Advantages

High productivity Although the cutting speeds and feeds for CNC machines are the same as for manually operated machines, much time is saved by rapid traversing and positioning between operations. Also a wider range of operations is possible on a CNC machine, avoiding the necessity to pass the work from one machine to another for, say, each of drilling, milling and boring. This reduces the need for expensive jigs and fixtures, and avoids reserves of work in progress between operations. In addition, CNC machines do not become tired and they maintain a constant rate of productivity. If the work is robot fed, they can work ‘lights out’ through the night.

Design flexibility Complex shapes are easily produced on CNC machines. In addition, contoured solids can be produced on CNC machines which cannot be produced on conventional, manually operated machines.

Management control Production rates, reduced scrap and improved quality comes under management control with CNC machines and is not influenced by operator performance.

Quality CNC machines have a higher accuracy and better repeatability than conventional machines. If the machine is fitted with adaptive control it will even sense tool failure or other variations in performance and either stop the machine or, if fitted with automatic tool changing, select backup tooling from the tool magazine before scrap is produced or the machine damaged.

Reduced lead time The lead time for CNC machines is much less than for other automatic machines. There are no complex and expensive cams and form tools; it is necessary simply to write a part program and load it into the machine memory. Complex profiles are generated using standard tooling.

Limitations

Capital cost The initial cost of CNC machines is substantially higher than for manually operated machines of the same type. However, in recent years, the cost differential has come down somewhat.

Tooling cost To exploit the production potential of CNC machine tools, specialized cutting tools are required. Although the initial cost is high, this largely reflects the cost of the tool shanks and tool holding devices which do not have to be replaced. The cost of replacement tool tips is no higher than for conventional machines.

Maintenance Due to the complexity of CNC machine tools, few small and medium companies will have the expertise to carry out more than very basic maintenance and repairs. Therefore, maintenance contracts are advisable. These are expensive, approximately 10% of the capital cost per annum.

Training Programmer/operator training is required. This is usually provided by the equipment manufacturer but can be time consuming and costly.

Depreciation As with all computer-based devices, CNC controllers rapidly become obsolescent. Therefore, CNC machines should be amortized over a shorter period than is usual with manually operated machines and should be replaced approximately every 5 years.

Open-loop control

This system derives its name from the fact that there is no feedback in the system, and hence no comparison between input and output. The system uses stepper motors to drive the positioning mechanism, and these have only limited torque compared with more conventional servo motors. Further, if the drag of the mechanism causes the motor to stall and miss a step, no corrective action is taken by the system.

Closed-loop control

The following figure explains closed-loop control system:

In a closed-loop control system the feedback continuously influences the action of the controller and corrects any positional errors. This system allows the use of servo motors to drive the positioning mechanism, and these have a much higher torque than the stepper motors used with the open-loop control system.

A character is a number, letter or other symbol which is recognized by the controller. An associated group of characters makes a word.

Word

A word is a group of characters which defines one complete element of information (e.g. N100). There are two types of word as follows.

Codes

A sample of a CNC program could look like this:

Block numbers

The block number is usually the first word which appears in any block. Blocks are numbered in steps of 5 or 10 so that additional blocks can be easily inserted in the event of an omission.

Preparatory functions (G)

These are used to inform the machine controller of the functions required for the next operation. Standardized preparatory functions are shown in the following table. In practice, the actual codes used will depend on the control system and the machine type. The codes that one system uses can vary from those of another, so reference to the relevant programming manual is essential.

