82 replies to this topic

[Dennis David]

I received this entry in my Guest Book recently. Any comments?

Please consider your proposed reason for
the split drive shafts on the P3. It was not for that reason, but what was it really for. I am open to any suggestion, but not that it was to sit the driver lower. As evidence of this, I ask you to recall that the preceding car, the Monza, had the same drive setup, and that the driver was offset - sitting over the driveshaft. There must have been a real reason - it was not likely that it was weight or unsprung weight, because drive shafts had to be heavier due to reduction gears in differential. Two bevel gears on the axle, and nearer to the ends of the axle, would not have helped, either. In the preceding Bimotore, the 1931/2 Type A car, this method of drive was used for the first time - apparently out of convenience - but something must have come to light that caused them to retain it. Please don't revert to driver seat height, it does not come into it at all.

------------------
Regards,

Dennis David
Yahoo = dennis_a_david

Life is racing, the rest is waiting

Grand Prix History
www.ddavid.com/formula1/

[Don Capps]

The P3. The P3, hmmm. The P3? That the same as the Tipo B?

I have the feeling - it will have to wait until later - that this just might be a Blue Whale sort of thing.

(The Blue Whale, the largest animal on earth, has a throat about the size of your fist? Why? Because God wanted it that way, that's why. So, whenever something is just that way because that is how it is, it is referred to as a Blue Whale...)

------------------
Yr fthfl & hmbl srvnt,

Don Capps

[Todd]

This is just a theory of mine, but could it have been a simple issue of torque wind-up? I've driven enough moderately powerful cars with open differentials and solid back axles to tell you that this can be a real issue. Suppose you're Jano. You've spent a few years building pretty powerful racing cars and wondering how to get them to hook up as strongly coming out of left hand turns as right hand turns. The twist of the differential is pressing on the left rear wheel under acceleration. You can't figure out how to equalize this without compensation springs which handicap other aspects of handling. When you are looking for a way to package a twin engined car, you stumble upon a solution to the disruptive torque effect of a single driveshaft on a solid back axle. Wouldn't you keep using the system in other cars until something better comes along? That is my theory. What do you think?

[Dennis David]

Interesting! Todd you need to spend some more time on this Forum. :-)

------------------
Regards,

Dennis David
Yahoo = dennis_a_david

Life is racing, the rest is waiting

Grand Prix History
www.ddavid.com/formula1/

[Don Capps]

Todd, Ditto on what DD said. That was very good thinking! Gee, very clever.

------------------
Yr fthfl & hmbl srvnt,

Don Capps

[Todd]

Thanks guys. I appreciate the invitation.

Dennis David,

I E-mailed in quiz #6 some time ago. When are you posting results? Great web page!

[Don Capps]

DD, Well try this:

"Producing a single-seat chassis created new problems. With a narrowed body, the driver had to sit on top of the transmission, which made the car much higher ans bulkier than it otherwise need to have been. So Jano produced a brilliant idea, derived from his twin-engine Tipo A - a divided drive arrangement, which shifted the differential from its usual place on the rear axle, and located it next to the gearbox in the middle of the car. From this differential, two propellor shafts emerged in a thirty degree vee towards the rear wheels. Each shaft drove one of the wheels through a light and simple bevel gear. The beauty of this arrangement was that it reduced the unsprung {intalics in the orginal} weight of the rear axle and differential and improved the handling of the car out of all recognition as well as greater ease in changing final drive ratios. And by replacing the single, centre-line propellor shaft by the open vee, it allowed the driver's seat to be placed slightly lower in the body, in the dip between the two angled shafts." David Owen, Viva! Alfa Romeo

Owen also notes that the Tipo B was approximately 220 kg lighter than the Monza.

I do believe, DD, your friend was in error and that Todd pointed out an unanticipated benefit from the layout.


------------------
Yr fthfl & hmbl srvnt,

Don Capps

[Todd]

Don Capps,

That makes sense too. The resulting suspension configuration of the P3 must have been a sort of elongated de Dion set up. Therefore it, conceptually at least, should share the reduced unsprung weight advantages of a de Dion. Of course by having a single, fixed and lighter drive shaft, a conventional de Dion does a better job of reducing unsprung and sprung weight and I suppose that is why it outlived the twin driveshaft configuration by so long.

[Dennis David]

Just to throw a wrench in the works Pomeroy in The Grand Prix Car Vol 1 :-) mentions only the weight savings and not the fact that the driver sits lower.

You know what I think? If you know Jano his ego was every bit a match for Porsche's. I think that he thought look how clever I am and just liked the idea over the actual benefit. What’s interesting is that the gear shift could be placed on either side.


------------------
Regards,

Dennis David
Yahoo = dennis_a_david

Life is racing, the rest is waiting

Grand Prix History
www.ddavid.com/formula1/

[Don Capps]

The "truth" is probably all of the above to some extent!

