Sikorsky Ilya Muromets CG
06-24-2015, 11:49 AM
#51
Reply to Burkhard Post #50, Off Topic.
Yes, a gyro works very well in yaw for a taildragger. My first experiment with a yaw rate gyro (mechanical gyro back in the 1980's) was to see if the gyro might help with spiral stability. Turned out that the yaw rate in the spiral mode was so slow that the gyro didn't sense the movement. But it sure did help with directional control on takeoff ! This gyro was a yaw damper only, with no heading hold. But it made takeoffs amazingly easy. I left the gyro in place for it's takeoff improvement and found that I had soon grown so lax that I could hardly manage if the gyro were turned off.
I have since adopted yaw rate gyros for most of my taildragger models. For the people who say that the gyro is a crutch, I say rubbish. The inertia of a model airplane is so much lower than a full scale that the divergence rate in a ground loop is far greater than on full scale. The rapid divergence of a model, plus the inability of the pilot to see or feel the initial stages of a ground loop, makes the model most unrealistic compared to full scale.
When we consider the feedback loop between airplane and pilot, the model is quite different from full scale. In full scale, we are flying primarily by attitude reference and accelerations. In a model, we are flying primarily by flight path direction. Two very different feedback schemes. Flight path is the integrated result of changes in attitude and acceleration, so the model pilot is always in a delay loop. Ever see a model pilot standing directly behind his model for takeoff? He is instinctively trying to see yaw angle changes so he can quicken his responses.
Especially in a fast model, the pilot is basically flying a projectile.
Yes, a gyro works very well in yaw for a taildragger. My first experiment with a yaw rate gyro (mechanical gyro back in the 1980's) was to see if the gyro might help with spiral stability. Turned out that the yaw rate in the spiral mode was so slow that the gyro didn't sense the movement. But it sure did help with directional control on takeoff ! This gyro was a yaw damper only, with no heading hold. But it made takeoffs amazingly easy. I left the gyro in place for it's takeoff improvement and found that I had soon grown so lax that I could hardly manage if the gyro were turned off.
I have since adopted yaw rate gyros for most of my taildragger models. For the people who say that the gyro is a crutch, I say rubbish. The inertia of a model airplane is so much lower than a full scale that the divergence rate in a ground loop is far greater than on full scale. The rapid divergence of a model, plus the inability of the pilot to see or feel the initial stages of a ground loop, makes the model most unrealistic compared to full scale.
When we consider the feedback loop between airplane and pilot, the model is quite different from full scale. In full scale, we are flying primarily by attitude reference and accelerations. In a model, we are flying primarily by flight path direction. Two very different feedback schemes. Flight path is the integrated result of changes in attitude and acceleration, so the model pilot is always in a delay loop. Ever see a model pilot standing directly behind his model for takeoff? He is instinctively trying to see yaw angle changes so he can quicken his responses.
Especially in a fast model, the pilot is basically flying a projectile.
Last edited by otrcman; 06-24-2015 at 12:08 PM.
06-24-2015, 12:16 PM
#52
Back to trying to help Sal with his IIya.
I concur with the idea of making a smaller model to evaluate the stability of the design and determine a safe CG. Calculations are helpful, but the more a design differs from other well understood designs, the more error might occur in the calculated predictions.
Even a simple "chuck glider" would be instructive. While the severely undercambered wings are a consideration, even a simple profile fuselage model with flat sheet wings will tell you a lot. A step up in fidelity could be achieved using a cruciform fuselage (profile in two axes) along with cambered sheet balsa wings and tail. All told, the model could be a small as 18" or 24" span and still be useful.
Dick
I concur with the idea of making a smaller model to evaluate the stability of the design and determine a safe CG. Calculations are helpful, but the more a design differs from other well understood designs, the more error might occur in the calculated predictions.
Even a simple "chuck glider" would be instructive. While the severely undercambered wings are a consideration, even a simple profile fuselage model with flat sheet wings will tell you a lot. A step up in fidelity could be achieved using a cruciform fuselage (profile in two axes) along with cambered sheet balsa wings and tail. All told, the model could be a small as 18" or 24" span and still be useful.
Dick
06-25-2015, 09:12 AM
#53
Dick, this is an excellent reasoning how flying models is different from full-scale flying. Seriously, we could collect some of your posts and make a sticky guidebook thread from them.
As to a smaller IM model, as I reasoned above I don't think a chuckie would be too useful since it won't show the implications of the pitch-instable setup. But as BMatthews said, a small R/C model would. Could be even bigger than 4 ft, how about 5? I would definitely build cambered-plate wings and tail because I think they are a vital part of the configuration.
On second thought I would try Depron foam. The fuselage is simple, anyway, so could be built easily (doesn't need to be profile or cross section). The wings and the stab could be built from foam sheets and get the camber with the help of kind of ribs in the undercamber. The ailerons would be just part of the sheet, the hinges would be scribed (as those for elevator and rudder). Stick-and-thread bracing would be needed to get a rigid structure, but it could be hot-glued like all else. I didn't try it myself but saw this building technique and was impressed. Install micro R/C gear without gyros and you're ready to go.
Edit: I forgot, only one motor needed.
Still I'd be with Sal (I suppose). If I had nearly built the big model and feel confident I can handle it, the perspective of building a test model would suck. It would feel like being hold from doing the real thing. That's not a flight test center approach, I know.
As to a smaller IM model, as I reasoned above I don't think a chuckie would be too useful since it won't show the implications of the pitch-instable setup. But as BMatthews said, a small R/C model would. Could be even bigger than 4 ft, how about 5? I would definitely build cambered-plate wings and tail because I think they are a vital part of the configuration.
On second thought I would try Depron foam. The fuselage is simple, anyway, so could be built easily (doesn't need to be profile or cross section). The wings and the stab could be built from foam sheets and get the camber with the help of kind of ribs in the undercamber. The ailerons would be just part of the sheet, the hinges would be scribed (as those for elevator and rudder). Stick-and-thread bracing would be needed to get a rigid structure, but it could be hot-glued like all else. I didn't try it myself but saw this building technique and was impressed. Install micro R/C gear without gyros and you're ready to go.
Edit: I forgot, only one motor needed.
Still I'd be with Sal (I suppose). If I had nearly built the big model and feel confident I can handle it, the perspective of building a test model would suck. It would feel like being hold from doing the real thing. That's not a flight test center approach, I know.
