43 replies to this topic

[Richard Border]

I don't understand the desmodromic valve cult. They were a good solution to a problem, which as I understand it, was valve spring material was not up to the job. The valve springs broke, and valves floated, desmodromic valves cured that. BUT now days on road cars and bikes valve springs are up to the job, and in F1 pneumatic valve "springs" work fine. What's the advantage you see in desmodromic valves?

[desmo]

Desmodromic valve cult? I think that desmo valve actuation acquired a technical cachet largely because it was seen on some really cool machines such as Ducati motorcycles and Mercedes racers. It may be a developmental dead-end now, but it will always be a compelling solution when valve gear are defining the upper limit of operation of a cammed valve system. Any reports of it's demise may be exaggerated.

[Ray Bell]

desmo's reasoning on this in the pneumatic valves thread is sufficient to explain that. The added weight and complication of desmodromic gear makes the present arrangement look pretty good, and even if it is reaching the ragged edge of its application, engineers are more likely to look for systems without camshafts at all.

[Top Fuel F1]

I know what desmodromic is in relation to the type of valve actuation used by Ducati; but I've been wondering what the word desmodromic actually connotates. It's not in the dictionary, so it must be a strictly engine engineering term. Depending on what the original intended strict meaning was, one could then see if the terminology would apply to any other scheme, other than what is used in the Ducati. For instance some one said that the desmodromic version used by Mercedes in the 50's was different than Ducati's. If desmodromic simply means a valve that is positively driven open and closed without springs, then it could be conceivable that some kind of solinoid scheme could also be termed desmodromic as well. Pls. advise. [p][Edited by Top Fuel F1 on 12-08-2000]

[Alvega]

Desmodromic valve actuation, like mechanical watches, vinyl discs or valve amplifiers, isn't perhaps the cutting edge of technologie, but sure has a lot of cachet.

[Darren]

Yes, dromos is Latin, but it does refer specifically in its origins to the racecourse at Sparta, so it might not be the etymological leap it appears. Still, we might have to blame the French for gifting us the word.

[irineos]

dromos is a greek word.

[Top Fuel F1]

Re: "Dictionary of Difficult Words - D Content", on the Internet

1. Word closest to dromic is dromos which = " n. avenue or entrance-way to a building; passage to a tomb; ancient Greek race course ".

2. Word closest to desmo is desmoid = " a. like a ligament ". Ligament = Something that forms a connection or bond.
An internet search on "desmo" commonly brings up Ducati things; but also some other items. So it may be a word who's meaning is in the Italian dictionary.

Rgds;

[Top Fuel F1]

Originally posted by Top Fuel F1
Re: "Dictionary of Difficult Words - D Content", on the Internet

1. Word closest to dromic is dromos which = " n. avenue or entrance-way to a building; passage to a tomb; ancient Greek race course ".

2. Word closest to desmo is desmoid = " a. like a ligament ". Ligament = Something that forms a connection or bond.
An internet search on "desmo" commonly brings up Ducati things; but also some other items. So it may be a word who's meaning is in the Italian dictionary.

Rgds;

Re: Oxford English Dictionary

Keeping in mind that I am trying to find out if there is any clear cut way to see if "desmodromic" is only unique to the form factor found in the Ducati valve scheme/architecture.

Anyway in the Oxford Dictionary I find:

1. Desmo = Combining form of Gr. xxxxxx(Greek) bond, fastening, chain, ligature, an element in scientific words of Greek derivation.

2. Dromic = xxxxxxxx(Greek) course, race course. Of pertaining to, or of the form of a race-course; applied to the basilican type of Eastern Churches from it's resemblance to the plan of a race course.

What I gather from this and from other examples Oxford gives is that "desmo" is used to indicate that "desmodromic" means a design belonging to or resembling the architecture of the Eastern Churches who's appearance resembles a particular shape of a race-course design. So the only significance of "desmo" is to mean that something pertains to another group of things. Assuming this is so, one would now have to use their imagination to figure out how (say the Ducati design) relates to this particular race-course form.

In addition, in a Latin Dictionary I found:

1. Dromas = Able to run well.

2. Dromos = An open space with room to move freely, a parade ground, etc. Oh, and also pertaining to the race-course of the Spartans.