Code number	Functiona
G00	Rapid positioning, point to point	(M)
G01	Positioning at controlled feed rate, normal dimensions	(M)
G02	Circular interpolation, normal dimensions	(M)
G03	Circular interpolation counter clockwise (CCW), normal dimensions	(M)
G04	Dwell for programmed duration
$\begin{array}{l}\text{G05}\\ \text{G06}\\ \text{G07}\end{array}\}$	Hold: cancelled by operator
	Reserved for future standardization: not
	normally used
G08	Programmed slide acceleration
G09	Programmed slide deceleration
G10	Linear interpolation, long dimensions	(M)
G11	Linear interpolation, short dimensions	(M)
G12	3D interpolation	(M)
G13-G16	Axis selection	(M)
G17	XY plane selection	(M)
G18	ZX plane selection	(M)
G19	YZ plane selection	(M)
G20	Circular interpolation clockwise (CW), long dimensions	(M)
G21	Circular interpolation CW, short dimensions	(M)
G22	Coupled motion positive
G23	Coupled motion negative
G24	Reserved for future standardization
G25-G29	Available for individual use
G30	Circular interpolation CCW, long dimensions	(M)
G31	Circular interpolation CCW, short dimensions	(M)
G32	Reserved for future standardization
G33	Thread cutting, constant lead	(M)
G34	Thread cutting, increasing lead	(M)
G35	Thread cutting, decreasing lead	(M)
G36-G39	Available for individual use
G40	Cutter compensation, cancel	(M)
G41	Cutter compensation, left	(M)
G42	Cutter compensation, right	(M)
G43	Cutter compensation, positive
G44	Cutter compensation, negative
G45	Cutter compensation +/+
G46	Cutter compensation +/–
G47	Cutter compensation –/–
G48	Cutter compensation –/+
G49	Cutter compensation 0/+
G50	Cutter compensation 0/-
G51	Cutter compensation +/0
G52	Cutter compensation –/0
G53	Linear shift cancel	(M)
G54	Linear shift X	(M)
G55	Linear shift Y	(M)
G56	Linear shift Z	(M)
G57	Linear shift XY	(M)
G58	Linear shift XZ	(M)
G59	Linear shift YZ	(M)
G60	Positioning exact 1	(M)
G61	Positioning exact 2	(M)
G62	Positioning fast	(M)
G63	Tapping
G64	Change of rate
G65-G79	Reserved for future standardization
G80	Fixed cycle cancel	(M)
G81-G89	Fixed cycles	(M)
G90-G99b	Reserved for future standardization

(M)indicates modal G commands which remain in force from line to line until cancelled.

^a Variations on standard codings occur not only between the different makes of controller but between different models by the same maker and even between different types of the same model. Thus again it must be stressed that the programmer/operator must work from the manufacturer’s manual for a given machine/controller combination.

^b Most control units use G90 to establish the program in absolute dimensional units, and G91 to establish the program in incremental dimensional units, however. FANUC control units use G20 in place of G90, and G21 in place of G92.

Dimensional words

A CNC machine tool will have axes which are addressed by letters (X, Y, Z, etc.). The CNC program will instruct the controller to drive the appropriate machine elements to a required position parallel to the relevant axes using dimensional words (X25.0 Y55.0) which consists of the axis letter plus a dimension. These dimensional words are sign sensitive (— or +). If there is no sign, the number is assumed positive by the controller. This adds built-in safety to a CNC program, ensuring that should a CNC programmer fail to input a minus value to a Z-axis command the tool should move to ’safe’ as shown below. That is, it moves away from the workpiece, as shown in the following figure.

Consider a line of program N10 G01 X25.0 Y25.0 Z30.0. This would result in the tool moving to the coordinates X25, Y25, Z30. However, since the Z word is positive the tool will go up and not down 30mm from the datum. To drill the component to a depth of 30 mm the line of program should read N10 G01 X25.0 Y25.0 Z-30.0.

Feed rate (F)

Words commencing with F indicate to the controller the desired feed rate for machining. There are a number of ways of defining the feed rate:

(a) Millimetres/minute: F30 = 30mm/min.

(b) Millimetres/revolution: F0.2 = 0.2 mm/rev.

(c) Feed rate number 1–20 indicates feed rates predetermined by the manufacture in rev/min or mm/min as appropriate. To use this system the feed rate command is prefaced by a G code. Only a few older machines use this system.

Spindle speed (S)

Words commencing with the letter S indicate to the controller the desired spindle speed for machining. Once again there are a number of ways of defining the spindle speed, each selected by use of the appropriate G code:

(a) Spindle speed in revolutions/minute (r.p.m.).

(b) Cutting speed in metres/minute.

(c) Constant cutting speed in metres/minute.

Tool numbers (T)

Each tool used during the machining of a part will have its own number (e.g. T01 for tool 1, T19 for tool 19). Most CNC machine controllers can hold information about 20 tools or more. This tool information will take the form of a tool length offset (TLO) and a tool diameter/radius compensation (see Sections 8.1.9 and 8.1.10). Where the machine is fitted with automatic tool changing it will also identify the position of the tool in the tool magazine.

