--------snip---------
There are two schools of thought on first flights.
Both have merit.

One is to "go for it!". Take off and climb to altitude where you can
become
safely aquatinted with her "personality" safely.

The other is to make several short hops down the runway to get the feel
first.

The latter, at first, scared the dickens out of me - just on principle.
Going from low and slow lift off to low and slow landing *seemed* like a
bad
idea. But in the end, I've come to think this is safer than I originally
thought, and had become my standard approach to testing a new plane.

I like it because 1) we are low and slow and if anything does go wrong, at
least we are low and slow. And 2) we are expecting to "abort" the take off
soon after lift off. We will not have the danger of the engine possibly
quitting on climb out, and the attendant difficulties that presents.

And 3!) it let's you skip the first flight! When you finally are
comfortable
with the plane and take it around the pattern for the first time, it's
really
not the first flight! (how 'bout that for a plan!)

And, frankly, this turned out not to be the pilot challenge that I first
thought it would be. Although YMMV?

-------snip---------

Personally, I agree and plan to use the aborted take-off method as well.
Actually, the plan has had many well known and respected advocates--IIRC,
Molt Taylor was among them.

Also, if the plan is to test a "custom built", or if there is any other
reason to question the weight and balance envelope, I plan to first test a
thrown model--prior to investing time in actual construction of a "real"
airplane. I would first re-read all of part 23 to glean any insight to
accumulated experience in defining the balance envelope. (I know, I really
have no intention to follow everything in part 23 either--for example, there
are specifications for the undercarriage and/or prop clearance that I may
find inappropriate for my application--experimental really is where we plow
the new ground!) Next, would construct a model of the wing only (with
dihedral, and a handle) and throw it with various weights and CG positions.
An excessive variation of airspeed and altitude due to fugoid oscillation,
as subjectively observed, would initially define the "natural" aft CG limit
of the wing by itself. The forward limit would be even more subjective--but
the basic objective of initial testing with something safe, light, simple,
and cheap should be fairly obvious. That should give some indication
whether the design actually has promise. If so, I would add a stick
fuselage and an empennage, and continue my subjective testing. If
satisfied, I could proceed with the main project; otherwise it might be time
to change the design and/or seek assistance.

The reason for this treatise is that I believe a lot really can be gained
from unmanned testing, and that it can be accomplished inexpensively and
with negligible risk of collateral damage.

However, (warning ... warning) the above applied only to conventional
aircraft, and even then does not address the required size of tail surfaces.
I am still looking for a "cheap and dirty" way to accept or reject a design
with regard to the tail's contribution to pitch and yaw stability. Also, I
also have NO intention of designing or building a canard aircraft because I
don't feel that I understand their principals well enough!

Peter