Physics Experiments

[dhk79]

There have recently been several questions raised about the origin of a 3D model and its relationship to the CG of the physics model. So as we delve into this mystery, here are some files to play with.

The attached three EAs are identical, except for the origin of the 3D model. All are based off of my UltraStick, but have default physics (i.e. the Nexstar's).

US1 - The origin is near the Mean Aerodynamic Center of the model (as originally modeled).

US2 - The origin is centered on lower edge of the firewall.

US3 - The origin is offset to the right (near the muffler).

These will give us something to experiment with as we try to get to the bottom of these questions.

[phrank]

Not to further pollute the Deuces Wild thread.

As a modeler gains more experience and fluency with the Aircraft Editor, of course they may place the aircraft center as they wish.
When they import the .kex, it's no surprise why the aircraft appears offset in the Aircraft Editor.
For the benefit of the newbs starting out their modeling adventures however, they should start off with a common frame of reference.

From the KE Aircraft Tutorial:

Next, verify the world center. The [0,0,0] world location will be the default Center of Gravity (CG) for the model. While the CG may be refined in-game, we suggest that you set an approximate CG location prior to exporting the model. Note that changing settings in the aircraft editor will modify the CG location.


Good Exercise for the inquiring minds.

[willsonman]

By complete coincidence I was researching CG for a scratch build and came across some great info on RCScaleBuilder.
"Basically, the CG should be set where the tail volume (later) can readily control the angle of attack of the wing, and the farther back it is, the easier it is for the pilot to control the airplane, and the slower the airplane can land - both good things. The airplane can land slower because the tail can push the nose up farther at low speeds.

The tail volume (remember that wing volume determines lifting ability) is the area of the horizontal stabilizer, including the elevators, times the distance the tail is from the wing, the "tail moment". This is obvious; the stabilizer is a lever and the longer it is, and the more force you can apply to it, the better it works.

The harder you can push the wing to a higher angle of attack, and control it there, the slower you can land. Also, the easier it becomes to stall the airplane, but that's a pilot skill issue. The more you can control the angle of attack, the more you can move the Center of Gravity fore and aft, safely. (Just as an aside, that's why flying wings were never practical; very small safe CG range. (we won't mention Spitfires here)).

There is a fairly simple math formula that tells you where the point is that the airplane is nicely balanced, controllable, and light on the controls, both. This is the formula, first in words:

The Center of Gravity should be at: one sixth of the mean wing chord, plus the result of three times the tail volume, divided by 8 times the wing area. That's straightforward. Here are some definitions.

Tail volume equals the tail moment times the horizontal tail (stabilizer) area.

The "tail moment" is the distance from 1/4 of the length of the wing mean chord, back from the front of the wing, to 1/4 of the stabilizer mean chord, back from the front of the stabilizer.

The "mean chord" is simply the average front to rear width of the wing, or stabilizer. You find that by dividing the area of one side of the wing by the length of that side ( I will assume that both sides are the same). I showed the location of this point on the WW wing in the diagram. You only need to know where it is on the wing because the CG is calculated along that line, and you have to balance the airplane out there, or project that line over to the fuselage to do so. The chord of the wing at the fuselage is usually different, so go straight over to the fuselage, and don't use the same percentage at both.

An airplane with a rectangular wing has the "mean chord" at all points along the wing - all the ribs are the same length. A tapered wing has the mean chord farther inboard, at the center of the area of one side of the wing. This works for a swept wing, too, but the CG point on a swept wing can actually be behind the wing, at the center line of the airplane.

The stabilizer mean chord is found the same way as the wing mean chord, but you probably will have to calculate the area all by yourself.

In the old days, I used "quadrille" paper to do this. that's the paper with all the little blue lines on it that make 1/4" squares all over the paper. I used a 3 view and traced it over the quadrille paper, and then counted squares to find the wing area. The scale of the drawing doesn't matter, as long as you make all measurements from the same drawing, and use the percentage on the little one to calculate the percentage on the model.

