The new three-face grinding blade design for dry cutting spiral bevel gears guarantees higher process capability, shorter handling sequences, and longer part runs.

In the past five years, dry-cutting technology has steadily moved to dominate production of automotive hypoid gearsets in the United States, while a parallel trend has occurred in Europe. The significant difference between wet cutting with high-speed steel and carbide cutting is higher productivity, better quality, increased ecological friendliness, and lower total cost of production.

A simple substitution of a high-speed steel blade with carbide will show the same shape of the flank, but the potential of dry cutting using carbide blades is not used. A protective coating of the carbide blades is mandatory. Normally, the protective coating consists of a Titan-Aluminum-Nitride-base and prevents thermal penetration into the outer layer of the carbide. This coating is critical to the dry-cutting process: without it, the underlying K-grade carbide substrate wears after only a few pieces.

Blade Geometry: An Introduction

Besides coating the blades, the cutting process itself needs to be taken into account. When chips are created, two technological angles are important: the rake angle, and the relief angle. In Figure 1 the influence of the rake angle can be seen.

Figure 1: Rake angle and shape of the chips.

The arrow in this figure indicates the actual cutting direction. The angle between this direction and the front side of the blade is the rake angle. The higher the rake angle, the less the chips are bent. The smaller the rake angle, the chips are more bent, and more thermal energy is inducted into the blade. The obvious fact is very important: after every pass through the gap, the chips need to leave the cutter head. If there is enough space between following blades, the shape of the chips is not very important.

In the case of a cutter head with a high number of blades, the room in between is limited. If single chips do not fall down, they will be scratched through the next gap, damaging the cutting edge of this blade. If the cutting edge loses its sharpness, more thermal energy is induced, which decreases the tool life. The other technological angle is the relief angle, shown in Figure 2.

Figure 2: Different relief angles.

Compared to the rake angle, the relief angle doesn’t effect the shape of the chips. But it can be seen that the thermal energy induced into the cutting edge is better transported to the center of the blade when the relief angle is small. If the relief angle is too small, the relief side will slide on the flank of the gear. If the relief angle is high, we have a locally high temperature on the cutting edge. Depending on the shape of the gear’s flank, a minimal angle between rake face and relief side can be calculated.

The relief angle is not allowed to be smaller than this minimum. Because of thermal aspects, the tool life is very sensitive to the relief angle. Another effect is based on a chemical aspect. Since there is an affinity between carbide and steel, the surface of the carbide is attacked and changes its chemical characteristics during the cutting process. In the case of a high relief angle, the contact zone is small, and the carbide is not much attacked. But now the temperature flow into the blade is limited. Using a small relief angle will transport the heat away from the cutting edge faster, but the contact zone of the carbide blade and the gear flank is bigger, attacking the carbide material to a greater degree.

Besides affinity, another physical effect is important. The more peaked the cutting edge, the less friction we have when making chips. This can only be achieved by increasing the sum of face angle and relief angle. But this advantage has the side effect of reducing the blade rigidity and of a slower transportation of thermal energy from the cutting edge into the carbide blade. There are several effects in the cutting process influenced by the technological geometry of the blade. Since the geometry of the gear is only given by the shape of cutting edge and not by the technological geometry of the blade, it is mandatory to optimize the rake and relief planes of the carbide blade.

Two-Face Grinding and Three-Face Grinding Blades

The geometry of a 2F blade and a 3F blade initially appears quite similar. Both have a front rake face, a cutting (pressure angle) flank, and a clearance (non-cutting side) flank. But in a 2F blade, the entire front rake face is pre-ground into the shaft of the blade during the blade blank manufacture. Typically TiAlN-based coating is applied, commercially available from several sources. After blade manufacture, the cutting profile (cutting and clearance sides) is ground into the precoated blade blank on a suitable precision blade-grinding machine, such as the Oerlikon B24 or B10. But the front rake face is never reground or recoated after blade blank manufacturing. While the pre-ground front rake face initially provides a simplification of the blade resharpening process, it does so at great cost once the carbide sticks are used in production (Figure 3).

Figure 3: Two-face grinding blade.

A 3F blade design requires only an uncoated rectangular carbide stick to be delivered for initial profiling. This greatly eases the manufacturing requirements of blade suppliers, as they no longer need to hold the very tight tolerances required when pregrinding the front rake face into the stick. A simple uncoated rectangular carbide stick, which is much simpler to manufacture, is all that is needed for the blade blank (Figure 4).

Figure 4: Three-face grinding blade.

But what does this mean for profiling and subsequent resharpening of 3F blades once they are in production? They must be reground on all three faces at every resharpening, and recoated afterwards.

Additional coating costs are incurred, and carbide blade supplies must be large enough to allow for “blade float” while the blades are out-of-house being recoated. These clear disadvantages initially stunted the acceptance of 3F blade technology in the United States.

However, these drawbacks allow for a few extremely critical opportunities when designing with 3F blades. The independent control that we have over the front rake face means that we can tune this angle to optimize the cutting performance. With 2F blades, the front rake angle is 12.00 degrees. Whether an outside blade or an inside blade, carbide or HSS, soft material or hard material, the front rake angle is always 12.00 degrees. Clearly this is not an optimal choice in every case. 3F blade geometry allows us to define this front rake angle as the cutting conditions demand for consistent tool life, surface finish, and part quality–from lot to lot, ratio to ratio, and job to job (Figure 5).

