More EM nerding

[QSHIPQ@aol.com]

On Sun, 14 Jun 1998 graydon wrote:

>>Each of those sections in the triple fin are much smaller than the 
>>ports feeding the EM.  I believe that if Audi did their homework on this 
>>EM then only one pulse at a time should travel through the collector.  
Maybe if you are designing or desire a constant pressure system.  This is a
pulse system.  Given a pulse system, there are a minimum of 12 pulse loss
functions per collector (x minimum of 5 in the audi manifold).  That's a lot
of homework, and I do believe that audi did their homework.  Measuring the
runner v the collector is only 2 of 3 variables.  A more accurate measure for
pulse converter theory is to measure the runner, the collector and the nozzle.

>>However, even if they did this correctly (and I doubt it - I suspect that 
>>other compromises were made for packaging here) the 3 little pie shaped 
>>outlets are too small.  They are little more than half the size of the EM 
>>inlet ports an dthere is only threeof them where we have 5 inlet ports.  
>>That's too small.   That would require that the gas 
>>increase its velocity almost twice in order to get through without heat
>>losses.   Since you never get anything for nothing, then we have a throttle 
>>here.  That causes an energy loss before the hot gas gets to the turbo.

Not with you here.  As a general rule short exhaust manifolds you go after the
least heat loss and maximum pressure pulse.  In a Pulse Converter, Area Ratio
= Anozzle/Aport.  .65 - .85 is considered optimum. Beyond that you get mixing
losses.    You go for uniform exhaust flow with equal length runners, you
offset those gains with more heat loss and not necessarily higher pressure
pulses.  Those 12 variables take a lot of design and fabrication to make sure
you go forward.  Given the amount of possible increase here, a very accurate
dyno test facility would need to accompany your talents.  And none of the
science from audi has been shared, and I don't expect it to.  And in EM design
(yes even today), the science and the art are about equal, hence the dyno.
Remember, the largest value of energy in a pressure wave is pressure, and the
smallest value is kenetic movement that is affected by pipe friction.
Maintain the velocity with heat.  Get the pressure wave, at the highest heat,
TO the turbine.  Shorties are really good for that.

>Just to get some definitions straight here, velocity is not equal to
>energy. Heat = energy.  Velocity is a property of the gas as it passes 
>through the exhaust manifold.  Just to recap our thermo here...  For a 
>subsonic flow (I haven't calculated the mach number, but I'm pretty sure 
>that we are dealing with subsonic flow here, at least in the EM)...

No, the intitial gas velocity at the exhaust valve opening is sonic.  Pressure
is the energy here.  Sorry if I indicated otherwise.


>At first glance this appears counterintuitive, but remember a converging 
>duct is basically a venturi and we all remember from our carb class that 
>the venturi generates low pressure which draws fuel into the throat of 
>the carburettor.
That would be the "nozzle" that you missed above.  

> volume.  You increase volume, you decrease velocity.  
>>Correct.  In a diverging duct (again with subsonic flow)

Not sure your definition of a diverging duct.  In pulse converter theory, one
speakes of diffusers and collectors (which is the pulse converter).  You get
too much volume in a collector, you get mixing losses.  Too much velocity
(?)...

> You increase heat,
> given volume, you increase velocity.  
>>Correct, I suppose, but we can't increase the amount of heat available.  
>>That's a relative constant.
Me, I'd want to make sure any heat from combustion isn't absorbed by some big
heat sink made of aluminum.  I can think of a few ways that the amount of heat
available could be increased if that's the goal.

> Can you improve what is there?  I
> suppose you could, buying the 2 piece exhaust manifold gets you the most.
But
> 'restriction' isn't necessarily a bad thing.  Betcha higher gains could be
>>If the "restriction" were a perfect converging/diverging nozzle, we would 
>>probably have a nozzle efficiency in the region of 90% to 99%, but even 
>>that means a loss, so why leave it there?   If the nozzle were close to 
>>99%, then you might be right, look elsewhere for bigger bang for the buck.

Doesn't need to be for the design of the component.  For a pulse converter,
way less than 99% is considered very efficient, better than 50% considered
good.  A short manifold has the advantage here of minimal energy loss.  You
start porting that manifold you start getting mixing losses at the collector. 

> you.  But then, you have increased runner length which means heat becomes a
>>I don't think that a runner length in the region of one or two feet will 
>>be an issue when you see 4 - 6 foot runners on all out racing engines all 
>>the time.  Check out Bell's book for a few pix.
All out racing engines are looking for all out gains.  For a street car, you
are offsetting those gains with a very efficient heat sink, a cast iron
manifold that is very short.  Assuming you to design the manifold correctly
(getting all the 12 pulse functions optimally), your gains are still minimal.
Specifically, for what the audi manifold IS, it's extremely efficient.  It
maintains high heat for high energy, it retains that heat over a long period
of varying operating time and conditions, and runner length maintains maximum
pressure pulses.  All good things.  You design an EM, unless it's cast, you
have a couple of losses that most likely offset your equal flow at the
turbine.  


>I don't have much faith in the Audi engineers any more than the Cadillac 
>engineers that I work with.  Most modern autos are designed with far too 
>many compromises.  There's rocket science in our labs, but it just 
>doesn't make it to the street.  :-(

I do think that a lot of time was spent on that EM.  A lot of time.  And a lot
of time on the IM as well (tho many years ago at vw).  That casting is
difficult at best for any piece of iron.  But, I've also spent a LOT of time,
evaluating whether to tackle the 'art' of EM design.  Not worth the effort,
IMO.  IF you want to do it, I encourage you to totally understand why audi did
what they did.  At 700hp race car, I might look at relocating the turbo, and
taking a crack (pun intended) at the EM design.  But the one that is there,
certainly can handle anything that will take 93 octane.

A very interesting and lofty discussion Graydon.  Bottom line:  You have more
physics background than I.  For 5 years tho, I've looked hard at the EM.  And
tho this may seem simple, I know it's not.  Cuz reducing reflection by equal
length runners, is offest by the temperature losses incurred.  My own
thinking, is that given the HP levels we are talking about here, the whole
project is at best, a wash.  Without proprietary science from audi, and the
ready available dyno to ck the 'art', a bunch of time on this seems wasting
good 'energy'.  To me, the best think one can do is NOT match the EM to the
head, to create a slight antireversion step.  Beyond that, you are beyond the
basics, and ANYTHING you do, needs to have a dyno associated with it.  Cuz
more or less of anything in that manifold, can easily take you backwards
quickly.  

Respect the artiste for what they did, question why, suppose the reasons.  But
bottom line, best left alone in favour of simpler and more proven tweeks.
Maybe another look at the downpipe for instance...   :)  (insert double
meaning)

My .05

Scott Justusson