Best Regards; [p][Edited by Top Fuel F1 on 12-20-2000]

[Wolf]

I don't know much about ethymology , yet as I gather from previous posts it's fairly obvious what the person 'inventing' that word was aiming at. It's something in a sense of 'linked path' or 'linked movement/running' and describes feature by which it to be distinguished from other valve mechanisms which use 'aids' (as sprigs or pressurized air) to ensure valve tracking the camsaft profile.

[danut]

Desmodromic is not strictly an engineering term. In fact I was quite surprised to meet it in a technical context as I was familiar with it from my cellular biology courses years ago in the medical school. there is a type of junction between cells called desmodromic, and as someone already explained it means something connected that moves together

Dan

[Top Fuel F1]

Danut:

It's good to keep in mind that "desmo" is just an element in scientific words of Greek derivation. Without counting too carefully I can see over 25 words starting out with "desmo".
These are mostly involved with biochem., animals, medical, biology, etc. However, I have located the word "dromic" which obviously has Greek origins. The definition in this case (see my previous Post) seems to have something to do with the form of a particular type of race course that was also seen in Eastern church designs.

Here's one that would be easier to work out a relation to the Ducati valve mechanism: "desmopelmous" = (A type of bird foot in which the hindtoe cannot be bent independently because planter tendons are united). That is to say something about working together.

Rgds;[p][Edited by Top Fuel F1 on 12-21-2000]

[danut]

Top Fuel,
I was expecting that desmo is a prefix to be found in many geek originated scientific words. As someone suggested that it is a technical word I just wanted to show it is not strictly so...
That desmopelmous is a really nice word, maybe one could create other new words based of desmo, like desmodrunk, meaning two or more guys gettting drunk together...

Dan

[Top Fuel F1]

Danut:

Of all the desmo words I have seen, I haven't come across the one you suggest. Although the two guys may be in the same state they can still act independently. So maybe it would not qualify. May be if you made it desmodrunkcidae.

[Yelnats]

Desmodromic = Greek = "I run on a halter or leash." First used in this context by the French automotive developers of this technology early in this century (before 1914).

This tidbit I recalled from the excellent F1 Survey of the Front-Engine Formula 1 Era entitled "CLASSIC GRAND PRIX CARS" by our own Atlas forum member, Karl Ludvigsen. Sutton Publishing U.K. 2000.

Can't recommend this book too highly. It saves a lot of searching through archival material, for not only Front Engined information but has excellent descriptions of the advantages of both configurations. It explains why our F1 cars are all rear engined and likely always will be where-as sports cars are enjoying a front engined resurgence.

[marion5drsn]

In an attempt to resolve the problem of the difficulty of grinding the closing cams of a desmodromic lobe I attempted to draw the lobe using an old lobe of Mercedes in 1939.
In this attempt I found that it is an exceedingly laborious task. And not easy to achieve with my limited knowledge. The problem seems to be that when transferring the curves from the standard lobe to the “odd” shaped closing lobe, the curves do not intersect in a smooth manner. This is easily seen when the layout and the finding of mistakes is done. Another problem is the closing lobe must of necessity be made in a much larger base circle to accommodate the core diameter.
It is no wonder that the lobes shown in some pictures are with weight reduction holes as these lobes are of considerable size. The ones I drew had a base circle of 42 MM (1.654”) Much larger than a Chevy cam lobe of about 23.9mm (.942”). This was not of sufficient diameter to support the torques and twists of a camshaft. Another thing I found was the amount of metal that would be out of balance with these arrangements. It being no wonder that camshafts may cause a lot of out of balance conditions. This is especially true if they are of the First (Primary Shake) and the Moments when the shaft is out of balance fore and aft.
It is a likely thing that Mercedes when making the M196 had a lot of problems getting a smooth transition between the opening and closing lobes. My guess at this point is that the draftsman who made the layouts found this out in a short time period. He may have made the closing cam first and then transferred this to the opening lobes this being the opposite method I choose.

