Let the conspiracy speculations begin!

[RainmanTime]

Often times, because of my work and the fact I sometimes work on classified projects, people accuse me of "working on reverse engineering UFOs". While I am flattered by this, I cannot (of course) confirm or deny anything about such accusations.

But when stuff I have worked on is clearly in the "white world" I am willing to share it. In this case, a project I worked on doing the design for the engine and throttle controls is undergoing US Navy testing at NAS Patuxent River....and some people seeing this UAV with its outboard wing segments folded and on a trailer think it is a UFO.....how quaint!

Military drone mistaken for ‘UFO’ along DC highways - DC Breaking Local News Weather Sports FOX 5 WTTG

RMT

 

[Einstein]

RMT

I'll have to agree with you on this one. Just knowing how dangerous UFO technology could be in the public's hands. I would have to say it would be in the best interest of all if everyone were to believe it was just a military drone.

 

[Gpa]

That's a great story Rainman. I'm curious though, since you worked on it, how does it handle Yaw with no tail.

 

[RainmanTime]

That's a great story Rainman. I'm curious though, since you worked on it, how does it handle Yaw with no tail.

Highly advanced control laws used on this airplane, and also used on the B-2. One thing you learn in aerodynamics is that any one control surface does not produce "pure" pitch, roll, or yaw moments. While an aileron mostly gives roll moment, it also gives a bit of yaw moment. You can convince yourself of this by looking at a single aileron deflected. It not only creates a lift force which, far out on the wing creates a roll moment, but it also creates a drag force. With no drag on the other side (the aileron is not deflected) you get a net yawing moment with the nose of the airplane moving in the direction of the deflected aileron. So the trick in the control laws that control roll and yaw is to know how to mix all of the trailing edge control surfaces to point your airplane's velocity vector exactly where you want, and generate just the right amount of pitch, roll, and yaw moments to complete the maneuver demanded by the pilot (or autopilot, which is the case on the X-47B unmanned combat aircraft).

That's why guys like me make "the big bucks".
(That's a joke in case anyone took it seriously. While I live a comfortable middle class existence, I bet there are people who believe in TT who earn more than me doing things like repackaging and selling bad debt!)
RMT

 

[Gpa]

Thanks for the reply Rainman,
I'm trying to learn a bit about aviation lately in my spare time (what little there is). Is this the same concept as fly-by-wire technology? I believe the F-16, F-117, F-18, and F-22 use a similar system (though they still have tails, I think)? There are probably more now. I have loved fighter jets since watching F-4 Phantoms taking off from the USS Enterprise. It was quite a sight back then and this is probably why I have always loved super roller coasters. It's the closest I will ever get to simulating being in one. I have been watching the History channel's Carrier Series lately too.

 

[RainmanTime]

Hi Gpa,

Is this the same concept as fly-by-wire technology? I believe the F-16, F-117, F-18, and F-22 use a similar system (though they still have tails, I think)?

Sorta kinda. The two are related. Let me explain. Fly-By-Wire (FBW) technology is what makes control surface mixing a lot "easier" to do, and much more powerful. But even in pre-FBW designs the designers could use mechanical and pre-computer electrical mixing of the control commands. Basically, FBW technology is were we completely replace a mechanical cable-and-pully connection from the pilot's control stick (or yoke) to the flight control actuators that move the elevators, ailersons, and rudders. We replace those push-pull cables with electrical wires that either carry analog signals (e.g. from the pilot's stick to the Flight Control Computer(FCC) is often an analog proportional signal coming from an LVDT sensor on the stick), or digital signal (e.g. from the FCC to the actuator control electronics via a high speed digital data bus like MIL-STD-1553 or more modern and faster busses).

