WANT TO KNOW!



Here is my theory... it all has to do with pressure drop across the pump. This is why my test tomorrow will include a pressure/vaccum gage on the inlet. In Gary's first video he was sucking fuel off the floor up to the pump and then generating 15 psi outlet pressure. That means that the top of the ball in the relief valve in the inlet of the pump is actually experiencing vacuum equal to the amount required to lift the fuel from the floor to the pump. When you combine that with the 15 psi outlet pressure pushing up on the ball from below, you have a situation where the relief valve will open easier.



So what about the second video... what was different? Placement of the fuel tank. What did this do?



Now the fuel is actually a little above the level of the pump so it is flowing into the pump with a little pressure behind it due to gravity. Now, think about how this affects the ball in the relief valve in the inlet of the pump. Instead of having a little vacuum pulling up on it, there is a little pressure pushing down on it. That means it is going to be a little harder to push open, which means in this configuration, you will need more than the demonstrated outlet pressure of 15 psi to open it, just as Gary demonstrated.



I couldn't figure out for the life of me why moving the pump to the rail would help the situation if the pump always cavitates at 15 psi and drops the pressure. But now after seeing Gary's second video and having to think through how the relief valve reacts to the pressure differential across the pump, it makes perfect sense!



In Gary's first video the pressure differential is greater, therefore the relief valve opens at a lower pressure. In the second video, the pressure differential is lower, therefore the relief valve opens at a higher pressure. Higher pressure than the pump will generate with the truck idling, meaning not more sudden pressure drop due to relief valve opening. This is a good thing!



I would be interested to know what would have happened if Gary had kept cranking on the clamp until near zero flow?



Now I am convinced that moving the pump down closer to, or below the level of the fuel is a good idea... especially if you stay with the OEM fuel pump! I would say by the same logic any pump will do better down there. Hey... isn't that what Gary said?



I do want to throw another little twist into the debate and it came from my own test today. I would submit that what you are seeing in the clear tube when you clamp down on it is the fuel going from laminar flow to turbulent flow. Fluid will flow very smoothly up to a certain velocity, but when it speeds up too much, it breaks up, loses those nice flow lines and becomes turbulent. As the clamp tightens up, it decreases the area that the fluid is flowing through. The only way to flow the same amount of fuel through a smaller opening is for it to go faster. That's what you are seeing in the tube as the fuel speeds up at the pinch point and breaks up into cloudy, turbulent flow. A little ways past the pinch point the fuel slows back down, goes back to the smooth laminar flow, but still carries the residue or cloudiness from breaking up at the high velocity.



Respectfully submitted.



Greg

GOOD POST!!!!!!!

Now you’re getting somewhere. At least you are trying to find out what it is!

 

[Gary - K7GLD]

At this point, based upon Greg's test findings, the basic issue and problem area, seems to be in the LP's internal bypass loop. I'd guess that in normal OEM configuration, once the LP gets into bypass mode at 15 PSI or so, the ball check valve opens as designed, but the high PSI/volume of fuel flow at that point rapidly begins the vaporization/cavitation action as I showed externally by similarly placing a restriction in the line of flow.



Once the vaporization and volume of resulting bubbles begins, it continues to flow and increase in a tight closed loop inside the LP, until the PSI is reduced externally by engine demand, or else the volume of bubbles finally causes pumping action to cease. This can be verified by the fact that once the system PSI is reduced, or the pump is shut down and then restarted, a large volume of captured bubbles immediately is forced from the internals of the pump.



If true, it seems the primary function of the LP bypass/regulator provides the potential for lots more problems than benefit, and some form of EXTERNAL bypass/regulation that returns flow and possible vaporization directly to the fuel tank, instead of back directly into the pumping cycle would be of benefit.



More testing along those lines would be helpful - but of course, many owners obviously and inderstandably prefer to trash the OEM LP, and simply go on to some variation of aftermarket pump...

 

[Fishin Guide]

I agree with Bob. Not cavitation in the "normal" sense. It does look like air seperation due to a low pressure area inside the pump aggravated by vacuum on the inlet side and then the turbulence of the rotor and the lack of proper flow through the bypass area of the pump.



I appreciate Gary taking the time to show the ill effects of different scenarios. At least we can see what is going on, even if I don't know what to call it. The trick now is how to reduce this effect at full outlet pressure while pulling through the same restriction as a stock module.



And it is prudent to check other pumps as well. Most of the aftermarket bypass pumps I have seen look like they dump the flow outside the inlet section, so that may help alone.