Miscellaneous functions (M)

Apart from preparatory functions there are a number of other functions that are required throughout the program (e.g. ’switch spindle on’, ’switch coolant on’, etc). These functions have been standardized and are popularly known as M codes. The address letter M is followed by two digits. The most commonly used M codes are as follows:

G41 Compensation to the left

The cutter is always to the left of the surface being machined when looking along the path of the cutter, as indicated by the arrows in the figure.

G42 Compensation to the right

The cutter is always to the right of the surface being machined when looking along the path of the cutter, as indicated by the arrows in the figure.

G40 Turning compensation on and off

Whenever G41 or G42 is activated, the cutter diameter is compensated during the next move in the X and Y axes, as shown. The preliminary movement of the cutter before cutting commences is called ramping on.

Similarly G40 causes the compensation to be cancelled on the next X and Y movements as shown, and is known as ramping off.

Ramping on and ramping off allow the feed servo motor to accelerate to the required feed rate, and to decelerate at the end of the cut whilst the cutter is clear of the work. It therefore follows that diameter compensation can never be applied or cancelled when the cutting tool is in direct contact with the workpiece.

To save repetitive programming on similar operations, CNC controllers offer a range of canned or fixed cycles. It is only necessary to provide dimensional data with the required G code. Some examples of canned cycles for milling are as follows:

G81 Drilling cycle

One of the most popular canned cycles is the drilling cycle G81. The machine movements involved in drilling a series of holes are shown and are as follows:

(a) Rapid to centre of first hole.

(b) Rapid to clearance plane height.

(c) Feed to depth of hole.

(d) Rapid up to clearance plane height.

(e) Rapid to centre of next hole.

(f) Repeat for as many holes as required.

The only data which needs to be programmed is:

Tool number

Hole position (X, Y)

Hole depth (Z)

Spindle speed (S)

Feed rate (F).

Material: aluminium alloy BS 1474: 1987 type 5154(A) condition 0.

Machining data

The tool change position will be X — 100.0, Y — 100.0. The appropriate M code in the program for tool changing will move the tool to these coordinates and retract the tool up the Z-axis to a ‘home’ position clear of the work to facilitate manual tool change.

Centre line path/point programming will be used, and to avoid complexity no diameter compensation will be shown in this example.

Dimensional data

The dimensions shown provide the X, Y, Z data for the 10mm wide and 8mm deep slot, and the 8 mm diameter and 12 mm deep holes. The numbers in the circles refer to the tool positions as described in the specimen program to follow. The Z datum (Z0) would be the top face of the workpiece.

Specimen program

The format of the following program example would be suitable for the Bridgeport BOSS 6 software. When evaluating the program, reference should be made to the coded information given in Section 8.1.7.

G41 Compensation to the left

G42 Compensation to the right

Examples of some of the canned cycles available when using CNC centre lathe controllers are as follows:

G68 Roughing cycle

This cycle is used for large amounts of stock removal. The cycle is activated by a G68 code. The machine moves are shown and are as follows:

(a) Rapid move to start point.

(b) A number of linear roughing passes (the number varies with depth of cut).

(c) A profiling pass, leaving on a finish allowance.

After the roughing cycle is completed the component would be finish turned to size.

G83 Deep hole drilling cycle

The deep hole drilling cycle is used to drill holes with a high depth/diameter ratio, and provides ‘flute clearance’ during the cycle. The cycle is activated by a G83 code. The moves involved are shown and are as follows:

(a) Rapid move to start point.

(b) Feed to first peck depth.

(c) Rapid move out to start point.

(d) Rapid in just short of first peck.

(e) Feed to second peck depth.

(f) Repeat movements (c)-(e) for each successive peck depth until required depth is achieved.

Component

Material: aluminium alloy BS 1474: 1987 type 5154(A) condition 0.

Tool and machining data

Specimen program

(a) The format of the program would be suitable for a GE 1050 CNC controller.

(b) Diameter programming is used in this example.

(c) Tool change position (home position) is X177.8 Z254.

Since the first edition of this pocket book was published, the power of desktop computers has increased out of all recognition and it is now possible to run full-scale CAD packages on such machines.

Centralized processing

Here, one computer (normally a mainframe) supports a number of terminals. Response times can be slower than for ’stand-alone’ systems, particularly if the system is heavily loaded. A schematic of a centralized system is shown below.

Shared centralized processing

This configuration is more costly than centralized processing since more than one central computer is used at the same time. It has two major advantages. Firstly the response time is improved and, secondly, if one computer fails the system is not totally disabled. A schematic of a shared centralized system is shown below.

Integrated system

With this configuration, a number of computers are networked together so that the system files can be shared. This arrangement also allows the plotter, printers and disk drives required to become a shared resource, resulting in more economical loading of the system, faster response time and less chance of total disablement. The initial capital outlay is lower, as the system can be started up with only a few workstations. Additional workstations can then be added to the network as and when required. A schematic of an integrated system is shown below.

Visual display units

Nowadays these have to be colour monitors since modern CAD packages have ‘layering’ facilities. For example, an architect can produce the ground floor plan of a building in one colour, and superimpose the first floor plan in another colour to ensure structural integrity. The plumbing and central heating layout can then be superimposed in another colour an the electrical installation in yet another colour to make sure that everything is in the correct place. The layers can then be printed individually or in combination as required.

When producing orthographical engineering drawings it is useful to use one colour layer for the construction lines an another colour layer for the final outline.

When producing PCB layouts, one colour can be used for the tracks on the top of the board and another colour layer for the tracks on the underside of the board.

Colour presentations are also used when producing publicity material, charts and graphs for inclusion in reports, also for overhead projector transparencies.

Input devices

Keyboards These are familiar to all computer operators and are user friendly. These are used to input commands, text and parameter values.

Digitizer tablet The data tablet detects the position of a mechanical scanning or handheld cursor. The tablet contains an electromagnetic, electrostatic or electroresistive grid which corresponds to the resolution of the visual display unit (VDU). Thus, X and Y coordinates on the tablet can be mapped directly on to the screen.

Joystick This is a two-axes control for providing a rapid movement of the cursor. It is familiar to players of video games. When used with CAD systems it allows more rapid cursor movement than is possible from the keyboard. However, the positional accuracy is low and is dependent on operator skill.

Mouse The mouse is now widely used in place of the joystick and is essential when using Windows versions of CAD software. The ‘click’ keys on the mouse allow the command to be entered quickly and easily once the cursor has been positioned.

Output devices

Hard copy can be printed from digitized data in a number of ways.

Flat bed plotter The paper is held to the flat bed of the machine by suction. The pen moves along the X- and Y-axis to draw the image. The carriage moves along the X-axis and the penhead moves along the Y-axis bridge which is located on the carriage. There is also a ‘home’ position to which the penhead is parked at the start and finish of the drawing. This allows the paper to be inserted and the finished drawing removed without fouling the pen. The penhead also returns to the ‘home’ position when the pen has to be changed to accommodate a different line thickness and/or colour. The pens are located in a magazine and are selected on command from the computer.

Drum or roller bed plotter The bed rotates to provide the X-axis movement and the pen carriage moves along the Y-axis parallel to the drum axis of rotation. This arrangement allows large drawings to be produced, yet little floor space is required. These plotters are not quite so accurate as flat bed plotters due to paper stretch and slip. They are also more limited in the range of materials that they will accept.
For small-scale drawings, conventional dot matrix, laser jet and electrostatic printers as used for normal text hard copy printing are suitable.

• There are substantial improvements in productivity during the design and draughting process.

• Drawings can be produced to very high levels of accuracy and consistent house styles.

• Product design can be improved.

• It is easy to make modifications without completely redrawing.

• The storing of digitized data on disk is simple and storage space is reduced compared with tracings.

• Colour coded hard copy can be produced whereas conventional dyeline prints are monochrome.

• Digitized data can be used for: costing and estimating, purchasing, production planning and control, control of labour costs, quality control.

• Digitized CAD data can be integrated with CNC machines (CAD/CAM) and also with robots to provide flexible manufacturing systems (FMS).