I am also wondering how much trouble Taglioni had when doing the Ducati? Altho he may have had the benefit of a computer. Mercedes did not! These cams lobes must be made to very close tolerances as is very apparent when drawing the lobes .00254
mm (.0001”). The tolerances of the clearance are also a close and critical thing. This being one of the things that Ducati seems to have a problem with. It is also apparent that the over the top of the two lobes is super critical in the smoothness of transition as already mentioned. I sure would like to have one of these on a Vinco optical dividing head. M.L. Anderson

[carlos.maza]

Desmo:

I know how Ducatti system works.
Was Mercedes' version different?
If so, could you explain us the differences? (or direct us to a link or another source?)

Thank you
Carlos

[desmo]

From what little I can tell the Mercedes system was essentially similar to the SOHC system used on the bevel-drive Ducatis you see in miniature in my avatar. Here's a photo I found of a model of the Mercedes system:

http://members1.chel...n/Mercedes1.jpg

This is from a Dutch website that has pretty much everything you'd want to know about desmodromics.

http://members1.chello.nl/~wgj.jansen/

[carlos.maza]

Thank you very much Desmo.

Carlos

[Top Fuel F1]

Yes, very good desmo!

Rgds;

[H. Eckener]

Hello all,

Here are two pictures (I hope they show up, first time posting and first time posting pictures) from H. Scherenberg's SAE paper from 1958, entitled "Mercedes-Benz Racing Cars-Design and Experience."
It is a very interesting paper on F1 design of the 1950's and specifically Mercedes's approach to it. It goes into some interesting details on engine development, ie the desmo valvetrain and direct injection, aerodynamics, suspension, and track analysis. Well the pictures are the most interesting, so enough talk.






PS. An interesting quote from the paper on the desmos specifically: "How the drivers liked the willingness of their cars to speed-up could be clearly seen after the various races, when the maximum tachometer needle reading was checked. The one in Fangio's car usually showed over 9000 rpm"

There is also a discussion section at the end, where they even have Mr. Arkus-Duntov of Cheverolet Corvette fame making a few comments.

[desmo]

Welcome to the Tech Forum and thanks for posting that. If it's true that M-B were getting 42% more valve lift than their competition, it's no wonder that they were mopping the racetracks of Europe with their competitors. 9000rpm was pretty good for the day, although M-B seems to have long been able to make high revving yet reliable engines. I had an old 220SE that was built in '65 and was redlined at 7000. Pretty remarkable for the day in a rather stodgy 4-door sedan.

[Yelnats]

H. Eckener Nice Photo and it's interesting to see that the big advantage off desmo valves systems is shown in the huge advantage in the negative (closing) deceleration curves being 228% greater than that supplied by spring closure. This surprises me and perhaps is an indication of the poor spring materials used in that era.

[H. Eckener]

Yelnats,

Not only would materials of the spring be a limiting factor, but also design and the general ability to do the computation for a valvetrain. Without computers, I would imagine very advance valvetrain modeling to be extremely limited. Or if instead you took the hands on approach of making various valve springs and valvetrains and testing them, it would be very time consuming and expensive.
What I would love to see, is a comparison between different types of systems, like a modern Ducati desmo setup, a pusrod setup, a setup before pneumatic valves and a modern pneumatic valve setup, plus at various engine conditions. It would prove very enlightening.
But, from the point of view of pure speculation I wouldn't be too surpised if a modern theoritical desmodromic valve setup would still be superior in terms of the accelerations it could impart to the valves, then the pneumatic valves of today.
But, then again I would be extremely impressed if the actual desmodromic valves of today could handle the revs. My hat would be off to the machinist as well for holding such tolerances.

[marion5drsn]

