Advances in Abrasive Technology for Grinding Gears from Solid

Past reviews of the status of grinding for the gear industry have inevitably focused on the process as one of finishing of steel gear profiles, diameters, bores, etc. in the hardened state [1]. Discussions have typically revolved around the relative benefits of grinding in comparison to hard cutting processes such as hard hobbing and power skiving, or to various honing or lapping technologies.

In general, grinding has been considered to offer improved surface quality, dimension accuracy, and process reliability, giving improvements in, for example, noise reduction, but at reduced removal rates, and a somewhat higher environmental impact due to the use of coolant. [2].

In general, little attention has been paid to the potential of grinding of the rough form in the soft state prior to heat treatment, which is currently dominated by hobbing, shaping, and broaching. This has been due in large part to the inability of traditional abrasive technologies to provide the necessary metal removal rates and power efficiencies to be competitive. However, recent advances in wheel technology, especially those related to dressable vitrified bonds utilizing engineered ceramic grains, appear to offer the potential for a significant shift in thinking of how to rough shape a gear form. (Figure 1)

Hobbing and shaping are cutting operations using well-defined tool geometries to create a specific chip configuration. Metal removal rates are of the order of 5-10 in3/in/min (50-100 mm3/mm/sec) and specific metal removal energy requirements of about 1.5 Hp/in3 /min (4 J/mm3) and as such are quite efficient, due to the relatively small numbers of large chips created. (Figure 2)

The disadvantages of a process like hobbing lie in the flexibility and reliability of the tooling. The tools, although often long-lived, are expensive, specific to a given gear, and require long lead times to manufacture. Cutting performance and quality also change as the tool wears.

By comparison, grinding with conventional abrasives has historically been limited by the metal removal rates (Q’) being too low, the specific grinding energy (U’) being too high, or the wheel life (G Ratio) being too little. Bonded CBN wheels have also been unable, practically, to achieve the high stock removal required even at high wheel speeds, while plated CBN wheels have also struggled to achieve the required removal rates, were extremely expensive, and, like hobs, were specific to a given gear profile. Grinding soft steel has also been a problem due to wheel loading unless accompanied by continuous dressing – e.g. continuous dress creep feed grinding (CDCF). CDCF was developed in the 1970s [4] using high porosity wheels to take deep form cuts, where it was found that by continuously dressing with a formed diamond roll dresser to keep the abrasive grains clean and sharp, the specific grinding energy was reduced significantly. In combination with good coolant access through the high-induced wheel porosity, CDCF allowed an order of magnitude increase in stock removal rates. Unfortunately although this was very effective on tough to grind metals such as Inconel, the level of wheel wear from continuous dressing (typically 1um/wheel rev at 1000 rpm), resulted in uneconomic G Ratio values compared to machining tool costs. This, combined with the need for a specific form diamond roll, again made CDCF impractical for rough grinding of gears.

For the last decade, a major goal of abrasive research at Norton has therefore been to develop a system that can achieve specific grinding energies U’ and metal removal rates Q’ approaching those of hobbing and shaping, but with a G Ratio that allows deep form grinding without continuous dressing, preferably without the need for high wheel speeds. This was recognized as offering the possibility for grinding in the soft state on standard finish gear grinders dressed using a common CNC contour diamond dress roll (Figure 3).

Grind Modeling

There are very good predictive models for grinding, based on un-deformed or active chip thickness theories, that help guide grinding system development. All these are founded on the fact that the grinding forces are a function of the thickness of chip created by the abrasive grains [6, 7].

