Author Topic: Prop tips and influence on thrust (Read 9151 times)

John Witt · « **on:** November 04, 2009, 10:37:07 AM »

Hi all,

I just have finished a Vector 40 from the Brodak kit. This airplane has a short landing gear and I'm faced with the need for a smaller prop diameter than the APC 12-6e that I've been using on my Panther. I have a 3 blade MA 11-7 that I was going to try initially, but my experience with this prop suggests that it consumes more power than the APC for equivalent performance.

I'm considering trimming the tips on the APC 12-6 to make an 11-6 which will have slightly more blade area than the APC 11-5.5. It seems to me that even though I decrease the swept area, the last bit of the APC's swept tip can't be contributing much to the thrust.

What do you think?

John

Alan Hahn · « **Reply #1 on:** November 04, 2009, 10:59:00 AM »

I am not sure whether the ARF/ARC gear is different from the kit, but you could consider somewhat larger wheels to get the clearance.

Igor Burger · « **Reply #2 on:** November 04, 2009, 12:02:07 PM »

Remember, it is power of 5 ... so that last inch on tip can contribude 40% of thrust.

However it depends on model, its size and weight and line length, it will save also 50% of battery, so it can be lighter and that smaller prop could be enough.

Dean Pappas · « **Reply #3 on:** November 04, 2009, 03:30:44 PM »

Hi John,
It is as Igor says ...
Here's an outfit that I have found will whip you up semi-custom aluminum gear for quite a reasonable price.
http://www.tntlandinggear.com/

Regards,
Dean P.

bob branch · « **Reply #4 on:** November 04, 2009, 06:37:23 PM »

John

You covered fillets! I bow down in humble worship

bob branch

John Witt · « **Reply #5 on:** November 04, 2009, 07:36:05 PM »

Bob, it's only an illusion created by the low photo resolution. The covering is Ultrakote, and the fillets and fuslage are painted. The actual finish isn't as good as it looks in the pictures. I had a learning experience...

Regards,

John

John Witt · « **Reply #6 on:** November 04, 2009, 07:43:29 PM »

I did a quick run on Javaprop (Martin Hepperle). Prop efficiencies:

12 in = 35.9%
11.5 = 35.4%
11 = 35.1%

The thrust went down slightly and the pitch angle increased slightly as the diameter went down (speed held constant at 10 m/sec)

About what you would expect, but the differences are smaller than I supposed. I didn't have time this evening to explore the differences in blade shape.

John

Dean Pappas · « **Reply #7 on:** November 04, 2009, 08:31:30 PM »

That sounds about right, John, but any time you run an RPM governed system and need more pitch to fly at the same speed, the tightness of the speed control gets looser: more whip-up in runway wind, and more slowdown in the climbs.
Wider blades with thinner airfoils might help recover some of the goodness, as well as a third blade; but in general, you should add more and more diameter until you think it is starting to kill the corner, then back off a little. We will all be flying storks before you know it!

Best Regards,
Dean P.

Howard Rush · « **Reply #8 on:** November 05, 2009, 01:00:39 AM »

35% sounds really low for prop efficiency.

Igor Burger · « **Reply #9 on:** November 05, 2009, 02:19:18 AM »

John, I think you have to count with speed 20-25m/s.

In any case, efficiency is number telling how much input power on shaft is converted to aerodynamic power. It does not tell you thrust itself.

So you can have two props 11" and 12" with equaivalent efficiency, but thrust can be 50% different.

However wider blades will lead to lower efficiency, especially if you load the prop because the aspect ratio is lower and induced drag on tips will be higher. It can have impact to unregulated power train, but we use governor, so higher drag (= torque) and lower efficiency will be compensated by higher input power. I do not think you can see any other difference - may be only higher reynolds number can improve it

And regarding bigger = better ... well it is not everytime so, bigger pulls better in climbing, but it has some disadvantages in wind and tubullence

Alan Hahn · « **Reply #10 on:** November 05, 2009, 07:42:56 AM »

well my 2 cents is that you need to look at prop efficiencies at the airspeed you intend to fly (probably 54-55 mph or ~24 m/s like Igor said) and adjust the rpm until you make the equivalent thrust at that airspeed--this of course is only comparing level lap airspeed.

Once you do that, then look at how the thrust varies as a function of airspeed around the level lap airspeed. The steeper the curve the better. Honestly I like PropCalc for this because it presents its data in exactly the format I need---prop thrust and power vs airspeed at a constant rpm. Also you can output a spreadsheet table of these items as a function of airspeed, then in a spreadsheet table, you can overlay different prop combo's.

This you need to fold in with level lap efficiency--since in electric we need to carry around a battery to match the energy needs of the flight and battery capacity equals weight (not to mention cost!!).

What makes CLPA so interesting is that it is the entire package (power, airframe, not to mention piloting!) that actually matters, so optimizing on one thing doesn't fix everything.

