Vous êtes sur la page 1sur 14

The Implementation of Optimum MRR on

Digital PC-Based Lathe System

Tian-Syung Lan1 and Kuei-Shu Hsu2


1
Department of Mechanical Engineering, De Lin Institute of Technology, Tuchen, Taiwan
236, R.O.C.
2
Department of Mechanical and Automation Engineering, Kao Yuan University.
Lu-Chu Hsiang, Kaohsiung 821, Taiwan, R.O.C.
E-mail: kshsu@cc.kyu.edu.tw
Abstract

Optimizing the profit of an individual cutting tool is crucial to the computer

numerical controlled (CNC) machining industry. In this paper, the mathematical

modeling, the dynamic solution, and the decision criteria through Calculus of Variations

are introduced to achieve the optimal material removal rate (MRR) control of a cutting

tool under the fixed tool life. To realize the optimum MRR, a commercialized lathe

system with a DSP (Digital Processor Controller) and a man-machine interface is

developed. Additionally, the implementation of dynamic MRR control for a real-world

industrial case is experimentally performed on our proposed digital PC-based lathe

system. It is found that the surface roughness of all machined work-pieces not only

stabilizes as the tool consumed, but also accomplishes the recognized standard for finish

turning. In this study, the adaptability of the dynamic control of optimum MRR as well as

the realization of the digital PC-based lathe system are absolutely guaranteed.

Keywords: material removal rate; digital lathe control, man-machine interface; surface
roughness
*
Corresponding author: Kuei-Shu Hsu
Mailing address: Department of Mechanical and Automation Engineering, Kao Yuan
University, Lu-Chu Hsiang, Kaohsiung 821, Taiwan, R.O.C.
E-mail: kshsu@cc.kyu.edu.tw

1
1. List of Symbols

a = average volume of material machined per unit part.

B = upper speed limit of MRR.

bM ′(t ) = marginal operation cost at the material removal rate M ′(t ) ; where b is a

constant.

bM ′ 2 (t ) = operational cost at time t.

c = overall holding cost of unit chip per unit time, including chip holding and finished

part holding costs.

M (t ) = cumulated volume of material machined during time interval [0, t ] .

M ′(t ) = MRR at time t.

P = revenue per unit part machined.

T = tool life for the dynamic model, which denotes the cumulated machining time before

a tool replacement.
~
t = time when the optimum MRR reaches the upper limit B .

2. Introduction

The material removal rate (MRR) is an important control factor of machining

operation [1, 2], and the control of machining rate is also crucial for production planners.

Since the modern computer numerical controlled (CNC) machines are extensively used to

perform various tasks, e.g., from job shops to flexible manufacturing systems (FMS), the

problem of optimal machining control of CNC has received considerable attention [3].

2
Therefore, the necessity of finding the optimal MRR control for maximum profit as well

as the implementation of dynamic optimization for an individual tool in the CNC

machine cannot be overlooked.

The mathematical modeling and the optimal solution of MRR were presented in [1].

This paper is aimed at developing a PC-based machining system to achieve dynamic

MRR control by manipulating the feed-rate in accordance with constant depth of cut. Fuh

et al. [4] designed a variable structure system (VSS) controller for CNC turning machines.

Rober and Shin [5] developed a PC-based open architecture controller that can override

the pre-programmed (???) feed rate of the CNC milling machines. In [6], Kim and Kim

proposed an adaptive cutting force control approach for machining centers based on

direct cutting force measurement. In addition, in [7, 8, 9, 10], the idea of adaptive control

has been applied to help select the optimal feed-rate. Generally, the basic objective of

adaptive control is to maintain consistent performance of a system in the presence of

uncertainty or unknown variation in plant parameters. However, these existing on-line

control schemes are all considerably expensive. Moreover, none of them is guaranteed to

achieve the maximum profit.

On the other hand, many researchers employed the direct-drive method to rotate the

ball-screw shaft, in which continuous adjustment of feed-rate can be guaranteed.

However, it has been reported that the performance of the direct-drive method is

hampered by some practical limitations. The CNC has an feed-rate adjustment function

that can be used to adjust the value of desired federate in percentage, which is often done

manually using a setting dial on the control panel. In order to incorporate the digital

PC-based system into commercialized CNC machine tools, this study develops the

interface between the digital PC-based control method and the CNC machines.