Last edited by UStik; 06-25-2015 at 11:44 PM.
06-25-2015, 05:19 PM
#54
As long as the wing, stab, and fuselage proportions are correct the neutral point will be proportionally the same regardless of scale, so a chuck glider should provide a safe CG for the larger model.
This is just the sort of plane you want to get a CG for before you fly, because if you choose a CG farther forward than necessary you will have to go to a lot of trouble to achieve it.
Fantastic project!
This is just the sort of plane you want to get a CG for before you fly, because if you choose a CG farther forward than necessary you will have to go to a lot of trouble to achieve it.
Fantastic project!
06-25-2015, 06:05 PM
#55
Thank you for the kind words, Burkhard. Unfortunately there are only a few of us pilot-engineer-modeler types who have an appreciation for the nuances of model flight.
One thing we haven't talked about for the IM is the possibility of a longitudinal "tuck" at low angle of attack & high speed. I know that some early airplanes such as JN-4 Jenny had this phenomenon. The normal portion of the pitch range seems quite normal, but the airplane becomes progressively unstable as speed goes up and AOA goes down.
Some old timer models have this same phenomenon, which seems to be connected to thin, highly cambered wings. I've never been sure whether the model tuck is due primarily to the pitching moment characteristics or to aeroelastic effects. Maybe both. Two model designs that come to mind are the Goldberg Sailplane and the Cleveland Playboy. The Playboy onset is fairly gradual and it is possible to do a shallow dive and maintain the dive while holding up elevator to keep things from getting worse. Then you can pull out of the dive by adding more up elevator. The Sailplane is a bit more dramatic. The onset is subtle and then dramatic. If you catch it soon enough it is possible to pull out, but I have found myself locked in a power dive with fluttering wings while holding full back stick to no avail. In my case, I was able to shut off the engine in time to slow down and afford a pull out.
Because tucking is a possibility with the IM, Sal, I would suggest that you keep in mind that chopping power is an historical solution if you find yourself in such a dive. With the Sailplane the problem is so severe that, if you let the model gain speed too much before takeoff, it will not become airborne. It seems to lock itself onto the ground in a tail high attitude no matter how much back stick is applied.
Dick
One thing we haven't talked about for the IM is the possibility of a longitudinal "tuck" at low angle of attack & high speed. I know that some early airplanes such as JN-4 Jenny had this phenomenon. The normal portion of the pitch range seems quite normal, but the airplane becomes progressively unstable as speed goes up and AOA goes down.
Some old timer models have this same phenomenon, which seems to be connected to thin, highly cambered wings. I've never been sure whether the model tuck is due primarily to the pitching moment characteristics or to aeroelastic effects. Maybe both. Two model designs that come to mind are the Goldberg Sailplane and the Cleveland Playboy. The Playboy onset is fairly gradual and it is possible to do a shallow dive and maintain the dive while holding up elevator to keep things from getting worse. Then you can pull out of the dive by adding more up elevator. The Sailplane is a bit more dramatic. The onset is subtle and then dramatic. If you catch it soon enough it is possible to pull out, but I have found myself locked in a power dive with fluttering wings while holding full back stick to no avail. In my case, I was able to shut off the engine in time to slow down and afford a pull out.
Because tucking is a possibility with the IM, Sal, I would suggest that you keep in mind that chopping power is an historical solution if you find yourself in such a dive. With the Sailplane the problem is so severe that, if you let the model gain speed too much before takeoff, it will not become airborne. It seems to lock itself onto the ground in a tail high attitude no matter how much back stick is applied.
Dick
06-26-2015, 04:40 AM
#56
You are so right, Dick. As Sal is with his statement. Sigh... ;-)
As to tucking, just the discussion last year about tucking (I think of a Mambo) was one example I thought of when remembering your reasonings. I didn't apply it to this case, though. I only thought of the center of lift moving forward when AoA increases. I excluded small AoAs because I would avoid them, anyway, with undercambered airfoils on wings and stab (to avoid the bottom side stall).
Undoubtedly, IM's wings will produce a big down-pitchting moment. Even if the stab might contribute and aeroelastic effects as well, I guess the wing to be the biggest player in that game. While the center of lift (CoL) is at 30% chord in slow flight, the airfoil measurements stop at Cl=0.3 where the CoL is at 65%. Here's where bottom-side stall begins and I don't know how that affects CoL.
But...this stall makes for quite noticeable drag, which may prevent even bigger airspeed. At least the original IM could have had so much drag and so weak engines that there was no risk of tucking under. In fact I even think they could have put up with pitch instability to have a really lifting stab due to lack of engine power in 1913. Anyway, Sal, you could lay out the drives of your IM model (battery voltage, motor Kv, prop pitch) to avoid too much speed, which would be 40 mph for a 0.3 lift coefficient. A dive wouldn't do much for more speed due to the big drag emerging then.
Aside from undercamber, the stab's size (33% of wing area like a very stable and damped model) might have been intended to avoid tucking. Maybe they knew about this problem, after all they used those thin undercambered airfoils exclusively. The more camber the more pitching moment, even at zero lift. Since the moment increases with square of speed, independent of AoA, it can be counteracted only by a big stab whose lift increases with square of speed as well (or by a reflex in the airfoil).
I like such speculations but by and by they begin to suck. Should I quantify things, a moment plan, a simulator model?
As to tucking, just the discussion last year about tucking (I think of a Mambo) was one example I thought of when remembering your reasonings. I didn't apply it to this case, though. I only thought of the center of lift moving forward when AoA increases. I excluded small AoAs because I would avoid them, anyway, with undercambered airfoils on wings and stab (to avoid the bottom side stall).
Undoubtedly, IM's wings will produce a big down-pitchting moment. Even if the stab might contribute and aeroelastic effects as well, I guess the wing to be the biggest player in that game. While the center of lift (CoL) is at 30% chord in slow flight, the airfoil measurements stop at Cl=0.3 where the CoL is at 65%. Here's where bottom-side stall begins and I don't know how that affects CoL.
