View Full Version : General stability equations

davidhills - n/a
24-Jan-09, 10:59 AM
Good Morning

my home built model hovercraft keeps falling over!!

It's roundish with a skirt peripheral circumference of 1.9meters

The lift fans should be able to carry about 10kg from the freezer back test I did

Which had an equivalent. hovercap of 70cu cms. The model and batteries weigh

4 kg. So there should be plenty of air flow.

One side drops down and the other then has a large gap where all the air escapes.

It’s a bag and finger design made from an inner tube for the bag and ripstop nylon for the fingers it’s ratio of depth is 1/3 finger to 2/3 bag (sort of). At least in terms of the over all ride height. I don’t think the skirt design is very good and is not stable but I am also no sure t the dynamic or static stability of the whole craft is helping.

see attached word doc for diagram

My question is what’s the maths for establish if the craft is at least possible of being statically stable.

I guess I need to know the location of the centre of gravity of the vehicle(which I can pretty easily figure out) and some thing about the peripheral lift. My gut feeling is that I don’t have a strong enough peripheral jet effect and the inner plenum chamber pressure is too high. So it’s like trying to balance a broom on its handle.

What I like to be able to do is design the layout so that it has a chance of being stable then measure some how the necessary stability values from the model so

1)it works

2)I understand why it works

Thanks All

David Hills

Paul Fitz - n/a
24-Jan-09, 07:28 PM
Whilst it is possible to estimate craft stability mathematically, it is extremely complicated and without recourse to known data and a barrage of expensive computers, probably very inaccurate. I wouldn't attempt the calculations!! http://hovercraft.org.uk/images/icons/smiley_icons/icon_surprised.gif)

Your post probably partially answers your problem.

A disc shaped craft is always the most unstable. Placing of the centre of gravity and centre of roll is critical and they often suffer from a nutational motion caused by aerodynamic effects of the skirt, especially when bag skirts are used.

I would suggest that the easiest way to resolve your stability problem is with a baffle. I gather from your attachment that the air feed is peripheral. You could therefore introduce a central baffle comprising a curtain skirt (or bag) which circles the centre of the craft, thereby producing an annular cushion plenum. The space enclosed at the centre must be vented to atmosphere to prevent it pressurising to cushion pressure. This will however reduce the cushion area, which will increase the cushion pressure required. The actual area required will need a little trial and error. I suggest you start with the baffle skirt set at about 30% of the radius from the craft centre. I suspect that you will need to move this to about 45% or more, but you will need to experiment.

Have fun.


davidhills - n/a
25-Jan-09, 10:45 AM
Thanks Paul

I suspect your right about the math’s, but it would be nice to be able to under stand why something work, if you can.

Can I be cheeky and ask for a sketch of your idea as I’m not exactly clear from your description.

I was thinking about putting a Peri-Cell off each feed from the bag, but that’s a lot of work.

see attachment

Ian Brooks - n/a
25-Jan-09, 10:55 AM
As Paul says, general (or dynamic) stability of a hovercraft is difficult to model mathematically, although it could be approached using a simulation package such as Matlab/simulink if the scope of your project warranted it. There are texts on the subject if you need them.

However, your craft is not even statically stable, and it should be possible deal with that using hand calcs. The concepts you need are center of gravity CG and center of pressure (CP).

Because the CG is generally above the CP, the craft will tend to be statically unstable, like a marble balanced on a football, and this is why your model falls over. To fix this, there has to be a way for the CP to move in response to the tipping of the craft - ie the CP has to automatically adjust itself as the craft rolls.

Put another way, if the craft rolls over, the cushion must generate a restoring moment which opposes the force causing the roll.

This is usually achieved in a modern small craft via contact of the skirt with the ground, although there could be some periferal jet effect if large lift powers are used. In fact, its a fallacy that skirted hovercraft are completely free of ground contact, and this albeit light ground contact force is used to provide stability.

Your segments should meet the ground at an angle no more than 45 degrees. Then, if one side drops, the segment contacts the floor and as the roll continues, progressively more contact area is produced. In effect this moves the center of pressure in the direction of the roll, creates a restoring moment and craft will settle, somewhat rolled over, but stable. The static stability can then be expressed as roll angle/roll moment.