It may be noted from these equations a necessary condition to achieve high Q’ value with low U’ is the creation of large chips (t’); this can be achieved in part through the manipulation of grind parameters, but the key factors for abrasive and wheel bond design are the cutting grain density (C) and grain shape (r). In addition, the chip size has an even bigger impact on the force/grain, emphasizing the fact that any system developed for creating large grinding chips requires the maximum grain strength and bond holding. (Figure 4)

Engineered abrasives are grains that have microstructures that have been produced with controlled crystallite sizes from the sub-micron to multi-micron level by processes other than traditional fusion and comminution. These include solgel/sintering and agglomeration techniques. For alumina grain it was known that reducing the crystal size from the macro scale equivalent to a single crystal per abrasive grain, common in fused material, to micron or ideally sub-micron crystalline structures, significantly enhanced grain properties such as hardness and strength. (Figure 5)

Over the last 30 years, Norton focused much of its research on the control of grain micro-structure and shape. The first milestone was the development of SG, a ceramic grain with sub-micron sized microstructure for maximum grain strength, hardness, and controlled breakdown by micro-fracture. This grain proved so tough that a later modification was made to the crystallite structure called NQ to maintain the hardness but increase friability to allow for micro-truing and reduced grinding forces under low metal removal rates. Most recently, Norton introduced Vortex, an agglomerate grain consisting of multi-micron crystallites. This resulted in a family of grain types provides a range of engineered breakdown levels from the sub-micron (SG) to the micron (NQ) to the multi-micron (Vortex). The Vortex grain also had the unique property of being both strong and permeable. This provides both a very free cutting action in its own right, and benefits when used as a co- or secondary abrasive filler alongside tougher abrasive grain, of providing structural rigidity to a wheel matrix combined with additional porosity / permeability for coolant access. (Figure 6)

Grain shape optimization was also developed in parallel based on the need to minimize the shape factor r in the chip model. The ultimate shape was considered to be that of a long cylinder. Conventional crushed grain has an aspect ratio of around 1.2:1–1.5:1 and an r-value of typically (7 – 11). Norton developed a family of cylindrical grains based on the hardness and strength of the SG microstructure, but with aspects ratios of 5:1 (TG grain) to 8:1 (TG2 or Altos) and r-values closer to 1–2. [9]. High aspect ratio cylindrical grain has a very beneficial packing characteristic. Standard crushed grain has natural packing volume of 50%; TG2 was found to have a far lower packing content allowing for dramatically more coolant access into a rigidly supported wheel structure. (Figure 7)

The study of packing of high aspect ratio objects is of broad interest across industry, from efficiencies of catalysts to the packaging of carrots. Figure 8, from a thesis by Zhang [10], shows the appearance of randomly packed cylinders with an aspect ratio of 10:1 and a packing density of 30%. The edges of the cylinders are exposed at the surface, providing a relatively low density of strongly anchored cutting points. The density of these points can be further reduced by the addition of lower aspect ratio filler such as Vortex mentioned above which maintains structural integrity and permeability, and resists anisotropic packing effect that might otherwise occur in, for example, pressing the wheel.

The final, crucial piece in wheel development is the optimization of the glass bond. Fine crystallite grain structures offer a larger, more reactive, and potentially stronger surface interface. Over the last 20 years, the research has focused on optimizing bond strength and in the process minimizing bond content to maximize freedom of cut and coolant access. The latest bond technology, launched this year called Norton Vitrium 3 has significantly greater strength than any previous glass bond system. (Another fortuitous benefit has been that the bond is fired at relatively low temperatures, significantly reducing the environmental impact of manufacture). The bond technology allows for narrow, long bond posts between regularly shaped grains, or a “spot-weld” type bonding at contact points between over-lapping cylindrical grains. (Figure 9)

To summarize, the resulting wheel formulation, engineered as a solution to grinding gears from solid, is an extremely porous wheel structure composed of high aspect ratio TG2 needles and high-permeability Vortex agglomerates held in an ultrahigh strength glass bond. The resulting cutting surface presents a relatively low surface concentration of cutting points (C) with a very low grain aspect ratio r from which it would be predicted to generate reduced specific grinding energy and much larger chips.

The most striking physical evidence for verification of the design concept is apparent in looking at the grinding swarf. Hashimoto [11] reported a comparison of grinding with wheels containing TG2 and conventional grain and found the chip size to be significantly larger with the TG2 grain (Figure 10). The appearance of swarf ground with TG2 has been found to resemble more that of machining processes than conventional grinding processes and is actually an important consideration in flushing from the grinding bed.