But...this stall makes for quite noticeable drag, which may prevent even bigger airspeed. At least the original IM could have had so much drag and so weak engines that there was no risk of tucking under. In fact I even think they could have put up with pitch instability to have a really lifting stab due to lack of engine power in 1913. Anyway, Sal, you could lay out the drives of your IM model (battery voltage, motor Kv, prop pitch) to avoid too much speed, which would be 40 mph for a 0.3 lift coefficient. A dive wouldn't do much for more speed due to the big drag emerging then.
Aside from undercamber, the stab's size (33% of wing area like a very stable and damped model) might have been intended to avoid tucking. Maybe they knew about this problem, after all they used those thin undercambered airfoils exclusively. The more camber the more pitching moment, even at zero lift. Since the moment increases with square of speed, independent of AoA, it can be counteracted only by a big stab whose lift increases with square of speed as well (or by a reflex in the airfoil).
I like such speculations but by and by they begin to suck. Should I quantify things, a moment plan, a simulator model?
Last edited by UStik; 06-26-2015 at 10:33 AM. Reason: personal year-2000 problem
06-26-2015, 05:16 AM
#57
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
Gents..
Oh why oh why did I pick this project!!!!! Too late now, I am well into it.. How about some wind tunnel testing??????
Check out the engine details on the build thread.. At the very least, it gonna look good!
Cheers, Sal
Oh why oh why did I pick this project!!!!! Too late now, I am well into it.. How about some wind tunnel testing??????
Check out the engine details on the build thread.. At the very least, it gonna look good!
Cheers, Sal
06-26-2015, 06:18 AM
#58
Yeah, looking great, you don't cease to amaze me! But don't take hard what we are ranting here, we're still trying to make things easier for you. The speculation is just the fun we have from your project, but we always come up with good news in the end.
Instable setup? Yes, but no problem.
No spiral stability? No big deal.
Tucking in a dive? Just chop power.
Unusual characteristics? You'll adapt easily.
And there are the modern aids you could use.
Lock gyro for elevator.
Rate gyro for rudder.
Lock gyro for ailerons.
Electric drive setup.
Hope to cheer you up!
Burkhard
Instable setup? Yes, but no problem.
No spiral stability? No big deal.
Tucking in a dive? Just chop power.
Unusual characteristics? You'll adapt easily.
And there are the modern aids you could use.
Lock gyro for elevator.
Rate gyro for rudder.
Lock gyro for ailerons.
Electric drive setup.
Hope to cheer you up!
Burkhard
06-26-2015, 08:48 AM
#60
Sal,
Yes, we will most certainly stick with you and your project. Very few people have your talent and determination to take on an airplane like the IM, and we relish the idea of being even a small part of it.
Regarding the "tuck". Perhaps I made too much of it. Without going into the technical details, I can tell you that your electric propulsion system will act to inhibit speed buildup in a dive. For this airplane, that's a big plus.
Dick
Yes, we will most certainly stick with you and your project. Very few people have your talent and determination to take on an airplane like the IM, and we relish the idea of being even a small part of it.
Regarding the "tuck". Perhaps I made too much of it. Without going into the technical details, I can tell you that your electric propulsion system will act to inhibit speed buildup in a dive. For this airplane, that's a big plus.
Dick
Last edited by otrcman; 06-26-2015 at 08:56 AM.
06-26-2015, 10:14 PM
#61
No Sal, not a wind tunnel test! (That was a good one.) ;-)
But I did a simulator model (no appearance, only physics parameters) because I need "tangible" inspections to make me aware of things. There are so many interesting aspects in the design that I see them but don't think about them. Quite like hidden in clear sight.
Sikorski really was a remarkable designer. Seems he knew what he was doing and din't need much trial-and-error (or had his fair share with his 21 previous designs). That's why it's so interesting to figure him out (at least trying).
Well, look at the wings. Actually it's like a quite square and symmetric biplane, apart from a bit negative stagger. Then there are those appendices to the upper wings. They carry, or they are the ailerons. These are appended to the airfoil with a reflex, effectively making for less incidence angle. Supposedly this makes the ailerons effective and lessens adverse yaw. It's also like a wing washout.
And that's required. The kind of undercambered airfoils used here has some remarkable traits. One of them is the behavior near (positive) stall. The airfoil has much drag in the first place, but while lift increases rapidly with increasing AoA, drag grows only slowly. Then 2 or 3 degrees before stall, lift stops increasing (but doesn't decrease either) and drag grows rapidly. That means if the airplane is pitched up to more than this magic maximum lift / minimum drag AoA, it loses roll stability, gets wobbly in yaw and roll. Maybe that's why even a Piper Cub has some washout.
Sal, I think this is another case where you can rely on Sikorski that he has taken all necessary precautions. The real thing flew, the model will fly as well (literally).
But I did a simulator model (no appearance, only physics parameters) because I need "tangible" inspections to make me aware of things. There are so many interesting aspects in the design that I see them but don't think about them. Quite like hidden in clear sight.
Sikorski really was a remarkable designer. Seems he knew what he was doing and din't need much trial-and-error (or had his fair share with his 21 previous designs). That's why it's so interesting to figure him out (at least trying).
Well, look at the wings. Actually it's like a quite square and symmetric biplane, apart from a bit negative stagger. Then there are those appendices to the upper wings. They carry, or they are the ailerons. These are appended to the airfoil with a reflex, effectively making for less incidence angle. Supposedly this makes the ailerons effective and lessens adverse yaw. It's also like a wing washout.
And that's required. The kind of undercambered airfoils used here has some remarkable traits. One of them is the behavior near (positive) stall. The airfoil has much drag in the first place, but while lift increases rapidly with increasing AoA, drag grows only slowly. Then 2 or 3 degrees before stall, lift stops increasing (but doesn't decrease either) and drag grows rapidly. That means if the airplane is pitched up to more than this magic maximum lift / minimum drag AoA, it loses roll stability, gets wobbly in yaw and roll. Maybe that's why even a Piper Cub has some washout.
Sal, I think this is another case where you can rely on Sikorski that he has taken all necessary precautions. The real thing flew, the model will fly as well (literally).
Last edited by UStik; 06-27-2015 at 12:09 AM.
06-27-2015, 01:44 AM
#62
A quick provisional report from the simulator, just can't wait.
But the reservations first:
The simulator is still a quite coarse rendering, so all has to be taken with a grain of salt (to say the least).
Effects of propwash on wing and tail have been excluded so far. Seem to be vital but are not yet understood.
