Author Topic: In flight speed control (Read 2406 times)

Larry Wong · « **on:** October 20, 2009, 10:29:49 PM »

I found this to be intersting. #^Don't forget to see video click red button..

http://www.clacro.de/Seite_timer.htm

Alan Hahn · « **Reply #1 on:** October 21, 2009, 07:47:10 AM »

It looks downright subversive!

I couldn't watch the movie---the codec apparently isn't supported on Macs (typical Windows stunt!).

One thing I was wondering is what line tension do you want. Remember when you are truly vertical, gravity lessens the tension by one "g", so the plane might now try to fly even faster than it does on level laps. I am not sure that is what we really want or not. Also I may not have read the text closely enough to get the answer to that question.

Igor Burger · « **Reply #2 on:** October 21, 2009, 07:55:40 AM »

exactly, strongest acceleration is on top of hemisphere, while we need it just after the turn up, means already at level, while overhead should be alreday back to level power or even less beceuse of upcomming descending :-)

it just reacts too late :-)

Alan Hahn · « **Reply #3 on:** October 21, 2009, 08:51:40 AM »

More than you want to see I bet, but here is an over-simplified physics calculation of what is going on in a wingover (first half from pullup to first pullout. What is over--simplified is that I have NOT included any speed loss due to the sharp corner, and left out any type of outward thrust that might be present.

Also I am not sure about the actual level lap thrust. Here I am using a value that a propeller calculation gave me (from my "Pointy-head alert" thread of a couple of years ago). I am pretty confident about using the thrust vs airspeed part of the plot, but its the absolute magnitude that I am not sure of.

I assume that the drag at 24m/s (=54mph) just equals the level lap thrust, and then scale the drag as the square of the airspeed.

The graphic includes 2 plots, the left side one is the quantities vs the angle in the climb (0 degrees is level lap height at the beginning, 90 degrees is overhead, and 180 degrees would be the pullout height (but remember, I don't model the corners at all in this plot). The right side graph are the same quantities as a function of time. Note the Line tension uses the scale on the left vertical axis, and all other quantities use the scale on the right vertical axis).

Just for a quantity check 4.5 N (my level lap thrust) is about a pound of thrust.

I plot Line tension (white), airspeed (red in m/s)), Altitude (green in meters), Thrust from prop (blue in Newtons --N), Drag (yellow, also in N), and total tangential force (violet in N---just = (thrust-drag+-component of gravity in direction of thrust).

As Igor mentions, notice the maximum decelerating force (violet trace) is at 0 degrees, where gravity is pulling all its weight against you, and drag is still high and thrust relatively low. Also note maximum drag is at the pullout where airspeed is the highest.

Enjoy!

Igor Burger · « **Reply #4 on:** October 21, 2009, 10:45:59 AM »

great ... now do some steps forward

if the prop thrust so nicely reflects the air speed, and thrust is is very well visible on shaft load (induced drag on prop blades due to slip ratio) then you can easily measure speed on motor power delivery ... and also esily compensate sliping ratio by higher rpm - means little positive feedback from power to rpm will do what we need

exactly that does this my solution:
http://stunthanger.com/smf/index.php?topic=12673.0

and here is how it looks in air:

Bob Reeves · « **Reply #5 on:** October 21, 2009, 12:38:46 PM »

What am I missing.. Sure looks to me like it's slowing down going up and speeding up going down.. Isn't this just the reverse of what we want?

Alan Hahn · « **Reply #6 on:** October 21, 2009, 01:08:22 PM »

It may be the reverse, but it may also be the truth!

My goal is to try and actually understand what our planes are actually doing.

I just reinstalled my altimeter and airspeed indicator in the eNobler. I have a cool new pitot tube to try out. Hopefully it will do a better job than my old setup which is shown here http://stunthanger.com/smf/index.php?topic=7574.0

Here is a gif of the pitot tube. The forward direction is to the right. The static pressure line comes off the angled tube, and the ram pressure off the center tube. I have branched the static line to the altimeter and to the static input of the airspeed sensor (these two things are basically identical pressure sensing devices). Hope to have some more fun soon!

Igor Burger · « **Reply #7 on:** October 21, 2009, 01:15:24 PM »

Bob, are you speaking about my video? Do not you hear sound of the motor? It is clearly to hear how it gets higher rpm uphill.

You can even hear how it gets higher rpm already in first corner of hourglass due to high drag in that corner (model does not have flaps).

Alan Hahn · « **Reply #8 on:** October 21, 2009, 01:44:17 PM »

Quote from: Igor Burger on October 21, 2009, 01:15:24 PM

Bob, are you speaking about my video? Do not you hear sound of the motor? It is clearly to hear how it gets higher rpm uphill.

You can even hear how it gets higher rpm already in first corner of hourglass due to high drag in that corner (model does not have flaps).