Limitations

• Hardware and software costs for CAD can be very high. Although the cost of individual components may be reduced, the continual increase in the complexity and sophistication of the software demands greater and greater memory capacity and greater power and speed from the CPU. This overtakes any cost reductions in individual components.

• Powerful CAD systems require regular maintenance, although they are becoming very much more reliable and maintenance contracts are available.

• Initial set-up and staff training can be expensive whilst libraries of conventions an symbols are prepared and digitized to suit company products.

• A suitable ergonomic environment is required to avoid operator fatigue and ensure economic use of the system.

Cartesian coordinate robots

Cartesian coordinate robots consist of three orthogonal linear sliding axes, the manipulator hardware and the interpolator. The control algorithms are similar to those of CNC machine tools. Therefore, the arm resolution and accuracy will also be of the order of magnitude of machine tool resolution.

An important feature of a cartesian robot is its equal and constant spatial resolution; that is, the resolution is fixed in all axes of motions and throughout the work volume of the robot arm. This is not the case with other coordinate systems.

Cylindrical coordinate robots

Cylindrical coordinate robots consist of a horizontal arm mounted on a vertical column, which in turn is mounted on a rotary base. The horizontal arm moves in and out, the carriage moves vertically up and down on the column, and these two units rotate as a single assembly on the base. The working volume is therefore the annular space of a cylinder.

The resolution of the cylindrical robot is not constant and depends on the distance between the column and the wrist along the horizontal arm. Given the standard resolution of an incremental digital encoder on the rotary axis and arm length of only 1 m, then the resolution at the hand at full arm extension will be of the order of 3 mm. This is two orders of magnitude larger than is regarded as the state of the art in machine tools (0.01 mm). This is one of the drawbacks of cylindrical robots as compared to those with cartesian frames. Cylindrical geometry robots do offer the advantage of higher speed at the end of the arm provided by the rotary axis, but this is often limited by the moment of inertia of the robot arm.

In robots which contain a rotary base, good dynamic performance is difficult to achieve. The moment of inertia reflected at the base drive depends not only on the weight of the object being carried but also on the distance between the base shaft and the manipulated object. This is regarded as one of the main drawbacks of robots containing revolute joints.

Spherical coordinate robots

The kinematic configuration of spherical coordinate robot arms is similar to the turret of a tank. These arms consist of a rotary base, an elevated pivot and a telescopic arm which moves in and out. The magnitude of rotation is usually measured by incremental encoders mounted on the rotary axes. The working envelope is a thick spherical shell.

The disadvantage of spherical robots compared with their cartesian counterparts is that there are two axes having a low resolution, which varies with the arm length.

The main advantage of spherical robots over the cartesian and cylindrical ones is a better mechanical flexibility. The pivot axis in the vertical plane permits convenient access to the base or under-the-base level. In addition, motions with rotary axes are much faster than those along linear axes.

Revolute coordinate robots

The revolute, or articulated, robot consists of three rigid members connected by two rotary joints and mounted on a rotary base. It closely resembles the human arm. The end effector is analogous to the hand, which attaches to the forearm via the wrist. The elbow joint connects the forearm and the upper arm, and the shoulder joint connects the upper arm to the base. Sometimes a rotary motion in the horizontal plane is also provided at the shoulder joint.

Since the revolute robot has three rotary axes it has a relatively low resolution, which depends entirely on the arm length. Its accuracy is also the poorest since the articulated structure accumulates the joint errors at the end of the arm. The advantage of such a structure and configuration is that it can move at high speeds and has excellent mechanical flexibility, which has made it the most popular medium-sized robot.

The wrist

The end effector is connected to the mainframe of the robot through the wrist. The wrist includes three rotary axes, denoted by roll, pitch and yaw. The roll (or twist) is a rotation in a plane perpendicular to the end of the arm; pitch (or bend) is a rotation in a vertical plane and yaw is a rotation in the horizontal plane through the arm. However, there are applications which require only two axes of motion in the wrist.

In order to reduce weight at the wrist, the wrist drives are sometimes located at the base of the robot, and the motion is transferred with rigid links or chains. Reduction of weight at the wrist increases the maximum allowable load and reduces the moment of inertia, which in turn improves the dynamic performance of the robot arm. The pitch, roll and yaw movements of the wrist can be seen on the robot shown.