bob@bobthecomputerguy.com (Robert Harris)
fangle  10 Jan 1999
- Virtually all of the complication of the early mechanical fuel injection 
  systems involves changing the fuel flow to follow the air flow.  This was 
  and continues to be a result of the carburetor induced mindset that to 
  respond to a power demand change, you must first change airflow to induce 
  the new fuel flow. 
- The diesel fuel system has always been a model of simplicity, but it is 
  probably too much of a mind stretch to use as the model.  But far closer to 
  home, racers have achieved excellent results with Hillborn derived constant 
  flow fuel injection systems.  It consists of a pump, a throttle valve, and a 
  couple of bleeds.   The fuel is volumetrically adjusted to the engine size. 
  During operation it is totally insensitive to air flow - a perfect Al Bundy 
  "Don't Know - Don't Care".  It is still in the top of the heap for shear 
  power and reliability - its major fault is over rich part throttle, but 
  someone invented a dejection system and is putting it in somekind of beat up 
  old furren kar - that's supposed to take care of that. 
- After coming over to the Dark Side, you realize that all the driver wants to 
  do is control the POWER by the throttle - not the air.  Since power is first 
  order related to the amount of fuel burned and second order related to the 
  amount of air, its backasswards silly to control the air and let the fuel 
  catch up.  Control the fuel and let the air catch up. 
- This mind reversal now reduces the fuel control to RPM vs Throttle Position 
  -easily done by a "stolen" diesel system or simple injectors where the 
  firing rate is controlled by rpm and the pulse width by the throttle.  The 
  rest is all air control. 
- As an exercise for the student, go rethink the fuel air mixture tables vs 
  power stuff.  BUT - keep in mind the tables are for Air Constant - Fuel 
  varying.  Re plot them holding fuel constant - air varying.  A very 
  different picture occurs.   Suffice to say that thru reasonable regions, 
  power will remain approximately constant and directly proportional to fuel 
  with much less effect of air changes.   Remember once more - the existing 
  table gives you power with the air held constant.  Rethink them with the 
  fuel held constant. 
- Other than greatly simplifying the fuel side of the equation, creating a 
  system where the natural dynamics go the way you want them, and being about 
  as contraryian as you can get - why come over to the Dark Side? 
- Simplicity.  Beautiful simplicity.  It is far simpler to coarsely meter the 
  air and finely correct it than the fuel.  Simple stepper motors can move air 
  valves about.  With the fuel metering fixed by the foot - fine steps of fuel 
  control are not needed to correct for whatever.  And, since absolute 
  accuracy is not needed for the air, much less sensitivity to radical cams. 
- Take for example, EGO feedback.  In the conventional scenario, the air leads 
  the fuel and finally somewhere after a bunch of time and changes, the 
  exhaust reflects the new mixture.  Ah duhh - how much time ( and instability 
  ) has lapsed since the induced change and the feedback.  Not real optimum 
  here. 
- The Dark Side - make the change - engine dynamics our way - EGO says "Too 
  Rich"  - adds air.  Finally stablizes.  Driver changes throttle - " Too 
  rich". The lag from change to sensor all ways results in a higher power 
  safer too rich mixture under acceleration and a cleaner leaner non back fire 
  ing ( damn spell checkers - won't let you create new words ) going down.  
  The EGO correction cannot counter the driver demands - it works with them. 
- With the Air Leading Fuel Lagging method, the driver has to change the air 
  flow.  Think about leaning out a mixture for cruise economy.  Conventional: 
  lean mixture - fuel flow drops, power drops - driver adds more throttle to 
  maintain power balance   ********  Decision Time ******  Does he want more 
  power to do something or is this simply in response to mixture tweaking - I 
  don't know - be safe yuch yuch yuch. 
- Dark side.  Power is set by fuel.  Computer cares less about driver.  Says 
  can I move leaner - **** Lean Misfire **** No -- I can add more air.   Yes -
  --Oops back off the air.   Driver assumes Al Bundy mode "Don't Know - Don't 
  Care" because the foot feed position and effective power is unchanged. 
- Welcome to my nightmare -  the Dark Side now returns you to your regularly 
  scheduled programming.  MTBE is good.   Klean Air is good.  Global Warming 
  exits.   The manufacturer knows best.  Buy Mushie Tushies - Pre prepared, 
  pre packaged, pre cooked, pre digested - Just open the package and flush it 
  down the toilet. 

bob@bobthecomputerguy.com (Robert Harris)
fangle  11 Jan 1999
- The concept of Anti-C is simply to break the domination of carb based 
  thinking and to computer control the air rather than the fuel.  The fuel is 
  arbitrarily set by the drivers throttle.  How simple can it get?  Watch. 
- Lets do a Dark Side 4 cylinder.  Acquire a Bosch KE fuel distributer and 
  associated fuel lines/injector.  Air body not desired.   In conventional 
  use, the plunger is driven by the air body in response to air flow.  Not us, 
  we'll violate its design and use a simple cam that the driver moves back and 
  forth. Thus the cam mechanically moves the plunger.  The cam will be mounted 
  on the primary air valve - also foot driven.  As the air flow profile across 
  a throttle blade is well known, the cam for the plunger can be cut to match 
  this profile. 
- The result is the no-brainer approximate matching of fuel percentage to air 
  flow percentage ignoring RPM.   To match for RPM, we pervert the Bosch 
  pressure actuator control.  Originally it was used to control mixture by 
  varying the flow through the lower chambers in response to a bazillion 
  electronic inputs.  The Dark Side uses it simply to compensate for RPM. 
- The warping plate uses a DC current of 0 to 120 milliampere to control its 
  position.  We'll simply call upon our friendly neighborhood electronic 
  wizard ( when he ain't whacked out ) to whip us up a couple of pots and two 
  or three transistors, re-invent the tachometer circuit and be done with it. 
- We have critical mass.   Throttle matching air flow approximately.  Fuel 
  matching rpm.   Just re-invented the Hilborn.  Tune accordingly for WOT and 
  the upper throttle regions. 
- We could get wet playing conventional C games at this point to makee nicee 
  the lower and mid ranges - or stay on the Dark Side.  Lets stay heretical. 
- What we need now is the ability to add/subtract some air to the system to 
  facilitate the lower mid range.  Remember, the power is set by the fuel and 
  the air roughly set to that level.  Enter the Ford Idle Air Controller.  Its 
  a solenoid driven thingee that by varying the pulse width, varies the air 
  flow past it.One or two of these off a truck motor ought to nicely increase 
  and control the air flow in the part throttle region.  Could in fact, be 
  driven as a Lamda Valve ( K-Lamda) is - using the same box.  One junk yard 
  box -unmodified and now the Dark Side tunes the engine using EGO and air. 
- Major myth concerning the EGO.  Its Sloooowwww - NOT.  Vizard claims it can 
  detect individual cylinder events at 6000 rpm.   What makes it appear slow 
  is that in a STOCK application using CATS, the mixture must be swung about 
  Stoic - both richer and leaner at a 1 to 3 times a second rate to make the 
  cats work.  Get some KLEAN AIR NAZI to explain why, but this slow swing has 
  absolutely nothing to do with the EGO sensor's response.  Since Komrade 
  Klinton does not write my paycheck, I am under no obligation to follow his 
  Smog Nazi's edicts and can thus use the EGO like God and Bosch intended and 
  implemented in the K-Lamdas - virtually instantaneous feedback. 
- What I've just described is a nicely compensated Dark Side constant flow 
  system that starts and idles like a K-Lambda and turns into a Hilborn at 
  middle rpm/throttle.   Air Flow measuring - we don't need no steeeenking Air 
  Flow measuring.   We can run any cam we want.  And yes, you could simply add 
  more control to the fuel and not futz with the air - but you forfeit all 
  that we think the Dark Side gains.  Each path to its own. 
- Last side note.  Computer authority over air flow is relevant only from idle 
  to mid throttle.  At some point, all systems revert to grossly metering fuel 
  to air and counting on tables or whatever.  This system and the Dark Side in 
  general, allows a smoother transition IMHO and a much simpler one. 


Low Fangled Auto Tech
---------------------

Date: Tue, 12 Jan 1999 07:25:41 GMT
From: bob@bobthecomputerguy.com (Robert Harris)
Subject: The Anti-C  -  Lean machine

For those who did the exercise for the students will have noticed that power
does not drop off as fast as economy increases.  Going lean definitely has
advantages for economy.

Using the normal control algorithm where the driver controls air and the
computer controls fuel requires that the computer reduce fuel and thus
indirectly power to lean the mixture out.  At some early point, the driver
begins to notice the slowing this induces.  Drivers response is to increase
the air flow.  Brain Phart follows: Is this to correct for leaning or is it to
respond to road conditions.  No choice - respond as if to road conditions.

Now, from the Dark Side prospective.  Remember driver controls fuel, computer
controls air.  When allowed to do so, the bot begins to lean the mixture.
Lean misfire is the lean limit.  As the mixture leans past stoic, the power
increases slightly - the result of more efficient combustion.  Vehicle speed
tends upward, driver responds by reducing fuel, the hunt for misfire
continues.

