161 replies to this topic

[Tim Murray]

Ferry Porsche tells the same story, worded slightly differently, in Chris Nixon's Racing the Silver Arrows:

Late in 1935 we made a change to the car which provided a dramatic improvement in its handling. One day I went with my father to the Nürburgring for some testing and took a good look at the car's behaviour during testing. When accelerating out of the corner, the inside rear wheel would spin furiously, trailing a stream of rubber smoke. We already had experience of this problem on some of our touring cars which had a locked differential and we had cured it by fitting a limited-slip unit. I pointed this out to my father and suggested the same cure for the Auto Union. He immediately agreed and we went back to the pits straightaway and put in a call to ZF in Stuttgart. They built up a limited-slip differential for us very quickly and the difference was quite extraordinary.

Throughout 1936 all the Mercedes drivers noticed how much better the Auto Unions accelerated out of the corners, but they didn't know why. At the end of the season Rudolf Caracciola came to visit us in Stuttgart and by then he knew the reason. Although he had lost the European Championship to Bernd Rosemeyer he was very good about it and admitted for many months he had no idea why the Auto Union was suddenly so superior!

I'm not sure I understand why a limited-slip diff would have improved the traction of a car which already had a locked differential . . .

Originally posted by Ray Bell

According to one book, the NSW 6/30 (1.6 litres) was 'very successful in competition'...

Maybe, as hillclimbs were very much in vogue in those times, this was the car spoken of and it's likely in that case that a limited slip diff would have been very helpful.

This is a good point. However, if the use of a limited slip diff in cars of this kind was not unusual, and its advantages were known, I do wonder why it wasn't already being used on GP cars.

[Ray Bell]

Originally posted by Tim Murray
.....I'm not sure I understand why a limited-slip diff would have improved the traction of a car which already had a locked differential.....

A locked differential will still mean that the inside wheel is spinning in a tight corner, though not spinning as furiously as it might if it had no restraint at all. Perhaps the limited slip differential allowed it to retain full traction without slippage?

This is a good point. However, if the use of a limited slip diff in cars of this kind was not unusual, and its advantages were known, I do wonder why it wasn't already being used on GP cars.

Cost? Lack of faith in the item? Maybe they didn't expect it to handle the power?

[David Beard]

Originally posted by Roger Clark

Karl Ludvigsen's "Mercedes-Benz Racing Cars" says that they first used the ZF in 1935. I would normally trust this work.

Is there a picture or diagram? Does it look the same?

[Roger Clark]

Originally posted by Ray Bell


A locked differential will still mean that the inside wheel is spinning in a tight corner, though not spinning as furiously as it might if it had no restraint at all.

How? Surely a locked differential means that the wheels must rotate at the same rate, and therefore both or neither must spin.

Originally posted by David Beard

Is there a picture or diagram? Does it look the same?

No, but there is a description in Pomeroy's "The Grand Prix Car" and it sounds the same.

Incidentally Karl L also says that during 1936 Daimler-Benz also used a fully locked rear end. At Tripoli in May Caracciola complained of understeer and the locked diff was replaced by a ZF on his car.

[David Beard]

Originally posted by Roger Clark

How? Surely a locked differential means that the wheels must rotate at the same rate, and therefore both or neither must spin.

With the sort of power and lack of grip these cars had, both wheels spinning coming out of a corner would have always been the case for a locked diff, hence the advantage of the LSD. In a straight line, however, the LSD would be no better.

After the war, who was using which sort of LSD, then? The options seems to be the ZF as discussed, which does away with the planet gear assembly, and various types retaining the star wheels but with addition of clutch plates. Then there is the Torsen type, containing lots of extra baffling gears.

And if we are to believe modern comics (which never tell us how things work), diffs are now stuffed full of microchips!

[Roger Clark]

Originally posted by David Beard


With the sort of power and lack of grip these cars had, both wheels spinning coming out of a corner would have always been the case for a locked diff, hence the advantage of the LSD. In a straight line, however, the LSD would be no better.

I've obviously misunderstood the whole thing.

I thought the whole point of an LSD was that it acts like an normal diff at ordinary cornering speeds, but locked if, for example, the lightly loaded inside wheel started to spin. In this case (I thought) the LSD was acting like a locked diff.