Sorry Igor, I thought Bob was referring to my plot, not your video.

My comment is that 99.999% of stunters (you may be that .001!) do slow down when climbing, and speed up when diving. The main issue is how much! That is what makes the difference--I think!

Bob Reeves · « **Reply #9 on:** October 21, 2009, 02:59:06 PM »

I was watching Igor's video without sound on.. the airplane is noticeably slowing up hill and speeding up on the down side. With sound on, motor is working well, maybe this is just proof of concept?

Alan Hahn · « **Reply #10 on:** October 21, 2009, 04:06:27 PM »

I think it is hard to actually NOT lose airspeed when you turn the nose up. I base this on some shakey calculations though.

If you figure you are making, lets say one lb of thrust on a 44 oz plane in level flight, and then pull the nose up, now you have almost 3 lbs of weight (~44oz) extra to ask the prop to deliver. That's quite an increase of rpm if the airspeed stays the same. We can argue about the initial thrust, except that by definition, it is more or less equal to the airplane drag at that starting airspeed. So the initial thrust is cancelled by that drag (more or less depending how much airspeed you lose in the corner itself, but that isn't good either!). So the fight against gravity is rpm on top of the cruise rpm.

Using PropCalc, I needed to increase rpm from the cruise 9000 rpm to ~12500 to get that kind of change. That's a lot.

Of course as the plane rises in the arc, it can drain some of the rpm off. Conversely if you don't care if the speed drops some in the initial part of the climb, you can overcompensate later, but I wouldn't want full power at 90 degrees straight up, I think that is too late.

Or my real guess is that we don't really to keep airspeed constant, we only need to keep it high enough to keep the airplane manageable. Obviously there are a lot of solutions already out there that do that---at least in relatively calm air!

Just my bloviating!

Igor Burger · « **Reply #11 on:** October 22, 2009, 01:04:41 AM »

It MUST lose some speed. It is from nature of the thing.

The principle is just simple positive feedback. If it needs more power, then I give hime even more (I set higher rpm on governor). If this regulation should be stable, then the compensation (higher rpm vs slippage) must me lower than the slippage. It is the only way how to keep it stable. If I make it more sensitive, then it just jumps to highest power and stays there (or lowest) or oscillates. This setup as you can see on video is most sensitive I could do and in some parts of figures you can hear some "pumping". So it is on edge of stability.

It is light model with almost no mass inertia, so it almost stalls on top of figures without such a regulation.

Dennis Toth · « **Reply #12 on:** October 25, 2009, 03:29:08 PM »

Guys,

Interesting stuff, I think this controller is getting closer to an aggressive 4-2-4. I also suspect that as the nose goes up around the corner the power is added along a curve rather than just 100% on. The curve approach would add power on as you go through the corner and then be there to re-accelerate the ship going vertical. We seem to talk about constant air speed but in stunt what we need is constant ground speed through maneuvers. It is the ground speed that gives us line tension. It will be interesting to see what Alan records but it would also be useful if at the time of the readings there could be some wind velocity readings that would give us some idea of how much ground speed is lost as the ship penetrate up wind and how much we gain going down wind.

This idea of re-acceleration after a corner or going into a strong head wind raise the question on what type of prop/rpm set up gets us the best line tension and control low rpm with large diameter and high pitch or high rpm smaller diameter and low pitch while maintaining the same load on the motor. This is the same question that the IC guys have struggled with since the start of the modern pattern. Problem is the IC system reacts slow in strong winds and this is were we can make improvements.

Best, DennisT

Dean Pappas · « **Reply #13 on:** October 26, 2009, 07:41:51 AM »

Hello Dennis,
Yes, the question of whether we want constant airspeed, constant groundspeed, both or neither is a complicated one. Consider what overhead 8s will feel like in strong wind with either setup, and you will realize this ain't a clearcut issue!
Dean P.

Alan Hahn · « **Reply #14 on:** October 26, 2009, 08:17:57 AM »

I think the biggest question--to me at least--is what are the successful setups (glow and electric) actually doing. I think there is a lot of misconception.

---and not just by us, but by a most CL flyers. (my opinion!).

We need to know what is working before programming in some feature which may go haywire!

For example, I don't think I want level lap speed at the top of the hourglass. That speed may feel pretty darn fast for that short leg.

You need "just enough", but of course "just enough" is a variable quantity.

Igor Burger · « **Reply #15 on:** October 26, 2009, 08:29:25 AM »

Alan, I think it is pretty clear, you need speed stability to some extent. How much should be adjustable. It depends on mass inertia, drag, speed atc.

Actually (with contant rpm on governor) we can adjust it only by size of prop. But when I used old 4-2-4 motors, we needed to play with ventury diameter, prop, tank etc. With piped setup we needed to find proper pipe baffles etc.