If this makes sense, you should be able to redesign the segments for stability, and predict the roll stablity to compare with experiment. If it doesn't make sense, PM me and I'll see how else I might help.


davidhills - n/a
26-Jan-09, 01:16 PM
Thanks, that makes a lot of sense.

my segments are almost vertical.

I had no idea that some of the stability came from ground contact



Paul Fitz - n/a
27-Jan-09, 11:40 PM
As Ian quite rightly states, the circular craft will be unstable because its payload will be central, with its centre of gravity above the centre of pressure. Theoretically, it might be stable when static, but as soon as you impart sideways thrust, the craft will tip and the CoG will move making it unstable. You could reduce the effect by lowering the hover height, making the craft larger and/or improving the effectiveness of the skirt.

My suggestion is to modify the skirt/cushion system to form an annular cushion (see diagram. This effectively moves the centre of pressure away from the centre of the craft further toward the side which is lower (on the circular path shown). The CoG will remain near the centre of the craft. This will improve the overall stability and improve the righting moment of the cushion. If the skirt contact point meets the ground at 45 deg (ideal) or more, this will improve the skirt recovery time.


kach22i - n/a
31-Jan-09, 09:25 PM
David, as good and as informative as the diagrams are, I think a photo would be worth a thousand words.

Over 20 years ago I tried making a hovercraft model and it did pretty much what yours is doing.

My mistakes:

1. I put the largest propeller I could find on the smallest engine and it would not create the static pressure required to keep lift going without falling to one side or another. All flow and no pressure.

2. The cushion height was too tall/high for the lift area.

3. The fingers went straight down, and not at 45 degrees.

4. It was a lot of work, but I learned a lot.

5. It was less work and money than messing around with a full sized craft, an education is an expensive thing.


http://www.hoverclubofamerica.org/forum/index.php?autocom=ga (www.hoverclubofamerica.org/forum/index.php?autocom=gallery&req=sc&cat=7&sort_key=idate&order_key=DESC&prune_key=*&st=0) llery&req=sc&cat=7&sort_key=idate&order_key= DESC&prune_key=*&st=0



I need more information about the lift fan set-up to help you any more.

Cheers, George/kach22i

NOTE: If you want to see more picture of this model just ask and I'll post, but it was a flop.

Sean Atterbury - n/a
1-Feb-09, 09:00 AM
I say that your skirt's too tall, how high do you want it to hover, two stories maybe 3. Cut the skirt down to say 100mm, thats my 10c

davidhills - n/a
1-Feb-09, 01:26 PM
I've scraped MK1

I now working on MK2

Hear some photo's MK1 and MK2



mark 1


kach22i - n/a
2-Feb-09, 09:56 PM
David, I think the problem was too high and no 45 degree angle. I kind of expected larger images showing the details you know are in question, so I'm just guessing blind.

I like what I see so far in the MkII, where did you source the fans and fan housings from? Can you provide a link or detailed picture?

Here are some pictures of what I think is good documentation, a standard for all to strive for.





Link-3 (Mark Porter)






kach22i - n/a
5-Feb-09, 09:19 PM
I've given this more thought.

As seen in the first posted diagram there is a web between the skirt and plenum chamber beneath the craft. Does this web have holes in it, is it constant? It needs to allow air flow to the middle of the craft.

To get any hovercraft to work there must be a capture bubble of pressurized air "under" the craft. This is what really supports the craft, not the fabric curtain all around it. The curtain is just supposed to hold in the air so that it maintains a lift pressure.

Two solutions:

1. Angle in the fingers so that the flow of air gets under the craft and builds lift pressure.

2. Provide a second or direct air (shaft) path from the fan to the plenum chamber under the craft.

NOTE: The point where the fingers touch the ground should NOT be outside of the edge of the hull above or it will give in, deform and not recover.

davidhills - n/a
6-Feb-09, 09:18 AM
The fans are from a ford escort they are 12 volt and pull 15amps

I did a freezer bag test to see what they could support and it was 10kg with both fans and 5 Kg with just 1 fan

It should way about 3 kg with the batteries