Looking for ideas?

I am looking for ideas here from all you "experts" (where an expert is described as - Ex (a has been) spurt 9 a drip under pressure))

Nothing like insulting the unpaid help! before they start :-)  happy New year everyone.

OK A friend has a model of a fair ground ride, this is a steam yacht, http://www.youtube.com/watch?v=zp76RTbqqLk

he wants to animate this and would like the motion to be realistic at the scale size. He had tried a number of mechanical solutions, cranks etc but they all give a jerky motion and the real thing is very smooth.

Any suggestions.

I am inclined to think this is a micro processor solution with either a suitable servo motor or similar driven to mimic the inertia of the real thing.

ideas folks?

sort by: active | newest | oldest
lemonie4 years ago
Yea, they always look "crappy" in Meccano. It would help if the "vehicles" were filled with lead, such that the actually swung. But you'd need a tricky drive mech'. That and the magnets Kiteman suggested.

L
rickharris (author)  lemonie4 years ago
Thanks for the comment I can't select a best from these as they are all pertinent to the problem and parallel my own thoughts on the problem I think I will have to do some experimentation.

I agree about meccano.

Once the drive is sorted out this will be a full scale model of a steam yacht .

My friend already has several models built of different steam rides, but this is proving to be the hardest to make realistic.
If you had a big and heavy eccentric mechanism under the table, you could get a "natural" movement, but it needs the "gravitational-swing"-motion and I don't know what geometries you would need.

L
rickharris (author)  lemonie4 years ago
On current models the biggest issue is the swing period is too short to look realistic for a model. the real thing swings fairly lowly - or appears to do so.

I think some kind of continuous drive to control the swing throughout it's arc will be necessary.
I think that I figured this out:
The swing of a pendulum is governed by it's length, so the swing on the model is mechanically "wrong" because gravity cannot be scaled with the model.
It needs to use your imagination to see the things like you're in an aeroplane and not looking at them actual size.

L
Add some damping to the mechanical system.
rickharris (author)  steveastrouk4 years ago
Thanks for the comment I can't select a best from these as they are all pertinent to the problem and parallel my own thoughts on the problem I think I will have to do some experimentation.

I am thinking I can use a dc motor as an intermittent drive to push the boat along as you might push a swing. A little reverse current could then provide the damping.
I used to have a "rocking" motor, used in shop displays, which was a pendulum and an electromagnet at the bottom of the swing. The pendulum had a soft iron lump on it, which swung through the magnet. A little flappy switch made brief contact as the lump swung through the core. Presumably, the core was briefly magetised in the opposite direction to the EM field and kicked out. The magnet switches off, the lump falls back through the core....
rickharris (author)  steveastrouk4 years ago
I have a false pendulum for a wall clock that is a bit like that - My only worry is the mass of this model might be fairly high. I think he works at 1/12 scale so if the real thing is around 20 feet high = 20 inches.

I will bear it in mind though.
iceng4 years ago
A uP driving a stepper can imitate the pendulum action.
Acceleration from top, speed at the bottom, deceleration to top & repeat.

Or a uP driving a DC motor with PWM on direct drive and current sensing
to detect reversal points and hall detectors for max swing limit edges for
the pendulum motion.

I think a servo would not do a good pendulum swing even with a uP.
One would have to reduce the DC supply voltage to avoid jerky moves.

A
rickharris (author)  iceng4 years ago
Thanks for the comment I can't select a best from these as they are all pertinent to the problem and parallel my own thoughts on the problem I think I will have to do some experimentation.

It would seem from others efforts that getting a model to mimic the real world in terms of a gently slow down at the top and increasing acceleration back down is hard - In general things tend not to look scaled.

I think as you say this will have to be driven all the time with some mimicking of the acceleration curves. No doubt there is some fancy maths but I think some experimentation may produce satisfactory simulation.
Kiteman4 years ago
Have you come across a "magbot"? It's a BEAM curcuit; a magnet would go inside the floor of the swing, and a coil in the base if the model. The current to the coil is varied, making the pendulum swing.
rickharris (author)  Kiteman4 years ago
Thanks for the comment I can't select a best from these as they are all pertinent to the problem and parallel my own thoughts on the problem I think I will have to do some experimentation.

I think (but don't know) the finished model may be too large to pulse from a coil and get realistic scale movement.