33 replies to this topic

[bobdar]

Does anyone have information re. crankcase vacuum in F1?? Assuming it is being utilized, are they using over-scavenge via a multi-stage dry sump suction section, or augmenting with a separate vacuum pump? Any idea of the amount of vacuum used?

[desmo]

The crankcases are run at "near vacuum"- the last psi or so isn't critical- each crank throw having it's own chamber with typically three Roots-type air pumps per chamber scavanging air/oil.

[bobdar]

Thanks, Desmo. So, they are drawing a partial vacuum of ~26" Hg, and using a 3-stage scavenge pump on each chamber. Pardon my ignorance, but what is the acvantage of the separate chambers??

[MRC]

Typically anything more than a half of an atmosphere is difficult to acheive with even the best sealing. I would be suprised if F1 cars were running 26" Hg, even under part load. If some vacumm is good, more vacuum does not always mean better than good. More vacuum means lower scavenge pump efficiency.

Seperating the oil in the pan, or 'baffling', allows you to keep the oil in closer proximity to the oil pickup, more of the time.

Desmo, how recent have you heard that any team was running a roots style oil scavenge pump? I was under the impression that they were all using gear type or gerotor scavenge pumps, but I don't know for sure.

mark

[bobdar]

Well, that makes sense, having multiple smaller compartments could reduce the amount of air entrained in the oil. But this is really the heart of my dilemma: it seems that if the crankcase is scavenged by the dry sump pump alone (and drawn down to a partial vacuum), there would be a lot of air introduced into the oil. An independent vacuum pump would reduce the aeration, but then the scavenge efficiency would be affected unless the oil reservoir was at the same partial vacuum. From Desmo's reply, it sounds as if they use scavenge pump only (no augmenting vacuum pump) and deal with the aeration. Correct?

[MRC]

Ya, I highly doubt that they are running an additional pump that's sole purpose is vacuum generation. Reducing crankcase vacuum will not reduce your entrained aeration. It might decrease the dissolved air, but who cares (going to make things worse). If the pickup/scavenge pump was operating at 100% efficiency (not going to happen), and you were picking up 100% oil (ie-no air, also not going to happen), you are still going to entrain air into the oil, because you are not likely going to pick up oil all the time. BTW, as you increase the crankcase vaccum, and thus decrease the pump front side absoulte pressure, you increase the sensitivity of efficiency loss, due to oil aeration. So, say you are at relatively low vacuum and at some rpm, and are picking up 0% aerated oil at 10gpm. Now you pick up 50% aerated oil, but might be doing it at 4.8 gpm (of oil flow). Now for very low vaccum, these numbers might be 9.9gpm for the first case, and maybe 2 gpm for the second. These numbers are pulled out of my ass, but you get the idea. Drastically increasing the vacuum, will make you suffer just when you need the pumping efficiency most. Hope that made sense. If you would like to see a graph of this, let me know, and I'll PM it to you.

[desmo]

The charactarisation of F1 crankcase pressures as "near vacuum" and description of the scavange pumps are from Peter Wright's book "Formula One Technology", which seems both pretty current and accurate in most respects- agreeing with any facts I've independantly obtained. He cites reduced aerodynamic losses as the primary motivation driving the employment of the rather elaborate array (15, usually!) of scavanging pumps. How does crankcase vacuum affect the bmep due to differing pressure ratios above and below the piston or does this cancel out through the four cycles?

[MRC]

I think the main benefit to the reduced crankcase pressure, is initial ring sealing. (I think there are a few SAE papers on this). The reduction of aero losses in the engine I don't think can be linked to just lower pressure every time. The reduction of cutting through strings of oil, and oil mist, are the main benefit to power increases, as far as the windage losses go. I also think that the amount of windage losses to get rid of is highly dependant on the specific engine being looked at. With an F1 engine, with the crank so close to the oil pan, I imagine that there is more to be gained with getting rid of some windage losses. There is really no good way to seperate all of the effects of crankcase vacuum either. I would have to agree with you that the vacuum is really cancelled out through the up and down strokes of the engine. Although I imagine that you could find a few people to disagree with me on that.

