 |
| Author | Post |
|---|
D.Bernardin
Member
|
Posted: Sun Dec 27th, 2009 06:54 pm |
|
At first, i beg your perdon for the "heaviness" of these articles. My thougt is not fluent. I promise to rewrite when that thing is more mature.
More i look at Yves Rousseau's flight video, more i see that diving is not only produced by the high traction axis at the beginning of the downstroke. It last during all the downstroke. And all the upstroke is flied nose up.
On the video it is difficult to see if an excessive correction of attitude is responsible of this movement. On the pictures you can observe that the Pulcimatic 2 (Vector?) is trimmed with a positive incidence of the tail, during launching at first and during downstroke too. Do I see it on upstroke ? Or am I wrong ?
At first i thought it was anything linked with the center of gravity position, but no. In the case of an excesively forward CoM, the trim of the tail would be contrary : negative incidence. Probably there is a voluntary intervention of Yves . Look at these pictures and video :
http://urvam.free.fr/modules/tinyd1/index.php?id=2
This diving at downstroke apparently eliminates the heaving. Probably Yves Rousseau was on the way to obtain an horizontal flight, more efficient... or worst. What a tragedy that he was wounded !
(EDIT)...
... The FAI official 64 meters of Yves Rousseau are surely more. I suppose that the judge has substracted a lot to eliminate honestly the speed adquired during launching, and perhaps the, more or less, 50 meters, wich was able to cover Pulcimatic, starting from 5 meters height, as a glider. Indeed the video shows a long flight that is not at all finished when the movie stops. And we can contemplate the fact that this result was obtained with a few hundred watts. (Edited on January 25)
Since the heaving is fought by reducing the AoA of the center-wing, lift repartition does not satisfy at the need of an elliptic one. Erich Von Holst's proposed that you should have a turning of the wingroot at the same time than the twisting. Would it be the manner to reduce induced drag ?
Look at von Holst's ornithopters video in this link of "the other forum" : http://www.ornithopter.org/video/shtml
and at this work about E. v. Holst and Herzog models :
http://www.ornithopter.de/english/herzog.htm
also the assertion attributed at von Holst in :
http://www.ornithopter.de/english/principle.htm
Since von Holst was a physiologist, not an aerodyamicist, and since i know his opinion through other people, we will take a little distance with this assertion. On the other hand, i believe that Horst ornitopters made a discrete loop.
Nevertheless this change of AoA at root, would only be interesting for an ornithopter without heaving. The movement of a great wing flapping slowly gives a vertical acceleration to the body wich consequency is a reverse movement of a growing part of the wing. Then Von Holst's conjecture is no more applicable.
The wing-mill effect makes easier the downstroke, but not immediatly, what a pity because it is the moment of a stronger resistance of the wings. But the combination of an up-phase with the downstroke does not look aerodinamically favourable.
Wing articulation is not at all a luxury; the dihedral-anhedral movement allows a reduction of heaving; although i fear, insufficient in the case of an orthogonal movement.
Last edited on Thu Feb 25th, 2010 05:20 am by D.Bernardin
|
D.Bernardin
Member
|
Posted: Thu Jan 21st, 2010 06:36 pm |
|
Now, look at UTIAS ornithopter, the DeLaurier's and Harris's creation. It has a pair of articulations : center-wing, in a ramming movement, flaps the external wings. Then when the center is going down, AoA grows, while external wings are going up with, quite sure, a reduced AoA. All the contrary when the center wing is going up : it has a reduced AoA and the tips an increased one. Lift repartion is bad and we can fear a lot of induced drag. This part due to the bad lift distribution is added at that part due to the change of AoA necessary in this case of orthogonal movement, and at the part due to the interrupted wing . Murray's estimation if i dont misinterprete him, is that span efficiency was halved in this case (i would tell : at less !).
Pulcimatic 2 was perhaps in a better situation, able to accelerate at down-stroke. All the contrary in the case of UTIAS Ornithopter : opposition between center and outer-wings must give a nearly constant speed at the all the plane : at first it looks pretty ... but... if center wing is needing more speed while outer wings has to cut it, the opposition does not cure nothing.
The fact that a lot of heaving remained could have been cured changing the proportions between center and outer wings. But it did not change the general problem.
Nor in this case, the struggle against heaving is orientated in the good direction.
http://www.ornithopter.net/index_e.html
gives you a good documentation about DeLaurier's ornithopter.
