Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site watdaisy.UUCP
Path: utzoo!watmath!watdaisy!gvcormack
From: gvcormack@watdaisy.UUCP (Gordon V. Cormack)
Newsgroups: net.auto
Subject: Re: Camless valve operation
Message-ID: <6852@watdaisy.UUCP>
Date: Fri, 11-Jan-85 18:57:35 EST
Article-I.D.: watdaisy.6852
Posted: Fri Jan 11 18:57:35 1985
Date-Received: Sat, 12-Jan-85 05:18:17 EST
References: <6848@watdaisy.UUCP> <1737@zehntel.UUCP>
Distribution: net
Organization: U of Waterloo, Ontario
Lines: 83

> > It has been suggested that the valves in an engine be operated
> > electrically rather than by a cam.  There is no doubt that such
> > operation could give more nearly optimal valve operation than
> > a fixed cam.
> > 
> > The above scheme has been implemented for the purposes of engine
> > testing (I can't remember the reference, it was either PM or C&D or
> > R&T in the last 10 years).  The trouble is that the solenoids to
> > operate the valves are tremendously inefficient and consume a
> > significant fraction of the overall output from the engine.  
> > 
> > If someone could invent an efficient mechanism to control the valves
> > like this, he would indeed have a better mousetrap.
> 
> This article, and the one suggesting the use of solenoids to operate
> springless, desmodronic valves, miss the point. What you have to do
> is separate the energy required to open the valves from the control
> of that energy.
> 
> Assuming that we stay with valves opened by engine-driven cams and
> closed by springs, how can we vary the lift and overlap of the valve
> timing?
> 
> Well, as it turns out, the elements of a system to do that have already
> been marketed although one beat a hasty retreat. For five years now,
> Alfa Romeo has had dynamic valve timing on their 4 cylinder, twin cam
> engines. The intake cam can rotate several degrees relative to its drive
> sprocket. It is set to be fully retarded at low speed to give a smooth
> idle and good low end torque. It advances to produce a good degree of
> overlap at high speeds for top end power. Orignally the cam timing was
> controlled by a govenor but now that they've gone to the new Bosch
> FI, it's controlled by the same module that controls the spark timing.
> 
> Cadillac's ill fated "modulated displacement" 4-6-8 engine (marketing
> wouldn't allow "variable displacement" because of the acronym) had
> an ingeniously simple method of shutting off the unused cylinders.
> A solenoid was positioned above the pivot of each rocker arm. If the
> cylinder was to be "on", the solenoid was extended forcing the rocker
> onto its pivot and the valve worked normally. If the cylider was to
> be "off", then the solenoid retracted allowing the rocker arm to "float"
> with the tip of the arm pivoting on the valve stem.
> 
> Of course, this represents the "bang-bang" approach to valve lift
> control: full lift or no lift at all. It seems that a variation on
> this idea could produce the infinitely variable lift desired.
> 
> I realize that this postulates a twin cam engine with rocker arms, a
> rather odious thought to those of us that admire the simplicity of
> the classic twin cam design. But at least that's much more in the realm
> of accepted engine design than solenoid-actuated desmodronic valve trains!
> 
>                      Jack Hagerty, Zehntel Automation Systems
>                       ...!ihnp4!zehntel!jackh

It is not I who missed the point.  First, the thing that is difficult
to vary with a cam is the duration of the valve opening.  It is not
clear that varying the timing is all that desirable without extending
the duration.  Second, I did suggest a cam-operated scheme that was
able to achieve this.  It looks like many of the components 
from the Alfa system you describe would be applicable.  Third, I
never suggested that valve lift should be modulated.  I think it is
more reasonable to consider the Cadillac system as bang-bang 
control over valve duration.  If the intake valve duration could
be controlled precisely, the intake valve could be closed when there
was enough air/fuel in the cylinder for the desired power.
Such a scheme could eliminate throttling and its inherent losses.

I would like to make a couple of semantic points.  First, I don't
know what "desmodronic" means, and neither does my OED.  Second,
I have long been on a campaign to stamp out the use of the phrase
"infinitely variable" for "continuously variable".  In this context
I am not really complaining, but the term first started to irk me
when manufactures of electric ranges began to tout them as having
"infinite heat" controls.

To get back to the point at hand, I believe there are many interesting
things one might do with a different mechanism to control the valves
(whether with or without a cam).  And I do not believe that this
newsgroup is a forum only for "accepted engine design".

Gordon V. Cormack,  University of Waterloo

gvcormack@watdaisy.uucp       gvcormack%watdaisy@waterloo.csnet