Halbach array fans for aeropropulsion

[DeltaV]

I'm not a materials specialist but what little In k now suggests the ideal material for these is liquid metal. It can be injection molded for fantastic uniformity between blades and the process is cheap enough that it's being used for things like skiis and snowboards. It's not a high temperature application, but none of these blades, for air or water; are going to get very hot.

http://www.liquidmetal.com/applications/

Sounds similar to this:
http://www.sciencedaily.com/releases/20 ... 172129.htm

Another approach:
http://www.modumetal.com/technology.php

[DeltaV]

It's not a high temperature application, but none of these blades, for air or water; are going to get very hot.

Unless you fire a REB down the central axis...

[GIThruster]

Sounds similar to this:
http://www.sciencedaily.com/releases/20 ... 172129.htm

Same stuff. The Liquid metal people all came from Cal Tech. It's just the last couple years they've commercialized the process to the point it's showing up in pop culture like skis. IIRC, there was a spot on NASA TV showing off the Head skiis using it. Because the metals are flash frozen into place, it has no metallic lattice, it does not conduct electricity and it does not suffer metal fatigue over time. Skis with a thin layer of liquid metal should never lose their bounce. They'll act like new skis for a lifetime.

Cool stuff. Stronger and lighter than composites. Looks like stuff to build your future spacecraft out of.

[rjaypeters]

Why is a hubless design desirable, especially for marine applications? Mostly anti-fouling. The original application was for bow thrusters which are used at low vehicle speeds in port where there is a lot of debris in the water. These hubless designs are more resistant to catching and keeping debris in the blades and a lot easier to clean.

Lack of hub also gains some efficiency because the working fluid which would normally flow around a hub doesn't have to flow around, but may pass straight through. The technical term is "entrainment", mostly caused by friction. The best example is the Dyson "bladeless" fan which uses entrainment more than their advertising literature suggestions (I don't know what "inducement" really means to them, but they are apparently drawing a destinction between air which flows inside the hoop and air flow outside the hoop).

The other fun part about the Dyson fan is they call it bladeless, but it uses a fan in the base! I just love advertising!

DeltaV suggested cascading hubless thrusters of different diameters on the same duct to use at different air speeds. An interesting idea, efficient from a packaging perspective; the problem is the duct losses for any given active thruster would be increased by the the drag of thruster blades which are not actively being spun.

[DeltaV]

DeltaV suggested cascading hubless thrusters of different diameters on the same duct to use at different air speeds. An interesting idea, efficient from a packaging perspective; the problem is the duct losses for any given active thruster would be increased by the the drag of thruster blades which are not actively being spun.

That's not exactly what I wrote, rjay. Here's what I wrote:

For the air turbine, with several stages inline, imagine each stage's rpm controlled independently to optimize overall thrust/efficiency at any flight condition.

Says nothing about any of the fans being stopped, or different diameters.

A gradual variation in duct diameter over the fan section might be appropriate, but CFD sims and experiments would have to determine that. This is not the compressor/combustor/turbine arrangement of a conventional jet engine, where large variations in diameter along the axis are necessary due to large changes in flow characteristics. The axial change would be much more gradual here since the Brayton cycle is not involved. Variable intakes and exhaust nozzles for the non-fan portions of the duct might be appropriate, depending on the operational envelope and vehicle design.

I'm envisioning several electric, counter-rotating stages per duct, closely spaced, each stage's rpm controlled independently. Non-rotating segments with variable stator vanes (common for conventional jet engine compressors) might be interspersed where appropriate, depending on CFD results. Keeping the stages closely spaced limits the turbulence seen by downstream stages and maximizes counter-rotation efficiency benefits.

Varying the blade angles for the spinning segments is possible, but I wouldn't add that complexity unless there was a good reason to do so. Now, if the rotors had variable blade angles, you could stop one and use it as a variable stator segment. Lot's of interesting possibilties when you go all-electric.

Conventional jet engines have two or three spools, giving two or three simultaneous rpms for the various sections. Generally, tailoring the rpm of each bladed segment to the local flow improves efficiency and performance. Pratt and Whitney's latest is the Geared TurboFan, which separates fan speed from spool speed through a gearbox. Each stage of an electric ducted fan can have its own rpm, optimized for the conditions at that stage's position along the axis. Said fan conditions varying much less radically than for a Brayton cycle engine's compressor/turbine sections, of course.

Reverse thrust via electrically reversing the rotor spin directions is also a possibility. Say your vehicle has several, parallel, multistage, electric ducted fans for forward flight. For VTOL, you might be able to use some of those for rear lift and some (running in reverse) for forward lift, with variable-geometry devices to direct the flows. Admittedly, a non-trivial integration challenge, but not that different from what F-35B already does.

This would free up space otherwise taken by dedicated lift fans/doors and maybe reduce lift system weight:

(Edit: fixed dead link, again)

vs.

[DeltaV]

Low speed, high thrust for large diameters ("lift fan"). High speed, lower thrust for smaller diameters, several cascaded along a duct ("turbine").

Rjay, I see now how this might have been confusing. I was referring primarily to rotor rotational speeds, not vehicle speeds, though they are related. Several cascaded stages for the longer duct "turbine". Only one (or maybe two, counter-rotating) for the shorter duct "lift fan".

[icarus]

In a nutshell:

The hubless idea is to allow the vorticity sheet from the blade 'lift' to roll-up and shed as strong, discrete tip-vortices. Manipulating the location of trailing vortices can minimise the 'induced drag' component contribution of these streamwise vortices.

There is more to it from the global flow viewpoint, (things like having a jet-like rather than wake-like core-flow is advantageous for avoiding vortex-breakdown) but it will probably bore most of you.

[rjaypeters]

DeltaV: Sorry, my mistake. Because of the principle I am about to write, I assumed the diameters would be different and fan stages would stop when not being used:

In general, one likes to disturb airflow the least. If one stage of blades will get the job done, then that's all one uses.

I note from the pictures you show (fictional aircraft and F-35), it looks like there is one stage of blades for each fan in the fictional aircraft and I can't tell how many stages the F-35 lift uses (but they are counterrotating).

If one wants to achieve a high compression ratio, then a hubbed design is probably necessary because highly compressed air will try to back flow through any open area between the blade tips. With greater rotation speeds, variable geometry blades are an increasingly attractive invitation to disaster, IMO.

You are right. It's down to the details, including firing a REB down the core of the engine. How hot does the REB make the surrounding air? It must be hot because we want the air to expand, right? And how does one get the REB to fire down the center of blades unless the REB "outlet" is positioned and aligned down the center of the blades? A curved duct, I suppose.

[GIThruster]

The answer above that a hubless core avoids fouling (rjay) makes perfect sense.

[icarus]

???

[GIThruster]

[]

[icarus]

???

[GIThruster]

[]

[icarus]

???

[DeltaV]

Rick, it's getting ugly. Can we please have some data?