Sorry about the confusion.

[David Beard]

Originally posted by Roger Clark


I've obviously misunderstood the whole thing.

Don't think so

I thought the whole point of an LSD was that it acts like an normal diff at ordinary cornering speeds, but locked if, for example, the lightly loaded inside wheel started to spin. In this case (I thought) the LSD was acting like a locked diff.
[/B]

I suppose the ZF would lock completely, but not necessarily other types?

Sorry about the confusion.

I'm confused now :

[Ray Bell]

The ZF is a softer lock than the Detroit Locker, for example...

It limits progressively, depending on need, whereas the Locker is either open or locked.

I see your point, David, that both wheels would be spinning anyway, but that would be losing traction with the kind of tyres they had in the thirties.

Torsens are a very modern invention, aren't they? Maybe from the seventies?

[David Beard]

Originally posted by Ray Bell
The ZF is a softer lock than the Detroit Locker, for example...

It limits progressively, depending on need, whereas the Locker is either open or locked.

I've Googled upon the Detroit Locker...seems to be on or off. Whether the ZF operates more softly, not sure...

Bonde, can you help?

[Bonde]

David,

You knew I was lurking, didn't you?;)

I must admit I can't help ypu off-hand - but now my curiousity is stirred, so I'll look into it!

However, looking at the bits of the ZF, I'm reminded of the good old free-wheeling device used in a bicycle hub, so I would venture to guess that the ZF lock-un-lock operation is at least somewhat progressive. I assume a 'bang-bang' operation would render the car very unstable and difficult to drive, which wouldn't be much of a problem on a lorry ambling slowly over rough ground, but would be on a racing car in a high speed turn...

[Tim Murray]

There is a bit about the ZF diff in the article below, which appeared in Motor magazine in December 1965. Go to the third image along if you don't want to read the whole article. It indicates that the locking effect is progressive as torque increases.

[VWV]

Thank you very much Tim for posting that article. I wish I could see more of these "old" technical articles as they were very much before my time.

[Tim Murray]

You're most welcome. The old Motor magazine, before it was merged with Autocar, used to publish some very interesting technical articles. Since you asked, here's another rather more light-hearted example from Christmas 1965 (which just happens to be next to the article on differentials in my cuttings file). It's based on the good old British put-down, "Nah mate, that thing couldn't pull the skin off a rice pudding!"

[David Beard]

Originally posted by Tim Murray
There is a bit about the ZF diff in the article below, which appeared in Motor magazine in December 1965. Go to the third image along if you don't want to read the whole article. It indicates that the locking effect is progressive as torque increases.

Thanks Tim

[David Beard]

I took this photo of a bench full of dismembered queerbox a year or two ago...I have only just noticed that the the ZF LSD under discussion also features...

[P 4 Staff]

Interesting thread.
I just seem to have too little time to read it all.
Anyway...I took a picture of the Brabham front suspension at Östereichring in 1972...if it´s of any interest.

Staff.

[Bonde]

I thought it appropriate to re-open this thread for the technophiles with a few questions:

When comparing F1 cars of the 60s and early 70s with later designs, even those sporting DFV engines, it appears to me that the oil tanks were a lot larger in the 60s and 70s than they subsequently became. For instance, on a March 701 the tank appears positively huge (though relatively small compared to that of a BRM P160, it always reminded me of a milk churn), whereas tanks appeared smaller on later DFV-powered cars. My questions then are:

Was the oil capacity requirement for the DFV (and other F1 engines of the period) reduced with development?

Did later versions of the DFV simply consume/leak less oil, or were the lubrication systems and lubrication cooling systems improved so as to enable less oil tankage?

Were the oils then used consumed more readily than modern oils?

Or am I just fooled by an optical illusion that makes the tail or nose mounted oil tanks appear much bigger than those installed in front of the engine or in the bellhousing?

[David Beard]

Originally posted by Bonde
I thought it appropriate to re-open this thread for the technophiles with a few questions:

When comparing F1 cars of the 60s and early 70s with later designs, even those sporting DFV engines, it appears to me that the oil tanks were a lot larger in the 60s and 70s than they subsequently became. For instance, on a March 701 the tank appears positively huge (though relatively small compared to that of a BRM P160, it always reminded me of a milk churn), whereas tanks appeared smaller on later DFV-powered cars. My questions then are:

Was the oil capacity requirement for the DFV (and other F1 engines of the period) reduced with development?

