has anyone here had any experience of digitally modelling physical material properties, e.g. the curvature of a piece of plywood, and creating a digital model of that (preferably in real time) in processing. My aim is to create a surface that is representative of the physical model and evaluate it in an engineering software, checking stresses etc. Presumably I need to create a surface of sensors that matches the size of the piece of material I'm modelling....or something like that.
Can anyone recommend any tools, bits of hardware etc (I have an arduino board and basic kit already) and any sites that have useful code for this sort of thing.
I don't have a whole lot of experience with arduino but have a fair bit of programming knowledge.
I was imagining I would measure it using some sort of sensory sheet or attach connected multiple sensors. I have seen examples of this in other software such as grasshopper and firefly but I am more interested in arduino and processing. See this video for an example of what I mean by a sensory sheet: http://vimeo.com/14637158.
CrossRoads, thanks for the suggestion but there's no benefit in modelling it in the computer (either in skecthup or any other digital modelling software) as I need a true representation of the material's properties which digital modelling can not achieve.
For the curvature of a piece of plywood (simple 2D curve), you could potentially build a connected "string" of potentiometers; the shaft of each connected to the body of the next; the resolution wouldn't be great (keep them as close together as possible?) then use an analog multiplexer setup to allow you to read each potentiometer via the analog inputs.
Another possibility might be to build a 2D "gantry" scanner, with an ultrasonic and/or Sharp IR sensor pointing downward to measure a 2D grid of height values (limited by the resolution of the sensor(s), motor drive mechanisms, and by the materials being measured).
There are other options (really, it would be best to bypass the Arduino, and invest in a Kinect to hack to the PC) - but at a certain point you are merely reinventing the wheel (perhaps badly, depending on your skills) - wasting time and money where it might be better to invest it in a commercial solution.
It all depends on what your end-goals are, of course (learning and experimentation, vs getting the data and working with it to solve some other problem or need).