Dave

 

[Gary - K7GLD]

I agree with Bob. Not cavitation in the "normal" sense. It does look like air seperation due to a low pressure area inside the pump aggravated by vacuum on the inlet side and then the turbulence of the rotor and the lack of proper flow through the bypass area of the pump.



I appreciate Gary taking the time to show the ill effects of different scenarios. At least we can see what is going on, even if I don't know what to call it. The trick now is how to reduce this effect at full outlet pressure while pulling through the same restriction as a stock module.



And it is prudent to check other pumps as well. Most of the aftermarket bypass pumps I have seen look like they dump the flow outside the inlet section, so that may help alone.



Dave

YUP - that pretty well sums it up - and for most, the fastest and easiest partial "cure" seems to be simply relocating the OEM pump down near the tank - after that, if problems persist, it becomes a judgement call as to how much more effort the OEM pump is worth...

 

[Misenheimer]

YUP - that pretty well sums it up - and for most, the fastest and easiest partial "cure" seems to be simply relocating the OEM pump down near the tank - after that, if problems persist, it becomes a judgement call as to how much more effort the OEM pump is worth...

Thank you Gary

You demonstrated that the pump will run all day long at 15 psi mounted on the rail and that is where both of mine will be soon.

Robert

 

[GHolcomb]

I thought you might be interested in some numbers I came up with in today's round of testing.



Same set up as Gary's. Today I added a pressure gage and a valve to the outlet and a vacuum gage to the inlet. I am pulling the fuel 14 inches up to the pump in the first test. In the second test I raised the fuel up to 1 inch above the pump. These numbers are rough but I think the relative values are interesting. I set the outlet pressure, then read the inlet vacuum, then ran the fuel into an old oil bottle for 10 seconds. Calculated gph from there.



Here are the outlet pressures, inlet vacuums and flow rates.



Lifting fuel 14 inches

1. 8 psi - 6 in-hg - 66 gph (wide open outlet)

2. 12. 5 psi - 3 in-hg - 37 gph

3. 15 psi - 1 in-hg - 17 gph (this was the max pressure I could get from the pump before the relief valve opened)



Fuel 1 inch above pump

1. 9psi - 6. 5 in-hg - 70 gph (wide open outlet)

2. 12. 5 psi - 3 in-hg - 48 gph

3. 15 psi - 1 in-hg - 29 gph

17 psi was the max pressure from the pump and it never did pop the relief valve, the flow just stopped.



Another number of interest is that the highest vacuum I could pull was 14-15 in-hg.



I think one of the most interesting observations is that in the second test at 12. 5 psi the pump is flowing 11 more gph (48 vs 37) and at 15 psi it is flowing 12 more gph (29 vs 17). Furthermore at the 15 psi mode the flow is not too far from double when you compare the first test and second test.



So... you couple the fact that on the rail, you remove the relief valve factor plus you flow more gph at the same pressure, that makes even a stronger case for relocating the OEM lp.



This test helps me understand a little bit more about how the performance of the pump changes when it is moved lower and closer to the tank.



Tomorrow I want to repeat these tests with a little piece of tubing inserted into the inlet of the pump to seal off the bleed port of the relief valve. That will take the relief valve out of the system. I did a quick test this morning with that configuration, right after I added the pressure valve and got 22 psi out of the pump before the flow stopped.



I think it will be an interesting test. I'll post the results if I get the test done.



Greg

 

[Gary - K7GLD]

Looks to me as though sleeving/defeating the internal bypass setup, and rigging an external one to divert excess flow back into the tank has good potential, especially after moving the LP down near the tank. Certainly cheaper and easier than most aftermarket options...



(EDIT)



And Greg, lest I forget, as one who also tends to do lots of experimenting on these trucks, I am VERY aware of the time, effort, and often, expense in providing what you are doing - plus, results from more than one source carry more weight with readers - THANKS!

 

[Fishin Guide]

Looks to me as though sleeving/defeating the internal bypass setup, and rigging an external one to divert excess flow back into the tank has good potential, especially after moving the LP down near the tank. Certainly cheaper and easier than most aftermarket options...





This is one of the reasons I am working on a custom bypass post filter. Stable fuel flow at around 20 psi, lower heat and make life easier on the 12V pump. AS it wn't be smacking the cam in order to maintain contact.



Dave

 

[rweis]

I remember an OLD OLD post that someone drilled the OEM lp and drilled it in such a way as to be able to thread the drill hole and physically lock the carter bypass closed.



