44 replies to this topic

[HistoryFan]

In Formula 1 we had a lot of interesting engines. The never ending story of the Cosworth DFV, the Pratt-&-Whitney turbine engine and so on.

 

But what are the most interesting pre-war GP engines and why?

[D-Type]

People were still finding their way then, there wasn't even concensus among the main manufacturers about the optimum number of cylinders and their layout.  Then there are the unusual ones you are asking about.

 

Here are a few for starters:

 

SEFAC - to quote David Hodges: "... a 2771cc parallel 8, comprising two blocks of four cylinders side by side, with camshafts rotating in opposite directions and linked by gears.  There were twin overhead camshafts, and another approach to desmodromic valves, which were positively closed".

 

Trossi Monaco - Hodges again: "The car had a Zoller-supercharged air-cooled 3982cc engine - a double piston two stroke, no less."  And Jenkinson describes it as an "8-cylinder, double piston, 2 stroke radial engine"  I'd love to know more.

 

Voisin - a straight-six engine with sleeve valves.

 

Maserati V4 - Hodges again: "a 16-cylinder engine that comprised two 26b straight eights on a common crankcase .  The right block had reversed porting, so that eight-into-one exhausts appeared on each side of the car.

 

Gobron-Brillié - four cylinder opposed-piston engine.

 

Fiat 806 - Fiat developed an opposed-piston two stroke engine but it was unsuccessful, sothey went on to produce the type 406 engine which Comarosti describes as  "... whose 1.5 litre 12 -cylinder engine was made up of two blocks of six cylindersset in a vee with two crankshafts.  The valves were4 driven by three overhead camshafts" and Hodges tells us that the central one operated the inlet valves.

 

BMW 328 - simply for its ingenious valve gear.

 

Then there was the car with a horizontal (ie vertical axis) rotary engine, but I forget the name

 

 I hope this gives you a starting point for your investigation.  Please let us know what more you manage to find out.

Edited by D-Type, 11 March 2014 - 19:22.

[bradbury west]

Duncan, you are a very genial chap, doing his work for him. Surely better for the originator to come up with a list of discoveries/options and ask for views. Not too hard, we all have a selection of books, or the search facility in this place........

In the spirit of goodwill I suggest the Alfa Tipo A, or as it became thus with the tipo B's arrival. I have never some across an article on the tipo A.

Also Guidobaldi, which we have covered at length variously here.  A post ww2 car, granted, but the engine was conceived pre war. Makes the Trossi look a good idea.....

As an aside, surely the level of "interesting-ness " is very much subjective.

Roger Lund

Edited by bradbury west, 11 March 2014 - 17:40.

[D-Type]

Just demonstrating what half an hour looking at a book can do. 

[Paul Parker]

New threads like this highlight the tragic loss of David McKinney even more, who would surely have had a candidate in mind.

[bschenker]

About betwen 1910 ~ 1915 Paul Henry Peugeot DHOC whit desmodromic valves engine. The rest is to find in the net or in books for the moment I have not time.

 

Henry has built also similiar motocicle engines.

 

-

[Allan Lupton]

Technically not quite all of Duncan's list were GP engines but that's more or less the list we'd offer. Many if not all have been discussed here in the past so we don't need to do more than link the old threads. e.g.:

http://forums.autosp...co#entry3491056

[D-Type]

About betwen 1910 ~ 1915 Paul Henry Peugeot DHOC with desmodromic valves engine. The rest is to find in the net or in books for the moment I have not time.

 

Henry has built also similiar motorcycle engines.

 

-

Was the question of whether the motorcycle engine possibly preceded the L76 ever resolved?

 

Incidentally, I don't think the valve operation was desmodromic, but I'm not an engine person.

[Roger Clark]

 

In the spirit of goodwill I suggest the Alfa Tipo A, or as it became thus with the tipo B's arrival. I have never some across an article on the tipo A.