Last edited on Fri Feb 5th, 2010 05:30 am by D.Bernardin
|
D.Bernardin
Member
|
Posted: Thu Jan 21st, 2010 07:42 pm |
|
Have a look at Jon Howes explanation of the loop wing's motion and its necessity for an economic flapped flight in ORNITHOPTER AERODYNAMICS, Ornithopter Wings Aerodynamics p.1 at the beginning (Jon's letter publicated by Patricia) and lower post illustrated with a graph of primordial importance : efficiency of flapping flight. If i have understood correctly, the wing accelerate first and the body catch later this acceleration (EDIT:or climb). Jon looks to speak about horizontal acceleration, but at the same time wings accelerate vertically, then body's reaction is a reverse vertical acceleration (EDIT:in other parts, vertical acceleration is evocated by Jon). Heaving remains but body and center wing go slower when the reaction drive them up... at down stroke ! Then on the all span, Cl can be more or less the same (rectangular distribution if there are feather's tips), or elliptically distributed. Also it can be the same at up-stroke and down-stroke, the difference of lift being the only result of the difference of speed between the two phases.
(EDIT):Note that the forward movement at downstroke must start from a point of the span that is never concerned by the "reverse" movement. So the orthogonal movement of the center wings of Jon Howes project is fully justificated. Would it be interesting to have a conical movement of the center-wings, for example in the case of an engine-driven flapper ? Possibly, for example the center-wings should move on a circular based cone, and further, cone should be elongated and inclinated by the sweep-dihedral articulation. Other movements are in the air, for example conical movement starting from the root with a refined deformation of the cone base : since we have a benefit of a vertical articulation (thrust and lift can be alittle more regular), why dont we use it at a sweep (horizontal) movement. Thrust gets the wing forward, drag backward. With or without skew-hinge.
Note that the forward movement at downstroke gets some limits : with a lot of inclination of the loop the path of the tips has the shape of a (rounded) saw-teeth. The angle of conversion of lift in thrust is reduced and the angle of conversion of lift in drag at upstroke is increased. Fortunately speed is more efficient than angle. It works by its square ! Don't forget that a "complete-total-true-ornithopter" has an advance ratio "at the limit". One of the reasons for what i prefer a partial ornitho !
Nevertheless it is important to understand that a low energy wake, only can be obtained by a low-loss plane, since to have a big thrust it should be necessary to change AoA, then Cl, between up and down phases : Back to the first step. Telling that i suppose that the release, of vorticity - on other words the startin'vortex, at each change of AoA - is a spring of induced drag. ¿ Or not ?
Last edited on Sat Feb 6th, 2010 02:38 pm by D.Bernardin
|
D.Bernardin
Member
|
Posted: Mon Jan 25th, 2010 08:14 pm |
|
Then, accelerate only the wing is easier at first step of the downstroke, and it consists in accelerate it forward and sweetly downward. For a rubber model it can be the beginning of a bad flapping cycle. For an human-powered-orni too. If a builder has made a machine without loop, he or she will have to use a lot of energy to win the resistance of wings-and-body to accelerate. Or shall have to seek some help : springs... or maneuvring for example.
Include with a good loop mechanism, the wings are not so easy to move. Their own mass is only a part of their inertia. A mass of air has to be accelerated at the same time. Classical flight mechanics take it in consideration in maneuvring calculus. For that reason, when wings accelerate forward, a no negligible reaction accelerate the body backward, and when wings accelerate downward, the body accelerate in the other direction. With a nicely articulated wing, at a few time, wings have lost their ability for a forward movement. At mid-stroke, more or less. Then, they accelerate all the aerodyne ... ¿forward? (note : the vertical acceleration is going on).
To answer, we have to look at a little phenomenon due to stability!
A plane, but i prefer to call it aerodyne, including animals in the case, when accelerates, climbs. Modellers, free-flight ones particulary, know the problem and not only inclinate traction axis of their propellers, but look for more efficient ways such High Traction Line.
I mind that it is the same for an ornithopter, and explain that currently it pulls it's nose up at downstroke. Watch other time at Yves Rousseau pictures, he has to give a strong angle at his stabilizer to obtain the light pitching down he looks for.
That tendency to pull nose up is not heaving; this one goes on, and adds his effects because verticals accelerations continue untill the downstroke is finished. And action-reaction too.
¡ Nearly the same things happen at upstroke ! With the difference that the force of acceleration actually "brakes", but the reaction accelerate the body (and center-wing : good, since there is heaving!).