Then go out and fly and see if it behaves like you want. That is the final arbiter. At least with the Chicago area, we have a lot of time to think and calculate until the next "spring"--like spring actually exists in this area!

Dennis Toth · « **Reply #11 on:** November 05, 2009, 08:04:31 AM »

John,

I have flown the MAS 11x7 3bld pusher and I like the way it pulls through the up top maneuvers. It would be worth getting a 3bld spinner (Duo bro has them) and giving it a try on your ship (I recommend getting the spinner because I and several others have found that it reduces the amp draw about 10% vs just a prop nut). You may have to repitch the MAS with a heat gun. I found the pusher was under pitched at my reference station 9 (I have found that I can set my pitch at station 9 for 11 & 12" props and it works fine) i.e. it measured 5" rather then the 7", I set mine to 5.7 ish and it hit the same lap time and amp draw(about 36 ish) as the 11 1/4" x 5.5 APCE - P (cut down 12x6 with raked tips) that I had been using. The three blade was quieter by just a small amount then the APCE-P 2bld and lots quieter then the 11 1/4"x6 Wood Zinger pusher (cut down 12x6P). All have lap times of 5.1 on 63' eye to eye lines (about 66' C to C for my ship).

Bottom line is you should give the three blade a try and see how it pulls your ship. If you need to adjust the speed use a heat gun to warm the blades (the MAS props take a bit more heat soaking to get them to take the repitch then the very thin APCE's) but it works fine.

Best, DennisT

John Witt · « **Reply #12 on:** November 05, 2009, 11:59:46 AM »

Is the spinner helpful because of turbulence around the prop hub?. The MA three blade has a large hub with a cylindrical shape and sharp edges. If I put much of a spinner (larger than, say 1" diameter, it wll start masking off the motor and cooling air inlet area.

I wonder if rounding the edges of the prop hub will do the same thing.

This is all somewhat theoretical at the moment. The Vector has not flown yet and the forecast is for crummy weather for the next week or so here in the Seattle area. Big winds today and tomorrow, plus 100% rain chance. Gives me something to think about whilst donning my Gore-tex.

John

Alan Hahn · « **Reply #13 on:** November 05, 2009, 12:15:35 PM »

Quote from: John Witt on November 05, 2009, 11:59:46 AM

Is the spinner helpful because of turbulence around the prop hub?. The MA three blade has a large hub with a cylindrical shape and sharp edges. If I put much of a spinner (larger than, say 1" diameter, it wll start masking off the motor and cooling air inlet area.

I wonder if rounding the edges of the prop hub will do the same thing.

This is all somewhat theoretical at the moment. The Vector has not flown yet and the forecast is for crummy weather for the next week or so here in the Seattle area. Big winds today and tomorrow, plus 100% rain chance. Gives me something to think about whilst donning my Gore-tex.

John

I have been flying my old e-Nobler recently in order to make some airspeed and altitude measurements. Since I am intrinsically lazy-- and my fooling around makes me take the cowl on and off-- I have been running the setup without the spinner, but with the backplate (for a 2" spinner). I really don't notice much difference between the measurements now and the ones last year with everything included. This of course is with an APC 12-6 Electric which has a pretty big blade angle next to the hub. Maybe that's a difference.

But one thing to consider is the spinner shaped prop nut you can by from DuBro--I have been using it on my Vector. It was the easiest thing to buy at the LHS. It also leaves a large enough gap for cooling air to flow into the nose, and still shadows the area around the prop hub. It is somewhat heavy though. Here's a photo of my Vector nose with the duBro prop nut. Ironically, the backplate you see is for a GreatPlanes 1 3/4" (I think) electric spinner. I left it on because I needed clearance for the prop hub and nose ring.

Dennis Toth · « **Reply #14 on:** November 05, 2009, 03:13:13 PM »

John,

I have been using a 2" DuBro plastic 3 blade spinner, it did need some sanding on the inside mount posts to fit the MAS 3 blade in but it will fit. I'm not sure why the amps drop so much with the spinner I am speculating that the hub and thick blade root (particularly with the 3 blade) on the MAS creates more drag that the spinner eliminates. If you have the air from the cowl ducted up toward the motor and allow for clean exit the motor cooling should not be a problem. The common approach is to split the cowl air inlet and direct half to the motor and the other half to the ESP, Timer and batteries. Another option that has been used is to install a small air scoop on top of the top block over the motor and have it blow air into the motor compartment. As long as you don't over prop there shouldn't be a problem, you want to keep the static amps around 36 ish as a starting point.

