Thread Tools
Aug 06, 2015, 02:47 PM
Registered User
Thread OP
Build Log

Skirtless hovercraft / Bixel hybrid

I've been playing around with two separate models recently and I've had an idea on how to potentially combine them into something (hopefully) pretty sweet.

First up is the Bixel ground effect vehicle. I ended up getting far more use of my build as a plane rather than a strict GEV. In fact with a large enough elevator it could do what I can only describe as almost 3D type flying, hanging off the props quite handily, rolling about while held nearly vertical, able to parachute down at walking pace, plus fly around alright in normal forward flight, and all of this on a planform with no dihedral and rudder/elevator only. That model eventually had one too many crashes and was beat up to the point of not really being worth trying to repair so it's been scrapped. The main problem I suffered with it was that all of my flying spots are covered in thick grass that endlessly stuck it down and made it very difficult to actually fly it in GE.

Bixel WIG/Ground effect vehicle (2 min 46 sec)

^That's without the oversize elevator, so it's limited in pitch authority.

I've also played around with a skirtless peripheral jet hovercraft as based on Windnseas thread here. I built a quick model and was quite pleased with how it turned out. Since that post I made a few refinements and have gotten far better lift height and better stability. In the video below I'm using my lightest battery with no propulsion so that represents a best case scenario from the weight point of view.

Peripheral Jet hovercraft model (0 min 14 sec)

I did put an EDF and rudder on that model but I managed at the same time to break the lift propellers and had to put on smaller ones to get it working so that plus the extra weight brought the lift height down considerably. Even with all that stacked against it the model worked fine on thick boggy grass, managing about 30mm of lift.

The largest challenge with the skirtless design once lift is sorted is it's pretty much purpose built for flipping over backwards as soon as it gets any amount of forward speed. One solution is to use a T-tail rear stabiliser as seen here.

If I've kept your attention this far you can probably see where this is going. The Bixel is exceptionally stable and simply will not flip backwards with the CG in the right place, but it relies on skidding along the ground and is prone to damage and getting stuck. The hovercraft on the otherhand can effortlessly float about over just about anything but will flip very readily.

Combining these models, I have what I'd class as a flying hovercraft. It's a rectangular skirtless hovercraft with outboard bixel wings at the rear. Planned specifications are:

50x100cm hovercraft with 25cm wide outboard wings, lift is a pair of counter-rotating 8x4 props. Height of the hovercraft block is 6cm and it will have an aerodynamic leading edge. Propulsion is a prop (likely 7x5) mounted on a pylon on the top surface of the hovercraft. Control is rudder and elevon with the elevons sitting off the back of the hovercraft to extend the actual bixel planform out.

CG for the hovercraft really needs to be geometrically in the middle, with the extended elevons this means the CG for the bixel is slightly ahead of the midpoint, this is slightly further back than I would like however I know the bixel will fly fine with the CG in that spot.

My main concern from the build point of view is attaching those outboard wings. As you can see from the internals of the hovercraft, there isn't any real solid structure in there, the sidewall is attached using spacers. I've been trying to think of any clever ways of putting in a spar across the outboard wings and the rear top surface of the hovercraft but it's difficult between the rear lift fan, the rudder and the need to keep a clean profile (ie I could just glue a balsa beam across the top but it will create a very wide step). For now I will reinforce the rear portions of the sidewalls so that they should be strong enough to cope with the leverage of the attached outboard wings.

From a piloting point of view I already have a decent idea of how this should work although it's complicated in the way that there's several distinct regimes of 'flight':

*Hovercraft mode: On the ground, lift fans providing all lift, it should be able to slide about quite happily using low throttle and rudder. I expect the high mounted prop to create a lot of nose-down with throttle tendancy, this can be dealt with at least partially using a throttle-elevator mix (worked well on the bixel). The inner sections of the elevons will deflect propwash when up to help with this.

*GEV flight: At a certain forward speed I should be able to turn off the lift fans and have it skim the ground with GE lift only. With the CG where it is it will want to leave GE by itself and will need to be kept there manually.

*Normal flight: I expect it to fly very well based on my bixel build, even better with elevons and rudder for full control.

Interesting situations I forsee: Trying to turn when moving fast in hovercraft mode, using the rudder....possible sideways flip?