Note for the Dark Side computer, there is no decision involved with the
drivers reaction.  All mixture changes make power changes - but, in this case
like with acceleration and de-acceleration, the driver induced change and the
computer change compliment each other and work with each other.

Thus, the Dark Side system can easily transition into lean burn and under near
steady state conditions, can lean the engine to just off lean misfire and
maximize economy without special programming.

The complication of going lean is that the flame speed starts to slow rapidly.
To compensate, ever increasing spark advance or other actions are needed to
compensate.  The Dark Side cries out for a reasonable implementation of the
Saab Trionic knock detect system.

One alternative to increasing spark advance is to INCREASE air temp when under
leaning conditions.  This speeds the flame front, which results in making less
negative power, and reduces the thermal loss's after ignition.   Suppose one
were to take a small amount of air, pass it thru a heat exchanger with the
exhaust and introduce it back as the enleanment air.  Now you would be
blending say 500 degree "superheat" air along with the normal feed air.  The
total air temp would rise significantly - giving greater volumetric efficiency
( hotter air - better apparent cylinder filling - less theoretical power
because of density - but we are not in that section of the power curve ),
mixture quality, efficiency and some part throttle recovery of "waste heat".

Since the Dark Side controls the air, the superheated air could be staged in
or out as needed without the driver leaving the Al Bundy mode "Don't Know -
Don't Care".




Low Fangled Auto Tech
---------------------

Date: Tue, 12 Jan 1999 08:39:46 GMT
From: bob@bobthecomputerguy.com (Robert Harris)
Subject: The Anti-C  -  EGR

Lets deal with the concept of span of control.   During enleanment, we are
adding air to the system outside of the drivers control.  Since this is
occurring at part throttle, light load, it does not take a lot to accomplish
this.   One strategy is to simply let the enleanment circuit/valve become
swamped and just become part of the WOT Hilborn simulation.  Its always there,
it just no longer can effect the system.

Now on to eating our own dung.   EGR is an unlovely, unloved thing - something
those who are enamored with the teachings of the Saints Carnot and Otto never
seem to understand.   For most (***** Not ALL ******) .  However, to the
initiated in the secret inner Klaven of the Dark Side, EGR is a lovely thing -
a secret weapon in the war of fangling.

The conventional view of EGR is that it is simply a hot, inert, dilutant gas.
This may be true from the thermodynamic point of view, but from the Dark Side,
this miss's many good things.

First, the components are far from inert.   Primary is Nitrogen - airborne
~79%, Carbon Dioxide and Carbon Monoxide - from the combustion completion of
the carbon portion of the hydrocarbon fuel, and WATER - from the combustion of
the hydrogen portion of the hydrocarbon fuel.   This water is produced at the
rate of approximately two gallons of water for every gallon of fuel.

Put this another way, every percent of EGR results isn two percent of water
injection.  Adding EGR adds water injection.   Can't think any other way when
twice as much water is present in the exhaust as fuel was consumed - unless
water in the exhaust recycled into the intake suddenly magically carnotly
changes into non water.

Next, when lean of stoic, EGR is a good source of superheated air - just the
thing to speed up combustion - well maybe.  Fact is that most of the heat
contained in the EGR and the superheated oxygen contribute to increasing the
flame speed, whilst the water and the nitrogen dilutant do the opposite.

Heywood, among other authors, reports that the maximum practical limit for a
reasonable engine is about 25% EGR - which quite happily co-incides with
Ricardo's 50% maximum water injection.  At part throttle cruise conditions,
the addition of massive EGR is a good thing for economy without detonation.

For the truly suspicious, you might want to run out what happens when you
raise the intake air temperature and add massive water injection.  Just don't
tell Smokey.  For our purposes, all Dave and I care about is increasing the
part throttle economy radically.

Low Fangled Auto Tech
---------------------

Date: Tue, 12 Jan 1999 16:28:34 GMT
From: bob@bobthecomputerguy.com (Robert Harris)
Subject: The Anti-C  -  Some Lengthy Unification Theory

The following will sadly lack in another C - Calculus, so for the
mathematically over-endowed, please bear with me.

The Carnot Cycle's major purpose (IMHO) is to teach engineering students in a
semester what Hienlien taught us in a sentence  -  TANSTAAFL  ( There ain't no
such thing as a free lunch  ---  for the SF challenged ).  Summed up, in a
closed system, the amount of energy available for work is the peak energy at
the top of the cycle minus the residual at the bottom - that's all there is.
Buried into the thinking relating to that is the standard educational caveat -
all other things being equal.   Change the caveat - confuse the poor Carnotist
into blatherdom.   The lean, The EGR, and The water ALL change the other
things being equal.