Best, DennisT

Paul Walker · « **Reply #15 on:** November 11, 2009, 01:19:16 PM »

And regarding bigger = better ... well it is not everytime so, bigger pulls better in climbing, but it has some disadvantages in wind and tubullence

[/quote]

I agree with Igor here. I also add that in dead air conditions, bigger is CLEARLY not better. The big prop creates too much turbulance, and a smaller prop makes less. I have done these tests in the past, and have conclusively proven to my self that this IS the case. However, not wanting to switch props all the time depending on conditions, I settled for something in the middle. That was on the gas planes.

I intend to repat these experiments with an electric plane as soon as my flying field is completed. Changing props from big to little is a bit more complicated with electric. I will need to switch motors as well to get the RPM I need for the small prop.

With the gas plane, I could fly a pattern standing "glued to the ground" (in other words, not backing up at all) in absolute dead air with a 9" prop on a 725^in plane. I am curious to try this on electric. To date, I have been down to as little as 11.3" prop on a 68 ounce electric with no apparent issues.

I'll keep you informed.....But getting the field done may be another year yet.

Dennis Toth · « **Reply #16 on:** November 11, 2009, 03:05:42 PM »

Paul,

Did the pitch have much effect on the dead air turbulence?

Best, DennisT

Igor Burger · « **Reply #17 on:** November 11, 2009, 03:25:22 PM »

Also good point Paul, but I can contribute also experience.

I have model with extremely large prop. I am speaking about my indoor with prop 10x4.7 and span only 32". I fly it in gymn, in dead calm and I never had such troubles.

It means that large prop itself is not the cause. But large prop on big model does it while small prop on the same model not. This we know from IC models as you wrote.

But there is one difference which can answer all. My indoor does not have flaps and it has low wing load, while large models with strong piped engines has relatively large wing load and uses flaps. Mean the wing does lot of lift and we know that we have strong tip whirpools on tip of wing at large lift coefficietnts and small on low lift coefficients as well as induced drag of the wing.

It can very well fit the experience:
-Small model and large prop, but low wing load without flaps does not make problems (low cl = low turbulence)
-large model with flaps and small prop also does not doo too much problems, bacuse model probably slows in corners due to low power and thus it does not have enough energy to make it
-large model with flaps and big prop does make problems, bacuse model does not slows in corners and prop will just win the battle with drag and pull it out of the corner, reason is mixed air on tips and also by loaded prop which will come to slippage

well - we will see more when you will do those test you mentioned, but I do not think you will find some surprise :-)

if this is the real reason, the better aspect ratio or some wingles etc can help

Howard Rush · « **Reply #18 on:** November 11, 2009, 04:09:03 PM »

Could the reason for success of the smaller prop be a higher velocity prop wake that causes the prop wake to move out of the path of the next loop?

Dean Pappas · « **Reply #19 on:** November 11, 2009, 04:47:34 PM »

I think Igor has it right:
More diameter always helps maintain speed in the middle of the corner, increasing lift and the wake turbulence that comes along for the ride, without option.
Dean

John Witt · « **Reply #20 on:** November 12, 2009, 06:32:28 PM »

It would be interesting to fly when the air is saturated and at the dewpoint. You would perhaps be able to see the tip vortices and perhaps the prop as well, as with the big birds. I bet the tip vortices would disappear in level flight and only show up in the hard turns, since you don't normally seem to run in your own turbulence when flying level --wing not loaded much.

John

Howard Rush · « **Reply #21 on:** November 12, 2009, 08:00:26 PM »

We live in a place where the dewpoint is usually near, so we can experiment.

RandySmith · « **Reply #22 on:** November 12, 2009, 08:34:26 PM »

I think there is NO question that wing turbulence is the biggest factor, and the more you move the controls the more turbulence you make.
The test I have done back this up,
1-When it is dead calm I get hit by my own wake mostly in vertical 8s and it seem to always happen right at the intersection point either going up or down or both, I very seldom get hit on top or bottom or any other place in between, it happens at the intersection where I am moving the flaps-elevator the most

2-Overhead 8s I hit my own wash in the intersection where I am moving the flaps-elevator the most

3-Triangles and squares, I only hit my own wake at the corners and never any other place..this is where I am moving the flaps the most

4-round loops, I can most always move out of the way, but if I get hit there it is most always at the bottom?? I have no reason why this would be so? If I ever get hit any other place in rounds it is going up where the plane was moved from level flight transitioning from level to a round loop

5-My wake turbulence is less when I go to a smaller prop, I also have a plane that feels much more tail heavy and is much easier to turn with the smaller prop..I do not have to move the handle nearly as much to get the same turn. So I see why Paul says this about the small diameter prop, however I am not sure if it is the prop? Maybe very interesting to load up the nose with weight so that I have to move the handle as much as with the large props.

I have also stood very close to the edge of the circle when Paul was flying (and many other) and there is a very large rush of air coming from the plane, I was very surprised at how much air a stunt ship moves in the bottom of a triangle or loop!

