Flying without a vertical stabilizer

[badpilot27]

I think mixing drag rudders into the ailerons should fix this in the final version.

I think you should try using different airfoils and twist the wing so that you get the prandtl bell shaped span load and place the elevons out ~70% (basically out board of the wing "tip" vortex) to get the induced thrust at the wing tip. I know you made a post earlier about not doing this because of the aspect ratio but the horten XIIIa had a similar aspect ratio and it worked. The smaller elevons will reduce pitch and roll authority though.

Also I'm not sure about this but I think the anhedral makes it more stable because when the plane yaws one wing will have a higher angle of attack then the other so it crates more lift and induced thrust which stabilizes the plane.

 

[Piotrsko]

Makes sense. The chuck gliders are by necessity mostly not full scale, though I have built several ones at the size of the rc plane I plan to build. The CFD is also run at the scale and speeds, pressures, and air densities I plan to fly, so it should have aerodynamic similarity so far as Reynolds number goes. The scale definitely still does make a difference, but if you input the right data in the software, you should get results that apply to your scale.

Sorry for the mis pointing. "Full scale" was intended to be a size capable of people carrying, many iterations up in size from model scale, but I see you have considered that also. Didn't think anything you're doing was going to hugely alter the Reynolds number, could be wrong there, particularly if the airfoil goes transitional at your speeds.

Was wondering if you still have the bulbous nose and canopy.

 

[telnar1236]

I think you should try using different airfoils and twist the wing so that you get the prandtl bell shaped span load and place the elevons out ~70% (basically out board of the wing "tip" vortex) to get the induced thrust at the wing tip. I know you made a post earlier about not doing this because of the aspect ratio but the horten XIIIa had a similar aspect ratio and it worked. The smaller elevons will reduce pitch and roll authority though.

Also I'm not sure about this but I think the anhedral makes it more stable because when the plane yaws one wing will have a higher angle of attack then the other so it crates more lift and induced thrust which stabilizes the plane.

Would be curious what airfoils you would recommend. I'm currently using thin symmetrical airfoils out of laziness more than anything since they're very easy to make out of a sheet of foam, though I don't want to make the wing too thick since it wouldn't really look like any of the NGAD concepts anymore.


As I've mentioned before, I'm coming as close to a bell shaped span load as I can with this low aspect ratio of a wing. Both the foam board model and the 3D printed versions have had significant twist in the wing, ranging between 3 degrees on the early ones and 11 degrees on all the more recent versions.

The Horten XIIIa is a more heavily swept design than I was aware of existing, but it's still enormously higher aspect ratio than what I'm building. It's worth keeping in mind, the entire fuselage on my designs is part of the wing as well. Almost all my planes are at least half again as long as they are wide. My elevons are also as small as they can reasonably be already without losing too much pitch authority or requiring so much deflection that they cause instability just by being in the trim state. The cranked arrow design requires a forward CG, so a fair of bit of elevon deflection just to keep it in the air. My current thinking is to keep the elevons purely for pitch control and to use spoilers for roll control, but that will only work if the plane is positively loaded.

In terms of the anhedral, I don't think that's it. The stability benefits show up regardless of how much wash out the wing has in my CFD and foam board models and if the wing was somehow producing induced thrust I would think it would be destabilizing to have that increase with lift on the advancing wing. In general, with swept wings, the advancing wing produces more induced drag due to its increase in lift which is what gives them directional stability.

 

[telnar1236]

Sorry for the mis pointing. "Full scale" was intended to be a size capable of people carrying, many iterations up in size from model scale, but I see you have considered that also. Didn't think anything you're doing was going to hugely alter the Reynolds number, could be wrong there, particularly if the airfoil goes transitional at your speeds.

Was wondering if you still have the bulbous nose and canopy.

Ah, ok, that makes more sense, yeah, very much not trying to understand, design or build anything full scale. I imagine NGAD is meant to be able to fly transonic and supersonic and I haven't even considered what that would look like for mine among many other things. I just think the concept art looks cool and want to build and RC plane that looks like it that also flies acceptably well.
I wasn't really thinking of the Reynolds number being an issue with scaling, at least at RC scales, as with scaling down. A lot of my foam board chuck gliders are pretty tiny and glide pretty slowly putting them firmly in a low Reynolds number regime where the air behaves more viscous than for my larger powered models. I think the increase in the ratio of viscous to induced drag from that probably at least partially accounts for their increased stability, though to be honest that's a bit of a shot in the dark. I'm still pretty confused as to why my little profile chuck gliders are so much more stable than their larger cousins.
The bulbous nose has progressively gotten smaller over time, and flatter, but it's still there. Pretty much all the concept art has it, so I want to include it if at all possible. This design is, after all, pretty much a case of trying to dramatically over engineer something to look the way I want it to

 

[badpilot27]

Would be curious what airfoils you would recommend. I'm currently using thin symmetrical airfoils out of laziness more than anything since they're very easy to make out of a sheet of foam, though I don't want to make the wing too thick since it wouldn't really look like any of the NGAD concepts anymore.

