I'm working on some Ford engines these days so I'm paying more attention to Kaase than I have before. Kaase doesn't put a lot of value on flow bench information because of the fact that a lot of heads with great flow numbers don't make power. He has some great examples of why that is on his website and in his interviews. Has a lot to do with the direction that the airflow is pointed.

JWB: That is very interesting. One of the data channels i monitor is engine vacuum. Monitoring vacuum over a run allows me to see if we are developing upper end issues as well as others.
This seems much better , could be really useful with tuning, monitoring .
Hmmm, wonder what is required in order to get this data .

The other concept to try and understand in a running engine is that air has mass, it weighs something and anything with mass has inertia. So when the valve opens air begins to move it wants to keep moving and when the valve closes the air piles up behind it. this is main cause of the pressure waves in an engine. Controlling these pressure waves and timing them correctly is how race engines obtain more than 100% volumetric efficiency. Lot of mechanics are using vacuum transducers with a scope to diagnose engines with these pulses. You can even pickup out of time camshafts if you are good. The red lines are when the injectors fire, the black lines are the pulses in the intake manifold. Notice how even the vacuum pulses are in a good running engine. If the engine has problems the vacuum wave forms will be disturbed. Older vehicles you would just take a compression test but newer cars sometimes this takes a lot of work. You can hook up a transducer in the intake or tailpipe in just a couple minutes.

The crankshaft rotational point and the amplitude of maximum vacuum from piston motion varies with:
1. connecting rod to stroke ratio
2. compression ratio
3, buffer volume upstream of the intake valve including intake port, manifold runner, plenum, etc.
4. buffer volume of exhaust gas remaining in the chamber from overlap
5. overlap triangle area (Vizard: the area is roughly proportionate to the square of the nominal duration)

Are these factors quantified, and blended into the prediction as contributors to spikes?

The 557 thread and discussion about airflow had me thinking about something we discussed on this thread many years ago.

To me Airflow cfm on a flowbench alone is a great tool when you think of CFM...but really in that it's simply a rate of flow in one direction at a standard pressure Of course its an effective tool for measuring total flow comparing one head relative to another....but it's really so much more than that.

Bench flow doesn't take into account (for example) the relative pressure (positive or negative) in the manifold at any given moment in a running engine or the effects of reversion or overlap and (also) the effect of the volume of the incoming charge REDUCING (pushing back) the instant the piston starts to move up from BDC on the intake stroke. This is why the mass and velocity of the intake charge "ramming" into the cylinder can effectively cause a supercharging effect.

More on that some other time..... but I acquired a lot of insight from a conversation with David Vizard about this.

But I'm sure a lot of Engine gurus have software programs and databases that simulate some if not all of these conditions, probably more from trial and error or for tuning and cam and manifold selection for a specific rpm range or racing condition.


For Example If the engine has 240 degrees of actual "off the seat" (lets call it @.050) duration, each valve is open almost exactly (720/240) = 20 seconds for each minute the engine is running...and that is (if you think about it) regardless of RPM. It also brings into perspective where in the rpm range you want to maximize efficiency (torque or 'packing the holes').

If it's say 270 degrees, the valve is open 22.5 seconds per minute. So in terms of port flow, the valve spends 1/2 of it's total duration only open between 1/3 open and 2/3 of peak. The rest is the ramp up, the ramp down and the 2/3 to peak and back to 2/3.

This is to me a good illustration of how little time there is to fill the hole....and that each intake valve is closed much more than 1/2 the time.

But if you examine the flow and you extrapolate the info here as a means of comparing heads.....it's an interesting way to look at it when you think about where the valve/airflow potential is relative to where the piston is in it's cycle; so a head that flows more at peak might not be as good as the head that flows a bit better between the 1/3 and 2/3 when you consider the amount of actual time that lift is filling (or emptying) the hole. And filling is relative to piston position/compression, intake charge, etc. I think guys who dominate stock and super stock classes understand this (or at least a particular head) better than most of us.

I recently did a lot of reading about the 1930's Auto Union and Mercedes Grand prix engine development which I found fascinating, I guess all this was in my head a while and the 557 thread got me thinking about it more.

I wrote this just in hope of inspiring some deeper thought about power production, I've always been more of a street strip type builder which in a way is more like a road race engine than a pure drag motor that is never below WOT and hopefully never much below 90% ve/torque peak.

Let the games begin....

How would one use this information to work with fuel injector timing? My car has a big solid roller cam with lots of overlap. I would like to be able to optimize the fuel injector timing to prevent fuel blowback in the manifold at low RPM. I have struggled with a good way to wrap my head around the numbers I would need to key in. Moving injector timing around doesn't seem to do much for how the engine runs at low RPM. But maybe using map sensor data and cam specs somebody could figure out when to fire the injector.

My car has port EFI but the manifold is dirty and fuel stained in the plenum like it had a carb. Presumably from the fuel blowback from overlap at low RPM. I would think some efficiency and maybe some power could be gained by optimizing injector timing.

How would one use this information to work with fuel injector timing? My car has a big solid roller cam with lots of overlap. I would like to be able to optimize the fuel injector timing to prevent fuel blowback in the manifold at low RPM. I have struggled with a good way to wrap my head around the numbers I would need to key in. Moving injector timing around doesn't seem to do much for how the engine runs at low RPM. But maybe using map sensor data and cam specs somebody could figure out when to fire the injector.

My car has port EFI but the manifold is dirty and fuel stained in the plenum like it had a carb. Presumably from the fuel blowback from overlap at low RPM. I would think some efficiency and maybe some power could be gained by optimizing injector timing.