Quiet Bird EDF

[Hogwerks]

My apologies folks! Life has been hectic lately and I forgot to circle back to the forums.
Quiet Bird did briefly (and dramatically) taste flight. The field I like to fly at has a nice hard mat runway so we decided to try a dolly takeoff. Unfortunately the inlets do indeed (not) suck so it got just enough speed to hang out in the stall regime and that's when the sharply tapered wing did its nasty work and generated tip stalls. She ended up flopping over on her back hard 4 times. Thankfully the parts are real sturdy so most of the damage is benign.
As dap35 pointed out yes indeed these inlets were used on the real design for lower RCS. After playing around with this, I suspect partly why it did not proceed further (at least as far as I can tell) is because the inlets were to high risk.
So as life settles back down for me I am circling back to figure out a plan of action going forward.

The headaches I have had with the design overall are as follows:
1) The inlets are pretty bad. Without getting into specifics, they violate most of the inlet rules I have ever known in full scale aircraft inlet design BUT again, its an early 1960s stealth design. Kind of expected this!
2) When translating the design from 3 views and the few pictures online of the halfscale RCS model into CAD some areas just don't close out properly in particular the bottom of the aft fuselage under the exhaust where the V-tails meet the fuselage and at the top of the fuselage where the inlet lip meets the "ramp" and blends out into the fuselage. In aircraft lofting speak we call this horrid intersection between inlet lip, inlet duct, and fuselage skin the "triple point" since 3 surfaces meet at the same point. It is usually hellashious to get right in the right CAD packages (NX & CATIA) and just plane awful in things like Solidworks and usually impossible in things like Onshape/Fusion (if you try to meet full scale standards).
3) The blending of the "ramp" from inlet duct to fuselage. The real airplane was short and stubby and this inlet is begging for more length to smooth things out.
4) The wing tapers too aggressively for docile stall characteristics.

Currently I am thinking I will relax the 'scale' look. This is something I have struggled with. How close should one be to the subject vs just make something that works better as an RC??? I could take the hypothetical posture of "okay, let's assume it was to go forward, what would have the next design iteration looked like if they tried to fly it in the 60s?" The answer to that....
1) Still sorting out how I would make the inlets better but in general it will involve some proportion changes
2) Pull the top inlet triple point farther away from the top of the fuselage to make the surface model more stable and detail design easier
3) Change the wing planform to something simpler; just 1 LE sweep angle = "1 panel" wing planform
4) Re-work the aft fuselage; get rid of the flat termination as depicted on the historical design and allow it to have a belly at the back; this facilitates structure for the V tails and gives room for surfaces to terminate more smoothly. See Kratos XQ-58's back end for reference.
5) As it is currently lofted (as it was originally) the canopy is seamlessly blended into the fuselage. Looks sharp, but is a real pain for surface model stability and down-stream detail design work. A middle of the road CAD package like SolidWorks does not have the brains to handle highly blended curvature continuous parametric shapes like this...frustrating...So I will break them up and make the canopy a separate more distinct surface feature (as is usually done).
6) Shrink the airplane down tightly around the 50mm EDF, make the airplane longer, and make it wider. Shrinking it will help lower the weight and drive a higher T/W even with poor inlet performance. More length will give me distance to smooth the inlet shaping for better airflow and a wider fuselage will again give more room for better inlet shaping.
7) Not obvious in the pictures but I did a 1 layer skin print with plenty of stiffeners to stabilize the skin. I'm going to go with "DirtyDee"'s suggestion on another forum to use a 2 layer skin and just a few stiffeners. Seemed to work better for weight with his designs so hopefully that will help me knock the weight down further aiding the T/W.
So in the end it might smell more like a cross between the historical Quiet Bird and Scaled Composite's Model 401 "Sierra" (which is really what this airplane SHOULD look like).

I hope all of you chew on this as you look at the B-21 Raider

 

[dap35]

I think there are a number of features that don't scale down in a linear fashion - especially EDF inlets. Most comercial EDFs tend to have cheaters. The best exception to this is the Avanti.

I would look at your current model and do some static testing with cheater holes or other modes to see if you can increase the static thrust. Because you loose so much on a smaller EDF, I would aviod downsizing to a 50mm.