I think that 'near vacuum' should still be taken with a grain of salt. The ONLY thing that is pushing oil into the oil pump, is the pressure in the crankcase. Reduce the pressure in the sump, and lose pump efficiency, there is no way around it. I'll believe a little more than a half of an atmosphere of vacuum, but not much more. Also, the vacuum will vary on rpm and load. How much, will depend on the degree of sealing. How much crankcase vacuum are you talking about, exactly?

[desmo]

It would be interesting to me to see a data trace from a pressure tap inside of of the crank chambers to see how much the pressure fluctuates as the volume under the piston changes as each pair rise and fall. It seems to me that bank angles would effect the degree of crankcase chamber volume change as well. Might most of the scavaging occur as the pressure intermittantly builds as the crank throw cycles through it's bottom? And might this pressure change be usefully modified by linking pairs of chambers using breathers from one to another?

[Wolf]

I think that Desmo addressed the very important issue, IMHO- the pressure difference between cylinders and the crankcase.

An excercise in futility but with figures like 3.0l, 18.5k rpm and 800BHP, I think BMEP could be *very roughly* estimated at almost 13bar (now, I don't have a number to compare it against, so feel free to correct me). Reducing the crankcase pressure by 0.5bar will increase the pressure difference by the same value so the new BMEP would be 13.5bar which translates into 3.5% BMEP, and hence BHP, increase.

But what remains unknown, even if I'm right, is how much of power is used to accomplish 0.5bar pressure decrease... :

[MRC]

What about the higher pressure difference acting on the piston, on the upstrokes?

[Wolf]

MRC- I belive a flywheel could be used to compensate (or other pistons' performance gain, part of it) that, and it would use insignificant changes, maybe a few tenths diameter increase.

[bobdar]

It seems to me that the forces cancel out across the total number of pistons/cylinders. If not, the engine would sustain rotation on crankcase vacuum alone.

[12.9:1]

bobdar wrote ___________________________________________________________________________
it sounds as if they use scavenge pump only (no augmenting vacuum pump) and deal with the aeration. Correct?
___________________________________________________________________________



The oil/air leaves the positive displacement pump i.e.: roots/gear at essentially atmospheric
{or a bit above} pressure then it go's straight into a centrifugal separator and on to the tank.


regards Eric

[Yelnats]

I find it hard to understand how an F1 crankcase could operate at anywhere near an absolute vacuum as this would reduce the vapour pressure and result in the oil boiling at much lower temps than if close to ambient pressure was maintained. Operating a crankcase at a "near vacuum" doesn't tell me anything about the pressure other than it runs at a lower than ambient atmospheric pressure. Similarly the term vacuum is correctly applied to a household vacuum cleaner that may run at only a 10 or 20% below ambient pressures.

[desmo]

Could effective bmep be improved, if even incrementally, by linking cranckcase chambers in such a way that as the crankpin in one chamber was descending and the effective volume under the pistons was diminishing, the crank was vented into another chamber whose crankpin was rising? I realize that an F1 crank with it's odd pin spacing doesn't have any diametrically opposite throws to maximize the potential effect, but I can picture a breather operated by a mechanical valve- perhaps a simple reed valve or valves might suffice- that might reduce pumping losses.

[palmas]

Dont know in F1, but we use vacuum pumps to remove gases produced by the oil (and air, if there is some). It helps the foam (litle as it has to be, otherwise there is no lubricating) expand in the sump and not in the engine. But this is big engines with shitty oil compared to F1.

[malbear]

A lot of drag racers and Motorcycle race teams use a simple means to achieve crankcase vacuum.
A small venturi is introduced into the exhaust stream and a tube runs from this venturi to a one way diaphragm or reed valve screwed into the crank case. Crank case or sump gas is sucked out by the venturi and cannot return because of the one way valve. The Britten bike suffered from oil consumption until the team was informed of this simple device.

Malbeare
sixstroke

[Chickenman]

Last year I was present in a Dyno shop when we were testing a Road Racing Big Block Chevy. One of the tests we did was to measure the effects of crankcase vacuum on power output. I can't remember the exact vacuum figure that we were able to make ...but it was quite high , 6 to 8 psi of vacuum in the crankcase to the best of my memory. Five stage Dry sump.

