New member intro
tsakiris at ed8200.ped.pto.ford.com
tsakiris at ed8200.ped.pto.ford.com
Tue Jan 10 13:57:17 GMT 1995
>I'm curious, Tony. When it comes to solving manifold problems,
>is everything modeled (i.e., boundary value problem with a
>pressure field...full blown numerical models) or are the basic
>properties/dynamics modeled empirically/heuristically?
>(what's the state-of-the-art?)
>
> John S Gwynne
> Gwynne.1 at osu.edu
It's somewhere in the middle. For some topics, theoretically derived models
are used, for others purely empirically derived approachs are used.
In my experience, the theoretical models are based on physical principles
but are very, very simple (conservation of mass, ideal gas law, etc.) I
think the simplicity stems from the time constraints. There is so much to
do. Making 4 million of something a year is much different than making 4000
a year. It can be very frustrating at times.
The majority of topics are dealt with empirically. There's a large
development community working in cars everyday, collecting data and monitoring
behavior. These observations get boiled down to algorithms. The reliance on
empirical methods is both (corporate) culture and resource driven. Real
measurements are often difficult to obtain. For example, try measuring the
air actually inducted into a cylinder, not just in a research lab, but on 20
different types of test vehicles (Escorts to F-250 trucks, 1.3 liter I-4s
to 7.5 liter V-8s) in the field.
In the past, I worked at a large supplier's R & D center. Now, I work in a
production environment, not in a plant, but dealing with daily engineering
problems for designs one to five years out. Very different worlds. Perhaps
the greatest challenge is merging the two and ensuring the efficient and
accurate transfer of information from one to the other (in both directions).
I don't want to go off on a topic outside the scope of the list. I'd like
to talk about it more, but I'll stop now to see how the above is taken by
the group. Maybe we can continue the discussion later.
Tony
More information about the Diy_efi
mailing list