Did later versions of the DFV simply consume/leak less oil, or were the lubrication systems and lubrication cooling systems improved so as to enable less oil tankage?

Were the oils then used consumed more readily than modern oils?

Or am I just fooled by an optical illusion that makes the tail or nose mounted oil tanks appear much bigger than those installed in front of the engine or in the bellhousing?

Good point

Some of the visible DFV tanks were sort of conical, weren't they?. The term "swirl pot" used to be around, didn't it? Something to do with de-airation?

And when did synthetic oils arrive...did they make a difference? Mobil1 certainly seemed to smooth out the wrinkles in my old very high mileage Saab 9000 Turbo.

[bradbury west]

Mobil 1 seemed to come to the fore in the v early '80s with Williams F1, although they had had their Mobil SHC?? well before that. Bear in mid Amsoil had been going for a very long time prior to that, their comment being that they spilled more than Mobil sold of Mobil 1. IIRC synthetic oils, or so I was told at Amsoil, was developed to support jet engines as normal oils, even earlier ester based stuff would "crack" under the temperatures experienced, to do with film strength etc, for which we used to demonstarte it with Falex wear tests, if that is the right term.

There was also a suggestion via Amsoil's man that synthetics were invented, or development accelerated, to prevent seizure from the frost etc on machine guns and other armaments of the Russian front in 39/45. We used to demonstrate it by asking the prospect to leave a provided bottle of Amsoil in his freezer for a coupe of days along with some of their regular oil in another provided ientical bottle, and then test them for low temp pour capability. It was interesting. It also had higher stability at high temp with lower? volatility. It was along time ago.

In 1982 a chemist at Mobil confirmed to me that they were obtaining 5-10 bhp more running on synthetic on a DFV at the time, a 505 bhp motor then. I had better results when Sammy Nelson did back to backs on his dyno with an FF engine

Roger Lund.

[Frank S]

During 1975 I ran a half-hour mid-day practice session at Riverside International Raceway, ambient about 90º F, with Castrol GTX 20-50 in the Vega engine. Oil temp was 225º at the end. While the next group did their half hour I made a quick oil and filter change to Gurney's synthetic, newly on the market.

Next half-hour practice, oil temp at the end was 210º F, no other changes than that ambient had increased by five degrees.

Worked for me. My cars were always street-driven as well as on track, so I drained that special oil and reserved it for track use. Seems to me it cost something like eight dollars a quart.

I wonder if that container of oil is still in the garage somewhere (probably over in a corner with the Vega Konis). I bet a couple quarts of Gurney oil would bring a price on eBay - in the original cans.

--
Frank S

[bradbury west]

That is xactly the key characteristic which Sam Nelson picked up on on the dyno test, more so that the apparent 3bhp gain,on a 105bhp motor, when the figures were converted.

He noticed that the fans for the cooling radiators were cutting in far less frequently, and that with a large number of readings taken at 100rpm , IIRC, increments from 4500 upwards . His argument was that it meant the dsigners could use a smaller radiator, which wold be lighter, but also which would need less exposure to the air, reducing frontal area etc.

The oil prsssure read higher and remained more constant at sustained testing intervals. Trucks used synthetic as it improves mileage on fuel etc. We used to enjoy showing how it reduced friction on larger multi cylinder engines etc. I was working closely with a Home Tune engineer who did all the clever stuff, and gained extra business. Synthetic works.

RL

[Bonde]

Originally posted by David Beard:
Some of the visible DFV tanks were sort of conical, weren't they?. The term "swirl pot" used to be around, didn't it? Something to do with de-airation?

Swirl pots are still around (and round ;) ). The oil (or coolant) is returned tangentially into the cylindrical or conical portion of the tank, thus separating oil and water by centrifugation. The conical tank is ideal because it best maintains the rotation of the oil and collects well at the outlet at the bottom apex under lateral and longitudinal acceleration. I'm not up to speed on current F1 technology, but I'm sure swirlpots are still used in both the cooling and lubrication systems, as they are on any other racing cars, the former integral with the header tank and the latter with the oil tank, whether located in front of the engine or in the bellhousing.

