[rainman]

Hi all,

I'm trying to simulate the above effect.

The product will be a platform and holding this platform will be 2 or 3 actuators on each side (made up of total 4 or 6). A person will be standing on top of this platform to experience the riding waves.

Basically for the start the wave will be a single direction, I will have the amptitube and also the frequency of sine wave. The "wave effect" has to be as natural as possible.

Next will be another similar wave profile that's coming from the side, hence creating a more rocky yet realistic effect.

I wished to get some ideas from here. It can be a "buy-off the shelve" product, an inverse kinetics software or purely a cam profile.

Will a normal geared servo actuator be "smooth" enough for this kind of profile? or pneumatic/hydralic is a better choice?

Thanks.
Rgds,
Rainman

[doorknob]

It sounds like an interesting project.

I wonder whether you could get some tips by looking at a commercial home theater motion platform such as the D-Box:

D-BOX / About D-BOX / Company Profile

I'm not sure of the details of their actuators, but a little bit of research might bring up some useful info.

[DangODave]

Hey, Rainman-

I would probably start by looking at what my requirements really are (reduced to accelerations, forces and so on) with some sort of simulation software. I have some experience with MapleSim, which is crazy expensive (unless you need it for work) and a free, open source project called Scilab that has simulation capabilities.

There's a learning curve, but especially in the case of Scilab, you can break your digital prototypes for a lot less than physical prototypes. There are good tutorials on the web, and it is really a matter of getting used to the software. Unless you are a mechatronics expert, I would look at online articles in, for instance, Machine Design magazine, on things like how much various drives transmit vibrations.

Is this a commercial project? You can get a non-commercial license of Mathematica for $295. It has a lot of good tutorials on simulation.

In any case, I think the project sounds fascinating, and my gut feeling is it is worth it to spend enough time with simulation and basic mockups until the gremlins reveal themselves. That's just me, though. And I am not suggesting paralysis through analysis. Ultimately, you gotta build it.

[Autonow]

While it may be simple to program the motion of an actuator. In the case of a wave the period will change with time. Your best bet is to create a feed back circuit to measure the frequency of the wave. you can use pressure (head) of the wave near the actuation point to trigger the actuators until the right amplitude is reached. A slight offset in the timing will kill a wave just as fast as making one, so a dynamic feed back system is required. Use a PLC and have fun with the program. it is not complicated and you can build a scale model for real time learning!

For actuators your best bet is hydraulic or pneumatic, They have the best energy density and speed. You do not need a lot of travel since you can build on the previous waves energy. Depending on wave size, pumps and valves may be a better solution.

[dan.andersson]

You really need to measure where the energy is in a wave, not only the frequency. And it's changing over time.

Controlling a contraption for a spectator or someone - that is the simple part.

I suggest:

Decide over your wave cycle time after when your wave program starts over again. You can also make a number of shorter cycles and then combine them randomly.

You need to record the real waves at a defined point, take out enough number of FFT analysis, doing it as fast as your rig can respond - say 2 samples per second. Then feed your FFT/sec to your rig.

This way will give you the waves, not only one sinusoidal wave that you call a wave. Real waves are complex and only sinusoidal in general.

Also, if your platform is to simulate a moving vessel/surfboard or whatever, you have different force vectors and you need a fresh set of FFT's at different speeds. Why different speeds? It might be to complex to calculate the interaction of speed as you need to calcilate the angle of impact, the mass of the vehicle, point of mass centre, density of water etc etc etc.

Make a recording instead!

Compare this to designing a "Tunnel Radio" that transmits on all FM frequencies at the same time. You create a recording of transmissions of all FM frequencies and then mix your audio warning message.

Make a wave recording, take out the necessary FFT per second calculations and use that to modulate your "wave". This way, you can jack up the wave height as you want it to be. You can replace wave height by accelerating the movements. Same as in real airplane simulators. You can use the same principle to simulate rough terrain.

Hmm, a speed boat simulator? Niice! Be sure to wear your kidney protection belt!

A nice oil ( hydraulic ) compressor can easily drive such a wave simulator but beware, wrong settings will create some interesting phenomena, and probably dangerous ones to.

REMEMBER! Mechanical and electronic stuff breaks up when you vibrate them. So do your internal organs and blood vessels if you are to hard on the controls!

Build a lot of safety protection gear while you can.

//Dan, M0DFI

[charles31]

Hi Rainman
Your application I suspect needs to major on human perception and for this the anticipation of the wave through visual cues is important. Sound, wind and cold salty spray are also needed to complete the picture. The person without sight of the waves simply gets tossed about whereas a sighted person can truly ride - or at least attempt to.
The ideal suspension is a vessel in a wave tank but I daresay this is not a practicality. A platform hanging from a trifiler suspension actuated by three servo-electric winches each compensated for their share of the static weight of the load by a pneumatic cylinder of appropriately adjusted pressure (or a spring of low spring rate) would be a method I would consider to get you pitch, roll and heave motion. A further two rope actuators could regulate the surge and sway accelerations while a 6th actuator establishes the yaw motion. Alternatively you could relax one or more of the latter three constraints and experiment with, for example, the polar inertia of the platform so that yaw motion in this case results from the rider’s actions.
I would adopt a working-man’s approach to signalling using a model wave tank (available at many universities) and wave height probes to signal the servos. A CCTV camera would magnify the wave images and project them in front of the rider. Of course all of this can be pre-recorded but you might wish to take the plunge and create waves in real time with one or two simple flaps in an adjacent miniature wave tank or analogue computer!
How did we get here from cnc?!! Good luck.