We saw a consistant 10hp increase throughout the usable rpm range and a similiar increase in torque when vacuum was applied to the crankcase vs the crankcase being vented to atmosphere.

The Dyno operator, who was very experienced, said that vacuum in the crankcase greatly reduces ring flutter, which of course results in better ring sealing and more power. All state of the art racing engines utilise vacuum evacuation of the crankcase for this very purpose. It's a well proven engineering principle.

[Bill Sherwood]

Hi Mal, I think they're called a 'Vacu-sump', and are common in a lot of classes in the US.

[Engineguy]

5 inHg (2.5psi) crankcase depression is enough to:

* eliminate the need for valve stem seals for drag reduction (assumes heads are vented to crankcase)
* eliminate the need for oil rings on the pistons (or very weak zero-drag oil rings)
* eliminate totally any oil contamination of the combustion chamber (slightest minute trace causes detonation which you must design out or "tune out" with the usual methods that reduce power)
* eliminate compression ring flutter and maintain perfect seal at WOT and near-WOT (assuming intake port tuning is somewhat correct)

With a system of check valves and reed valves, this level of depression can be obtained "passively" from an extractor tube in the exhaust collector (siphon effect). No problem to design the dry sump pump to assure the depression. I believe one section of Cosworth pumps have been of the roots type since way back.

[Mark Beckman]

Originally posted by Chickenman
Last year I was present in a Dyno shop when we were testing a Road Racing Big Block Chevy. One of the tests we did was to measure the effects of crankcase vacuum on power output. I can't remember the exact vacuum figure that we were able to make ...but it was quite high , 6 to 8 psi of vacuum in the crankcase to the best of my memory. Five stage Dry sump.

We saw a consistant 10hp increase throughout the usable rpm range and a similiar increase in torque when vacuum was applied to the crankcase vs the crankcase being vented to atmosphere.

The Dyno operator, who was very experienced, said that vacuum in the crankcase greatly reduces ring flutter, which of course results in better ring sealing and more power. All state of the art racing engines utilise vacuum evacuation of the crankcase for this very purpose. It's a well proven engineering principle.

Personally my favorites are the round reed valves found on Mazda engines late 1980's meant for induction of air into the exhaust manifold to help burn fuel particles in the exhaust for lower emissions. These on the exhaust (at the collector) and common brake booster hose with one way valves from the inlet manifold give me vacuum whether on or off the throttle to a bit more than 5 inHg to help remove the "oil rope" off the crank (oil that stays trapped in the drag behind crank throws) and thats where part of the HP increase comes from.

I set my Brother-in-laws Jet racing boat Chev 400 up last year with this system and showed 15 Hp increase on the dyno .

I hear top Drag Racers actually turn their crank seals around they are getting such vacuum.

[J. Edlund]

The Ferrari Tipo 049 did use 12 gerotor pumps for scavenging, two for each chamber + one for each head. The pressure pump is can be either a gerotor or a gear type pump.

The engine has no "sump", the crankcase is formed so it follows the maximum radius of the crankshaft. The oil/air pickup is placed tangentially in the crankcase, like a port in the chamber.

The engines are fitted with an oil/air separator which separates the air from the oil, oil is fed to the tank. Air is also fed to the tank, but at a higher place. The tank is ventilated into the plenum chamber to prevent oil getting out on the track. The tank feeds the oil pressure pump.

Crankcase pressure was mentioned by Lamborghini in SAE paper no. 942518 to be 0,06 MPa.

[hydra]

This may sound like a silly question, but are (optimized) low-pressure crankcase setups streetable in the long term? And why don't any OEMs use the vacu-sump principle to reduce windage somewhat?

[J. Edlund]

Originally posted by hydra
This may sound like a silly question, but are (optimized) low-pressure crankcase setups streetable in the long term? And why don't any OEMs use the vacu-sump principle to reduce windage somewhat?

Some supercars use it.