Transitioning from hovercraft lift to GE lift - given my track record, I'm going to mess this up at least once. Either turn off the fans too early or forget to turn them back on when landing.

Well, I'm sure you're heartily sick of reading by now, so I'll finish up by saying I'll start the build tomorrow and document my progress. Summer here has been cancelled yet again so it may be a while before the wind calms down enough for any testing.


Edit (September 2017): So I just noticed that for the past two years this post has been up, the link that was supposed to point to my original bixel build, was linking to the single engine hovercraft thread by mistake. Fixed now, better late than never?
Last edited by FiftySlicks; Sep 28, 2017 at 08:26 AM.
Sign up now
to remove ads between posts
Aug 08, 2015, 03:15 AM
Registered User
It should be a really fun project which I'm sure will work.
Ive often considered fitting wings to one of my models, but have tried to stay focused on ground effect which is the real trick. There have been a few attempts over the years to produce an aircraft with an air cushion landing system, but they all relied on using a skirt which was their downfall. The secret is to use the entire lifting surface for the air cushion to keep the cushion pressure low and therefore power for lift to a minimum.
A model can have a really good power to weight ratio so your idea should work well even if its not that efficient. Please keep us posted.
Aug 10, 2015, 01:19 PM
Registered User
Thread OP
Only had a single afternoon to build so far and thankfully the most annoying part of it is over. Cutting out all those spacers and getting them lined up and glued into place was pure tedium. In fact I made a slight slip up with a few of them and part of the sidewall is a little misaligned in one place but it's a difference of a few mm in a model that's a full metre long so I can live with that.

Material of choice is 6mm Depron aero which I've grown to like far more than the standard Depron. In return for a little less stiffness and direct strength, it's lighter (15% or so) and far less brittle meaning it's overall more resilient and easier to work with. Most of the construction is hot-glued together. The spacers that support the rear half of the sidewall where the outboard wings will be attached are doubled up to 12mm thickness.

To make the bevels on the base perimeter and sidewalls I lay the sheet on the edge of a thick piece of mdf, then use a metal ruler on the top surface to act as a guide for hot-wire cutting. I use a little battery powered hot wire cutter to get a clean cut. Ideally you would use two metal guides for the wire as imperfections in the wood give a slightly wavy cut in places, but for this application it's fine.

I still need to add some spacers on the front and rear walls, the problem is that I don't want to restrict airflow to the corners too much so they will need to be a lower profile and shorter.

Next up is installing the motors and beginning work on the top plate. It's a large area that has to include cutouts for the lift fans and the motor pylon as well as providing a hatch to access the battery so I don't have to remove the entire top plate to change it. I anticipate the battery being located somewhere slightly ahead of the midpoint, the weight of the battery plus the propulsion motor will counterbalance the outboard wings/rudder/elevons to get the CG right. At this point a lot of things are going to be temporarily taped into place so I can get the balance right.

If all goes well the hovercraft section should be done fairly soon so we'll see what that can do.
Aug 11, 2015, 03:20 AM
Registered User
Thats pretty big and light, my guess is you'll be flying most of the time. Even with just a light headwind it will lift off at walking speed so concentrate on stability and control you'll need it
Aug 14, 2015, 06:38 AM
Registered User
Thread OP
I had this back-and-forth decision with how big to make it, since the depron comes in 120x80cm sheets it's very easy to make large sections with it. The length is manageable but the width with the outboard wings on will be getting a bit much, hopefully I can come up with some way of making the wings foldable inwards at some point. I'm planning on a 2200mah battery but if it ends up too light to be manageable I have plenty of heavier ones. The bixel is stable out of ground effect but certainly not efficient as it's just a big ol' flat plate so it needs a lot of speed.

Build progress: Hovercraft is almost ready to roll, er, float. Need to cut the stabilising ports and add the tape 'skirt' and a temporary rudder to try it out. Propulsion motor is mounted as a puller for cooling.
Aug 28, 2015, 12:05 PM
Registered User
Thread OP
Quite a bit to report so I'll split it across a couple of posts.

