oXs glider performance measurement

Development & General Chat for the superb openxvario project.

Moderator: rainer

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Sun Aug 07, 2016 5:05 pm

jhsa wrote:... will give it a shot and report in the right thread wherever that is ;) :)

João
IMU-Vario off-topic postings have a long tradition in this forum. Everyone posted offtopic until today ;)

User avatar
jhsa
Posts: 16676
Joined: Tue Dec 27, 2011 5:13 pm
Country: Germany

Re: oXs glider performance measurement

Post by jhsa » Sun Aug 07, 2016 5:07 pm

Maybe because there isn't really a thread for it? ;)

João
My er9x/Ersky9x/eepskye Video Tutorials
https://www.youtube.com/playlist?list=PL5uJhoD7sAKidZmkhMpYpp_qcuIqJXhb9

Donate to Er9x/Ersky9x:
https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=YHX43JR3J7XGW

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Wed Aug 31, 2016 4:18 pm

I have just made a test with a vertical dive, comparing the two pressure sensors for VSpd and ASpd. The result is remarkable - i never thought, they are so close. Mstrens did a marvelous job! VSpd is converted in km/h in the csv.
VSpdASpd.png

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Sat Oct 01, 2016 12:47 am

I finally got some data out over the sea with my 3.8M X2 e-glider. It weighs about 1700g. The tide was out when I started, but my beach was shrinking by the time I finished. It was mid-day, with very light on-shore winds, 13C, pressure 101.6kPa, water surface temperature about 14C. The air was very smooth, but there was the occasional bump even out over the water. I also discovered damp, salty sand sticks to everything!

I'm not nearly as good as Bernd at damping out the phugiod oscillation. I had a silent centre on my vario, which wasn't a good idea. It is very hard see the oscillations, particularly when the glider is at altitude. Next time I'll have a continuous tone vario, which might help me smooth out the pitch oscillation.

On flight 2 I forgot to turn on the logging until near the end, and it got appended to Flight 1. I also forgot to make sure the airspeed was zeroed that flight, so I don't think that data is of any use.

I had the three oXs performance measurement variables assigned to AccX, AccY, and AccZ. I believe AccX is TEST3: elapsed time since Tzero (in 1/10 of sec), AccY is Glide Ratio (1/10th), and AccZ is Average Vspd since Tzero (cm/sec? Doesn't seem right.)

SE is my flight mode switch. 0 is Cruise, flaps neutral, -1 is Speed mode, and +1 is Thermal mode. I also have variable amounts of camber in thermal mode on the right slider, which I did vary on some of the later flights. I haven't really spent any time on the file, but I think there might be some useful info in it. I'll have to see if averaging the airspeed and Vspd over longer, reasonably stable times, gives decent numbers.

Kevin
Attachments
e-Xplorer-2016-09-30.csv.zip
(307.82 KiB) Downloaded 24 times
Tide coming in.jpg

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Sat Oct 01, 2016 2:49 pm

Kevin,

nice flight, nice location, nice plane, nice data, good pilote - all in all impossible for me to compete ;)
Circus.png
I have analyzed 3 intervals, where you can see a very good performance, especially the glideratio with nearly 50 km/h is amazing:
Kevin1.png
Kevin2.png
Kevin3.png
I think your system works very well and the best way to analyze the performance is, to calculate constant flight intervals in the csv-file in a spreadsheet. I calculate the average speed, the flown distance and together with the depleted altitude you have all you need.

Mstrens made a oXs testversion, where the measurement can be started with a PPM-command. I am still testing with it, but it looks very promising. When you notice, that your plane is in constant flight, you start the measurement and oXs does the calculation and transmits the performance data over telemetry. There is still a little deviation, it needs more tests. Did you install a PPM input in your oXs?


kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 2:52 am

Thanks for looking at the file Bernd, I'd almost given up on it after some more time on it today. I am very disappointed by the oscillations. It is impossible to see them from the ground, but the glider does it continuously. I ma very impressed you are able to damp them out listening to the vario. I guess we must do it almost subconsciously when we are sitting in the airplane, sort of like keeping a bicycle upright or standing up.