Now since I have those my active timers, I see that even in calm air in gymn I need to limit its gain, because it is not good to limit the throttle too much or accelerate too much. It is even worse in wind and tubulence as dean wrote. So the magic of electric is, that you can have ANYTHING ... but you must set it properly for your needs :-)

Dennis Toth · « **Reply #16 on:** October 26, 2009, 10:29:47 AM »

Guys,

I agree that we need to have the "boost" adjustable. As Alan and Dean suggest some parts of the pattern are more comfortable to fly at a little slower speed (like the first corner of the outside squares). Maybe the approach is to have the controller rpm gain / loss adjustable so that one could add what they feel comfortable with. A good starting point might be the +-500 rpm target for a 4-2-4. It seems the idea would be to be able to adjust line tension as needed to maintain good control while flying at the slowest ground speed to achieve this.

Best, DennisT

Bob Reeves · « **Reply #17 on:** October 26, 2009, 12:00:05 PM »

Quote from: Alan Hahn on October 26, 2009, 08:17:57 AM

I think the biggest question--to me at least--is what are the successful setups (glow and electric) actually doing. I think there is a lot of misconception. ---and not just by us, but by a most CL flyers. (my opinion!).

We need to know what is working before programming in some feature which may go haywire!

For example, I don't think I want level lap speed at the top of the hourglass. That speed may feel pretty darn fast for that short leg.

You need "just enough", but of course "just enough" is a variable quantity.

If we get any flying weather Joe and I hope to help the data pool.. we have a pipe setup and 4 stroke ready and waiting.

Peter Germann · « **Reply #18 on:** October 27, 2009, 03:36:01 AM »

Quote from: Igor Burger on October 26, 2009, 08:29:25 AM

Alan, I think it is pretty clear, you need speed stability to some extent. How much should be adjustable. It depends on mass inertia, drag, speed atc.
Now since I have those my active timers, I see that even in calm air in gymn I need to limit its gain, because it is not good to limit the throttle too much or accelerate too much. It is even worse in wind and tubulence as dean wrote. So the magic of electric is, that you can have ANYTHING ... but you must set it properly for your needs :-)

Well, since you all seem to agree that we in fact do need "somehow" variable power, why not define what would be best by using an R/C (2.5Ghz) system to control power in flight and a data logger to record what will be found? I would not be surprised to learn from such an analysis (because of too many parameters and a lot of unknown interdependencies) that the design of a truly superior (v.s. IC) autonomous auto-throttle might be rather difficult.
Whether such finding then should launch a debate over the legalizing 2.4 Ghz throttles or not is of course another question...

Kind regards, Peter Germann

Dennis Adamisin · « **Reply #19 on:** October 27, 2009, 07:36:19 AM »

It is also possible to deal with the causes of the slowdown in the corner. As Alan pointed out, nose up or nose down - weight comes into play - less weight is better. Aerodynamically the drag caused by induced drag (drag due to lift) and control deflections are the source of most of the slow down. Efficient cornering can reduce the drag associated with the turn: higher AR, less "numb airfoils (avoid the currently fashoinable ice-cream cone shapes) and less control deflection, especially flap deflection can go a long ways to creating a bird that retains energy through manuvers...

Active throttle could be an enhancement to deal with the symptoms of lost energy through the manuvers.

Dean Pappas · « **Reply #20 on:** October 27, 2009, 08:54:44 AM »

Hello All,
Peter ... how delightfully subversive of you! Yes, something like this would be helpful, and will most likely highlight the difficulty we will face.

Dennis is right, of course. A super light airframe with minimized induced drag and a large prop moving a large mass of air through a small change in velocity (to make the same level flight thrust) will enjoy better characteristics than a heavy, large flapped ship with a small diameter prop. The smaller diameter requires more delta-V for the same thrust, so the thrust versus airspeed slope will be flatter. That is true now, and will be true even when we have all sorts of powerplant "smarts". It's just that that an optimised plane will not need quite so much smarts.

Regards,
Dean P.

Igor Burger · « **Reply #21 on:** October 27, 2009, 09:48:01 AM »

Quote from: Dean Pappas on October 27, 2009, 08:54:44 AM

a large mass of air through a small change in velocity (to make the same level flight thrust) will enjoy better characteristics than ... ship with a small diameter prop.

hmmm ... mnmmnmnmnmnm ... well ... I do not think so

:-) ... I am not going to argue now, but remember, that I did not agree too much ... I think we will speak about it after som years again ... :-)

Igor Burger · « **Reply #22 on:** October 27, 2009, 09:59:17 AM »

Quote from: Peter Germann on October 27, 2009, 03:36:01 AM

using an R/C (2.5Ghz) system to control power in flight and a data logger to record what will be found?

Yes Peter, but HOW you want to do it? By another hand? or another flyer? Just by feeling? I cannot imagine doing it and controll the pattern at the same time.