 

Roger Lund

Luigi Fusi wrote a very good book on the Tipo A

Edited by Roger Clark, 11 March 2014 - 19:25.

[kayemod]

Duncan, you are a very genial chap, doing his work for him.

 

I think that History Fanboy's computer must be a very unusual one. From the numerous vague one-line thread openers we've seen, it obviously can't access Google, or even Wikipedia.

[Roger Clark]

 

 

Incidentally, I don't think the valve operation was desmodromic, but I'm not an engine person.

That is correct.  I don't know who first suggested that the Peugeot had desmodromic valves but he's been shown to be wrong.

[bschenker]

 

 

Mayby not a clean solution but it's desmodromic. I speack only from memory in this moment I dont remenber in what's book and the place who I have this book.

 

This was one off the solutions adapted by Peugeot, ther are others. Exemples the 4 valves for cylindre, also in this im not sure at all or one off the solution.

.

Edited by bschenker, 11 March 2014 - 20:40.

[bradbury west]

Thanks for the book info, Roger

RL

[D-Type]

Cimarosti (English 1997 edition) P42 includes a cutaway of the Peugeot valve gear.  It has the stirrup-shaped cam follower enclosing the cam.  But it also has conventional valve springs to close the valves similar to the right hand sketc above.  And it has the smaller 'helper' springs to help opening.  So, definitely not desmodromic.

 

The L76 engine had hemispherical combustion chambers, overhead valves operated by two overhead camshafts.  This is why it is considered the forerunner of all subsequent racing engines.  Had the L76 used desmodromic valve actuation this would have been adopted by at least some, but maybe not all, subsequent racing engine designers and not left to Daimler-Benz to rediscover and reinvent for the W196.

[Siddley]

 

Trossi Monaco - Hodges again: "The car had a Zoller-supercharged air-cooled 3982cc engine - a double piston two stroke, no less."  And Jenkinson describes it as an "8-cylinder, double piston, 2 stroke radial engine"  I'd love to know more.

 

I'd guess that by radial, he means 'radially opposed' - with two crankshafts - one at the top and one at the bottom ( or left and right, depending on how it was mounted ) with the pistons meeting in the middle. Pretty much the same as the Rolls Royce ' multi fuel' engines used in British AFV's of the 60's and 70's
Requires a supercharger as there isn't any primary ( crankcase ) compression.
I'm pretty sure Junkers used the same layout for an aero engine, it's in one of my Bill Gunston books somewhere...if I could find it...

Edit, having just Googled 'Trossi Monaco' I find out I'm completely and utterly wrong. It really is a radial. What were they thinking ?

Edited by Siddley, 11 March 2014 - 22:47.

[D-Type]

If you look at this cutaway,  it appears to have eight, or in fact eight pairs of cylinders arranged radially on a central crank.  Being a two stroke I expect thepaired cylinders were connected somehow

[bschenker]

The left hand solution was used by Peugeot (it's must be patented) mayby only testet with some problems so I think the finaly version (no more desmodromic) from Peugot was the right side one, remember it's was 1912 material and production was an other thing as in the fiftys, also Ducati has first some similar solution, sure with one cam for open and one to close the valves.

 

A storysite with some of the desmotromic solutions

http://www.italian.s...e/engine05.html

 

Since i startet to work (mecanic in end 50) searching for the desmodromic story the Peugeot solution is always nominated.

 

.

Edited by bschenker, 12 March 2014 - 01:42.

[Allan Lupton]

If you look at this cutaway,  it appears to have eight, or in fact eight pairs of cylinders arranged radially on a central crank.  Being a two stroke I expect thepaired cylinders were connected somehow

Yes, as I wrote in the thread I linked in post 7 above, it was a duplex two-stroke. That involves pairs of cylinders with a combustion chamber in common with the two pistons controlling either intake ports or exhaust ports. They are almost, but not quite, in step so the exhaust opens and closes slightly earlier than the inlet.

