What caused this flight behavior?

[dongfang]

Hi,

I have made my 2nd water rocket plane. See pictures.

It has a 55 cm wing span, a 40 cm chord at the root, and weighs about 475 grams all-up, except for the reaction water.

It is designed to climb vertically on the water rocket (like the space shuttle without side rocket boosters, and with an internal tank instead of external), and land in glide flight.

The concept is basically feasible - see my 1st generation plane, for example, at http://www.youtube.com/watch?v=d3ND3dC-Myw

Now, this 2nd generation one has a much larger rocket and a real structured wing. Unfortunately, it behaves strangely. I have added a picture sequence of a flight, supposedly taken at 13 fps.

The plane rises OK on the first part of the flight - then it seems to stumble over the air, and pitch down sharply. This causes it to slow down considerably. and return to a slow horizontal flight.
So, the trajectory is like an "S" seem from the side, with entry at the top and exit at the bottom.

The trust vector of the rocket goes through the CG.

The GC is somewhat ahead of where one would want to put it according to calculations like this (http://www.mulert-fotografik.de/schwerpunktdelta.html), but the bottle fuselage is much fatter than that of a Concorde, and generates more lift. At one launch with very low pressure, it glided stably and horizontally, but very slowly (should have stalled really, but didn't) and at a steep trajectory. Like an inefficient glider going too slow.

I didn't touch the stick any of the flights, except a slight pull-up at the last half second! The plane just happened to come in the right attitude to survive.

Any idea what may cause these crazy flights?

I have a few:

- Bad CG. But to which end is it off?
- Drag imbalance. The drag of the wing turns the plane over. I will try to block the passage of air through the tail, introducing some drag above the centerline. Might make a difference.
- Airfoil has one center line, the rocket has another, and they are not parallel. Could fix this by bending the rocket nozzle up by a few degrees.

Anyway, if anybody has a qualified guess, I'd be happy to hear about it.

Kind regards
Soren

[dongfang]

Hi,

Here's a flight with the observed behavior - flying an S. After the last picture, the plane pulled out of the dive, and with my help, made a safe landing.

Regards
Soren

[FlyingW]

Soren,

Excellent and interesting project. Haven't seen a water-powered rocket for may years. Cheers for bringing the technology back!

Could there be a change of CG as the remaining water runs out of the body?

Good luck,

Paul

[dongfang]

Hi,

Thanks for the nice words ;)

Well the plane has a well-placed (on the left-right axis and the top-bottom axis, as seen from behind the rocket nozzle) CG when dry.

Adding or removing water in the rotation-symmetrical bottle should not make a difference. The water will get distributed symmetrically around the "dry" CG, and the "wet" CG will be the same as the dry one.

Of course, there is the possible issue of water slosh in flight. But I don't consider that right now, as the flight trajectory does not look random and erratic. It is consistently - wrong.

Regards
Soren

[HerkS]

The craft flys as if it is extremely unstable in pitch. This is usually caused when the longitudinal balance point is too far back. However a large fuselage (the container) also contributes to longitudinal instability. Can you provide more details? Close up pictures of the aircraft and it's control surfaces. Location of the balance point when no water remains.

[BMatthews]

The movie was great but from the wrong vantage point. You'd have been better off to post a movie showing the flight pattern from the ground. It's pretty much impossible to try to analyse it from the video. And the stills you posted don't really provide the motion related clues to help much either.

All I can really suggest is the old tried and true standard of going out somewhere and doing some dry test glides. Preferably from a hill or rise so you extend the glide time and can better analyse the flight.

Or shoot it up again and film it from the ground well away from the launch point so we get a "side view" of the flight as much as is practical.

Keep in mind too that unless you push the model to horizontal BEFORE it gets lower than its stall speed then the natural action will be to nose down to get back to the trimmed flying speed. To avoid this you need to watch the launch and push forward to get the nose down early enough that there is still good flying speed at the point that it is back to level. From there it'll glide down normally. If you just rely on Mother Nature to pull the model out then it will do just what it is doing. It'll slow down until almost stopped and then swap ends and dive until flying speed is reached and then slowly raise the nose back to level. Any airplane will do the same under the same conditions.

[dongfang]

Hi,

The onboard movie was from my old rocket plane. That flew OK, but didn't really perform. The problem is with my new one.

I have added a few more pics of the plane, and I will try to get a side view video at the next launch attempt.

Regards
Soren

[JetPlaneFlyer]

When the bottle is part full the water will sit at the bottom (rear of the model) and could very likely be causing the CG to shift too far back. Have you checked that the CG is far enough forward for every possible 'water fill' condition? The CG shift as the water is expelled may well limit the length of bottle that can be used.