They had problems duplicating the drill angles exactely, but had successfully drilled the OEM lp frame a couple fo times and threaded it and hit the bypass valve exactely right and then threaded in a screw making it impossible for the bypass valve to open.



It might have been in the "pumps , lines, and whatnot, I don't remember.



BUT - now when someone comes up with the next latest and greatest lp it can be tested before everyone jumps on it as the "silver bullet"



Bob Weis

 

[Gary - K7GLD]

I remember an OLD OLD post that someone drilled the OEM lp and drilled it in such a way as to be able to thread the drill hole and physically lock the carter bypass closed.



They had problems duplicating the drill angles exactely, but had successfully drilled the OEM lp frame a couple fo times and threaded it and hit the bypass valve exactely right and then threaded in a screw making it impossible for the bypass valve to open.



It might have been in the "pumps , lines, and whatnot, I don't remember.



BUT - now when someone comes up with the next latest and greatest lp it can be tested before everyone jumps on it as the "silver bullet"



Bob Weis

YUP - I remember that one too - but not what the thread name was - actually, inserting a proper sized sleeve into the output side to block flow thru the bypass valve sounds like a good way to go - the only unknown is how that might affect the pump motor, since it's cooling comes from fuel passage thru it - but then, since the only actual flow occurs when the bypass valve is open, probably not much problem...

 

[GHolcomb]

I've just been reading through the "pumps, lines and whatnot" posts. Interesting to see the history of this pump issue.



My thought was to put the sleeve into the input side of the pump to seal off the flow or at least limit the flow through the relief valve in order to diminish the effect of the relief valve opening. I was thinking that if you could eliminate the sudden drop in pressure and flow at 15 psi, that would eliminate one of the pumps problems. I was also thinking that by leaving the output port open, you would still get fresh fuel to the pump motor.



Forgive me for being new to this issue, I've only had my truck a couple of months, but as I understand it the goal is to keep a constant pressure (13. 5 psi?) at the vp44 through out the flow range from idle to wot. Correct?



If correct, why is that the goal? I have read about the cavitation issue with the vp44 so I guess having some pressure at the pump inlet is important, but why 13. 5. I've read that 13. 5 is where they calibrate the pump. If so, I'm guessing it is important to run it on the truck that way, from idle to wot?



If you could measure flow relatively easily, would it be sufficient to know that your lp was keeping up the right flow at say 3 or 4 psi, just enough to prevent cavitation?



In a stock configuration, is 45 gph the max flow for the vp44? And we need a lp that will deliver that 45 gph at 13. 5 psi, ideally?



Just trying to get this nailed down a little more as I do a little more testing. I have a high pressure fuel pump that I want to play with as well to try to simulate what you guys have done. I am thinking that the regulator approach is the best way to go if your goal is constant pressure throughout the entire flow range.



Thanks for your patience!



Greg

 

[Gary - K7GLD]

Greg, opinions vary on what is actually needed in terms of PSI to properly feed a VP-44 - keep in mind that the VP series of pumps are used in a number of vehicles and configurations besides the Cummins - and in some of those, NO LP is used at all - and all fuel movement is accomplished by the VP's own built-in mechanical vane type pump. That same pump is all that quite often has kept various Dodge owners moving LONG after the regular LP has failed.



Thus, my own feeling that if I could maintain a rock steady 5-10 PSI right at the VP-44 inlet, I'd be happy as a clam! As far as the specified - or at least mentioned - 13. 5 calibration PSI used for VP-44 adjustment and tuning, I suspect that has been used purely to match up with what LP the Cummins has been traditionally furnished with - if it had been supplied with a 5 PSI LP, THAT would probably then be the rebuilders parameter.



As it stands on my own truck, I have a Walbro LP installed, and currently am seeing a very stable 18-20 PSI right at the VP-44 inlet - slightly higher than my original goal - but perfectly fine as far as I am concerned...



And NONE of that horrible whatchamaycallit stuff this thread relates to... :-laf

 

[rweis]

Are the mechanical pumps subject to CIWYM? like the RASP? probably, but I do not know.



Bob Weis

 

[Gary - K7GLD]

Are the mechanical pumps subject to CIWYM? like the RASP? probably, but I do not know.



Bob Weis

IF you're referring to the mechanical pump built into the VP-44, probably not - don't they run their excess pressure back to the fuel tank by way of an internal valve? They also use spring loaded vanes, so in some ways should function with the operating characteristics of a positive displacement gear pump...