I was witness to a famous Northeast flyer who was going thru the pattern in dead calm, He turned a hard bottom sq. corner and in the dead calm you could see the exhaust trail spinning in a circle, when he came back and made the corner in it the plane dropped straight to the ground, It is clear this was vortices from the wing. This doesn;t rule out the prop wash, but it does tend to make me think prop wash alone is not the entire reason for our woes in flying thru our own wake.

More later

Regards
Randy

Igor Burger · « **Reply #23 on:** November 13, 2009, 01:37:40 AM »

Yes, exactly, it is interaction between wing and prop. Large prop gives more energy than smaller one - energy is what can move the air. Wing is tool to do it - its induced drag.

Dennis Toth · « **Reply #24 on:** November 13, 2009, 06:49:12 AM »

Randy,

Interesting observations, one explanation of why the small prop has less wake could be that because there is less control input to make the turn the ship creates a smaller wake. It also seems that lighter ships would also create less wake, something to try.

Best, DennisT

RandySmith · « **Reply #25 on:** November 13, 2009, 02:00:29 PM »

Quote from: stuntguy13 on November 13, 2009, 06:49:12 AM

Randy,

Interesting observations, one explanation of why the small prop has less wake could be that because there is less control input to make the turn the ship creates a smaller wake. It also seems that lighter ships would also create less wake, something to try.

Best, DennisT

Hi Dennis

I can "almost" confirm that too. I have flown many heavy and light SV-11s and Impacts, the heavy ones most always seem to make more turbulence than light ones

Randy

John Witt · « **Reply #26 on:** November 14, 2009, 07:15:41 PM »

As I understand it, you have to move a volume air equal to the wing load to hold the airplane up. In a square corner, the plane is pulling maybe 20 g times say 4 lbs = 80lbs of air. That's quite a large volume of air. Likewise, to turn the plane you have to react a mass of air equal to the plane's mass, polar moment of inertia and acceleration.

The density of air is 1.2 kg/cubic meter, so for a 60 oz airplane, you must react downward a 1.7 kg/sec of air to fly level, which is 1.4 cubic meters of air. No wonder it makes a breeze. The air motion will be in the opposite direction that the plane is moving (dow if the plane is moving up).

BTW, in pilot training when you fly light airpanes, wake turbulence is a big deal. Minimal turbulance behind a heavy jet until it rotates, at which point the wings start carrying the load and the vortices begin. So you try to stay above the wake at all times.

John

Igor Burger · « **Reply #27 on:** November 15, 2009, 02:00:50 AM »

Quote from: John Witt on November 14, 2009, 07:15:41 PM

a 60 oz airplane, you must react downward a 1.7 kg/sec of air to fly level

Right ... if you move that air one meter down ... frorce F= mass * aceleration ... so if you move that mass of air two meters per one second, you need only 1/4 of that and if you move it half meter you need 6.8kg

so how far your model reacts the air downward?

.... (hmmm great question anyway

)

Howard Rush · « **Reply #28 on:** November 15, 2009, 03:34:00 AM »

I did some cyphering, and it appears that Igor is correct, as usual. Work-- vortex energy in this case-- is force x distance. Induced drag is the amount of vortex energy an airplane puts into the air per unit distance along its path. Not counting induced drag due to lift to overcome gravity, the difference between the same airplane's induced drag going around a corner between the big prop and the little prop is proportional to the square of airspeed in the corner. So if the little prop allows the airplane to slow down more during the turn, the little-prop-powered airplane will put less wake into the air than same airplane with the big prop. Everything else cancels out. I figure (and please check this) that vortex energy per unit path length = induced drag = 2 m^2 V^2 / (pi e b^2 rho r^2) , where m = airplane mass, V = airspeed, e is the efficiency number = 1/the fraction of the airplane's induced drag over that of a wing with the same aspect ratio and an elliptical lift distribution, b= span, rho = air density, and r = loop radius. I doubt if the difference in prop wake energy between the two props matters much.

If you can do something to move the airplane's wake away from the maneuver path, the wake won't bother the airplane. For example, if somehow the prop blast entrains the wake and blows it away from the center of the loop, the airplane might not encounter the wake except at eight intersections, as Randy notes. So maybe this happens, and the little prop is better at blowing the wake away. I sure don't know.

John Witt · « **Reply #29 on:** November 15, 2009, 09:30:09 AM »

This business about turbulence at the intersections is interesting. Seems to me there are a couple of components. The airplane is rotating around the spanwise axis (nose up or down), so at the intersection first the rotation must be stopped and the second rotation in the opposite direction started, so a fairly sharp control deflection is put in. The old idea of short nose and tail arms to minimize the polar moment may be a result of observing this. So the energy required to rotate the current style long fuselage airplanes ( where's my Stuntwagon) is not insignificant, as evidenced by the turbulence. Second then the wing is now required to produce the force toward the center of the loop, so we are back to the v^2/r acceleration load for the wing.

