slate blue

Colin I got some dimensions from the seller, this tank is the one I own, it is from the 2007 Honda and has a intergrated breather tank. I bought this one over the other one as I thought when I mold my tank I can just use that part of the Honda tank as a mold. Save where you can. I hope it is worth the purchase to sell the technology, it may also help with my carbon lay up.

Height 700 mm width 140 mm and depth 170 mm, well I wont have the 700 mm, it shows how important height is when it come to getting the air out. Other than the front guard area I can't think of another area with that much height in a street car. That area is where Ferrari put their tank in the 599, so you will have much more choice in a racer.







Greg

slate blue

Quote:
Originally Posted by ptuomov
Here's a quick measurement of the pump displacement. I did this with a caliper and third-grade geometry in the garage, so there's some margin of error. It shouldn't be too far, though.

dimension, inner gear, outer gear
depth, 2.1, 2.1
width , 0.7, 0.7
inner, 0, 1.1
outer, 1.1, 0.7
displacement cc, 0.8085, 1.323
#teeth, 9, 11

cc per pump revolution
21.83

pump sprocket teeth
35
crank sprocket teeth
24
Drive ratio
0.686

cc per crank revolution
14.97
gallons per 1000 crank revolutions
3.95

.

The option I referred to was to use the early oil pump. That is after everything is set up. At least we have that option, lucky to have it in my book.

We could also work out how much oil these Nascar pumps can supply of just ask the manufacturer? That way we can get a comparison.

Greg

ptuomov

Does the old pump fit into a new block?

slate blue

By Ptuomov

Surely does.

Greg

ptuomov

What's the old pump displacement?

GlenL

How old and how new?

I put an '84 in my '80 and they're the same (or I thought they were the same)...except you've got to drill a hole to allow the shaft lube to drain into the sump.

ptuomov

I want to know if the old pump and new pump have the same displacement per crank revolution. By old I mean 928.107.018.06 from 78-79 and by new I mean 928.107.008.02 from 80-95. And does the old pump fit in the S4 block?

[EDIT: apparently the displacements are listed in the 1980 service tech bulletin, does anyone have that?]

Bill Ball

The "old" (82-84) pump is the same except 21mm thick instead of 23. From 1985:

GlenL

Sooooo...

It's thicker "out" and not into the block and the oil shaft drain hole started in '82. Correct?

slate blue

Glen I believe they just bored the pump out more, thanks for posting that Bill, I couldn't find it anywhere. So roughly 9% less flow.

Also I measured the guard/fender area for the dry sump tank and you can achieve over 600 mm in height. That is a mighty big improvement over the wet sump pan and very close to the Honda F1 tank at 700 mm, so it should supply excellent air free oil.
Greg

Kevin Johnson

Ok, I pulled the pump on my early motor and measured the three trapezoids that combine to form the volume displaced per revolution in the Gerotor.

The edges of the teeth seal in such a manner as to create two trapezoids to add together. Surface area in mm^2 are 51 and 46.2 so 97.2. The third trapezoid intrudes on this area so it must be subtracted so -15.75.

This yields 81.45 mm^2 or .8145 cm^2.

The height of the (early) gears is 2.1 cm

The volume displaced per revolution is 1.71045 cc

Drive ratio .686

So, 1.1733687 cc per rpm or 1173.3682 cc per 1000 rpm.

1 gallon = 3785.4178 cc

So, .30997 gallons per minute per 1000 rpm.

So, 2.1697 gallons per minute at 7000 rpm. Or 8.6792 quarts per minute, approximately.

Edit: I see the problem. There are 10 chamber transfers per revolution of the rotor. So, 21.697 gallons per minute. Thought I was losing my mind there.

slate blue

Anybody got an early oil pump in excellent condition that would like to sell?

Greg

Kevin Johnson

Ok, when I remove the gerotor gears it becomes even more interesting. The area that the oil is displaced into is open and communicates before and after the "teeth" or lobes mesh. It is clear that this is a hybrid type of gerotor that does not truly displace -- out of the pump -- a certain amount of oil per revolution. I think the 21.7 gallons per minute is a theoretical maximum flow at zero pressure. The design is probably why cavitation damage is not commonly seen in this pump.

