25 replies to this topic

[Manfred Cubenoggin]

Hi, fellas!

Don't get over to these here parts very often...

Re the Porsche 917 in its various guises...

The original endurance cars featured the 12-cylinder engine with assorted displacements upwards of 5-litres and power
figures pushing 600 bhp nominal. Of course, all were air-cooled via a singe fan mounted on top of the engine which fed
ducting chambers around the cylinders and all.

That's all well and good but...

The CanAm variants of 1972-73 featured turbo-charging and vastly increased power values. For short periods, I understand
that the 917/30 with boost turned WAY UP could briefly register 1500 bhp.

I have a question re the cooling of the CanAm turbo engines. It appears to my eyes that the engine configuration was
largely static from its debut in the endurance racers right through to the turbo models. Same fan position and size and all.

How did Porsche manage to adequately cool an engine that produced 100%+ more power(and power is heat)and make it live?

Thanks for all replies...

[gruntguru]

I have no specific knowledge of these engines but would suggest that (for one thing) a large valve timing overlap would greatly assist by scavenging heat from the hottest surfaces in the engine.

[bigleagueslider]

Consider that ultimately every race car engine is air-cooled. It's just that current race engines first transfer engine heat to the liquid coolant flow, and then the liquid coolant transfers that heat to the airflow passing over the radiator core.

[desmo]

Couldn't it be accomplished by putting in a bigger fan and ducting that flows more, lots of circulating oil and larger oil coolers? Even nominally air cooled engines are to a significant degree liquid cooled by their circulating oil.

[mariner]

The top mounted cooling fan was pretty big and its gearing could be altered fairly easily I would think. Also they had a big oil cooler and some pretty large bore oil lines to get heat out of the oil.

[thegforcemaybewithyou]

Maybe cooling through a rich mixture is part of the solution?

[J. Edlund]

Consider that ultimately every race car engine is air-cooled. It's just that current race engines first transfer engine heat to the liquid coolant flow, and then the liquid coolant transfers that heat to the airflow passing over the radiator core.

 

That may be the case, but air has a poor heat transfer capacity which means we need a rather large area to transfer a given amount of heat. It is in other words far easier to cool a radiator with air that provides a large area, than a small engine cylinder and head. If you look at air- and watercooled WW2 aero engines you will notice one difference between these, the watercooled engines had smaller displacements for a given output, something made possible due to their better cooling.

[Manfred Cubenoggin]

Thanks for all the replies thus far, gents. Hope for more!

[kikiturbo2]

I think you are underestimating the cooling trough engine oil and oil coolers...

[Manfred Cubenoggin]

Cooling via the circulating oil may well be the key. In a long-since binned issue of Road & Track that I owned back in the day, there
was an extensive article on the Porsche 917/10 that debuted in the 1972 CanAm series. The feature included photos plus sketch art from
Werner Buher and some technical data. I seem to recall...and was shocked to read...that the dry sump oil tank had a capacity of some
50+ litres which to my mind, was vastly more than any other typical race car of the day in any formula. Can anyone confirm that capacity?

I've Binged assorted sights and can't seem to find a figure on that detail.

[Manfred Cubenoggin]

My bad...

Just found the Werner Buher sketch art for the 917/10 on my very own HD. The capacity for the dry sump tank is noted as
21 quarts. I believe that translates to 19.9 litres. So much for my memory...

But is that not a lot of oil to carry? I remember my old FF's carrying something on the likes of four to six litres of oil in the tank.
Granted that the Porsche engine is vastly bigger.

I'd post the sketch but have long since lost the touch about posting images at TNF and I suspect it's likewise here.

[NeilR]

I have photo's of the 917/30 somewhere, will try and link them tomorrow. It had a substantial oil circuit.

[bigleagueslider]

My bad...

Just found the Werner Buher sketch art for the 917/10 on my very own HD. The capacity for the dry sump tank is noted as
21 quarts. I believe that translates to 19.9 litres. So much for my memory...

But is that not a lot of oil to carry? I remember my old FF's carrying something on the likes of four to six litres of oil in the tank.
Granted that the Porsche engine is vastly bigger.

I'd post the sketch but have long since lost the touch about posting images at TNF and I suspect it's likewise here.

Yes, 21 quarts (or over 5 gallons) is a significant volume of lube oil for a race engine. But compare that to the 20+ gallons of gearbox lube oil carried by the CH-53 helicopter. The CH-53 helicopter gearbox is oil-cooled, but the heat rejection thru the lube oil flow is only about 4-5% of engine power, versus something like 10-15% of engine power in the 917 oil system. While the specific heat of water is much better than oil, oil has the advantage of being able to handle much higher temperatures than water. A liquid/air heat exchanger with oil at 300degF and air at 80degF is quite efficient.