Now to the observations:
It needed 4 degrees washout to get a straight, forward stall. Should be the same as the reflexed outer wings mentioned above.
Aileron effect is good. Not that much adverse yaw. Top aileron required, hence hold gyro could be useful. Not really needed though, all is slow motion.
Yaw as well, rudder effect is good, cancelling adverse yaw is easy. Seems full rudder is needed if outer engine is dead, but not sure yet. Rate gyro (damper) not needed, not even for take-off.
The C/G has been set close to the trailing edge, aft of the neutral point, so this is an unstable setup (-14% static margin).
There's very, very much pitch damping. The airplane doesn't seem to be unstable at all, just neutral. (Dick, seems your example applies here.)
In the normal pitch range (no steep climb, no real dive) it just maintains attitude. To my surprise, it climbs with full power and descends with little power, like an ordinary Cessna. So it seems like flying an approach or a climb all the time, even in level flight: set and maintain proper pitch and control flight path with power.
It's a draggy airplane, needs power in turns. Still a glide is easy, just a bit steeper than usual.
Only a bit, please! Dick, it seems this airplane tucks, however only at dive angles bigger than 45 degrees. It starts slowly and pulling out is easy and quick. The airplane needed a stab incidence bigger than wing incidence to balance the aft C/G so I think that's the cause of tucking and the damping makes it easy to cope with it.
More awareness might be required to avoid too much pitch up. Due to drag and airfoil the plane flies not that fast and needs big AoAs, near the magic value mentioned above. If gone beyond that point it doesn't wobble due to washout, but drag increases, speed is lost, and stall appears. Here's where the instability actually appears, gets noticeable!
Stall is severe in the simulator, and that could be realistic considering the airfoil. Recovery is standard, if you are used to pushing the elevator in such a situation, that is. And severe means some altitude is lost (one to two wingspans, it seems).
Landing approach can be done without power, just keep speed with down pitch. Decent power will flatten the approach.
There's a bit of a flare, but unconsciously pitching up during the approach (like a beginner) is fatal (IM is not a Cessna). You must fly the airplane to the runway. It is not flared to three-point attitude, at least not in the simulator. If it's slow it needs power for flaring and chopping power let's it plop down (good, it will stay on the ground). If it's a bit faster, just flare with a bit elevator but don't stall it. Only little elevator needed!
Take-off is a bit special as well. Like with any taildragger, the tail is lifted with down elevator. But IM seems to need it or it won't take-off at all, just stick to the ground. Maybe that's different when some propwash is set in the simulator. But I think the real reason is the steep attitude of the simulator model which keeps wing and stab (even more so) in a stalled condition.
IM's tailskid is high, so just don't flare more than needed for a three-point, and maybe the tail will lift easily in the takeoff run.
My bottom line:
IM seems to be a majestically flying airplane if all is set properly and as long as you're on top of it. (Dick, I hear you.)
Staying on top means finding the proper pitch and power for each situation. Straight and level is really easy. Turns, takeoff, and landing are critical (more than with ordinary airplanes).
A pitch lock gyro would be handy as an aid in staying on top, even far away and in gusty wind.
Such a gyro would be a big help as long as not all is set properly yet.
That means stab incidence for me. The simulator demonstrates that the rear C/G is not bad, but - as always - then the plane has to be balanced with decalage. Obviously, a stab incidence bigger than wing incidence should not be dreaded in this case.
But the reservations first:
The simulator is still a quite coarse rendering, so all has to be taken with a grain of salt (to say the least).
Effects of propwash on wing and tail have been excluded so far. Seem to be vital but are not yet understood.
Now to the observations:
It needed 4 degrees washout to get a straight, forward stall. Should be the same as the reflexed outer wings mentioned above.
Aileron effect is good. Not that much adverse yaw. Top aileron required, hence hold gyro could be useful. Not really needed though, all is slow motion.
Yaw as well, rudder effect is good, cancelling adverse yaw is easy. Seems full rudder is needed if outer engine is dead, but not sure yet. Rate gyro (damper) not needed, not even for take-off.
The C/G has been set close to the trailing edge, aft of the neutral point, so this is an unstable setup (-14% static margin).
There's very, very much pitch damping. The airplane doesn't seem to be unstable at all, just neutral. (Dick, seems your example applies here.)
In the normal pitch range (no steep climb, no real dive) it just maintains attitude. To my surprise, it climbs with full power and descends with little power, like an ordinary Cessna. So it seems like flying an approach or a climb all the time, even in level flight: set and maintain proper pitch and control flight path with power.
It's a draggy airplane, needs power in turns. Still a glide is easy, just a bit steeper than usual.
Only a bit, please! Dick, it seems this airplane tucks, however only at dive angles bigger than 45 degrees. It starts slowly and pulling out is easy and quick. The airplane needed a stab incidence bigger than wing incidence to balance the aft C/G so I think that's the cause of tucking and the damping makes it easy to cope with it.
More awareness might be required to avoid too much pitch up. Due to drag and airfoil the plane flies not that fast and needs big AoAs, near the magic value mentioned above. If gone beyond that point it doesn't wobble due to washout, but drag increases, speed is lost, and stall appears. Here's where the instability actually appears, gets noticeable!
Stall is severe in the simulator, and that could be realistic considering the airfoil. Recovery is standard, if you are used to pushing the elevator in such a situation, that is. And severe means some altitude is lost (one to two wingspans, it seems).
Landing approach can be done without power, just keep speed with down pitch. Decent power will flatten the approach.
There's a bit of a flare, but unconsciously pitching up during the approach (like a beginner) is fatal (IM is not a Cessna). You must fly the airplane to the runway. It is not flared to three-point attitude, at least not in the simulator. If it's slow it needs power for flaring and chopping power let's it plop down (good, it will stay on the ground). If it's a bit faster, just flare with a bit elevator but don't stall it. Only little elevator needed!
Take-off is a bit special as well. Like with any taildragger, the tail is lifted with down elevator. But IM seems to need it or it won't take-off at all, just stick to the ground. Maybe that's different when some propwash is set in the simulator. But I think the real reason is the steep attitude of the simulator model which keeps wing and stab (even more so) in a stalled condition.
IM's tailskid is high, so just don't flare more than needed for a three-point, and maybe the tail will lift easily in the takeoff run.