Todd, spot on to what crossed my mind when I looked at the Owen book. And, per usual, DD brought up perhaps the pertinent point: Jano and his ego.

------------------
Yr fthfl & hmbl srvnt,

Don Capps

[Dr.DeDion]

Another possible explanation involves the cryptic statements that I've seen in a couple of articles about the P3 to the effect that they had "further developments" in mind for the powertrain of the car which were stymied by the financial straits that the company was in at the time.I've sort of assumed these developments involved taking advantage of the fact that the 8C engine was essentially two four cylinders placed end-to-end with the camshafts and superchargers driven from the middle.This would theoretically allow them to drive the transmission from the center of the engine as well, like Colombo did on the Bugatti 251 of the 1950s.The flaw in this explanation is that they would have had to design a new chassis and transmission for this new layout anyway so why include one element of the new design in the P3?

It seems more likely that the combination of the slightly lower seat, slightly less unsprung weight(though I have my doubts about that one) and worthwhile reduction in torque reaction that Todd mentioned provide the real reason.

Here is another P3 mystery about which someone may have information or theories.Although some accounts of the 1935 German GP state confidently that Nuvolari's car had the 3.8 liter engine from the upcoming 8C-35 tipo C, other accounts describe it as having the same 3.2 unit that the other 1935 P3's used.The 3.8 engine would have put Nuvolari's car within 100hp of the Mercedes and therefore partially explain how he beat them that day, but I have my doubts.

During its competition career the tipo B's engine was enlarged from 2.7 to 2.9 and finally over 3.2 liters.While the P3's engine was successively enlarged, the gearbox was not.When breakages became a problem one gear was removed so that the same gearbox could contain just three larger, more robust gears.But even this was not enough to eliminate breakages of the 3.2 version's transmission to an acceptable level.That day at the Nurburgring, a circuit that was hard on the gears, Nuvolari's was the only 3.2 Alfa which did not succumb to transmission failure.I doubt that even Nuvolari's uncanny gearshifting wizardry could have nursed a 3.8 P3 to the finish line in first place.On the other hand, it is also hard to believe that he drove that race with less than 300hp at his disposal.

[Don Capps]

Doc D,

First, those are some points about where the Tipo B might have been headed. I had overlooked the configuration of the block.

Second, I can't find anything to confirm that it wasn't anything but a 3.2-litre for the 1935 German GP. I know that the 3.8 made an appearance at Monza, but unaware of it being used prior to that.

Third, Perhaps it was just Nuvolari's day and not von Brauchitsch's. TGN was eating huge gobs of time out of MvB's lead at the end. I often wonder what it was like in the boardroom of the Continental tire folks the next day.

------------------
Yr fthfl & hmbl srvnt,

Don Capps

[Dr.DeDion]

I suppose they couldn't make Manfred take the blame for spinning his tires too much; he was the son of Field Marshall von Brauchitsch after all(Well...let's try nephew).But he never would have displaced Tazio from the lead on lap 12 if the Nurburgring's state-of-the-art high pressure refueling system hadn't developed a blockage just when the Ferrari crew needed it.(Hmmm..?)Von B. should have cruised to victory with that minute and a half head-start.

There is a funny description of the scene in front of the pits as they played the German national anthem before the start."Then the Nazi and German flags were run up the flag pole, and everyone stood and saluted--all except Chiron, who did a little dance!"
(That's from Motor Sport's original 1935 account of the German GP.)

[This message has been edited by Dr.DeDion (edited 12-10-1999).]

[This message has been edited by Dr.DeDion (edited 12-12-1999).]

[Don Capps]

Doc D,

Ah, Leapin' Lou, Master of Jumping the Start.

Somehow I missed the Chiron dance.

The mysterious malfunction was noted for the record I believe...

------------------
Yr fthfl & hmbl srvnt,

Don Capps

Semper Gumbi: If this was easy, we’d have the solution already…

[Dr.DeDion]

Am I repeating myself again?But it's been months since I posted about that pit stop!Oh well, sorry.

[Ray Bell]

My initial enquiry was to try and establish what theories existed about the reason for this development on the Monza and P3 (or more correctly Type 3). I excluded seat height at the beginning for very sound reasons, and you have only to look at photographs to see that seat height was a non-issue.
The torque reaction theory is definitely on the right track, but there may be something more. I have reasoned that the only component that was lighter was the axle shafts, but that everything else was heavier - drive shafts and torque tubes duplicated, drive shafts heavier to cope with multiplied torque (same would apply to axle-mounted bevel gears), so it was not unsprung weight.
As for the 3.8 engine, I read in Grand Prix Racing Facts & Figures (1894-1963) that Seaman co-drove a 3.8-engined P3 to victory in a Donington race. Admittedly this was later than the 1935 Nurburgring event, but it shows that the use of such an engine was not unusual late in this car's development.
Does anyone know where the example that lived in Australia from 1938 to the 1960s and was then sold to Hon Patrick Lindsay is now, and does it have an engine again?
For the benefit of those not knowing its Australian history, the engine was sent to Italy in mid-1939 for a rebuild and was never to return (some talk of torpedoes, I don't really believe that...), so in 1952 (after 12 years of being known as the 'Nomotore Alfa' it was fitted with a 4.3-litre Alvis 6, then in 1957 with a GMC 6, then a Chev V8, racing in this form until 1961.

