Positive displacement pumps explained (hopefully).

Joined
Dec 23, 2006
Messages
12,378
Location
Canuck - moved to —> California —> Texas —> ???
Reading some of the claims on this board about positive displacement pumps and how they affect the oil flow and pressure made me do some digging to try and clarify few misconceptions.
I'm not going to go off with a lengthy post as I'm by no means and expert on the subject, but instead people can read the below link and I'm going to quote several points from that source.

Quote:
Flow Rate and Pressure Head

The two types of pumps behave very differently regarding pressure head and flow rate:

The Centrifugal Pump has varying flow depending on the system pressure or head.
The Positive Displacement Pump has more or less a constant flow regardless of the system pressure or head. Positive Displacement pumps generally gives more pressure than Centrifugal Pump's.


Quote:
Capacity and Viscosity

Another major difference between the pump types is the effect of viscosity on the capacity:

In the Centrifugal Pump the flow is reduced when the viscosity is increased
In the Positive Displacement Pump the flow is increased when viscosity is increased
Liquids with high viscosity fills the clearances of a Positive Displacement Pump causing a higher volumetric efficiency and a Positive Displacement Pump is better suited for high viscosity applications. A Centrifugal Pump becomes very inefficient at even modest viscosity.




http://www.engineeringtoolbox.com/classification-pumps-d_55.html


In other words the two pump types serve different design characteristics and needs. In systems where the flow is not critical and can vary greatly or even be stopped, a centrifugal pump is ideal because it will maintain a certain pressure regardless if the fluid in the system is flowing or not, such as in the cooling system or house water pipes. In these pumps the pressure is constant and the flow varies.

A positive displacement pump is used when a certain amount of fluid has to be delivered, or constant amount of flow maintained. In these pumps the flow is constant and the pressure varies. That is why these pumps have pressure relief valves, to protect the system and the pump from over pressurization. Because these pumps will keep on moving the fluid until something gives way.
 
Thanks KrisZ, I suspect that you need to start wearing asbestos underwear, as soon the "feeling" and "intuition" will attempt to over-ride the most simple science.
 
OK, somebody grabbed the bull by the horns and continued the discussion started in the previous thread.

Great opening post
thumbsup2.gif


Another topic that ties into this, since we know how a positive-displacement pump operates, is consumption downstream of the pump. That is, the engine's willingness to ingest the oil being fed to it, or, in the case of bearings, as was mentioned by Shannow, the draw of oil; the "self pumping" nature of bearings as they operate. As speed increases, so of course does oil volume provided by the pump (assuming we are operating under the pressure that the relief opens at) and so does the consumption rate of the bearings, which draw in more oil.

I have heard (though do not have first hand experience with) of an engine's bearings at higher RPM consuming an increasingly greater volume of oil resulting in an observable loss of oil pressure. Example, say you have an engine with a 7,000RPM redline and a pump with a 65lb relief spring. Oil pressure at 4,500RPM is 50psi. Relief pressure is reached at 5,500. But when closing in on 7,000RPM, observed pressure on the gauge reduces from 65psi to 55psi, which is counter-intuitive given that volume moved by the pump increases with engine RPM in a linear fashion. The explanation being that the bearings self-pumping nature, which also increases with RPM, effectively eclipsed the pump/RPM relationship. Any thoughts on this?

My experience with SBF and SBC engines has been that once the oil is hot, the increase in oil pressure is very progressive from idle on, with the relief being engaged somewhere north of 4,500RPM. Which jives nicely with the premise that while volume increases in a linear fashion with RPM, so does the consumption of oil by the engine in the hydrodynamic areas, which is why we don't see oil pressure double when RPM is doubled, IE, if you have 25psi hot at 850, you don't have 50psi at 1900.
 
