cavitation- a deep dive.

[Geoff Weeks]

Not to muck up Larrys 300 build thread any more I am going to start an new one.

1st I would like to say that I wasn't trying to be disrespectful to anyone (esp Paul) but did want to understand the thinking.

I did a deep read on the subject, and found we are both wrong. Liner cavitation has nothing to do with the water pump at all. Not design, not placement on the block, not how well it bleeds itself of air, nor air itself.

We were both partially correct and partially wrong.  Paul is correct that vaporized coolant, not air, is the cause. It isn't due to flow or pressure however, according to what I have read on the subject.

 Liner cavitation is caused by high frequency sound wave from the liner itself, cause by the combustion and movement inside the liner. When these high frequency sounds get transferred from the liner into the coolant, they make high and low pressure points in a wave, as the liner resonates. When at the low pressure stage of the wave the coolant can vaporize and when at the high they "burst" back to liquid.  This, over time can etch the liner in a tale tail pin hole shape. 

 For this reason, cavitation damage is limited to one plane of the liner, and 90 deg around the liner there will be none, it is all in how the liner resonates.  This also explains why it occurs more or less evenly across all liners in the engine without regard to how close or far they are from the water pump. If it were caused by vapor pockets circulating in the coolant, we would expect the damage to be random around the outside of the liner. It is also why some manuals state "with minor cavitation damage, the liner can be reused if turned 90 deg in the block".

Centrifugal pumps can not pump air efficiently and even less so with a liquid "head" above the outlet. As stated it is why they are not "self priming".  For this reason, there is a way provided to expel air when the outlet is below the top of the impeller housing. Eliminating any trapped air is important, but has no effect on liner cavitation. It can effect the pump itself and other parts of the engine. It is partially why all trucks for over 50 years use a "bottom fill" system (shown in the 300 thread as a Marmon attachment) so the coolant rises from the lowest point in the engine and pushes out the air as level rises.

Paul stated "I must have an open mind to learn" and I do, which is why I ask anyone to back up their statements, I can't learn if I don't know the "why". An open mind lead me to a few hours of reading on the subject. I have already apologized to Paul in the 300 thread if I seamed a bit too "confrontational" but I need to understand the "why" of something to be able to retain it. His 

If you want more info or possibly a better explanation then I can give, do a web search for "causes of liner cavitation in heavy diesel engines" and select the ones Not from some YouTTube, but from engine mfg.  

To quote Paul "

 

So cavitation occours when the liquid, in this case water, separates into gasses

Oxygen and hydrogen in the case of water

This happens when the delivery side of the pump isn't restricted enough or supply side is to restricted "

It isn't that the coolant separates into Oxygen and Hydrogen or even that the mixture of antifreeze separates at all, but rather that it changes from liquid form to gas form while still being coolant, this happens at the minute scale where the wave caused by the high frequency sound from the liner meets the coolant on the outside of the liner. Much like an ultrasonic cleaner does.  It does this while at pressure inside the block. Cummins low flow has a block coolant pressure of ~45psi @2100 rpm! That engine still has liner cavitation issues. It isn't that the flow is restricted in any way, either on the low pressure or high pressure side. 

Until I read up on it, I never really gave a thought to why it is on one side of the liner and 90 deg it doesn't happen, I knew that to be true, both from reading and pulling liners out of engines, but now I know the "why" and it help me to better understand the problem.

Likely more than most want to read or know, but I did put in the time to understand it.

The most important point I learned is the "vapor bubble" is both created and destroyed at the liner itself and not the result of being moved about by the circulation of the coolant. 

Edited 7 hours ago by Geoff Weeks

[other dog]

That's what I thought, I just didn't say anything...

 

 

 

🤣 

All I know is that I went to the dentist one time and he said "you have a cavitation here, you need to cut back on the M&Ms" 😀 

[67RModel]

So based on this analysis what is the conclusion as to why some engines suffer more so than others due to cavitation issues? Is it the specific harmonic that a particular engine creates inside of itself as well as the engine's specific liner design / material?

Also does cavitation affect parent bore engines? Or does it but there is so much material between the wall of the water jacket and the wall of the cylinder that it never becomes a problem?

Edited 7 hours ago by 67RModel

[67RModel]

I was considering purchasing a 40-60hp Ford farm tractor of mid 1980s to mid 1990s vintage. They all had the same 3 or 4 cylinder wet sleeved engines. And from my research there was was repeated mention of cavitation and/or "porous blocks" in those series/vintage of Ford tractor engines. I don't know if those two terms are synonymous or different issues that are conflated by people that don't actually know. Anyway I decided I didn't want to take a chance on something that was coming up on 40 years old that had known issues when they were brand new.