The stabilising ports are 6x4cm and are positioned with their outer edge underneath the prop tip. Each port is ~7% of a single prop area which doesn't seem like much but they catch the most intense propwash and certainly do their job well. When it's running over grass you can see the propwash creating little dimples/deflections directly under the ports. I was going to experiment with different areas but it hovered really well as it is so I doubt I can get much of an improvement. In dead calm on a level surface the hovercraft section would levitate perfectly with no tendency to drift or roll, the stability was really spot on.

Electronics are 12A ESCs for the lift motors and a single 20A ESC for propulsion and BEC for powering the servos. RX is a Lemon 6 channel model, they are excellent receivers for the money, far better than Orangerx in my experience. I couldn't seem to hunt down the case for the lemon though so I used an old orangerx case. I attach most things to foam using these wonderful adhesive cable tie base things, they are really useful things to have around, though I hot glue them down since the adhesive on them isn't great in the long term.

I taped on a temporary rudder for testing out the hovercraft on it's own before adding the wings:

Peripheral Jet Hovercraft test (1 min 11 sec)

That's me doing my part to maintain local air quality then!

The dust might make for a cool effect but it also coated the lovely clean foam, thankfully this was easy enough to wipe down. Then I discovered....well, a warm still evening is great for testing, but it's also great for all manner of flies and insects to come out, an open air prop must chop quite a number of them up over it's time, so when you get a craft like this sucking in loads of air....yeah. I'm not generally insect-phobic but this was not much fun to clean up. I think the beautifully round circle under the prop tips is pretty much part of the foam now. Yuck.

Next up: Building the outboard wings and seeing if this thing can fly.
Sep 01, 2015, 04:40 PM
Registered User
Thread OP
This post is about a week late, have been busy and somewhat unmotivated recently, but the short version is that this big lump o' foam flies!*

*With many oddities, quirks and cases of 'didn't quite expect that!'

I was struggling to think of a way to attach those outboard wings, as they need to be secure, support the elevons properly and also able to be removed for transport. While sitting looking at the hull I realised the perfect mounts were already there: the tops of the rear spacers. The wings are nothing more than foam with balsa bits stuck to them and they line up with the spacers and are taped into place. If I was to build this again I'd tweak the alignment of things to get it a bit neater but I've used copious amounts of tape to seal everything up.

Servos of choice are Hextronik MG14, digital metal gear. Plenty of grunt and durability for the large control surfaces here. I've secured them well using a built up foam mount and balsa plates for reinforcement where the control horns go through the surfaces. The leading edge was originally a failed experiment with making an aerodynamic cover for a quadcopter.

Onto the tests: I had a few things working against me. The lift motors simply weren't enough to overcome all the added weight and it can't get 'above' the grass properly, this was made worse by the tape skirt that forms the peripheral jet slot coming unstuck and losing even more lift. Still I got enough lift to be useful, it drags a bit on the really thick grass, but generally was fine. To get the CG right (more on that later) I had to add some nose ballast which didn't help things either.

RC Flying Hovercraft test (1 min 23 sec)

The handling of this thing is like nothing else. Where do I start....I've got the lift fans set to the left slider (index finger) on the radio, the rest of the controls are conventional with rudder/elevon. It took me some time to get used to it.

*At low speeds the elevons do absolutely nothing, rudder has to be used both for yaw and roll. When trying to pick up speed the torque roll (I assume) tends to make one side dig in, this has to be corrected for first with rudder, then as speed builds up with the elevons. It can turn when moving quite fast along the ground but this requires aileron to be applied opposite from a usual coordinated turn which takes some un-learning.

*It's very evident when it begins to pick up GE lift on top of the fan lift, however it needs tons of forward speed to sustain this, turning off the lift fans early results in an ungraceful slide to a stop.

*Forcing it out of GE early (before it has enough speed to really sustain it) usually results in a pitch up and stall. It then parachutes down back onto the lift fan air cushion. To get it out of GE and flying I have to build lots of speed and then gently nudge it up, then only turn off the lift when it's well clear of the ground. I then enter a very delicate equilibrium: It has lots of lift but also lots of drag and the T/W ratio isn't very high. To get it to climb consistently I have to maintain a narrow angle of attack; too shallow and lift reduces too much even if it picks up more speed; too steep and the massive increase in drag will stop it almost dead in the air and it parachutes down slowly.

*It has a great deal of positive roll stability (ie acts like dihedral/self rights) due to the sweep of the planform, it is generally quite stable and controllable as it gains speed, the problems start if it's going too slowly and stalls out.

*CG: Done some quick testing. The rear limit is definitely at the midpoint of the hovercraft section, any further back results in it pitching up and stalling/falling back as soon as it gets forward speed. Which is good in a way, the CG can't really be too far back for flight because it can't get into flight with it too far back. CG further forward isn't ideal for the hovercraft but tames the handling a bit.

So, what I've learned is, I need more power. More for the lift fans to clear it properly above the grass, and more for propulsion to help drag it out of GE and flying properly. I've got new lift motors installed but have had problems with them, short version is I may have done goofed with that

More work to do on the hull, I'm going to double up the thickness of the sidewalls so the tape skirt can stick to the bevel better.

The good news is this concept works and is pretty awesome, the bad news is I'm going to be grounded by wind for at least the next week. Ah well, more time to build.
Sep 01, 2015, 10:20 PM
Registered User

Well done Fifty

Good to see you getting away from scraping gravel.

Just finished my version of hovering Bixel type craft. Have tested controls in hover mode before fuselage complete. Able to manipulate forward, aft and yaw movement.

Details to follow later.

Could you give us specific specs on yours? Weigh, engines, wattage draw, thrust measurements and such.
Thanks, Fred
Sep 04, 2015, 06:16 PM
derk's Avatar
Really cool mix of designs, I'm really wanting to build a good skirtless hovercraft, glad to see more successful ones being built
Oct 08, 2015, 04:28 PM
Registered User
Thread OP
Quick update: After several weeks with very little in the way of building or flying time, things are getting back to normal for me and I've resumed work on this and several other stalled projects.

Main upgrades are far better lift and propulsion motors, all D2836/9 950kv with 8x4.5 props for lift and a 9x6 for propulsion. Also improved the tape skirt so it holds it's shape better. Lift is better than before over grass although still not as much as I'd like. Thrust is much better which means I've finally had a proper chance to play with it flying.

Rc Flying Hovercraft 2 (2 min 48 sec)

Put the better handling on me spending more time learning this thing and also on a few pennies taped to the nose to bring the CG into the not as stupidly twitchy region.

On order is an even more powerful propulsion motor as the current one still struggles in trying to gain height without overheating. I'll be cutting new top plates to accommodate 9" props for lift and also probably a twin rudder to improve low speed turning.
Oct 08, 2015, 06:58 PM
Rapidly aging user
raw89's Avatar
Great vid Slick....looks cool flying....can you post some pics of the top & bottom in your current configuration?
Oct 09, 2015, 04:20 PM
Registered User
Thread OP

If there's anything about the design that isn't clear I'll get some extra photos or do some diagrams. The top is 4 separate panels, one for each prop cutout, and two smaller ones at the motor pylon. I vastly underestimated how much thrust I'd need when building and planned only a 7" prop max or 8" at a push so the motor mount has been bodged to extend upwards and the support beams are to stop it flapping all over the place. One change I need to make is to adjust the propline upwards, under full power there's enough flex for the prop to point downwards noticeably, meaning if I want a straight thrustline it'll need to be adjusted upwards a bit.

There isn't much to see on the bottom. Covered it in packing tape for protection, much of which (along with a lot of the tape) is coming off and needing redone.

When I get the 9" lift props on I'll hopefully get enough clearance to get this thing down a beach and over some rougher ground. When it's fully dialed in I'll rope someone in to help film in better and post it over in the scratchbuilt foamies section to see if it gets their interest.
Oct 19, 2015, 11:35 AM
Rapidly aging user
raw89's Avatar
Slicks - thanks for the info & pics....meant to get back before this.
One thing I found out on my skirted craft is to keep the lift fan opening about 1/2" smaller than the prop dia so they run under the opening about 1/4" on each tip. I get less air blowing back out of the fan opening by doing this.
Here's a (crappy) pic....
Feb 20, 2017, 01:53 PM
Registered User
Thread OP
I've made a bunch of changes to this thing in the past year all of which have considerably upped the performance. I've been keeping a detailed journal of observations that I will probably publish as a blog post soon since it's a little too long for this thread.