This is a good read from Kingsbury, one of the leaders in making bearings (to the point that machine designers install/specifiy Kingsbury, rather than design their own).

http://www.kingsbury.com/pdf/catalog-fixed_profile.pdf

P3
Quote:
The inlet pressure to the bearing has little effect on the flow rate and only needs to be 0.25 bar but can be higher. To increase the flow rate and lower bearing temperatures, the bore profile should be modified


Quote:
The clearance between the bearing and the shaft is a critical parameter for successful performance
and expected operation. The clearance must be large enough to allow for thermal expansion and
provide space for the hydrodynamic oil film. In general, tighter clearance promotes stability
characteristics, while looser clearance provides for lower oil film temperatures


Interesting that they also have statements on heat generation within the oil film, but that's a topic for another play-date.
 
Only reason big rig use those rotary type pump is because they re proven!are they better ?rofl highly unlikly.[censored] trucking still mainly use 15w40 and god knows there are way better oil but 15w40 is proven.
 
Positive displacement (PD) oil pumps are the reason the flow restrictiveness of oil filters doesn't matter 99.99% of the time. Because all the volume leaving the PD oil pump will go through the filter and engine. The only time it doesn't is when the PD oil pump goes into pressure relief mode, which usually only happens when the oil is very cold and thick, and the engine RPM is high. With hot oil (200 deg F and above), most engines will not hit the pump's pressure relief point.

Most oil filters are all within a few PSID (delta-P) of each other at any give flow through them. Generally, the engine's oiling circuit is 15 times more flow restrictive than an oil filter, and the oil pressure you see on the dash is 95% due to the restriction of the engine's oiling circuit and the PD pump forcing the oil flow through the engine's tight bearings, etc.

If an oil filter was super restrictive, then it would made the PD oil pump operate at a higher output pressure to force the positive displacement volume. A more restrictive oil filter puts the oil pump a little closer to the pressure relief point, but the filter would have to be super restrictive and have a very high set bypass valve also.
 
Originally Posted By: OVERKILL
OK, somebody grabbed the bull by the horns and continued the discussion started in the previous thread.

Great opening post
thumbsup2.gif


Another topic that ties into this, since we know how a positive-displacement pump operates, is consumption downstream of the pump. That is, the engine's willingness to ingest the oil being fed to it, or, in the case of bearings, as was mentioned by Shannow, the draw of oil; the "self pumping" nature of bearings as they operate. As speed increases, so of course does oil volume provided by the pump (assuming we are operating under the pressure that the relief opens at) and so does the consumption rate of the bearings, which draw in more oil.

I have heard (though do not have first hand experience with) of an engine's bearings at higher RPM consuming an increasingly greater volume of oil resulting in an observable loss of oil pressure. Example, say you have an engine with a 7,000RPM redline and a pump with a 65lb relief spring. Oil pressure at 4,500RPM is 50psi. Relief pressure is reached at 5,500. But when closing in on 7,000RPM, observed pressure on the gauge reduces from 65psi to 55psi, which is counter-intuitive given that volume moved by the pump increases with engine RPM in a linear fashion. The explanation being that the bearings self-pumping nature, which also increases with RPM, effectively eclipsed the pump/RPM relationship. Any thoughts on this?

My experience with SBF and SBC engines has been that once the oil is hot, the increase in oil pressure is very progressive from idle on, with the relief being engaged somewhere north of 4,500RPM. Which jives nicely with the premise that while volume increases in a linear fashion with RPM, so does the consumption of oil by the engine in the hydrodynamic areas, which is why we don't see oil pressure double when RPM is doubled, IE, if you have 25psi hot at 850, you don't have 50psi at 1900.


We left off doing calcs from a link to an article on the Machinery Lubrication website.
We got widely varying calcs based on how the force against the bearing was estimated. I used the max torque. Someone else used the max impulse force.

Shannow stated that torque is not a force. I disagree with that. Torque is a force acting on a lever. Get out the statics books.

I still maintain that although bearing oil feed requirements are linear, the total is normally in drops per minute or small fractions of gallons per minute which is miniscule compared to the output of the oil pump. The bulk of the oil is excess leaking out of the bearing before the hydrodynamic area.