I don't really have any airfoil recommendations since I still have to wait at least year or 2 to have access to the math required to make a good airfoil that is available in high school and that's only if I even get accepted into TJHSST, otherwise ill probably have to wait until collage. what I would do though is probably copy or use very similar airfoils to that used on nasa's prandtl-d wing and have the center line of the fuselage have an airfoil similar to the wing root of nasa's prandtl-d wing.

Also I know this was only a paper plane, but the horten XIIIB proposal had an even lower aspect ratio and probably the same span load.

 

[Piotrsko]

I'm still pretty confused as to why my little profile chuck gliders are so much more stable than their larger cousins.

You're demonstrating what is understood to be "scale effect", the bane of designers, software and wind tunnels: works great at one random size but not at others for no apparent reason. However that reason eventually arrives as the art progresses. I recall when bumble bees weren't supposed to fly because they violated the known laws of physics. Key point: the word known.

What happens on the larger ones if you install a flow disrupter on the furthest aft part of the wing trailing edge to induce just a wee bit of drag? Like a 1/4 sq inch piece of foamboard glued to the top side of the airfoil or even glued perpendicular to the trailing edge.

 

[telnar1236]

You're demonstrating what is understood to be "scale effect", the bane of designers, software and wind tunnels: works great at one random size but not at others for no apparent reason. However that reason eventually arrives as the art progresses. I recall when bumble bees weren't supposed to fly because they violated the known laws of physics. Key point: the word known.

What happens on the larger ones if you install a flow disrupter on the furthest aft part of the wing trailing edge to induce just a wee bit of drag? Like a 1/4 sq inch piece of foamboard glued to the top side of the airfoil or even glued perpendicular to the trailing edge.

I'll need to try that on my newest version. It has spoilers that are about 1 in^2 on each wing and are aligned with the trailing edge, so after trying it in its base configuration, I should just be able to program in my transmitter to have them partially open and see what happens.

 

[telnar1236]

I don't really have any airfoil recommendations since I still have to wait at least year or 2 to have access to the math required to make a good airfoil that is available in high school and that's only if I even get accepted into TJHSST, otherwise ill probably have to wait until collage. what I would do though is probably copy or use very similar airfoils to that used on nasa's prandtl-d wing and have the center line of the fuselage have an airfoil similar to the wing root of nasa's prandtl-d wing.

Also I know this was only a paper plane, but the horten XIIIB proposal had an even lower aspect ratio and probably the same span load.

Looking at the Prandtl-D airfoils, my tip airfoil is already very close to what it uses on the 3D printed versions, and my root airfoil is essentially a thinner version of what it uses. Using a thicker airfoil more analogous to what the Prandtl-D uses did help in some early XFLR-5 runs, but I haven't looked at it further since the airfoil is so thick it completely buries the rest of the fuselage.
For the Horten XIIIb, that is much more similar to my design in terms of aspect ratio. But, crucially, at that aspect ratio, it looks like they turned the fuselage and cabin into a vertical stabilizer.
As someone who has gone through the years of school and pursued a career in aerospace, it's definitely worth sticking it out through all the math, though I don't think I got to actually calculating aerodynamic forces by hand until my third year in college. And at a practical level, that is something we basically never do. At a hobby level, you definitely don't need the math just to play around and see what works. A lot of practical hobby-level design is rule of thumb for pretty much any design.

 

[L Edge]

I'll need to try that on my newest version. It has spoilers that are about 1 in^2 on each wing and are aligned with the trailing edge, so after trying it in its base configuration, I should just be able to program in my transmitter to have them partially open and see what happens.

How are you going to program that in the transmitter?
Are you using 2 servos or 4 for the spoilers?
Are you going to use a FC?

