I have just completed a close inspection of the MGA 1500 oil pickup and pump,
have also done some measuring, testing and calculating. So here it is:
Mechanically the parts looked good, no surprise there. But I wanted to know
the flow rate of the pump. So I put a 16 ounce capacity tub inside of a 2
gallon pan, filled the little tub up with oil, stuck the pickup end of the
oil pump in the little tub, and proceeded to turn the pump drive at about 2
turns per second with a small wrench. I tilted the pump so the oil comming
out the top would dump outside of the little tub into the bigger pan, and
primed the pump. I then filled the little tub to the brim, cranked the pump
20 turns in about 10 seconds, and refilled the little tub while carefully
measuring the amount of oil displaced. And the answer is 10.5 ounces of oil
for 20 turns of the pump, or 0.525 ounces per turn.
The oil pump is geared to run at camshaft speed. At 7000 rpm engine speed
the pump turns at 3500 rpm displacing (get this) 57.4 quarts per minute!
Trust me, the numbers work. The pump is a positive displacement type, so
the flow is not affected much by pressure changes. Discounting the flow a
bit for internal leakage of the pump would still give about 56 quarts per
minute delivery. At that rate it sucks up the entire contents of the oil pan
in just over 4 seconds. Now the oil system hits relief pressure at about
2500 rpm with hot oil, meaning that 20 quarts per minute is all it takes to
provide maximum oil flow through the engine. Beyond that, any excess flow
dumps over the relief valve directly back into the sump, generating heat in
the oil equal to the power running the pump. Some hydraulics engineer out
there will likely tell us how much power is consumed running the pump and how
much heat is generated in the oil (which in turn may infer why race cars like
oil coolers).
Since maximum oil flow through the upper engine commences at 2500 rpm, if
there was an oil return flow problem it would show up quickly at any speed
above 2500 rpm. And since I do not see oil starvation at highway cruising
speed, I conclude that there is no problem with the oil return flow. That
translates into at least 20 quarts per minute of return flow capacity,
circulating the entire contents of the oil pan every 12 seconds at 2500 rpm.
The 20 quarts per minute of oil flows from the pump across the rear of the
block, then via the external lines through the oil cooler and the filter then
to return full flow back into the engine block. It then flows into the main
bearings where it continues to flow, among other places, through drilled
holes in the crankshaft from the main journals to the rod journals. Not
being satisfied with having checked all the oil passages prior to the last
reassembly, I went on to test the actual flow through the engine to the rod
journals.
The oil pan and pump were already removed, and the connecting rods were
disconnected from the crankshaft. Using an oil gun with a rubber nose piece
from a compression tester, I injected oil (3/4 of a quart per stroke over
several seconds) into the block at the oil pump mount. A generous flow of
oil came out in roughly equal amounts from all four of the crankshaft rod
journals at the same time. It was a messy job, but I am now satisfied beyond
any doubt that there is no obstruction of oil flow to any of the rod
bearings.
The hypothesis at this point is that the pump must be cavitating at high
speed, and that there must be some partial obstruction of the pickup tube.
So on to an even closer inspection of the oil pickup assembly.
My original 1500 style parts have a side entry from the oil pump extention
into the screen assembly. Inside the screen a tube with a 90 degree bend
turns to face nose down at the bottom of the screen assembly. Both this
curved pickup tube and the pump extension have 1/2" bore for the oil to flow
through. The bottom piece of the screen assembly is perforated (as well as
the sides), but has a blank sheet metal area directly under the end of the
pickup tube, with fine mesh screen directly on the plate, and the end of the
tube TOUCHING the screen (there are contact marks on the screen). The end of
the tube has two semi-circular notches at opposing sides, these notches
measuring 0.33" across. The two notches combined equal one round hole 0.33"
in diameter. There is no other flow area between the tube and the bottom
plate. Imagine having to push nearly one quart of oil per second through a
hole that small with just 14.7 psi of atmospheric pressure (vented
crankcase).
OK, now that same hydraulics engineer out there should take a stab at how
much warm 10W40 oil (or worse 20W50) will go through a 0.33" hole with 14.7
psi pressure pushing it. Me thinks it won't make 1 quart per second. I
think the conclusion is natural. Somwhere around 7000 rpm (or maybe a bit
more) the oil pump cavitates, and the moment it does, the oil flow drops
dramatically, and the oil pressure nearly disappears instantly. When the rpm
drops below that magic threshold, the flow returns just as quickly. In the
few seconds between, my rod bearings retire.
The fix (I am sure) is to cut 1/8" off of the bottom end of the pickup tube,
providing an entry area around the end a bit greater than the cross section
of the tube. The good news is that the crankshaft wasn't damaged this time
around, cleaned and polished up nicely, will get a new set of standard size
rod bearings and will be back in service in a few days. Next race date is a
week from Saturday, but you know I'll give it hell before then. TIME WILL
TELL!
Thank you all for the help. Drainback suggestions score low. Pickup
congestion suggestions get the cigar.
Barney Gaylord -- 1958 MGA
|