Why birds do not suffer of heaving, or just a few ? It is a scale effect : flapping is quicker, time is shorter, acceleration acts, but a few time, thus with less results. Bodies inertia, concentration of mass, plays its role. Very important in the "mechanics" of upstroke ! Manned ornithopters have much more time to fall down by a lack of lift during this bloody phase ! But is it so important ? Sure it is, but if the wing adapts itself at this movement the loss is not so big, it is a kind of gliding, not very good, but it works.
About this theme look at "Jon Howes project" p.1 "derivation of glide propulsive efficiency", a sketch and a curve... and an interesting commentary.
Last edited on Sat Feb 6th, 2010 06:47 pm by D.Bernardin
|
D.Bernardin
Member
|
Posted: Sat Jan 30th, 2010 01:04 pm |
|
Following mindings concern the heaving in the case of an articulated wing. I make a distinction between heaving that produce a diferential movement of body and tips along the path, and pulling nose up, or pitching down, where all the aerodyne follows an upward or downward way. Perhaps i am wrong, then tell me it.
Since the sweep movement has a limit, reached at mid-stroke (more or less), and flappin'one has not finished yet, thus we enter in the bad phase of the flight. Center wing and body, are rowing an up-phase while outer wings follow their own down phase. Then, it is not welcome that center wing pursue accelerating. An accelerated part of wing making an upward movement! : either we have to reduce AoA and we are in the same case than with an orthogonal movement; or we produce a lot of "mechanical drag". Wich choose ?
I believe we have to conceive a mechanism that pursues a tip forward movement, producing a reaction of body and center wing, accelerated,yes, but not as much as the tips. It must last untill flapping approach the BDC, and then, there is no more reason to worry about that, (but we have to preocupate about upstroke)!
It was the way explorated by von Holst with his "Swan". Rods gave a forward-backward movement at a free center wing, at the same time they moved it up and down. It was one of the early successfull ornithopter, flying a long time with a little weight of rubber : 9 g. for a 185 g. machine, if i am not wrong.
Successfull too Jon's indoor, with very different mechanisms : flights of 4-5 mn, what a result. But ¡hem!, i am speaking about very light machines, not about manned ornis. Do they were concerned. Sure they could suffer of bad sides of heaving, since they had a poor body inercia and a slow flapping. The "proof" is that both the authors have minded about the problem and used "solutions"! E. v. Holst with his sweeping movement controlled from the mechanism, and with an orthogonal articulation at mid-span to reduce heaving.
http://www.ornithopter.de/english/herzog.htm
Jon Howes used a controlled out of phase dihedral motion. His early concept of a skew hinge gave only a dihedral-anhedral movement joined to a control of incidence : nothing of a sweep one. Perhaps the wing was sufficiently flexible to sweep by it's own ? Yet the research of smoothness was on the rails. The man powered project, now has a sweep movement joined to the dihedral and the incidence one.
Look at Jon's page : http://www.ornithopterresearchgroup.com/view_topic.php?id=115&forum_id=8 Last edited on Sat Feb 6th, 2010 06:37 pm by D.Bernardin
|
D.Bernardin
Member
|
Posted: Sat Jan 30th, 2010 05:02 pm |
|
Now, why would'nt we look for heaving suppression ? Is it so difficult ? Do we afford to do it ?
We must avoid the counter action of a part of the wing. Although a study of a cylindrical-conical version of Harris-DeLaurier orni could give hopes. Or a conically modified Buckley's machine !
¿ Could it be ?
Nevertheless, I advocate for a multi-articulated wing. Don't the pterosaurs have been the larger flying animals ? But my ability to study a so involved solution is limited. Then i prefer to trust in a partial ornithopter... that could be a good testing bench for a pair of articulated wings : at first one articulation, then two, until we have a smooth motion and a nearly constant lift, so we can reduce the fixed part and enlarge the flapping system !
But i am not intenting to sell you my merchandises. I intent to think about flapping wings problems to get a clear( !) idea of how to reach a nice efficiency. Then let me tell you about a doubt :
Up to now, i have considered that signature of a flapping wing in the wake is determined by induced angle, so with a more or less constant span and constant A/R, determined by Cl and then by AoA. That is what mean for me the research of a constant downwash. The loop accelerating the tips at downstroke and accelerating them negatively at upstroke is the way. What happens if we discover that induced angle and Cd-i constant are not sufficient, and we have to maintain induced drag constant ? We have obtained a nice lift distribution when the machine is gliding; if the speed, then induced drag, has anything to see with the downwash we would have to reduce (not to much, thrust would suffer) the AoA or the area of the tips to improve lift distribution at downstroke. As long as heaving remains we cannot increase the angle of attack of the center-wings, as v. Holst suggested. And at upstroke ¿ reduce the span ? It is a big problem of tuning ! I ask everybody if you have an answer about it ? My feeling is that it is difficult that induced drag be the same at up and downstroke, in compensation a constant AoA would allow not to have vortex release, thus approach fixed wing conditions, but perhaps i am wrong !