I a have turned off the centre dead-band on my vario for test flying, so hopefully I can get closer to your smooth flying by using the vario tone. I'm also going to try putting my Eagle Tree Guardian in my eSupra, to see how it does. I can't see how to use it on my Xplorer. The fuse is very small, and I have an SBus set-up in it.

I think I might try a mix on my Taranis that feeds the airspeed back into the pitch trim. It will be on a switch! I'll try to set it up so I can vary the target airspeed with my S1 knob. This has all sorts of potential for bad things to happen, but I'll see if I can make it work - maybe on a foamy glider over land first!

And I've changed my airspeed to m/sec, instead of stupid knots. I meant to do that before, but there is always something I forgot.

Yes, even over 14C ocean the air is not perfectly smooth at noon on an unstable day as the tide comes back in. It was pretty smooth though. Yes, tther was very light lift over the beach and along the tree line. It is likely better in spring, when the water is colder than the air.

I did not put a PPM input on my oXs, but I suppose it can be added? Another thing to figure out! This is much harder than I hoped.

Kevin

Edit: The config file says the Average Vspd since T0 is supposed to be in cm/sec, but it doesn't seem to be. I believe I have it assigned to AccZ, but I have values like -0.6, which is only 0.006m/sec (1.2 ft/min)? Maybe it is m/sec? 0.6m/sec would be 120 ft/min. That sounds closer?

mstrens
Posts: 1066
Joined: Fri Dec 27, 2013 7:49 pm
Country: -

Re: oXs glider performance measurement

Post by mstrens » Mon Oct 03, 2016 6:41 am

If I remember correctly, when you use the Frsky fields Accx, AccY or AccZ, the values send by openTx are divided by 100 when displayed on the Tx.
Best is to use fields Temp1 and Temp2 because for those fields openTx displays the values it get.
If you have to use a field Acc, then you can change oXs code in order to multiply the internally value by 100 just before sending.

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Mon Oct 03, 2016 8:56 am

Kevin,

i understand your mind regarding the phugoids. I had exactly the same concerns. But it is only an exchange between kinetic and potential energy. It will happen always and why not live with it. We want to measure the real performance of our planes, and if they tend to oscillate, let us measure it oscillating. Any manipulation of the pilot or a gyro during measurement manipulates the measurement. I am very relaxed meanwhile. It is yet to proove, if the phugoids have a negative influence to the performance.

The consequence of "living with the phugoids" is to measure in longer intervals, as i did manually above in your log and as mstrens testversion does automatically, when you start and stop the measuring with a ppm signal.

Your setup is already able to measure the real performance, the rest is some fine work.

If it were easy, everybody could do it ;)

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Mon Oct 03, 2016 2:51 pm

Additionally a log with the performance of Kevins flight calculated the original way inside oXs:

Red=Average sink m/s --> longtimecalulation over 140 sec. was -0.40 m/s
Blue= glideratio (left scale x 10) --> longtimecalulation over 140 sec. was 24.3
Kevin4.png
If we want to stay with this method, we have to stabilize the flight. I am openminded, but i prefer the easier way without additional variables and devices. What do you think, Michel and Kevin?

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 3:13 pm

Thanks again Bernd! I was trying different speeds and flap settings during those flights, so I'll have to see what configuration that was. Too ambitious for the first go! I really want to get enough points to make a polar curve.

I would still like to get a polar plot of the glider performance without the phugoids. It would be interesting to compare with and without, just to see what the performance impact is. There are a number of Rx with gyros built in, and talk of allowing them in competitions, so it would be nice to know if there is an advantage. Plus the aero tools model the performance without phugoids, and I would like to get a polar to compare. The oscillations have to hurt the performance, but I don't know how much?

There is also the oscillation in yaw, which according to Marc Pujol's analysis causes a pretty sizeable performance hit.

I can't seem to get the Test values to show up as telemetry values on my Taranis, so I ended up using the Acc outputs. The other values (Glide Ratio (1/10), and Seconds since T0) don't seem to be divided by 100 in the log file. In any case, my AccZ (Average Vertical Speed since T0), seems to be in m/sec, which is perfect anyway. Now if I can get the speed to stop oscillating...