By doing this replacement, now we can unlock the "magic" of software inside the FCC to drastically reshape how the vehicle performs through the techniques of augmented stability. We can develop the algorithms in FCC software that close control loops much more quickly than a pilot could. This is my specialty in the area of detailed design, both inner loop stabilization and outer loop autopilot guidance algorithms. This is also why aerospace engineers take, on average, 3 more math classes than all other engineers, because the vector calculus and LaPlace Transform and Fourier Series concepts are needed to develop robust algorithms, as trial-and-error (something Einstein seems to think will always work) just does not work in this case.

RMT

 

[Einstein]

RMT

This is also why aerospace engineers take, on average, 3 more math classes than all other engineers, because the vector calculus and LaPlace Transform and Fourier Series concepts are needed to develop robust algorithms, as trial-and-error (something Einstein seems to think will always work) just does not work in this case.

So what if the algorithm doesn't work? Oops! Back to the drawing board. Good thing it's only a drone. Of course then another algorithm would need to be developed. That kind of looks like trial and error to me. But I am curious about the extra math courses you took. I would like to look over the text material. Maybe you could refer me to something along the lines of what you had to learn.

Here's another link on that UFO-drone with some real good pics of the drone in flight and on the flatbed truck.

 

[RainmanTime]

Einstein,

So what if the algorithm doesn't work? Oops! Back to the drawing board. Good thing it's only a drone. Of course then another algorithm would need to be developed. That kind of looks like trial and error to me.

Well, not surprising it would look that way to you given you have no experience in developing highly complex systems based on principles of physics. But what you do (which is trial and error) is so far from developing and testing a complex engineering product so as to be primitive. It is not called trial and error when we can derive the 6DOF equations of motion for the vehicle we intend to build, build a scale aerodynamic model of it, take it into a wind tunnel to quantify its aerodynamic coefficients, and validate how it flys before we ever build a flying version of it. That is not trial and error, it is engineering.

But I am curious about the extra math courses you took. I would like to look over the text material. Maybe you could refer me to something along the lines of what you had to learn.

The two extra math classes that Aero majors take that other majors do not are listed here in the Cal Poly Math Department curriculum as MAT 317 (LaPlace Transform & Fourier Series) and MAT 318 (Mathematical Analysis of Engineering Problems - A better description is really Calculus of Vector Valued Functions)

Math & Stat Undergraduate Courses Listing | Cal Poly Pomona

And then there is the ARO core class which is really yet another advanced math class in disguise ARO 406 (Advanced Dynamics & Vibrations of Aerospace Systems) which you can read about in the curriculum here:

http://www.csupomona.edu/~aro/current/documents/AROcatalog.pdf

As for books for the two math courses, there was a single book that covered both of those classes:

Advanced Engineering Mathematics by C. Ray Wylie & Louis C. Barrett.

The textbook for ARO 406 was a custom textbook developed by my instructor, and I still use it to this day.

RMT

 

[Darby]

So what if the algorithm doesn't work? Oops! Back to the drawing board. Good thing it's only a drone.

Good thing it's only contrarianism for the sake of contrarianism. Otherwise we'd be stuck in the middle of the 16th Century with your "science".

Ray didn't say it this time but we've both said it before: military aircraft are, by design, unstable. Stable aircraft do a great job of flying straight and level. Military aircraft that fly straight and level are called KIA's. But instability requires computers especially in tailless aircraft. Human pilots cannot match the power curve of instability when the plane is maneuvering to avoid "bad stuff". Tails on aircraft are also radar magnets. Survivability requires low radar cross-sections, high speed maneuverability and computer controlled flight surfaces to maintain stability - unless we want to go back to WWII where "winning" the war cost 10,000 aircraft losses on each side and 200,000+ air crew casualties. We cannot afford to toss up 10,000 $25+ million per unit aircraft as triple-A fodder.