Recently, the profit optimization of an individual cutting tool in machining

3
operation has become an important issue for CNC machining industry. In this paper, the

mathematical modeling and dynamic MRR control to achieve the optimal machining

profit for a cutting tool during a fixed life time is introduced. To realize the dynamic

solution, the control scheme on a commercialized lathe system with DSP (Digital

Processor Controller) and a Man-machine Interface is developed. Additionally, a

real-world industrial machining task is performed on our proposed digital PC-based lathe

system.

3. Theoretical Background

In the previous research [1], the cutting process is regarded as a continuous

single-tool turning operation without breakdown. The data for both the fixed tool life

time and the upper MRR limit are obtained from the maximal machining conditions

suggested in the machining handbook. In general, if the machine is operated within the

tool life time, it will not break even with the highest machining rate.

The machining costs are divided into two different categories [1, 2], the operational

cost and the holding cost. While the marginal operation cost is a linear increasing

function of the production rate [11], the operational cost of the machine is directly

proportional to the square of the MRR [1, 2]. In most manufacturing cases, all chips from

cutting and the finished products are usually held and stored at the machine until tool

replacement. All parts are paid at a given price after machining operation [1, 2].

M (T )
In [1], P is used to describe the contribution of one tool under machining
a

T T
operation with a fixed life time T. ∫ 0
bM ' 2 (t )dt and ∫ 0
cM (t )dt represent the

operational cost and the overall holding cost during time interval [0, T ] , respectively.

4
Therefore, the mathematical model and its constraints [1] are listed as below.

∫ [bM ' ]
⎧ ⎧ M (T ) T ⎫
⎪Max ⎨P - 2
(t ) + cM (t ) dt ⎬
⎪ ⎩ a 0


⎪ s.t.
⎪⎪
⎨ M (0) = 0

⎪ M (T ) is free

⎪ 0 ≤ M ' (t ) ≤ B

⎪⎩ ∀ t ∈ [0, T ]


Let M * (t ) be the optimal solution of the model, and assume that time interval

(0, ~t ) is the maximal subinterval of [0,T] to satisfy Euler Equation [11, 12]. There are

two feasible cases to be discussed.


3.1 Case 1: M * (t ) will not touch B before T.

In this case, it is assumed that M ' (t ) will never reach the upper speed limit B

before tool life time T . From Euler Equation [11, 12], the transversality condition of

salvage value for free M (T ) [11, 12], and the boundary conditions, one can obtain the

optimal solution [1] for Case 1 as

′ c 1 P
M * (t ) = t + ( − cT ) (1)
2b 2b a

c 2 1 P
M * (t ) = t + ( − cT )t (2)
4b 2b a

Before solving the other case, one property is proposed [1, 2] and discussed as follows:

5

PROPERTY: If M * (t ) touches the upper MRR limit B , it will stay to be B from the

touch time ~
t to the end of the event time T.


Proof: From Eq.(1), M * (t ) is a strictly increasing linear function of t. And it holds

′ ′
for the subinterval of [0, ~
t ) subject to 0 ≤ M * (t ) ≤ B . Since M * (t )

cannot contradict Euler Equation [11, 12] to be a decreasing function of t,

the property is verified.


3.2 Case 2: M * (t ) will touch B before T.

In this case, we assume that M ' (t ) will reach the upper limit B at time ~
t . From the

transversality condition of salvage value for free end value M (t~ ) [11, 12], and the

PROPERTY, one will have

~ 1 P
t = T − ( − 2bB) (3)
c a

Therefore, the optimal solution [1] for Case 2 is shown as follows:

⎧c 1 P
t + ( − cT ) , if t ∈ [0, ~
t]
*′ ⎪ 2b
M (t ) = ⎨ 2b a (4)

⎩B , if t ∈ (~
t ,T ]

⎧c 2 1 P ~
*′ ⎪ 4b t + 2b ( a − cT )t , if t ∈ [0, t ]
M (t ) = ⎨ (5)
⎪ ~ ~
⎩M ( t ) + B(t − t ) , if t ∈ (~
t ,T ]

3.3 Decision Criteria

From Eq.(3), two decision criteria are classified.

