Crashing and Learning

by MyGeekShow | July 11, 2015 | (11) Posted in Projects

Ummm... lots of crashing! Maiden flights can be painful.

This week I flew the Solar Prototype on it's maiden flights. Sadly, it was the worst airplane I think I've ever made! I had to use the APM's stabilize flight mode just to get it in the air, and then it quickly crashed 35ft into a tree.

I learned a lot from this attempt, specifically that I needed some kind of roll control via dihedral or ailerons, and that a 10cm (4in) is too small of a chord for this wing at this speed.

I may be on vacation next week (I'll be right back!!!) but meanwhile I'll be getting a new wing and doing some additional research. A new plane will be built!

-Trent

Follow the mission and learn more on: http://www.mygeekshow.com/24-hour-solar-uav/

COMMENTS

Aviation Rehab on July 15, 2015
Totally enjoyed ur show u must let me know What result u get eventually!
Aviation REHAB out...
Log In to reply
grether2000 on July 12, 2015
I like the idea of working on a solar R/C plane, but I think you need to step back and start with some known good design criteria. Borrowing from existing good gliders design is a start, working the numbers for weight, wing loading, stall speed, and power. Then test and measure those, and compare with solar panel output.

As the full sized attempts have shown, it takes really good efficiency all around and is still very hard to do right.
Log In to reply
HilldaFlyer on July 15, 2015
Trent - great project, and best wishes. Did I hear you say that you wanted to fly during the day while charging the battery and then continue to fly all night on the battery? That is a pretty tall order. I encourage you to keep going because I really like the envelope pushing. One thing you should consider is what battery will keep your plane aloft for the duration of night? Another thing, just to keep the plane aloft during the day consuming 10A and charge a battery is a solar panel the size of your car hood. What is the weight of the solar panel? This setup can be run static on a picnic table. Good luck -
Log In to reply
grether2000 on July 12, 2015
This might be a good source for solar cells. They specifically have an R/C line available that is lighter at the cost of long term UV protection. http://www.powerfilmsolar.com/products/oem-comparison-chart/
Log In to reply
Krumple on July 11, 2015
From what I could see of your airfoil it was semi-symmetrical. You need at least a Clarke-Y (flat bottom) for good lift and low stall speed, even better would be a under-cambered airfoil. You also need some kind of dihedral or polyhedral unless you want to be forced to control the airframe ALL the time. No dihedral = no self correction = any correction for wind etc will have to come from the pilot. I suggest studying some of the old free flight airframes for ideas.

Good luck.
Log In to reply
Holski77 on July 13, 2015
People are over simplifying this. Re # is a ratio of momentum and viscous forces. If your claim is that he needs to raise or lower this number then you must think about what you are saying and realize that it just means he must make his plane larger are smaller. He already know what size he want his plane to do be, and he already knows he want to fly at a low speed (basically means Re is not going to change much). Other design changes need to be made, and Re is the wrong place to start.

Airfoil performance is determined only by mach and Re. Once you know these numbers you can find out what kind of lift and drag coefficients you will have at certain angles of attack. All that is needed is to find an airfoil that will work well at the cruise Re and mach you want your plane to fly at. Look for a good L/D value here, XFOIL is a free program and is perfect for finding these 2D lift coefficients. For modeling purposes you can compare these 2D coefficient to each other since they will both decrease about the same amount when translated to 3D. You will need to come back and find CL after you have picked a 2D cL and wing geometry.

Once you have gone through the above process just use the lift equation to size your wing.

L = dynamic pressure * Cl * wing area

Dynamic pressure = 1/2 * rho * V^2

Are you going to hand launch your plane or use landing Gear? This will determine your dynamic pressure at take off assuming you can't change the ambient pressure of your local park. If you hand launch your plane use your tossing speed to determine wing area. If this lift is less than 1.2 times the weight of your plane then you are in trouble. You can increase Cl by adding flaps, You can use landing gear to reach a higher takeoff velocity, or you can increase wing area.

If you are wondering where aspect ratio comes in, it determines how much 2d cL (lower case normally used to 2D) is lowered when translated to a 3D CL. This is a simplified explanation of the solution if you want to learn more i suggest some in depth reading from some aeronautical textbooks like those from Raymer.

