dave.williams@chaos.lrk.ar.us (Dave Williams)
fangle  02 Mar 2000
[dave williams said]
- Thinking about the four stroke cycle this morning while in the shower.
- The pressure differential across the exhaust valve is greatest when the 
  valve first opens.  Blowdown is rapid, then pressure remains relatively 
  constant across the rest of the exhaust stroke, as shown by typical 
  indicator diagrams. 
- I haven't had a chance to plot the pressure drop across lift yet, but I 
  strongly suspect the majority of blowdown occurs before the valve reaches 
  fully open. 
- This leads me to think that low lift flow is critical on the exhaust side, 
  both for blowdown and to get a sharply defined pulse into the exhaust system 
  for tuning. 
- Flow velocity across the low-lift exhaust valve is, according to Taylor, in 
  the "critical speed" zone, where the exhaust gas behaves more like a liquid 
  than a gas.  This indicates conventional low pressure flow bench testing may 
  not be greatly relevant to exhaust flow during the most active portions of 
  the exhaust cycle. 
- I have a 1960s SAE paper on the flow bench development of AMC cylinder 
  heads.  The paper says their bench used 100PSI compressed air instead of 
  vacuum pumps because they had a large supply of easily-available air. Note 
  the later dogleg AMC heads had outstanding exhaust flow and only mediocre 
  intake flow. 
- I've also noticed many modern-design cams have shortened the exhaust timing 
  by comparison to the intake timing.  This reduces overlap, which has strong 
  influences at low and mid range RPM, and looking at it from a pressure 
  standpoint, the cylinder pressure at intake valve opening is likely to be 
  about the same as it would be with a longer exhaust duration. 

"Leone, Thomas (T.G.)"
fangle 02 Mar 2000
[dave said]
- > The pressure differential across the exhaust valve is greatest when
  > the valve first opens.  Blowdown is rapid, then pressure remains
  > relatively constant across the rest of the exhaust stroke, as shown
  > by typical indicator diagrams.
- > I haven't had a chance to plot the pressure drop across lift yet,
  > but I strongly suspect the majority of blowdown occurs before the
  > valve reaches fully open.
- > This leads me to think that low lift flow is critical on the exhaust
  > side, both for blowdown and to get a sharply defined pulse into the
  > exhaust system for tuning.
- > Flow velocity across the low-lift exhaust valve is, according to
  > Taylor, in the "critical speed" zone, where the exhaust gas behaves
  > more like a liquid than a gas.  This indicates conventional low
  > pressure flow bench testing may not be greatly relevant to exhaust
  > flow during the most active portions of the exhaust cycle.
[tom replied]
- You're basically right, but here are a few nit-picks.  Pressure during the 
  middle and end of the exhaust stroke is only constant if there is little 
  exhaust tuning (highly tuned engines show very large amplitude waves).  
  Blowdown occurs mostly at low valve lift at low RPM, but not at high RPM.  
  Since the physics of the blowdown are time-based and the valve opening is 
  crank angle-based, at high RPM the exhaust valve opens significantly before 
  blowdown is complete.  For example, I pulled out some data on a four valve 
  per cylinder engine with mechanical bucket valvetrain, and cylinder pressure 
  is still coming down about one fifth of the way through the exhaust stroke 
  at 6000 RPM (it would be even later on a 2-valve engine).  So flow at higher 
  exhaust lift is very important at high RPM.  Low pressure flow bench results 
  are useful mostly at low RPM. 
- Also, it is not clear to me that a sharply defined exhaust pulse is 
  necessarily a good thing.  It is only good if your valve overlap is arranged 
  to take advantage of the exhaust tuning, which probably means that it will 
  only offer an advantage at one RPM.  A sharper pulse might mean a narrower 
  RPM band where you see a benefit.  The most important part of "tuning" an 
  exhaust system is avoiding a high pressure pulse during valve overlap.  This 
  usually means keeping the pipe length between cylinders long enough to 
  prevent the blowdown of one cylinder from arriving during valve overlap of 
  another.  This does not require sharp pressure pulses or any wave dynamics 
  at all. 
[dave said]
- > I've also noticed many modern-design cams have shortened the exhaust
  > timing by comparison to the intake timing.  This reduces overlap, which
[tom replied]
- Most recent cam design changes are probably driven by emissions rather than 
  wide-open throttle performance.  Meeting the upcoming emissions laws 
  requires very fast catalyst warmup, which most manufacturers are doing with 
  massive spark retard for the first 30 seconds or so after cold start.  Lower 
  valve overlap is critical for good combustion stability under these 
  conditions.  The conventional wisdom is that the lower valve overlap hurts 
  wide-open throttle performance mostly at high RPM, not low RPM (this may or 
  may not be true with highly tuned exhaust). 

William A Williams 
fangle 02 Mar 2000
- You are correct about the criticality of the blowdown phase of the exhaust 
  process and the amount of flow through the valve at very low lifts. Work by 
  Kadenacy in the 20's and 30's showed that if you can open the exhaust 
  rapidly enough the cylinder could be pulled well below atmospheric prior to 
  the intake opening. It proved difficult to implement in practice but several 
  engines were built and developed. See "Scavenging the Two Stroke Engine" by 
  Paul Schweitzer, chapter 13 for details.

dave.williams@chaos.lrk.ar.us (Dave Williams)
fangle  03 Mar 2000
- -> Done properly, a cat pre-turbo might well add some heat to the
  -> exhaust stream (good) and ALSO act as a decent "pressure recovery
  -> accumulator"--thus making the turbine operate more efficiently !
- Robert and I have been discussing the pros and cons of installing a cat in a 
  racing exhaust system.  Done properly it could keep the exhaust velocity up 
  in the back half of the system, affecting tuning. 
- -> According to Vizard--30� valve seats flow significantly better at low
  -> valve lifts. Also--once the absolute pressure ratio from the cylinder
- Yes.  Ford big blocks use 30 degree intake seats, as do Pontiac V8s, and 
  some Olds V8s.  Some *other* Olds V8s use 30 degree seats on the exhaust 
  side instead of the intake side; so far, that's the only application I've 
  found of a 30 degree exhaust seat.  Oddly enough, the "C" casting heads I 
  have for Thunder had 30 degree exhaust seats.  I have already recut the 
  heads for oversize valves with 45 degree seats. Hmm... 
- -> is more properly refered to as "choked" flow than as "critical" flow.
  -> The REAL interesting trick would be to get the seat/bowl/port shape
  -> configured so that the flow in the exhaust port would go supersonic
  -> downstream of the choked point during the blowdown part of the cycle!
- Yes.  Once you toss the "average gas speed" idea and start looking at the 
  sharp pressure curve that the port actually sees, various ideas start to 
  take shape...