As discussed over on the www.practicalmachinist.com Deckel forum, the horizontal spindle of my Deckel FP2NC has been making squealing noises for the past couple of months that have gotten worse over time.  The noises have persisted through two cleaning and regreasing cycles and the replacement of thrust bearing components.  I borrowed a vibration analyzer from work to see if I could figure out what is going on.

The first photo shows my hand holding the analyzer and the accelerometer (vibration sensor) held to the bottom of the quill housing with a magnet.

Deckel_FP2NC_Spindle_Squeal_01.jpg

The second photo shows a closeup of the accelerometer with the quill partially retracted.

Deckel_FP2NC_Spindle_Squeal_02.jpg

The resulting vibration spectra for situations where the squeal happens (quill fully extended) and does not (quill partially retracted), when running the spindle at 3,150 RPM (top speed), are compared below.

Deckel_FP2NC_Spindle_Squeal_03.jpg

When the spindle is squealing, there are very distinct vibration peaks at around 3.6 kHz, 10 kHz, and 19.2 kHz.

Notably, when the squeal is not too strong, lightly grasping the quill as shown in the next photo is enough to kill the squeal.

Deckel_FP2NC_Spindle_Squeal_04.jpg

The next spectra show that the squealing frequency is independent of the spindle speed.  This means that it is not caused by issues with the rolling elements.

Deckel_FP2NC_Spindle_Squeal_05.jpg

Since the frequencies are much higher than anything that could be associated with the rolling elements anyway, and since grasping the quill as shown above quiets the noise, I suspected that the quill, which is a tube, was ringing like a bell.  I looked up the theory of the in-plane vibration of the thin-walled ring and plugged in the numbers for the quill (80-mm OD, 65-mm ID, modulus and density for steel), and I obtained the following frequencies for the 2nd, 3rd, and 4th modes:  3.5 kHz, 10 kHz, and 19.1 kHz.  These are remarkably close to the measured values for the spectral peaks.

In the 2nd mode, the tube oscillates between a flattened oval and a tall oval shape.  In the 3rd mode, it oscillates between two orientations of a tri-lobed (like a Wankel engine rotor) shape.  In the 4th mode, it oscillated between two orientations of a four-lobed (square with very rounded corners) shape.  The 1st mode is the breathing mode, that is, uniform circular expansion of the tube like uniformly stretching a circular rubber band.  I do not seem to be getting the 1st mode in my measurements.

I also did a bump test, where I took spectra as I tapped on the quill while the spindle was stopped.

Deckiel_FP2NC_Spindle_Squeal_06.jpg

The peak around 3.5 kHz is apparent, but the higher peaks are not.  There are peaks at a few hundred Hz and 6.4 kHz that are not present in the measurements with the spindle running.  These must be some other mode that does not get excited when running.

The vibration that is happening with the spindle running is very much analogous to that of a wine glass when a wetted finger is run around the rim.  I am hoping that returning to the OEM Isoflex Super Tel grease will alter the "finger friction" enough to stop the ringing.  However, it is also a very natural phenomenon intrinsic to a tube structure that doesn't have much damping, and therefore may not go away easily.