Question 6 of 7: Numerical Control of Machine Tools

Question 6: Numerical Control of Machine Tools (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Part (a) — the advantages and disadvantages of numerical control

Numerical control replaces the machinist's continuous manual guidance of the tool with a stored program of coded instructions that a machine control unit executes through servo-driven axes. Everything that follows — good and bad — is a consequence of that single substitution: the skill and the judgement move off the machine and into the program, the fixture and the tooling.

The advantages begin with flexibility with automation. A conventional automatic machine achieves its speed through hard tooling that must be rebuilt for each new part; an NC machine achieves it through a program that can be changed in minutes, so short runs and frequent changeovers become economic. This is the central argument, and the others follow from it.

Reduced non-productive time. The saving is not principally in cutting; feeds and speeds are limited by the same tool life on either kind of machine. It is in positioning, in setup, in tool changes on a machining centre with an automatic tool changer, in fewer setups because several faces are machined in one clamping, and in the elimination of layout, marking-out and trial cuts. Non-productive time typically dominates the cycle on short runs, which is precisely where NC pays.

Greater accuracy and repeatability. The positional accuracy of the machine does not decay with operator fatigue or attention, so dimensions repeat piece to piece and lot to lot. Scrap falls, inspection can be reduced to a sampling basis once the process is qualified, and interchangeability improves.

Complex geometry becomes routine. Contours, pockets, angled holes and sculpted surfaces that are impractical or impossible to machine manually are simply another set of coordinates to the control. Aerospace structural parts and mould and die cavities exist in their present form because of this.

Shorter lead times and lower work-in-process. Fewer setups and fewer operations mean fewer queues; parts spend less time in the shop and less capital is tied up between operations. Reduced tooling cost, because programmed motion replaces jigs, templates and drill bushings, and a simple locating fixture usually suffices. Reduced fixture and part storage. Easier design changes, since a dimension change is a program edit rather than a new jig. Better process planning data, because the program itself is an exact record of the method and the cycle time is known before the first part is cut. And less operator skill required at the machine, which matters where skilled machinists are scarce.

The disadvantages are equally real. The capital cost is high — typically several times that of the equivalent manual machine — so the machine must be kept loaded, which means multi-shift operation and careful scheduling; an idle NC machine loses money far faster than an idle manual one. Higher operating and maintenance cost, with specialised electronic and servo maintenance skills that a small shop may not possess, and a slow, expensive repair when the control fails. The programming function must be created and staffed: part programmers, post-processors, CAM software, program proving and program storage and revision control are a new overhead that did not exist before, and errors in a program can wreck a part, a tool or the machine itself. New planning discipline: tooling must be pre-set and identified, material must be to size, and the shop must be able to deliver the right fixture and cutter at the right time, because the machine cannot improvise. Higher utilisation demanded of the supporting departments, and a certain de-skilling and consequent industrial relations sensitivity, since the craft content of the operator's job is reduced. Finally, NC is not economic on very long runs, where dedicated special-purpose or transfer machinery has a lower cost per piece, nor on the very simplest one-off jobs, where the programming effort exceeds the machining effort.

Part (b) — the jobs for which numerical control is most appropriate

The characteristics that make a metal-machining job a good candidate for NC are the mirror image of the advantages just listed. Taken together they describe the mid-volume, high-variety, high-complexity region of the manufacturing spectrum.

• Small to medium lot sizes, produced repeatedly. Roughly speaking, lots of a few to a few hundred pieces, with the part recurring on a regular cycle so the program is written once and used many times. Very large runs belong on dedicated equipment; true one-offs may not repay the programming.
• Frequent changeovers and a wide part mix. The benefit of NC is realised in the changeover, so a shop that changes what it is making often gains the most.
• Complex part geometry. Contoured or sculptured surfaces, many holes in a precise pattern, pockets, angular features, or profiles that would require a tracer, a rotary table or an elaborate jig on a conventional machine.
• Many separate operations on one part. Parts requiring drilling, boring, milling, tapping and counterboring benefit disproportionately, because a machining centre with an automatic tool changer collapses several conventional setups into one.
• Close tolerances, and dimensions that must repeat. Where accuracy and consistency carry value — interchangeable assembly, tight fits, mating parts — NC removes the operator variability that otherwise dominates.
• Expensive workpiece material or high value added. Titanium, nickel alloy and large aluminium billets punish scrap heavily, and a proven program is a far more reliable protection against scrap than a skilled operator on a long, tedious cycle.
• Frequent engineering changes. Where the design is still moving, editing a program is trivial compared with modifying hard tooling.
• Long, tedious or fatiguing manual cycles where operator attention would otherwise be the limiting factor, and parts requiring many machined surfaces relative to the raw stock removed.
• Short lead times demanded, since NC eliminates the jig and fixture manufacturing lead time that conventional tooling requires.

The classic industrial examples are aerospace structural components, mould and die work, hydraulic and pneumatic valve bodies, machine-tool components, prototype and short-run production of all kinds, and any family of similar parts that can be grouped by group technology so that one fixture and one program family serves many part numbers. The counter-examples are equally instructive: a plain turned shaft made a million times a year belongs on a multi-spindle automatic, and a single bracket needing two drilled holes belongs on a drill press.