Article by Kxxx x. xxxxxxxxx

Flow testing was done on a 1972 style (-72) Sportster gear type oil pump from a KHK.
Flow rate was the target of the experiment.
And there were some concerns about oil scavenging ability and the introduction of froth in the oil.
An oil bypass was installed in the pump to see if possible frothing could be separated out (keeping suspected thick froth out of the engine on cold startup).

All testing was performed at real time to simulate actual oil at operating temperature with a blend of WD-40 and 5W-20 motor.
More specifically, a blend of Fleet Farm 5W-20 semi-synthetic oil containing ~ 30 wt%.
WD-40 was prepared to provide oil having a real time viscosity of ~ 20 cP, to simulate 50 weight oil at 100C.
The final run conducted was with straight Mystik JT8 15W-50 at room temperature.

(data courtesy of Kurt Melancon))

The pump output collection consisted of 2 streams - the bypass circuit and the engine feed.
The feed circuit had a modded restriction terminal “cow udder” as shown below having 2 discharge orifices of 0.040” ID and 1 discharge orifice of 0.060” ID.
These restrictions represented a 0.060“ ID pill installed in the end of the pinon shaft and 2 x 0.040” ID factory piston squirters on KHK cylinders.

Directly adjacent to the pump body was a pressure gauge, followed by an adjustable valve and finally the udder.
The adjustable valve allowed the udder to be throttled back so output from only the bypass could be observed, if desired.
A spring-loaded pop-off valve (steel ball on seat) controlled the output ratio between the bypass and engine oil.
The pop-off valve spring pressure was adjusted by turning a nut on the rear of the pump as shown below.
The pump was driven by a lathe and was conservatively operated at only 500 rpm for the majority of the testing.
The final day of testing finished off with several runs at 1250 rpm.

“Cow Udder” manifold for feed output
^{1) 2)}

It is well known that gear pump output, under controlled conditions, correlates linearly with pump speed.
So we can extrapolate pump output to higher operation speeds with good confidence.

The primary take-away from this final run was that the heavy yellow foam collected was down-right scary when one thinks of a cold engine full of this stuff.
Note: the volume fraction of bypass oil increased dramatically with the heavy oil, which is exactly why the bypass is present.
Up to this point we had tested exclusively with 20 cP WD-40/oil blend.
This made nice little jets of fluid in the collection vessels and carried essentially no air whatsoever. But the 15W-50 frothed into a scary mess.

This final experiment also verified that at a fixed pump speed, the same volume of oil was moved regardless of whether it was the 20 cP WD-40 blend or the heavy motor oil.
But the partition between feed and bypass changed dramatically.

The data table and plot below show the results from the final screening testing performed.
Perhaps the single “take home” from testing is, that at 3000 rpm, the pump passed more than 1.5 qt/min.
So there is certainly a generous amount of oil going through the engine.
In spite of the generous supply of oil passing through the KHK engine, under real world operation, very little engine heat is bled off through the oil, as evidenced by the fact that the oil tank temperature remains very low. In stock flatheads, there is just no way to get a continuous oil flow on the piston crown, where all the heat is (KHK piston squirter provides a 0.040” stream directly across the base of the cylinder, with nothing directed upward). Simply stated, the heat is up top and the oil is down low. On the KHK the oil tank temperature remains at almost exactly the temperature of the heavy aluminum primary cover. If I start the bike at 70F, and ride it for 5 mi the primary cover and oil tank are each barely getting luke warm. After 30 mi they are both about 135F, and nothing but dramatically higher air temp pushes the temp of either component much higher. In all the KHK driving I’ve done, I’ve never observed an oil temperature over 165° F, and that was on a 95F day. To limit the amount of bypass oil, I have tightened the bypass slug spring to be almost solid, so the large majority of oil pumped should be going through the engine, not bypassing.

I took the liberty below of converting your gal/min table from Sportsterpedia to qt/min for comparison to my results. You can see why I like the 1977-1985 gerotor pump since the 3000 rpm output is about identical to the -72 gear pump, according to my interpretation.