in theory if you do that you'll knock the plane's dynamic balance out of kilter, tending to make it roll out on inside maneuvers and roll in on outside ones.

In practice -- I dunno.  Start with the number of answers you want.  Multiply this number by 0.8333 and subtract 1/4.  Round up to the nearest whole number, and find that many experts.  Ask your question, and stand back to record your answer(s).  If any of them are hot heads, keep your first aid kit handy.

On airplanes with a higher length/wingspan ratio (think jet fighters from the 1950's) there's a problem that can develop called "inertial coupling", which I suspect is made worse if you hang the weights off of the principal axes.  Like I say, I'm not sure if it makes a difference with our planes, or if it does if it makes enough of a difference until you're really really good.

One thing to keep in mind, when you add weight, it creates a moment( proper term?) to the CG. I think of this as an imaginary line from the weight to the CG. so adding weight to the trailing edge tip creates a moment arm to the cg which is actually acting diagonally instead of in line with the desired direction. the other thing to consider is that adding an ounce at the trailing edge is pretty much like adding about 1/8 ounce to the tail as far as pitch CG is concerned. So its not a very efficient way of balancing the airframe .
The other consideration is that it will in some way affect the roll of the airframe because of the moment line drawn from the weight to the CG.This COULD create some roll yaw issues. Now the question is, are we good enough to recognize it when it happens. I do know this, even if you cant tell what is wrong, an airplane that is not trimmed well is much harder to fly well, and that you can tell.

I'll agree and disagree with some of the comments to date...

First, the tip weight is part of the structural unit called the model. Its basic function is to balance other things. The inboard wing flies slower than the outboard, gradually going faster from inboard tip to outboard tip. At some point betweeen tips, lift to each side is equal. Unequal panel spans - outer panel shorter - is one way to tackle that. "Excess" tipweight outboard - to meet the lift difference - is another.

The weight of the flying lines is not centered on the model, either. The part of that weight carried by the model is all on the lower-lifting panel side. Again, tipweight helps balance that..

These are useful to prevent lift from rolling the model about (around) that heavy cluster of stuff called the fuselage/engine/tail, etc.

Second, as mentioned, we can identify a moment about the cg. Since the wing and other pieces are a one-piece, rigid structure, there is only one center of mass. It may not lie on the principal axes of the structure, exactly, but should be very close to them.

A moment is force acting at a distance from a real or 'virtual' pivot point. Sayyy, whaa??? Examples: Lift is a force; the point across the span where lift is equal on both sides can be considered the point where it all acts. If force through that point goes through the cg, the cg moves in the direction of the force. If it doesn't go through the cg, a moment is created trying to roll the model in response to the greater force on one side.

For a model in flight, the cg (Center of Mass, actually) is the equivalent of the pivot point. Balancing the model about (on) the fuselage long axis for the lift, drag, thrust, etc., conditions reduces or eliminates the tendency of a temporary moment to cause unwanted roll. "Hinging" is a roll/yaw reaction caused by severe moments appearing at high-g maneuvering when the total forces and the Center of Mass pivot don't "agree."

Third, as mentioned, the longer the arm between the weight and the pivot point, the greater the "moment" a given weight provides. But, adding tailweight doesn't really cause a dynamic moment. The weight just becomes a part of the rigid structure. It DOES adjust the Center of Mass of the rigid structure... Which is what you want to do...

So, the least added weight that does the job will be as far from the (incorrect) cg as possible, and as near the line of the dynamic center of forces as possible. (Dynamic - since it is the changes in lift -and its drag- that can cause problems, static (non-moving) solutions just might not be enough.) Weight added to the left or right of the fuselage axis can create an unwanted "moment" in high-g conditions.

In fewest words: ballast in the tail on the fuselage centerline.

IMHO...

I think you're right Howard, but I'm not sure.  When you need things balanced down to a gnat's whisker then lining up the major axes of inertia with the rotational axes is vital (car wheels need this done during balancing -- not only do they need to be balanced around the axis of rotation, but the inertial ellipsoid needs to be lined up with that same axis, otherwise it tries to shimmy.).

