Milling Sculptured Surfaces

The challenges that die and mold makers face daily are leading many of them to consider high-speed machining. While manufacturing dies and manufacturing molds are completely different jobs, they have one key thing in common: the intricate task of machining sculptured surfaces. The trend toward high-speed machining has been bolstered by the development of milling centers and turning centers capable of higher feed rates and spindle speeds. Currently, cutting tool manufacturers are pushing the development of new tool materials and coatings to keep up with these high-speed machines.

Many components go into a successful high-speed machining system. For example, high-speed machining actually begins with the CAD/CAM system as it generates cutter paths. Accuracy requirements, cutting strategy, and interpolation (linear, circular, or spline) determine the amount of data the CNC will have to process later. The high-performance milling system will be only as effective as its weakest component. Too often, that component is the cutting tool.

Dr. Taylan Altan, director of the Engineering Research Center for Net Shape Manufacturing at The Ohio State University, Columbus, OH, and his team of researchers have started a program to evaluate tooling requirements for high-speed milling. Companies such as Dapra Corp., GE Superabrasives, and Ingersoll Cutting Tool Co. have joined in this effort to advance cutting tool technology.

Generally, high-speed milling is best suited to finishing operations. In die or mold making, the stringent requirements for surface finish often make a polishing operation necessary—an operation that in many cases is still done manually. But if a finish-milling operation performed at high speed requires subsequent polishing, then the purpose of high-speed machining—to get the finishing job done as proficiently and as quickly as possible—is defeated.

Increasing the number of finishing paths improves surface finish. The width of cut combined with the tool radius determines the theoretical surface roughness. Since the maximum cutter radius is limited by the part geometry, especially fillet radii, the only way to minimize the theoretical surface roughness is to minimize the width of cut. If the width of cut is decreased by 50%, then the number of cutter paths automatically increases by 100%. This improves the surface finish, but it also doubles the time it takes to finish the part. To avoid increased cutting time, higher feed rates must be applied. These higher feed rates require higher spindle speeds to ensure constant chip thickness, and higher spindle speeds automatically result in higher surface cutting speeds. Higher temperatures and accelerated tool wear are the unavoidable consequences. Therefore, new materials must be developed to withstand these more severe cutting conditions.

The Tools 
Ball endmills, bullnose endmills, and flat endmills can be used to finish-mill sculptured surfaces. Flat endmills can be used for this purpose solely on a 5-axis milling center, a piece of machinery used by only 6% of American die- or mold-making companies. The bullnose endmill, a tool whose cross section falls in between the rectangular cross section of a flat endmill and the semicircular cross section of a ball endmill (Figure 1), is rarely used in the United States.

The bullnose endmill’s scarcity in the United States is surprising, because its geometry offers some advantages over the more popular ball endmill. There is no area on the bullnose endmill’s cutting edge where the cutting speed is zero. However, on a ball endmill, the tip is moving in a linear motion and the cutting speed is zero. This is why metal surfaces generated with the tip of a ball endmill have a dull appearance. Nevertheless, since the ball endmill is currently the tool of choice for finishing sculptured surfaces in America, the researchers have restricted their experiments to this tool.

The cutting speed of a ball endmill changes constantly along the cutting edge. It is zero at the tool tip and reaches its maximum at the outer diameter of the tool. If a 1"-dia. ball endmill is rotated at 10,000 rpm, the cutting speed ranges from 0 to 2,600 fpm. Also, the chip thickness changes along the cutting edge (axial chip thinning) and, at the same time, with the engagement angle (radial chip thinning). This is illustrated in Figure 2.

To further complicate matters, the ball endmill must remove stock left by roughing or semifinishing operations. Since the stock allowance is not uniform, the chip thickness changes constantly along the cutter path.

As a result, cutting conditions in terms of cutting speed and chip thickness change constantly, and several wear mechanisms occur on a single cutting edge. To improve the design of milling cutters, these tool wear mechanisms have to be further investigated.

Ball Endmill Wear 
The Ohio State University’s Engineering Research Center for Net Shape Manufacturing surveyed 99 die and mold manufacturers in the United States, Canada, and Mexico. Of these firms, 75% said their top priority was longer tool life. There’s good reason for this concern. When finishing a large die or mold, it’s best to avoid replacing a worn tool in the middle of a cut, because there will be a mark on the surface where the worn tool was removed and the new tool installed. It’s usually best to remove the 0.01" to 0.02" of finishing stock in one operation with one insert. To achieve this, a tool often must last for several hours of cutting.

Different wear mechanisms attack the cutting edge. While abrasion and adhesion are the two main causes of tool wear at conventional cutting speeds, the temperatures generated at higher cutting speeds lead to chemical interaction between the workpiece and the tool material. The transition from predominantly mechanical to predominantly chemical wear is a function of temperature and the combination of cutting tool material and workpiece material. Titanium alloys are particularly prone to material interaction with tools, because they are highly reactive to all metals even at comparatively low temperatures. Therefore, when machining a titanium alloy, a tool wears quickly even at relatively low cutting speeds.

