two ignition ideas

[Helge Hafting]

In <MAPI.Id.0016.00796e63687269733030303830303038 at MAPI.to.RFC822>, on
08/05/98 
   at 05:38 PM, "Chris Conlon" <synchris at ricochet.net> said:

>Hi eveyrone,

>A couple of people (on my other lists) have mentioned various engines
>with 2 plugs per cylinder. (Not even counting rotaries, diesels, etc.)
>Apparently in some designs the second plug fires much later, and helps
>reduce emissions significantly. This led me to a couple of weird ideas
>for ignition systems.
You don't need an extra plug if all you want is to fire twice.  There  are
ignition systems that fire the same plug several times.  Look at:
<http://www.msdignition.com/2mcpulse.htm>
for an example.  2 plugs in the same cylinder are usually fired 
simultaneously.  This ignites the charge in two places making it burn
faster.  This gives more pressure while the piston still is at the top,
making some more power.  Using 2 plugs may also allow a somewhat higher
compression without getting detonation.

Firing several times in one cycle avoids misfiring problems and make sure
everything is burned.  This gives less emissions and slightly more power. 
The effect is most noticeable at low speed/idle.  (There isn't much time
for multiple firing at high rpm's anyway.)

>First, take a basically stock ignition system, and build a small inverter
>to supply 3kV-5kV at a few mA, and a HV capacitor of hopefully several
>hundred mF. Using some HV diodes, wire this in parallel with the existing
>ignition system. The idea is to have the main spark fire, and then have
>the inverter/capacitor continue a lower-current, long duration spark
>across the now-ionized gap. (You'd need one per cylinder, or some kind of
>icky switch, etc etc.)  Hopefully you'd get more complete combustion,
>reliable ignition, maybe a bit lower emissions.

>Has anyone tried this? The idea seems simple enough that I imagine it's
>been tried and discarded for some reason or other. (No major benefits
>over a well-done normal ignition, would be my guess.)
Price is one thing - you are introducing a lot more components.  And will
it work at all?  The "ionized path" may very well be blown away by
combustion pressure stopping your low-current spark.

>Second, something made me think that it would be clever/cool to use
>short-wave UV light to ignite the charge. It generates ozone, after all,
>which ought to ignite things nicely. I'm not sure how you'd make part of
>the cylinder head transparent (a quartz glass plug?), or how you'd
>generate the UV pulses reliably. (Some kind of solid state laser is
>probably the only thing reliable enough that I can think of. Are there
>long-life UV flashlamps?)  A drawback or benefit is that the charge would
>no longer be ignited from a single point, but rather a whole region would
>be ignited. You could control this somewhat by altering how the light
>shines through the port in the head. Would it be good or bad to have the
>whole charge ignite at once? 
I guess igniting a region instead of a point will be like having several
plugs - you'll get a faster burn and that is usually an advantage. I
imagine igniting everything at once will be just like detonation - no
good.  You want a flame front that moves through the charge in a short
time, but not zero time.  One way of speeding it up is to start the fire
in several places.

Helge Hafting
-- 
-----------------------------------------------------------
helge.hafting at daldata.no
-----------------------------------------------------------