Engineering the TOADCam

[Padesatka]

Or towed cam is a video camera "glider" sytem. Consider it the aerial trailer, which I intend to one day put to full-scale applications. My aerobatic airplane is overpowered for cruising flight, yet has no room for my cooler of beer. Until I tow it in a sleek "trailer" that is jettisonable, even though it goes wherever the plane goes, except in a spin or tailslide (jettison before such maneuvering, and please safely drop and collect your litter).

Anyway, at the cheaper scale of RC aircraft, there is the ready capability to film always-locked and stabilized gunsight footage of any model with the excess thrust. Like all good things, there is attached a penalty, which is a bit of a drag.

I would welcome any assistance in designing and maybe building this non-existent but interesting thing. I visualize two towropes attached to the wingtips of the subject, or tow plane. We are trying to comunicate roll and yaw information through the two lines in order to achive the result of a vehicle that automatically flies a loose slot formation.

Think "control line" with me for an instant. The bellcrank. Now, we mechanically drive ailerons with it, see? -and tune the vehicle to follow the leader with the most pleasing delay in roll input. Through yaw springs on the bellcrank that softly marry the glider fuselage to the towropes, yaw is communicated. Pitch is communicated through a spring damper.

The whole pachage looks to me like a sleek, stubby flying wing with a camera behind a clear plastic bullet-nose. The towropes attach to bridles that enter the leading edge of the wing at about 1/4 span.

Or maybe it should look different, fly different. Maybe this is impossible. All I want to know is, what do you think?

[Padesatka]

OK maybe it needs a gyro. I need a gyro myself now and then.

I'm pleased with this thread so far. Let's keep these ideas coming, gang!

http://static.flickr.com/45/114311317_0b938a3805.jpg

edit: Here's a sketch

[Sparky Paul]

The towed planes I'm familiar with have a long tow rope, with the towed plane flying its own path due to its aerodynamics.
A short tow rope might get into oscillations, and a tow rope which moves surfaces on the towed vehicle -will- get into oscillations!
I believe the distance between the two planes reduces the effect of the towed plane trying to fly its own path due to its stability parameters, as these are damped in the tow rope before they can affect the towing plane.

[Padesatka]

In WWII, manned gliders were "automatically" flown on tow, for hours, and in instrument conditions, without much if any glider pilot intervention. With one single tow rope, these crates actually did follow-the-leader mechanically and without pilot help in cruise.

With a quick Googling, you can learn that the tow ropes were not proportionally very long. And if you viscerally imagine "being" the camera platform I'm writing about, I think you would prefer a shorter tow rope if those 2 lines really became (in realistic visualization) your only input- Your whole experiential world as a trailing camera pod.

There is enough information and propulsion available from 2 wing-tip tow ropes to follow the leader almost anywhere. That is, through any forward flight. With ostentatious and unecessary thrusters, the "towed" vehicle could of course keep tension on the lines through any condition or maneuver, opposite the inertial path of the towplane. But I don't think it's necessary for some spectacular flying that leaves out zero -airspeed, spins, and tumbles.

"and a tow rope which moves surfaces on the towed vehicle -will- get into oscillations!"

If left undamped. Elastomers, springs, programming, even rate gyros can surely smooth that out. The near-term benefit is spectacular videography. Scaled up, Imagine small cams trailing a Zlin 50 through her paces at a Red Bull demonstration.

And if it all goes wrong, the tow lines have stretched-out parachutes in line. Worst case: Flutter, lock-out snap roll on tow? No problem: When any high tension causes the weak link at towplane end to let go, then the sled comes down under 2 canopies... and the towplane flies on. This may be the routine before landing. For takeoff, the vehicle is carried inside, along with the spooled towline, to be released in flight out the hatch along with the camera pod. Because the weak links are relatively weak, we can route the towlines along the subject planes wings with small bits of masking tape, so that deployment to trailing mode happens predictably. This involves the same simple concepts used in wire-guided-missile launch and control. I'm thinking of spooling the lines right around the bullet-nose.

