Friday, March 25, 2011

Adventures in Aircraft Model Validation

After several months of working on the Smoking Ace, I have learned one thing. Balsa is not your friend.
Aluminum is not so bad, but it is what it is and you need to make sure you bolt it up.

We have been working on conreol systems for the plane. It works like a normal pusher. The simulator models that we use suggest that it is pretty predictable to fly. I hope that is the design and not the simulator dumbing things down, so that my feelings are not hurt. Our primary model is X-Plane 9.3. It was pretty easy to build a model in the system, for what it is worth. I could plot out a simple airfoil and plot out the weights and measures of the plane as we are predicting them. I had to use a custom airfoil, our low moment  airfoil is well understood, but is not super common in general libraries. We use the e204 with a 16" chord.

We chose 16" because it allows us to get 2000 square inches of wing area in a reasonable span. No, there is
nothing reasonable about wings that are longer than I am tall by a ways. That would be fine, but there are two of them. Which means we have to take it outside to assemble the plane. As a reference, this is gives us a do not exceed gross vehicle take off wing loading of 50.2 oz/square foot.This is a little high, but for a 14' wingspan it is not out of bed.This is all in the batteries. This is an electric plane, with a 65cc equivalent motor driving a 22" fan.

This configuration should give us around 30 pounds of thrust. Hopefully, this will be more than enough to keep it flying. The plane's primary mission is to fly straight and level and take pictures and instrument readings. I think that the configuration will be successful at this. The Smoking Ace turns like a school bus, so it would be unexpected for us to try and snap roll it.The sim says behaves like a normal pusher configuration.

I mean that unlike a puller plane, when you put on the power you have to put on elevator or the nose dives. To compensate for this a bit our elevator is set at 4 degrees nose down. This us effective at counteracting the
offset of the line of thrust to the center of gravity. There is no vertical tail in this configration. So we have a soft feeling rear end. Combined with our shortish, blended fuselage, means that we have to use the inner  ailerons sometimes in flight as poor man's elevator. Long wings and an aggressive sweep makes this an effective technique. Double, mid-span control surfaces help to give us an effective rudder. Differential roll inputs counter-act the other roll inputs and change the drag profiles of the wings and developing a yaw moment with little roll.

In the next generation of the aircraft, there will be duckerons on the wing tips. Yeah, I know this is not an
approved, aerospace term. Which was discussed in an earlier posting. Properly termed, they are drag rudders.

Using X-Plane sizing them was pretty easy. Sure I can estimate that a drag force at the tip of the wing will
 result in a given yaw moment. However, it is hard to estimate the tip effects on approach when the drag rudders are too close to the wing tips. We found that they should be one chord or so from the wing tip. Farther out made them more effective, but it also made them impractical during low altitude maneuvers. They were so effective that they burned off airspeed and slammed the plane into the ground. Watching your model cartwheel wing tip over tip is exciting, but depressing.

NOTE: You have to release the brake when you are in X-Plane. Otherwise it has some crazy effects... I would have  thought that if you kept making the same mistake, they could figure that out and say... release the brake... or if you throttled up and the plane was not moving... some other visual cue may help.

On the other hand, I kind of thought that most of the planes were kind of lackluster to fly. The differences
between a Piper Cub and a 747 were not as striking as I would have hoped. A Cub should have felt a bit under powered and then feel kind of flippy. When it got into its acrobatic range. Where the tail and rudder became effective nearstall conditions.

If you want the latest revision of our model, you can get it here. If you find something interesting, or if we made a mistake please feel free to comment at .