The other thing that's a possibility is that the low wing means that the thrust line is not aligned with the centre of drag, meaning that there will be a downward pitching moment as the speed builds up... but i'd check the CG first.

Steve

[HerkS]

When the craft is in the gliding phase, I suspect that all water is now gone from the container. So when it is now gliding, I think water is not a factor.

The very large forward fuselage (water bottle) ahead of the wing now acts as part of the wing. This makes a simple flying wing balance calculation incorrect by a large factor. I have calculated roughly from your picture, the effect of this - having no accurate measurements. Based on my analysis of this factor, I suggest that you try a flight with the balance point moved forward to the red line on this picture.

Do not expect this to be an efficient gliding machine. A much larger wing would be required to offset the effect of the container. But then you would not get so much height on the launch.

[dongfang]

Hi,
\
Thanks for the replies, both!

When most of the water is still in the tank, it actually flies OK. It´s around burnout it starts to get strange. And yes, all water is expelled from the tank during the rocket burn.
The gray PVC pipe with the nozzle interfaces to the launcher, and also mixes air and water. This eliminates transient effects that sometimes happen when just expelling pure water and then air, and throw the rocket off course.

I totally agree with the off-center center-of-drag idea. I have tried to add an air brake to the tail, and I will see if it makes a difference.

Today, I tried to test fly it again, but I didn't have any luck: One one the plastic clevises broke in the -3 deg C cold as I tried to adjust it, and had no spares with me.
I did make a few test tosses again though. It actually glides OK, with some tendency to stall and roll on the side. I think I will try move the CG as HerkS suggests.

So the order of experimentation will be:

1) Launch with the tail air brake installed
2) Launch with the shifted CG and the air brake

If that does not solve it (and the plane is still alive):
3) Heat and bend up rocket nozzle 4 degrees, remove air brake, try again.

If this ultimately works out, I'm already considering a longer bottle and canards... should get me a better energy/weight ratio.

Greetings from snowy Switzerland!

Regards
Soren

[HerkS]

Soren, I took time now to study all three of the videos that you posted on YouTube.

In all of your launches you appear to be getting strong aerodynamic flutter of the control surfaces. This is what damaged your servos as you show in the number 2 video.

I am not sure how best to prevent this, but the most important thing you can do is to make the wing more stiff. If this was my project, I would probably put carbon fiber strips on top and bottom of the wing just forward of the hinge line.

Happy New Year greetings also to you from Virginia USA. Not so cold here today.

[dongfang]

Hi HerkS,

What you saw on YouTube was my first, quite simple rocket plane. It flew well-behaved, but had the flutter problem.

The problem in this thread is my second generation water rocket plane. I took pretty much the corrective actions that you also suggested, and the new plane has structured (what is it called really??) wings of 2 layers of Depron with 2 CF tube spars, ribs and all. This plane has some stability problem.

Hope that cleared it up ;) I should maybe have made it clearer from the beginning that the YouTube vid was from a different plane.

When I once tried to fly the 1st plane without the camera on its back, it made a negative pitch 180 deg turn at launch, and disintegrated upon impact :)

Regards
Soren

[HerkS]

Hoping to see more of your flight experiences -- perhaps after the atmosphere warms a bit

[dongfang]

Hi,

OK, I actually got a few flights logged today despite the cold - and the plane is still alive!

The cold caused the launcher to seize repeatedly after the first launch, and the check valve in the pump to freeze stuck too. Enough to look after.

The air brake experiment: The plane just flipped over in the +pitch direction instead of minus. We decided that is was unstable, and moved the camera forward to the other side of the R/C, to shift the CG forward. I will balance the plane on that pipe again later, and post a pic.

Second and last launch of today: The boost phase was perfect, but I guess I was too slow to get the plane into horizontal and on into a glide after that. It went into a spin of some kind, where it did not respond to controls at all.
I have seen that pattern before from my 1st water rocket plane, and the solution there was a larger vertical stabilizer. So, I will put a larger one on the new plane too.
These planes are pretty unforgiving; it is impossible to recover control of them once lost. At least for me, who is not so much of a flying ace really. In fact, today I thought I must had lost RX power, so little response did I feel.

So, it appears that shifting forward the CG solved the problems during boost phase. I actually flew the plane with a motor on board once (pusher prop inrunner in a slot in the bottle right before the tail, and with the end of the bottle cut off) and it was flyable OK. I just need to get it more monkey proof.