For the past two weeks I have been drawing a 330 degrees duration lobe for a desmodromic system and the more I work on it the more I appreciate people like Taglioni. After doing this I see no problem in the acceleration of the valve with modern materials. The big problem seems to be is the lifting of the valve at these extreme durations. This due to clearance ramps.This is true after you get past 300 degrees of duration. Getting them to work below 300 degrees is no problem. My guess is that the GP engines of today have cam durations at or greater then 300 degrees. The model I am using is the one from the 753 Porsche. One of the big problems is the clearance ramp in the extreme duration lobes. The opening lobe must match the closing lobe to a great degree. It seemed to work on paper but whether it would work in practice is problematical. These particular lobes are engineered to work on the same shaft as opposed to engines as the Norton, which has cams turning in opposite directions. Requiring eight camshafts on a Vee engine. With modern computers it should be no be problem to make Desmo cams work better than Air Springs. It is just a matter of the engineers making up their minds just to do it.
One big advantage of the Desmo over Air springs is the acceleration rates at the opening side can be much greater than with ordinary lobes. You just aren’t pushing around the springs. I can’t see any reason why with all the money that F1 throws around this isn’t redone, especially by Mercedes. Another thing is the difficulty of the valve lash. These engines need hydraulic lash adjustment very much. It’s no wonder that Ducati owners have a lot of trouble with this item. I did not allow any different dimensions on the opening and closing sides of the two lobes altho this would be easily accomplished. Also it is easy to see just why Cadillac used finger followers on their newest V-8, with the lash adjusters on the end of the finger opposite the valve stem. I wish I had a picture of this method.
M.L. Anderson

[leegle]

If I understand what you are describing here Marion the Cadillac system would be very similar to that used by Opel and Holden in the engine we know in Australia as the Camira engine. The hydraulic adjustment was in the fulcrum point with the cam lobe operating between that and the valve.;)
Getting back to F1 engines would weight be an issue with desmodromic operation?

[desmo]

On my old SOHC M-B you adjusted the valve lash by raising or lowering the fulcrum point on the cam followers. It was a good solution for the application, easily done and it kept it's adjustment for a long time. At the interval to check, they were usually still right on spec. I believe all F1 engines use shims, which are good for racing but a real pain. My roommate and I built a basket Ducati desmo single which was shim adjusted and it was really a chore to get all the clearances right.

[H. Eckener]

Marion5drsn,

Although, I strongly agree with you that an automatic and accurate means of lash adjustment would be required and that certianly on paper the desmodromic valvetrain looks very appealing. I am still a little bit skeptical of it in practice. I am curious, how are you modeling this? Could you possiblely post some more details or a greater description of what all your model takes into account to the list. I would find it most interesting.
Didn't Cosworth in the pre-pneumatic valve days try to develop desmos again? Anyone know what came of it? Perhaps this is entirely incorrect, but could a hydraulic lash adjustment system suffer from cavitation or a situation where the adjustment "pumps" itself up and causes interference problems. Maybe I am just thinking too much along the lines of hydraulic lifters.

[Marco94]

Check out this link to a Ferrari desmo item. Don't know about any Cosworth projects of that kind.

[Marco94]

H. Eckener,

What is the number of the SAE paper? I'll see if I can get a photocopy of it somewhere.

Further to M-B's system, for 1956 they wanted to use a 4-valve lay-out with springs again. According to "Die letzte Silberpfeile" at least. I am not sure if anyone ever saw a picture of the combustion chamber, but it uses a very roof-like top and a matching piston shape. Not a nice shape for a combustion chamber.

[marion5drsn]

Desmodromic the word and what it means when used in conjugation with valves in a piston engine. What the desmodromic cam system does that is more apparent than first noted is that the two rockers completely restrict the valve jumping out of its racecourse! If the valve wants to jump up out of its designed path it cannot do so without breaking or bending the rocker of the valve-closing lobe. The same holds true of the closing rocker arm doing the same to the opening cam lobe. This is way the word is used and the person who coined it really knew his stuff as the cam followers are literally running within a fenced racecourse. It may not look as a racecourse looks to us but that’s the way it is! The fence is the opening and closing ramps plus the part of the lobe that is somewhat semi circular in shape. Both lobes forming the fence. Unless something breaks it will not jump, float or other undesirable things that can happen to valve springs especially if metallic. M. L. Anderson

[H. Eckener]

Marco94,

I don't know of any specific SAE paper on the Cosworth desmodromic valvetrain attempt. But, if you go to the link the actual desmo posted, and then go to the section titled "systems" and check out Cosworth, you will see a few none SAE refrences that you could follow on Cosworth specifically. There seemed to be some extensive testing and two patents assigned to Ford.

http://members1.chello.nl/~wgj.jansen/

[desmo]

Here's the text from the cossie bit on the desmo site:

"Cosworth in the eighties and Ferrari in the nineties also tried but rejected desmodromic valve control."
Source: Classic Grand Prix Cars : The Front-engined Formula 1 Era 1906-1960 by Karl Ludvigsen (cop. 2000) p. 31 (in section "Introducing Desmodromy")