P
1. When ≤ 2bB , it means that ~
t ≥ T . This contradicts the assumption of Case 2.
a

6
That is, the optimal control of MRR will not reach the upper speed limit within the

tool life time. The optimal solution is Case 1.

P
2. When > 2bB , it means that ~
t < T . That is, the optimal control of MRR will
a

reach the upper speed limit within the tool life time. The optimal solution is Case 2.

4. The Implementation

A schematic diagram of the experimental set-up is illustrated in Figure 1. The

overall experimental set-up consists of a commercialized lathe system, a DSP processor

controller, and a Man-machine Interface system. The characteristics of each component

are given as below:

4.1 The Digital Lathe System

The layout of the digital lathe system via PC-based control for realization of the

optimum MRR control through the Man-machine interface controller is illustrated in

Figure 1, in which the digital lathe system consists of four components: the user interface,

networking, simulation, and Lathe control scheme.

Pulse
PMC32-6000 Input
Motion
Control Encoder
Feed AC Servo Motor
Output
PC PCI
Bus
Frequency
Input
Frequency
Human Machine
Transformer Rotation
Interface
232/485 Rate
Spindle induced Motor

7
Figure 1: Layout of the digital lathe system.

The characteristics of each component are described as below:

1. The user interface provides programs with the following functions:

● Generation of the desired reference input and calculation of the feedback

information generated from the commanded input for the feed actuator servo

systems.

● Providing the parameter setting of spindle motor system based on the theoretical

development.

2. The networking provides communication protocol programs with the following

functions:

Modbus, the industrial communication network equipment, generally consists of

one master side and many slave sides maneuvered by a specific address. In the process

of communication operation, the master side delivers a serial command signal to the

slave side and waits to receive feedback signals from the slave side after the

subsequent command is provided. The slave side obtains the request signal for a start

to verify the assigned number with its address number, and the master side demands to

deliver the data signal with the assigned number correctly. In other words, if there is

error message in the communication process, the slave side does not deliver any

returned pass corresponding data to the master side and send a message error to check

the mistakes instead.

Generally, the Modbus has two parts: RTU and ASCII modules. RTU module

supports the binary data format and 16 bit unsigned integer and it is used to cope with

large data material by big-endian. The CRC (Cyclical Redundancy Check) is adopted

8
to examine error in communication format. In this paper, the ASCII module is used in

less quantity with character format communication and LRC (Longitudinal

Redundancy Check) in error examination.

3. The digital Lathe control scheme for maneuvering the cutting has the following

characteristics:

A servo motor with screw gears for driving the cutting tool base in the feed

directions, and the frequency transformer with induced motor manipulated for the

spindle rate via 232-485 converter by the PC-based are included in the proposed lathe

re-equip system.

The PMC32-6000 motion control card supports the pulse output. There is only

one output signal wired for each channel, and the card also supports quadrature

encoder which encodes the position of the spindle shaft of the motors for encoding,

pulse/direction counting or up/down counting.

4.2 The Man-Machine Interface

The PMC32-6000 DSP-Based motion control card was designed to provide a

powerful computation process. The code composer by TI is used to develop and debug

the DSP hardware, and the Labview by NI is selected as the man-machine interface. The

MATLAB is utilized to plot the experimental results. The DSP processor on the mother

board is a TMS320C31 with 1 Mbyte of SRAM. The mother board has eight independent

analog-to-digital converters (each with 12-bit resolution) with a programmable sampling

frequency up to 50 KHz. The unique architecture is designed so that at the instant that the

sampling interrupt occurred, all eight input values will be converted and can be read

simultaneously by all DSPs. The PMC32-6000 has six 12-bit independent analog output

9
ports in total. In this manner, the computation and communication power can be tailored

to individual applications.

A man-machine programming interface provides the operator to set machining

parameters of the turning process by spindle rate as well as the feedforward speed is

shown in Figure 2.

Figure 2: The man-machine programming interface

The BORLAND C++ BUILDER is utilized as the human-machine interface, and the

programmed interface is shown as Figure 3.

Figure 3: Window of the programmed human-machine interface

10
5. The Experiment

A numerical example of a single-tool finish-turning operation for specific S45C

steel fixture plates from AirTAC Corporation in Taiwan is referenced for the experiment.