[Don Capps]

At this point Don bows out to Ray - by the way, Welcome to the Forum! - and Todd and the rest of you would actually understand all that sort of thing. I am good up to a point, and I think we passed it....

------------------
Yr fthfl & hmbl srvnt,

Don Capps

Semper Gumbi: If this was easy, we’d have the solution already…

[Ray Bell]

I'm resurrecting this old topic in the hope that one of the several new members (Karl, Keir, John?) might be able to shed light on the question.
I have established, by the way, that both driveshafts (and thus pinions) turned the same direction.

[ZippyD]

Ray,
I think you may be missing the point on upsprung weight. The slit axles were probably designed for one the same reasons we use multiple valves: It's easier to stop and start two 16 ounce valves that one 32 ounce valve.
It may not be upsrung weight that is the issue but reciprocating mass. Perhaps as the engine became more powerful the larger and therefore heavier drive shaft needed for the increased torque was a detriment to performance. So why not divide the drive shaft into two lighter units that would respond to throttle inputs more quickly? Just a thought.

------------------
"Pete, Do you sometimes get tired? Lately I have been getting tired. Very tired."

[Leif Snellman]

Cyril Posthumus gives the following explanations in his book "Classical Racing Cars":
- Reduced unsprung weight.
- Better power application
- Convenient access for changing final drive ratios without disturbing the rear axle.

Doug Nye in " Famous Racing Cars" also mentions the unsprung weight and confirms that the driver seat was as high on the "Monoposto" as on a car with normal prop-shaft.

[Ray Bell]

It's clear to me that the two final drive assemblies on the rear axle - although individually lighter than one in the middle, were not only heavier in total, but closer to the wheel and therefore of greater import when considering damping axle movement. This is why, when you ally this also to the weight of the individual torque tubes and drive shafts, I cannot accept unsprung weight as being an issue.
Just how better traction resulted is not so clear. Had the drives been contra-rotating, then we would have had a cancelling of forces and allowed the axle to do its thing.
It's possible that the quick changing of ratios was an issue, though I'm not so sure that the inventiveness that went into this design could not have come up with a European version of the Halibrand.
And just how quickly could one change the ratio up there ahead of the torque tubes and in the bowels of the chassis and controls?
I will consult Ray Wamsley (the gent who fitted the Chevy to one) about this.
Certainly, they had some traction advantage. It would not be worth the engineering without that, and they had the opportunity to abandon it at any time after the Monza had it not worked.
Keep at it, boys, we've got to work it out one day!

[Ray Bell]

Here's a thought:
What if the torque tubes, acting as they did at the outer ends of the axle, better placed the weight they transmitted nearer the wheels?
Would that work out? Who's the Einstein?

[Art]

Split drive shafts.

Presuming that both drive shafts are on the same side of the Pinions? The drive shafts would have to turn in opposite directions or we would have opposite rotation of the rear wheels? With opposite rotation of the drive shafts we would have equal torque load on both rear wheels. One drive shaft would have to drive off ov the front of one pinion and the rear of the other pinion for the same drive shaft rotation to work. I bet no one understands this mess. Make a drawing on paper it helps.

[Keir]

Hey guys,
This really isn't my field, but while looking through some old tomes, I found a possible link.
The double setup might have been a test bed for the Alfa bi-motor which ran the following season. After 1935 Scuderia Ferrari was doing all the official racing for the factory and I'm sure that Enzo had some influence on the design.
Other than that, I really can't see much of an advantage to the design. Ride height?
Maybe, but it would be minimal.


------------------
"I Was Born Ready"

[Keir]

Or maybe this,
Getting back to the Bi-motore.
The twin shafts must have limited axle hop, with those two engines working at either end of the car. So, with no axle hop, you get less or no wheel spin. But a well designed and calibrated system would work just as well. This could have been just a parts availability situation. Alfa had something that worked, why throw it away?
Question, Questions??

[Ray Bell]

The original bi-motore came into the equation, certainly. This rear end was used on that car for the first time. Then it was translated to a single gearbox setup on the Monza, then the P3.
I have a diagram (someone reading this kindly sent it to me) and it shows the driveshafts both going to the same side of the ring gear, so they do turn the same direction. Pinions are at the front on both.
I think this was dropped for the later bi-motore, but I'm not sure.
Ride height, I stress, is a non-issue. In the Monza the driver sat on top of it all, while the P3 was by no means low. Keep thinking...