It seems fairly easy to set-up the Taranis with two logical switches, one when the airspeed is greater than your desired value, and one when the airspeed is less then the value, and feed those into an elevator mix that drives the elevator the required direction. There are adjustable delays and speed available in the mixer, and it is easy to activate the mix with a switch. What could go wrong! Well, increasing oscillations, pitch divergence, etc., but other than that?!

I'm not sure if the heading information from the gyro would be of any use in setting up a mixer on the rudder? Have to remember to turn the mixers off before trying to turn or land or anything...

I've used all the global variables already. Never enough of those things.

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 3:51 pm

Carbo wrote:Additionally a log with the performance of Kevins flight calculated the original way inside oXs:

Red=Average sink m/s --> longtimecalulation over 140 sec. was -0.40 m/s
Blue= glideratio (left scale x 10) --> longtimecalulation over 140 sec. was 24.3

Kevin4.png

If we want to stay with this method, we have to stabilize the flight. I am openminded, but i prefer the easier way without additional variables and devices. What do you think, Michel and Kevin?
I just looked at that section, and I made two 180 degree turns over that period! I would think that has to hurt the performance?

That was all in cruise mode, flaps in profile. The speed varied continuously, with a peak of 22 knots (40.7kph) and a minimum of 16.8 knots (31.1kph). Just visually, the average speed looks like about 19 knots (35kph).

When I look at the period without turns from 11:03:55.180 to 11:05:00.020, the heading varies from 260 to 320 degrees, and the speed varies from 17.4 knots to 21.2, with a calculated average of 19.2 knots (9.88m/sec, 35.6kph). The average glide ratio is 24:1, and the average sink rate is 0.412 m/sec (81.1 ft/min).

Surprisingly close to the numbers including two turns!

Kevin

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Mon Oct 03, 2016 4:34 pm

kcaldwel wrote:I just looked at that section, and I made two 180 degree turns over that period! I would think that has to hurt the performance?
Surprisingly close to the numbers including two turns!
Obviously you are a good pilote. I learned better turns during the measurement flights, my turns where dolphin style before.

That is by the way another argument for the version where the pilot defines the measurement interval. If you have no GPS onboard, you can not see the turns afterwards. And even the best pilot can not avoid some loss in turns.

If you want to try it, you need a free channel on your RX and a single cable to bring the PWM signal to PIN 3 of your arduino. It can also be used to switch and control the airspeed compensated vario and to control the vario sensibility.

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 5:10 pm

I have taken all the aileron differential out of my sailplanes, actually measuring the degrees of throw up and down on the ailerons and flaps so the angles are the same in all flight modes and flap settings. I find that makes the turns much more efficient and easier to fly. If you have any differential in the actual surface throw, you are changing the average wing camber and therefore the pitch trim every time you deflect the ailerons. That is hard to correct for, and I found the glider would pitch down every time I started to roll into a turn. It is much better with no differential, and the rudder is far less drag to correct for adverse yaw. I don't use aileron to rudder coupling either, but then I don't seem to fly straight much either!

I'd have to put a Y-lead in my SBUS, and add another SBus to PWM converter in my Xplorer to get another PWM channel available. It is simpler in the Supra. I wonder if it is easier just to parse the csv file afterwards? I'm not that good at telling how steady the glider is in flight obviously! I know they "notch the barograph" when they do full scale tests. A sudden pitch down and up at the start of each straight run gives a reference in the data.

I think I need electronic help to steady the glider. It may be unrealistic, but it would show the maximum the glider is capable of. Does anyone see a fatal flaw in trying the logical switches using the airspeed and pitch mixes?

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 6:47 pm

I've got a few more points that look like they might be reasonably good.

One is at high speed with camber in reflex, from 11:25:57.130 to 11:26:15.050:

average speed = 33.16knots (17.06m/sec, 61.4kph)
L/D = 9.22
Sink rate = 1.85m/sec (364 ft/min)

One is at low speed with camber at my minimum thermal flap setting, from 11:39:49.200 to 11:42:04.980:

average speed = 13.80 knots (7.10m/sec, 25.56kph)
L/D = 19.29
Sink rate = 0.368m/sec (72.4 ft/min)

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Mon Oct 03, 2016 8:11 pm

It's only 3 points, and all at different camber settings which should all have their separate polar curves, but it is a start!