The size of the squares doesn't matter, either, as long as they are the same for the wing and the stabilizer. I added a square if the wing filled at least half of it, and ignored it if less than half. Nowadays, I let my CAD program tell me the area of the surface.

The "Tail Moment" is the distance from 1/4 of the way along the wing mean chord to 1/4 of the way along the tail mean chord, straight back Don't slant the measurement in to the tail from the wing. Multiply that distance by the stabilizer area, and you have the tail volume.

Now, find 1/6 of the length of the mean chord, and make a mark there. Don't confuse this with the 1/4 mean chord point used above.

Next, multiply the wing area by 8

Next, multiply the tail volume by 3, and divide that number by the (8 times the wing area) number. This answer is usually about the same as 1/6 of the mean chord, but don't confuse the two. It will be more if the tail is large, and less if the tail is small (we still won't mention the Spitfire, here).

Finally, add the 1/6 of the mean chord number to the (3 x tail volume divided by 8 x wing area) number found above.

That sum is how far along the mean chord your CG should be. It wille be more or less at the center line of the fuselage, depending on the shape of the wing. Project it straight over, perpendicular to the fuselage, to find the balance point at the center line of the airplane.

Here is the formula

CG = 1/6 mean chord + (3 x tail volume) divided by (8 times wing area).

I can't make the formula line up in this web site, so you'll just have to line it up yourself.

Just for the record, 20.7 lbs /cu. ft is relatively light wing loading, about the same as the typical "sport 40". Having the balance point relatively far back will make that weight a lot easier to achieve. The moral of this story is that a large rear end is nicer to use. Don't extend that to anything else."

I also found that typical sport models have a 15lbs/cu.ft. loading and scale is 25lbs/cu.ft.

I suppose you could do this calculation and adjust the CG in RF to get to this location or set it as the pivot point. I think one of the frusterating things in the editor is the location of the fuel tank does not affect the CG in the editor but it can drastically change how the plane flies. Not sure what is up with that. Same with batteries in EP aircraft.

[dhk79]

Very good data.

Originally Posted by willsonman

I think one of the frusterating things in the editor is the location of the fuel tank does not affect the CG in the editor but it can drastically change how the plane flies. Not sure what is up with that.

There actually is a reason for that. CG calculations are made "dry". The default plastic fuel tank weighs next to nothing. Replace it with a heavy steel tank and try your experiment again

[willsonman]

Did not know that. Thanks!

[flip3d]

Ok guys! Here's the deal. US1 flys acceptably well. US2 never leaves the ground due to WAY nose heavy condition or the equivelent as determined by the software. US3 flys acceptably well except for an off axis wobble in rolls as one would expect. The off axis wobble can be removed by placing the CG offset by -2.50in. This leads to a predictable heavy side in the flight model in all other aspects.

An edit of US2 is provided with alterations to the position of the graphics model to closely parallel US1. Exact figures are given in the model description.

To my eye... they fly the same.

Your thoughts please?

[dhk79]

Flip, did you set up the physics? Or were you trying to fly the Nexstar default?

[flip3d]

Originally Posted by dhk79

Flip, did you set up the physics? Or were you trying to fly the Nexstar default?

The objective is to follow a routine editing process and keep the two models identical except for the design origination position. This has been set. I will take the models through various levels of flight performance once again using two models as a basis for comparison.

US1 and US2v1 seem to have no discernable differences. If anybody sees something that I've missed, please speak up.

[flip3d]

At this point I will use US1 and US2 and eliminate US3.

The wireframe will be altered to fit the graphics model(s). Appropriate airfoil cross sections will be applied and controll throws will be adjusted. The Ultra Stick by all appearance is a simple high speed "turn and burn" type of aircraft. The benign flight characteristics will be replaced in order to test the graphics placement differences at the limits of the simulator.

[dhk79]

Actually the point of the exercise is not to eliminate one of the test subjects, but to figure out if having the 3D model's point of origin other than at near aircraft center makes a difference in RF.

