question re oil flow engine vs tank

[andyt11]

When I do an oil change I empty both the engine case and the tank. When I pull the plug on the engine case, I only get a couple of liters, the rest stays in the tank.

First question, what stops the oil from the tank getting into the case?

Second question, when I refil the tank and start the car, is it fair to say there is no oil in the case?

Third question, when I do a vavle adjust, do I only need to drop the engine case oil, not the tank? (Oil is still rather recent).

Thanks in advance.

[HarryD]

This is a diagram of the oil flow in your 911:



As you can see, the oil, when it leaves the tank, has to go through the pump, through the oil passages and then into the sump. This is why the oil does not drain out of the tank into the sump whenthe car sits, too much resisitance to flow there.

There is oil in the case but it is trapped in the passages and some is in the sump.

In theory, you "should" be able to get by with just draining the case and leave the oil in the tank. I never tried it so I can't say for sure. I have done an adjustment on one side by jacking up just that side and removing covers on that side only. No problem there with all the oil on board.

A better approach would be to wait for the next oil change interval (you do change your oil every 3,000 miles, don't you?) and do it then.

[andyt11]

Thanks for the info, much appreciated.
It gets oil very regular yes, it's just unusual circumstances this time. I may well just give it new oil.

So with regard to start up after an oil change, I guess we are saying that the oil in the tank is very quickly pushed through the pump upon start up and you very soon get back to a situation where you have a couple of liters i the case and about 8/9 liters in the tank.

[Wil Ferch]

I've posted this before....for our cars that have front oil coolers and a total system capacity of 13 qts..... the "change-out" volume is about 10 qts. However.....

1.) If the car sat a long time ( overnite or longer)....about 6-7 qts comes from the engine case...and 3-4 qts from the tank.

2.) If the car was recently run before the oil change....just the opposite.... about 6-7 qts from the tank and 3-4 qts from the engine. This is the most frequently seen scenario... AFAIK.

[Grady Clay]

Harry,

Thanks for the timely diagram.

In ‘theory’ all the oil stays in the tank and the engine sump is ‘dry’.
In practice, there is a lot of oil ‘in process’ in the engine when running.
The least amount is when warm, level and at idle.
When cold, a lot of oil sticks to the internal surfaces of the engine and is ‘slow’ returning to the scavenge pick-up in the bottom of the engine case.
The level in the tank is proportionally lower (usually off the scale).

Imagine your 911 sitting, idling, on a 45º slope to the side.
A LOT of oil will be in the cam-box and oil return tubes (full, in fact) and the tank proportionally lower (usually off the scale).
This same phenomenon occurs when you are in an extended corner at 1G (not unreasonable in a 911).

When you are cruising at top speed (say 7000 rpm), the pressure pump is a ‘positive displacement pump’.
It pumps so much oil every revolution.
That oil must traverse (by feeble gravity) back to the sump.
That has a lot of oil traveling in the engine (compared to idle).
The level in the tank is proportionally lower (usually off the scale).


Now, what keeps the oil in the tank when parked is the clearance ‘resistance’ to the viscous flow past the oil pump to low restriction places like bearings and cam squirters.
With hot oil, there can be a lot of oil flow.
Over an extended time (week+) there can be a lot of flow.

So … if you change your oil soon after parking with the engine at idle, level and hot, you will get most oil out of the tank and little out of the sump.
If you change oil after parking hot and waiting a month, you will get most oil out of the sump and still some out of the tank.


All this brings up some important engineering points that Porsche didn’t realize in the original design and finally implemented with the GT2/GT3 dry-sump oil systems.


First is the distance between the oil tank and the inlet to the pressure pump.
Shorter is better and a ‘tall’ tank provides more ‘head’ to the oil.

Next is the lower effectiveness of the front cooler considering the scavenged oil is about 40% air (the size ratio between the pressure pump and scavenge pump).

Third is the lack of an oil/air centrifugal separator to insure pure (no entrained air) oil to the bearings.

Fourth is our old oil tanks must carry about twice the oil (compared to a GT2/GT3) in order to not cavitate (suck air) under some conditions (the well-intentioned '72 is the worst offender).

It is senseless to have the oil filter in the scavenge circuit and not protecting the engine from oil tank contamination.

There are more.



All this brings me to lobby some of our very talented aftermarket suppliers to engineer an up-dated oil system for our nice old 911s (and 914-6s).
The oil tank should be in place of the engine oil cooler (like GT2/GT3).
The centrifugal separator should be integral with the pressure regulating valve (so it regulates pressure by bypassing the seperated air).
The pressurized and separated oil should go to the front cooler (with additional pressure by-pass for cold start-up).
The oil filter should be just prior to the engine lubrication.
There should be adequate screen filters to protect both pumps from foreign objects (meaning if it passes the screen, it won't damage the pump).
There should be improved (single) thermostat function for faster warm-up.
All this can be done with a new external oil tank and no changes to our engines.
In 20/20 hindsight, this system was within the capabilities of Porsche in the '50s, perhaps pre-war.