Re my original question: Do I understand correctly that the switch to synthetic olis enabled smaller tankage due to reduced loss and degradation?

I suppose ever tightening of engine manufacturing tolerances and improved materials and surface treatment technologies have also enabled less oil to be consumed and carried.

[Macca]

Originally posted by Roger Clark
The ZF apparently required very great care during assembly and its performance deteriorated markedly when the plungers became worn. These things are acceptable on a racing car but not, usually, on a passenger car or commercial vehicle.

Having just rediscovered this thread, this made me think of Surtees' Ferrari at Monaco in 1966.
He told Ferrari the 246 was faster than the 3-litre, but Dragoni made him drive it and BJ said he would have to pedal it so hard it would break - which it did, apparently the LSD. Maybe putting the power on earlier out of the hairpins just wore it out too quickly.

At Spa and Brands that year, the works Cooper-Maseratis had dreadful handling problems as the track got more oily (which it seemed to do do at all the races that year - Brabham said he'd never known as oily a track as Zandvoort, which is why Clark could pass him and pull away........but maybe it was a self-inflicted wound for Coopers, with all those leaky Maserati engines that year!). Roy Salvadori said in his book that they thought it was the LSDs at first and gave ZF an earfull, but then found it was a tyre problem and switched makers.

Paul M

[Wolf]

Well, as for ZF diffs my 'Bible' doesn't mention any GP cars (there are few pics of details of A-U cars and fragmental fuel mixture data, which I have posted earlier), but shows 3 kinds of locking diffs (Prometheus, ZF and Rheinmetall). As it's relatively close to being contemporary to A-U and M-B cars, I'll assume that they were used in those cars and could confirm they ZF is the same type shown in pics posted here.

I don't have the scanner, but I could try to have the pics and acompanying text (explanations, summary) scanned if it's of interest. Due to the age, the book is quite fragile (literally) so I'm not too keen to part with it, but will do so if needed.

[Frank S]

.
Here's a mysterious piece of equipment in the Aston Martin pit at
Sebring, 1958. The Astons both failed, relatively early. What do
you suppose the man with a leg over the counter is holding to
his face? Is he the boss, shouting directions through a megaphone,
observing personnel limits? Is he preparing to drown his sorrows
in a large bottle of ... ?

What?



Medium crop and full view:

.

--
Frank s

[bradbury west]

With a wide hair parting like that, it looks like Reg Parnell at first glance.

RL

[Sandeep Banerjee]

Yep, that's gotta be Reg Parnell cause I saw him in a Motorsport Quarterly video titled 'Astons in Action' where he was shown barking orders to the crew-members during pitstops at the Le Mans 24 hours that year with that mic.

Sandeep

[David Beard]

Prompted by that excellent poster Mr Bonde, on another thread, I thought I would give this one another bump.

This photo is from the marvellous Motor Sport Racing Car Review, by DSJ, 1957. (I will remove it if anyone is unhappy on the copyright front)
It shows the rear suspension of the eight cylinder Gordini. Now… I understand trailing arm suspension, and I understand how a Watts linkage usually works to give transverse axle location. But I don’t understand what Jenks is on about here when he refers to the “trailing lower arm of the Watts linkage”
Any ideas?

[Tim Murray]

From The Racing Car by Clutton, Posthumus & Jenkinson:

Suspension of this new [Gordini] was unusual and decidedly novel, being independent to all four wheels by an arrangement common to front and rear. Each wheel was carried on a vertical post that was pivoted on two tubular L-shaped arms, one trailing and the other leading. These arms were pivoted on the chassis and provided a vertical motion to the wheel hub. Torsion bars parallel to the frame were coupled to the pivoting arms by links.

So, seen from the side, the leading and trailing arms could be described as a Watts linkage arrangement. The leading arm certainly isn't very clear in the photo, however.

[Bonde]

Thanks for the kind words, David - though I havn't had the time to make many posts lately (a current racing car project taking all my time and then some), I'm not too sure any of my posts really warrant the term "excellent"!