By the middle of WWII it was apparent to everyone that aircraft abilities were outstripping pilot's abilities to control them. Several propeller driven fighter aircraft types (Bf-109, FW-190, Ta-162, P-51D, P-47D, P-38L) were able to enter the trans-sonic and on occassion super-sonic zone in high power dives during that war. Many pilots died not because they were shot down but because they did not have the physical strength or the instant brain power to recover their aircraft once it entered the trans-sonic zone and the control surfaces froze (actually stalled). That would have required computers that were able to stay ahead of the aircraft and make corrections before the situation became unrecoverable and loss of the aircraft and crew was inevitable. Even in the early to mid jet age pilots were not able to push their aircraft to their actual limits because the control surfaces were not computer controlled.

But you don't care. It's just contrarianism for the sake of contrarianism.

 

[Einstein]

RMT

Well, not surprising it would look that way to you given you have no experience in developing highly complex systems based on principles of physics. But what you do (which is trial and error) is so far from developing and testing a complex engineering product so as to be primitive. It is not called trial and error when we can derive the 6DOF equations of motion for the vehicle we intend to build, build a scale aerodynamic model of it, take it into a wind tunnel to quantify its
aerodynamic coefficients, and validate how it flys before we ever build a flying version of it. That is not trial and error, it is engineering.

I do have lots of self taught on hands experience developing lots of electronic gizmos. A scale aerodynamic model is a form of trial and error. I use the same principle in building electronic gadgets. I'll construct an experimental electrical circuit on a breadboard just to see if there are glitches or performance problems or even if the circuit will work. If the circuit works as intended, then I'll spend the extra time to make a printed circuit board, allowing me to make the device in a more permanent form.

Thanks for your input on the math books. Although I don't see anything in the curriculum that I'm not already familiar with. I'm just hunting for something new. Apparently I've been taught a version of calculus that no one else uses. In the past couple of weeks I've gone over three other calculus books looking for clues. Nothing found yet. Although that person you mentioned, Miles Mathis did mention a similarity between the calculus he was taught and what I was taught. So apparently I'm not the only one with this weird knowledge. But I don't think he knows it yet. The only thing is, I haven't found a textbook that reflects this knowledge. But then my calculus instructor didn't really require us to read our book. The class was mostly all lecture. We had to do the problems out of the book. But those were done with what we learned in class. So I'm going to keep hunting. Just got a translated copy of Newton's Principia. So maybe I'll find some clues in there.

 

[Einstein]

Darby

Good thing it's only contrarianism for the sake of contrarianism. Otherwise we'd be stuck in the middle of the 16th Century with your "science".

I'll bet you think I choose a contrary point of view just to pick on you. Maybe on occasion that might be true. But the truth is I value your input, even if I don't get the kind of response I was looking for. Some of our discussions are quite thought provoking. So keep up those thought provoking responses of yours. I do like the informative writing style you use. Always a pleasure to read.

 

[RainmanTime]

A scale aerodynamic model is a form of trial and error. I use the same principle in building electronic gadgets. I'll construct an experimental electrical circuit on a breadboard just to see if there are glitches or performance problems or even if the circuit will work. If the circuit works as intended, then I'll spend the extra time to make a printed circuit board, allowing me to make the device in a more permanent form.

Nope. Not the same thing. With a wind tunnel model we are not "trying to see if it works". In fact, we can't because there is no engine installed. Thrust being one of the primary forces acting on an airplane, we could only do trial and error with a wind tunnel model if it had an engine. No, with a wind tunnel model we are not doing trial and error at all, because wind tunnel models are too expensive to take such a primitive approach. What we are doing is, in fact, comparing actual performance against the performance predicted by equations and by CFD analysis. This data then goes into other mathematical models, such as structural and control models, from which we develop new design features. While you may wish to believe it is trial and error, it is not because we actually have a name for it: guided empiricism. This approach starts out with physical theories that have been validated, whereas trial and error starts out with nothing but guesses. In fact, there are certain aerodynamic derivatives that quantify an aircraft's aerodynamic performance that cannot even be verified in a wind tunnel. Primarily, what we call the dynamic derivatives. Good thing theory informs us how to predict them.

RMT