All data compiled are transformed into SI units as well as US dollars and listed as

follows:

P = 5.45 dollars , a = 17355 mm 3 , b = 1.7 × 10 −8 (dollars - min)/mm 6 ,

c = 6.625 × 10 −8 dollars /(min − mm 3 ) , B = 16470 mm 3 / min , and T = 40 min .

The desk-top lathe control system is proposed. The feedrate is generated by the

servo-drive which controls the servo-motor, and the spindle speed is controlled by the

inverter.

The φ 25mm × 180mm S45C work-pieces are selected for the experiment, and the

length of turn for each work-piece is 150mm . The spindle speed ranges from 1328 rpm

to 1340 rpm , and the feed rate is selected as 2 mm sec .

6. Results and Discussion

The surface roughness of all finished work-pieces is measured at the front, middle,

and rear regions. The surface roughness of all thirty-three finished work-pieces under the

digital lathe system is plotted as Figure 4.

11
Figure 4: Plot of the measured surface roughness

It can be seen from Figure 4 that the surface roughness tends to increase as the tool

life is consumed. However, the growing rate of surface roughness decelerates and

stabilizes as the tool life consumed. Besides, it is also found that the machined

work-pieces are all within a surface roughness of 2.4 μ m which satisfactorily matches

the recognized standard for finish turning. With the experimental result, the adaptability

and applicability of the dynamic machining model are guaranteed.

7. Concluding Remarks

In recent years, a significant progress in CNC machine tools has been made as high

productivity and machine tool lift as important role factors in manufacturing processes.

The MRR is an important factor of machining operation, and the control of machining

speed is crucial for production planners. This paper not only introduces the idea of

dynamic machining optimization for an individual tool during a fixed life time, but also

12
contributes a reliable and applicable technique in approaching dynamic MRR control for

maximum profit.

A digital controlled PC-based lathe system is used to realize the dynamic MRR

control. It is shown that the machined parts are all within a surface roughness of 2.4 μm .

The experimental results demonstrate that the adaptability and applicability of the

dynamic machining model are promised.

Acknowledgement

This research was partially supported by the National Science Council in Taiwan through
Grant NSC 92-2213-E-244-004.

References

[1] Yet, L. J., Lan, T. S., The optimal control of material removal rate with fixed tool life

and speed limitation, Journal of Materials Processing Technology, 121 (2-3) (2002)

238-242.

[2] Lan, T. S., Lan, C. H., and Yet, L. J., Dynamic machining project control model under

order quantity and deadline constraints, Journal of the Operations Research Society

of Japan, 45 (1) (2002) 83-92.

[3] Wang, J., Luh, P. B., Scheduling of a machining center, Mathematical Comput.

Modelling, 23 (1996) 203-214.

[4] Fuh, K. H., Chen, C. T., Chang, Y. F., Design and implementation for maximum metal

removal-rate control of a constant turning-force system, Journal of Materials

Processing Technology, 57 (1996) 351-359.

[5] Rober, S. J., Shin, Y. C., Modeling and control of CNC machines using a PC-based

13
open architecture controller, Mechatronics, 5 (4) (1995) 401-420.

[6] Kim, T. Y., Kim, J., Adaptive cutting force control for a machining center by using

direct cutting force measurement, International Journal of Machine Tools and

Manufacture, 36 (8) (1996) 925-937.

[7] Masory, O., Real-time estimation of cutting process parameters in turning. ASME

Journal of Engineering for Industry 106 (1984)218-221.

[8] Daneshmend, L.K., Pak, H.A., Model reference adaptive control of feed force in

turning, ASME Journal of Dynamic Systems, Measurement, and Control 108 (1986)

215-222.

[9] Tomizuka, M., Zhang, S., Modeling and conventional/adaptive PI control of a lathe

cutting process, ASME Journal of Dynamic Systems. Measurement, and Control 110

(1988) 350-354.

[10] Toutant, R., Balakrishnan, S., Onyshko, S. and Popplewell, N., Feedrate

compensation for constant cutting force turning, IEEE Control Systems Magazine

13(6) (1999) 44-47.

[11] Chiang, A., Dynamic Optimization, McGraw-Hill Inc., New York, N.Y. (1992).

[12] Kamien, M. and Schwartz, N., Dynamic Optimization, New York, Elsevier (1991).

14

Vous aimerez peut-être aussi