[Art]

Ray Bell

If we have the ring gear teath pointing inward on both sides. And the pinion at the front of each ring gear and both drive shafts turning the same direction. One pinion will put a downward pressure on the ring gear teath and the other will put an upward pressure on the other ring gear teath. So we just set there wondering why the car wont move.

[Art]

Better yet.

Here we sit on the starting line 800 hp at 17,000 rpm we pop the clutch. 3 drivers 12 pit crew members and 150 fans are injured from shrapnel from every thing that let go. And being that you have the gear box between your legs. Well you take it from there.

[karlcars]

Griff Borgeson interviewed Jano so I checked on his comments about the Tipo B/P3, but he seems to have been more interested in engines than chassis. He bought the lower-seat argument, which I agree is a minor consideration at best.
Angelo Tito Anselmi and Valerio Moretti have produced a book on Jano's Alfas which is as close as we are likely to get to the last word on this topic. They say that the aim of the design was two-fold:
1. To reduce the unsprung weight of the rear axle. In this the design was definitely successful. Shifting the differential forward to the axle pivot point made it virtually sprung weight. The two axle shafts could be quite light, because they rotated at gearbox-output speed; the gear reduction was at the hubs. Completely eliminated were the half-shafts and the heavy central housing of the conventional differential.
2. To move weight forward in the chassis. Jano saw this as a means to reduce the typically heavy oversteer of the cars of this era. He shifted the mass of the differential gears well forward, significantly lightening the rear of the car. Thereby he created a more balanced and better-handling chassis with the tires available at that time.
I guess we have to give Jano some credit as a chassis engineer as well as a fine engine designer.
Interesting question!

------------------
Karl Ludvigsen

[Ray Bell]

Thanks, Karl, I'll mull over all of that for a while. I must talk to Ray Wamsley about the torque tubes etc and establish how big they were.
Certainly, balance by moving the differential forward could have been a factor, but again, in the car it was first built for there was plenty of weight in the front. And Wamsley has already told me the car handled better with the GMC, which was lighter than the Alvis engine it had in the early and mid fifties.
Guess I'm still puzzling!

[Art]

Dennis David.

Could you put an enlarged cut away picture of the P3 from your GP History on the board? I have some questions and don't want to put my foot in my mouth with out a better look. A bare chassis picture would be better if you have one.

Art

[Dennis David]

I'll see what I can do but it may take me a couple of days. Also Karl the book on Jano, is that Italian text only? If you ever have time I would love it if you could post your favorite books that you use in you're research or just enjoyment.

------------------
Regards,

Dennis David
Yahoo = dennis_a_david

Life is racing, the rest is waiting

Grand Prix History
www.ddavid.com/formula1/

[karlcars]

Yes, the Jano book is only in Italian at present.

I like the 'favourite book' idea. When I get a chance I will start a thread with this theme. Or someone else may like to!

------------------
Karl Ludvigsen


[This message has been edited by karlcars (edited 02-20-2000).]

[Ray Bell]

Seeing as Patrick has accepted the challenge, I have resurrected this thread yet again...

...and here are some comments from a parallel (but later) thread:

I dragged out a fresh horse since the old one was already glue and being used somewhere in a classroom:


quote:
--------------------------------------------------------------------------------
The most unusual feature of the car was the transmission, which consisted of a single differential behind the gearbox from which came two propellor shafts inside torque tubes, splayed out to form a V, and each leading to a bevel gear enclosed in an alloy housing situated under the chassis frame and just inboard of each rear wheel. This made for a very light back axle with low unsprung weight to assist road-holding. Peter Hull, Alfa Romeo, 1971
--------------------------------------------------------------------------------


...and this as well....


quote:
--------------------------------------------------------------------------------
It is at this point that the design became wholly unconventional. A differential was attached to th the rear of the gearbox where it gave the advantage of being part of the the sprung weight of the car. Two shafts diverged from the differential, each driving a crown wheel and pinion in an alloy casing adjacent to each back wheel; an empty axle tube joined and supported the two casings. Various theories have been advanced for Jano's reasons in embracing such a complicated and unusual transmission. It has been suggested that it enabled the driver to sit lower in the car, but a study of the car does not support this. It has also been suggested that the system was lighter than the conventional axle, but any weight saving would have been marginal. Another theory is that it made it easier for the axle ratio to be changed, but this is debatable and a small advantage for such a complication. Apart from the transmission, the chassis was entirely conventional. The rear frame was upswept and there were semi-elliptic springs all round. David Venables, First Among Champions, 2000
--------------------------------------------------------------------------------


