stunthanger.com
Design => Engineering board => Topic started by: Peter Germann on January 08, 2017, 03:32:22 AM
-
To determine the gyroscopic force (Precession) induced by the rotating elements of an F2B drive train, Wolfgang Nieuwkamp suggested using a rotating pendulum to define the combined moment of inertia of rotor (crankshaft), spinner and propeller. He furthermore offered to calculate the resulting precession torque in level flight of a typical F2B airplane.
I have built such a device consisting of a triangular base plate (6.5 Gr. / 0.23 oz) being suspended by three thin textile threads of 6 ft. (1.8 m) length. The components installed on the baseplate were:
• Rotor (bell) of a Hacker A 40 motor (72 Gr. / 2.54 oz)
• Spinner and prop driver Tru-Turn turbo cool 2 in (35 Gr. / 1.23 oz)
• Propeller APC 12 x 6 EP (28 Gr. / 1 oz.)
When excited to rotational oscillation of +/- 15 °, I was able to take the time for one full (1.45 sec) cycle which I then sent to Germany. Based on above component weights, a lap time of 5.2 sec and 9’500 RPM, Wolfgang calculated a value of 0.18 Nm for the precession torque resulting from use of the those components. For clockwise tractor prop flying the precession torque generates a nose up force and with pushers it forces the nose down.
-
Interesting method. :)
The equations can be found here:
http://www.me.utexas.edu/~me244L/labs/filar/filaroverview.html
OT, but while googling for them (and not knowing what to call it) I also came across something called a "harmonograph" that is kind of cool.
https://www.youtube.com/watch?v=63TK_-defac
It might be interesting to do the same test, but with a crankshaft and some typical "wet" props instead of the rotor.
-
From testing done years ago by WB Netzeband, Pete Soule and others, on a wet motor most of the precession comes from the prop. Typical prop shapes of the time, mostly wood, didn't make much difference, but lighter props are much preferred. The current APC props put the mass a bit closer in which is helpful. From testing Mark Rudner has done, the outrunner can/magnets may actually be generating more force than the prop. They were test for eF2D at high rpms. One motor, a Turnigy 28-36(a relatively lot of weight for its size) made the plane uncontrollable. He's tried an inrunner motor with better results.
-
Nice, so last time calculation was .270Nm for 24g prop and 9000 rpm:
http://stunthanger.com/smf/stunt-design/stab-incidence/
so recalculated for weight and rpm which you measured it gives .270/9000*9600/24*28 = .336Nm
since prop and motor moment inertia was aproximately 10:1 it gives aproximately .3 for prop and .036 for motor
I expected overestimation for prop formula factor 2, means it should not be .3 but .15 instead
so now my value .15 + 0.036 = .186Nm
and you measured 0.180Nm ... means I was hell good ;D with my estimation of mass distribution in blade and finaly we have calculated and measured numbers matching
-
So now the fun starts... The typical D over q for wheel is .25, say .2 for a faired wheel and strut. Let's say each fairing has a frontal cross section of 2in.^2. Multiply by two and then by the distance of the gear from the CG and what's the nose-down moment due to the landing gear at F2B dynamic pressures?
-
You are certainly doing a great job of quantifying this fact.
CL planes set up the normal way definitely precess inward on down control.
The Al Rabe rudder control was a good way to deal with this.
Anybody who has done aerobatics with a real prop-driven airplane knows it's real.
Maybe you CL electric stunt designers could add a second motor driving a flywheel the opposite way to neutralize the main motor.
-
Maybe you CL electric stunt designers could add a second motor driving a flywheel the opposite way to neutralize the main motor.
Maybe, but about what axis, and with what angular momentum? Maybe not what you think.
-
Maybe you could put the flywheel under the real motor and make the front end look like a real engine plane. \
You're aero engineer. I'm just a model flyer.
-
Nice, so last time calculation was .270Nm for 24g prop and 9000 rpm:
http://stunthanger.com/smf/stunt-design/stab-incidence/
so recalculated for weight and rpm which you measured it gives .270/9000*9600/24*28 = .336Nm
since prop and motor moment inertia was aproximately 10:1 it gives aproximately .3 for prop and .036 for motor
I expected overestimation for prop formula factor 2, means it should not be .3 but .15 instead
so now my value .15 + 0.036 = .186Nm
and you measured 0.180Nm ... means I was hell good ;D with my estimation of mass distribution in blade and finaly we have calculated and measured numbers matching
BTW, you can skip the bifilar pendula and use Pete Soule's approximation that gives you a radius of gyration of .225* diameter. I thought it was far too hand-wavy to work, but I measured a wide array of props and it's surprisingly close for a large number of cases (of constant-modulus materials, of course - those crazy guys with hollow props might have to keep measuring with a torsion or bifilar pendulum, or finite-element-models like back in caveman days).
I did the same thing back in the day (2004) and got something a little larger that yours for my particular set of parameters, about 33 inch-ounces (or .233 newton-meters in olde-worlde units). My RPM was 11,000 rpm and my inertia was 1.28x10-4 slug-ft2, and physics hasn't changed since then.
Brett
-
Maybe, but about what axis, and with what angular momentum? Maybe not what you think.
I briefly considered using a control-moment gyro with a spring and a damper to create artificial yaw damping, but you could use it in any axis you thought you needed, I suppose.
Brett
-
Let's not get too whacked out over the gyroscopic precession. Remember, what matters is not the just the precession, but the total sum of all the moments on the airframe. So if the precession is nose up, you can move the wing slightly below the fuselage center line. Then you can move the thrust line above the centerline. Then you put the wheels on the bottom and they create a nose-down moment.
So basically, the modern typical Nobler-derivative planform has it just about worked out, LOL!
To say that there's a nose up moment for precession in level flight so I need stab incidence is way oversimplifying it.
The dynamics of a CLPA aircraft are incredibly difficult to quantify. Until someone is willing to give me and some of my fellow aerospace engineers here a grant for several million dollars, supercomputer time for CFD and some dynamic analysis at Calspan we're all pretending we know.
And with that said, there is, and always has been - a lot of snake oil in this hobby. Head games too.
My $.02 is that if you build a straight-as-@#$% airplane of a proven design, build it light with the correct control throws and have a reliable, correctly sized power plant... you're pretty well there. From that point it's about trimming and what you do with the handle.
All in my humble opinion and I will respect and defend your right to disagree because we're all in this together,
Chuck
-
Let's not get too whacked out over the gyroscopic precession. Remember, what matters is not the just the precession, but the total sum of all the moments on the airframe. So if the precession is nose up, you can move the wing slightly below the fuselage center line. Then you can move the thrust line above the centerline. Then you put the wheels on the bottom and they create a nose-down moment.
So basically, the modern typical Nobler-derivative planform has it just about worked out, LOL!
To say that there's a nose up moment for precession in level flight so I need stab incidence is way oversimplifying it.
If you had bothered to read the 2000-word description of the issue I wrote for SN about 15 years ago - where I noted the other asymmetries you mentioned - you would see that even with all of that, they still generally need some down elevator (built in or rigged in).
Brett
-
If you had bothered to read the 2000-word description of the issue I wrote for SN about 15 years ago - where I noted the other asymmetries you mentioned - you would see that even with all of that, they still generally need some down elevator (built in or rigged in).
Brett
Cool. Do you remember which issue?