ET: Engine (Twin) Oil pump volume

[greg brillus]

The volume of oil/air returned by the scavenge pump should not be underestimated.......This applies to the stock pump but even more so with a 2 start pump.....If the oil tank does not breathe well enough, then oil can and will spew from the filler cap........the returning oil within the filler neck is too close to the cap.......this volume needs to be vented......I'm sure on the twin racer I ran years ago, I used a small brass fitting taped into a plain filler cap and ran a hose up forward and down, not dissimilar to a fuel tank breather line. I know it is common among some owners on here to blank off the chain oiler T 29 fitting, but to me this is a mistake.......its main purpose is a tank breather and secondly to lube the chain, if one wants to do so........The ET183 seals in the timing cover should be installed and checked one at a time......this way, you can press the cover on with the dowels centralising the cover......be mindful about whether you are using a stock cover gasket, a cometic one or no gasket at all, as this will dictate the installed height of the cover.......You want the cover to show not more than a 1mm gap at the top, adjusting the amount is easy enough by trimming down the height of the Simmons nuts on each spindle, or if the gap is too small, install a 5/16" plain washer under said nut........carry this out on all 4 seals and with the cover screws tightened, the crush on the rubber seals will not be excessive.......All too common to remove a cover and the seals are crushed way too much with the rubber deformed badly.......remember there is near nil oil pressure, so a crush of 20 to 40 thou (0.5 to 1.0 mm) is plenty.

 

[bmetcalf]

Regarding venting the UFM, Greg brought this up a few years ago and it prompted me to see if improvements were available. I was in the habit of tightening the cap as much as I could. When I tried just tightening to the first detent and it allowed some breathing and my motor leaked less.

@davidd has made some diagrams that show how the chain oiler bleed works and I need to find that and decide what to do with the now-plugged port at the back of the UFM.

Edit: https://www.vincentownersclub.co.uk/threads/oil-tank-breather-advice.17137/page-3#post-154442

It looks like the easiest mod is to leave the port plugged and make the cap vent clear or bigger.

 

[vibrac]

The volume of oil/air returned by the scavenge pump should not be underestimated.......This applies to the stock pump but even more so with a 2 start pump.....If the oil tank does not breathe well enough, then oil can and will spew from the filler cap........the returning oil within the filler neck is too close to the cap.......this volume needs to be vented......I'm sure on the twin racer I ran years ago, I used a small brass fitting taped into a plain filler cap and ran a hose up forward and down, not dissimilar to a fuel tank breather line. I know it is common among some owners on here to blank off the chain oiler T 29 fitting, but to me this is a mistake.......its main purpose is a tank breather and secondly to lube the chain, if one wants to do so........The ET183 seals in the timing cover should be installed and checked one at a time......this way, you can press the cover on with the dowels centralising the cover......be mindful about whether you are using a stock cover gasket, a cometic one or no gasket at all, as this will dictate the installed height of the cover.......You want the cover to show not more than a 1mm gap at the top, adjusting the amount is easy enough by trimming down the height of the Simmons nuts on each spindle, or if the gap is too small, install a 5/16" plain washer under said nut........carry this out on all 4 seals and with the cover screws tightened, the crush on the rubber seals will not be excessive.......All too common to remove a cover and the seals are crushed way too much with the rubber deformed badly.......remember there is near nil oil pressure, so a crush of 20 to 40 thou (0.5 to 1.0 mm) is plenty.

Thanks Greg
The JE stem is oil tight!

 

[Cyborg]

the flats on the pups are only acting as timing valves, have no influence on the volume displaced.
feed pump is 0.625" with ¼" stroke, so 12.56cc/revolution. pump runs 1/15 crank speed (single start pump) so 12.56/15 = 0.838cc/crank revolution.
you can convert to whatever unit needed from there.
so at 1000 crank revs/min gives 1000 crank rev/min x 0.838cc/crank rev = 838cc/min = 0,8litres/min.
therefore, for every factor of 1000rpm, the volume will be that factor x 0.838L/min.
i.e. @2500rpm volume will be 2.5 x 0.838 = 2.1L/min.

wasn't all too clever with maths so if its wrong then feel free to correct.

If you are calculating the volume by just using the diameter of the rotor and its stroke and not factoring in the flat at the end of the rotor and the fact that the rotor is also moving back and forth as the ports open and close, then the results will be off.
Have a look at DanQ’s post above and apologies for the run on sentence.

 

[DucATIRadeon]

If you are calculating the volume by just using the diameter of the rotor and its stroke and not factoring in the flat at the end of the rotor and the fact that the rotor is also moving back and forth as the ports open and close, then the results will be off.
Have a look at DanQ’s post above and apologies for the run on sentence.

no offence, but the pump goes up and down as a piston, yes? the flats only act as port timing, and do not affect the volume displaced.
so then the volume is pi/4 x dia^2 x stroke...
the volume of the segment (between flat and circumference) is almost the same as the volume lost by the port timing (literally a few tenths of stroke), they don't affect each other.
so as I see it, the displaced volume if still 0,838L/min @ 1000rpm, but if you insist on the flats and timing, make it 0,75L/min. result is still the same: its not much volume (compared to modern stuff which is like apples and pears).

