Using speed variation to counteract chatter in a lathe

[Norppu]

Chatter can be a real PITA for small lathes that are not very rigid. It may also become a problem when the work piece has a tendency to "sing".
The project, i am about to start, deals with the resonance type of chatter.
It is a simple micro controller powered device that alters the VFD speed periodically. The speed variation depth and frequency can be adjusted as well as the center frequency. The device is a simple add-on to any VFD that can be controlled by push buttons and whose speed can be controlled with a voltage (potentiometer). There should also be a suitable (12VCD) power output for this device. Most of the cheap chinese VFD's can be used with this one.

I have uploaded a video HERE.

Later on there will be the full documentation, program (source) code and the gerber files available.

 

[benmychree]

One can do the same thing with any lathe, if it chatters, slow it down!

 

[Bob Korves]

One can do the same thing with any lathe, if it chatters, slow it down!

John, I think his idea is to vary the speed of the work slightly but constantly, up and down, to never let it dwell long enough at a specific harmonic speed to build waves. By the time the resonance starts, the speed has changed to a different value that will not support or reinforce the chatter. I assume the spindle speed would only vary slightly up or down, probably not even noticeably to the operator, and probably have a knob the operator could use to search and find a slightly different and continuously varying speed that does not build or reinforce harmonics, while the lathe is cutting. Is that correct, Esko?

 

[jwmelvin]

Later on there will be the full documentation, program (source) code and the gerber files available.

Thank you, this is excellent of you.

 

[gradient]

I tried something like that a couple of years ago with our Tormach mill. It works but you have to find the sweet spot depending on the type of material and how well it is supported. Below is the code I used if anyone wants to try it out. It is based on the method used by Haas machines and a random number generator referenced in the code. As they say, no warranty is expressed or implied; your mileage may vary.

%
OMD_face_mill_test_Rand
( POSTPROCESSOR: )
( GENERATED BY SprutCAM and MD)
( DATE: 5/31/2016 )
( TIME: 7:36:27 AM )
(Tool) (1) (Diameter)(4.00) (Fly Cutter) (Operation) (2D contouring)
#<XSIZE> = 8.00
#<YSIZE> = 7.04
#<TOOLDIAM> = 4.0
#<IPM> = 9.0
#<RPM> = 490
#<NEWRPM> = #<RPM>
#<VARIATION>=0.2 (ie. 20%)
#<DELTA> = [#<RPM> * #<VARIATION>]
#<minRPM> = [[1 - #<VARIATION>] * #<RPM>]
#<maxRPM> = [[1 + #<VARIATION>] * #<RPM>]
#<XSTART> = [-#<TOOLDIAM>/2]
#<POSITION> = 0.0
#<INC>=[[#<IPM> * 100/60] / [#<RPM> * #<VARIATION>]]
#<_SEED>=1
#<_RANDOM>=0
(www.cnczone.com/forums/haas-mills/59194-code-RANDOM-number.html)
o300 sub
#<_SEED>=106*#<_SEED>+1283
#<_SEED>=#<_SEED>-6075*FIX[#<_SEED>/6075]
#<_RANDOM>=[[#<_SEED>/3037]-1.0]
o300 endsub
G90 G64 G50 G54 G80 G17 G40 G49
G20 (Inch)
(2D contouring)
G30
T1 G43 H1 M6
(Fly Cutter)
S#<RPM> M3 M8 (set RPM, spindle on, coolant on)
G0 G94 X[#<POSITION> + #<XSTART>] Y[-#<YSIZE>/2]
(1/2 tool diam left of part, Y mid)
Z0.1 (rapid to z=0.1)
G1 Z0.0 F#<IPM> (straight line cutter move to z=0, @ xxx IPM)
o100 while [#<POSITION> LT [#<XSIZE> + #<TOOLDIAM>/2]]
o300 call (call ran num gen, ret -1 to +1 in #<_RANDOM> )
#<NEWRPM> = [#<RPM> + [#<_RANDOM> * #<DELTA>]]
S#<NEWRPM>
X[#<XSTART> + #<POSITION>]
#<POSITION> = [#<POSITION> + #<INC>]
o100 endwhile
G0 Z0.1 (rapid to z=0.1)
G30 M5 M9 (home tool, turn off spindle, turn off coolant)
M30 (program end)
%

 

[MarkM]

Is there any real difference than having a good reliable vfd set up with a speed pot. All this is in the manual for the vfd if there is interferance to add chokes and such. Am I missing something here?

 

[benmychree]

John, I think his idea is to vary the speed of the work slightly but constantly, up and down, to never let it dwell long enough at a specific harmonic speed to build waves. By the time the resonance starts, the speed has changed to a different value that will not support or reinforce the chatter. I assume the spindle speed would only vary slightly up or down, probably not even noticeably to the operator, and probably have a knob the operator could use to search and find a slightly different and continuously varying speed that does not build or reinforce harmonics, while the lathe is cutting. Is that correct, Esko?

My approach to that problem is old school, one just lightly holds a metal object up against the vibrating part, what ever is at hand, maybe a lead hammer up against the rotating object or boring bar, whatever and manipulate its location until the chatter stops, or, slow it down. I remember that years ago, I saw ads for anti chatter rollers that were positioned on long turnings to suppress chatter, never saw one used.

 

[Norppu]

The idea ist to vary the speed constantly around a given center speed. The center speed, the variation frequency (how fast the speed varies) and the amount of variation are continuously adjustable. I am not sure whether the randomness makes any difference here. On my manual experiment I found out that it really does not matter how the speed changes as long as it is changing all the time.

 

[mikey]

Interesting project, Norppu. My understanding is that chatter can be addressed by altering speed, feed rate or the lead angle of the tool. Depending on the situation, one or more of these three factors may need to be altered to eliminate chatter. It will be interesting to see if simply focusing on speed alone works.

 

[FOMOGO]

This is madness! Madness I say!! You know, that just might work. Now back to your regular programing. Mike