It looks like best L/D and min sink speed may be quite close together, near 8.3m/sec? More data needed.

If I could get the oscillations to stop, I might be able to get enough points for a reasonable polar in a couple of flights for each flap setting. For now the data is so noisy, it would take long periods at each speed to get a reasonable data point.

Oh yeah, my oXs airspeed is set for "indicated", but this data at least was so close to standard conditions and sea level, the difference would be minimal anyway.

Kevin
Attachments
eX2 Polar V1.png

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Thu Oct 06, 2016 1:03 am

Well, I tired my simple logic switch airspeed autopilot today. And the winner is - pitch oscillations! I tried a few different ramp speeds, and had it set-up so I could vary the amount of trim the logic switches added from the ground, but the results were always the same: rapidly increasing pitch oscillations. The logic works right - adds down elevator if the speed is below the target deadband, or adds up elevator if the speed is too high, but it doesn't help in practice. I think the elevator would have to work on the rate of airspeed change, not just a simple on/off.

But I did get a chance to compare the airspeed control of the Eagle Tree Guardian and trying to hand fly with the vario, in reasonably bumpy air. I've attached some plots, the first one is an entire flight with some autopilot engaged segments, some hand flying segments, and the logic switch airspeed set-up in the middle.

The Eagle Tree Guardian does a pretty good job of controlling the airspeed, on a day with some wind and turbulence at our inland flying field. I'm hoping it will be even better over the sea. I can adjust the flight speed target with the elevator trim, so I should be able to get multiple points at pretty stable airspeeds per flight. That should make it pretty quick to build up some decent polars at different camber settings.

I think I've even figured out how to hook up the Guardian in my SBus X2, and cram it in the small fuse. I'll only have it hooked up the rudder and elevator in the X2, but hopefully I can keep the wings level by hand.

Kevin
Attachments
Hand vs ET airspeed.png
Hand_vario airspeed.png
ET Guard Airspeed.png

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Thu Oct 06, 2016 1:19 pm

Kevin,

the phugoids are a serious opponent. I made exactly the same experience. My next approach, was to control not only the amount of correction, but also the period of time, to avoid overcorrection. While i was testing this, i lost my plane nearly twice and decided to change my mind. In other words: the phugoids won the game.

But i will follow your way very attentive, eventually we will know sometime, how much performanceloss the phugoids generate.

Bernd

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Thu Oct 06, 2016 3:47 pm

It is amazing how much the gliders are oscillating, because I cannot really see it from the ground. It looks like it is flying along in a well-trimmed glide, but the speed and sink rate are consistently oscillating. I am not very good at trying to damp it out by hand, even with the vario.

Yes, the logic switch speed non-controller got quite exciting quite quickly! I lost sight of the glider yesterday after trying to watch the airspeed on the telemetry screen a bit too long. And I accidentally shut my Tx off with my X2 not that high over the forest while playing with switches and trim buttons the other day. My fail-safe deploys full flaps, and puts it in a turn, which is good in some circumstances, not so good in others. It takes an agonizingly long time for the Tx to come back on, and then of course there is a switch warning that has to be cleared before it will transmit. This is all done with some danger to the gliders.

I might try the Guardian in my eSupra fiirst, because it requires less dismantling and modification. We're into our fall storms here, so the flying days are much further apart now.

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Thu Oct 06, 2016 4:43 pm

It's a very long time since I studied aircraft dynamics, so I've been doing a bit of reading on pitch phugoids. The long period pitch phugoid on a full-scale sailplane is a bit longer, approximately 17 seconds. An RC sailplane appears to have a 10 or 11 second period typically. The pilot in a full-scale glider easily and subconsciously damps out the long period phugoid in normal circumstances, but it can be an issue in cloud flying as legally practised in Europe. On an RC sailplane the pilot doesn't seem able to easily damp it out, except for Bernd using the vario tone and concentration.

Pitch phugoids have caused a number of crashes in full-scale airplanes, mostly in unusual circumstances, listed on the wiki page:

https://en.wikipedia.org/wiki/Phugoid

It is interesting that the Airbus software has anti-phugoid code.

