opjose
Well-known member
You are talking about "predictive" modeling.
To an extent:
I've taken an existing user created model, added in the physics per weights and dimensions ( This is the Funtana 1.00 up on the swaps ), then compared it against my real world aircraft.
My real world aircraft exhibited some quirks which the Realflight model did not exhibit.
So I adjusted the Realflight model to exhibit the same quirks.
Note: One of the surprising things I found, was that to obtain the same behaviour the Realflight model had to have it's C.G. set further BACK than in my real world plane...
This gave me a "test bed" with good fidelity, so to speak and a known good starting point.
I then used that to attempt predictive changes to my actual plane...
e.g. How it would behave with different engines, battery packs, throws, etc.
Realflight was spot on predicting all of the effects and characteristics produced by the changes...
But you have to remember that I started with a RF model that was adjusted correctly to begin with...
So "Catch 22".
----
Many, many, many moons ago I was asked to do some predictive modeling for satellite look-down angles. The intent was to produce an accurate prediction of what a satellite would see over it's ground track, after a specified number of future orbits, at a specific time and date.
The problem was that in order to get accurate predictions, you had to have an accurate model of the earth's gravity "geode" which changes all the time.
When I check to see how this was done, I found that the gravitational geode, was mapped via satellites...
Duh....
So the predictive modeling of the lookdown, was directly dependant upon a predictive gravitational model... ARRRRGGGH!
In other words, to get an accurate predictive model, well you needed to start with an accurate predictive model...
Catch - 22.
----
For all of those who don't know what Catch - 22 is, well you're just too young to understand any of this anyway!
Read the book and watch the movie, you'll enjoy both.
To an extent:
I've taken an existing user created model, added in the physics per weights and dimensions ( This is the Funtana 1.00 up on the swaps ), then compared it against my real world aircraft.
My real world aircraft exhibited some quirks which the Realflight model did not exhibit.
So I adjusted the Realflight model to exhibit the same quirks.
Note: One of the surprising things I found, was that to obtain the same behaviour the Realflight model had to have it's C.G. set further BACK than in my real world plane...
This gave me a "test bed" with good fidelity, so to speak and a known good starting point.
I then used that to attempt predictive changes to my actual plane...
e.g. How it would behave with different engines, battery packs, throws, etc.
Realflight was spot on predicting all of the effects and characteristics produced by the changes...
But you have to remember that I started with a RF model that was adjusted correctly to begin with...
So "Catch 22".
----
Many, many, many moons ago I was asked to do some predictive modeling for satellite look-down angles. The intent was to produce an accurate prediction of what a satellite would see over it's ground track, after a specified number of future orbits, at a specific time and date.
The problem was that in order to get accurate predictions, you had to have an accurate model of the earth's gravity "geode" which changes all the time.
When I check to see how this was done, I found that the gravitational geode, was mapped via satellites...
Duh....
So the predictive modeling of the lookdown, was directly dependant upon a predictive gravitational model... ARRRRGGGH!
In other words, to get an accurate predictive model, well you needed to start with an accurate predictive model...
Catch - 22.
----
For all of those who don't know what Catch - 22 is, well you're just too young to understand any of this anyway!
Read the book and watch the movie, you'll enjoy both.
Last edited: