Intelligent Manufacturing Laboratory at Purdue University

Adaptive Control of Cutting Force for End Milling Processes

Motivation

Fixed controllers are designed for nominal plants and effectiveness diminishes as the
cutting parameters vary.
Varying gains and dynamics require conservative feedrate settings for fixed control
systems
Maintaining desired cutting forces prolongs tool life increasing productivity
Standard MRAC has the following limitations:
- Introduces instability for nonminimum phase zeros
- Large input signal oscillation possible

Approaches:

Develop an adaptive control scheme to maintain constant cutting force during end
milling for time-varying cutting conditions.
Augment the performance index of the controller to reduce input signal oscillation.
Implement the adaptive control scheme on a high speed milling machine through a
PC-based open architecture.

Results:

An adaptive controller has been designed that will maintain a constant cutting force
for milling operations by controlling the feedrate of the table.
Extended MRAC designs offer an improvement over the Standard MRAC system in
its ability to control nonminimum phase plants.
Augmented Performance Index can effectively reduce input oscillation which is
detrimental to machine life.
A PC based open architecture controller was designed to control CNC machines, thus
allowing the testing of new algorithms.
Experimental implementation was performed using a workpiece containing two
changes in the axial depth of cut (Figure 2). While machining with a fixed parameter
controller showed step changes in cutting force due to the depth of cut changes
(Figure 3), the tests performed with the developed adaptive controller for various
radial depths of cut under varying machining conditions produced constant cutting
force (Figure 4).

Intelligent Manufacturing Laboratory at Purdue University

Tachometer

Spindle
Motor
Cutting Tool
Amplifier
Workpiece
Dynamometer

Lead
Screw

Table
Motor

Tachometer

Feed Table
Amplifier

PC #1

PC #2

Force Monitoring

Adaptive Controller
User Interface
Real-Time Display

Figure 1. Experimental set for adaptive cutting force control on a high speed milling
machine.

50 mm
50 mm

50 mm
Axial DOC

10 mm

20 mm
10 mm

Direction of Travel

Figure 2 - Cross Section of Workpiece Used in End Milling Experiments

Intelligent Manufacturing Laboratory at Purdue University

Input Signal
1

(V)

0.8
0.6

DOC=10mm

DOC=20mm

DOC=10mm

0.4
0.2
0
0

10

15
Time (sec)

20

25

30

20

25

30

Force Level
2500

(N)

2000
1500
1000
500
0
0

10

15
Time (sec)

Overshoot: 30 %, Settling Time: 3 Seconds

Sampling Period: 20ms, Cutter Speed: 1500rpm
Figure 3. Constant input force response without adaptive control.

Input Signal
1

(V)

0.8

DOC=10mm

DOC=20mm

DOC=10mm

0.6
0.4
0.2
0
0

10

15

20

25

15

20

25

Time (sec)
Force Level
2500

(N)

2000
1500
1000
500
0
0

10
Time (sec)

Figure 4. Input signal and force level with the adaptive controller
(0.080" DOC, W=10,000)