Quickly reviewing the IC process.  Fuel is mixed with air, introduced into a
chamber, compressed and then combusted.   So far, chemical energy is converted
to heat.  This energy is used to expand a mass of gas, resulting in pressure
against the piston, which blissfully unaware of the laws of thermodynamics,
follows the laws of physics and moves the way the pressure wants it to go.

Chemical energy to thermal energy to mechanical energy.  The whole purpose of
the chemical and thermal exercise is to put pressure on the piston.

The maximum thermal energy of a fuel is released at stoic - you have burned it
all.  This results in a given mass of gas at a given volume depending on
temperature.    This translates to a certain pressure based on mass, volume
and temperature.  Call this the starting baseline.

Adding additional mass of unburned air to the baseline does not cause the
release of any additional thermal energy - stoic burned it all remember?  What
it does do is increase the mass, thereby increasing the volume and the
resultant pressure.  Because the thermal energy is unchanged, the temperature
will fall somewhat.  Practically speaking, we can keep adding air and
increasing efficiency out to a point where the mixture will no longer burn -
Lean Misfire.  Dark Diesels go much past this point.  Note that with the fuel
held constant, power ( the result of pressure - remember? ) increases right
out to about the point where it won't sustains combustion.  Something that
gets ignored.  Thus ends the first violation of the Carnot Caveats - the mass
remains unchanged.

Next, lets look at the stoic/lean composition of the exhaust gases's again.
Nitrogen, Carbon Dioxide, Oxygen and Water.  Now below the temperature of
dis-association, the pressure contribution will be determined by the moles
present and volume of each molecule of each type.   Bigger molecular size,
bigger pressure contribution.  So we change pressure not only by changing
temperature, or the number of molecules present, but by changing the size of
the molecule.

Consider now water injection.  Water adds molecular mass - a diluent just like
additional air.  Has exactly the same effect.  Ricardo documents two
experiments.  One, the fuel held constant, the second power held constant.
The results were simple fuel constant power went up proportionate to the
water.  Power held constant, water replaced fuel.  Worked thru a broad range.
Now if I could only find the rest of my old books without someone glomming
onto them ...   Please note - no magic.  Both involve the addition of mass and
the reduction of temperature that results.  But pressure moves the piston -
not temperature.

So far we have seen that adding mass to the mixture without changing the
chemical energy results in additional **** Part Throttle **** power or
efficiency.  Now consider that water expands much more under temperature than
the other three. In the pressure equation, adding size to a molecule adds
pressure -  Amagat's Law of Partial Volumes and all that.  To the rate that in
a boiler, saturated steam pressure peaks at only ~705 F and the resultant
pressure is ~3206 PSI.  I leave it as an exercise for the student to explain
that why if water makes so much pressure at so low a temperature and is a
major constituent of combustion product, does the peak pressure of an IC
engine seldom exceed 1500 PSI and that at 3 to 4 times the temperature.   We
now bring on the handwaving Carnotist/Ideal Gas apologists for their high
volume,  high decibel, orthodox  performance.

Long pause for mega tirades

Now the final part to consider is the temperature.  All devotees of WOT now
that reducing the temperature reduces detonation and increases power because
more fuel can be burned and thus more gas mass to move the piston.  This works
- no problem.  But, not at part throttle.  Peak power is always a trade off of
quantity for quality ( read efficiency ) with quantity beating quality any
day.   But moving away from this region, the hotter the air the better - right
up to detonation.   What we really need is an interheater.  Why so boss?

Start with going lean or wet results in slowing the flame front.  More advance
needed - more negative power, longer high temperature burn cycle results in
greater heat rejected to the cooling system and lower efficiency than could be
expected.  Adding heat to the air speeds the combustion process back up,
reducing the loss's.   BTW it  LOWERS the theoretical peak efficiency by
reducing the range of the cycle, but since Carnot ain't filling the tank,
screw his followers.  It also increase's the effective volume and reduces the
part throttle pumping loss's.  The higher temperature results in a more
homogeneous mixture and a better burn.  This quickly becomes a trade of
quality for quantity - exactly what we want at part throttle.

Ideally, the part throttle air would consist of say a 500F mixture of air,
fuel, water for detonation controls and maybe passed thru a turbocharger
acting primarily as a turbo driven homogenator and to force this thin mixture
into the cylinders.  Haven't I seen that somewhere before?   Nah ... only a
Dark Sider would reverse enough thought trains to come up with that.