4-Axis vs. 5-Axis Milling Machines: The Numbers Game

David Leeson
article by David Leeson

As with so many fabrication industries, the temporal and economic advantages afforded by digital CAD/CAM technologies have revolutionized manufacturing in dental laboratories. Increasing hardware and software sophistication joins decreasing upgrade costs in a capitalistic one-two punch that continues to supplant cast-and-burnout production methods of prosthetics and components.

To compete for business from dentists accustomed to short in-lab working times and minimized costs, laboratory decision makers may be tempted to purchase the latest in cutting-edge technologies under the assumption that better equipment will translate to larger profit margins. However, when it comes to milling machines, the bigger-is-better axiom is anything but certain. The manufacturing freedom granted by that extra axis comes at significant cost beyond the one-time purchasing fee, and any decision to acquire such a machine should be preceded by a thorough cost-benefit analysis.

Any machine created to mill in a three-dimensional space must start with the three core linear axes: a horizontal axis (x), a depth axis (y), and a vertical axis (z). Facing a vhf CAM 4-K4 Impression Milling Machine (vhf camfacture AG; Ammerbuch, Germany), one can see this principle in action: The milling tool moves left-to-right (x-axis) and up-and-down (z-axis), while the machine’s disc clamp moves forward-and-backward (y-axis). From this simple configuration, the tip of the tool can reach any point within the work cube.

Of course, it’s safe to presume no lab in history has been asked to deliver a restoration in the shape of a perfect cube. For the tip of the tool to reach a point on the surface of a convex shape below the height of contour, rotational axes that position the angle of the milling block are necessary. A fourth tilt axis (a-axis) is added in the case of the vhf K Series, around which the milling disc rotates (Fig. 1). With this 3+1 arrangement of axes, most shapes encountered in restorative dentistry (crowns, bridges and so forth) can be created by milling everything above the conceptual "parting line," then flipping the disc along the a-axis and milling everything below the same line.

But to really break the bonds of logistics and experience unchained milling autonomy, a fifth rotary axis (b-axis) is prescribed: a second rotational axis perpendicular to the a-axis (Fig. 2). With the tilt and rotary axes working in unison, any facet of the milling component can be oriented upward; and with the sudden ability to orient any facet toward the milling tool while the x-, y-, and z-axes control spatial relation, no geometry, no matter how intricate, is beyond the machine’s capability to mill. In addition, while a 4-axis apparatus performs "indexed milling" — that is, incrementally tilts-and-pauses along the a-axis while the tool lifts-and-places — a 5-axis mill provides the opportunity for "continuous milling." As in the case of the vhf CAM 5-S1 Impression Milling Machine (vhf camfacture AG), by altogether removing the disc clamp in exchange for the rotary capabilities of the fifth axis, the a-axis can be indexed while the b-axis continues to spin at high speeds. This allows the tool to remain in constant contact with the milling component while the rotary axis does the work of moving points along the component’s 3-D surface to the tool’s sculpting tip.

At first glance, the unlimited shape creation and apparent increase in speed efficiency might seem to make the purchase of the 5-axis milling machine over the 4-axis an obvious choice. True, the cost bump is significant: a vhf K Series runs just under $30,000, while the extra axis in the S Series pushes the price to $50,000. But how much money does a dental laboratory stand to lose by outsourcing — or worse, turning away — cases that require unrestricted milling capabilities? Won’t the additional business rapidly pay for the extra cost of the machine?

To answer these questions, the obvious starting point is to determine what the extra monetary and logistical costs really are. Unfortunately, cost-boosting disadvantages follow closely behind the advantages of that fifth axis. For example, the swiftness of its "continuous milling" capabilities is often hampered by the inescapably complex nature of organic shapes. In the industrial world, many of the shapes requiring fabrication can be made simply by running the 2-D shape of the tool surface across a 3-D path. Like in the case of a centrifugal jet engine impellor: The path of a milling tool can be matched to the form of an impellor blade, and the final shape of the blade made with a single movement through the metal merely by adjusting the orientation of the tool continuously throughout. This is because the shape, although complex, was designed by the human mind with an eye toward machining ease. Mother Nature, on the other hand, does not engineer with the same concern for tool paths. When mimicking organic shapes – say, those in a 3-unit bridge – the forms cannot be sculpted by varying the relative angle of the tool in a single pass. Rather, the buccolingual surface must be generated using a surface machining technique, in which a shape is created through multiple passes of the tool’s spherical tip. Any arbitrary shape, within the limits of the tip radius, can be created using this method; however, in order to create a smooth surface, the tool must make passes at increments as small as 50 microns (Figs. 3a, 3b). This is a time-consuming reality, whether machining with five axes or three; and one which, provided the component is free from undercuts, nullifies any speed advantage afforded by extra axes.