So perhaps the judges should be looking for the bump as evidence that the pilot makes his intersections match. I know I can't help thinking that my flying is improving when I actually manage to find my own wake.

John

Howard Rush · « **Reply #30 on:** November 15, 2009, 12:43:33 PM »

You could calculate the power required to rotate the airplane and compare it to the induced drag.

I have felt that bump, but only with a wind component of a couple of knots from the left to match my maneuver drift.

RandySmith · « **Reply #31 on:** November 16, 2009, 10:06:52 PM »

It would be also interesting to see how much the Stab and elevators disturb the airflow..
Since the wing vorticies seem to roll in at one another, I wonder If Howard may be interested in designing a counter movement ,tip anti turbulator device that works opposite of the flaps???

Or a least something that can straighten out the vorticies.....

Could be a revolution in stuntship device design

Randy

Igor Burger · « **Reply #32 on:** November 17, 2009, 02:14:26 AM »

>>>It would be also interesting to see how much the Stab and elevators disturb the airflow..<<<
Nothing ... at least compared to wing. I always carefully analyze every model, so I can tell that while lift coefficient on wing goes close to 2, lift coefficient of tail is only 0.18. Means less than 1/10. Airfoil drag of tail is higher, (wing cd~=0.015 and tail cd~=0.04) but anyway, tail does only little of that.

>>> something that can straighten out the vorticies<<<

We can do LOT, but all of that is know:

1/ good airfoil - we are speaking abut airfoils in degning forum ... I wrote that we have to minimize drag IN CORNER and not in level. I analyzed my airfoil in corner and I designed all to keep him in good conditions:

picture shows airfoil in corner with flap at 30deg

I designed the linkage and elevater deflections to keep the airfoil in that low drag buble just under the lift cl=2 in the corner.
You can see that little exceeding can lead to far higher numbers of drag. Means deflections and weight, area, flap size etc must be well matched.

2/ proper planform will reduce induced drag, high aspec ratio will help, may be we will see some shift to high AR on electric ships soon 

3/ we can still use something like winglets which effectively enlarge the AR. = minimize induced drag.

Alan Hahn · « **Reply #33 on:** November 17, 2009, 07:48:01 AM »

Igor,

Out of curiosity, what do the polars look like at a small positive AoA (consistent with lift needed for level flight) but with the flap not deflected. I am interested in how much CL and CL/drag change when you "pop" the corner.

Also curious as to what AoA you need to get through a 10 foot corner.

I realize this doesn't come from your polar directly, but thought you might have some "back of the envelope" thoughts on the actual ranges the wing sees in a corner.

Igor Burger · « **Reply #34 on:** November 17, 2009, 10:38:40 AM »

Here is the same airfoil in level without flaps.

The Drag coefficient is little lower, but you must see that the difference in drag is greather that coefficiets indicate, because of induced drag.
While lift for level is somewhere at ~0.1, the lift for corner is 2. So if you calculate induced drag, which is = something * Cl^2 / AR, then you will see that the induces drag itself will be 400 higher in corner that in level. It is not big number, but anuway, my calculation for this modes says, that drag of wing in level flight is 1.3N and drag in corner is 2.8N. The same goes on with the tail.

Angle of attack is relatively difficult to speak about. It is not so clear what exactly is angle of attack, model flies on circular path and thus it is not clear what co call AoA. I calculated 25% of MAC as the refence point, then it looks like the AoA is somewhere at 6-8deg. However the corner is very dynamic, elevator is deflected slowly, and path of the model is also changing in time. Data of this model shows, taht instant deflection of elevator can make AoA up to 18 degrees, while model flying on circular path will stabilize at 8 degrees (due to circular air flow and thus lowered AoA on elevator) and continue on 3.5m radius. It is not so easy to say what and when and how.

Dean Pappas · « **Reply #35 on:** November 17, 2009, 10:52:16 AM »

Hello Igor,
Is this simulator from the Martin Hepperle site?
If not, where from?
Thanks,
Dean P.

Oh yes, I think you are right about the direction of E-Stunt design: we do not need to build in drag to fight the inadequacies of wet engines, and so we will build in efficiency "in the corners". If it comes along with higher drag in level flight, then who cares?

Alan Hahn · « **Reply #36 on:** November 17, 2009, 11:18:47 AM »

Quote from: Igor Burger on November 17, 2009, 10:38:40 AM

Here is the same airfoil in level without flaps.

The Drag coefficient is little lower, but you must see that the difference in drag is greather that coefficiets indicate, because of induced drag.
While lift for level is somewhere at ~0.1, the lift for corner is 2. So if you calculate induced drag, which is = something * Cl^2 / AR, then you will see that the induces drag itself will be 400 higher in corner that in level. It is not big number, but anuway, my calculation for this modes says, that drag of wing in level flight is 1.3N and drag in corner is 2.8N. The same goes on with the tail.