Ahh, CATIA.... It has been close to 30 years since I worked with it. Such a revolution.

 

[L Edge]

Did you take any videos of those 4 flop overs?
Already exploring NGAD stuff dealing with EDF's. Lot's of problems.

 

[Hogwerks]

Did you take any videos of those 4 flop overs?
Already exploring NGAD stuff dealing with EDF's. Lot's of problems.

I did not. Don't have a camera and can't hold a phone while (attempting to) flying.

In regards to the NGAD configs I'll summarize it this way = Tails are very useful.

If you get rid of the tail, you still need something to generate yaw forces. Keep in mind stability and control (S&C) has a lot to do with "time to double". How much time it takes for the given moment (pitch roll yaw) to reach double amplitude. There is significant mis-understanding in amateur aero circles (sim and RC/drone especially) when it comes to making unstable aircraft fly.
People throw around static margin without understanding that static margin is a static point reference and all configs behave in non-linear dynamic fashion.

Now back to the public NGAD configs. First, ask yourself, is this the real config or just art? So far only art has been officially released, so the config you see could be 100% unflyable because it is just art like you would see in a movie. 2nd, so far all released ones are low aspect ratio high sweep tailless configs (some with canards but another topic). There is a boogy man for those planforms called the "pitch up boundary". It is a harsh boundary usually plotted as quarter chord sweep vs aspect ratio. Get close to the curve and you better keep that AoA LOW and steady...get beyond that curve, and you are toast. Just saying "Fly by wire" wont fix it. The time to doubles get so fast no flight controller will stop that bird from swapping ends (without a tail)....notice you don't see any video of things like P-175 "Polecat", B-2, XB-35, nEUROn, Taranis, X-47A, X-47B, or X-45 do any aggressive movement. Note X-47A only flew once and it was a very scary flight. Hence X-47B's planform change (to tame the pitch stability).
And then there are all the fun roll-yaw-pitch coupling characteristics....someone give me a desperate "nooooooo!!!!!" gif.

So if one wants to pursue those configs watch the sweep angles and aspect ratios and make sure you have something some where that can be used as a yaw moment generator (control effector). B-2 uses split drag rudders. X-47B used combinations of different control surfaces and effectors in a blended manner to provide adequate yaw control. Lots of videos and pics out there depicting its controls.
So far the European "NGAD"s have tails!

 

[L Edge]

Like to show you a video of my X-47B I designed. It can be done. There is no gyro or FC being used. Concept is based on 2 servos used as elevons and a TV nozzle offset so when ailerons are moved, the TVN applies a yaw for the turns. It took a bunch of runs to get the correct length of arm to provide the right value.

Did this 8 years ago. Stop the video at the beginning and you can see the setup.

Since this is my prize possession, fly it only a few times a year. This one was a month ago.

NGAD is the future so until the RC groups' understand what the problems are and solve it, no progress. I believe the 3D printing method will help solve a number of problem areas.

 

[Hogwerks]

Like to show you a video of my X-47B I designed. It can be done. There is no gyro or FC being used. Concept is based on 2 servos used as elevons and a TV nozzle offset so when ailerons are moved, the TVN applies a yaw for the turns. It took a bunch of runs to get the correct length of arm to provide the right value.

Did this 8 years ago. Stop the video at the beginning and you can see the setup.


NGAD is the future so until the RC groups' understand what the problems are and solve it, no progress. I believe the 3D printing method will help solve a number of problem areas.

That is an excellent example of just because there are challenges does not mean it cannot be done. There are just creative compromises! Looks like it flies nice!
The full scale X-47B as an outboard spoiler so through drag from the spoiler and deflection of the ailerons you get yaw control. Works, just means some other compromises.
Utilizing the EDF nozzle for yaw control works, but it now means you have coupled thrust to yaw control. Sometimes for some folks (like in the RC hobby case) that is just fine...for say the real X-47B on approach to land on a $13B aircraft carrier with 4,500 people on board, one might not be so keen on having that 45,000lb jet moving at 150mph being 100% dependent on that engine working 100% all the time...what if the engine has an issue???? That's a possible full scale design constraint...not always relevant to RC! The joys of scale modeling! Which is why I am enjoying transitioning from designing full scale airplanes to RC. More fun and less to worry about!