Kenku

Picked up an SAE paper the other day; by Mazda, from '90... paper #900032 if you have access to a library with the SAE transactions. It's modeling power output of peripheral port motors, followed by coming up with an "optimum" port timing. What's interesting is that the port timings they come up with are completely different than any other published ones I've yet seen. The "baseline" port shape is about the same as the published values for the MFR ports on Paul Yaw's site, but the "optimized" port is... well, I'll just come up with a little chart.



Baseline Optimized

Exhaust Open 73 BBDC 73 BBDC

Exhaust Closed 65 ATDC 55 ATDC

Exhaust Area 9.2cm^2 12.5cm^2

Intake Open 100 BTDC 80 BTDC

Intake Closed 75 ABDC 80 ABDC

Intake Area 20cm^2 25cm^2

Runner Diameter 43mm 46mm



Note that there's both less port timing and more area... and all the port timing changes made for less overlap. The ports are a lot more rectangular, extending out towards the edges of the rotor housings farther than the more square baseline ones. The model says that 30cm^2 ports will get even *MORE* flow across the 7-9k RPM band, but only by around 1% over the 25. And I have to wonder if there's room to *fit* 30 cm of port area without going the whole width of the housing.



Oh yeah, also note that the port they publish for the "baseline" one opens the intake 14 degrees earlier than the MFR one Yaw has the data for, but is otherwise identical. This might be explained by the fact that all the MFR housings I've seen for sale are supposed to be for earlier motors, so perhaps there's even development work done between the later MFR housings and the ones we're familiar with.



They don't actually publish very detailed dyno results. The "optimized" timing beats the "baseline" port by a good 2-5% volumetric efficiency from 7500 RPM on up, and trails by 1-2% below. The one dyno plot they have is of a 3-rotor motor, and seems to indicate a gain of 30hp at 8500 and 50hp at 9000, while only losing 10hp at 7k or below. Absolute power numbers... aren't on the graph, regrettably. And even the comparison numbers are estimates, as the graph is of a type used to merely show a trend. They *DO* have real test results in it to validate the model, though.



The capper to this is that the time of the paper (and some of the talk about variable length intake tracts on a 4-rotor later on in the paper) seem to indicate that this was either being developed for the R26B LeMans motor or for a follow-on effort... either way, it makes me *REALLY* curious.



Thought some of you might be interested, or maybe some experienced builders would have comments.

j9fd3s

there is a papre on the 787b, is this a different one?

Kenku

Yup, it's a different one. They don't even mention any race cars in this one, I'm just reading between the lines.

j9fd3s

cool, post pics? scan it? transcribe in stone?

Kenku

Been meaning to get around to scanning it, actually... I'll see what I can do with regards to scanning and posting it when I get home from school.