[Geoff Weeks]

4 minutes ago, 67RModel said:

So based on this analysis does what is the conclusion as to why some engines suffer more so than others due to cavitation issues? Is it the specific harmonic that a particular engine creates inside of itself as well as the engine's specific liner design / material?

Also does cavitation affect parent bore engines? Or does it but there is so much material between the wall of the water jacket and the wall of the cylinder that it never becomes a problem?

I went looking for the causes, not solutions but you bring up good questions. I would add one more, does a dry liner engine have a similar problem as a parent bore? Do either of them have a problem at all? You don't hear much about dry lined engine leaking coolant between the dry liner and block/crankcase. I know it happens with freeze cracks.

It would follow that the places in the block that hold the liner would effect the resonance of the liner. So a liner pressed in at the very top and bottom might resonate differently than one with the top press fit is lower in the block.

The irony about learning is: the more you know the more questions it brings up!

Mostly I hear about it in wet lined engines, but those are often the most powerful engines as well, so the cylinder pressure and piston speed may be a factor? 

I bet there could be a collage level course on the causes and solutions alone!

[mowerman]

lol I have no argument,,, here… I’ve never even heard of an air pocket in the water pump till now… like most people… assuming after running an engine a few minutes after new water pump ,,, air is all pushed through to the radiator 

[Geoff Weeks]

7 minutes ago, 67RModel said:

I was considering purchasing a 40-60hp Ford farm tractor of mid 1980s to mid 1990s vintage. They all had the same 3 or 4 cylinder wet sleeved engines. And from my research there was was repeated mention of cavitation and/or "porous blocks" in those series/vintage of Ford tractor engines. I don't know if those two terms are synonymous or different issues that are conflated by people that don't actually know. Anyway I decided I didn't want to take a chance on something that was coming up on 40 years old that had known issues when they were brand new.

IDK, I know a neighbor had a problem with a IHC. It was wet lined . Coolant filter and SCA's were the std answer in trucks at that time. He, like many farmers never gave it any thought until there was a problem. Never used SCA's or coolant filters. 

[Geoff Weeks]

2 minutes ago, mowerman said:

lol I have no argument,,, here… I’ve never even heard of an air pocket in the water pump till now… like most people… assuming after running an engine a few minutes after new water pump ,,, air is all pushed through to the radiator 

If the MFG was any good, they likely would have done the "thinking" for you and provided a way to vent most if not all the air. Cat 3406 has the bypass to the 'stat come down vertical, so a small hole at the top of the higher pressure side will vent to the T stat housing. 

 If the mfg did provide a method either internal or external, it is important not to plug what they put there.

[mrsmackpaul]

So there's lots to take away from this 

Water doesn't get trapped in the pump housing causing cavitation 

The gases, as was describes to me by pump experts (that is their lively hood) aren't hydrogen and oxygen as I think ?? I assumed they were

This may only explain cavitation in engines, but irrigation pumps still suffer from cavitation and have motors no water at all goes through

Restricting the delivery side of the pump stops cavitation and increases flow

Maybe it was my assumption from years of experience and lessons learnt that is the pump delivering more than it can draw 

It was certainly explained to me along those lines and proved to be correct over the years 

Poor design in motors causes this cavitation problem and not trapped air 

Also as long as a centrifical pump has the inlet and labyrinth seal on the inlet end covered and no air leaks on the inlet and a supply great enough to supply the pump, the pump housing will expel all air with seconds of running 

 

÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷

So funny line above to slightly change what I'm saying in reply to this thread, to ask questions or maybe some thoughts 

Were on the liner did this 90° issue happen

If it isn't hydrogen and oxygen gas, then what gas is it ?

÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷÷

I feel it maybe still the pump delivering more than it can supply 

Here is my thoughts, based no science at all

The coolant system cant be at 45 psi all over as suggested, need pressure change to get flow

So maybe 45 psi at the thermostat housing or radiator 

Maybe 25 psi at the pump inlet

But for flow the pressure has to be lower in one spot and higer in another for water or coolant to flow, other wise it is just sitting in a puddle or a bath tub or dam or pond

I do feel that there was a time were manufacturers got to smart for their own good and changed the make up of metals in the late 70's and early 80's and every car came with rust built in 

I wonder if engine manufacturers had a similar issue and materials weren't as good 

Porus blocks and transmission housings were pretty common for a while

So with a possible cheaper stinger use of materials and the demand for more power requirements and then more cooling requirements 

Was this the "perfect storm" for  cavitation to become a issue in some manufacturers engines ?