To sum everything just saying you need to change Re simplifies everything because it completely skips airfoil determination and sizing based on mission, and so so much more. Even my description is too simple. If you find a good airfoil, a great L/D you can't just mount the airfoil at that angle on your fuselage and call it good. Re is different at takeoff, that means max L/D angle will change a little, so you will have to increase alpha for more lift. Oops, now when you trim your aircraft our for best incidence your fuselage will be flying nose down through the air, that's gonna add a lot of drag. Drag, oh drag... haven't even mentioned that yet.

Before you get everything right you will run in a lot of circles, but when you finish you will have the best plane for the job.

It is important to notice that when you get to this point that your airplane is not something you created, every aspect of its shape would be influenced by aeronautics. If you enjoy making your own inventions fly just because you want to see it, there is little need to go this in depth, just pick a wing loading and go. There is a reason airbus planes look a lot like Boeing; they are designed to do the same job. I think its funny when people argue over which company is better. Their engineers are trying NOT to compete with each other. Why try and beat a 747 when you can do a whole different job, like carry more people, or carry the same amount of people farther, or faster. If you want to watch this happen just look at what is happening in the regional jet market right now.


END RANT

Log In to reply
rcspaceflight on July 11, 2015
Definitely go with polyhedral. There is a reason why (pretty much) all gliders use it. As Krumple said, it will self correct with will be much more efficient than using ailerons. Even if you have computer controlled ailerons, dihedral/poly will be more efficient.

Remember, every movement of a control surface equals drag. Drag equal less speed. Less speed means you'll have to use your motor more. The more you use your motor the more you use your batteries.

You do have the right idea by having a long wing with a short chord. But you went too extreme with it.
Log In to reply
The-One-Who-Never-Crashes on July 11, 2015
I am really enjoying your video series and hope that you can get this project finished soon!

I would recommend a long (~4-meter wing) with an undercambered airfoil and polyhedral. You might want to look at some high-performance gliders for inspiration (such as David Windestal's glider in the Flite Test Glider Tow episode).

Another suggestion that I had was modeling your plane after an existing design or even modifying an R/C glider for testing.

I hope this helps!
Log In to reply
enivid on July 12, 2015
Keep up the good work! Before rebuilding, I think you should also do a little bit of math around what surface area you'll need on the plane itself to sustain good charging. The plane design is also a function of how much solar charging surface it can provide to itself.
Log In to reply
Stingray Aviation on July 12, 2015
I have a pretty effective way to design wings you might like. Start with this calculator that will give you the Reynolds numbers you fly at and some other data:
http://www.aerospaceweb.org/design/scripts/atmosphere/
Since Re is (density*velocity*mean aero chord)/dynamic viscosity, you can solve for the best wing chord for your Re range by inputing the other data from the calculator and working the formula backwards. Now that you know the wing chord you want, just decide on a wing area, aspect ratio, or span, calculate the other dimensions from that, and you have your planform. Next, come up with a list of different airfoils that you think might work well in your flight envelope, and compare their performance at your Re range. I would strongly recommend XFLR5 software for this. It takes awhile to learn but is extremely useful and can be found at www.xflr5.com. Whichever airfoils the program shows perform best, use those on your wing. This method might be a little complicated but it yields a very efficient wing and is totally worth the work.
Log In to reply
FlyingPooch on July 15, 2015
Firstly I'd like to make clear that I have absolutely no aerospace engineer wisdom to offer to your endeavors, but I'd like to offer my sincere well-wishes to your project.
Secondly, I'd like to express to you my admiration for your ability/willingness to put your project out there on public display - exposed to the masses - willingly accepting praise and criticism alike. An admirable trait, no doubt. Admittedly one that doesn't come naturally to me.
So again, I'd like to offer my congratulations on your progress so far as well as commend you on your willingness to experiment and learn in a crowdsourcing-type arena.
Enjoy your vacation!
Log In to reply
Tri blades on July 15, 2015
love the idea keep them coming

Log In to reply
Joseph S. on October 9, 2016
We've all been there bro. :) I was flying a quad I'm my front yard and I was high up and it was a toy and then it browned out or something and it went in this tree that was like 70 feet up. Stayed there for three days until my dad and my neighbor(it was his tree) combined their poles and we funnily got it down. It snaked down into the pavement. Charged t and it flew awesome.
Log In to reply

You need to log-in to comment on articles.


Crashing and Learning