Desmo Drom i

7/22/2019 Desmo Drom i

1/5

,

og

DesmodromicsMick Walker explains how Ducati beattbe world with spring-less valve operation

HE WORD desmodmmic wil! not befound in the Oxford or any otherEnglish dictionaries. It was coinedrom two Greek words, meaning 'controlledun' and its mechanical function is to elimi-

ate one of the phenomenon of valve float,r 'bounce'. This happens at high revs whenhe valve springs are unable to responduickly enough to close the valve back onheir seats. 11le desmodromic idea was toeplace troublesome springs with ae(\. jcal closing system much like thatsedl open them.Eliminate the springs and you eliminate

the bounce and get a higher-revving engine- in theory.This had been known since the early daysof the internal combustion engine but formany years no designer managed to success-fully harness it. One of the first examples ofcompletely positive mechanical valve oper-ation was used by the French Delage concern

in its grand prix car of 1914 - and theFrench knew the method as desmod-romique.The Delage engine employed four valvesper cylinder in pent-roof combustion cham-bers. Twin overhead camshafts were used

and each valve pair was actuated through astirrup-shaped tappet with an integraI, dup-lex bridge-piece at its lower end. Two camswere used: one depressed the tappet toopen the valve via the upper member ofthebridge, the second cam raised the tappet toclose the valve through the medium of thelower bridge member. Between this lastitem and acollar on the valve stem a small'toleranee spring' was interposed to ensurethat the valve was pulled positively on to itsseat. lt would not have been possible tomanufacture sufficiently accurately then toprovide positive seating without somedegree of freedom.The Delage system was not particularly

James L Norton - tbe great Britishdesigner conceived bis version of tbepositive valve operation in tbe late1920's.

SIC MECHANICS

successful and was not developed, but inthe 1920s several other manufacturers triedvariants of it. One layout, the Vagova,utilized a cam track embodying inner andouter cam forms which guided a rollerattached to one end of a 'rocker', the otherend of the rocker was forked to embracethe valve stem. To provide the requiredfreedom for seating, the pivot of the rockerwas given a small amount of spring-Ioadedfloat parallel to the valve axis.Alternative methods investigated were touse the cam track in conjunction with a tap-pet for direction operation of the valve, orto retain the rocker but incorporate twocams, one bearing on each end of therocker. A more complicated version wasthat used by Bignan-Sport, aIso during thetwenties. It featured a bevel-gear drivenface cam or swash plate provided with avee-section periphery. Motion was con-veyed from cam to valve by a crossheadrunning on a guidc; the crosshe;ld carricd apair of oppositely inclined rollers, one ofwhich ran on each face of the swash-plateperiphery.

In Britain another fonTi of desmodromicoperation was patented in the early 1920sby none other than the legendary James LNorton, founder of perhaps the greatestname in British motorcycling. This wassimilar to the Vagova design in that itemployed a cam track and rocker. How-ever, instead of Vagova's spring loadedrocker trunnion the Norton rocker endentered a slot in the valve stem and a smallleaf spring was fitted on the rocker's uppersurface. Complete closure was ensured bythe provision of an adjustable abutment inthe valve stem above the rocker. Ingenious,but one cannot visualize slotted valvesstanding up to todays thermal and mechan-ical stresses.A complex variation of the Bignan-Sportface cam theme was patented by the cargiants Fiat, also in the late 1920s. A springloaded crosshead was mounted on thevalve stem; at each end of the crossheadwas a roller, and the rollers ran between apair of faced cams mounted coaxially withthe valve and driven by spur gears. 11ledirect ion of rotation of the rollers would, ofcourse, be suddenly reversed as they trans-ferred from the opening to the closing facecam, so that the rate of wear of themechanism might weil be high.The double cam layout has had manyadherents in the past. In 1916 J M Brewsterinvented an overhead valve arrangementoperated by rods. 11le opening cam was fol-lowed by a roller ended tappet connectedto the valve opening rocker by a jointed ~

25


2/5

push-pull rod. On the tappet was a trans-verse peg engaging with the forked end ofone arm of a bell crank lever. The other armof the bell crank carried a second rollerwhich bore on the closing cam. But this lay-out inc1uded more pivots and arms thanwould appear desirabie, though it wouldprobably have worked weil enough at thelow engine speeds of the day.A spring loaded trunnion appeared alsoin the Vareille design in which, instead ofthe cam track as on the Vagona, there weretwo cams between which the rocker endwas sandwich ed. A shackle connected therocker's other end to the valve stem. Tech-nically, though, it would have been morepractical to have had a forked rocker endand abutments on the valve; otherwise theVareille system was robust and simple -even by present day standards.But one of the neatest and most sensibleof the early designs - and which was similarto what both Mercedes and Ducati usedyears later, came from Ballot, a name onceweil known in French motoring circles dur-ing the veteran and early vintage period.This comprised a double cam and three armrocker of Y form. One arm bore on the

Tbe granddaddy of all desmodromicsystems, F H Amott's valvemechanism was patented in 1910. Itincorporated a cam track, a bell cranklever and a tappet.

