- This topic has 18 replies, 8 voices, and was last updated 14 years, 5 months ago by
.
-
Posts
-
I sent this to SEN, George Batiuc, Paul Crowy and Buzz Lightyear. JIM
USA Team Selection Committee,
We have not seen Dallas on the flying field lately but he and his senior design team at CSUN have been at work. After the last finals, Dallas and I conceptualize a new bunt method that would be much more toleration to release speed and attitude. He was able to use this as his senior project. Now 2 weeks from the finals, it is time to reveal the iPlane! The name was given by the design team and it stuck, I would have preferred iGlider. As you can see from the photo on the NFFs forum, the college kids used the computational power along with the GPS and attitude capability of the iPhone. The iPhone maintains all it’s phone capability, it plugs easily into the fuselage. The team did added a speed module used by the RC people. The mechanics and flying surfaces are M&K long with impulse hook. The model is a bit over weight (34 grs) but initial tests look promising.
The question for the TSC is this. While the model is being accelerated for launch, the iPhone collects attitude pitch, roll and yaw) and speed data and performs real time processing, continually establishing the optimum stab and rudder positions for the launch / bunt. At the time of line release, the processing stops and the last program (rudder and stab servo positions relative to time) is used by the model for the bunt. DT via standard timer or RDT.
Dallas and I believe this meets all the rules, ie this is not a closed loop control system because the processing stops just before the official timed portion of the flight begins. I’m requesting a TSC ruling on this at or just after the livotto’s contest. If the TSC rules the model is not legal, I would not expend any more energy on it before the finals.
Respectfully, Jim Parker
Sorry, I hope I figure this out soon. JIM
Jim,
Whilst I applaud what you and Dallas have achieved this is using rather old technology (and WHO would use Apple euipment when there are such good MicroSoft applications about – but I digress).
I had similar thoughts a few years ago but immediately saw the advantage of these new smart fabrics. You know the sort of thing, used for iPod and mobile phone connectivity to your coat (or shorts even). Anyway, I found someone who could incorporate a suitable PIC processor and figured you could add the sensors to a sytem built into your penant (all availble from Denmark would you believe !). After much work I had a prototype which worked like this :-
1) Vasilly re-latchable hook working as normal but additionally arms the system.
2) As the penant steams, senors pick up the ambient conditions and attitude of the model well away from the model’s influence.
3) Now the neat trick. If conditions (ie no thermal) or model attitude are poor, then the hook re-latches and the penant sensor continues to gather data. When both you and the software are happy you can launch with optimised trajectory as you describe.
4) The penant hardware uses the model’s battery so it not active when on the ground and so processing is not an issue !One of the hardest problem is the colouring of the material. As you know I live in Scotland and there are matters of decorum. So far I have only been able to achive an Edmonson tartan but soon I hope to have the Ancient Hunting McGregor style you have come to expect from me.
The other point is keeping the sensors clean. At our recent team selection meet I had two gnat stikes directly on the pitto tube ! The software thought all was OK and released the model thus explaining my poor scores in Round 2 and 3. More work is of course needed.
CHE
This is extremely exiting stuff. I can’t wait to see the system Dallas and friends have put together.
Jim: maybe you can get Dallas to post some details including pictures?
FWIW, I have also experimented with accelerometers, although i didn’t put an entire phone in my models. I have developed a small dedicated datalogger, which in addition to the usual altitude and temperature, also records acceleration in all 3 axis.
The first prototype used an accelerometer with a peak sensing range of +/-6g, which turned out not to be enough, the sensor peaks (clamps) during the launch. so for a while, i used this only as an altimeter.
recently, the accelerometer component became available with an increased sensing range, now +/-18g, and i put together another prototype. i hope to be able to fly this during the upcoming week in lost hills and can’t wait to see the acceleration graphs for a typical launch.
for anyone who is interested, here are some pictures of the unit and how it interfaces to the iPhone. i have also written a graphing application that displays altitude, temperature and soon also acceleration on the iPhone.
http://gallery.me.com/rlimberger#100078
http://gallery.me.com/rlimberger#100160
if the new sensor works out, i plan on using it as an addition to my current servo hook system, to provide the electronic equivalent of the inertia hook used by Jim, M&K, etc.
-r
ps: before certain people get carried away (again), NO, this sensor is NOT connected to the timer and does NOT participate in a closed loop system.
OK– the collage kids had to set up my aol photo gallery. Here’s the link to the iPlane photo.
Its all about October. Thermals, JIM
i will definitely protest this! this is yet another example how Steve Jobs destroys our sport! first carbon fiber, now this…
I think I will wait till the Blackberry Storm comes out at then end of the year to make my modifcations. Better global capabilities.