...<snip> ...

as a comment, using your extra induced drag of ~1.5N (2.8N-1.3 N) which is acting over ~0.2s (24 m/s speed with 3 m radius 1/4 circle or so will decrease the velocity of a 48 oz (1.4kg) ship by ~0.2m/s. This is just m*(v_f-v_i)=F*(t_f-t_i)--just the change of momentum=impulse.

This is relatively trivial in the big picture. Also may be consistent with my not seeing any noticeable velocity decrease in the corner on my airspeed indicator equiped eNobler. The major loss of speed is in the climb to ~45-60 degrees or so.

I haven't posted my latest flights since I have only recently (last flight actually) gotten the airspeed indicator and altimeter to "behave". Unfortunately I didn't use my video camera to capture that flight--it is very useful to synchronize the video with the data recorder to try and really nail down the start of a maneuver--which in a corner means at the level of 0.1 s or so.

Igor Burger · « **Reply #37 on:** November 17, 2009, 01:03:48 PM »

Yes Dean, it is Martin's JavFoil. Good tool for playing, because contains interactive tool for modifying, allow playing with boundary layer etc.

Igor Burger · « **Reply #38 on:** November 17, 2009, 01:14:56 PM »

>>>which is acting over ~0.2s<<<

And that is the point. I think it is easier to invest some time to proper design to minimize that drag and not to count that some magic device will be able control the motor in 0.2s. The complete drag in level is ~5N and in the corner ~10N. Means the dufference is in 5N during 0.2s. While the gravity acts minimally one whole second (5x longer) by force 17N what is over 3x more. So it is far far stronger braking. But that one second is enough also for relatively simple regulation which can fit to small PIC device. It is duable.

Alan Hahn · « **Reply #39 on:** November 17, 2009, 01:23:19 PM »

Quote from: Igor Burger on November 17, 2009, 01:14:56 PM

>>>which is acting over ~0.2s<<<

And that is the point. I think it is easier to invest some time to proper design to minimize that drag and not to count that some magic device will be able control the motor in 0.2s. The complete drag in level is ~5N and in the corner ~10N. Means the dufference is in 5N during 0.2s. While the gravity acts minimally one whole second (5x longer) by force 17N what is over 3x more. So it is far far stronger braking. But that one second is enough also for relatively simple regulation which can fit to small PIC device. It is duable.

I am not sure where you get the extra 5N in the corner, but assuming you are right, that still makes a trivial velocity change of less than 1 m/s. So I think we are agreeing!

In any case, this is interesting (even if the original thread is about the prop tips!), that speed loss in the corner is a non issue (unless one manages to really stall out the plane).

Igor Burger · « **Reply #40 on:** November 17, 2009, 01:46:03 PM »

Exactly ... IF airfoil works at good conditions, you can see on that my earlier polar, that if you exceed limiting weight, or you try to fly tighter radius, then drag coefficient can go to large numbers.

BTW thise numers I see in my analyze. It consists of:
wing 2.8N
tail 1.9N
fuselage 0.6N
landing gears 0.04N
lines 2.5N

and the rest is induced drag between parts

the same for level:
wing 1.3N
tail 0.5N
fuselage 0.6N
landing gears 0.04N
lines 2.5N

may be it is not 100% excact, but at least it is good number for further estimation ... I think 5N in level and 10 in corner is good estimation

Howard Rush · « **Reply #41 on:** November 17, 2009, 04:54:28 PM »

Very nice, Igor. Is the camber in the airfoil in the turn calculated from the turn radius or the pitch rate?

Igor Burger · « **Reply #42 on:** November 17, 2009, 05:03:29 PM »

Thanx, but I am not able to make complete simulation, so I do not know maximal pitch rate, it is just 3.5m radius, which I expect is the tightest controllable radius and all numbers in my calculation is related to that. Means I expect that the model can instantly enter and also escape the radius to the straight path and it does it at constant speed. It is simplification, but ... well my job is something else

Howard Rush · « **Reply #43 on:** November 17, 2009, 05:05:12 PM »

Quote from: RandySmith on November 16, 2009, 10:06:52 PM

Since the wing vorticies seem to roll in at one another, I wonder If Howard may be interested in designing a counter movement ,tip anti turbulator device that works opposite of the flaps???

Or a least something that can straighten out the vorticies.....

Could be a revolution in stuntship device design

Wingspan is the best trick anybody's come up with. Winglets can do almost the same thing, but the vertical surfaces would thicken the plot in sideslip.

John Witt · « **Reply #44 on:** November 17, 2009, 07:06:03 PM »

The tip vortices, as you all know, are caused by the high pressure under the wing rolling over the tip into the lower pressure area on top. This lead to the design of tip plates, but, I think I remember reading that most of the vortex evolution occurs in the free air space behind the wing, so tip plates don't help much.