__________________
Yr fthfl & hmbl srvnt,

Don Capps

The truth is rarely pure and never simple. -- Oscar Wilde
When I get new information, I adjust my position accordingly. What, sir, do you do with new information? -- John Maynard Keynes


IP: Logged


Dennis David
Member
10-06-2000 05:19
--------------------------------------------------------------------------------
My opinion is it's a case of Jano saying "See how clever I am" without any real benefit.
__________________
Regards,

Dennis David
Grand Prix History

Life is racing, the rest is waiting





Ray Bell
Member
10-06-2000 06:47
--------------------------------------------------------------------------------
Funny thing, I saw a book this morning that mentioned that they got to the first outing with the car and found it had a very 'whippy' chassis... stiffening was added overnight.
That severe upsweep at the rear must have contributed to this, and it was needed solely to give clearance to those cwp housings at each end of the axle.
Then, almost incredibly, in 1934 they fitted quarter eliptic springs... mounted at the REAR of the chassis and running forward to the axle!
I still put my case thus:
Saving in unsprung weight... not possible. Two cwps, two torque tubes, two tail shafts... and these closer to the ends of the axle, where their presence was least needed.
Driver height... absolute balderdash, as above.
Ease of ratio changes.... not really, they had to dismantle the rear end to get at the back of the gearbox to do it, from memory.
The most likely theory is that, after the twin contra-rotating whole deal of the twin-engined predecessor, Jano found traction levels were great. He has then mis-identified the reason as being the twin drive shafts, while all along it was a cancelling out of torque reactions by the contra-rotating drive shafts of his super-complicated design.
There is also a possibility that the proximity of the torque tubes to the ends of the axle that they applied their load better than did a single one in the middle. For those not understanding this, a torque tube takes the torque reaction of the cwp (the diff, remember, is in the back of the gearbox here) and uses it to try and lift up the car. Which in turn loads the axle and gives better grip.
Anyone else?
Hans wasn't here when we had this going....
__________________
The stories of Youth, of the burden of Time, and the death of Devotion...




John Cross
Member
10-08-2000 08:16
--------------------------------------------------------------------------------
I think you are under-estimating the size and weight of a conventional live axle. Pomeroy was certainly impressed by Jano's design - he says "the unique transmission system undoubtedly contributed materially towards weight reduction , as well as having certain other advantages".

Let's remember that we are comparing the Tipo B system with a live axle + torque tube, as used on Jano's Alfa P2 and Monza. This is a rigid assembly that is attached to the car at the front of the torque tube via a spherical bearing, inside of which is a universal joint. The longitudinal tractive forces (acceleration and braking) are transmitted from the rear wheels to the chassis at this joint. We all know how weak a T-shaped axle + tube structure is compared to a triangulated structure - the axle is acting as a beam when transmitting the forces to the diff which then feeds them to the tube.

With the triangulated design used on the Tipo B, the heavy centre of the axle has been replaced by a light steel tube and the two light alloy bevel housings have no diff - each pair of bevels only has to transmit half the torque. The 'half shafts' were just short stubs and this saved further weight. Also the diff unit was placed before the final drive reduction, so it was transmitting quite a lot less torque and was much lighter than normal. Plus it was sprung weight, of course.

15 years later some research was done that showed that a conventional live axle has a tendency to 'hop' when going over a bump, with the wheel centre traversing an ellipse. If the weight of the central diff is relocated nearer the wheels (as on the Tipo B), this tendency was eliminated.

So you have to think of the complete structural assembly compared to an axle + tube - both statically and dynamically it is superior.

On the separate points:

1) The driver is not seated lower, although the oil and petrol tanks probably were.

2) Saving in unsprung weight - YES.

3) Easier gear ratio changing - I don't know. Seems plausible, but Ray may be right.

Regarding the mounting of the rear springs - these gave an exceptionally high roll centre with improved stability (according to Pomeroy, anyway). Perhaps Jano knew what he was doing after all ...

It would have been interesting to see how widely the Tipo B axle would have been adopted had independent suspension (and then de Dion) not been so close.
__________________
I'll be back...





Ray Bell
Member
10-08-2000 22:47
--------------------------------------------------------------------------------
Please, let's eliminate the differential itself from this discussion... it weighs so little it's pointless discussing.
I don't really think the crown wheel and pinion assemblies could afford to be much lighter... let's forget the 'half the torque' bit, at times they got the lot. Also, shorter axles would need to be larger (maybe) because they don't have as much compliance and therefore have to take the power by virtue of larger size rather than allowing wind-up.
I'm well aware of what torque tubes are and do, and the same applies to them.
What you have contributed here is the bit about the eliptical wheel centre travel, this might be a bit of a clue.
Personally, I believe he was right onto something with the Tipo A and its contra-rotating shafts and self-cancelling torque reactions. But did he know that was happening?
The tube in the middle wouldn't have been significantly lighter, and the 'light alloy housings' were still duplicated.
__________________

So let's see what Patrick can add...