The increased oil requirements you've seen as engines approach 7k rpm is probably cap walk increasing the clearance in the bearing allowing more excess oil leakage.
 
Here is real data from the on-board pressure and oil temperature sensors on my Z06. The oil temperature was a constant 200 deg F, and this is how the oil pressure changed with engine RPM.

Z06PressurevsPRM200DegF.jpg
 
Quote:
In the Centrifugal Pump the flow is reduced when the viscosity is increased
In the Positive Displacement Pump the flow is increased when viscosity is increased


That kind of blows the "thinner gives you better flow at startup" argument out of the water if the engine has a positive displacement oil pump...
 
Only if you misunderstand what they mean by flow. They aren't talking about the oil pump efficiency.
Since almost all automotive engines use a Positive displacement -type oil pump

a 70psi stream of honey isnt going to flow as well as a 65psi stream of water.
(using extreme example)

and to put it crudely.. in same cases pressure is resistance to flow. not volume.

for example you could have a turbo engine pushing 16psi boost.. then you take it apart and port/polish it up.. and its 13psi boost.. is it because its flowing less.. no its because there is less resistance to flow.

Originally Posted By: Merkava_4
Quote:
In the Centrifugal Pump the flow is reduced when the viscosity is increased
In the Positive Displacement Pump the flow is increased when viscosity is increased


That kind of blows the "thinner gives you better flow at startup" argument out of the water if the engine has a positive displacement oil pump...
 
Originally Posted By: turtlevette


We left off doing calcs from a link to an article on the Machinery Lubrication website.
We got widely varying calcs based on how the force against the bearing was estimated. I used the max torque. Someone else used the max impulse force.


Yes, saw that, very stimulating discussion! Was hoping to see some further discussion on that, you think there's perhaps some merit to copying and pasting the relevant sections from that thread to this one?

Quote:
Shannow stated that torque is not a force. I disagree with that. Torque is a force acting on a lever. Get out the statics books.


Wikipedia agrees with you:

http://en.wikipedia.org/wiki/Torque

Originally Posted By: wikipedia
Torque, moment or moment of force (see the terminology below), is the tendency of a force to rotate an object about an axis, fulcrum, or pivot. Just as a force is a push or a pull, a torque can be thought of as a twist to an object. Mathematically, torque is defined as the cross product of the lever-arm distance vector and the force vector, which tends to produce rotation.

Loosely speaking, torque is a measure of the turning force on an object such as a bolt or a flywheel. For example, pushing or pulling the handle of a wrench connected to a nut or bolt produces a torque (turning force) that loosens or tightens the nut or bolt.


I am, as I am sure you are, interested in seeing what Shannow meant in his statement in the other thread about Torque not being a force.

Quote:
I still maintain that although bearing oil feed requirements are linear, the total is normally in drops per minute or small fractions of gallons per minute which is miniscule compared to the output of the oil pump. The bulk of the oil is excess leaking out of the bearing before the hydrodynamic area.


Well, I think we can determine that once we get the equation sorted out, no? Remember of course that there is not only the crankshaft that consumes oil by virtue of the hydrodynamic regime and leakage but also the spray out the sides of the bearings (which I am certain is a significant volume, since I have seen it in a cross-section video and it was quite the thing to behold), the camshaft bearings, which likely mimic Shannow's turbine shaft examples far more than the crankshaft does, and multiple cams are multiple points. Also, in a pushrod hydraulic lifter scenario, each lifter acts as a mini oil pump shooting out the pushrod/rocker assembly and this of course also increases with RPM.

So I think we can all agree that there are multiple leakage/consumption points, a few of them who's consumption most definitely increases with engine RPM like pump volume.

Quote:
The increased oil requirements you've seen as engines approach 7k rpm is probably cap walk increasing the clearance in the bearing allowing more excess oil leakage.