I feel there is probably a lot of contributing factors here and not just one cause 

14 litre Cummins nearly doubled in out put in little over a decade 

IH were under financial strain and trying desperately to keep up and had a bad history of not developing products fully 

Thanks for the info Geoff, all very interesting stuff

 

Paul

Edited 3 hours ago by mrsmackpaul
Correct my fat fingers and poor grammar

[kt_Engineer]

5 hours ago, Geoff Weeks said:

Not to muck up Larrys 300 build thread any more I am going to start an new one.

1st I would like to say that I wasn't trying to be disrespectful to anyone (esp Paul) but did want to understand the thinking.

I did a deep read on the subject, and found we are both wrong. Liner cavitation has nothing to do with the water pump at all. Not design, not placement on the block, not how well it bleeds itself of air, nor air itself.

We were both partially correct and partially wrong.  Paul is correct that vaporized coolant, not air, is the cause. It isn't due to flow or pressure however, according to what I have read on the subject.

 Liner cavitation is caused by high frequency sound wave from the liner itself, cause by the combustion and movement inside the liner. When these high frequency sounds get transferred from the liner into the coolant, they make high and low pressure points in a wave, as the liner resonates. When at the low pressure stage of the wave the coolant can vaporize and when at the high they "burst" back to liquid.  This, over time can etch the liner in a tale tail pin hole shape. 

 For this reason, cavitation damage is limited to one plane of the liner, and 90 deg around the liner there will be none, it is all in how the liner resonates.  This also explains why it occurs more or less evenly across all liners in the engine without regard to how close or far they are from the water pump. If it were caused by vapor pockets circulating in the coolant, we would expect the damage to be random around the outside of the liner. It is also why some manuals state "with minor cavitation damage, the liner can be reused if turned 90 deg in the block".

Centrifugal pumps can not pump air efficiently and even less so with a liquid "head" above the outlet. As stated it is why they are not "self priming".  For this reason, there is a way provided to expel air when the outlet is below the top of the impeller housing. Eliminating any trapped air is important, but has no effect on liner cavitation. It can effect the pump itself and other parts of the engine. It is partially why all trucks for over 50 years use a "bottom fill" system (shown in the 300 thread as a Marmon attachment) so the coolant rises from the lowest point in the engine and pushes out the air as level rises.

Paul stated "I must have an open mind to learn" and I do, which is why I ask anyone to back up their statements, I can't learn if I don't know the "why". An open mind lead me to a few hours of reading on the subject. I have already apologized to Paul in the 300 thread if I seamed a bit too "confrontational" but I need to understand the "why" of something to be able to retain it. His 

If you want more info or possibly a better explanation then I can give, do a web search for "causes of liner cavitation in heavy diesel engines" and select the ones Not from some YouTTube, but from engine mfg.  

To quote Paul "

 

So cavitation occours when the liquid, in this case water, separates into gasses

Oxygen and hydrogen in the case of water

This happens when the delivery side of the pump isn't restricted enough or supply side is to restricted "

It isn't that the coolant separates into Oxygen and Hydrogen or even that the mixture of antifreeze separates at all, but rather that it changes from liquid form to gas form while still being coolant, this happens at the minute scale where the wave caused by the high frequency sound from the liner meets the coolant on the outside of the liner. Much like an ultrasonic cleaner does.  It does this while at pressure inside the block. Cummins low flow has a block coolant pressure of ~45psi @2100 rpm! That engine still has liner cavitation issues. It isn't that the flow is restricted in any way, either on the low pressure or high pressure side. 

Until I read up on it, I never really gave a thought to why it is on one side of the liner and 90 deg it doesn't happen, I knew that to be true, both from reading and pulling liners out of engines, but now I know the "why" and it help me to better understand the problem.

Likely more than most want to read or know, but I did put in the time to understand it.

The most important point I learned is the "vapor bubble" is both created and destroyed at the liner itself and not the result of being moved about by the circulation of the coolant. 

I did my entire Master's research thesis on trying to simulate cavitation on wet line Diesel Engine cylinder liners at ~1500 Hz in lab conditions. As you mentioned the phenomenon occurs due to cylinder vibration (due to combustion) causing the liner to move away from the coolant for milli-seconds creating a vacuum like condition. The static pressure dropping below the vapor pressure of the fluid causes the coolant to form micro-bubbles (cavitate) which implode when the cylinder liner moves back to its original position. The micro-jets from the implosion travel at high-velocity impinging on the cylinder wall damaging the liner. The additives in coolant thinning out starts cylinder line corrosion eventually. The dual action of corrosion and cavitating bubble collapse accelerates the damage. 