[oldtimer]

I've really been enjoying this thread. On the Auto-Union thread I posed a query about the effectiveness of the de-Dion rear suspensions of the 1938 Auto-Unions and the Mercedes. No response, but now I KNOW there are people who can discuss these things, so when you're finished with the Alfa, chaps, can we pick up the other thread?

BTW, going back a long way in the thread, the first P3/Type B had the gear lever between the drivers legs.

[Ray Bell]

Glad to, oldtimer, where are the drawings?

[dbw]

i'll jump in here....at this point, i haven't had my hands on a p3 rear end yet[but i have been near one]..i can speak to the construction of a gp bugatti rear end however....while not quite state or the art by the late '30s,it represents a fine piece of engineering well executed.it has an alloy diff case with elegant half axle trumpets in high strength steel...a typical 3/4 floater with tapered and keyed hubs... a open drive shaft with a pressed steel torque arm....bugatti even added the brake drums to the alloy paddle wheels to keep the duplication of components...however, as an assembly,this thing is one heavy mother lump!!!!i can imagine the removal of differential[planet gears,axle bevels,crufiform,carrier,etc] combined with shorter half shafts and typical italian application of as much alloy as humanly possible could cut significant weight tho still visually appearing heavier...alfa was also always concerned with axle placement and torque wind-up..the triangulated unit seems a superb way to locate an axle still attached with leaf springs...still conjecture on my part..sooooo...i do, however, have some friends that have actually had one of these things apart in their shop...i'll check with them and get their considered opinion....

[Ray Bell]

I reiterate... two crownwheels, two pinions, two torque tubes, two tail shafts, double the number of bearings... it all must weigh something. So the axle shafts are shorter... but maybe they had to be thicker because there wasn't enough length to allow compliance...

Certainly, it's all tied up in torque reactions, but I can't see exactly where. And in the Tipo A, Jano actually had the whole thing fighting itself in corners and over bumps... check it out!

Contra-rotating drives would have been a saving grace.

[Patrick Italiano]

I've been surprised by this thread appearing so quickly after Ray Bell put the argument on the table, and I was quietly researching when I discovered already 30+ messages about this.

I'll just put on some related issues, certainly not resolve the case. I tried yesterday night to go to some sources and to make links with former projects, i.e. mainly the Tipo A Monoposto.

First, what do people involved say? Answer: quite nothing.

The book 'Le Alfa Romeo di Vittorio Jano' quotes the integral version of Borgeson's interview to Jano in 1964. When it comes to the P3, Jano, seemingly bored or tired, just says 'nice car', nothing more, while Borgeson would have wanted more comments. Dead end.

The other witness is Fusi. He really says little about that. In the same book, in a chapter titled 'Vittorio Jano, my master', he mentions the transmission peculiarities, not the reason. I stress this point, because right aside to his text, there's a drawing whose caption does mention "lowering the driver's seat", but Fusi's text not. The captions are from the book's editor, in this case Angelo Tito Anselmi.

In 'Tutte le vetture dal 1910', in the Tipo B chapter, there's also no reference to lowering the seat as a reason.

All other comments are from serious people (including Peter Hull in his GP Car Profile about the P3), but not from those who actually designed the car.

But Fusi also wrote a detailled book on the Tipo A, which might help.

The Monoposto Tipo A had, as Ray already stressed it, two parallel 6C engines, complete with gear
box, each driving one wheel. So was the first version of that car, but Jano seemed concerned with one engine revving slower than the other in curve, although the drivers didn't complain. Somehow, his 'mechanical sense' was upset by this, and he first tried to fit a differential in each crownwheel connecting the two halfshafts. Strange idea, but it didn't work. It worsened the situation, because Fusi reports that it 'interfered with the engine speeds'. Jano then found his solution fitting freewheels to each transmisssion shaft, allowing thus both engines keeping the same speed in corners.

That amazes me actually, because I don't understand why he accepted to loose all engine-braking possibility to cure a problem the drivers didn't complain about. Plus the fact that, if you apply the throttle with such a device I still don't understand why the hell they would both rev at the same speed.

But astonishly the car is reported to have had a good handling (despite its weight - 950 kg), and I suspect that Jano himself didn't know why exactly. Fusi himself reports that, while the car actually existed, Jano pointed out a theoric flaw of the twin-shafts setup.

He probably transferred the axle layout to the single engined car for only half-explained reasons, on which we are still arguing 70 years later. Actually, Fusi's report stresses the Tipo B as being the follower of the Tipo A (as the name indicate).

That's what I conclude from the origins of the design.