From other point of view it is true that considering the momentum at root, the best shape of lift repartition can be different of an elliptic one (Prandtl-Jones); i will have a look at Jones'papers, NACA tech. note 2249 (1950) and NASA tech. mem. 81174 (1980). Is it the same for a nearly rectangular repartition ? More span, less lift at the tips, the wing is not harder to move but has got more A/R. Looks good, at the condition of a moderate use !
Anyway, i dont want to design a perfect ornithopter, just one that works ! Don't forget there are phenomenons like Magnus lift and extra lift due to unsteady aerodynamics that can destroy our beautyfull constructions (of a constant washdown); does it produce a little vortex release at each endstroke ? And it remains difficult to reach the perfect AoA of a flexible wing, even with an active control !
"Nobody is perfect" told Charlie Brown. "Eso si que es cierto" contestaba Mafalda.
Last edited on Sat Feb 6th, 2010 06:56 pm by D.Bernardin
|
D.Bernardin
Member
|
Posted: Thu Feb 4th, 2010 06:24 pm |
|
Let's get back at our subject ! Aerodynamics is in another thread, but indeed all that is so intricated !
What happens at upstroke, yes it is quite a glide, see Jon Howes'page, posted: Tue Mar 18th, 2008 11:03 am
http://www.ornithopterresearchgroup.com/view_topic.php?id=115&forum_id=8
Eta=((L/D)^2)/((L/D)^2+1) and results that efficiency of gliding is nearly 100% with only L/D 10 or better 15.
But there is a lack of lift. The body, in the general case, should dive, then center-wing producing lift... and thrust. We encounter the problems of downstroke, but seen like in a mirror. The articulated wing, skew-hinge one e.g., looks to work correctly until it reaches (approximatly) mid-stroke : reaction to the braking tips accelerate opportunely the body and the center-wing. The same as at downstroke : we should have a continuous sweep, so reaction would follow accelerating body until the stroke is nearly finished, so our bird is quite gliding a few time.
Other time i advocate for a research of heaving suppression ! One articulation is barely sufficient for a 3 m. bird. For a manned orni we need more.
But i suppose that some heaving remains (for a century at less), then... There is a nice question :
What about the loss of altitude of the body, next height gain, are we wasting energy ? At first look, no, diving generates speed and you should be able to turn it in height later. A square fuselage with sharp angles is not recommended ! But always there are other reasons of loss than fuselage drag, and sure not all the potential energy is converted in cinetic one, nor the contrary.
Nonetheless, it is not the worst side of heaving.
Last edited on Fri Feb 5th, 2010 10:49 am by D.Bernardin
|
D.Bernardin
Member
|
Posted: Sat Feb 6th, 2010 03:13 pm |
|
I have described a first cycle. What happens at the next one ?
At the end of the "second" up-stroke, kinetic energy of the body falling by lack of lift, or creating what is lacking (As you want it Charlie Brown), will remain and be used at beginning of next down-stroke. Heaving will come later but not stronger. Since this energy helps to maintain the body steady and flapping is more efficient, energy is recuperated.
Yet starts heaving, then it is important to follow it with a fore movement of the wings and a reaction of the remaining.
When they reach BDC, an amount of kinetic energy is used to fight tendency at a fall due to up-stroke.
We get it good ! We heave, but not so much ! Let me believe that it is not sufficient for a "bird" of 8-16 m. span. And less for a 28 m. one, the case of UTIAS students machine ! But let them fly, the movement of their wings still is an half mystery for me !
http://www.flikr.com/photos/hpoproject/3820643290 , their photo gallery and
http://www.hpoproject.ca , their page.
http://rowingbike.com/site/EN/Technical/Rowglider/ , page of Derk Thijs of Thijs productions, "Gold Sponsor" of the project, gives an idea of the mechanism, but it can be Thijs's idea. Anyway its a second hand information. First published by Kjell in the "other place". Best information is in the hand-sketch at right side of the document.
Last edited on Sun Feb 7th, 2010 07:13 pm by D.Bernardin
|
Current time is 07:42 pm | |
|
|
|
|
