Hi all,
We are wanting to decide the best method for STA roll shape plots for paper machine press and calender stack rolls. Our 4 analysts use 3 different methods and we would like to decide on one for most measurements so comparisons can be made accros machines. We dont do this as a routine (yet), only if there is an issue. My thoughts are accelerometer in line with nip, integrated to velocity (Displyed in RMS?)so low freq eccentricity and higher freq calender barring can be seen.50 averages. I used to use an LVDT in line with nip mounted from one housing to the next,p-p displacement 50 averages but this was a pain to set up (back in the days of 2110, home made power supply and circuit for an LVDT robbed from somewhere) and the lvdt we used probably didnt have a great freq resonce to higher freqs. Good for 14 lobed press CC rolls but recently we have been getting 92 lobed queen rolls (about 900Hz)
So the range is about 4-1000Hz
Any thoughts would help and the reasons behind.
Thanks in advance.

Are you actually wanting to profile the roll shape itself? I think you'll go back to the drawing board if that's the case.

Hi Sam. Not too concerned about profiling actual roll shape. Would need displacement for that. More about detecting, identifying which component, then trending. Cheers