My bottom line:
IM seems to be a majestically flying airplane if all is set properly and as long as you're on top of it. (Dick, I hear you.)
Staying on top means finding the proper pitch and power for each situation. Straight and level is really easy. Turns, takeoff, and landing are critical (more than with ordinary airplanes).
A pitch lock gyro would be handy as an aid in staying on top, even far away and in gusty wind.
Such a gyro would be a big help as long as not all is set properly yet.
That means stab incidence for me. The simulator demonstrates that the rear C/G is not bad, but - as always - then the plane has to be balanced with decalage. Obviously, a stab incidence bigger than wing incidence should not be dreaded in this case.
06-27-2015, 06:47 PM
#63
Excellent report, Burkhard. And encouraging as well.
One thing this tells me, Sal, is that you should probably have some means of adjusting the incidence of the horizontal stab. I don't know if you had planned to permanently mount the stab or to screw it down. Screw type attachment might be advisable so you can add or subtract shims if necessary.
Dick
One thing this tells me, Sal, is that you should probably have some means of adjusting the incidence of the horizontal stab. I don't know if you had planned to permanently mount the stab or to screw it down. Screw type attachment might be advisable so you can add or subtract shims if necessary.
Dick
06-28-2015, 06:40 AM
#64
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
Hi Dick.. Too late, I designed the stab with two carbon fiber tubes as spars, more or less equally spaced from each other. I use short carbon rods as joiners that correspond with the tubes in the fuse.In other words, the incidence I have is the incidence I gonna have. I should have thought ahead and made it adjustable. My bad....Cheers, Sal
06-28-2015, 09:35 AM
#65
Sal, that should be no big deal, I dare saying that this time. I'm quite confident that the original was well thought out and the geometrically equal (just smaller) model will fly well. The big elevator should be able to balance the model. I just can't prove that since both my stability calculator and the simulator simply assume a nice "round" wing airfoil polar, a symmetric stab airfoil, and positive stability. I can somewhat circumvent them but not completely.
But negative stability is basically rendered, as the simulator flights show. That's why the simulator model now takes off easily after I made the tailskid higher. The model picks up speed until the tail raises automatically and takeoff occurs. It's quite surprising to see that, and it's good.
Several things are counterintuitive on IM, but only because we're not used to it. The airplane seems even quite good-natured as long as the normal flight envelope is observed. If you fail to do that it's a beast, though. So finding a good balance may not really be a problem but you might rather have a hard time getting used to the things that are counterintuitive in the first place. After that (programming the flight computer between the ears) flying the IM should be easy.
You could use the simulator model - even if it's crude - to practise. You should set up the real model to assist you in any case.
I would set different elevator throws up and down, maybe by dual rate. Up should be small (10° or less) to avoid stall, down should be big (25°) to balance the airplane.
Permanently pushing the elevator stick (when the airplane is not yet balanced) is cumbersome, hence a pitch hold gyro. Takeoff (raising the tail) might be tricky, both with unbalanced and balanced airplane, so I think about switching it on only after takeoff (and maybe before landing as well).
Adverse yaw is big but easily balanced by rudder. Now top aileron required even in standard turns is really big, and I feel a bit annoyed if the aileron-rudder mix is on and the rate of turn seems to be diminished by top aileron. It feels better if the airplane is banked with the mixer on and top aileron held with mixer off, but switching the mixer is annoying as well. In the simulator, working the rudder by hand is the best solution, but in reality it's rather a bank-hold gyro.
Maybe just for the record, there's another idea, not by me but by Blaine Beron-Rawdon whose stability calculator I'm using. He suggested a rate gyro to stabilize a thermal glider in circling (http://envisiondesignusa.com/evdusa/...mentation.html, mind the title). Interestingly, a two-channel (aileron) gyro is put on the airplane's yaw axis so top aileron is about proportional to rate of turn. By adjusting gyro gain, different bank angles and rates of turn can be established. If the gyro axis is inclined 30° backwards (like in my glider, next three pictures: http://time.hs-augsburg.de/~erd/Mode...ar.html#Fixtur) it will even make for some roll stabilization in straight flight. All this would be nice also for IM but it has to be set up / adjusted so I'd put the gyro on the roll axis and switch hold mode on for the first flights.
There are no such solutions for other "problems" like down elevator required in some situations, power needed in turns, minimum speed avoidance, and such. That would require a programmable flight computer but is a completely different department as well as FPV. I think only about means to make all, and especially the first flights safer.
But negative stability is basically rendered, as the simulator flights show. That's why the simulator model now takes off easily after I made the tailskid higher. The model picks up speed until the tail raises automatically and takeoff occurs. It's quite surprising to see that, and it's good.
Several things are counterintuitive on IM, but only because we're not used to it. The airplane seems even quite good-natured as long as the normal flight envelope is observed. If you fail to do that it's a beast, though. So finding a good balance may not really be a problem but you might rather have a hard time getting used to the things that are counterintuitive in the first place. After that (programming the flight computer between the ears) flying the IM should be easy.
You could use the simulator model - even if it's crude - to practise. You should set up the real model to assist you in any case.
I would set different elevator throws up and down, maybe by dual rate. Up should be small (10° or less) to avoid stall, down should be big (25°) to balance the airplane.
Permanently pushing the elevator stick (when the airplane is not yet balanced) is cumbersome, hence a pitch hold gyro. Takeoff (raising the tail) might be tricky, both with unbalanced and balanced airplane, so I think about switching it on only after takeoff (and maybe before landing as well).
Adverse yaw is big but easily balanced by rudder. Now top aileron required even in standard turns is really big, and I feel a bit annoyed if the aileron-rudder mix is on and the rate of turn seems to be diminished by top aileron. It feels better if the airplane is banked with the mixer on and top aileron held with mixer off, but switching the mixer is annoying as well. In the simulator, working the rudder by hand is the best solution, but in reality it's rather a bank-hold gyro.
Maybe just for the record, there's another idea, not by me but by Blaine Beron-Rawdon whose stability calculator I'm using. He suggested a rate gyro to stabilize a thermal glider in circling (http://envisiondesignusa.com/evdusa/...mentation.html, mind the title). Interestingly, a two-channel (aileron) gyro is put on the airplane's yaw axis so top aileron is about proportional to rate of turn. By adjusting gyro gain, different bank angles and rates of turn can be established. If the gyro axis is inclined 30° backwards (like in my glider, next three pictures: http://time.hs-augsburg.de/~erd/Mode...ar.html#Fixtur) it will even make for some roll stabilization in straight flight. All this would be nice also for IM but it has to be set up / adjusted so I'd put the gyro on the roll axis and switch hold mode on for the first flights.