Kevin Johnson

Ok, just managed to remove the key so I can follow the drilled passage in this "communication" area. Tres cool. It looks as though the pump has a simple deaerator built into it. Displaced air would be less dense and the pressure gradient in the relieved communication area beneath the gears would push the air back and towards the center of rotation -- right where the drilled passage is and an annular groove is cut into the inner gear. After lubing the shaft bearing which acts as a sort of one-way valve as well, the vent passage goes into the block -- don't have time to trace it right now.

Yet another engineering cue that the Porsche engineers understood that the engine would be (was) drawing in highly aerated oil from the sump. They apparently counted on the deaerator to be able to handle the load but it was not enough in the turns at high rpms. The oil circuit in the bedplate clearly assumes neat "idealized" oil.

By increasing the volume of the pump they apparently were trying to shift the equilibrium so that there would be more of the outerphase (oil) available per chamber movement and which could be "condensed" into a sufficiently "neat" flow to protect the bearings. Looks like they almost succeeded in the later engines at high rpms judging by the GTR.

It would be interesting to compare the oil pump castings from early versus late to see if they modified the deaerator circuit.

Edit: In another thread Jim Bailey mentions that some blocks are not machined to allow the oil (and air) to drain from behind the seal and so the pressure will pop the seal out.

In this case the lubrication of the oil pump shaft with aerated oil is acceptable as it receives nowhere near the loading of the crankshaft bearings.

In doing a patent search, it appears this technology was quite well understood at the time of the design of the 928 engine. In particular, Mosbacher in the late 1950s and early 1960s calls out general deaerator designs like this:



United States Patent 3,045,778

Issue Date: July 24, 1962


United States Patent 3,137,234
Issue Date: June 16, 1964



Later patents also use this idea:


United States Patent 4,940,115
Sugden July 10, 1990
________________________________________
Differential with bleed holes for deaeration of oil for lubricating journals
Abstract
A lubrication system (40, 60) for an integrated drive generator in accordance with the present invention includes a source of pressurized oil; a carrier (12), rotatably mounted on a axis of rotation, having an inlet (32) for receiving pressurized oil from the source and an outlet (38) for discharging a mixture of oil and air, and rotatably supporting a plurality of planetary gears (14) as the carrier rotates; a plurality of first ports (26), each port being associated with a different planetary gear for supplying only oil to a journal (31) rotatably supporting the associated planetary gear upon rotation of the carrier when pressurized oil is supplied to the inlet, each first port being disposed at a radial position on the carrier at a point displaced from the axis of rotation; and a deaerator (42), in fluid communication with a mixture of air and oil moving radially outward from the axis of rotation during rotation of the carrier to each of the first ports, for removing air from the oil so that only oil moves radially outward to the first ports and for discharging a mixture of oil and air at the outlet.
________________________________________
Inventors: Sugden; Kenneth B. (Rockford, IL)
Assignee: Sundstrand Corporation (Rockford, IL)
Appl. No.: 07/324,751
Filed: March 17, 1989
________________________________________
Current U.S. Class: 184/6.12 ; 184/6.23; 464/7; 74/467
Current International Class: F01D 25/00 (20060101); F01M 9/10 (20060101); F01M 9/00 (20060101); F01D 25/18 (20060101); F16N 39/00 (20060101); F01M 009/10 ()
Field of Search: 184/6.12,6.23 74/467,468 464/16,7
________________________________________
References Cited [Referenced By]
________________________________________
U.S. Patent Documents

4104933
August 1978 Campbell
4459869
July 1984 Bucksch
4793440
December 1988 Iseman

Foreign Patent Documents

966346 Jul., 1957 DE
5524279 Oct., 1989 JP
8804747 Oct., 1989 WO

Primary Examiner: Lazarus; Ira S.
Assistant Examiner: Cariaso; Alan
Attorney, Agent or Firm: Antonelli, Terry, Stout & Kraus