I think you remember that my solution in real model. I dah it on contest in Sazburg. It worked nice, but I needed to swith it off in turbullence in Landres, because I had feeling that I cannot concentrate when it tried to cancel all effects of the mixed air. I cannot imagine to do it myself.

Unfotunately I did not have time to continue with testing, but I think I will do it upcomming seasone. So may be in Gyula ...

Dean Pappas · « **Reply #23 on:** October 27, 2009, 02:39:37 PM »

Hi Igor,
Well, the first thing we would conclusively learn from Peter's experiment is that the power corrections would indeed need to be very fast in order to make us happy: very likely it will need to be too fast for manual control.

We agree that constant airspeed isn't necessarily the best, I think we are looking for different compromises.

In any case, I look forward to many discussions, maybe in person too!
Regards,
Dean P.

Keith Renecle · « **Reply #24 on:** October 27, 2009, 11:48:10 PM »

Igor said "Yes Peter, but HOW you want to do it? By another hand? or another flyer? Just by feeling? I cannot imagine doing it and controll the pattern at the same time."

I think that the guys like Windy that tried remote throttle control, found that a switchable 3 position throttle worked quite well, so for such an experiment as suggested by Peter, I would try a switch in my non-flying hand with preset high/low rpm for the two extremes. If you think of the 4 simultaneous controls used for 3-D heli flying, then this should not too difficult.

You can get PIC chips with built-in RF, so it could be done without having to modify existing R/C equipment. Igor??

Keith R

Peter Germann · « **Reply #25 on:** October 28, 2009, 06:12:40 AM »

Quote from: Igor Burger on October 27, 2009, 09:59:17 AM

Yes Peter, but HOW you want to do it? By another hand? or another flyer? Just by feeling? I cannot imagine doing it and controll the pattern at the same time.

I am referring to a feasibilty study and for the time being, Igor, there is not too much need to worry about how to do it in competition because they will not let you fly with it (in Gyula) anyway... As far as what can be done with two hands simultanously, you may want to go to YouTube and watch a violinist from close. Or someone playing the Organ, with two hands and two feet. Or, as Keith mentioned, observe the fingers of a 16 year old flying a 3-D chopper 50 cm abiove ground, inverted.

Only one way to find out more, I guess.

Kind regards,

Alan Hahn · « **Reply #26 on:** October 28, 2009, 01:02:37 PM »

Quote from: Peter Germann on October 28, 2009, 06:12:40 AM

I am referring to a feasibilty study and for the time being, Igor, there is not too much need to worry about how to do it in competition because they will not let you fly with it (in Gyula) anyway... As far as what can be done with two hands simultanously, you may want to go to YouTube and watch a violinist from close. Or someone playing the Organ, with two hands and two feet. Or, as Keith mentioned, observe the fingers of a 16 year old flying a 3-D chopper 50 cm abiove ground, inverted.

Only one way to find out more, I guess.

Kind regards,

16 year old flying -----that's the rub!

Igor Burger · « **Reply #27 on:** October 28, 2009, 02:39:50 PM »

Peter, I know what you mean, but I mean rather thinking what and when and not to repeat learned things with some feedback from observed motion. Ok so someone can try it :-). I have seen single button senders and receivers for remote control, and if necessary I can make a timer with such input - I can just diconnect current sensor from my existing 4-2-4 timer and thatt's it :-)

Erik Janssen · « **Reply #28 on:** October 30, 2009, 01:38:29 AM »

I have built a fully automatic sensor timer that is allowed by the rule book, the details can be found in the engineering section of this forum.

Section 4.2.2.g) For power sources other than piston engines, engine power controlling systems whether pilot operated or automatic, shall be permitted.

Peter Germann · « **Reply #29 on:** October 30, 2009, 09:01:51 AM »

Quote from: Erik Janssen on October 30, 2009, 01:38:29 AM

I have built a fully automatic sensor timer that is allowed by the rule book, the details can be found in the engineering section of this forum. Section 4.2.2.g) For power sources other than piston engines, engine power controlling systems whether pilot operated or automatic, shall be permitted.

Good to see you're back in business, Eric.

Yes, you project is legal and, to be honest, it was the intention of that rule to promote the kind of development you are working on right now.

How does your setup react in terms of braking in dive? What happens in the down segment of wingover? Does it reduce power quick enough to keep the airplane from diving too fast? I mean, if it would actually reduce the airspeed in that particular dive considerably, then tight, flat and level pullouts from wingover dives might actually become easier to fly...

Anyway, thank you for sharing the the result of your research with the community, I look forward to hear more about it.

regards, Peter Germann
p.s. I take your remark re the 16 years "stagnation" over at the engineering board as a big compliment for those who, in whichever form, actually influence, run and govern the event.