Figure 5

Since we regrind this front rake face, however, we are forced to recoat the blade after every resharpening. Here, what appears at first to be a severe drawback to the 3F approach is actually one of its greatest strengths: not only is the front rake face recoated, but also the cutting and the clearance flanks. With a fully coated blade, the cutting results improve dramatically. The inert TiAlN coating on the relief surfaces inhibits the tremendous flank wear, which is a key weakness of a 2F blade design. Surface finish improves, and blade “run-in” becomes trivial, with no need for additional edge preparation equipment.

Contact pattern, tooth topography, and size remain constant, with almost no change during the entire run of the cutterhead. And tool life, as a rule of thumb, doubles. Technology Applied: Awesome Cutting Results

Based on this analysis, technical experts at the Klingelnberg-Oerlikon group proposed an interesting experiment: could 3F blade technology also satisfy the process requirements of U.S. automotive axle producers?

With all of this information in hand, Klingelnberg-Oerlikon–together with a tier-one U.S. axle supplier — developed a 3F blade design to duplicate a high-volume production job currently cut with a 2F carbide blade system. After a preliminary trial using a single blade group, we placed a full cutterhead with 3F technology into production. The results were even better than expected.

The job selected for this experiment was a 225mm ring gear with 41 teeth. This job has high production numbers and is representative of the modern face-hobbed designs in use throughout the U.S. axle industry (Figure 6).

Figure 6

Typical tool life for this job with 2F blade design is 250 parts per cutterhead, on average, with some heads running as many as 300 pieces, and others as few as 180 pieces. The primary reason for pulling 2F cutterheads in normal production is a loss of blade pressure angle on the inside blade, causing the drive side contact pattern to “dive,” moving low toward the root of the ring gear. This condition is difficult to correct when the parts are lapped and, as such, must be monitored closely.

The 3F trial proceeded using a 17/88 Spiron cutterhead with a full set of optimized 3F blades, trued on a CS200 CNC cutter setting machine. The strong and stiff Spiron head construction allows us to true the assembled Spiron head radially without losing axial runout accuracy. The cutting parameters used were exactly the same as is normally used for 2F cutterheads, with identical feeds and speeds. Supplier personnel were in charge of all normal production activities during the 3F run — mounting the cutterhead, loading and unloading parts, and inspecting the finished pieces (Figure 7).

Figure 7

Klingelnberg-Oerlikon personnel were present for data collection and record keeping. We cut 605 pieces. We pulled the cutter for deteriorating coast side surface finish. We input KOMET corrections only once, to zero the machine after piece #1. The initial correction achieved a sum of the errors squared of less than 100. This shows the excellent ability of 3F blade design to emulate in-production jobs currently cut with 2F blades.

600 Pieces – No Degradation

The results of the painted and rolled parts go hand in hand with the sum of errors squared results, if not going a step further to highlight the robustness of the 3F process. All pieces rolled matched identically with the soft roll master. The typical 2F blade geometry currently in production shows tremendous pattern degradation in as little as 50 pieces. After more than 600 pieces, the 3F process showed no pattern degradation whatsoever. Blade wear was predictable and even. The outside blade and the inside blade had equal wear: a condition almost unheard of when cutting with 2F blades. Again, this is due to the ability to independently optimize the front rake faces of both the inside and outside blades with a 3F design.

The flanks of the 3F blades were barely worn, showing only smooth, even wear. The bulk of the wear was concentrated on the blade tips. 3F blades need only 0.4 mm of grinding stock removal during resharpening, which is less than half of what 2F blades require. This keeps 3F resharpening cycle times even with 2F blades, despite the addition of a third face to grind. And, over their lifetime, each set of 3F blades will produce more than four times the number of gears as a set of 2F blades (Figure 8).

Figure 8

Conclusion: 3F Blades

3F blade design can emulate tooth topographies and contact patterns of gears currently in production with 2F blade geometry. The improved quality and reduced variation of parts produced with 3F blades positively impacts downstream operations, ensuring faster and more-consistent lapping results. The additional degrees of freedom that 3F blade design provides allow optimization of the inside and the outside blades independently for even tool wear, consistent results, and a robust process. 3F cutter designs guarantee the user with longer part runs with higher process capability.

Another strong manufacturing benefit is the elimination or reduction of the sources of variation in the cutting process. Because every active face of the blade is accurately resharpened, variation in the blade blanks themselves is monitored and controlled before every cutterhead goes “out on the floor.” Loosely controlled processes, such as edge preparation, are rendered unnecessary. And the full coating of the entire active portion of the blade reduces the impact of slight variations present in blade grinding (wheel wear, surface finish, etc.). All of this translates into more-stable, more-predictable, and more-efficient production of bevel and hypoid gears.

In the future, new designs can be developed using the KIMoS hypoid gear calculation program that is tailored to take full advantage of the capabilities of the 3F system. For example, Spiron cutterheads are available in both a 17/76 and a 19/88 construction. These increased numbers of blade groups translate to higher cutting efficiency, and shorter cycle times. And contact pattern shape and inclination — also known as “bias” — which is so critical to the lapping development and noise behavior of a gearset, can be more completely controlled when designing with a 3F blade “from scratch.”