Back on topic: As far as I can tell from the description and picture, the Gordini suspension is essentially like a double trailing arm suspension (Porsche, BRM V-16 etc.), except one arm is trailing, the other leading. Seen in plan view, this would to an extent locate the upright as a simply supported beam (supported at each end), as opposed to the 'normal' double trailing (or leading) arm suspension, which is fully cantilevered, making for a stiffer and/or lighter installation. It would appear to me that the Gordini arrangement could only be used at the rear, because the hubcarrier will rotate (as does the interlinking crank in a Watts linkage), creating massive changes in caster angle, as the hubcarrier moves up and down. Although not having sketched it out, I assume off hand that the roll centre will remain at ground level since the camber change will still be equal to roll angle.

[David Beard]

Thanks Tim and Anders

I think I see now. The Watts linkage isn't providing the usual transverse location, but longitudinal.

[philippe charuest]

a watt linkage provide the transverse and the longitudinal location and the vertical movement is perfectly vertical or perpendicular to the whole sheebang ,in a watt linkage theres no variation at all of the caster angles. in the case of a front suspension with watt linkage the hub is on a bearing and need a third arm (or steering rod ), that was use on many indycar roadster in the 50s

[Tim Murray]

Philippe, I accept that the hub moves up and down entirely vertically. But surely, to achieve this the upright has to rotate around the hub, hence (as Anders said) drastically altering the castor angle.

[philippe charuest]

Originally posted by Tim Murray
Philippe, I accept that the hub moves up and down entirely vertically. But surely, to achieve this the upright has to rotate around the hub, hence (as Anders said) drastically altering the castor angle.

the question is ,does it still call a "castor " if it doenst induce any movement in any plans except vertical .of course no additional trailing arm/rod especially the steering one can be fix to the rotating hub

[Bonde]

Philippe,

I think you may be referríng to a Watts linkage for lateral location, which, as you correctly state, can be arranged such that it doesn't induce caster change, and thus can be used at the front as well, caster not being significant to most rear suspension arrangements.

Trailing/leading arm suspension arrangements will induce some caster change, even if parallel as on a VW, when the car pitch angle changes, but apparently not enough to worry millions of VW Type 1s - and thousands of Formula Vees. The Citroën 2CV and its derivatives experience significant caster change - equal to the angle of the arc of the single leading arm - but that never seemed to bother anyone at the speeds (or lack of!) the "Deuce" ambles...even in 24 hr. races...

[Tim Murray]

In The Grand Prix Car 1954 to 1966 by Leonard Setright there is a cutaway drawing of the Gordini by the great Theo Page. This shows that the front suspension differed from the rear in having the trailing arm connected to the top end of the upright and the leading arm to its lower end. So, assuming these leading and trailing arms were approximately parallel to the ground in the static position, as the wheel rose in bump the top end of the upright would move forward and the bottom end would move rearward, hence reducing the designed castor angle. The poor drivers must surely have got some very funny messages coming through the steering, to put it mildly.

[philippe charuest]

http://www.brockeng....hanism/Watt.htm
Watt's Linkage. i thinq the misundersting come from the fact that we dont talk of the same thing .

[Tim Murray]

No, that's exactly what I understand by the term Watts linkage, and thank you for finding such a nice demonstration of its operation. You can see how much points B and C are moving relative to the central point P (which would be the hub on the Gordini) and hence the variation in steering castor angle which would occur as the suspension moved.

[Roger Clark]

In 1956 John Bolster drove the Gordini. He wrote:

"It is true that the employment of fore and aft arms for the suspension imposes a change of castor angle during considerable deflection. Under the extreme conditions of an ultimate travel from full rebound to full bump, the maximum castor is 8 degrees and the minimum is 4 degrees. In practice, no noticeably variation is apparent to the driver, and certainly the immense rigidity of the whole structure ensures that the desired limits are not exceeded. My only criticism concerned a tendency for the front wheels to bounce on very bumpy slow corners, and this will be cured by setting up the front dampers. The rear end stays glued to the road under all conditions and never "breaks away", furthermore, freedom from wheelspin is a great attribute."

[Tim Murray]

Very interesting - thanks Roger. I had been trying to do a rough estimation of the castor angle variation, so it's nice to have some genuine figures. I still feel that as 8 degrees is twice 4 degrees, the steering self-centring force at 8 degrees of castor will be twice that at 4 degrees, which I think would definitely worry me as a driver. Could it be that anyone who could cope with the evil-handling Bloody Mary in the way John did would perhaps have been relatively untroubled by cars with lesser handling defects?