Performance of Norton Vitrium 3-Engineered Ceramic Wheels

Tests were carried out at the Higgins Grinding Technology Center (HGTC) to simulate the grinding of a half form of a 3D.P. Spur gear from solid in 8620 steel in the soft state. The details are given in Table 1.

The grinding performance of the TG2 grain (TG280 – F20VTX2) with the Vortex porous filler in the Norton Vitrium 3 bond is presented in Table 2.

The G Ratio values at 200–750 were one or two orders of magnitude higher than those expected for conventional grinding operations and approached values more typical of CBN. The specific grinding energy, while still higher than hobbing, was half that of conventional grinding processes. One critical factor for success was high-pressure coolant. The tests above were carried out using oil-cleaning jets at 1000 psi.

Recently published data by Heinzel [13] has demonstrated effectiveness down to 500 psi. with appropriate positioning. Gleason [14] reported on successfully grinding bevel and hypoid gears from solid and emphasized the need for appropriate wheel cleaning (Figure 11).

Field Study and Summary

There are already a number of proven field applications illustrating the potential using the Norton Vitrium 3 technology as illustrated in Table 3 [14]. The immediate benefits have been achieved by job shops requiring fast turnaround times that cannot wait for the time required to manufacture a hob or shaper, or have the grinder capacity that they can grind components in the soft and hard state on the same machine saving the cost of investing in a hobber and increasing shop flexibility. The introduction of Norton Vitrium 3 technology is expected to broaden the economic benefits of grinding into higher production operations. It is also to be expected that rapid bond and grain advances will occur in what is becoming a highly competitive abrasive market, as other wheel and engineered grain technologies emerge such as the recently introduced Cubitron 2, a polygonal shaped engineered grain from 3M [15]. This should be a consideration in specifying both the cutting process but also machine requirements such as grinding spindle power, when buying new machines.

References

[1] B. Karpuschewski, H.-J. Knoche, M. Hipke, “Gear Finishing By Abrasive Processes” CIRP Annals – Manufacturing Technology 57 (2008) 621–640
[2] F. Klocke, E. Brinksmeier, K. Weinert, “Capability Profile of Hard Cutting and Grinding Processes” CIRP Annals Keynote 2005 Antalya, Turkey – August 21-27, 2005
[3] Gleason Corp, Coniflex Tools Manufacturing Engineering Sept 2007 Vol. 139/3
[4] S Salmon “Modern Grinding Process Technology” Book McGraw Hill Inc Publ. 1992
[5] Hofler Machine Catalog “Profile Grinding Machine Rapid 2500 – 6000” 2010
[6] M. Shaw. “Principles of Abrasive Processing” Book Publ. Oxford Series on Advanced Mfg 13. Clarendon Press 1996
[7] S Malkin “Grinding Technology Theory and Application of Machining with Abrasives” Book Publ Ellis Horwood 1989
[8] J.A.Webster. M Tricard, “Innovations in Abrasive Products for Precision Grinding” CIRP Innovations in Abrasive Products for Precision Grinding Keynote STC G August 23, (2004).
[9] M.J. Jackson, M Hitchiner “High Performance Grinding and Advanced Cutting Tools” Springer Briefs in Applied Sciences and Technology 2013
[10] W. Zhang “Experimental and Computational Analysis of Random Cylinder Packings with Applications” PhD Diss. Louisiana State University 2006
[11] F Hashimoto “Key Manufacturing Technologies for Ultra Large Components” 2012 Saint-Gobain Grinding Research Symposium Northboro MA, November 8th, 2012
[12] Anon “Less is More in New Abrasive Bond” GSF June 2013, p20-21
[13] C. Heinzel, G.Antsupov, “Prevention of Wheel Clogging in Creep Feed Grinding by Efficient Tool Cleaning” CIRP Annals-Manufacturing Technology 61(2012)323– 326
[14] Gleason Corp. “Bevel Gear Manufacturing: Grinding from Solid” Commercial brochure Sept. 2009
[15] 3M Corp “ Bonded Abrasive Wheel” Patent Appl W2011/109188 A2 Sept 9, 2011.