 

[NoSeeUm]

Vaporization

Air ingestion (Not really cavitation, but has similar symptoms)

This is what I have been trying to say! Some people have such a closed mind they will not listen to ANYONE!!!!!!!

INGESTED from WHERE?



I showed you vaporization inside a *closed* tube - WHERE was the air "ingested " from?

and

What you see at the pinch point of the second video on the suction line is what I would consider cavitation and a pretty darn good example of it. What is happening is for the same flow rate above and below your pinch point, the velocity must increase and the pressure must drop in the region of the pinch. The drop in pressure allows entrained air to come out of solution. I wonder if you could get the same thing to happen if you pulled a little vacuum on fuel.

I had to re-think this and I believe I was wrong. As some people above have stated, the pinch point in the 2nd video is not really cavitation. Cavitation is really more about the formation and collapse of vapor bubbles. The corresponding shock waves blah blah blah... .



IMO nor is it strictly vaporization. I think the effect could probably be closer to being classifided as de-aeration.



The reduced pressure at the pinch point allows the air entrained in the fuel to form bubbles. Once the bubbles are formed they can not re-collapse immediately. More or less the same effect you get when you pull the cap off of a Pepsi bottle.



Kind of a shame quite a bit of the focus of the video(s) has been on the term "cavitation" instead of the cause and effect.



Thanks again Gary, great work there;

Jim

 

[NoSeeUm]

3. Now to take the relief valve out of the equation, I cut a short (~3/8" long) piece of clear tubing and wrapped it with electrical tape. My goal was to press this into the input and effectively seal up the bleed port for the relief valve. If it didn't seal, at least restrict the flow enough to reduce the effect of the relief valve. Put the short piece of tubing in place and put the fitting back into the inlet. Fired up the pump, cranked down on the output, no sudden pressure drop, cranked some more, pump working harder and harder. Finally stopped the flow all together with the pump still working hard. Never did experience the sudden pressure drop.

My first thought when I saw the first video. That is why I suggested Gary do an amp draw test.

So Greg... how do you explain the second video? INQUIRING MINDS WANT TO KNOW!



Here is my theory... it all has to do with pressure drop across the pump.



That means it is going to be a little harder to push open, which means in this configuration, you will need more than the demonstrated outlet pressure of 15 psi to open it, just as Gary demonstrated.



I couldn't figure out for the life of me why moving the pump to the rail would help the situation if the pump always cavitates at 15 psi and drops the pressure. But now after seeing Gary's second video and having to think through how the relief valve reacts to the pressure differential across the pump, it makes perfect sense!



In Gary's first video the pressure differential is greater, therefore the relief valve opens at a lower pressure. In the second video, the pressure differential is lower, therefore the relief valve opens at a higher pressure. Higher pressure than the pump will generate with the truck idling, meaning not more sudden pressure drop due to relief valve opening. This is a good thing!

Yeah, the internal bypass is effected by the pressure on both sides of it. With the discharge pressure constant, as you pull vacuum on the suction the differential pressure across the pump increases. I have held this theory for some time that the additionally D/P is also what accelerates the wear on the bearings and causes early failure. My only guess is that spring in the internal bypass weakens over time.

I do want to throw another little twist into the debate and it came from my own test today. I would submit that what you are seeing in the clear tube when you clamp down on it is the fuel going from laminar flow to turbulent flow.

I am going to guess the below quote, from my ealier post is closer to what is happening, although the disruption of laminer flow is an issue. One has to explain the formation of air ( or possibly vapor) bubbles in a sealed envoirment.

What is happening is for the same flow rate above and below your pinch point, the velocity must increase and the pressure must drop in the region of the pinch. The drop in pressure allows entrained air to come out of solution. I wonder if you could get the same thing to happen if you pulled a little vacuum on fuel.

The drop in pressure at the pinch point is the fulcrum of the mechanism.



LOL

Maybe Gary could get a hyperdermic needle and sample them thar bubbles to see if they burn or worst case just taste bad.



This is good stuff Greg;

Jim

 

[bigredblur]

I had another thought that may or may not have been addressed already in this thread. The VP and the LP both use the fuel tank as a "cooler". If the tank gets low on fuel, there goes our fuel system cooler. Also, for the LP, if the bypass valve kicks in, then all it's doing is recirculating warm fuel within itself and continuing to warm it more, right? Could this also assist the vaporization process? If so, I'm guessing we would receive enormous benefit from adding that "external to the pump" bypass valve that is set to a psi lower than the LP bypass. At least we would get cooling from the tank in that setup.

