Open BrianClement opened 7 years ago
BrianClement
commented
7 years ago [Richard - 1/02/2017 10:53:49] I'll add a capture LVDTs of the RAMS closing during a "manual" close (operator driven from the USV dialog -not a scan), and store the data as acsv. This will give us some initial data, and allow tuning of the RAMS, offset closure delays(i.e. starting the slow RAM first), and total closure delay.
BrianClement
commented
7 years ago [Richard - 1/02/2017 21:29:12] Have 2 csv files in my Bitbucket, RAM_POT_Opening|Closing.csv, samples at 1khz over 2 seconds - so 2000 readings. First 2000 are Pot0 (Upper?) and second 2000 are Pot1 ... in lines of 10. The Lower pot starts closing 500ms earlier than the top - so should be easy to sort which is which.
BrianClement
commented
7 years ago [Grant - 2017-02-14 2:55:51 p.m.] Pot0 is the lower pot.
DB0084 [Created: 2017-01-18 by Jonathan]
How smoothly and synchronously the USV rams contact the sample is important since (a) if too violent may move disc possibly resulting in collision, damageor dropping out of frame and (b) impact measurement quality.
As for frame deflection, it would be good to be able to at least monitor (if not provide feedbacksufficient to tune the process) the ram contact process.
This might be done using:
(a) accelerometer(s) attached to either the ram or the frame (b) laser distance sensor (c) lvdt's already attached to each ram for thickness measurement (10kHz sample rate)
Since for (c) all the equipment exists already, it makes a good place to start. All that is needed is to start collecting lvdt data before the contact is actually made (currently lvdt datastarts to be collected at the same time as usv pulse is generated).
GE will try to log lvdt data for a complete ram cycle and we'll take it from there. It would be good if this were done for both an optimal and a far-from optimal device tuning.