I was just reading an article on the new Ducati 1098 sportsbike and there was another explanation of their clever Desmodromic valve system. Every time I come across this, it strikes me as a very elegant engineering solution.
For those not familiar with it, the Desmodromic system employs no valve springs. Instead, the valves are both opened and closed by positive action from a relatively simple cam and rocker arrangement. In performance terms, it allows the engine to rev safely, but I would have thought it would have efficiency benefits too, as the cam is not having to "overcome" the resistance of the valve spring on every engine cycle. On high revving engines which need stiffer valve springs, I would have thought it would have particularly useful benefits.
Anyone know why this hasn't made it into the car world, or even to other motorbikes? I believe Mercedes have used a similar system in a couple of their racing cars over the years, but that seems to be it.
Cheers
DP
|
|
It's expensive. One would think it must also have some sort of very short spring arrangement to take up any wear or compensate for manufacturing inaccuracy. If these are absent, any wear would result in a loss of compression.
|
and not necessary. At very high revs the valve springs just can't move the valves fast enough. Solution - desmodronic valves. F1 cards use them. You and I don't need them.
JH
|
|
The Hondas of the 1960s used to rev to 22000+ rpm and did so quite happily with normal valve springs. Not sure why you'd need desmodronic valves in that case. I believe the standard cam belt change interval for a Ducati is 2 years/12000 miles - sounds expensive to me......
|
The Hondas of that era had tiny, tiny valves (literally, the stem and head are like a 4-inch nail) which have very low intertia, so were not subject to bouncing even with relatively weak valve springs.
As the valves get bigger they have more inertia and are more likely to float or bounce, so a desmo system makes more sense.
The other advantage of a desmo system is that pushing conventionally-sprung valves open creates a lot of power-sapping friction. Because desmo systems have no springs there's less reciprocating resistance ..... but it's very complex and expensive to engineer
|
|
I think even Ducati have admitted that it was a marketing ploy by the late 1970s. Probably an advantage over valve springs in the 1960s though.
|
|
|
|
Cam belt? my desmo had a shaft and bevel gears
------------------------------
TourVanMan TM < Ex RF >
|
|
Checkout the maintainance intervals and costs of Ducatis and that may give an idea why it didn't catch on.
|
Thanks all for the replies.
Maintenance costs on Ducatis are high mostly because of rip-off parts prices. This was covered well in Ride last month where they showed that an in-tank fuel filter from Ducati for a 916 costs more than £40, but one from a car spares shop to fit a Vauxhall Vectra from the same manufacturer with the same part number costs a tenner. There's also the cambelt changes which most recommend replacing every 6,000 miles. I don't think any of it relates to the Desmo system as such, and they could still run longer lasting chains to the cams instead of belts.
I would be really interested to know how much more efficient it is than a conventional sprung valve, as it strikes me as quite a lot of wasted energy pushing a heavily sprung valve open.
Some really interesting thoughts here - cheers!
DP
|
Normal valve action (valves with springs) is a little more complex than mentioned here, the actual course of events is:
1) The cam pushes the valve open against the spring.
2) The spring then slows the valve down as it nears the top of its lift.
3) The spring then pushes the valve shut aginst the cam.
4) The cam then slows the valve as it closes against its seat.
A desmodromic gear driven system, must do both the accelerating and decelerating of the valves without the assistance of the stored energy within the springs. I do not have any figures to hand, but the difference in energy to operate them between 'normal' and desmodromic systems is probably quite small. I think that the main benefit of desmodromic valves is simply one of control at very high revs; there are no springs involved to give rise to 'valve bounce'.
|
Indeed, a good deal of the energy which is put into the spring as it is being compressed is given back to the cam as it relaxes. The energy lost is mainly to friction, which can be minimised with good design.
One of the limitations of valve springs is surge - this is the phenomenon where the valve spring begins to resonate itself (i.e., the excitation frequency matches a natural frequency of the valve spring) beyond this point, the valve is no longer under the control of the spring. One very effective solution o this problem is to replace the metal valve spring by an air spring; as per Formula one practice.
Although the cam on a conventional engine feels difficult to turn at slow speeds - and you can feel the valve springs driving the cam every so often as you turn - these forces will not be dominant at high speeds, when the inertia forces required to accelerate and decelerate the valve dominate.
My view is that there is no real justification for the complexity and expense of desmodromic valves, and they are a bit of a gimmick.
Number_Cruncher
|
|
|
|
|