Why it isn't used by OEMs? Cost I suppose. For example, Toyota accept one dollar in increased engine manufacturling costs for one percent of reduced fuel consumption.

[ciaoduc1]

I imagine cost is definately the reason.
I don't know how many engines your average big car manufacturer makes per year but I bet it's in the upper hundreds of thousands. Just multiply an additional cost of, say $200 per engine...you're talking 10's of millions of dollars for only a 1-2% (just a guess) gain in hp.

[hydra]

I guess I should have added that I KNOW that cost is a major reason... My question was oriented more towards high-end cars than regular bread and butter cars, although I'm sure that the reduction in FMEP by using low-tension oil rings might make it worth their while. I was more concerned with high mileage durability and ability of the pump/gaskets/etc to maintain useful vacuum for 200,000km+

[275 GTB-4]

Originally posted by hydra
This may sound like a silly question, but are (optimized) low-pressure crankcase setups streetable in the long term? And why don't any OEMs use the vacu-sump principle to reduce windage somewhat?

Ummmm BMC A Series exploited the horsepower gain possible from this phenomenon many years ago...people have found that removing the standard PCV valve can result in a loss of power.

[Catalina Park]

Originally posted by 275 GTB-4


Ummmm BMC A Series exploited the horsepower gain possible from this phenomenon many years ago...people have found that removing the standard PCV valve can result in a loss of power.

Yes, the power goes down and the oil leaks out!

[275 GTB-4]

Originally posted by Catalina Park
Yes, the power goes down and the oil leaks out!

I don't think the two are connected Mike...you are guarrantedd to lose A Series oil no matter how hard you try!!

[J. Edlund]

Originally posted by 275 GTB-4


Ummmm BMC A Series exploited the horsepower gain possible from this phenomenon many years ago...people have found that removing the standard PCV valve can result in a loss of power.

For road cars there are also other problems with this, poor crankcase ventilation leads to oil sludge which can clog the oil pickup with serious engine damage as the result. A problem which have increased with the use of low tension piston rings. There's also the emission problem...

[shaun979]

In closed throttle situations at upper RPM. Doesn't the high amount of vacuum in the cylinder overcome the vacuum in the crankcase and head and pull oil into the cylinder if seals are not installed? Or are there check valves somewhere along the way that prevent this? Even with a check valve there would be some oil that gets pulled into the port and cylinder no?

[Engineguy]

Originally posted by shaun979
In closed throttle situations at upper RPM. Doesn't the high amount of vacuum in the cylinder overcome the vacuum in the crankcase and head and pull oil into the cylinder if seals are not installed? Or are there check valves somewhere along the way that prevent this? Even with a check valve there would be some oil that gets pulled into the port and cylinder no?

I haven't thought about this for a long time, but drag racers (before they used vacuum pumps) used a system that pulled a vacuum in the crankcase using intake manifold depression at closed throttle (that makes the vacuum the same at both ends of the valve guide since the rocker area was open to the crankcase for oil drainback) and exhaust header collector siphon effect at wide open throttle. I'll let you figure out where they cleverly put the reed valves and oil traps to accomplish this. I would assume high RPM closed throttle never happened (i.e. kick the clutch in at finish line) .

[J. Edlund]

Originally posted by Engineguy

I haven't thought about this for a long time, but drag racers (before they used vacuum pumps) used a system that pulled a vacuum in the crankcase using intake manifold depression at closed throttle (that makes the vacuum the same at both ends of the valve guide since the rocker area was open to the crankcase for oil drainback) and exhaust header collector siphon effect at wide open throttle. I'll let you figure out where they cleverly put the reed valves and oil traps to accomplish this. I would assume high RPM closed throttle never happened (i.e. kick the clutch in at finish line) .

The same setup as you mentioned is today used in most road cars, the siphon effect in the collector have however been changed to occur in the intake before throttle/turbo/supercharger instead.

The system contains one way valves, restrictors and oiltraps with drainback to below the oil level. The one way valves control the flow so gas only can be evacuated from the crankcase while the restrictors limits the crankcase depression (too low can cause problems for the oilpump and that the oil would follow the gases out).