Certainly a solid theory/possibility. If I was observing it on a Modular I wouldn't be as inclined to think that the case due to the huge multi-bolted and pinned caps but on an SBF or SBC that is certainly a possibility. Reminds me of the cap and crank walk BuickGN's little 3.8 was known for. That's one of the reasons he posited that he needed to run 20w-50 in it (anything thinner and the engine ate bearings), because the crank and caps moved around so much it needed the heavier buffer to prevent the journals going through the oil wedge and marking the bearings.
 
Originally Posted By: ZeeOSix
Here is real data from the on-board pressure and oil temperature sensors on my Z06. The oil temperature was a constant 200 deg F, and this is how the oil pressure changed with engine RPM.

Z06PressurevsPRM200DegF.jpg



Thanks for that
thumbsup2.gif
Supports my observations that I posted earlier and in the other thread and my position on the matter as to how often the engine is on the relief.
 
Originally Posted By: turtlevette
Shannow stated that torque is not a force. I disagree with that. Torque is a force acting on a lever. Get out the statics books.


Throw some dimensional analysis into the equations that you were using, and you'll find with your method of equating torque (fxl) with "f", you are out by an "l" component...therefore the formula as you applied it is incorrect...formula is, but can't use any number or colour that you like, or it becomes nonsense.

Originally Posted By: turtlevette
I still maintain that although bearing oil feed requirements are linear, the total is normally in drops per minute or small fractions of gallons per minute which is miniscule compared to the output of the oil pump. The bulk of the oil is excess leaking out of the bearing before the hydrodynamic area.


Go to the simple bearing calculator that I posted, and install the bearing dimension of your choice, and see what the side leakage (aka make-up requirement) actually is.

It can be converted to "drops per minute"...but will be lots and lots of them...

As to your leakage statement...it's how they work...look at my first post in this thread, the bearing manufacturer gives it away when they state that to reduce bearing/oil temperature, you increase the radial clearance (which increases the amount of fresh oil mixed with the working oil, dropping the temperature. I've used as little as 0.002" on a 14" shaft to drop the oil temperature 15F)

Originally Posted By: turtlevette
The increased oil requirements you've seen as engines approach 7k rpm is probably cap walk increasing the clearance in the bearing allowing more excess oil leakage.


That requires an
lol.gif
, or a David Copperfield smiley.
 
Originally Posted By: Shannow


Originally Posted By: turtlevette
The increased oil requirements you've seen as engines approach 7k rpm is probably cap walk increasing the clearance in the bearing allowing more excess oil leakage.


That requires an
lol.gif
, or a David Copperfield smiley.


Some engines actually are known for cap/crank walk (the little DSM engine in the Eclipse is a crank-walk legend!) which would in fact increase bearing clearances. By how much? I'm not sure. I'm inclined to think that based on our other discussions about bearing consumption that it is as I noted in my original post, the consumption of the bearing increase out-pacing the increase in volume of the pump per RPM, but I don't know for sure, which was why I posed it as a question.
 
Fair cop OVERKILL, older style blocks tend to flex, cranks tend to bend/whip as well.

My first thought would be that the work carried out against the oil (internal friction of the lubricant being sheared) increases with revs, meaning that the "working" oil temperature is higher, and thus the bearing working viscosity lower...leading to more leakage/make-up requirement.

Oh, and I forgot to add the link to the calculator

http://www.tribology-abc.com/calculators/c9_3.htm
 
FYI - "Fluid Power Formulas"
http://www.engineershandbook.com/Tables/fluidpowerformulas.htm

Fluid Power in Horsepower - HP
Horsepower = Pressure (PSIG) × Flow (GPM)/ 1714
HP = PQ / 1714

So if an engine is putting out 12 GPM with engine oil pressure of 70 PSI at near red line, then the HP going into pumping this oil volume is:

HP = (70)(12)/1714 = 0.49 HP

If cruising down the road at 2500 RPM with 45 PSI of oil pressure (assume 4 GPM), the HP going into pumping the oil is:

HP = (45)(4)/1714 = 0.11 HP
 
Back
Top