Of course, as I always tell my wife, I could be completely wrong.

 

[Gary - K7GLD]

Kind of a shame quite a bit of the focus of the video(s) has been on the term "cavitation" instead of the cause and effect.

ABSOLUTELY!



The disruptive diversion of focusing upon the exact and "proper" technical terminology, rather than the central and obvious importance of what is happening and how it can be corrected tends to get me a bit ...



"Cavitation" seemingly has several close relatives - and our LP's rather obviously are the victims of one of them - thanks to Greg's excellent followup testing - the RIGHT way to contribute to a thread like this - we now have an even clearer picture of what is happening! Oo.



It sure beats a cryptic, one-liner criticism that offers no observable advancement of the central issue...



Simply moving the OEM LP down lower in position and closer to the fuel tank will likely erase the VAST majority of LP issues - after that, it's a judgement call as to whether to apply more effort, or move on to an aftermarket solution...

 

[GHolcomb]

From yesterday's testing:



Same set up as Gary's. I have a pressure gage and a valve on the outlet and a vacuum gage to the inlet. I am pulling the fuel 14 inches up to the pump in the first test. In the second test I raised the fuel up to 1 inch above the pump. These numbers are rough but I think the relative values are interesting. I set the outlet pressure, then read the inlet vacuum, then ran the fuel into an old oil bottle for 10 seconds. Calculated gph from there.



Today I inserted a 9/16" long x 3/8" OD X 1/4" ID section of clear tubing into the inlet port of the pump. I have inserted those results below.



Here are the outlet pressures, inlet vacuums and flow rates.



Lifting fuel 14 inches

1. 8 psi - 6 in-hg - 66 gph (wide open outlet)

8psi - 6 in-hg - 60gph

2. 12. 5 psi - 3 in-hg - 37 gph

12. 5psi - 3 in-hg - 39 gph

3. 15 psi - 1 in-hg - 17 gph (this was the max pressure I could get from the pump before the relief valve opened)

15psi - 1 in-hg - 16. 8 gph (the max pressure I could get before flow stopped was 19 psi and never did see the sudden drop due to relief valve opening)



Fuel 1 inch above pump

1. 9psi - 6. 5 in-hg - 70 gph (wide open outlet)

9psi - 6. 5 in-hg - 66 gph

2. 12. 5 psi - 3 in-hg - 48 gph

12. 5 psi - 3 in-hg - 48 gph

3. 15 psi - 1 in-hg - 29 gph 17 psi was the max pressure from the pump and it never did pop the relief valve, the flow just stopped.

15 psi - 1 in-hg - 31 gph 19 psi was the max today



Just for info I also did a little amp draw comparison, based on some other comments I was reading. My amp meter only reads to tenths of an amp, but you'll see the picture.



Measured with 12 volt input



8 psi - 2. 7 amps

12 psi - 2. 8 amps

15 psi - 2. 9 amps

17 psi - 3. 0 amps (sleeve installed)

19 psi - 3. 2 amps (sleeve installed, flow almost stopped)

3 psi - 1. 8 amps (no sleeve, relief valve popped)



Commentary:



1. Effect of the sleeve - In my test today, I used the 3/8" OD tubing which only has a slip fit into the inlet of the pump. I also shortened it a bit to let it have a little room to move. I think that even with a little looser fit, this is still limiting the amount of flow you get through the relief valve and certainly reducing the sudden pressure and flow drop that we were seeing. I have another idea I want to try... I want to find a ball bearing that is the right size to drop into the bleed port on the inlet side of the pump. Then I want to find a metal sleeve to slide into the inlet (I'm thinking 3/8" OD brake line). This will keep the ball bearing in place and on top of the ball in the relief valve. This idea seems a little more robust in my mind for long term use than the clear tubing!



2. Relocation of lp - I would sure be tempted to say for my application, I will just put the sleeve in my OEM lp and leave it mounted on the engine. However... if you go back and look at what happens to the gph numbers at the same pressure when you get the pump at the same level as the fuel it sure seems like a no brainer to relocate the lp as well!



I would like to move on to evaluating failure modes of the lp that are not related to the issue with the relief valve. My brother has a lp we took off his truck that was pumping 0 at idle and -0 at 2000 rpm. I hope to have that pump in my hands next week. I'd love to hear about any other failure modes you have seen for this pump.



TDR - What a great resource for information!!



Greg

 

[PC12Driver]

Nice work!



**applause**