Jim,
As you pointed out on SEN, that once the model is relased by the helper, it’s considered to be air born and all the FAI rules apply, albeit still being tethered to a tow line. The iPhone, if I understand correctly, uses on-bord sensor information to calulate the optimal stab’s angle of attack during OLA. It’s definitely a meritious student project! But, it probably violates the FAI principle of prohibiting closed feed-back loops on-board to effect the model’s trim. RDT is a different creature – performing a non-reversable flight terminating operation. This, of course, reflects my own personal understanding of the rules.Rene,
Steve Jobs created carbon fibre ? I think not !
Chris, Steve invented *everything* don’t you know that? 🙂
On a more serious note, and in an attempt to turn this silly thread into a more useful discussion, i was wondering about the “closed control loop” rule.
while it makes sense in it’s intent, i wonder if the actual circle tow is already a violation. couldn’t one argue that circle towing the model is a closed loop? …where the control flow is like such:
1. pilot senses flight behavior (eyes)
2. pilot makes control decision
3. pilot issues control action (alter tension on towline)
4. hook position changes, which causes control surface change
5. flight behavior changes
6. start at #1maybe i am missing something here but isn’t that a closed control loop?
because of the sketchy interpretation of the “closed control loop”, i will only do straight tow from now on – to avoid any risk of being protested.
-r
Really! Obviously the flier is controlling his/her towed model through line tension. In contrast, closed feed back loops use inputs from on-board sensors to induce changes in the model’s surfaces (wings, stab or rudder). The iPhone, with it’s sensors, clearly falls into this category.
The real question, however, is when does the flight begin– either at the instant the helper releases it or when the pennant drops off the tow hook.
You might recall that about six months ago, on SEN, I raised the idea of using a radio signal to release a model over head without a bunt if it was in an enourmous thermal. (Powerful thermals can make bringing a model around rather tricky). My premise was that during towing the model is not considered to be flying (the clock has not started), so that a RC signal was legal. Well, a lot of people thought this was a really bad idea, and I retracted it after realizing that fliers, using RC, could keep the line attached to the model if they did not like the launch or the air – and earn a second attempt.
Jim Parker pointed out that a towed model is considered as being in a state of flight (supported by it’s lift) although the clock is not yet running.
This point can be illustrated by absurd counter examples. Suppose the flight begins only when the pennant drops – then no FAI rules apply during towing. One could, for example, enlarge or reduce the model’s area above or below 32-34 dm^2; apply external propulsion systems (rockets?) during OLA or even have a helper, using RC, fully control the model during the OLA. (My RC line releasing proposal above definitely falls into this category). I doubt if anyone would consider this to be free flight anymore.
The fact that no thermal detection equipment can be attached to the tow line also support the idea that towing should satisfy FAI rules.
Aram,
Why assume that FAI rules necessarily would not apply just because the flight hasn’t started? In other words, what’s so inconceivable about keeping the same rules we have for towing without calling it part of the flight? Or add or modify rules as necessary to prevent RC or CLC or whatever is considered bad behavior. I must be missing something that’s painfully obvious because every time I bring this up it kills the conversation thread.
Jim,
In terms of on-board iPhones, establishing when FAI rules apply is essential. As the FAI rule book does not explicitly state that the rules only apply from the instant of launch (pennant detachment) this has led some, including me, to erroneously exempt the tow period. The absurd examples listed above illustrate what would happen under this misinterpretation.RDT is a different issue. Now that it has become legal outside F1C, people are concerned about its usage as a back door for getting a second chance under the current 20-second grace period. RDT could be made riskier by shortening the grace period – an idea I floated in the latest SEN (#1248). Another possibility is to explicitly ban RDT as a way to get a second chance for non-power models. The possible collateral damage by F1As in a cruise setting (zero decalage) or an F1B model (beginning of the burst, VIT setting) is rather small – given their velocities and weights. This debate is currently raging on SEN and might trigger a FAI rule change.
Aram,
I’m not making my point clear enough. You stated: “Suppose the flight begins only when the pennant drops – then no FAI rules apply during towing”. That appears to be correct as the rules currently exist. My question was if you’re talking about a rule change making the glider flight start when the watch starts, then for the reason you stated, make another change saying that the rules apply to “towing”. I don’t see the problem with that, but again maybe I’m missing something.
I’ve also followed the raging discussion about the 20 second attempt, 5 sec attempt, bad launches by your spouse, cheating, etc., with great interest, and even put in my 2 cents worth with the result previously mentioned. Those comments were aimed more at pointing out that a consequence of divorcing the tow activity from flying by rule is that the word “attempt” would no longer apply to towing, meaning the 20 sec attempt would only apply to the flying part, same as rubber and power (I know you know this and pointed it out). No attempt, then no limit on number of tows or crashes or bad launches by your mom or brother, which seems like a good thing to me. Of course the rules on towing could say you only get two tries. But why? Does it hurt me if CHE tows four times not finding a thermal and get’s exhausted and dehydrated? OK, please fire away.
- You must be logged in to reply to this topic.