Laser vibrometry for vibration measurements on rotating milling tools
The knowledge of the dynamic behaviour of a milling process is very important for finding an optimum process window. In today’s manufacturing industry the machining parameters are often predicted using experimental data from non-rotating spindles.
Many times the predicted machining parameters prove to be ineffective and inaccurate which lead to reduced quality of the machined surface, tool wear, noise or at worst spindle failure. Spindle and tool vibration measurements during operational conditions are therefore desirable.
Laser vibrometry is a well-established non-contact, non-disturbing method commonly used for measurements of vibrations on static objects. The non-contact and remote nature of the technique offers the possibility to measure vibrations on rotating high-speed spindles and tools. However, two major problems occur when measuring on rotating surfaces, namely (1) speckle noise and (2) crosstalk between the vibration components. These two effects make vibration measurements on rotating spindles difficult to interpret and must always be considered, especially at higher spindle speeds.
In this project the speckle noise and the crosstalk between the velocity components of a rotating spindle is studied experimentally. We have shown that by polishing the measurement surface optically smooth we are able to avoid the speckle noise and the crosstalk problem. By using this approach, the vibrations as well as the roundness and the radial misalignment of the measured rotating spindle can be resolved.
In an experiment, shoulder cutting tests of an aluminium workpiece were performed using a 16 mm R390 mill with two cutters in a five-axis high-speed vertical milling machine centre, Liechti Turbomill ST1200. The machining conditions were set as following; spindle speed: 19000 rpm, axial depth of cut: 5 mm, radial depth of cut: 1 mm, and feed rate: 6200 mm/min.
In the figure the measured displacement of the cutting tool, Dy together with the measured cutting force, Fy measured on the workpiece are shown as a function of spindle rotation angle. Here, the radial misalignment error and the out-of-roundness of the tool are subtracted from the raw laser vibrometry signal, hence the true cutting vibrations are resolved. It can be seen that cutter one is engaged between 80 and 110 degrees and after the first cutter disengagement the tool continue to vibrate freely until the next cutter edge engage between 260 and 290 degrees. Further, the amplitude of the first cutter is higher than for the second cutter in both tool displacement measurement and cutting force measurement. The difference can be explained by the actual asymmetries of the cutters shape and positions.
Updated: