telnar1236

Elite member
Turns out I'm just bad at designing wing apex fences. I reread the NASA paper and realized I'd missed a couple things about the geometry. The redesigned wing fence creates a 16% boost in lift which is closer to what I was looking for (the equivalent of cutting 360 grams out of the weight in terms of landing speed). It also reduces the pitching moment by 36% which will reduce the needed deflection of the elevons and result in an even greater increase in lift. Paired with weight reductions and the increased drag from the wing fence I think this design will make the plane much easier to handle. I'll still include the current wing design with the final STLs for people with 500' runways since it will bleed less speed in turns and handles very nicely once in the air, but for most people I think the wing with the retractable fence will work better.
1732903825963.png

1732903859891.png

You can see how much better developed the vortex is with this new design and how the flow stays attached for most of the length of the wing as opposed to just closer to leading edge as on the previous designs.
 

JetCrafts

Active member
what software do u use
Turns out I'm just bad at designing wing apex fences. I reread the NASA paper and realized I'd missed a couple things about the geometry. The redesigned wing fence creates a 16% boost in lift which is closer to what I was looking for (the equivalent of cutting 360 grams out of the weight in terms of landing speed). It also reduces the pitching moment by 36% which will reduce the needed deflection of the elevons and result in an even greater increase in lift. Paired with weight reductions and the increased drag from the wing fence I think this design will make the plane much easier to handle. I'll still include the current wing design with the final STLs for people with 500' runways since it will bleed less speed in turns and handles very nicely once in the air, but for most people I think the wing with the retractable fence will work better.
View attachment 246900
View attachment 246901
You can see how much better developed the vortex is with this new design and how the flow stays attached for most of the length of the wing as opposed to just closer to leading edge as on the previous designs.
what software do u use
 

JetCrafts

Active member
I don't think I'm going to manage to reduce the landing speed as much as I want just based on reducing the weight, so I've been looking into ways to improve the aerodynamics. Delta wings aren't really compatible with conventional flaps since they would just make the plane pitch down so I need another option.

One possibility is to put a set of split flaps in the middle of the wing where they won't cause much pitch change.
View attachment 246883
The issue with this kind of flap design is that they don't actually do that much. Wings fly primarily due to the low pressure on top of them and the low pressure on top of the flaps is shared with the bottom surface of the wing. Therefore, they are mostly air brakes, only boosting lift by about 5% and slightly making the plane want to nose up decreasing the loss of lift due to trim forces on the elevons. They may be worth adding still since every little bit helps, though, especially if combined with reduced weight and other high lift devices.
View attachment 246884

I did some reading and also stumbled on a second possibility. Wing apex fences look very strange, but like flaps they increase the lift at a given angle of attack.

View attachment 246877
View attachment 246878

Since the wing has enough lift to fly slower, just while at a higher angle of attack than the plane lands at, the apex fence could be a good option and easier to design and more effective than the flaps. It works by generating and capturing a vortex similar to what delta wing generates at higher angles of attack which increases the lift at the cost of drag. Here's a paper from NASA describing how they work: https://ntrs.nasa.gov/api/citations/19940019630/downloads/19940019630.pdf

Here is some CFD showing it a little more clearly. The wing with no fence doesn't generate much of a leading-edge vortex at all at 12.5 degrees AOA.
View attachment 246879
Adding in a small partial fence creates a vortex over the inboard part of the wing and boosts lift by about 5%.
View attachment 246880
And adding in a fence over most of the span generates a much larger and stronger vortex and boosts lift by about 10%.
View attachment 246881
The NASA paper shows that the lift increase should be able to be much bigger (about 20%) at least at the Reynolds numbers they tested, so I'm still trying to refine my designs. I'm trying to keep them relatively low profile to reduce how much of the wing the mechanism to deploy and retract them would take, but I think I should be able to improve performance by increasing the height and angle of the fence closer to 5 degrees or possibly by splitting the fence into a couple smaller ones. Tilting it back by 10 degrees should also help a tiny bit. In the last CFD of the full fence, the outer section of the fence doesn't actually generate much of a vortex, probably due to its smaller angle. I was trying to replicate a gothic wing fence with a smaller outboard angle (amazing name on the part of whoever designed it) that should get better performance, but that may not work the same on this particular plane. It's also possible that wing apex fences just work much better on full scale planes or I did a bad job designing it.

View attachment 246882
(From the NASA paper I found)
u can try dog teeth aka leading edge extensions https://en.wikipedia.org/wiki/Leading-edge_extension
 

telnar1236

Elite member
what software do u use

what software do u use
For the analysis shown above, the free version of Simflow to run the analysis and Paraview to generate the plots. It's a bit of a learning curve to use, but for wings that generate things like vortices it's the only way to get particularly accurate results.
u can try dog teeth aka leading edge extensions https://en.wikipedia.org/wiki/Leading-edge_extension
Those mostly improve the high angle of attack characteristics of a wing but don't do much to boost lift at lower angles of attack. As a delta wing, this wing can already reach 40 degrees AOA and get plenty of lift to slow down, but the issue is that the plane cannot land at those angles of attack because the tail would hit before the landing gear. So what I'm looking for is a way to shift the lift curve to the left so that it generates more lift at lower angles of attack. Basically flaps on a delta wing. That's what the wing apex fence does very effectively. This is an incomplete plot from an earlier iteration of the wing fence that shows what I'm trying to do.
1733074877574.png
 

telnar1236

Elite member
What, you mean you can't do this?



Yeah, if you do my a+b approach, you could come close. Want to hear it? Guaranteed less than 150 ft for landing. And that includes acrobatics including the "Cobra" if you can practice enough.
The issue isn't so much that the plane can't perform that maneuver as the height of the landing gear and how brittle 3D printed structures are. I can't land at more than 14.5 degrees AOA and I can't come it at a higher AOA and then drop the nose and let it "plop" down since it would break
1733078782604.png
 

L Edge

Master member
The issue isn't so much that the plane can't perform that maneuver as the height of the landing gear and how brittle 3D printed structures are. I can't land at more than 14.5 degrees AOA and I can't come it at a higher AOA and then drop the nose and let it "plop" down since it would break
View attachment 246950

A side note where NYGAD planes have no rudders and 2EDF's. In my research, found that the drag brakes that are used for yaw stability (deflected up to 25-30 degrees) can be set up as speed brakes for landing and this is accomplished by using a V-Tail mode using 2 rudder inputs.
Now it all falls into place. So, it would work on a single engine EDF controlling pitch and yaw. Tahh-dadh!!!!!!!!

Would even reduce your rollout.

Be fun to try a parachute upon landing.
 

telnar1236

Elite member
Even with all the changes to the aerodynamics, and changes to the wings, I think the only way to make this into the plane I want is going to be to make it bigger. I kept refining the wing design and wing apex fence to increase lift and ended up with a final design with 37% more lift - but it still would have had a 44 mph stall speed. Better than the 52 mph of the v1, but not ideal still. The higher aspect ratio longer wing would have boosted lift and the fence would have boosted it further and reduced the trim needed to keep it at its landing angle of attack, but I wasn't too thrilled with the looks. And it still wouldn't have had a reasonable landing speed. I still have the v1 airframe, so I might try printing the new wings to test it at some point, but it won't be the final version.
1734496452961.png

1734496495630.png

So I'm making the plane 29% bigger which will increase the lift by 66%. I'm also adding in some of the things I learned from my v1 design that would have required a redesign of most of the plane but that could have significantly improved lift. Here are the significant changes.
  1. Increase in size of 29% (53" to 68"). The weight will also increase from 2500g to about 3200g, but that should be acceptable.
  2. Increase in fan size from 70mm fan to 80mm to power larger airframe. I'm now using an FMS fan with 3260 g of thrust which should give overall improved performance.
  3. Redesigned wing so that only the wingtip is cambered. The v1 design had a continuous increase in camber from the center line of the airframe to the tips which reduced drag at higher AOAs but also reduced the overall angle of attack of the wing and killed lift at a given AOA. This meant that when the nose was 12.5 degrees up, the wing was only seeing about 11 degrees AOA. With just the wingtip cambered, it still gets many of the benefits and looks about right, but should get more lift.
  4. The wing has a built in 1.5 degree incidence angle. This increases lift at the landing angle of attack and reduces drag at higher speed since the fuselage is flying straighter. This was one of the biggest lessons from the v1 and something that would have required a major redesign of most of the fuselage.
  5. The main gear is moved aft to the scale position. In the v1 it was forward of the scale position to allow for the cheater inlet, but in this larger design, it can be where it should be. This has two benefits. First, it allows a more aft CG which should make the airframe more responsive and reduce the required trim which will improve lift. It also allows landing at a higher angle of attack.
  6. Redesigned and improved inlets. With the bigger airframe I won't need the cheater inlets anymore which should improve thrust at speed and overall efficiency. They are a bit larger than scale, but I'm still happy with how they look.
  7. Larger hatch that makes the battery easier to place. In the v1 it was always a struggle to fit a 6s pack in a space designed for a 4s pack originally.
  8. Nose gear uses more conventional arm which does not require servo to move to allow retraction. The previous system was meant to be mechanically simpler and rely on programming to move the servo, and it worked fine, but it was a pain to program and calibrate the system, wasn't reliable, and put too much strain on the servo.
  9. Many smaller changes - it's essentially a new design so the structure and internals will all be somewhat different.
It's definitely a work in progress again and a big step backwards, but this should be a big improvement. The landing speed will now be a full 15 mph slower than the v1 at 37 mph which is very much acceptable for a jet this size. The top speed will also be higher at almost 130 mph.
1734497027704.png

Honestly, this is the scale I should have built it at in the first place. When I was planning this build, it started off as a new home for my 80mm fan before I decided to change it to a smaller 4s design. Then when the issues with the fan led me to increase to 6s the weight ballooned back up. This goes back to an adequate size for a 6s fan.
 
Last edited:

JetCrafts

Active member
thx for the simflow suggestion been searching for something
For the analysis shown above, the free version of Simflow to run the analysis and Paraview to generate the plots. It's a bit of a learning curve to use, but for wings that generate things like vortices it's the only way to get particularly accurate results.

Those mostly improve the high angle of attack characteristics of a wing but don't do much to boost lift at lower angles of attack. As a delta wing, this wing can already reach 40 degrees AOA and get plenty of lift to slow down, but the issue is that the plane cannot land at those angles of attack because the tail would hit before the landing gear. So what I'm looking for is a way to shift the lift curve to the left so that it generates more lift at lower angles of attack. Basically flaps on a delta wing. That's what the wing apex fence does very effectively. This is an incomplete plot from an earlier iteration of the wing fence that shows what I'm trying to do.
View attachment 246949
m
 

JetCrafts

Active member
For the analysis shown above, the free version of Simflow to run the analysis and Paraview to generate the plots. It's a bit of a learning curve to use, but for wings that generate things like vortices it's the only way to get particularly accurate results.

Those mostly improve the high angle of attack characteristics of a wing but don't do much to boost lift at lower angles of attack. As a delta wing, this wing can already reach 40 degrees AOA and get plenty of lift to slow down, but the issue is that the plane cannot land at those angles of attack because the tail would hit before the landing gear. So what I'm looking for is a way to shift the lift curve to the left so that it generates more lift at lower angles of attack. Basically flaps on a delta wing. That's what the wing apex fence does very effectively. This is an incomplete plot from an earlier iteration of the wing fence that shows what I'm trying to do.
View attachment 246949
this is a nice thing to think of a solution to during history class
thx
 

telnar1236

Elite member
what abt canrds hmmmmmmmmmmm
Honestly those would probably help a lot. Convair did actually propose variants of the F-106 with canards, so it's not that out there either.
1734569315175.png
1734569290425.png

I'm just not a fan of the way they look even if they would help a lot aerodynamically. I thought about retractable canards, similar in concept to the glove vanes on an F-14, but by the time they were small enough to fully retract, they wouldn't help that much with lift either. It ultimately comes down to the v1 being the right size for a 4s plane and not a 6s plane which means it's too heavy for its wing without major aerodynamic modifications which would mean it wasn't an F-106A anymore.
 

telnar1236

Elite member
thx for the simflow suggestion been searching for something

m
No problem. It's a bit of a learning curve, but definitely worth it. The risk of CFD is garbage in, garbage out, so if an answer doesn't seem to make sense or the mesh doesn't represent the real geometry, it's probably a safe bet that the answer it gives will be wrong.
 

JetCrafts

Active member
Honestly those would probably help a lot. Convair did actually propose variants of the F-106 with canards, so it's not that out there either.
View attachment 247147 View attachment 247146
I'm just not a fan of the way they look even if they would help a lot aerodynamically. I thought about retractable canards, similar in concept to the glove vanes on an F-14, but by the time they were small enough to fully retract, they wouldn't help that much with lift either. It ultimately comes down to the v1 being the right size for a 4s plane and not a 6s plane which means it's too heavy for its wing without major aerodynamic modifications which would mean it wasn't an F-106A anymore.
tru thats why the hmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
 

telnar1236

Elite member
Based on some of @quorneng comments in another thread, I was somewhat curious how the flow would behave in the EDF ducting. He suggested trying to build the plane big enough of to avoid needing cheater inlets, so one of the reasons for scaling up the design is that it means the inlet area is sufficient to supply the fan without cheater inlets. At 143% FSA, the inlets (and overall plane) are larger than would by typical for an 80mm fan. I tried to model fairly smooth ducts, but also didn't spend a ton of time optimizing to keep constant curvature or similar and allowed the area to vary slightly, primarily with the goal of simplifying the printing process. This is the duct design.
1735015898543.png

I don't really have a good baseline to compare it to or much experience with inlet ducting from a more formal engineering perspective but thought it would be interesting to run CFD to get an idea of what the flow would look like anyway. In the interest of not making my computer explode and having a life doing stuff other than running simulations, it's a symmetry model of only half the duct, only continues to the front of the EDF unit, does not make any attempt to simulate rotational flow from the EDF, and as a whole is a pretty bad representation of the actual physics (and has a bad mesh too). However, it does give a baseline of what's going on.
1735016213718.png

Probably the most interesting plot is the total pressure plot since it gives the best idea of what the losses are in the ducting. Overall, I don't think they're that awful and it looks like they're pretty continuous instead of occurring at any one location which is a good indication that they're primarily due to the length of the duct and not due to bad turbulence somewhere. There is room for improvement on the outside lip of the inlet, though where it looks like flow separation occurs. All the pressures are kinematic pressure in m^2/s^2 if anyone is wondering.
1735016329833.png

The only other thing of much interest is just how abruptly the pressure drops going into the inlets. Anyone who has flown EDF jets much knows it has to be happening like this from hands on experience, but it's still neat to see.
1735016876008.png
 

L Edge

Master member
Wow, decided that the F-106 is a bird that you pretty well defined and is meant for a good pilot. If you go to "you tube" and review the EDF's and turbine F-106 bird's, the takeoffs and landings require a good feel for the bird and really "eat up the runway".

Found one video that had a short t/o and landing that was relaxing. Picked this one since he had a 80 edf and he mentioned that he recommends a rudder for easier landings. The guy that flys it is a top heli pilot as well as excellent pilot, so you can see.


Have you ever thought of changing the delta angle to give you more area?
 

telnar1236

Elite member
Wow, decided that the F-106 is a bird that you pretty well defined and is meant for a good pilot. If you go to "you tube" and review the EDF's and turbine F-106 bird's, the takeoffs and landings require a good feel for the bird and really "eat up the runway".

Found one video that had a short t/o and landing that was relaxing. Picked this one since he had a 80 edf and he mentioned that he recommends a rudder for easier landings. The guy that flys it is a top heli pilot as well as excellent pilot, so you can see.


Have you ever thought of changing the delta angle to give you more area?
He's definitely a good pilot and that's a beautiful model. Can't tell if he was just trying to fly it gentle and scale, but it looked a bit underpower. The Dan Savage plans seem to be a pretty common starting place. I used them as a reference when building my 70mm version and to confirm the CG. I think the wing on the F-106 just isn't very good at producing lift at lower angles of attack which makes the landings a pain.

I played around with the idea of reducing the sweep/making the wings longer, but any appreciable improvements would have really hurt the looks. Even combining the reduced sweep and a wing apex fence, I couldn't get the landing speed down to a reasonable value on paper without a total redesign.