The best-known example of this in the UK was the rather pedestrian Trojan, but the Z13 from Brno, Czechoslovakia, was rather racy getting 68 h.p. from 1500 cc (H-O Neubauer writing in Georgano)

Edited by Allan Lupton, 12 March 2014 - 08:47.

[Roger Clark]

People were still finding their way then, there wasn't even concensus among the main manufacturers about the optimum number of cylinders and their layout.  

Designers used different configurations be cause they could; the reasons for choice might be theory, tradition or economics.  However, the straight eight was the dominant engine for most of the inter-war period.

[Allan Lupton]

People were still finding their way then, there wasn't even concensus among the main manufacturers about the optimum number of cylinders and their layout.

There still wouldn't be consensus about the number of cylinders, and certainly not about their layout, were it not for the over-prescriptive anti-competitive rules of the modern entertainment business that has taken Grand Prix racing's place.

[Roger Clark]

There still wouldn't be consensus about the number of cylinders, and certainly not about their layout, were it not for the over-prescriptive anti-competitive rules of the modern entertainment business that has taken Grand Prix racing's place.

I'd like to think so but I fear they'd all copy the most successful - witness the number of 10-cylinder engines that appeared in the late-90s.

[Allan Lupton]

I'd like to think so but I fear they'd all copy the most successful - witness the number of 10-cylinder engines that appeared in the late-90s.

That'd only be those without the intellectual capacity to do something first.

That's before you get on to such things as the experimental re-invention of a long-defunct valve system against which rules were hurridly rewritten.

[scheivlak]

Edit, having just Googled 'Trossi Monaco' I find out I'm completely and utterly wrong. It really is a radial. What were they thinking ?

 

"Aerodynamics are for people who can't build good engines"?

[HistoryFan]

I think that History Fanboy's computer must be a very unusual one. From the numerous vague one-line thread openers we've seen, it obviously can't access Google, or even Wikipedia.

 

The problem is that I have just a little knowledge about the pre-war-GP-races. So I have no idea for what I have to search for. I never heard of the Trossi car, but with them know knowing, I can research about it. We have perhaps 100 or more pre-war-GP-engines, I cannot know which of these were very interesting. So my goal was to hear some interesting engines with one or two keywords and I can research on that. And perhaps such threads will also start interesting discussions. I don't see where is the problem, if you have another opinion, you don't have to read or answer my threads.

[D-Type]

The problem is that I have just a little knowledge about the pre-war-GP-races. So I have no idea for what I have to search for. I never heard of the Trossi car, but with them know knowing, I can research about it. We have perhaps 100 or more pre-war-GP-engines, I cannot know which of these were very interesting. So my goal was to hear some interesting engines with one or two keywords and I can research on that. And perhaps such threads will also start interesting discussions. I don't see where is the problem, if you have another opinion, you don't have to read or answer my threads.

I think what irritates people and elicits some of the tart responses you get is that your posts never say what information you have already found.  For example, when I  went through David Hodges A-Z of Grand Prix Cars and copied out some information for you I had absolutely no idea whether I was totally wasting my time reproducing information you already have or whether I was providing you with useful information.  And I still don't know - other than that [not surprisingly] you have never heard of the Trossi.

 

TNF is a community of enthusiasts who will help out other enthusiasts.  It is not a tap that you can simply turn on and receive a one way flow of knowledge and information that some have spent many years acquiring.

[HistoryFan]

Voisin - a straight-six engine with sleeve valves.

 

 

 

I thought that this propeller was used to power a water pump and was not part of the engine...
 

[D-Type]

I thought that this propeller was used to power a water pump and was not part of the engine...
 

Yes, another interesting feature in addition to the sleeve valves.  If you dig deeper you may find other interesting features about the car and itsengine.

