[2] If variables such as cutter geometry and the rigidity of the machine tool and its tooling setup could be ideally maximized (and reduced to negligible constants), then only a lack of power (that is, kilowatts or horsepower) available to the spindle would prevent the use of the maximum possible speeds and feeds for any given workpiece material and cutter material.
Outside of the context of machine tooling, "speeds and feeds" can be used colloquially to refer to the technical details of a product or process.
The most common materials are available in reference books or charts, but will always be subject to adjustment depending on the cutting conditions.
The conditions are a tool life of 1 hour, dry cutting (no coolant), and at medium feeds, so they may appear to be incorrect depending on circumstances.
These cutting speeds may change if, for instance, adequate coolant is available or an improved grade of HSS is used (such as one that includes [cobalt]).
The American Iron and Steel Institute (AISI) determined machinability ratings for a wide variety of materials by running turning tests at 180 surface feet per minute (sfpm).
The machinability rating is determined by measuring the weighed averages of the normal cutting speed, surface finish, and tool life for each material.
Mechanical arrangements to effect CSS have existed for centuries, but they were never applied commonly to machine tool control.
The introduction of CNC-controlled lathes has provided a practical, everyday solution via automated CSS Machining Process Monitoring and Control.
By means of the machine's software and variable speed electric motors, the lathe can increase the RPM of the spindle as the cutter gets closer to the center of the part.
The required feed rate can be extremely variable depending on the power of the motor, the hardness of the wood or other material being machined, and the sharpness of the cutting tool.
In woodworking, the ideal feed rate is one that is slow enough not to bog down the motor, yet fast enough to avoid burning the material.
The right feed rate is usually obtained by "feel" if the material is hand fed, or by trial and error if a power feeder is used.
A slower feed rate usually results in a finer surface as more cuts are made for any length of wood.
Stainless steel is one material that hardens very easily under cold working, therefore insufficient feed rate or incorrect spindle speed can lead to less than ideal cutting conditions as the work piece will quickly harden and resist the tool's cutting action.
e.g. for a cutting speed of 100 ft/min (a plain HSS steel cutter on mild steel) and diameter of 10 inches (the cutter or the work piece) and, for an example using metric values, where the cutting speed is 30 m/min and a diameter of 10 mm (0.01 m), However, for more accurate calculations, and at the expense of simplicity, this formula can be used: and using the same example and using the same example as above where: Feed rate is the velocity at which the cutter is fed, that is, advanced against the workpiece.
The ratio of the spindle speed and the feed rate controls how aggressive the cut is, and the nature of the swarf formed.
Speed-and-feed selection is analogous to other examples of applied science, such as meteorology or pharmacology, in that the theoretical modeling is necessary and useful but can never fully predict the reality of specific cases because of the massively multivariate environment.
Just as weather forecasts or drug dosages can be modeled with fair accuracy, but never with complete certainty, machinists can predict with charts and formulas the approximate speed and feed values that will work best on a particular job, but cannot know the exact optimal values until running the job.
In CNC machining, usually the programmer programs speeds and feedrates that are as maximally tuned as calculations and general guidelines can supply.
The operator then fine-tunes the values while running the machine, based on sights, sounds, smells, temperatures, tolerance holding, and tool tip lifespan.
Under proper management, the revised values are captured for future use, so that when a program is run again later, this work need not be duplicated.
For example, an effort called the Machine Tool Genome Project is working toward providing the computer modeling (simulation) needed to predict optimal speed-and-feed combinations for particular setups in any internet-connected shop with less local experimentation and testing.
The work is typically done in engineering laboratories, with the funding coming from three basic roots: corporations, governments (including their militaries), and universities.
The recommended speeds and feeds provided in this book were the result of extensive testing to determine optimum tool life under controlled conditions for every material of the day, operation and hardness.