Our stunters don't rotate that fast, so my knee jerk response is to think it doesn't matter -- but I'm no Expert flier to be able to tell the difference when 1/4 ounce goes into or out of my tip weight box, much less a gram or so.  So what do I know?

Chris has an interesting issue.  I'm embarrassed not to know anything about it, although I think I'm supposed to.   Not knowing anything is no bar to my opining, however.  Of the three principal axes, the pitch axis of a conventional stunt plane is the axis with the middle value of moment of inertia.  Roll has the least, and yaw has the most.  Rotations about the axis with the middle value of moment of inertia are unstable, so it may not hurt anything to move the axis.  Brett should comment. He steers satellites, so he knows from rotational inertia. I asked him about this before.  I think he said it's no big deal. 

  Well, I just watch other people steer them, to be entirely technical.

     This is really two different (and interesting) issues. The original is a shift of the principle axes from the geometric pitch axis.  I tried an experiment intentionally moving the pitch principle axis by adding an ounce near the LE on the outboard and an ounce near the TE of the inboard wing. I  I was moderately careful about ensuring that the CG didn't move. When I flew it it had a small but distinct change that looked like it was rolling and yawing differently than before and in the direction I expected for the . I didn't spend any time trying to retrim it. I did take a stab at the values of the principle axes and then some idea of the rotation of the principle axes - on the order of a few degrees as I recall. I have some notes somewhere, or its in an old Compuserve forum post - never let anyone say nothing ever disappears from the internet!

   The question at hand was essentially the same as above - did it make any sense to shift the tipweight fore and aft to adjust the longitudinal balance. I came to the conclusion that the value of the tiny weight saving you get VS the shift of the principle axes was definltely not worth it. You might shift your 3/4 ounce of tipweight by 4" and have a microscopic change in the longitudinal CG but shift the principle axes enough that you would have to completely retrim the airplane in some strange way. As opposed to putting a tiny bit (maybe 3 grams in this case) of weight on the tail.

    One huge problem with the entire analysis is that we really don't have any good idea of the geometric pitch axis is really the principal axes in any case. Particularly when people pull little tricks like putting all the heavy wood in the outboard wing to "save weight". Or when you have intentionally crooked fuselages, or very small misalignments , either intentional or unintentional. That's a bad idea on many levels but it certainly makes it entirely impossible to determine where the principle axes are relative to the geometry.

   The second issue is trying to rotate about the intermediate axis. Of course if you really torqued it perfectly around the inertial intermediate axis it wouldn't do anything funny but that's not going to happen. Without any notes in hand I can only say I came to the conclusion that the other dynamic issues (like those arising from constraining the motion in the models yaw axis) was going to swamp the effects of the intermediate axes.

   And in any case, most people never get the basics (i.e tip weight, rudder, leadouts) even remotely close to right, so the errors in the aerodynamics and mechanical setup entirely swamp any attempt to do useful experiments or take advantage of any of these effects. I can't overstate this - most people's airplanes are SO far off that just going through the very simplest straightforward adjustments carefully would tremendously improve their performance.

    Brett   

So I gather from reading all of this that side mounting engines would receive a similar penalty?

  Presumably, although an inverted engine probably moves the principle axes about the same amount. Only difference is that it's a rotation around the Y axis instead of the Z axis which probably makes less difference still.

    I think the key feature of this is that the other trim settings - either erroneous, or dogma-derived - are generally so far off that you can't tell the difference. And given the likely variations in the mass properties arising from other effects, we probably have no idea whether a shift of the principle axis from the effects we are discussing makes it closer to the geometric axes of symmetry, or further away!

     Brett

OK so following all of this a control line canard would seem to have a real tip weight placement problem that is only matched by its lead out guide issue - both placed out on forward facing probes I would imagine? But then you would still have the mass of the main wing well off the principle axis  to contend with.

A good reason that canards are not well liked for stunt?