[Lee Nicolle]

The joke of all of this is, yes you can cool and engine with oil, never as efficient as liquid cooling. Porsche [and VW]  have proved this with liquid cooled cylinder heads for over 30 years now. That and it is far lighter, simpler and cheaper. Those huge fans coupled with multiple oil pumps are heavy and consume considerable power too. A simple 3 stage dry sump pump consumes about twice the power alone to that of a production oil pump. Plus the tank, lines, filters and extra oil volume is all heavy and space intrusive too. BUT then you can drive the car and engine very hard hopefully not worrying about oil surge and aeration. Plus far less oil and crankcase pressure means less leaks and the crank is not throwing excess oil on cylinder walls yet alone buckets of oil around the valve stem seals.

 

The more oil you put in an engine is more oil that has to be controlled. That alone has always been a disadvantage of simple OHC engines,, all that oil flailing around to lube the cams relies on very good stem seals to stop the engine drinking oil. Plus to keep said oil contained in the engine is not that easy either. Yes ofcourse it happens though replacing stem seals and hard gaskets is something that happens regularly after 100000km, or on a race engine far more often.

[johnnycomelately1]

for what it's worth, here some informations, which I have read somewhere, but I can't vouch for them, so use at your own risk:

 

the cooling fan was powered via a shaft drive from the center of the engine ( the engine also had a center power take-off, to combat torsional oscillations in the crankshaft)

 

shaft OD & material:

 

N/A engine: 22mm / steel

turbo:          24mm / titan

 

max. airflow rate:

 

N/A: 2400 l/min

turbo: 3100 l/min

 

power consumption of the fan:

 

N/A: 17 hp

turbo: 31 hp

good points and is the obvious answer to a good question.

i guess there was also a different fan shape also.

the oil solution i am not so sure about as from my memory :was there any oil squirters etc??

[johnnycomelately1]

of course the fin area may have been increased as well???

[VWV]

Porsche presented a SAE technical paper on the 917, which from what I have read online goes into amazing detail on the design of the 917. I have never actually read the paper, I have only read someones review of it.

 

Any chance a SAE member could post a scan of it????

[johnnycomelately1]

Porsche presented a SAE technical paper on the 917, which from what I have read online goes into amazing detail on the design of the 917. I have never actually read the paper, I have only read someones review of it.

 

Any chance a SAE member could post a scan of it????

Yell louder VWV

[gruntguru]

I had a quick look and didn't find it. Post a link or title when you make that kind of request.

[bigleagueslider]

There is a 1979 SAE paper (780718) written by Hans Mezger that provides technical details of the 1100hp 5.4L turbocharged, air-cooled, flat 12 engine used in the 1973 917 for CanAm racing. The engine weighed 285kg. The cooling fan ran at up to 9000rpm and delivered a volume of 2.8L/second/hp.

[gruntguru]

Yes, I read that paper and it doesn't have much information on cooling the 917 beyond that already posted by TC3000 in post #6.

[Manfred Cubenoggin]

Am keeping an eye on my thread, chaps, and have appreciated the inputs thus far. TYVM.

[bigleagueslider]

Yes, I read that paper and it doesn't have much information on cooling the 917 beyond that already posted by TC3000 in post #6.

You should pay closer attention to what you read in posts. The SAE paper I noted states the cooling airflow for the 630hp N/A 5.0L flat 12 917 engine was 2400L/second and it was 3100L/second for the 1100hp turbocharged 5.4L flat 12 engine. The numbers for cooling airflow volume noted by TC3000 in post #6 were in L/minute.

 

A couple other interesting statistics listed in the SAE paper I noted:

 

- The CR of the N/A engine was 10.5:1, but the turbo engine with a modest 1.4bar boost only had a 6.5:1 CR.

- The turbos used ball bearings and turned up to 90,000rpm, which was pretty impressive for 1973.

[Manfred Cubenoggin]

Many thanks to all whom have contributed data to my thread.

I've been examining cross sectional views of the 917 engine and am trying to determine the step-down/step-up ratio of the cooling fan re the crankshaft rpm.
From what I can make out, I'm estimating a step-down of apx 3:1(ie, 6000rpm at the crankshaft = 2000rpm fan speed). Could anyone confirm?

Thanks.

Edit: Just noticed something with respect to the cooling fan on the 917 variants. It would seem that the original configuration featured on the enduro racers
had a six blade arrangement whereas in the saved images on my HD I just noted moments ago have the turbo 917/19 & 917/30 with eight blades. Additional flow
potential seems the obvious goal.

Edited by Manfred Cubenoggin, 08 September 2015 - 15:27.

[bigleagueslider]

I believe the 917 fan ran at a higher speed than the crankshaft. However, I don't think it was anywhere close to 3:1.

[Tim Murray]

In the 'Cars in Profile' publication on the Porsche 917 written by Paul Frere, the following appeared in the Specifications listing:

Forced air cooling by plastic axial blower (with cushioned hub from 1970) running at 0.895 times engine speed (1.12 times engine speed in 4.5- and 5-litre turbo-charged engines, 1972).

This profile was published in January 1973, so covers the 917/10 but not the 917/30.