Now, it did work well, and we have to know why.

I think that, although not supported by Fusi himself, the seat position isn't such a stupid issue some people have written here and in many other reports. The driver doesn't actually sit between the shafts, nor did he in the twin engined car, in the same way. But, looking close at cutaway drawings, it seems reasonable to think that he would have sat even higher with a conventional axle, since his position should then have cleared enough room for the shaft's upward motion. In other words, statically he sits over the triangle, but he doesn't prevent it to move upwards over bumps. This might be seen as marginal, OK.

The second point is about weight. Overall weight is higher for sure, but unsprung weight isn't, and more convicingly is it better centered. That claim can be found at the end of the Tipo A book by Fusi. Indeed is the differential almost sprung in its position, close to the central spherical bearing. When I compare the technical drawings of both the Tipo B rear axle (diff and crown wheels) to the drawing of the Monza axle (conventional), it's obvious that the Tipo B construction is way lighter than the Monza's, probably compensating for the twin tube / twin crownwheel arrangement. The shafts in the V are much smaller and are hollow, they transmit only half the torque each. The casing of the two crownwheels are also much smaller and lighter built than the unique unit of the Monza. There's a reason : the triangular design is much less stressed than a conventional one. Don't forget that the car is actually 'pushed' by the axle front spherical bearing. The two crownwheels are also smaller than those of a conventional car. Actually 31 teeth on the Tipo B comparing to 50 on the later Tipo C. I think, but I never had the parts actually in my hands, (I conclude from an accurate study of the technical drawings of all those parts), that tha Jano design may be not lighter overall than a coventional, but its rear end probably is, which means less unsprung weight.

The third point is the ease of changing the final drive. I've seen pictures of Tipo Bs during work at the gearbox, and I don't exacty realize the difference between changing the final drive on both designs. But indeed, it seems rather easy on the Tipo B, since the axle is easily dropped, and then opening its front end seems also easy.

I had read the torque reaction already somewhere before some of you quoted it here. Sorry, Ray, it makes no sense to me. Both axle turn indeed in the same direction, and the lift of one side of the axle has to be identical with both a triangular and a conventional axle. In both cases, the axle is retained to the body in three points, disposed identically : one in front and one at each spring leaf. Thus, torque reaction is just identical. But, as I said above, structurally, the triangle can cope with the same momentum while being lighter.

I'm too sorry I have no scanner to show you the material. I can try to have something scanned tomorrow by a friend, if I manage to. However, the only picture I've found on the net of the design is on Dennis David's page, but about the Bimotore.

http://www.ddavid.co...a_bimotore2.jpg

Just ignore the second engine behind the gearbox, and you see how the P3 was done.

[oldtimer]

Originally posted by Ray Bell
Glad to, oldtimer, where are the drawings?

And what would I be doing with pre-WW2 Mercedes and Auto-Union drawings?

[dbw]

just a thought....how much time lapsed between the p3 and the swing axle rear of the 2900?...was this a design that never got a chance to evolve as beam axle cars were on the way out?...was it a potentially great notion snuffed out by change?.....comments...?

[Patrick Italiano]

Originally posted by dbw
just a thought....how much time lapsed between the p3 and the swing axle rear of the 2900?...was this a design that never got a chance to evolve as beam axle cars were on the way out?...was it a potentially great notion snuffed out by change?.....comments...?

Here are the dates of the Alfa GP cars designs of the 30s.

8C2300 (to become 'Monza' after winning the Monza 10 Hours GP)

Design started late 1929 for a sports car, initially not seen as saleable, eventually produced at 188 ex.. The first chassis weren't numbered. Initially called just 8C2300 'Corsa': light version of the spider sports car (2 seater), first appeared at the Mille Miglia in April 1931. First GP at Monza in May 1931. Raced afficially (through Scuderia Ferrari) until 1933.

Monoposto Tipo A : twin engined single seater with two 6C1750 engines from the sports cars, placed side to side, each one with its gear box. Designed late 1930, first entered early 1931, last race at Coppa Acerbo, august 1931.

Monoposto Tipo B (unofficially called P3). Designed end 1931, first raced in June 1932. Two series of cars: one in 1932, the second for the 1934 season. Difference in frame construction (that's, I guess, what has been reported as 'strenghtening', but, sorry, it was 2 years later). For the 1935 season, the 1934 cars were converted to quarter-elliptical rear suspension (february 1935), Dubonnet front axle (end april) and hydraulic brakes (needed for the front Dubonnet).

Bimotore : racer built at Modena early 1935 for fast races. First entered at Tripoli 1935. New frame, Tipo B-like, lenghtened. The second engine is in the centre of the transmission V. The axle tube is replaced with swing arms (the second engine is right there), thus it features an IRS.