There are no such solutions for other "problems" like down elevator required in some situations, power needed in turns, minimum speed avoidance, and such. That would require a programmable flight computer but is a completely different department as well as FPV. I think only about means to make all, and especially the first flights safer.
Last edited by UStik; 06-28-2015 at 09:49 AM.
06-28-2015, 04:58 PM
#66
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
guten Morgen Burkhard..
Responding from your second to last post. The wing has polyhedral, The upper outer panels that support the ailerons have a dihedral of a few degrees, in addition the aileron,
which is very deep, has what I would call inverted undercamber. I would guess that this provides some level of washout, which is a very good thing.
In addition, the wings are parallel to each other, since there is a great deal of positive incidence, the upper wing is pushed back some to form a reverse stagger.
Here is a photo that shows the Ilya Muromets in flight. What is interesting is that, like I read, it flies nose down.
More later.
Cheers, Sal
Responding from your second to last post. The wing has polyhedral, The upper outer panels that support the ailerons have a dihedral of a few degrees, in addition the aileron,
which is very deep, has what I would call inverted undercamber. I would guess that this provides some level of washout, which is a very good thing.
In addition, the wings are parallel to each other, since there is a great deal of positive incidence, the upper wing is pushed back some to form a reverse stagger.
Here is a photo that shows the Ilya Muromets in flight. What is interesting is that, like I read, it flies nose down.
More later.
Cheers, Sal
06-29-2015, 04:08 AM
#67
Guten Morgen Sal,
I ignored the polyhedral so far, sorry. It makes for some spiral stability both in the calculation and the simulator model so there's less top aileron required. But the wingspan is just too big and the amount of top aileron required is still bigger than usual. A roll gyro would be nice.
Again an interesting picture of IM, quite nose-down. Since the simulator model does likewise I suspect that was a fast fly-by (for demonstration, maybe without load). At least nobody stands on top of the fuselage like in the other picture.
Anyway, the simulator seems to be set up (wing incidence) and trimmed realistically. With the rearward C/G, I just increased stab incidence until I got comfortable behavior in powered flight. That turned out to be also the correct trim for power-off glide (22 mph). It was just hard to fly the model "tail heavy", that is with too little stab incidence. Easier to begin with is too much stab incidence ("nose heavy", as always), which could be achieved by down elevator on the real model (hence pitch gyro).
I had started with a wild guess of drive power. Now I did an educated guess, assuming an 8 cell LiPo for your AXI 4130/20 and 16x8 prop. That turned out to be a good choice, but my assumption was rather bad. 8 cells make for about 12.5 lbf thrust per motor, meaning more than 1:1 thrust:weight ratio. Other than usual I wouldn't say "better more than less power" in this case. Of course you can always throttle back, but one of Dick's hints applies: You'll see the result of too much power quite late, only when tucking begins. The more speed the earlier (dive angle) tucking begins. Still it is stopped easily with up elevator, but here 25° throw is advisable and you don't have time to flip the dual-rate switch. Of course it's nice to see the model leap off the ground but that's not realistic anyway and not needed. The motor is rated for 6 to 8 cells LiPo, and 6 cells will do nicely. Well, I should stop rambling and just ask what kind of cells and how many you intend to use.
I flew the simulator model a bit this last weekend. The more I fly it the more I love it. It really was a matter of getting used to it, like learning where to fly and what to avoid on a new airfield. The airplane does nothing unusual, it's just satisfying to fly it around, knowing and avoiding the limits without having to think about them. It's a warm feeling to know you're on top of it, all is under control. It's like a tamed wild animal in the circus, the tamer can treat the tiger like a kitten as long as he doesn't the slightest mistake, and he knows he is far from that. (I know I'm exaggerating/rambling, but I like to play with the English language for practice.)
My funny assessment in my first post couldn't be wronger: The model is not sensitive to wind. In a 3 Beaufort wind with appropriate gusts, the simulator model is remarkably steady. Actually you notice the wind only by a bit yaw wobbling, which could be eliminated by a yaw damper gyro.
Looking forward to your further information.
Burkhard
P.S.: Nose-down in fast flight means you have to avoid any nose-up attitude in any flight regime, even in touch-down. Just do a wheel landing and let the tail settle. Touch-down speed is less than 20 mph and is quickly dissipated (short roll). If the model is trimmed as decribed above, you may leave the elevator alone and control approach path with power, if there's no pitch hold gyro. More and more I think such a gyro would be needed for the first flights but no longer when the model is trimmed. It doesn't help in avoiding too much pitch, anyway. You can (and maybe should) fly this model like an old rudder-and-throttle (class 1) model. It needs more speed (hence power) in turns and even flare is done with power. Other than the old class-1 models, this one needs that because it's on the verge of stalling if trimmed for glide, what is recommendable because it's then trimmed for powered flight as well.There'll be just enough speed reserve for a nudge to flare and to cope with gusts. I believe this behavior was meant when they wrote IM was hard to fly.
I ignored the polyhedral so far, sorry. It makes for some spiral stability both in the calculation and the simulator model so there's less top aileron required. But the wingspan is just too big and the amount of top aileron required is still bigger than usual. A roll gyro would be nice.
Again an interesting picture of IM, quite nose-down. Since the simulator model does likewise I suspect that was a fast fly-by (for demonstration, maybe without load). At least nobody stands on top of the fuselage like in the other picture.
Anyway, the simulator seems to be set up (wing incidence) and trimmed realistically. With the rearward C/G, I just increased stab incidence until I got comfortable behavior in powered flight. That turned out to be also the correct trim for power-off glide (22 mph). It was just hard to fly the model "tail heavy", that is with too little stab incidence. Easier to begin with is too much stab incidence ("nose heavy", as always), which could be achieved by down elevator on the real model (hence pitch gyro).
