Hmm I guess I did word that with a fair bit contradiction...
I was trying to solely focus on mixture-ratio since that is the main
purpose of a O2 sensor on a car.
To me a purely open loop control in this situation would be using only the
position of the valves and a lookup table of flow rates to control mixture
ratio.
Adding in data from onboard pressure sensors on the injectors, or onboard
flowmeters even to provide feedback on mass-flow rates would be most basic
form of
closed loop control for mixture that I can think of.
I mentioned further characterizing of parameters on-top of that basic
closed loop system if more accuracy was needed. This additional open loop
bits is only
possible on rocket engines opposed to car engines, since in car engines you
do not know your inputs well enough to make those assumptions.
I am kinda curious what more could be practically be done to close the loop
on mixture ratio.
On Wed, Dec 9, 2015 at 7:01 PM, Ben Brockert <wikkit@xxxxxxxxx> wrote:
On Wed, Dec 9, 2015 at 10:12 PM, Wyatt Rehder <wyatt.rehder@xxxxxxxxx>
wrote:
You can characterize your mixture ratio pretty accurately on the ground,and
have a pretty solid expectation that it is going to be the same inflight.
Main reason this works is again you have a high level of control overyour
propellants.
So liquid engines have been doing closed loop control for quite awhile.
These statements are contradictory. Expecting high levels of
characterization from ground testing to be sufficient knowledge for
control is the basis for open loop control. Measuring and responding
to what's going on in flight is closed loop control.
Closed loop control over every variable of significance is required
for high performance. For a liquid rocket, those variables are, at
minimum, thrust, attitude, and mixture ratio.
Using car oxygen sensors in a rocket engine is still an obviously bad
idea though.
Ben