I'm probably wrong, as I'm a minor in the hall of wisdom here

 

[vibrac]

I venture that if 'modern stuff' had roller big ends they would have a low oil rate as well rollers don't like rolling in a large wave of oil
I guess the universal adoption of relatively cheap shell big ends has fostered the thought that high oil flow is necessary for low stress engines
when you ride an early constant loss oil twenties bike with a hand pump in the tank and you see the drip of oil and the speed of bike and engine you get some idea of how little oil a roller bearing engine needs

 

[Cyborg]

no offence, but the pump goes up and down as a piston, yes? the flats only act as port timing, and do not affect the volume displaced.
so then the volume is pi/4 x dia^2 x stroke...
the volume of the segment (between flat and circumference) is almost the same as the volume lost by the port timing (literally a few tenths of stroke), they don't affect each other.
so as I see it, the displaced volume if still 0,838L/min @ 1000rpm, but if you insist on the flats and timing, make it 0,75L/min. result is still the same: its not much volume (compared to modern stuff which is like apples and pears).

I'm probably wrong, as I'm a minor in the hall of wisdom here

I take no offence…. if Dan’s number of 230ml/min @ 2800 RPM, (which is apparently quite close to what is posted elsewhere) is anywhere near accurate, then that’s a long way from 2500 = 2.1L/min. I may be completely nuts when it comes to my understanding of the pump, but I’ve committed to an actual flow test. I’m not in a mad rush to do it, but I did drag a piece of aluminum stock out from underneath the bench. I’ll make a housing that will hold the pump, accommodated the ports, and that grenade pin like screw that fits into the cam. Just for sport, I’ll wager a cheeseburger & beer that the number will be closer to Dan’s.

 

[DanQ]

I venture that if 'modern stuff' had roller big ends they would have a low oil rate as well rollers don't like rolling in a large wave of oil
I guess the universal adoption of relatively cheap shell big ends has fostered the thought that high oil flow is necessary for low stress engines
when you ride an early constant loss oil twenties bike with a hand pump in the tank and you see the drip of oil and the speed of bike and engine you get some idea of how little oil a roller bearing engine needs

From what i gather, 'modern stuff' with roller big ends, such as Harleys, have a restrictor jet in the end of the crank to stop the big ends from being flooded. The Harley pump we fitted puts out 1400ml@2800 RPM vs 230ml/min of a standard vin pump, plugging the quill with a 0.7mm jet was enough to restrict the oil flow to the big ends on a vin. The main advantage of greater oil flow is that is allows us to direct oil to other places inside the engine, such as the underside of the pistons, at the valve springs and onto the cams.

 

[DucATIRadeon]

The flats on the pumps are only acting as timing valves and have no influence on the volume displaced.
The feed pump is 0.625" with ¼" stroke, so 12.56cc/revolution. The pump runs 1/15 crank speed (single start pump) so 12.56/15 = 0.838cc/crank revolution.
You can convert to whatever unit is needed from there.
So at 1000 crank revs/min gives 1000 crank rev/min x 0.838cc/crank rev = 838cc/min = 0.8litres/min.
Therefore, for every factor of 1000rpm, the volume will be that factor x 0.838L/min.
i.e. @2500rpm volume will be 2.5 x 0.838 = 2.1L/min.

I wasn't all too clever with maths so if it's wrong then feel free to correct it.

made a setup for the oil pump to confirm (or correct) my reasoning.
1 mistake I made: 12.56 should be 1.256cc. (pi/4 x 1.5875cm^2 x 0.635cm = 1.2567cc).

primed the pump, emptied and dried out the beaker, and counted 100 squirts, 3x.
in all 3 times I get 125ml for every 100 squirts = 100 pump revolutions.
1 pump revolution is therefore 1.25ml = 1.25cc.
as the calculation above for ⅝" diameter and ¼" stroke gives 1.256cc displacement, the 1.25cc/revolution is pretty reliable figure if one would consider losses here and there.
this is for a standard oil pump. a 2-start oil pump will give exactly the same results per pump revolution, but as it is a 2-start pump, the volume per crank revolution is 2x that of a single start. feel free to perform your own test, it should be the same.

so: (1.25/15) x 2800rpm = 233cc/min which is very close to what was stated earlier, so if anything this test has confirmed 1) the volume measured and calculated support the previous stated volume/time by (DanQ?), and 2) the reasoning made about the volume of segments is inaccurate/untrue. sorry I don't know how else to put it, its not an attack!!

the setup in the vice. I used betray 6-1 as a fluid and as a nice colour to clearly visualise the level in the beaker. pump used is the old pump, still ok fit only the splines are worn somewhat. won't be using it in the rebuild (already have both single and double start pumps).

some silicone sealant to minimise the air leaks at the suction side

M8 bolt with ¼-26tpi (BSF?) threaded on the end as a means to rotate the pump. the M8 nut acts a locking nut so the bolt doesn't come loose while turning counter clockwise direction, same direction as the pump would rotate by the crank.

 

[Cyborg]

All good… thanks for the info.
I’m 75% of the way to completing a test, so will soldier on. I’m using a new pump, but doubt that would made much difference because there isn’t much if any pressure involved.