Successful flight trial completion of unmanned aerial vehicle, MAGMA (2017) | Hacker News Reader

Successful flight trial completion of unmanned aerial vehicle, MAGMA (2017)

baesystems.com

44 pointsjdmark25 comments

Needs a (2017) on the title.

Last video seems to be from 2019: https://www.youtube.com/watch?v=-WP40eL-xS4

Last news on it looks to be from 2021: https://www.thedefensepost.com/2021/09/08/bae-next-gen-aircr...

Kind of a cool concept. The lack of updates makes sense given the title of that second link... The next time we hear about it will be if it crashes over some war zone.

Year added above. Thanks!

Long time HN user, but new account to break real life / online links.

I'm one of the designers of the aircraft, if anyone has questions about flow control, integration, "what's the point", I can try and answer them for you.

Interesting project! Three questions:

1) Doesn't this idea mean that if the engine quits the aircraft is uncontrollable? I presume it's (just) dynamically stable, but that would very much be a 'bad thing'

2) While I can understand that replacing mechanical control surfaces with aerodynamic ones would be a significant weight saving, how much is it really a drag saving? The inviscid layer is tiny at high Reynolds numbers and aren't you going to require the rest of the fluid drag terms in order to effect a movement?

3) What is the maximum amount of control authority you have, and how do you decouple it from engine thrust (c.f. accelerated stalls and all that)?

1) Yes - if the engine stopped on a fully fluidic aircraft you'd lose the ability to run the controls. The applications we typically considered for the technology are in addition to normal flappy control surfaces. E.g. you use these for a load alleviation function on civil airliner, or take off with normal control surfaces and switch to fluidics for when you want a lower RCS.

2) Fluidic controls can be (on paper at least) lower drag than the conventional alternative, to be honest that side of it isn't my area, but there are plenty of papers on the subject.

3) These actuators were designed for a 10deg/s roll and pitch rates. For the roll device the slot length is around 100mm long, longer slots / higher momentum coefficients would yield a faster roll rate. There's a "more scientific" video on youtube https://www.youtube.com/watch?v=bI6ouhXjjk4

And you are comfortable actually answering those questions in public? Audacious, to say the least...

Yep, there's nothing in discussion here that isn't openly published academic research. Granted its a niche area of research, but that doesn't automatically mean it is sensitive.

Just wanted to be sure, I tend to be rather paranoid in that regard! :-)

The maneuvering thruster system sounds similar to that of the Harrier VTOL. Which makes sense, because BAE Systems made the later Harriers. The Harrier used that because it had to have some control at very low speeds, where the control surfaces are ineffective.

That system was very difficult for humans to fly. Pure-thrust vehicles are unstable. Computers have far less trouble, as we see with quadrotor drones.

Interesting, I wonder if this is more like a reaction control system with discrete jets, or if it's more focused on boundary layer control (or some combination thereof). By strategically causing flow separation on wing elements at specific times, a little bit of air could have a huge effect on the aerodynamic forces on the vehicle. The bare metal on the leading and trailing edge surfaces of the wing make me thing the latter is included in the design.

Edit: yes, they're doing boundary layer control, just not in the way I thought. This article has a diagram: https://www.baesystems.com/en/article/magma-the-future-of-fl...

> The maneuvering thruster system sounds similar to that of the Harrier VTOL.

It doesn't really, to me. The big idea here is using blown air to avoid having physical surfaces deflect-- neither the engine nozzle or any actual control surfaces on the plane. The Harrier uses fairly conventional bending nozzles to vector trust, and uses fairly conventional control surfaces at speed.

> That system was very difficult for humans to fly.

Anything helicopter-like is difficult to fly, and the Harrier is a pretty mediocre "helicopter" with relatively poor response times. In principle blown-air deflection of thrust could be more responsive for VTOL-type aircraft, but that's not even what this is aiming at.

(The deflection attained of thrust is not sufficient for VTOL use at this point-- nor is deflecting thrust in one degree of freedom from one point sufficient for hovering flight.)

> Pure-thrust vehicles are unstable.