[HistoryFan]

In 1909 Benz had built an engine with 21.5 litres! It was based on the GP engine which was driven to second place by Victor Hemery in the 1908 French Grand Prix. I don't know whether this 21.5 litres Benz was entered in any GP race but that must be the car with the largest number of litres. But what was the largest numbers of litres entered in a GP race? And how much litres had the Benz engine at the 1908 French Grand Prix?

[HistoryFan]

Okay I found that thread:

 

http://forums.autosp...gest-gp-engine/

[D-Type]

The left hand solution was used by Peugeot (it's must be patented) mayby only testet with some problems so I think the finaly version (no more desmodromic) from Peugot was the right side one, remember it's was 1912 material and production was an other thing as in the fiftys, also Ducati has first some similar solution, sure with one cam for open and one to close the valves.

 

A storysite with some of the desmotromic solutions

http://www.italian.s...e/engine05.html

 

Since i startet to work (mecanic in end 50) searching for the desmodromic story the Peugeot solution is always nominated.

 

.

 

My apologies!  The valve operation 0f the L76 was desmodromic. 

 

Cyril Posthumus says:

An unusual method of ensuring positive valve actuation was the use of bronze 'D' or stirrup-shaped tappets surrounding each cam, which both opened and closed the valves wit assistance from light springs - an early form of desmodromic operation which Peugeot dropped a year later.  Zuccarelli and Henry also exploited valve overlap between the opening of the inlets and the closing of the exhausts to improve breathing and combustion.

 

Does anybody know how the springs that helped open and close the valves worked?  How did the springs know when to act to help open or close a valve and and when they weren't to act as they would oppose the valve movement as it was now doing the opposite and closing or opening respectively?

 

H-F, I think the 21 litre 'Blitzen' Benz was intended primarily as a record breaker and not as a Grand Prix car.  Although it was raced by, amongst others, Barney Oldfield, it was not in the Grand Prix or the Grand Prize (as far as i know).

Edited by D-Type, 17 March 2014 - 23:55.

[bschenker]

I think it's was a problem about wat's in 1912 was posibily in material and productions tecnic. The springs needed to compensing the jocke between the parts. When you look on the first solution, you cant find the spring on the top side and you have a positiv and a negativ cam (hope I can say this name). I think in thats time the precision was not on the same level than leater by Mercedes. From this the change without the negative cam in change for the top spring. Then at least changing to a normal system with standart spring typ (in the same fusion), for a stable function.

 

I think also that's in 1912 the internal pression was not enough to close definitiv the valves, like in the F1 - Sportprototyp from Mercedes. Mayby for the lower turning engines, mayby for lower compresion in thats time or both of them (and other like fuel ecc).

 

(Sorry my english is not the best but I hope it's cant be unterstand)

 

.

Edited by bschenker, 14 March 2014 - 00:57.

[RCH]

Duncan, along with most commentators, refers to the BMW 328 engine as "ingenious", this has always puzzled me. Wouldn't the really ingenious answer have been a twin overhead cam design? Whilst we can in no way denigrate the engine I have often wondered why nobody tapped the designer on the shoulder and said "Fritz, have you seen what Alfa Romeo are doing?" 

 

Or am I missing the point completely?

[D-Type]

Rod, I won't dispute what you say.  I included the 328 in my list as I felt that HistoryFan might find it interesting and enjoy researching the details - particularly as he is German and apparently more at home in that language.  I believe a similar system was used by contemporary Talbots and [I think] 40 years later by the Triumph Dolomite so it must have had domething going for it.  Presumably it was simpler (ie cheaper) to manufacture than a full DOHC set up.

 

I also wondered whether HistoryFan would consider the Auto Union's multi-element Hirth crankshaft interesting, but decided to leave that one to somebody else to suggest. 

[Siddley]

Did face cams ever make it into multi cylinder engines ? Chater Lea used them in their ( 1920's ? ) motorcycle engines. I'd imagine them being a very poor choice for anything other than a single cylinder or perhaps a twin, but it is the era of crazy ideas

[fbarrett]

Maybe Fiedler (?) designed the 327/328 that way because he had to use the existing block (with cam) but was allowed to replace the head.

