overlap

[MadBill]

Yes, 8° more, 42° vs. 34°*. (*Recall my previous comment about my version DM under-estimating optimum EVC.)
Here's the graphs at 6750 & 8000 RPM. Note the exhaust dimensions are just roughed in, not optimized. We may be able to prevail on Erland Cox to present similar data from the DM-5 program for comparison.

Overlap Graphs 014.JPG

Overlap Graphs 015.JPG

[nitro2]

Bill you mentioned it being a high effort 604. Back on Pg. 18 for the optimal setup, 1051 ft.lb for 604 ci gives a specific TQ of 1.74 at 6750. That's pretty "high effort" in light of previous discussions on how much specific TQ top level engines make. The pressure traces as shown, don't really correspond to that high of a specific TQ, and the exhaust trace and exhaust stroke cylinder pressure trace show there's more on the table from the exhaust stroke alone (even at 6750).

About how much peak TQ are you actually expecting from this engine on the dyno ?

[MadBill]

I did mention that (at least my version of) DM seems at least 5% optimistic re torque predictions. As far as pressure traces Vs. predicted torque, I would assume the biggest disconnect between DM and reality would be re the data going into the B.M.E.P. calculation rather than the resultant predictions, but....
I'll have to go back a ways in my (chaotic) files to find sims of the engine as last dynoed. Will let you know what I turn up.
The builder is hoping for ~ 1300 HP and 950 lb-ft. in the next configuration. Of course, when there are no program inputs for fuel composition, chamber shape, friction factors, crankcase vacuum, etc., actual projected numbers are of less value than directional trends.
More to come...

[nitro2]

Sorry, forgot about DM reading high. I guess it's trying to simulate most dynos (they read high too ).

I think most sim soft is fairly decent at directional trends, at least up to a point, though an experienced engine builder is also fairly decent at knowing directional trends even without using sim soft. So in a way there is no magic bullet with sim soft, but a lot of convenience, and some fun.

The magic bullet part comes in by knowing the details of the processes, accurately. Looking at the cylinder pressure over the exhaust stroke on the DM traces if that was to be considered accurate or even fairly close to accurate, then there is an easy 25 to 50 HP sitting right there by correcting that. If it is not accurate or fairly close to accurate then there is no 25 to 50 HP sitting right there.

That's where the details come into play as opposed to just directional trends, if the exhaust stroke pressures are accurate you do a little work (on the actual engine) and you get a big , but if the exhaust stroke pressures as simulated are fantasy well you do a lot of work (on the actual engine) and only get this and this The other option is to do nothing and get nothing.

I'll have to get my extra strong glasses out and look at the overlap part next.

[302ford]

If you look into the chamber a few degrees before TDC you would see three distinct pressures. Negative pressure in the exhaust system, ambient pressure in the combustion chamber, and slightly higher ambient pressure behind the intake valve. Overlap allows these three pressures to work in unison switching places with changing rpm's to fill and evacuate the cylinder.
The key is controlling and timing them.

[nitro2]

If you look into the chamber a few degrees before TDC you would see three distinct pressures. Negative pressure in the exhaust system, ambient pressure in the combustion chamber, and slightly higher ambient pressure behind the intake valve. Overlap allows these three pressures to work in unison switching places with changing rpm's to fill and evacuate the cylinder.
The key is controlling and timing them.

True, but there are actually 2 keys, the second is controlling and timing the pressures, the first is understanding them. It's not just a simple as it looks.

A simple example of how one can be tricked by looking at dynamic pressures using the "understanding" which is suitable for steady state flow (such as a flowbench). Imagine a pipe a few feet long and a pressure sensor in the middle of the pipe. You only see the sensor reading, you don't know anything else but what the sensor says. The sensor initially reads atmospheric pressure, but then a pressure pulse, say 3 psi above atmospheric pressure, is applied at both ends at the same time (but you don't know this, all you know is the sensor reading), so in a very short span of time you see 6 psi above atmospheric pressure on your sensor. Naturally you think there is some tremendous flow going on because you see your sensor reading go from atmospheric pressure to 6 psi very quickly.

In the pipe above, if the pressure pulse was 6 psi and had been started at just one end of the pipe, the sensor reading of 6 psi would rightly suggest a lot of flow. However, when there is a 3 psi pulse coming from each direction generating the 6 psi reading at the sensor, there is actually no flow happening at your sensor.

It's necessary to keep track of what is going on, not just take the sensor reading at "face value". The same applies to intake and exhaust systems, since pressure pulses have several sources and they interact to generate a "reading". The "face value" of the reading is not the whole story.

[MadBill]

So then Clint, can we expect shortly to see some kind of bi-directional laser anemometry incorporated into some of your pressure sensors?

[John Wallace]

Clint, would the pressures be additive like that?
( would think 3 psi on one end and 3 psi on other would be 3 total?)

[MadBill]

Remember this example is two pressure pulses in phase, so when the meet, they add together...

[nitro2]

Clint, would the pressures be additive like that?
( would think 3 psi on one end and 3 psi on other would be 3 total?)

Wave pressures add.

[nitro2]

So then Clint, can we expect shortly to see some kind of bi-directional laser anemometry incorporated into some of your pressure sensors?

We'll get right on that.

Usually (but not always), it is easy enough to figure out what's going on without anything beyond the pressure sensor and a bit of thought. In cases when it is not so easy, then repeating the test a couple times, but with the sensor at different locations, is generally enough to spill the beans.