Oh, and BMatthews: OK I see the information here was not enough for a real judgment... in the flights of Jan 2., the plane dipped the nose during boost phase, when the rocket was still pushing, and the speed was well about stall speed. Today, I might well have stalled it after rocket burn-out, because I did not get it to horizontal soon enough.

Regards
Soren

[HerkS]

Hi Soren,

This is a very interesting project. Thank you for telling us about it as you make your modifications and test flights.

[KenmoreKid]

Soren,

I'm glad to see you have made progress on stable flights with your water rocket. But if you insist on flying in -3 C temps, perhaps add some food-safe anti-freeze to your water :-)

And for better understanding of aeronautical stability, find a copy of Harry G. Stein's "Handbook of Model Rocketry".

That book taught me about stability criteria and how to determine them for model rockets. Folks on RCGroups always talk about CG (Center of Gravity, aka Center of Mass, aka balance point) but I never see any references to the other critical item - Center of Pressure (CP). I learned about that from Mr. Stein's book.

So if you just look up the rule of thumb for a delta wing you are not getting the full story.

As the Center of Gravity, or more correctly the Center of Mass, is the "geometrical centroid of all the mass" the Center of Pressure is the "geometrical centroid of all the aerodynamic forces". For stable flight, the CG needs to be in front of the CP. If they are reversed the rocket or plane wants to fly backwards, and if they are at the exact same spot you've got a critically unstable bird that will tumble.

Think of a bow and arrow. The arrow has a heavy nose (the stone or metal tip) and a large but light tail - made from feathers in the old days. The heavy nose pulls the CG toward the front, while the large area of the feathers pulls the Center of Pressure toward the rear.

Mr. Stine also showed me how to figure out the approximate CP without fancy math or computers.

Make a cardboard silhouette of the rocket or plane and find the balance point of the silhouette. That's a good first approximation of the CP. You can do better by adding the other fins not seen in the silhouette, etc..., or cut out both side and top view and then balance them together. If you have a computer CAD system you can of course do this mathematically, but the cardboard is quick and easy, and less subject to math errors.

A second cool trick I learned from Mr. Stein to determine true stability before flight. First find the balance point (the CG) of the real rocket loaded for flight and tie a string (1 or 2 meters is enough) to that exact point. Then twirl the rocket around your head in a horizontal circle. Not too fast, just fast enough to keep the string taught and the rocket off the ground. If it follows it's nose, you are stable. If it goes backwards, you are unstable. If it tumbles, you have CG=CP and are at critical stability/instability and will need an onboard computer system like the B-2 in order to fly :) If you get dizzy and fall over, you are twirling too fast :eek:

For your water rocket I'd try this both fully "fueled" and completely empty. Then add enough nose weight to stay stable in both conditions.

I first learned this stuff back in 8th grade when I tried to scratch build a model rocket based on the Space Shuttle. My first prototype went horizontal and circled around about 15 feet over our heads. That was scary enough to prompt further investigation on my part and I found Mr. Stine's book.

Good luck.

- Neal

[dongfang]

Hi all,

Well I tried to fly it again Saturday. I have gotten the boost phase stable now, but as soon as the rocket burns out, I lose control!

I guess its worst problem simply is its large and light fuselage (soda pop bottles). It simply competes with the small wing for aerodynamic attention / forces?

I will definitely try adding weight to the nose, until it stops blowing about (and maybe simply falls like a bomb, haha). It that works well, I might try extend the bottle and shift the weight further forward.

Meanwhile, I am building a copy of first water rocket plane (the one from the YouTube vid in the beginning of this thread). The original one was lost in a crash - but they take only a few hours to build. It's going to be fun having something again that actually works...

And I will make up a number called the exposed bottle area : wing area ratio, and see how much difference there is in that between the 2 planes. It is simply the area of the bottle that is not covered by the wing in a bottom view, divided by the wing area.

Maybe the ultimate shuttle plane will have 3 or 4 parallel stacks of smaller diameter, heavier bottles that can withstand more pressure.

I will make a comparison of the ratios, location of CG etc. of the 2 planes - one that flies and one that does not - when I have both ready. I will do the cardboard thing that KenmoreKid suggested, too.

Kind regards
Soren

[HerkS]

Soren, glad that you are now working on this again.

Looking forward to your experiments. I do think nose weight will help.

[maxadventure]

Just found this thread, brilliant work!

I'm somewhat surprised that it appears you are using elevator/ailerons on version two? I would think it would be advantageous to use the radio's elevon mixing function and mechanically simplify the design. Just a thought.