The answer to that is no, it does not make the slightest difference. All three models fly exactly the same and with none of the problems Flip mentioned (This is why I was asking if he set up the physics or was looking at the defaults).

Here's my take on the three.

[flip3d]

Originally Posted by dhk79

Actually the point of the exercise is not to eliminate one of the test subjects, but to figure out if having the 3D model's point of origin other than at near aircraft center makes a difference in RF.

The answer to that is no, it does not make the slightest difference. All three models fly exactly the same and with none of the problems Flip mentioned (This is why I was asking if he set up the physics or was looking at the defaults).

Here's my take on the three.

These models are much improved over the originals. Wireframe adjustments, component frame corrections, engine got BIGGER??, and WHATS UP WITH 12OZ. OF BALLAST IN THE TAIL???????????

Ok! US3 is a no go because the -2.5 in. side offset in the graphic has been removed making it a moot point since that was the only difference from US1. I trust that no similar alterations have been made to the others.

Otherwise you would be surprised at how similar your wireframe is to mine. That's where the similarity ends.

My objective is to make duplicates of the SAME graphics model with DIFFERENT origination points and edit them IDENTICALLY into very capable sport/aerobatic models. You must push the limits in order to make different flight characteristics apparent (Not proven at this point).

Question: Do Ultra Sticks knife edge like a stone?

[steve19]

The title of the thread "physics experiments" gave me bad memories so I decided not to delve into the mystery.

[flip3d]

Originally Posted by steve19

The title of the thread "physics experiments" gave me bad memories so I decided not to delve into the mystery.

Yeah baby!

[dhk79]

These models are much improved over the originals. Wireframe adjustments, component frame corrections, engine got BIGGER??, and WHATS UP WITH 12OZ. OF BALLAST IN THE TAIL???????????

Originals??? The original EA files that I posted had no defined physics. They were just RF defaults and that was noted in the post. This was done so others could play and see how to set up the physics with origin offsets and find out how it effects the end result.

Ballast??? What do you think? It's to adjust the balance and CG. As an engineer, I refuse to wave a magic wand and say move the CG back 3". You can't do that with a real plane, so I don't do it in the sim unless there is a valid engineering reason to do so. If you need to adjust the CG, you first move heavy components like batteries and if that's not enough you add lead.

Ok! US3 is a no go because the -2.5 in. side offset in the graphic has been removed making it a moot point since that was the only difference from US1. I trust that no similar alterations have been made to the others.

No alterations were made to any of the AVs. The offsets were just accounted for in the physics set ups. Look at the locations for the root frame, you'll see that they are different in each. But even with being different, they all fly the same.

Otherwise you would be surprised at how similar your wireframe is to mine. That's where the similarity ends.

I'd be more surprised if they weren't the same. The physics wire frame is supposed to match the 3D model after all, and this is a very simple plane.

My objective is to make duplicates of the SAME graphics model with DIFFERENT origination points and edit them IDENTICALLY into very capable sport/aerobatic models. You must push the limits in order to make different flight characteristics apparent (Not proven at this point).

I'm not sure I understand this as I think that's exactly what was done.

Question: Do Ultra Sticks knife edge like a stone?

For the most part - yes. The rudder does not have the authority to hold a knife edge with a standard power plant. If you over-size the engine, you can just barely pull it off. This RF model performs dead on to its real life counterpart.

[flip3d]

The original EA's were unretouched representations and as such were excellent test subjects. The AV's presented do not meet that criteria. The editing software that I use is strictly limited to the Aircraft Editor as supplied with RF. I do not have the tools necessary to move centerlines in graphics representations.

It is interesting to note that all three AV's have the same roll/yaw characteristics to the right. This is exactly how US3 (original) behaved when the lateral CG was corrected. US1 and US2 (after correction for the "nose heavy" condition) did not exhibit this characteristic.

While I like the improvements in the AV's such as the component frame alterations, I'll have to stick with the original EA's.