Jeff,
This is a wonderful example - thank you for the link to the Yanmar data.
If you know what the cause is, the results outside th published points
can be very closely approximated.

Let's start by looking at the torque curve only:
First - notice the surpressed zero of the plots.
Second - notice that if you were to plot this data with zero-zero at the
lower left, you would notice the the decline in the torque curve from
2400 to 1800 wa less than 5% and probably changes less than that to 1200.
Third - notice that the torque and power curves from 24-3600 are not
far off linear.

Explanation:

Diesel engines have two factors at work (no carburettor curves to deal
with here).
-First is the fuel delivery as provided by a mechanical injection pump
system. At a given rack (fuel control input - not governor or throttle)
the fuel delivered at each injection is substantially identical
regardless of crankshaft speed.
-Second is a feature of all internal combustion engines called
volumetric efficiency [VE] (this is how much air is pumped through the
engine as a function of the swept volume [aka displacement])

Simple rules: (unfortunately 100% true)
-Fuel is torque unless you run out of air (that starts happening at the
torque peak - cams are usually optimised for speeds more to clients
requirements). Maximum fuel delivery is usually determined by the
mechanical capability of the engine components.
-Air is horsepower. As available air declines because of loss of VE
with increasing crank speed, that effectiveness of combustion is lost to
the point that the engine wiil start to smoke.

Did I lose anybody - speak up - time is short and you are probably not
the only one with a question.

Matt Colie

Jeff wrote:
I appreciate that torque and HP are essentially the same, such that one
curve can be derived from the other. It was a revelation that this
conversion can be derived geometrically! However, is it really true
that given two points you can get the full story? What if the torque
peak and HP peak came close together, say at 2500 rpm; then you would
have little info as to how it behaves at 1200 rpm.

Or, using these curves as an example, the straight line approximation
between the two points may be valid, but it says nothing about what
happens outside this range.
http://www.yanmarmarine.com/products...5_TechData.pdf

I'm not trying to debate, I really want to know how to make use of these
curves.

jeff

Matt Colie wrote:

Skip,
As an engine professional for many decades I can tell you that all you
need is the two published numbers I hope are in the engines
specifications.
That would be the torque peak and speed and the horsepower peak and
speed.
For a naturally aspirated diesel this is very close to what a dyno
test would give you.
Turn the HP back into torque (rev*tor/5252=hp)_ and plot it out as two
straight lines (it isn't, but it won't matter much). If I can find it
and you are interested, I could try to post the one page explaination
of this.
Ta Da - power curve

Matt Colie - still trying to get out of Detroit and back to tidal
water after 30+ years.