The pitch phugoid is a constant angle of attack manoeuvre, where speed and altitude exchange potential and kinetic energy. The pitch input to stop it might work better exactly opposite of what I had set-up with the logic switches: Down elevator at the peak airspeed, up elevator at the minimum speed. Do I feel brave enough to try that?!

CG location will change the phugoid damping characteristics, but I'm not sure you can "cure" it with a change in CG position, except possibly by having the CG slightly behind the aircraft neutral point. Of course the aft CG introduces pitch instability which has it's own issues. I have successfully flown an RC sailplane at very unstable CG positions with a simple rate gyro on the pitch control, so that may be an option for testing.

The creator of XFLR5 has done some pitch phugoid modelling and testing with RC sailplanes, with similar results to what Bernd and I are seeing:

http://www.xflr5.com/docs/XFLR5_Mode_Measurements.pdf

Kevin

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Thu Oct 06, 2016 4:51 pm

I am often thinking about this RCG discussion. Maybe i understand it wrong. What does he mean with this sentence:
"Stability is part of trimming process which you ran into phugoid pitch stability"
Is there a chance to have a "natural" pitch stability, had or have the A2 gliders this kind of pitch stability?

I just saw your answer, before i was able to hit the Submit button to ask. Voodoo feeling :lol:

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Thu Oct 06, 2016 5:38 pm

I never understand most of the posts by "old"! The phugoid on typical RC gliders doesn't seem to increase under normal conditions, so it is damped to some extent. The disturbances come from the pilot or the air often enough to re-excite the oscillation before it can decay. I suspect FF gliders are the same. They will oscillate in speed. If the air is very smooth (early morning), the disturbances and oscillations are smaller, so you can easily trim for closer to stall speed. In more turbulent air, the excitations will be big enough that the glider will stall on the slow part of the oscillation if it is trimmed too close to stall speed to start with, and then the oscillation quickly gets out of hand. So you trim faster for more turbulent conditions, nothing new in that. I think he was playing in a shallow pool...

I might actually try reversing my logic switches. Down elevator at the peak of the airspeed band I want to stay in, up elevator at the low speed. It can't be any worse than the way I tried yesterday! I vote for pitch divergence this time though. Without detecting the rate of change of the airspeed, I don't think you can stop the oscillation, and I can't see how to do that without a LUA, Even then, getting it to work sounds like it could take quite awhile. I've talked myself back into going ahead with the Guardian, since I know it works.

Kevin

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Wed Oct 19, 2016 11:16 pm

I got a couple of flights with the Eagle Tree Guardian in my home-built e-Supra out over the sea yesterday. It was quite overcast, 11.5 degrees, very light wind, and some rain showers. I have only glanced at the file so far, but the speed stability looks pretty good. This is all in cruise camber. I tried to do about 20 seconds at each speed. There was some rain drops on the wing, which probably didn't help the performance.

I was going to do some thermal and speed camber settings, but I managed to dunk the glider in a 150mm deep salt water pond. The tide wasn't very low, so I had quite limited sand to land on. I had a bit of an argument with the autopilot during the last turn back, and I ended up trying to stretch a glide across the pool, which didn't work out. One day I will remember these are motor gliders, and use the motor to get out of the trouble I sometimes fly myself into!

A quick drive home, a wash in fresh water, and then some drying out, and everything still seems to work, amazingly enough. My boots are still pretty wet though.

I'll try to get some data extracted from this file soon.

Kevin
Attachments
e-Supra_Test-2016-10-18.csv.zip
(123.37 KiB) Downloaded 23 times

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Thu Oct 20, 2016 8:01 am

kcaldwel wrote:One day I will remember these are motor gliders, and use the motor to get out of the trouble I sometimes fly myself into!
Sorry for your mishap, but LOL, you are not alone, remember my landing in the wineyard behind a mobile home. Your declaration above has now prominent places in my poetry album and my ToDo list :D

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Fri Oct 21, 2016 4:58 am

I'm honoured to be in a poetry album!