See: http://www.ddavid.co...lfabimotore.htm

Monoposto Tipo C : designed also in 1935, first raced in september that year. Features independent suspensions all round. Front with trailing arms and a transverse lever acting on a coil spring enclosed in a bottle working as hydraulic shock. Rear suspension by swing axle and trabnsverse leaf spring. Gearbox suspended to the frame in unit with differential. The 8C2900 sports car is actually a 2 seater Tipo C.

Tipo 158 : designed in 1937 (some design already from 1935). The layout is similar to the Tipo C, while the frame cosntruction isn't. Similar front suspension, but with transverse leaf spring, rear with swing axle up to mid-1951, while De Dion tested in 1940 and fitted to some 159 late 1951.

The IRS of the Tipo C was undoubtly better than the Tipo B triangle for reducing unsprung weight. The Tipo B design is part of a car concept where chassis flex was normal, with live front and rear axles. Chassis had to become stiffer with independent suspensions (control of geometry) and power increase.

As I posted yesterday, I suspect that Jano designed the Tipo B rear end layout on base of the Tipo A experience, probably for 'bad' reasons. Otherwise, this thread would have been solved by the claims of the designer themselves...

[Ray Bell]

Thanks for your efforts, Patrick, but I still don't believe that the axle and torque tube total weight was lighter. Even lighter bits don't make up for the fact that there was now two of everything.
Stubborn... but this thread has been on the board since before I became a member, and I think it has to remain until I can get some realistic answers.
How I wish I could examine a car and weigh the bits!

And, again, I'm sure Jano was onto something when he had those contra-rotating shafts on the Tipo A, but that he possibly didn't know what he was onto. I'll have to analyse those freewheels again...

[Don Capps]

Another for DCN.

[David McKinney]

Originally posted by Ray Bell
Does anyone know where the example that lived in Australia from 1938 to the 1960s and was then sold to Hon Patrick Lindsay is now, and does it have an engine again?
F

I must have missed this first time round
The car in question, telaio 5002, spent the late 1980s in the US, the early 1990s in Japan and has since returned to the USA. I haven't updated my records on the car since 12 months ago, when it was being advertised in the trade in California. And yes, it has an engine, a proper 2.9 job, though what its origin is I'm not sure.
The other Australian P3, the car Lex Davison raced, was still in Japan in 1999

[Ray Bell]

This car has never been mentioned on the forum, I'm quite sure...



It's said to be a streamlined version of the Tipo B and to have won at Avus...

There are other cars worth looking at on the page...

http://www.classics.com/retro01.html

[dmj]

Ah, that fascinating Guy Moll car... I would also like to know more about its history.

[Vitesse2]

Alfa originally intended three cars would be constructed, but this was the only one built, as Jano was apparently not convinced of their worth. They were suggested by Guidotti, and designed by Ing. Pallavicino, aerodynamicist at the Caproni aircraft company, who had served with Guidotti's brother in the Great War. After a dispute over money, the details of which are unclear, the design was apparently provided for free, within ten days!

The engine was bored out to 3165cc, giving 265bhp, while to save weight (!!) the rear brakes were reduced in size! The car was tested unpainted on the Milan-Varese autostrada on May 24th 1934, driven by both Moll and Varzi, achieving 175mph/280kmh. Varzi was unhappy with the car's handling and was adamant he would drive a conventional P3 at AVUS - Chiron also had a similar car. The junior driver Moll was therefore left with the streamliner.

In the race, Stuck sped away in the wet, building up a 1 minute lead by the end of lap 1, chased by Chiron, Varzi and Moll in that order, but as the track dried Moll speeded up, passing Varzi on lap 4 and Chiron on lap 7. He then took the lead when Stuck's Auto Union stopped for new tyres - two laps later the clutch gave out on the German's car and Moll went on to complete the win.

After that, the car seems to have been discarded - I don't think it ever raced again.

[Ray Bell]

I've never heard of it at all... but that's nothing new.

And thanks for the detail, Richard. Nice to know it was a success.

One thing about it is the similarity in appearance with the rebodied Indianapolis Ballot that ran at Lobethal in 1940 with a Ford V8 engine. Anyone with the April (? May?) 1940 Motor Sport can turn up a pic of that car and see what I mean. Just the front of the car, by the way, not the tail.

[Vitesse2]

May 1940 Yes - I can see the similarity, although the cowl on Gullan's Ballot comes slightly lower than on the Alfa.

I think Moll was rather lucky to win, since Mercedes had withdrawn just before the start, zu Leiningen's Auto Union retired and Momberger wasn't really up to the job in the third Auto Union. Varzi was half a minute down at the finish, with Momberger behind him, followed by Howe's Maserati 8CM, Nuvolari with his leg in plaster and Pietsch in his old Alfa Monza.

That of course was the race where de Paolo's 4-wd Miller blew up so spectacularly ....