update on tbi setup

[James Boughton]

>I used the following specs:

>engine capacity = 2000 ccs
>number of cylinders = 4
>maximum rpm = 6600 
>tb id = 40mm = 0.04m
>throttle shaft area = 8mm x 40mm

>I also got my hands on an efi plenum for the engine, which came on the 
>later model twin cams, and this has a cross-sectional diameter of 36mm. 
>As the plenum was damaged, and able to split into two parts, i decided i 
>would use the bottom half, which holds the fuel rail and injectors 
>(thereby getting out of having to build one from scratch), and put the 
>throttles on this.

Do you intend to have any ability to adjust length for development?  If not,
I would suggest a length shorter that what you think is correct.  Power can
fall off very dramatically once the runner becomes too long, but tends to
fall off much more gradually if the runner is too short.

>vol of air for each cyl = (500cc/2) * (6600/60) = 27.5 l/s = 0.0275 m^3/min
>area of tb = (area of tb id) - (area of shaft) = 9.366 x 10^-4 m^2

In actuality, the average area of the intake system should be used if you are
going to use the calculations in the SAE papers by Engleman and Eberhard.

>air speed = volume / area = 0.0275 / (9.366 x 10^-4) = 29.36 m/s

Very high!

>I also found that the runner cross-sectional area is about 8.4% larger 
>(average) than the tb effective area, though the throttle shaft has been 
>machined so that there is less exposed metal at wot.

The larger runner area should help with the high airflow rates, but
unfortunately, now you are creating some of the highest air velocities
in region of high turbulence.  This cannot be good.  I worked with an
airflow engineer that would cringe at the process of speeding the air up
and slowing it down in the intake system.  He believed that the process
should be a consistent decrease in area except in the port (which is a
different story.

>It was mentioned in a previous post by James Boughton that a target air 
>speed of 15 m/s is ideal over the intake air event (or so i read, please 
>correct me if i'm wrong). The figure above is about twice this, and 
>assuming the valve is open for 180 degress (half of intake cycle) , is this 
>about right?

The figure used is 60 m/s over the intake cycle which is 4 times the figure you
are using because the intake event is only one of the 4 events of the 4-stroke
process.  This would put your throttle bodies at about twice what I would suggest.
It is really the throttle blade that causes most of the problem, but what can you do?

>What is the reasoning behind the 15m/s figure above? am i trying to keep 
>the air laminar in the intake?

15m/s or 60m/s over the intake event was an empirical number used by
a number of intake engineers over the last decade or so.  This is changing
as engines get better, but since no one understands the mechanisms by
which this phenomenon works it is hard to say why.  I dealt with an engine
that was over 100m/s (I will always quote velocity over the intake event), but
that was through development.  We started out at around 60m/s.  I think
part of the historical increase is due to the improvement in ports.  Our ports
tended to have >0.9 discharge coefficients which I believe helped out a lot.
So if you are going to run high velocities, I would invest in some very good
port work.

>Any ideas and comments most welcome. Thanks.

>		 Bruno. (b.marzano at student.canberra.edu.au)

I might be able to give you my suggestion for length later if I can dig up the
spreadsheet, or the SAE paper.  It has been a while, but I need to do this 
for a friend of mine anyway.

Regards,

Jim Boughton
boughton at bignet.net