I pulled some points out of the file in areas where the airspeed was the most stable. Now I'm trying to decide which is the most realistic way to plot the data. If I use all the points I have, I have a weird bumpy polar (V1). I think this may actually be the first sign of the low Re effects at work. From a paper on the SuperGeeII DLG polars using the wind tunnel data from the UIUC on Dr. Drela's airfoils, the polar ends up lumpy as well as opposed to the nice smooth curves from linear theory. I guess I could spend the time doing a polar on the Supra wing using the wind tunnel data too, but it would take some time.

If I remove all the "outlier" points, and do a x^2 curve fit, I can get a reasonably smooth curve (V2).

The lumpy polar shows a best L/D and min sink (-0.300 m/s) at the same speed of 7.54m/sec (27.1 km/h). There is a second point at 8.81m/sec (31.7 km/h) that has an L/D of 23.4 and a higher sink rate of -0.376m/sec. All these are at cruise camber, so I expect the min sink rate would improve with more camber.

Kevin
Attachments
SuperGeeII Polar.jpg
Lumpy SuperGeeII polar
eSupra Polar V2.png
eSupra polar with outliers removed to make a more classic smooth polar.
eSupra Polar V1.png
sSupra lumpy polar.

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Fri Oct 21, 2016 7:33 pm

Here are two scans from a brandnew article in a german modeller magazine about telemetry in scale gliders. This glider is equipped with a compensated vario "Braunschweiger Duese" and an airspeed sensor with pitot tube. Phugoids are not mentioned in the text.
Aufwind1.jpg
Aufwind2.jpg
The measuring flight shows a large phugoid, even the compensated vario shows it, the compensation is not very good. A compensated vario is also not helpful to stabilize the flight. Therefore the pilot was listening to the speed announcements.

But how does one manage to get such a perfect polar from such a measurement?

FabFlight
Posts: 2
Joined: Wed Mar 19, 2014 5:23 pm
Country: -

Re: oXs glider performance measurement

Post by FabFlight » Fri Oct 21, 2016 8:18 pm

John Ellias' polars weren't perfect either, but better than what Kevin's shown. Notheless you're on the right track. I think if you put out another session you'll be able to remove the right outliers and get more good data.
http://www.xcsoaring.com/techPicts/%20E ... 20test.pdf

kcaldwel
Posts: 40
Joined: Wed Mar 12, 2014 7:49 pm
Country: Canada

Re: oXs glider performance measurement

Post by kcaldwel » Sat Oct 22, 2016 12:19 am

Well, I would debate that John's data is better. He used a GPS for speed information, which gives ground speed, not airspeed, and has a very slow update speed. The wind speed variation with height would not be taken into account at all. I doubt he would even see the pitch phugoid from the GPS speed data. He was manually trying to fly airspeeds, so I suspect there were rather large variations. which were all massaged out in the data reduction. Making the data fir a nice parabolic curve isn't necessarily very realistic, and will give incorrect results. Even at full-scale Re, there are kinks in the polars from Re effects.

He was also flying at much higher Re (9.8lb. glider), where the low Re effects will be less than at Supra airspeeds and Re.

Kevin

Carbo
Posts: 249
Joined: Fri Aug 02, 2013 6:55 pm
Country: -

Re: oXs glider performance measurement

Post by Carbo » Sat Oct 22, 2016 1:05 pm

GPS speed will unfortunately not give usable results. Every flight with GSpd and ASpd logged, shows this clearly, there are a lot of examples.
@Kevin: I have calculated nine smooth sections of your log in a spreadsheet, can you please paste the results in your polar. I am interested, if there is a deviation to your method.
Kev_eSupra.png
The spreadsheet contains your log and the nine calculated sections:
e-Supra_Test-2016-10-18mod.zip
(159.75 KiB) Downloaded 18 times

RightRudder
Posts: 241
Joined: Tue Jan 15, 2013 9:41 pm
Country: -

Re: oXs glider performance measurement

Post by RightRudder » Fri Dec 16, 2016 8:12 pm

You guys are one step away from being able to build a netto vario. If you can reliably measure the glider polar then oXs can compensate the vario not only for total energy but also for the natural sink rate at the given airspeed. Couple this with an accelerometer and even a super netto vario is possible!

Post Reply

Return to “OpenXVario - an open source vario supported by the open source firmwares!!”

Who is online

Users browsing this forum: No registered users and 1 guest