This page describes design evolution from the Mark 2A design.
The changes to the design from 2A to 2B are limited to the propulsion arrangements (and that I made a front compartment cover).
As discussed on the Mk2A page, the Mk1 didn't examine propulsion, and therefore Mk2A was built with the propulsion system mounted on a sub-frame that would lift out of the hull. The idea was that if my propulsion arrangements didn't work, I could build an alternative sub-frame and slot it in. Mk2A got a very rough-and-ready mount for some GWS ducted fans.
Mark 2B replaced the GWS fans with Wemotec mini fans and Graupner speed 480BB motors. They are built into nice pods, and the pods themselves are on a very solid frame that fixes securely into the hull by a combination of physical interlocking and two M5 nylon wing-bolts. Even if the bolts detach, the thrust from the fans (when pushing forwards) keep the sub-frame in place - to remove it, it needs to be pivoted up and back (against the fan thrust) and then slid forward.
The Mk2A page has cad drawings of the hull, and
these are some miscellaneous bits and pieces I drew out in devising the
propulsion for the Mk2B. This is at the scale of 1 pixels=1mm.
What's innovative is that this hovercraft has brakes, and can reverse with almost as much power as it runs forward. On the back of each propulsion pod is a thrust reverser, so the thrust from the fans can be directed backwards or forwards.
To control this arrangement, I built and programmed an on-board microcontroller that mixes the desired speed and turn channels from the receiver to control two speed controllers (for the fans) and two servos (for the thrust reversers). It's actually even cleverer than that - it also monitors the lift motor and factors that into the calculations, and finally it receives input from an on-board gyro, adjusting steering depending upon signals from the gyro. There's enough complexity in it that the mixer has got a page of its own where I describe my design goals and concept of the mixer itself and the propulsion and steering system more generally, since that is inextricably linked with the design of the microcontroller system. If you're interested (and speak PIC assembler), you can download my source code too.
However, a word of caution. I don't think I'd actually advocate building this. I did it to see if it works, and while the answer is that yes, it does work, but probably not well enough to justify the complexity. Mark 2C will probably be 2B but without the reversers. With brushed motors, the easiest solution to controlling this layout craft is a simple delta / V-tail mixer and using marine speed controllers which will give reverse.
The thrust pods
are mainly built up from thick soft balsa with lots of
sanding. Since I'm not very good at accurate sanding, there's also quite a lot of
filler, where I over-sanded, so built up with filler again for a second
attempt!
(Note: the photo here shows the pods before the thrust-reversers
were fitted.)
The fans are mounted to a frame of spruce and beech (beech
verticals, spruce horizontals) with some ply diaphragms in the vicinity of the
mounting points. It's very solid, expecially as I set a balsa web between
the main verticals turning them into a 30mm deep I-beam with beech top and
bottom flanges and a balsa web. This is probably overkill, but the mk2A
suffered from an incident of running at speed under an overhang that clipped
the top of the fans and ripped them from the mounting. That won't happen now.
The thrust reversers are themselves a second version. The first version used quarter-circle gates, as shown in the following pictures. It's very efficient when fully reversed, but very non-proportional, and very difficult to arrange servo linkages to (since everything crosses over something there isn't a clear path for a linkage - I'd need to put the control horn on the hinge pin above or below the whole unit).
The second version is much like a backwards rudder. It's easy to arrange the linkage, and much more proportional. It's slightly less efficient driving forwards than the first version and quite a lot less in reverse - I lose about 30% of the thrust in reverse, I think. However, I am yet to finish the internal face of the pods, and I have an idea about improving reverse efficiency.
Centred, with equal thrust forwards and backwards:
full forwards and full reverse:
I'm still getting used to it, and changing the code in the on-board fly-by-wire (particularly the gyro contribution). It behaves lots different to the mark 2A, and to other hovercraft I've seen. Turning is very sharp, and it tends to pivot about the centre of the craft rather than sweep into a turn. At the moment the gyro authority is very high- if you put some turn on then release the stick, it stops pivoting immediately, absolutely rock solid, as if it ran into a wall. Very odd, for a hovercraft. I don't think a hovercraft with gyro-controlled rudders would do this, because if you released the stick, there'd be no thrust and relatively little authority from the rudder, regardless of what the gyro tried to do.
I need to:
back to the top hovercraft page
back to Ian's contents page
back to www.astounding.org.uk top contents page
To comment on anything (please do) email ian@astounding.org.uk