[Wolf]

But why would anyone go for that kind of suspension to achieve such hub trajectory? Wouldn't parallel, equal length wishbones have served the purpose more adequately (track change presumably being of less consequence than castor change)?

[Bonde]

Wolf, that's actually a very good question! Perhaps Gordini just wanted to be...well...different - a trait that appears to be not uncommon in French automotive design history...

[Wolf]

Off top of my head, I can't recall any GP cars with Kamm tails- am I missing something or the principle isn't all that it's cracked up to be?

[Macca]

Well, there was the 1963 Scirocco-BRM, as well as Bruce McLaren' 1964-65 Tasman Cooper, the 1963 F1 Cooper and sundry USAC cars - but I don't expect any of them believed they would get much effect from a slim single-seater with the driver's head and rollover bar obstructing the airflow.

So that leaves the Lotus 49B at Monaco in 1968 as the first deliberate and effective version; but it was rendered obsolete at the next GP by Brabham and Ferrari.


Paul M

[David Beard]

How do we define a Kamm tail on a single seater...an abrubt cut off at the rear of an otherwise gently tapering shape? The Sharknose 156 Ferrari must be a good candidate, emulating the sportscar Ferraris of the same period with a narrow version of the same form?

[David McKinney]

The 1955 rear-engined F1 Cooper-Bristol would surely be the best example of a Kamm-tailed single-seater (though it did have full-width bodywork)

[Wolf]

Thanks. Well, I was thinking of Kamm tail in 'closed' form (like 'bobtail' Cooper)- I'd think those 'open' tails would primarily serve to draw out the hot air from gearbox/engine directly in front.

[David Beard]

Changing direction...

In the article in this month's Motor Sport magazine about the P153 BRM, Jackie Oliver is quoted as saying that the front wheel bearings were needle rollers "in the quest to reduce friction". Struck me as an odd thing to say: how is a needle roller bearing (whether caged or full compliment) going to be better in this respect than the normal taper roller or angular contact ball bearings you usually find in a hub assembly? To the best of my knowledge needle rollers are most often used where space is at a premium, and frequently as an upgrade on a plain bush. What was Oliver on about....or have I been missing something for years?

[Macca]

Originally posted by David Beard
How do we define a Kamm tail on a single seater...an abrubt cut off at the rear of an otherwise gently tapering shape? The Sharknose 156 Ferrari must be a good candidate, emulating the sportscar Ferraris of the same period with a narrow version of the same form?

Taking the Kamm tail as it was meant to be, achieving both clean seperation of the airflow and a degree of thrust from the eddys against the flat rear; then there were no single-seaters I can think of withe the correct form, as all the likely candidates (sharknose, 1963 FJ Cooper, BRM P261, etc.) have either rounded corners or air-extraction openings, or both.

And the Bobtail of the Cooper was attributed either to making the (rear-hinged) tail stand open without support, or just to be able to get it in the van that was originally carrying F3 500s.


Paul M

[Wolf]

Going back a bit to diffs, I have two or three points...

I think I saw someone elsewhere mention that Salisbury type diffs made it to F1 only in 1974 or thereabouts (with a book as a source, I think). So, I would think Hewlands used two types of LSD with their gearboxes, and one of them could be adjustable. There was a mention of Chapman saying he went with Hewland in '68 because their gearboxes offered two different locking 'coefficients'.

My last post on the subject, I mentioned two other types of 'locking diffs' that were used prior to '42- Rheinmetall seems to be working similarly to modern 'torsen' diff, but looks a bit simpler (I reckon still too cumbersome for racing applications), while Promethäus type looks like slightly more complex ZF.

As for ZF- I'm still mesmerized by its simplicity, and would like to use it in my car- does anyone still make them, and, in other case, where do I get an old ones? I'd like to put ~180BHP through it (if anyone knows a road car that used it)...

I've heard they seem to be quite agreeable when driven 'hard', but I'm a bit puzzled as they should have more lock with more torque being fed through them. That would mean the car should understeer in tight corners- the slower one goes the more it locks (in lower gears more torque is fed through LSD).