Experiments conducted by Herbert Schulz at the University of Darmstadt, Germany, revealed just how high temperatures can get on the rake face of a milling tool at high speeds. The experiments were run at the following parameters: workpiece material, 1045 steel; cutting material, P-20/P-25 carbide; cutter diameter, 40mm; width of cut, 1.5mm; DOC, 5mm; feed per tooth, 0.31mm. At a cutting speed of 100m/min., the temperature reached 550° C; at 400m/min., 600° C; at 800m/min., 700° C; and at 1600m/min., 1100° C. These tests and others reveal that cutting materials must combine high hardness with chemical stabilities at temperatures of 1,000° C or higher to survive in a high-speed milling environment.

It is nearly impossible to accurately measure temperatures and stresses on the cutting edge of a tool in actual cutting operations. Installation of sensors in the cutting zone would affect the conditions they were intended to monitor. However, investigations of tool wear and tool design can be supported by finite element analysis (FEA). This software-based process-simulation program can predict the temperatures, forces, and stresses to which a cutting tool will be subjected in a machining operation. Formerly a tool of universities only, FEA is currently being evaluated and used by several cutting tool companies. Figure 3 shows a simulation of a PCBN tool cutting P-20 steel.

Put to the Test 
To gage the effects of high-speed milling on various tool materials, the Engineering Research Center for Net Shape Manufacturing conducted experiments on P-20 mold steel with a hardness of about RC 30. P-20 was chosen as the work material because it is a major player among materials for die and mold making; about 20,000 tons are output each year for manufacturing molds. The experiments were conducted on a LeBlond Makino A55 horizontal machining center. A 7 1/2" x 7 1/2" flat workpiece was machined in a climb-milling mode at 0.02" DOC and 0.02" width of cut. The 6"-long ball endmill was tilted at a 30° angle to avoid engagement of the tool tip.

Inserts for the 1"-dia. ball endmills were provided by Dapra Corp., among them uncoated carbide inserts as well as carbide inserts coated with titanium nitride (TiN), titanium carbonitride (TiCN), and titanium aluminum nitride (TiAlN). In addition, GE Superabrasives provided BZN-6000 PCBN tool tips brazed onto plain carbide substrates. (BZN-6000 is a high-CBN-content tool material.) Only inserts with flat rake faces were used; none had chipbreakers.

Calculations of cutting conditions were based on a constant chip thickness of 0.0025". At a cutting speed of 1,000 fpm, spindle speed was 5,300 rpm and the feed rate was 104 ipm; at 1,800 fpm, spindle speed was 9,500 rpm and the feed rate was 187 ipm. Also, one PCBN insert was run at 2,600 fpm, with a spindle speed of 13,700 rpm and a feed rate of 271 ipm. Cutting speeds were evaluated for the highest point of engagement on the ball endmill. Each test condition was run only once per tool. To avoid the influence of tool runout on wear measurements, only tools with a single cutting edge were used. The inserts were considered worn out as soon as the maximum flank wear reached 150µm.

At a cutting speed of 1000 fpm, accelerated flank and crater wear caused the uncoated carbide inserts to perform more poorly than coated carbide and PCBN (Figure 4). Surprisingly, TiN-coated carbide ran much better than the TiAlN- and TiCN-coated tools. The apparent explanation for this is that, to take advantage of the special properties of the advanced coatings, the machining setup must be extremely rigid.

However, in die- and mold-making operations that involve deep pockets, longer tools are needed, which reduce rigidity. The 6"-long tools used in the experiments were probably not rigid enough to take advantage of TiAlN and TiCN coatings. The PCBN insert had less flank wear than the other tools but chipped at 45,000 linear inches.

The 1,800-fpm test results show wear on coated inserts developing at similar rates. Even under these conditions, TiAlN and TiCN could not outperform TiN. Since PCBN wear was developing very slowly, the test was stopped after 100,000 linear inches. At this stage, the flank wear was measured at 80µm.

In the test on PCBN at 2,600 fpm, the tool lasted for about 30,000 linear inches before reaching the 150µm maximum flank wear. This indicates that, for approximately the same tool life, productivity can be increased by 30% when using PCBN instead of coated carbide inserts, when machining time is taken into account. Overall, along with being more productive than the coated carbide tools, the PCBN tools also produced the best surface finish. The roughness measurements were taken after the inserts reached the end of their tool life.

It is misleading to rely solely on roughness to evaluate the performance of a cutting material. A worn tool may impart a better surface finish than a fresh one, because fresh inserts cut more defined scallops, resulting in higher surface roughness. However, since worn inserts also have an offset cutting edge, they leave stock on the surface that will have to be removed manually.

Based on the results of the tests, PCBN appears to be the optimal cutting tool material for the high-speed milling of sculptured surfaces. When using PCBN, it is important to keep in mind that it is sensitive to the sudden changes in chip thickness that often occur when milling fillets. Therefore, one should program tool paths that avoid abrupt variations in stock removal. Furthermore, engagement of the tool tip on ball endmills is not recommended. When using ball endmills to finish large P-20 steel molds with slightly curved surfaces, at least 4-axis milling capability is suggested. Die and mold makers can complement a high-speed machining setup with PCBN cutting tools to obtain higher productivity, better surface finishes, and fewer tool changes.

About the Authors 
Peter Fallböhmer is currently leading the high-speed milling effort at the Engineering Research Center for Net Shape Manufacturing at The Ohio State University, Columbus, OH. Bob Scurlock is application specialist with GE Superabrasives, Worthington, OH.