I had started with a wild guess of drive power. Now I did an educated guess, assuming an 8 cell LiPo for your AXI 4130/20 and 16x8 prop. That turned out to be a good choice, but my assumption was rather bad. 8 cells make for about 12.5 lbf thrust per motor, meaning more than 1:1 thrust:weight ratio. Other than usual I wouldn't say "better more than less power" in this case. Of course you can always throttle back, but one of Dick's hints applies: You'll see the result of too much power quite late, only when tucking begins. The more speed the earlier (dive angle) tucking begins. Still it is stopped easily with up elevator, but here 25° throw is advisable and you don't have time to flip the dual-rate switch. Of course it's nice to see the model leap off the ground but that's not realistic anyway and not needed. The motor is rated for 6 to 8 cells LiPo, and 6 cells will do nicely. Well, I should stop rambling and just ask what kind of cells and how many you intend to use.
I flew the simulator model a bit this last weekend. The more I fly it the more I love it. It really was a matter of getting used to it, like learning where to fly and what to avoid on a new airfield. The airplane does nothing unusual, it's just satisfying to fly it around, knowing and avoiding the limits without having to think about them. It's a warm feeling to know you're on top of it, all is under control. It's like a tamed wild animal in the circus, the tamer can treat the tiger like a kitten as long as he doesn't the slightest mistake, and he knows he is far from that. (I know I'm exaggerating/rambling, but I like to play with the English language for practice.)
My funny assessment in my first post couldn't be wronger: The model is not sensitive to wind. In a 3 Beaufort wind with appropriate gusts, the simulator model is remarkably steady. Actually you notice the wind only by a bit yaw wobbling, which could be eliminated by a yaw damper gyro.
Looking forward to your further information.
Burkhard
P.S.: Nose-down in fast flight means you have to avoid any nose-up attitude in any flight regime, even in touch-down. Just do a wheel landing and let the tail settle. Touch-down speed is less than 20 mph and is quickly dissipated (short roll). If the model is trimmed as decribed above, you may leave the elevator alone and control approach path with power, if there's no pitch hold gyro. More and more I think such a gyro would be needed for the first flights but no longer when the model is trimmed. It doesn't help in avoiding too much pitch, anyway. You can (and maybe should) fly this model like an old rudder-and-throttle (class 1) model. It needs more speed (hence power) in turns and even flare is done with power. Other than the old class-1 models, this one needs that because it's on the verge of stalling if trimmed for glide, what is recommendable because it's then trimmed for powered flight as well.There'll be just enough speed reserve for a nudge to flare and to cope with gusts. I believe this behavior was meant when they wrote IM was hard to fly.
Last edited by UStik; 06-29-2015 at 05:14 AM.
06-29-2015, 05:14 AM
#68
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
Burkhard..
I was planning on using a 5 or 6 cell, 4500-5000mah LiPo pack. Of course, 4 of them, one for each motor. The packs will be easily accessible as they will be installed between the engine mount supports, under each V8 engine. This is another engineering problem I have to work out once I focus on that area again. Right now I am working on the wing panels. I am installing all the hard points for struts, landing gear, etc. I should be covering wings soon.
Cheers, Sal
I was planning on using a 5 or 6 cell, 4500-5000mah LiPo pack. Of course, 4 of them, one for each motor. The packs will be easily accessible as they will be installed between the engine mount supports, under each V8 engine. This is another engineering problem I have to work out once I focus on that area again. Right now I am working on the wing panels. I am installing all the hard points for struts, landing gear, etc. I should be covering wings soon.
Cheers, Sal
06-29-2015, 05:38 AM
#69
Sounds good, Sal, I imagine the problem but it's good to have the batteries as far forward as possible.
As to 5 or 6 cells, both will do. 6 cells will have more zip, more like we are used to today. 5 cells just suffice and will feel more like the original with scarce power, especially since the lower voltage means less top speed. If I'd want to be proud to be the tamer of this beast, I'd take only 5 cells. Only moderate climb, only straight ahead. (In the old times, if someone had tried climbing and turning at the same time, they said at the funeral: He wouldn't have become a good pilot, anyway.) Even full-power do no turns steeper than standard! For peace of mind, meaning a bit speed and climb reserve, I would take 6 cells. A good compromise could be 5 cells with a 16x10 prop, though.
Edit: Yes, only a compromise. A bit more speed solves the turn problem (and the dreaded downwind turn I forgot to mention so far), but there's not more power which would be needed for better climb. (Since you can't pitch up further to avoid stall you must fly faster in order to climb faster.) I'd take 6 cells and 16x8 prop.
6s with a 16x10 would be even better. Even more speed, so steeper turns and faster climb as well. Just watch out for tucking (the counterintuitive negative stability). And the drive would be noticeably more efficient in cruise flight, could mean 5 minutes more duration.
I guess less than 10 Amps are needed for cruise flight in any case so you should have safe 20 minutes per flight. Since last year I use telemetry (remaining-charge display and warning) to monitor duration.
Hey, wouldn't it be nice to have telemetry for the flight data? A speed indicator with voice output and stall warning? I know they didn't have that in 1913 but they were in the airplane, not on the ground.
As to 5 or 6 cells, both will do. 6 cells will have more zip, more like we are used to today. 5 cells just suffice and will feel more like the original with scarce power, especially since the lower voltage means less top speed. If I'd want to be proud to be the tamer of this beast, I'd take only 5 cells. Only moderate climb, only straight ahead. (In the old times, if someone had tried climbing and turning at the same time, they said at the funeral: He wouldn't have become a good pilot, anyway.) Even full-power do no turns steeper than standard! For peace of mind, meaning a bit speed and climb reserve, I would take 6 cells. A good compromise could be 5 cells with a 16x10 prop, though.
Edit: Yes, only a compromise. A bit more speed solves the turn problem (and the dreaded downwind turn I forgot to mention so far), but there's not more power which would be needed for better climb. (Since you can't pitch up further to avoid stall you must fly faster in order to climb faster.) I'd take 6 cells and 16x8 prop.
6s with a 16x10 would be even better. Even more speed, so steeper turns and faster climb as well. Just watch out for tucking (the counterintuitive negative stability). And the drive would be noticeably more efficient in cruise flight, could mean 5 minutes more duration.
I guess less than 10 Amps are needed for cruise flight in any case so you should have safe 20 minutes per flight. Since last year I use telemetry (remaining-charge display and warning) to monitor duration.