The biggest Engine OEM in US (who funded the research) created the cavitation in their lab using ultrasonic transducer vibration at 20,000 Hz on test specimen. They tested out different additives to see the improvement of wear on test specimen. Since actual Engine vibration occurs at 1500 Hz the fluids (with additives) developed in lab were not showing identical life improvement (on liner) as was seen in the lab (on specimen). I was tasked with creating cavitation at 1500 Hz and testing out different coolant additives to see the improvement. This was something I worked on over 25 years back.

[Geoff Weeks]

The ~45 psi is straight from the Cummins Big Cam manual. it is measured at the lower water rail (where heater lines hook up) there is a "compucheck" fitting there to measure it. 

Yes, we both agree that there has to be a difference  in pressure for there to be flow. I always knew block pressure was higher but until I got deep into Big Cams, I didn't know it was that high. Pressure in the upper water rail is considerably less, as there are restrictor washers between the head and upper water rail (on Big Cam 4's). It was Glenn Akers who clued me on the higher pressure. I wasn't saying it is that high all over, just in the block. The Big Cam 4's and NT88's use waterpumps that differ from the earlier Big Cams, The 4 water pump looks similar to the -3 and down but has an extra bleed hole and is turned at a fast speed (smaller pulley) then the -3's and down. Many a person has made the mistake of installing the wrong pump and having problems because of it.

The upper water rail, bypass, return from the radiator as well as the "bottom fill" line from the top tank are kept at radiator cap pressure, most often 8-12 PSI. The Big Cam 4 and NT88's are a whole 'nother topic as they have essentially 2 coolant loops that are in motion at all times, and come together in the water pump.

Antifreeze mixture when it changes from liquid to gas doesn't change the chemical make up. Just like R 134a is the same chemical when it is in a gas or liquid form. 

While writing this, KT replied, and is better suited to explaining, I would love to hear more from him.

Here is the liner re-used and spec for block pressure (max pressure 50 PSI). The cavitation happens 90 degs from the crankshaft centerline, so to re-use you rotate the liner so the damage is inline with the crankshaft.

Sorry about one right side up and the other upside down in the attachment, I can't seam to flip just one.

Cumminsspec.pdf

Edited 1 hour ago by Geoff Weeks

[other dog]

4 hours ago, mowerman said:

lol I have no argument,,, here… I’ve never even heard of an air pocket in the water pump till now… like most people… assuming after running an engine a few minutes after new water pump ,,, air is all pushed through to the radiator 

That's me Bob. I've never rebuilt an engine, but I've changed many water pumps, radiators, hoses, etc. and all I ever did was fill it back up with water and antifreeze, run it a while, then when it cooled off again just top the radiator off and it was done. 

This has become so technical I'm gonna step back and not even make any more jokes...

[Geoff Weeks]

20 minutes ago, kt_Engineer said:

I did my entire Master's research thesis on trying to simulate cavitation on wet line Diesel Engine cylinder liners at ~1500 Hz in lab conditions. As you mentioned the phenomenon occurs due to cylinder vibration (due to combustion) causing the liner to move away from the coolant for milli-seconds creating a vacuum like condition. The static pressure dropping below the vapor pressure of the fluid causes the coolant to form micro-bubbles (cavitate) which implode when the cylinder liner moves back to its original position. The micro-jets from the implosion travel at high-velocity impinging on the cylinder wall damaging the liner. The additives in coolant thinning out starts cylinder line corrosion eventually. The dual action of corrosion and cavitating bubble collapse accelerates the damage. 

The biggest Engine OEM in US (who funded the research) created the cavitation in their lab using ultrasonic transducer vibration at 20,000 Hz on test specimen. They tested out different additives to see the improvement of wear on test specimen. Since actual Engine vibration occurs at 1500 Hz the fluids (with additives) developed in lab were not showing identical life improvement (on liner) as was seen in the lab (on specimen). I was tasked with creating cavitation at 1500 Hz and testing out different coolant additives to see the improvement. This was something I worked on over 25 years back.

Who knew we had an expert lurking among us! Please correct me if I have gone a stray in my understanding!

[other dog]

4 minutes ago, Geoff Weeks said:

Who knew we had an expert lurking among us! Please correct me if I have gone a stray in my understanding!

OK, I - oh wait, you're probably talking to kt_engineer. Never mind.

[other dog]

That was the last one, promise. I couldn't resist 🤣