-The single top mounted motor had the power for flight but the high thrustline proved very difficult to handle. Replaced with two smaller motors at the leading edge, with differential thrust for yaw control.

-Vertical stabilser/rudder removed and replaced by vertical winglets on the outboard wings.


RC flying hovercraft 3 (3 min 4 sec)

Much better takeoff and flight performance, no more having to build speed for ages to lift the nose. Not having to fight the thrustline pitching the nose down constantly is a revelation, this is much more in line with the sort of performance and agility I thought I would get. Note the hover performance here is very poor for various reasons that have now been resolved:

-4 large heavy plane ESCs replaced with 4 tiny lightweight multirotor OPTO ESCs with dedicated BEC for power. Saved ~110g of weight by itself.

-Several efficiency improvements: Minimised air leakage, improved tape skirt stiffness, strengthened top plates to prevent the prop cutouts from ballooning out under pressure. All makes for better use of the available lift power.

Static hover height is up to about 50mm with some minor improvements still to do. I'm up against the current draw limits of the battery more than anything else now. Looking at the weather it may be a while before I can get this out again but I'll keep this thread updated.
Oct 16, 2019, 05:40 PM
Registered User
Thread OP
Quote: may be a while before I can get this out again but I'll keep this thread updated.
Hmm someone's been slacking off with this promise

RC Flying Hovercraft 4 (1 min 33 sec)

Last flight from what feels like 'just a little while ago' and it's nearly 2 years. This whole hobby fell by the wayside due to personal circumstances, I've been getting back into things and trying to make up for lost time.

The above video is with the new lighter ESCs so hover height is a bit better but still never as good as I wanted. I reverted back to a vertical fin instead of the plate winglets. I felt the latter were if anything giving too much yaw stability and made coordinated turning difficult with the differential thrust from the motors, also on the ground the winglets were a problem when trying to do drifting turns as they would catch the sideways airflow and roll the craft into the turn and dig the inside edge into the ground. The fin has much less side area so this is lessened, although to keep the thing level in drifting turns you need to roll with elevon opposite to the rudder (differential thrust) input, which is a very strange feeling on the sticks.

As I posted on the other thread I removed the wings and front motors to test out the hovercraft by itself:

Triple Motor Skirtless Hovercraft (re-upload) (7 min 15 sec)

First up youtube seems to have truly butchered the quality of this, I used to put these together in windows movie maker (snort!) and the quality seemed to be ok when uploaded, I'm now using videopad and it's way worse on YT compared to playing the video file. I am using the 1080/60 quality setting and it's not my connection at fault, I even deleted the first upload and redone the video at a much higher bitrate, this file was about 3 times larger than the first and the quality is still just as bad on YT.....I'm pretty well versed in photography but not video, apparently even when using the recommended codec YT can still wreck the quality pretty bad when it re-encodes it.

Then again if you're viewing on a phone or tablet maybe it looks fine compared to up close on a big monitor. I'm not going to try another upload but know that the cameras were delivering more detail than is shown.

Performance was pretty reasonable given the state this thing was in, there are very many improvements to be made in the (maybe upcoming) next version. At the motor mount, I prioritised an unbroken elevator and split surface rudder rather than the reverse as commonly found on 3D planes. Yaw authority is fine with large throws on the rudder and the elevator keeps the nose up at low speeds and the nose down at higher speeds. Bit of a flaw in that the motor mount is stuck down to the top plate of the hovercraft, this inflates out when the lift motors are powered and points the motor up a bit which needs more up-elevator to compensate. It would be better mounted straight to the bottom plate of the hovercraft but I wasn't thinking this far ahead.

I have a plan of remaking this with all the improvements I know from my other hovercraft builds, I can certainly tidy up the air inlets, peripheral slot and stabilisation ports to wring a bit more lift out of the motors. I was thinking of going all in on hover height with a bigger craft, maybe 100x70cm or even larger, it would be relegated to calm days only but that's fine as I have another single motor craft completed that is better suited to wind.

Also have some sketches of another flying version, no promises about it being built but.....I'll keep this thread updated

Quick Reply

Thread Tools