________________________________________
United States Patent 4,681,189
Krisiloff July 21, 1987
________________________________________
Dry sump lubrication system for an internal combustion engine
Abstract
A dry sump lubrication system for an internal combustion engine which includes a cam housing, a crankcase and an oil sump therein. The lubrication system includes an oil tank, a conduit connecting the oil tank to the engine and an oil pump for pumping oil from the sump. The pump may include a plurality of pumping stages. An air separator is directly connected to the pump to remove air from the oil. Oil is conducted from the air separator to the oil tank. Separated air is conducted from the air separator to a canister which is vented to the atmosphere. Oil particles will be condensed out of the air in the canister and will be conducted back to the engine through a conduit which is connected to the engine housing, which conduit includes a restrictor whereby engine pressure will not be equalized with the atmospheric pressure in the canister. A pressure relief is also provided for the oil tank by means of a one way check valve which connects the oil tank to the canister.
________________________________________
Inventors: Krisiloff; Steven (Indianapolis, IN)
Appl. No.: 06/804,673
Filed: December 4, 1985
________________________________________
Current U.S. Class: 184/6.13 ; 123/196A; 184/6.21; 184/6.23; 55/437; 96/174; 96/215
Current International Class: F01M 11/06 (20060101); F01M 11/00 (20060101); F01M 1/10 (20060101); F01M 1/00 (20060101); F01M 1/12 (20060101); F16N 39/00 (20060101); F01M 011/08 ()
Field of Search: 184/6.13,6.21,6.23 123/196A,196R,90.38,90.33 55/159,182,350,437,189
________________________________________
References Cited [Referenced By]
________________________________________
U.S. Patent Documents

1717096
June 1929 Czarny
2024336
December 1935 Cavanaugh
2268653
January 1942 Flowers
2373360
April 1945 Walsh
2432130
December 1947 Serrell
2443875
June 1948 Spangenberger
2538983
January 1951 Sharples
2575315
November 1951 Edwards
2581886
January 1952 Rockwell
2747514
May 1956 Edwards
2755888
July 1956 Cunningham
2888097
May 1959 Scheffler
3045778
July 1962 Mosbacher

Primary Examiner: Croyle; Carlton R.
Assistant Examiner: Obee; Jane E.
Attorney, Agent or Firm: Jeffers, Hoffman & Niewyk

Kevin Johnson

The following files are snapshot images taken from the video. They show that it is very possible with careful observation to judge both the areas that the violent surging air and oil come from as well as the distinctive foam that is ejected by the cam bearings and flung by the cam lobe. This distinctive fine foam appears after a delay that can vary in length of time. This is typical of aerated oil being ingested by a pump.

http://www.crank-scrapers.com/A-928.jpg

http://www.crank-scrapers.com/B-928.jpg

http://www.crank-scrapers.com/C-928.jpg

We can see that the engineers knew a lot of aerated oil was making its way back to the sump. The screening over the early sumps is an attempt to release trapped air from the oil. The clover-leaf fitting that sits next to the sump floor is an attempt to stop foamed oil from being directly ingested by the pickup for as long as possible. We know that the Porsche engineers would be aware of the quick benefit of adding a perimeter ledge around the sump well. It would trap a substantial portion of the oil that easily spills out during a turn producing only a 30 degree angle of repose.

We now know that the gerotor pump in at least some of the 928 engines was not a positive displacement pump with a certain amount of oil being pushed into the oil galleys or even out the relief valve with every revolution of the pump. In at least some of the pumps a deaerator relief passage was also provided.

What seems most likely is that for levels of lateral acceleration up to perhaps 1.5 Gs that the thrashed oil and air does return to the sump in sufficient quantities to keep the pickup covered. I think somewhere between sustained 1.5 Gs and 2 Gs the pickup does start to become uncovered. At 2 Gs it seems reasonable to assume that the pickup is more uncovered than not.

At some point even the deaerator in the pump (at least some of them) is overwhelmed by the sheer volume of entrained air. At that point the hope must be that the multiple bars of pressure that can be generated by the pump is sufficient to push the air into solution at the rate of 9% per bar.

If that pressure was maintained throughout the circuit at all times the 2/6 bearing could hold out longer but there are a series of cross sectional decreases from the exit at the pump to the various arms of the circuit. In a dynamic hydraulic circuit this leads to pressure drops and that in turn leads to a supersatured solution releasing dissolved air to come into equilibrium with the new pressure.

There is a lot of debate over why the 2/6 circuit in particular gives so much trouble. It could be because it is the first to get a good dose of highly aerated oil. It could be as Mike Simard thinks, that there is some sort of Bernoulli effect happening at the entrance to the circuit. At some point it doesn't matter because fixing it will simply transfer the problem to another branch, ultimately. And there is no guarantee that with the 2/6 bearing protected that another bearing will not fail EVEN MORE QUICKLY than the 2/6 bearing did. I really think this is the sort of demonic situation the Porsche engineers were faced with.

Assuredly this is a highly complex problem. It is very understandable that people want to bypass it by installing a dry sump. Just be careful there too. Some new Vettes pop the engines despite having OEM dry sumps. Proper design is critical.