The few RC planes that actually are stable using that system impressed me with the added fact is they are partially deployed on takeoff and after the gear is upped, they fold back in. Then it started to work when they made a turn.
Finally, realized that it is used to provide yaw at lower V1 and V2 speeds if there is crosswinds or gust.

Boy, a FC would sure make this an easy approach to solve it. But like you, I want a method so I can fly the plane.

 

[telnar1236]

How are you going to program that in the transmitter?
Are you using 2 servos or 4 for the spoilers?
Are you going to use a FC?

The few RC planes that actually are stable using that system impressed me with the added fact is they are partially deployed on takeoff and after the gear is upped, they fold back in. Then it started to work when they made a turn.
Finally, realized that it is used to provide yaw at lower V1 and V2 speeds if there is crosswinds or gust.

Boy, a FC would sure make this an easy approach to solve it. But like you, I want a method so I can fly the plane.

The current plan is to have a total of 4 servos for the control surfaces on the plane, one for each side of the elevons which will deflect only for pitch control, and one for each spoiler which will deflect only for roll control. The elevator servos will just be on a Y lead. Each spoiler servo will be on its own channel, and I plan to program the transmitter so that the spoiler only deflects up, but has no downwards deflection. Basically, moving like ailerons, but it only deflects when the stick is pushed in the direction it would deflect upwards.
Unfortunately, this will only work when the plane has a positive angle of attack, so no inverted flying for me, but the current plane is also only stable at positive angles of attack, so I'm not really losing anything.

Currently, I'm trying to get around needing a flight controller for my smaller foam board versions, since, like you, I want to fly the plane. I would also like to be able to possibly release plans, that don't require a ton of computer/FC setup if it flies well enough, since most people will be more interested in a simple foam board plane than in a complex 3D printed build.

For the larger 70mm EDF I plan to eventually build, I expect I will use a flight controller though. A large part of the reason I'm trying to find a configuration that is aerodynamically stable at a wide range of attitudes is to try and simplify the final programming of the flight controller. I very much would like to be able to use just a gyro and accelerometer and not need to find a way to read my actual angle of attack and sideslip angles. If I just need to overcome adverse yaw, but not inherent instability, that makes my life so much simpler, since the gyro can then just smooth the flight like in a normal plane, and the accelerometer is only needed to figure out if the plane is positively or negatively loaded and therefore which direction the controls should deflect to counteract adverse yaw and perform a given maneuver.

 

[Piotrsko]

Why wouldn't your system work inverted? Seems like you'll be deflecting the surfaces as much as you would be deflecting them for stability while upright. Never had a plane that flew inverted without some "nose up", even my mostly zero zero symmetrical combat wings. Would it be twitchy? Probably, but not anything a decent flight controller shouldn't handle if the controller can fly inverted

 

[L Edge]

Did the CG of the chuck glider match the calculated CG of your model? Is it anywhere near where the spline and wing meet?

 

[telnar1236]

The big issue with inverted flight would be adverse yaw. The goal with using spoilers instead of ailerons is to have a proverse yaw effect so that the roll axis is controllable. When upright, or more specifically when the plane has a positive angle of attack, regardless of its orientation relative to the ground, deflecting the spoiler will reduce lift and increase drag on the same side, so the plane will both yaw and roll into the turn, avoiding adverse yaw. In contrast, when the angle of attack is negative, the spoiler will function more like a split flap. The wing gaining lift will also gain drag and then adverse yaw is back with a vengeance and it's anyone's guess which way the plane will go.

The wing itself also relies on a fairly complex twist to give it stability with such a low aspect ratio. When the fuselage is at 0 AOA, the wing root has 3 deg positive angle of attack and the wing tip 8 deg negative angle of attack. This means the wing is always producing stronger aerodynamic forces than the fuselage when the plane is at positive angles of attack, but the situation is reversed at negative angles of attack and the twist becomes destabilizing. I'm trying to find ways to reduce the reliance on this wing twist for stability, but I'm not sure how possible it is.

 

[telnar1236]

Did the CG of the chuck glider match the calculated CG of your model? Is it anywhere near where the spline and wing meet?

The CG needed to be way ahead of where most CG calculators would put it. Pretty much right where the wing meets the spline, yes. My more complex Simflow model was pretty much spot on with the neutral point, but XFLR5 also got it very wrong and recommended a CG that was way too far aft. I think that vortex lift on the wing near the root results in the COP shifting forward as the plane pitches up. Between the chuck gliders and the powered model, the CG was identical.

 

[L Edge]

The CG needed to be way ahead of where most CG calculators would put it. Pretty much right where the wing meets the spline, yes. My more complex Simflow model was pretty much spot on with the neutral point, but XFLR5 also got it very wrong and recommended a CG that was way too far aft. I think that vortex lift on the wing near the root results in the COP shifting forward as the plane pitches up. Between the chuck gliders and the powered model, the CG was identical.

??

 

[telnar1236]

??

For compound deltas, like my cranked arrow design, the outboard section of the wing increases in lift slower than the inboard section and stalls sooner. In general, for delta wings, there are two mechanisms of lift generation. First, there is the normal linear lift, same as any wing generates. Then there is vortex lift which is generated by the powerful leading edge vortex. Vortex lift is why delta wings are so good at high angles of attack. In general, you need a leading edge sweep greater than 60 degrees to get the full benefits of vortex lift. So, for my design, since the inboard section has a very swept leading edge, it generates vortex lift very effectively, but since the outboard section only has a sweep of about 45 degrees, or even lower in some iterations, it mostly relies on the vortex from the inboard section to avoid stalling. So, the inboard section of the wing, which is also further forwards on the plane, increases in lift faster making the plane want to pitch up. To counter that, the CG needs to be a fair bit more forwards. This is really only going to be an issue with compound delta wings like my design, and unless it's a cranked arrow wing, it shouldn't really apply.

This simulation shows the vortex pretty clearly. It's very powerful over the inboard section of the wing which is more forwards. However, on the further aft outboard section of the wing, the air pretty much flows straight over the wing except for at the very tip where you get another vortex, this time contributing to induced drag.

As opposed to on a plane like the F-104 where the air pretty much flows smoothly over the wing except for the wingtip vortex (in this picture you can see how wingtip accessories like dummy tanks or missiles help reduce the strength of the wingtip vortex and actually improve aerodynamic efficiency - an effect present in both RC models and the real F-104).

As far as progress on building the next version, I haven't really had the couple hours to just sit down with a hot glue gun and some foam to finish putting together my next testbed. It's about two thirds finished, so I might be able to get it done this week, but things have been pretty busy in real life on my end. Probably the most notable progress I've made was finding this 360-degree render of a 6th gen fighter from Collins.

Collins 6th Gen Fighter Aircraft

After modeling it in CFD, it looks like this design is more stable than anything else I've looked at so far and is even approaching the stability of a more conventional plane. It looks like it probably comes from a different set of geometries to the set I've been honing in on, and I'm not convinced by some aspects of its looks, but it's so much more stable I might end up going with it.

 

[Scotto]

Just for fun I tried yawing with the nose. A better pilot would have had no problem flying knife edge with it. I tried cutting off half of the vert stabs but immediately crashed. On a full scale fighter maybe drag rudders would work better up front? It couldnt have turbulant air going in the engines though. Im clearly not an aerospace engineer.
In some of those concept videos Im pretty sure I saw a retractable vertical stabilizer. If the point of not having one is radar I dont see why it would hurt to have one that you just use for take off and landing. And maybe extreme maneuvers when everything sees you anyway.
One other thing NASA was just bragging on a new design software. I dont know anything about it, but it might help you. Its on github. https://github.com/OpenMDAO/Aviary You look like you are doing great with what you have, though. Keep it up.

 

[Piotrsko]

Generally want drag inducing stuff behind the CG. In front it tends to do odd reactions.

@telnar1236 thanks for the data on variable sweep, didn't know that.

 

[L Edge]

@telnar1236 :

Did you see BAE's GCAP plane of future which still has rudders? Inlets are still below cockpit and work their way up? Larger wing which is good.
Had fun putting prop and EDF planes back together. Just enjoyed bunches of flights.
Finalized second approach, so now back on construction of udder rudder scheme.

 

[telnar1236]

@telnar1236 :

Did you see BAE's GCAP plane of future which still has rudders? Inlets are still below cockpit and work their way up? Larger wing which is good.
Had fun putting prop and EDF planes back together. Just enjoyed bunches of flights.
Finalized second approach, so now back on construction of udder rudder scheme.

I've seen the design, to be honest, I'm not a huge fan of how it looks - I think some of the previous concept art for the Tempest was prettier, but both have a lot in common with the F-35. Definitely looks like it would fly well as an RC plane though. Looking forward to seeing how your udder rudder design flies, looks like it could be promising