2¢

Best,
Grady

[dshepp806]

Good read, Grady. Thanks!

Doyle

[calling911]

I think I agree with every thing you said here Grady.. but it wont just be a tank replacement if I understand it correctly.. you cannot just put a paper element filter in the scavenge line... it has to be a positive flow to the filter and past it of course.. so it will require an electric pump.. which presents a problem.. how would you control how much oil goes back to the engine without flooding it and defeating the purpose of a dry sump?

Its a mess... but as I stated in my thread.. I think the thing to do here is install a screen filter in the scavenge line and run a filter with no bypass and watch to make sure you warm up your car carefully and change the filter often (which should be cheap, since the only filter I know of without a bypass is a fram... and even that Im not sure of yet)...

EDIT: I take it back.. it looks like you wouldnt have a flood problem after looking at the diagram again..but you'd be positive feeding the pressure pump and I dont know if that could cause any issues or not.. Im just an armchair engineer. But these electric pumps are common for transmission and rear end cooling in race cars. One could achieve this by somehow gutting a filter, using it to bypass the original filter location and then putting the pump and a remote filter in the return line to the engine... Actually, you could run two filters with no issue as you are just filling the tank with the factory location filter..

I think Im going to simply go with the screen option...

Originally Posted by Grady Clay

Harry,

Thanks for the timely diagram.

In ‘theory’ all the oil stays in the tank and the engine sump is ‘dry’.
In practice, there is a lot of oil ‘in process’ in the engine when running.
The least amount is when warm, level and at idle.
When cold, a lot of oil sticks to the internal surfaces of the engine and is ‘slow’ returning to the scavenge pick-up in the bottom of the engine case.
The level in the tank is proportionally lower (usually off the scale).

Imagine your 911 sitting, idling, on a 45º slope to the side.
A LOT of oil will be in the cam-box and oil return tubes (full, in fact) and the tank proportionally lower (usually off the scale).
This same phenomenon occurs when you are in an extended corner at 1G (not unreasonable in a 911).

When you are cruising at top speed (say 7000 rpm), the pressure pump is a ‘positive displacement pump’.
It pumps so much oil every revolution.
That oil must traverse (by feeble gravity) back to the sump.
That has a lot of oil traveling in the engine (compared to idle).
The level in the tank is proportionally lower (usually off the scale).


Now, what keeps the oil in the tank when parked is the clearance ‘resistance’ to the viscous flow past the oil pump to low restriction places like bearings and cam squirters.
With hot oil, there can be a lot of oil flow.
Over an extended time (week+) there can be a lot of flow.

So … if you change your oil soon after parking with the engine at idle, level and hot, you will get most oil out of the tank and little out of the sump.
If you change oil after parking hot and waiting a month, you will get most oil out of the sump and still some out of the tank.


All this brings up some important engineering points that Porsche didn’t realize in the original design and finally implemented with the GT2/GT3 dry-sump oil systems.


First is the distance between the oil tank and the inlet to the pressure pump.
Shorter is better and a ‘tall’ tank provides more ‘head’ to the oil.

Next is the lower effectiveness of the front cooler considering the scavenged oil is about 40% air (the size ratio between the pressure pump and scavenge pump).

Third is the lack of an oil/air centrifugal separator to insure pure (no entrained air) oil to the bearings.

Fourth is our old oil tanks must carry about twice the oil (compared to a GT2/GT3) in order to not cavitate (suck air) under some conditions (the well-intentioned '72 is the worst offender).

It is senseless to have the oil filter in the scavenge circuit and not protecting the engine from oil tank contamination.

There are more.



All this brings me to lobby some of our very talented aftermarket suppliers to engineer an up-dated oil system for our nice old 911s (and 914-6s).
The oil tank should be in place of the engine oil cooler (like GT2/GT3).
The centrifugal separator should be integral with the pressure regulating valve (so it regulates pressure by bypassing the seperated air).
The pressurized and separated oil should go to the front cooler (with additional pressure by-pass for cold start-up).
The oil filter should be just prior to the engine lubrication.
There should be adequate screen filters to protect both pumps from foreign objects (meaning if it passes the screen, it won't damage the pump).
There should be improved (single) thermostat function for faster warm-up.
All this can be done with a new external oil tank and no changes to our engines.
In 20/20 hindsight, this system was within the capabilities of Porsche in the '50s, perhaps pre-war.

2¢

Best,
Grady