-~Vagova utilized a cam track androcker; tbe rocker trunnion wasspring loaded to ensure completerocker closure.

opening cam, the second on the closingcam and a third had a forked end whichactuated the valve through abutments.Complete closure was ensured eitherthrough a spring above the fork or by allow-ing some spring controlled movement bet-ween the two cam follower arms of therocker.One of the last designs to appear prior tothe onset of the barren years for desmod-romics (1930-54) was the work of G AMangoletsi, then a weil known sand racer

from Southport and later head of the G MCarburettor Company in the 1950s. Notunlike the Ballot valve gear, Mangoletsi'sscheme was even simpier in that the fol-lower of the closing cam was embodied inthe valve actuating arm. The opening camfollower ran on the same pivot as the arm,and a spring between the two componentscompensated for cam irregularities ofmanufacture or wear.Valve position was set relative to therocker arm by means of adjustable collarson a threaded valve stem. Spring washersbetween the collars and rocker abutments

provided a degree of freedom to ensurecomplete closure.

.> fll \.~1914 French Delage employed adesmodromic layout of two cams anda stirrup member; each stirrupoperated two valves.

Tbe Mangoletsi system used a leafspring to take up any irregularitiesbetween tbe proffies of tbe two cams.

A LL THE activities in the 1920s mightsuggest that, apart from Brewsterand Delage, no one had done any-thing earlier on positive c1osure. However,that wa..


3/5

doubt the reason it is not in wider usetoday).Interest in desmodromic valve operationwaned until in the immediate post-war daysit had almost been forgotten when in 1954the giant Mercedes Benz concern decidedthat something better than spring returnwas required for its new W196 21/2 litregrand prix engine. The German companytherefore evolved a desmodromic layout,.the success of which was evident if oneconsults the car's 1954 racing record. TheuJ1


4/5


5/5

Even so desmodromic Ducatis took second,third and fourth places in the race.Even better was to come. The nextround, the Belgian GP, was held over thesuper fast Spa Francorchamps circuit. Herespeed, rather than riding ability was theprime requirement and the Bolognadesmos clearly demonstrated their poten-tial by taking flrst, second, fourth and sixthplaces with world champion Ubbiali rele-gated to fifth spot!And it is generally accepted that ifDucati's leading rider Alberto Gandossi(who had not ridden in the Isle ofMan) hadnot fallen in Ulster, he, rather than Ubbiali,would have been champion at the season'send. The flnal found at Monza had seen the

mighty MVsannihilated, with Ducati takingthe flrstfive places. A brand new prototypetwin cylinder model flnished third.Unfortunately just when Taglioni's brainchild seemed likely to totally dominate theultra lightweight racing category, a changein factory policy meant only limited partici-pation the following season. Even so an 18-:,old youngster named Mike Hailwoodmm'his flrst GP victory (at the Ulster GP)on a desmo single that year.Stan Hailwood, Mike's father, had been soimpressed with the desmo's performanceduring 1958 that he bought Mike a coupleofworks 125 desmo singles, and even com-missioned 250 and 350 twin cylinder ver-sions. These latter two designs would nodoubt have had a better chance of repeatingthe single desmos success had they beendeveloped by the factory, rather than beingpurchased, largely untested, by the Hail-wood equipe.During the period 1958 to 1961 manyothers tried unsuccessfully, to emulateTaglioni's success. Firstly, Mondial wouldmost likely have raced a desmodromic 250in 1958 had they continued racing. Benelliexperimented but gave up. A desmodromicversion of the Manx Norton was built, butthis was not proceeded with. Even privateconcerns, s\lch as Velocette specialistsB~ of Ilford, Essex, converted a 500'. pm to desmodromic valve operation,6tIf.none were anywhere as successful asDucati.But probably Taglioni's crowning glorycame in 1968 with the introduction of theflrst mass production motorcycle (or forthat matter vehicle), with desmodromicvalve operation.For a man who believed that the simplethings were usually the best, the newcomerswere eactly the same as their Mark 3 ohcbrothers from the cylinder barrel down-wards. The mechanical operation for valveopening and closing was similar (althoughwith only one camshaft) to the type thatTaglioni had devised for his earlier desmoracing machines. The main difference wasthat in the roadsters the valve closing wasassisted by springs, unlike the racers.However, the springs used were verymuch lighter than the ones on non-desmomodels (in fact, they werefrom the 125/160 roadsters ). And as proof of Ing.Taglioni's simply-is-best-approach, many ofthe cylinder head's less critical componentswere borrowed from existing productionmodeIs giving both Taglioni's productionstat! and potential owners a far easier timewhen spares were needed.

CLASSICMECHANICS

Cementing tbe legend, Mike Hailwood retumed to Ducatis - and tbe Isle of Man- to create a sensation by winning tbe 1978 Formula I TT.Compared with the valve spring enginesvalve seat pressure of 80lb/square inch, thedesmo was Bib/square inch makingkickstarting easier and givingmuch less seatwear - both important advantages for road-. going units.Amazingly the only real difference bet-ween the standard valve spring Mark 3 andDesmo engine was, that the latter had fourrockers (in place of two on the Mark 3) anda four lobe camshaft, in place of two. Eventhe valve head diameter and materials wereidentical but the method of valve stemcollect retention was different!The commercial success of the single

cylinder desmo roadsters meant tha t whenthe factory introduced its V-twin range inthe early 1970's it was not long before

desmo versions of these too began toappear.

B Y 1980 aU Ducati productionmotorcycles had desmodromic valveoperation and this is the picturetoday, with Ducati, now owned by Cagiva,still exclusively marketing a range of bikeswith one feature which no other producerin the world offers - and that magical wordDESMO. This sales success is surely lastingproof of an idea which had its roots flrmlyin the veteran and vintage days of motor-cycling, but even today is viewed by manyas being amongst the very vanguard of twowheel progress - a rare achievementindeed.

29

Documents

Desmo Drom i