Winglets extract some of the energy in that flow and turn it into forward thrust, but I remember reading that they work only over a narrow range of airspeed and angle of attack. Maybe a variation of winglets would really help the wake turbulence problem.

Then we'd have something really complicated to stretch the Ultrakote over.

John W

Dean Pappas · « **Reply #45 on:** November 17, 2009, 07:17:34 PM »

Howard, you are so right! tip plates on a Stunter would be a nightmare in strong wind.
Now I can't help but think about how to put useful side force generators on a Stunter. It's all your fault!

By the way, my quick thumbnail on the math says that in a corner at or near the max Cl, the plane might lose about 1/16th of its speed and 1/8 of its lift. The slowdown is negiligible near max Cl/Cd.

take care,
Dean

John Sunderland · « **Reply #46 on:** November 18, 2009, 03:33:26 AM »

I believe that a whale tail, ( airfoiled stab and elevator) 67:37 with less flap % and repeatable power in most all conditions with maybe a 50 rpm goose per push based on temp and ambient conditions and vector/velocity considerations with a factor for loss of lift variable + or - based on percentage airfoil or variable increase as a percentage of deflection from flap from root to tip ratio and that number.....to elevator ratio. 24 or 25 %.........We can make these decisions with a knife at the field. I'm sure Dean will straighten me out here!

Dean Pappas · « **Reply #47 on:** November 18, 2009, 08:04:29 PM »

I think it's worth trying, John.

Dean

Howard Rush · « **Reply #48 on:** November 18, 2009, 08:32:26 PM »

Igor, I get 52 Newtons of induced drag using my airplane's numbers at a lift coefficient of 2. That is just at the start of the corner before the airplane slows down. Does that look correct?

Alan Hahn · « **Reply #49 on:** November 18, 2009, 09:01:26 PM »

Quote from: Howard Rush on November 18, 2009, 08:32:26 PM

Igor, I get 52 Newtons of induced drag using my airplane's numbers at a lift coefficient of 2. That is just at the start of the corner before the airplane slows down. Does that look correct?

Howard,

On a 2kg ship, lift is on the order of 400N in a 20 g corner. So that would be ~8x your 50N induced drag.

Admittedly I am not sure exactly what I am looking at on these polars, but if Cl=2, isn't Cd ~0.014 (Igor's plot)? Wouldn't that make the induced drag of the wing ~2.8N (scaling from Cl to Cd) ?

RandySmith · « **Reply #50 on:** November 18, 2009, 10:20:01 PM »

Quote from: Howard Rush on November 17, 2009, 05:05:12 PM

Wingspan is the best trick anybody's come up with. Winglets can do almost the same thing, but the vertical surfaces would thicken the plot in sideslip.

Hi Howard

That is the reason I have my ships at the highest aspect ratio and the most span I can get within what works for me, Any higher aspect ratio and it is too much in high winds. I don't need any more lift since I already have much more than the plane needs now. What I have seem to be the best compromise in my experience.
The wing tips design on several of the SVs is also an attempt to help keep the turbulence cleaner, seems to me to work at least on a noticeable level.

I probably need to built a SV with a 6 to 6.5 to 1 AR for calmer weather contest.

Regards
Randy

Igor Burger · « **Reply #51 on:** November 19, 2009, 02:08:08 AM »

Hmmm ... that number 50N looks proper. I see I clipboarded my drag number from cell before adding induced drag ... sorry for confusion. Fortunatelyy someone reads what I write

, thanx Howard.

Yes it is right, 2^2 (lift SQR)/pi/5.5(AR) ~= 0.25

means the drag on 0.5m^2 at 25 m/s will be (1.2/2) * 25^2 * 0.25 * 0.5 ~= 50N

Anyway, the speed loss in corner 0.2s is still less than speed loss in climbing

Alan Hahn · « **Reply #52 on:** November 19, 2009, 07:56:32 AM »

I see I need to study my aerodynamics more!

Alan Hahn · « **Reply #53 on:** November 19, 2009, 09:27:31 AM »

Quote from: Igor Burger on November 19, 2009, 02:08:08 AM

Hmmm ... that number 50N looks proper. I see I clipboarded my drag number from cell before adding induced drag ... sorry for confusion. Fortunatelyy someone reads what I write , thanx Howard.

Yes it is right, 2^2 (lift SQR)/pi/5.5(AR) ~= 0.25

means the drag on 0.5m^2 at 25 m/s will be (1.2/2) * 25^2 * 0.25 * 0.5 ~= 50N

Anyway, the speed loss in corner 0.2s is still less than speed loss in climbing

If the model weighs ~60oz (1.7kg), then the airspeed loss would be ~6m/s, out of the original ~24-25 m/s. This isn't negligible.

I just don't seem to be seeing such a dramatic airspeed loss. Maybe I am not even making 10 foot corners though!