With a minimal towed cam applied in the full-scale world, the pilot would probably never notice the drag change, but could release the trailing vehicle at will. Pilots must understand that any tumbling maneuver, spin, tailslide, is a bad idea with the tagalong attached. The fatal zooming camera closeup on the truly stupid pilot, or impact on some other critical aircraft component would be in poor taste.

But enough dreaming, let's get it flying. I'm prepared and inclined to build a prototype considering any and all advice given here.

[Sparky Paul]

Since you've worked it all out, have fun.

[Ensignnolo]

The view from it might look something like this, depending on how close to the ground you got!

http://www.afrlhorizons.com/Briefs/Jun02/MN0203.html

Ensignnolo

[Padesatka]

"Since you've worked it all out, have fun."

But I haven't! Still brainstorming, and open to ideas.

"it might look something like this,"

Thanks, Ensignnolo, that's right. All I need in addition to what you linked is the the maintenance of roll orientation with the towplane. I think this would be more aesthetically pleasing with a slight delay, proportional to airspeed, in the control inputs- A virtual camera view from a chasing aircraft.

I'm hopeful that there is an elegant solution enabling smooth control of a towed glider passively through 2 tow ropes.

Pitch input through the tows is simple, with relative position from CG the primary variable. I'm hoping good pitch input can be made from gromets in the leading edge of each wing. There is actually a lot of postioning flexibility there, with taper and sweep still unestablished. Then there are the variables of tow length and CG.

Yaw, or differential force on the towropes might be best mixed into rudder and elevon. With a programmable interface, the possibilities are there for a towed camera platform that can fly any trailing position, like a slalom waterskiier behind a motorboat.

But I'm hoping to arrive at a design for a mechanical mixer that will provide the gentle inputs necessary to follow the flight path of the towplane, and damp any excursions from it.

Another variable is towrope elasticity, not only for issues of slack rope but control lag. This should be part of the damping or passive stability scheme.

A more fantastical design, single towline, would just lock in on LEDs on the towplane's wingtips, and a processor in the trailing pod would fly hopefully flawless visual aerotow. But that's beyond my budget right now. I'm hoping to come up with a trailing aerovehicle that inherently, passively stays roughly in a stable tow position while the lead/subject aircraft maneuvers.

[Padesatka]

It's of little advantage for the pod to be very sleek. Light in weight and minimal drag, but of low mass to surface area. Maybe a flying tetrahedron, because towline tension is so important. So maybe a big stick-construction wedge shape, something like the first third of the Facetmobile (http://members.aol.com/slicklynne/facet.htm), with blunt trailing edges, would be a good airframe for prototyping. Something like a badminton birdie with roll control.


http://members.aol.com/slicklynne/FMX4IF1.JPG

[BMatthews]

I suspect that a large part of any success in this method would center around flying the tow plane in a very gentle manner that involves very large radius turns. The reference to the full sized gliders being towed for hours at a time is most likely while on cross country transit trips where no turns were made other than minimal course corrections.

The 3 or 4 line control inducing "control line" system may be the answer to slightly more aggresive flight paths but I still think Paul is right about this being very much an issue. One very big problem is how to communicate rolling inputs since a roll would simply wind up the control/tow lines. It's for reasons like this that you're going to be limited to gently banked turns regardless of any control inputs or damping or other solutions.

But within those requirements I suspect that a follower type craft would be possible.

FWIW you do not need wingtip to wing tip

[Padesatka]

http://static.flickr.com/41/114713274_34761f155a.jpg

Gentle? The heck with gentle. But I do understand the need to avoid spins, slides, tumbles and the like until the smarter, powered free-flight version evolves. That one would "watch" reflective strips stuck to the subject plane, to lock on to any formation position, or pan between. Somewhat useful in flight testing.

Roll would be used largely for stability, and for intentional variations with the subject. This towed camera platform is rollicking along back there, but always hanging in stable because within the subject's flight envelope right up into high AoA, it is inherently so.