Hey, wouldn't it be nice to have telemetry for the flight data? A speed indicator with voice output and stall warning? I know they didn't have that in 1913 but they were in the airplane, not on the ground.
Last edited by UStik; 06-29-2015 at 07:40 AM.
06-30-2015, 06:11 PM
#70
Join Date: Oct 2002
Location: Chilliwack, BC, CANADA
Posts: 12,425
Likes: 0
Received 22 Likes
on
19 Posts
It's possible that the picture above was taken of an empty plane other than the crew. With a load of passengers or bombs it would need to fly at a higher angle of attack and likely the fuselage would not be so tail high.
06-30-2015, 06:40 PM
#71
Also, it may be coming in for a landing!
The stab is huge, the tail arm is long, and the wing chord is very narrow. All those things suggest you'll be fine with a balance point very far aft of where we usually balance our planes. Stab incidence looks to be about the same as the wing incidence, so I suspect you'll be fine with it. I agree that the reflex on the ailerons is likely to be helpful.
Is your model going to a have a 190+ inch wingspan? Cool!
Jim
The stab is huge, the tail arm is long, and the wing chord is very narrow. All those things suggest you'll be fine with a balance point very far aft of where we usually balance our planes. Stab incidence looks to be about the same as the wing incidence, so I suspect you'll be fine with it. I agree that the reflex on the ailerons is likely to be helpful.
Is your model going to a have a 190+ inch wingspan? Cool!
Jim
07-01-2015, 06:25 AM
#72
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
Jim
The model is 1/6 scale with a wingspan around 195 inches. The tail and wing incidence are the same as far as I can figure. I am using a number of three views
and they all seem to show the same configuration. Yes, the stab is huge.. The model comes out to around 6 feet.
The caption for that photo states that this is the way the Ilya Muromets flies. It is more likely a take off photo.
Cheers, Sal
The model is 1/6 scale with a wingspan around 195 inches. The tail and wing incidence are the same as far as I can figure. I am using a number of three views
and they all seem to show the same configuration. Yes, the stab is huge.. The model comes out to around 6 feet.
The caption for that photo states that this is the way the Ilya Muromets flies. It is more likely a take off photo.
Cheers, Sal
07-01-2015, 06:29 AM
#73
Thread Starter
Join Date: Jan 2005
Location: Holbrook, New York NY
Posts: 119
Likes: 0
Received 0 Likes
on
0 Posts
Hello All
Due to the lack of interest in the build thread, I will end the thread and just continue here in "Aerodynamics" for those who are interested.
Cheers, Sal
Due to the lack of interest in the build thread, I will end the thread and just continue here in "Aerodynamics" for those who are interested.
Cheers, Sal
07-01-2015, 08:52 AM
#74
Hi Sal, as a last try you could ask a moderator to move your build thread to the RC Scale Aircraft forum since all WWI prospects are there. Where it is now seem to be only WWII aircraft discussed.
As to the picture, this IS the way it flies, as I said earlier. I just suspected that they made a fast fly-by without load for a good (spectacular) picture (tail up and low AoA exaggerated). But even with its comparatively big weight the IM model flies quite nose-down in cruise (at least in the simulator).
All three-views show the fuselage (and the cabin) level, with the tailskid a bit off the ground. In this position the propellers have about 4 degrees up-thrust and I measure about 8 degrees wing and stab incidence (or even more). That means the wings are near stall AoA-wise and I think that is take-off (and landing) position - with the tailskid off ground. 4 degrees pitch-down make the prop disks vertical and make for a good AoA and efficient flight.
Keep in mind that the airplane is not pitch stable. There's a small speed "window" of apparent neutral trim, but if you leave that window the instability will make things worse in no time. Too slow makes for pitch up (rears up like a horse), too fast for tucking (like a dropped bomb). Must and can be counteracted with elevator, of course (pitch-hold gyro should bear the workload).
As any airplane, it has to be trimmed for a certain speed, for instance down elevator for faster speed. Actually, one might think it would help if the stab is blown by the prop. More propwash would give more stab lift. Even this unstable airplane needs some down elevator at higher speed, but not much. I think that's why the stab sits so high and the propwash goes below it. (Propwash is deflected down by the undercambered wing.)
As to the picture, this IS the way it flies, as I said earlier. I just suspected that they made a fast fly-by without load for a good (spectacular) picture (tail up and low AoA exaggerated). But even with its comparatively big weight the IM model flies quite nose-down in cruise (at least in the simulator).
All three-views show the fuselage (and the cabin) level, with the tailskid a bit off the ground. In this position the propellers have about 4 degrees up-thrust and I measure about 8 degrees wing and stab incidence (or even more). That means the wings are near stall AoA-wise and I think that is take-off (and landing) position - with the tailskid off ground. 4 degrees pitch-down make the prop disks vertical and make for a good AoA and efficient flight.
Keep in mind that the airplane is not pitch stable. There's a small speed "window" of apparent neutral trim, but if you leave that window the instability will make things worse in no time. Too slow makes for pitch up (rears up like a horse), too fast for tucking (like a dropped bomb). Must and can be counteracted with elevator, of course (pitch-hold gyro should bear the workload).
As any airplane, it has to be trimmed for a certain speed, for instance down elevator for faster speed. Actually, one might think it would help if the stab is blown by the prop. More propwash would give more stab lift. Even this unstable airplane needs some down elevator at higher speed, but not much. I think that's why the stab sits so high and the propwash goes below it. (Propwash is deflected down by the undercambered wing.)
Last edited by UStik; 07-01-2015 at 09:05 AM.
07-02-2015, 12:34 PM
#75
Good evening Sal,
hope your amazing IM build is going well. Just want to show how the model might look in flight. The simulator model looks crude but shows the main characteristics, like attitude in level flight.
The flight characteristics are crude as well, especially the propwash effects can not be rendered correctly. But the model could be set up especially vicious, just to train your reflexes. So the simulator is an option in addition or as an alternative to a smaller R/C model.
Best wishes, Burkhard
hope your amazing IM build is going well. Just want to show how the model might look in flight. The simulator model looks crude but shows the main characteristics, like attitude in level flight.
The flight characteristics are crude as well, especially the propwash effects can not be rendered correctly. But the model could be set up especially vicious, just to train your reflexes. So the simulator is an option in addition or as an alternative to a smaller R/C model.
Best wishes, Burkhard