In our book "desmodromie" (cop. 1990) we mention Cosworth with regard to DVG on three places and we discuss two Ford patents also(see pictures).
P. 52/53 we discuss the patent of Peter William Bothwell (see first left picture).
In those days there were rumours telling this patent was intended to sell to Cosworth.
P.79/80 we discuss the cooperation between Cosworth and Ducati (Bordi) in 1987.
(design of the 4 valve 851)
From Bruno de Prato (in Moto Technica no. 1 '87) we learn that Cosworth extensively tested their desmo-motor but eventually closed the subject.
The Cosworth engineers regarded the 4-valve desmo-cilinderheads thermodynamically ideal. It could make the DFV/DFY 3000 F1 engines competable again, but the mechanical complexity and the limited space would bring too much complications.

From John Blunsden in his book "The power to win : The design, development and achievements of the Ford Cosworth DFV, DFL and DFY V8 racing engines", we learn that the desmo-programme was initiated by Duckworth in 1979, prompted by the difficulties in trying to solve the valve spring surge problems.
The desmodromic programme was cut short because of underestimation of the subject and doubts about the the significance with regard to cost and necessary man-years.
It's a pity that there is no drawing or picture of the desmodromic test rig in this chapter.
From the patentfigures - although they are of later ('83; '85;'86) date - we may conclude how the tested system(s ?) could have looked like.
We are waiting patiently for the confirmation (in text, photographs and drawings) from Cosworth or from other good authority.
More research to be done.

[815]

its going through my mind for a while:
T/fore,is anyone out there,who can answer:

1.Desmodronic is in German 'Drehschieber' (Einlass) ?
2.Before Ducati used this system,same design was used by
MAICO,the German motocycle manufacturer? (early seventies)
Or was it the other way round,that MAICO used it after Ducati?
Thanks for a little help.

08 15

[leegle]

About 1973 I seriously considered buying a basket case 250cc Ducati desmo and it was not new. I understood ducati had been making them for years.

[karlcars]

"Drehschieber" refers to a rotary valve, not a desmodromic system.

Many thanks to Yelnats and Desmo for the references to my book!

The photo posted by Desmo from the Dutch site shows a test rig for the desmo system used in the Scarab engine. This was a kind of copy of the M-B system but it had a very crude and heavy clearance adjustment. For its desmo engine M-B developed a super-sophisticated system of twin eccentrics inside the pivot of the bell crank that closed the valve. When I figured this out from drawings of the engine in 1956 and published it in SCI Mercedes had a huge internal inquiry to find out who had spilled the beans!!! The whole M-B desmo story is in my book "Quicksilver Century."

[desmo]

A book I can highly recommend. I always liked eccentrics for adjustments, elegant and usually very simple to adjust. The Duck basket I built was pure mid-60s style although I cannot recall- if I ever knew- what exact year it was. Great little bike, fun to ride and beautiful music from the pipe.

I can't recall Maico ever producing 4-strokes, although I could certainly have missed it. When they changed their name to M-Star it all became rather sad to me.

[marion5drsn]