So in the most basic form the Klingon (as I now think of it) chases your tail by knowing where your wingtips are up ahead. There is instantaneous information regarding acceleration/deceleration of the subject/leader. Also instantaneous yaw information. Lagging pitch, lagging roll.

Seems like enough to work with using mechanical mixers no more complicated than mech v-tail mixers.

I'm going in the direction of ballistic shuttlecock for the airframe. If the effective surface drag/mass is less than the subject, the towed camera pod can only have an impact with the subject if falling into the subjects burble in stalled or tumbling flight. In other words, slow-deceleration stalls could probably be filmed.

Imagine being a flight student who can review every lesson from a DVD offering interactive perspective changes from within and outside the aircraft.

Considering airlineand aerospace markets, the applications for an inflight inspection (damage control?) from a deployed "little buddy" are interesting.

But more implementable for me right now with consumer RC planes, radios, and cameras. And groups like this. Thanks for the input!

[Padesatka]

Your link is insightful, Ensignnolo because thinking about this as something of a guided bomb shape is helpful. In reaction to the dual towrope (actually looped through the camera pod) the pod maintains position with 3-axis elevons. The fixed ring provides drag and ballistic stability, and has toed-in surfaces: It's a bernoulli ring that could possibly resist flying in the wake of the towplane by "feeling" it and "bouncing" back out to the intended position relative to the wake.

Rather than avoiding wake interactions, maybe the pod can be shaped to passively bounce out on contact with the jiggly (bad ride, bad video) wake. The whole concept is tow-it-and-forget-it along with very simple and fail-safe hookup to the towplane.

[BMatthews]

Sorry, double posted.

[BMatthews]

Gentle? The heck with gentle. ....

..... Lagging pitch, lagging roll.....

Actually your Klingon concept may be the way to go but with a few corrections. With the high degree of vertical area the trailer would not need to roll at all. It would have enough vertical area to "knife edge" it's way around the turn without banking. Just don't start rolling the towplane... :D

Your comment about lagging is where it all comes to a head. The problem is that the trailer needs to follow in the travelled path of the towplane and not to react at the same time as the tow plane. Otherwise if the tow plane turns a typical model sized 90 degree turn you would find the trailer wanting to turn 90 degrees at the same time and suddenly it's trying to fly alongside the towplane rather than politely behind it. Of course it won't be THAT strong an effect but the idea is still there and this is largely why you can't get crazy with the idea of the trailer being flown by the towplane maneuvers directly. Think of a long ladder firetruck with the rear steersman and what happens when in the old Keystone Cops movies when the back end suddenly goes wonky. To avoid this the trailer needs to follow more or less the path of the towplane rather than maneuver at the same time as the towplane. So the inputs need to either be delayed by the amount of time required for the trailer to reach the same point as where the tow plane started doing whatever. But at the same time the towplane is already pulling off in a different direction since it has already dove or turned or whatever a second before. All of this is why it's just practical and may well not even be possible to fly aggresively with the trailer attached.

Your Klingon concept shows how fore riggers can be used to communicate both the pitch and yaw to the trailer. What about taking that one step further? One line off the tow plane to a similar trailer but that has only the one fore guide. Set the lift high and the center of mass low so the trailer does not need to roll or bank but just tows happily along. The fore guide then needs to be adjustable in pitch to set the best angle of attack for the trailer for a given speed range and to allow for any line sag. But it would provide the steering needed to place the trailer in the right path.

If it was alters to move the pod down a little within the oval wing that would take care of the stability issue. Then swap the twin guides for a single attachment point off the top of the nose on the pod. An extension boom coming forward a bit more would enhance the corrective inputs. Longer would be a stronger effect and shorter would be less leverage. The tow line would also provide the tracking stability for the trailer such that it would not requre any tail surfaces. But the CG of the trailer would still need to be about at the 25% chord point of the wing to work. It's still meant to provide lift after all. The way you showed it in your sketch it'll tend to try to lawn dart itself. You still need to make it follow the same rules as the airplane.


I think that's about as good a solution you'll get without getting into a heap of complex solutions.

[Padesatka]

Excellent thoughts, Bmatthews. I'm pondering them...