[AAGR]

Correct - that's absolutely what happened. Incidentally, Triumph never produced anything like the 327/328 unit, which had pushrods. In the 16-valve Dolomite Sprint (which wasn't a race car engine, though Broadspeed turned it into a very effective one), there were two lines of valves, one being operated directed from a single overhead camshaft, the other by rockers working off that same camshaft ....

[D-Type]

  Memory is fallible.  I know the Dolomite Sprint had an unusual and ingenious way of utilising a single OHC and assumed it was similar to the 328

[RCH]

I had assumed that cost must have come into it somewhere, it just seems to me that by eliminating pushrods and rockers there is a beautiful simplicity of design with overhead camshafts when it comes to achieving desirable features. Perhaps I spent too much of my youth playing with Hillman Imps.  Although setting the tappets is a downside...

 

Was there ever a pushrod block that was converted to take upstairs camshafts? I'm thinking MGA, Triumph "Sabrina"? Talking of Triumphs the Dolomite Sprint engine is another puzzle. Why not just go straight to 2 camshafts to drive the 16 valves?

 

Sorry I've taken us well off topic... again

[Michael Ferner]

Ford's Model T, A and B blocks were used to build DO engines en masse - go to Speedy Bill's Museum of American Speed site and feast your eyes on some ingenious solutions...

Edited by Michael Ferner, 16 March 2014 - 09:40.

[Michael Ferner]

Not going for a DO setup has pretty much always one source of inspiration... money! The most simple and thus least expensive way to build an engine is to place the camshaft(s) as close as possible to the crankshaft, and drive it/them by means of a simple cogwheel off the main shaft. Driving an "upstairs" shaft for the cams means more engineering, more parts to manufacture and assemble and to possibly fail. Add a second upstairs shaft, and you have twice as many "problems"...

Edited by Michael Ferner, 16 March 2014 - 09:51.

[Allan Lupton]

Maybe Fiedler (?) designed the 327/328 that way because he had to use the existing block (with cam) but was allowed to replace the head.

Quite so and there was also the Testadoro head on a Fiat 500 sidevalve engine - in both cases everything below the head gasket face could remain unaltered.

[RCH]

Yes... but... Well what could be simpler than a SOHC design like the Coventry Climax/Imp set up? You've simply got a camshaft in a different place driven by a chain rather than a cog; the camshaft working almost directly on the valve via a tappet and shim. No pushrods or rocker assemblies to worry about. The only downside I can see is adjusting the tappets (which Vauxhall came up with a neat solution for) and possible reliability issues with a chain. Nowhere near as complex as the multiplicity of pushrods and rockers needed for the BMW design.

[Michael Ferner]

In general, I sympathize with your views; some of the rocker arm designs I've come across have struck me as excessively complex solutions to a fairly basic problem. However, not being an engineer or even mechanic myself, I believe the potential (timing) problems multiply with the distance between crank- and camshafts. For instance, a faulty rocker will at worst ruin one cylinder, while a camshaft out of tune will likely mess up the entire engine. And even if chains or belts are a cheaper solution than a proper gear train, they add reliability issues to the whole timing process with, again, likely fatal results in the case of failure.

[AAGR]

Sorry, O/T again.

 

1)  The Triumph 'Sabrina' was a totally unique engine, with no relationship to the TR's 'wet-liner' production engine.

2)  The Dolomite Sprint engine didn't have two overhead camshafts for cost reasons. Advice on designing it like it was, incidentally, came from Harry Mundy and Walter Hassan (both of Jaguar at the time, and therefore linked with Triumph through British Leyland).

[RCH]

It would seem that with both BMW and Dolly Sprint designs not going the DOHC route was down to cost, yet I still find it surprising that the plethora of pushrods, rockers and various odds and ends was that much cheaper than a simple straightforward extra camshaft.