Here is some data from a flight last week where I think the airspeed and altimeter are nominally working. This is the wingover. The blue vertical line (at t=178 seconds) is where I think I hammer the plane (Nobler). The white trace is airspeed (mph-- right vertical axis for scale), the red is altitude (feet -left vertical axis for scale), and the purple is the power into the motor (kwatts--right vertical axis for scale).

A couple of caveats--I still feel a little uncertain about how well the sensors are working. For example the airspeed is indicating ~46mph level, while my lap times show ~54mph. The altimeter accuracy is something on the order of 4 feet. Also each tick is 0.1 second.

The max power into the motor coincides with minimum airspeed , which seems to occur a little before the top of the wingover. I wish the data was a little cleaner.

I need to go out and hammer some corners and use my video camera to synchronize when exactly the corner occurs!

Igor Burger · « **Reply #54 on:** November 19, 2009, 09:54:04 AM »

It is extreme pesimistic estimation I do not know if the spanwise flow can apear immediatelly and also if you need cl=2 for some radius, than you can do 10% larger corner what is hard to see at cl=1.8 what makes induced drag only 35N

Howard Rush · « **Reply #55 on:** November 19, 2009, 01:23:05 PM »

My calculation was pretty crude. I used the 5.25 sec/ lap cruise speed to calculate 52 N. Induced drag would decrease as airspeed decreases. It's proportional to the square of airspeed. Also, this is a pretty dynamic case, and I only understand the static case.

Dean Pappas · « **Reply #56 on:** November 19, 2009, 01:28:55 PM »

Hi Gang,
I went back to von Doenhoff and re-did the math. Sure enough a 650 square inch wing (0.4 sq-m) with an aspect ratio of 5, when pushed hard (Cl=2) will create about 50 newtons of induced drag. That could easily lead to a loss of 1/4 of the airplane's speed! The 50 newton figure is probably underestimated by 10 or 20% because I used the ideal case of an elliptical lift distribution.

Thanks for prompting me to drag the book out of the closet, Howard!

Dean P.

John Witt · « **Reply #57 on:** November 19, 2009, 06:06:06 PM »

An interesting set of wing tips on the racing Mustang "Precious metal". Now what would happen if you mirror-imaged those tips so you had one up and one down on each side.

http://stunthanger.com/smf/index.php?topic=14690.msg133449#msg133449

John W

Howard Rush · « **Reply #58 on:** November 20, 2009, 10:52:37 AM »

My guess is that you'd keep damaging them getting them in and out of the car.

Howard Rush · « **Reply #59 on:** November 20, 2009, 10:53:54 AM »

Quote from: Dean Pappas on November 19, 2009, 01:28:55 PM

Hi Gang,
I went back to von Doenhoff...

George Aldrich's cousin, by the way

Dean Pappas · « **Reply #60 on:** November 20, 2009, 11:33:06 AM »

Hi Howard,
Aldrich's cousin? No kidding! Now that's some neat toy airplane trivia.
Did this come up in conversation with George the departed or what?

This thread has turned into a thematic mish-mosh and I'm not likely to mess with it other than look for a way to rename it: the subject is good and it evolved quite comfortably. I take it that we all agree that if we could maintain airspeed throughout the corner then it would be easier to fly well.

later,
Dean P.

Howard Rush · « **Reply #61 on:** November 20, 2009, 11:52:03 AM »

Quote from: Dean Pappas on November 20, 2009, 11:33:06 AM

Aldrich's cousin? No kidding! Now that's some neat toy airplane trivia.
Did this come up in conversation with George the departed or what?

Drifting further afield from the thread's intent or interest.

Yes, back when he was alive. On one of these fora I mentioned Abbott and Costello's airfoil book, as it's known in the trade. Somebody came on and sternly corrected me regarding the author. Then George piped up and mentioned the kinship. Awhile later, I met a Roger von Doenhoff at work. I asked him the obvious question: whether he was related to George Aldrich. (When I met Curt Nixon, I similarly asked him if he was kin to Mojo.) Roger didn't know George, but admitted to being Albert's nephew. I put the two of them in touch.

Dean Pappas · « **Reply #62 on:** November 20, 2009, 01:18:41 PM »

Leaving Skid Roper out in the weeds, yet again.

Dean

John Sunderland · « **Reply #63 on:** November 22, 2009, 11:51:04 PM »

Quote from: John Witt on November 19, 2009, 06:06:06 PM

An interesting set of wing tips on the racing Mustang "Precious metal". Now what would happen if you mirror-imaged those tips so you had one up and one down on each side.

http://stunthanger.com/smf/index.php?topic=14690.msg133449#msg133449

John W

We tried this on a Coyote back in 95....it was a hog in the wind.

Author Topic: Prop tips and influence on thrust (Read 9151 times)

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn

Alan Hahn