Efrag
A couple of thoughts on the principle-
1. I think you will find that this approach is very similar in concept to knock detection in modern automobiles/ecus. That is, you are looking for a vibration signature that is buried within a lot of machine noise. Without any benefit of test, I suspect that you will find it very difficult to differentiate 'offending' signatures like chatter from other signatures like drive harmonics or system coupling/resonance. Even varying the spindle speed, the cutter geometry, or the stock material will have a significant effect. Long story short, my guess is that without a whole lot of effort, you are going to get a ton of false positives. So far as I know, the big machines do not leverage this approach- they use path speed optimization from the toolpath generator, and also strain gauges on the drives. If an accelerometer was simple and effective, it would be in common use.
2. I dont believe a stamp has the power required to oversample and interpret the vibration signatures. You will probably need to look at the dsPic or equivalent that is designed to do digital signal processing. Or, there are a number of OEM knock detection chips for automotive that you can repurpose- assuming that they are configurable enough to catch the events in question.
I have done a lot of work in knock detection in engines- and it is far more complicated than just looking for a voltage spike. And, the signature of the event changes for each engine geometry and placement of the sensor.
Good luck-
Rob  |