After a great deal of work I finally drew a closing cam that looks a lot like the on shown in the M-B and Ducati Photos. In this I found the real reason for the large size of the closing lobe, it is due to the necessity of arriving at the correct shape of the lobe which must be almost twice the diameter of the opening lobe. If made too small it is much too sharp at the blend points where the lobe changes from opening phase to the closing phase. One must always keep in mind the two rises and drops must correspond to a very close degree and in linear measurement! The Master Cam Lobe must be within about .0001” or about .0000039mm. On the production cams I doubt that any more than .001” or .0000394mm would be acceptable. This is on racing cams; on garden-variety cams more might be tolerated. Another thing is the point that the rockers, and especially the closing rocker, must be very closely machined to assist in the sealing of the valves as there should be no burning of the valves and the valve seats. It is no wonder that Ducati has a reputation for needing valve adjustment about every 6,000 miles (9662k). Altho this may be due to a poor design of the adjustment mechanism. There just isn’t much room to make a large adjustment mechanism! The one piece seems to have a beating down of the soft looking “Lock”. Go to www.mbpducati.com to get a good view of this fix. In my communications with Henk Cloosterman in Holland I found out a lot of things about books of Desmodromic cams and valve systems. Design of Machinery by Robert L. Norton, which has two chapters on cam lobes are very revealing. As I have stated before these cams seem to need in a desperate way some form of hydraulic valve lash mechanism similar to the ones used by Ford S.O.H.C. and D.O.H.C. V6 and V-8 production engines.
These types of cams need a lot of engineering to achieve success and it’s is no wonder to me that only M-B and Ducati have succeeded in this field. Up until the last several years I would have given Ferrari no chance at all in this. It takes almost infinite patience to even make the two lobes and then to make the rockers work in conjunction with this is a real bear! One must also recall that M-B and Ducati at first had no computers to do the extremely laborious task of the arithmetical calculation of the lobes. ……In the case of a 330-degree lobe duration one must make about 36,000 plunge cuts to form the Master cam. If Continuous Numerical Control is used then much of it can be put into the program and lessen the laborious tedium. Norton’s book does not mention the use of a C.N.C. controlled rotary table, which would also smooth out the scallops of plunge cutting. This in the making of the Master Cam Lobe(s). I really do not believe that M-B even approached the degree of machining necessary to achieve a decisive cam lobe, as they didn’t have the computers to do this!
Another thing about the designing the lobe is the use of Critical Path and polynomials seem to produce a lobe with far too much lift for the duration of the lobe valve lift. The way I see it one does not need to worry about spring harmonics to the degree as when using metallic or even air springs. Altho even rocker arms can act as springs. The big worry seems to be in getting the valve to seat properly without banging it down on the seat. Notice that Ducati uses small assist springs to achieve the final closure of the valve.
This due to the condition of the exhaust valve in particular grows a great deal in the warm-up period of the engine and this must cause trouble when designing the lobes and the lash problem.
I would give a good group of engineers about 6 months to design, manufacture and debug a Desmodromic system starting from scratch. Anything less would be a big plus. The debugging time being the biggest part of it. One must remember that it takes a completely different cylinder head, timing gear, pinions and idler wheels to drive the mechanism.
These cams are heavy, as the closing lobe might be as large as 1.181” (30mm) radius.
Just what this would do to the engine vibrations is a problem of unknown quantity, as they would be turning as fast as 9,000 rpm or more. M.L. Anderson

[desmo]

Thanks for the good info, Marion. I think you are spot on about using CNC machining to produce the closing cam, although I wonder if EDM or perhaps even investment casting might be useful alternatives. The old bevel-drive Ducatis were, I believe, run without the "clothespin" closing springs in racing trim, as they were said to be neccessary only at low rpm.

[Yelnats]

Just a thought about tolerances in pre computer Desmo valve systems. I wonder if the old timers resorted to hand lfitting of the cams as was often done for valves and bearings in the earlier years of engine building? Perhaps they cut these cams to a rough fit and then hand mated the surfaces together with lapping compound and elbow grease.

[desmo]

Marion mentioned the desirability of having a hydraulic lash adjuster for the closing cam/follower. Scarbs sent me this image he got of an alternate approach to desmodromics where the closing force is hydraulically communicated to the valve intermittantly, it's timing controlled by a cam working on a cylinder (Thanks Craig and Andrea!



If one could devise a hydraulic actuator that could be electrically controlled that could cycle fast enough to follw the demand, perhaps one could lose the closing cam mechanism.

Not a new idea, this appearing to be the work of Manfred Lorscheidt, Mercedes desmo proponent in the '50s. This may be worth it's weight and complication by dint of potentially reduced friction effective pressures compared to a conventional sprung valvetrain.

Could this be Renault's much speculated about innovation?

[Marco94]

H. Eckener,

I am sorry to have formulated my question so badly. I wanted to have the number of the SAE paper about the M-B engine, not the Cosworth. If you have that number I can check my library for a copy.

Marco.

[marion5drsn]

Yelnuts; I would not be a bit surprised that this was done. I didn't happen to think of it! Marion L. Anderson

[Gunner21]

In my humbel opinion I still think that the camless and valvless ball check intake is the way to go. More power and less weight and no blower needed. It may also work on the exhaust side of the engine.

Gunner21