In addition to the organic shapes in dentistry, engineering limitations also hinder the perceived speed increase of a 5-axis machine. Like a chain’s weakest link, a milling machine is only as fast as its slowest moving part. In a Haas DT-1 equipped with a T5C 5-axis rotary table (Haas Automation; Oxnard, Calif.), for example, the x- and y-axis pack 2,550 lbs. of thrust each; the z-axis, 4,200 lbs. This provides linear motion up to 2,400 ipm at an acceleration rate of more than one g, allowing the tool and component to move very fast through the work cube. Conversely, physical limitations in the rotary gear box mechanism provide the a- and b-axis with only 210 ft-lbs. and 60 ft-lbs. of torque, respectively, resulting in a significantly reduced pace the machine must keep when the two additional axes are in motion.

With added degrees of freedom comes added maintenance as well, particularly in the form of sensitivity to inaccurate calibration. Each axis in a milling machine is accurate to within a certain distance range; and while there is some automation to ensure this range remains minimal, most dental milling machine providers limit this automated calibration to finding the axes’ center of rotation. This is only one facet of the machine’s complex mechanism, and does not truly test or adjust the positioning accuracy thereof.

To sustain the level of precision we demand at Glidewell, laser interferometers are required that can guarantee positioning inaccuracies remain inconceivably small — as tiny as +/-2 microns per linear axis. Still, inaccuracies are additive in nature; and since the milling tool mills at the vertex where multiple axes intersect, a rise in the number of axes necessarily precipitates a rise in the range of possible inaccuracy. If, for instance, the maximum 2-micron inaccuracy was made across each of the three linear axes, the tool would mill at a point 3.46 microns off the intended one. Similarly, accuracy ranges in a rotary table also compound and are typically the largest component of a positioning error, because the numbers quickly increase depending on the distance the programmed point lies from the center of rotation. Each of the previously mentioned Haas T5C’s additional axes provides a potential increased range of inaccuracy that builds upon the standard 2-micron inaccuracy of each linear axis. Vigilance in calibration becomes an imperative with more-complex milling machines.

Another financial pitfall lies in the complexity of the software needed to drive the machine. Although extremely rare, more moving parts increase the likelihood and severity of a collision. To combat this potential outcome, advanced CAM software and post processors sophisticated enough to keep all the parts moving in sync are required. But while these tools are adept at avoiding catastrophe, they are ineffective at avoiding sticker shock. Software driving a 5-axis milling machine ranges from $7,000–$20,000; and with annual maintenance packages billed at a percentage of the initial purchase amount (the standard fee structure, no matter how many axes), the costs quickly multiply beyond the already augmented purchase price.

Of course, while the extra costs and complexities of 5-axis machining are daunting, the astute laboratory decision maker might dismiss them with a sigh of resignation. After all, these must all be chalked up to the cost of doing business in an industry where five axes are a prerequisite to staying competitive. Right?

Therein lies the silver lining. The vast majority of dental restorations can be milled with a 4-axis milling machine (Fig. 4). While five axes are required for shapes of excessive complexity (e.g., mesh frames for partial dentures or screw-retained bars and bridges where the screw holes are arranged at disparate angles parallel to the a-axis tilt), most shapes in restorative dentistry consist of surface points easily accessible by a 3+1 axis configuration. This includes crowns, bridges, copings, hybrid abutments, temporaries, veneers, and others that comprise the bulk of laboratory cases. Commonly, a 4-axis machine’s supposed inability to mill one of these restorations stems not from a hardware inadequacy, but rather from bugs in the CAD software. For instance, certain design software may fail to orient the insertion direction of a 3-unit bridge toward the vertical before outputting the file. When exported to the 4-axis mill, the shape appears to have undercuts that can’t be reached without a rotary axis; in reality, a simple rotation of the virtual model would render the 4-axis machine perfectly adequate to the task.

In the end, the decision to buy a 4- or 5-axis milling machine should rest heavily on the number of cases rolling through the laboratory that require the additional axis; and that number is often inflated. A 5-axis milling machine carries with it a significant increase in both upfront and continuing costs, moneys that may be better allocated to outsourcing the physical milling of complex cases if these restorations represent a low percentage of the laboratory’s overall case load.