Although many people talk about high-speed machining, many machine tools are advertised as high speed and there are conferences devoted to HSM, the term is not well defined. Some definitions focus on surface speeds, others on spindle bearing rotation and still others on spindle speed along. However, a better approach may be to define HSM based on stability phenomenon.

One common definition is based on the surface speed—the relative speed between the tool and workpiece. This is essentially a thermal limit because many of the wear mechanisms are temperature dependent.

The surface speed definition is favored by toolmakers and metalcutting researchers. In a classic paper from the Annals of the CIRP from 1992, H. Schulz and T. Moriwaki produced a figure similar to Table 1, indicating the ranges for conventional machining (green), a transitional region (yellow) and a high-speed region (red) based on surface speed. Surface speed in m/min. is determined by v=πdn, where d is the diameter of the milling tool or workpiece, and n is the spindle speed. Conventional metric units for cutting speed are m/min. For example, a 25mm-dia. milling tool rotating at 15,000 rpm has a surface speed of 1,177.5 m/min.

A second common definition uses the DN number, which is based on the rotation capability of the preloaded spindle bearings. Spindle builders and machine tool designers usually prefer the DN number as an expression of high speed. The DN number is the product of D, the diameter in mm of the main bearing bore, and N, the maximum spindle speed in rpm. For example, a spindle with a 60mm main bearing bore and a 20,000-rpm maximum spindle speed would have a DN number of 1.2 million.

Although laboratory spindles have achieved much higher numbers, commercial high-speed spindles have remained below 2 million DN for a number of years. One of the main reasons is heat in the bearings. The bearings are preloaded (the bearing is arranged so the balls are always in compression, because that increases the bearing stiffness and accuracy). The preload mechanism must fight against the centrifugal forces caused by the balls circulating at high speed. Because ceramic balls are stiff and have a low mass, they are often preferred to steel ones. As the balls rotate in the race of the bearing, they are squeezed and released, and that work shows up as heat.

There is also heat from the rotor in an integral spindle and from the drivetrain in a nonintegral spindle. Therefore, the preload mechanism must allow for thermal growth of the spindle while maintaining preload—a significant design challenge.

Toolholder makers often look at spindle speed alone when defining HSM. Toolholders and cutting tools are rated to be safe up to a certain spindle speed, meaning the tool and holder bodies will stay together, and all of the mechanical connections will remain intact. Such a rating usually means the tool or holder has already been spun up to twice the rated spindle speed.

The definition I prefer is related to the appearance of stability phenomenon. The "stability lobes" appear when the tooth passing a frequency reaches a significant fraction of the dominant natural frequency. In low-speed machining, the selection of the spindle speed does not significantly change the dynamic performance of the machine. The stability limit, the DOC above which chatter occurs, seems like a constant. When the tooth passing frequency reaches a substantial fraction, say ¼, 1⁄3, ½ or 1⁄1, then the selection of a correct speed can easily improve the metal-removal rate by a factor of two or more.

For example, when milling aluminum with a 2-flute tool and a dominant (most flexible) natural frequency of 1,200 Hz, the stability lobe effects are strongest at 1,200 × 60 ÷ 2 = 36,000 rpm. Other significant stable zones would occur at 18,000 rpm, 12,000 rpm and 9,000 rpm.

For tool systems with lower dominant natural frequencies, the spindle speeds where these effects appear can be much lower. When titanium milling with a 6-flute tool and a dominant natural frequency of 250 Hz, the most stable pocket would appear at 250 × 60 ÷ 6 = 2,500 rpm and still be HSM from the dynamics point of view. CTE

About the Author: Dr. Scott Smith is a professor at the William States Lee College of Engineering, University of North Carolina at Charlotte, specializing in machine tool structural dynamics. Contact him via e-mail at [email protected].