Blowing air to adjust airflow at the trailing edge doesn't necessarily produce anything too hard to control-- it's pretty similar to conventional control schemes. On the other hand, this aircraft will not be controllable at all without the active systems gating the airflow, while aircraft with control surfaces can often fallback completely to dumb control cables.

> The Harrier uses fairly conventional bending nozzles to vector trust, and uses fairly conventional control surfaces at speed.

On the Harrier, those are the big rotating lift nozzles. There are also small attitude control nozzles ("puffers") at the nose, tail, and wingtips. The puffers are used in hover and at very low speed.

This new drone is pumping out supersonic air through control jets and bending the jets with a fluidic amplifier. That's new. May or may not be useful.

[1] https://i.imgur.com/78bVjbC.png

These aren’t maneuvering thrusters where thrust moves the plane. These are changing the airflow on wing to act like control surfaces. Pilot should fly them the same way as now, with computers making them work and act like normal surfaces. This isn’t replacing aileron with thruster but replacing with inverse aileron.

The next generation might use airflow control surfaces.

Right. It's a fluidic amplifier. Steering a fluid jet with a smaller fluid jet is known, but not used much.

It's useful for reducing radar cross section. Moving control surfaces are a headache for RCS management. The F-22 uses "cat eye" slits between its control surfaces to minimize these radar blooms, but it'd better to do away with them altogether. It could also remove the need for those huge aerodynamic fairings on the CS actuators on jetliners, but then add some complexity and remove fuel volume in the wings.

I am unsure of how the trailing edge tech works, are they blowing the air out of trailing edge for roll control, and is it blowing aft or directionally? Is the blown air itself the force moving the wing or is it just redirecting flow over the wing? I am thinking of diminishing control as speeds lower, which is already a characteristic of regular control surfaces.

This is super interesting, I would love to build it in an RC plane.

It acts like an aileron. If you blow air up from the right wingtip trailing edge, the right wing will go down.

What is key here is that it affects the airflow and the effect is bigger than just the momentum of the blown air itself.

The key concept is Coanda effect: https://en.wikipedia.org/wiki/Coand%C4%83_effect

There's a diagram of how this is implemented in the aircraft: https://www.baesystems.com/en/article/magma-the-future-of-fl...

For roll, air is blown out of the TE aftward. Per Coanda and the lift equation (lift is proportional to square of flow speed), this blowing will generate higher lift on one wing, imparting a roll moment on the aircraft. For pitch, air is blown up at the engine's exhaust surface to point the flow up/down. It is shaped so that with 0 blowing, it's pointing down.

The air is blown out aft, but the jet is blown adjacent to a curved surface and tends to stick to it. This entrains the flow over the wing and creates a change in lift distribution just like a normal aileron.

The thrust vectoring is similar but slightly different, again the blown air is used to change an existing airflow.

Not really, thrust vectoring is separate from what they're doing here. Thrust vectoring allows you to literally use thrust from the engine to push the plane in different directions, this is a bit more subtle. Basically you can use little puffs of air to control the flow of air over the wings and other lifting surfaces and this can be used to increase or decrease lift in certain places. The actual force produced by these little air jets it totally negligible but by changing the way air travels over the lifting surfaces it can change the lift generated, allowing you to get rid of moving mechanical control surfaces. This requires very careful engineering and a very detailed computer modeling to get right which is why it's only really happening now.

similar to the DARPA active flow control demonstrator that just received is designation (X-65):

https://www.darpa.mil/program/control-of-revolutionary-aircr...

I'm missing the basic premise here - it uses small jets of air to vector thrust and whatnot right ? instead of flaps? Because fewer moving parts?

This seems like it would involve a massive number of small fans or jets, which sound like more moving parts than flaps would be.

What gives?

As I understand from the article it bleeds air from the exhaust for this purpose. So I assume no more than a few valves, and a computer.

There will presumably be a slight drop in thrust, but I expect that's too small, and too short-lived to be material.

interesting concept - would have thought the radar return from the twin , almost vertical, tail plane was more significant than that of any other control surfaces they may have been able to eliminate.

To be fair maybe they can eliminate that part of the structure later on.

Reply on news.ycombinator.com