CNC finishing did not exist for most optics until COM and its industrial collaborators invented, patented, and commercialized the magnetorheological finishing (MRF) process. As exemplified by a line of MRF machines sold by QED Technologies, Inc., of Rochester, NY, the MRF process is capable of rapidly polishing out and figuring a variety of materials from a few mm to over 1 m in diameter. Plano, spherical, aspherical, and cylindrical optics with round or non-round apertures may be finished to better than 0.1 µm p-v form accuracy in minutes with a resulting surface micro-roughness of <1 nm rms.

QED Q22-Y MRF (magnetorheological finishing) machine for CNC polishing. Over 50 machines have been sold to industry for the precision finishing of optics used in high-end cameras, in semiconductor photolithography, and in other complex optical systems applications.

The key to MRF is a ribbon of abrasive-doped magnetic fluid that moves over a wheel and into contact with the surface of a spindle-mounted, rotating part. The fluid stiffens by four orders of magnitude in the contact zone, due to the presence of a magnetic field, turning from the consistency of honey to that of clay. Shear stresses between polishing abrasives in the fluid and the part surface cause material removal. The removal mechanism in MRF is unlike any other, and it results in a pit-free and scratch-free surface with high resistance to laser-induced damage.

Highest Quality Conventional Polish	MRF Polished Part
Dark field microscopy for acid-etched fused silica surfaces. A collaboration among Lawrence Livermore National Laboratory, Zygo Corp., and QED Technologies found that clean surfaces resulting from MRF are critical to enhanced UV laser damage resistance.

The removal spot is characterized with interferometric precision, and sophisticated computer algorithms allow a machine operator to define the MRF process goal. This may be the uniform removal of material from the micro-ground work piece surface to eliminate sub-surface damage, or the preferential removal of material to eliminate global surface figure errors. COM, ARL, and QED Technologies, Inc. won the DoD Defense Manufacturing Technologies Achievement Award in 2000 for MRF, and the R & D 100 Award for the Q22-Y MRF (Magnetorheological Finishing) System in 2001. [The Q22-Y, a new machine for finishing prisms, square and rectangular flats, and cylinders was introduced at the Optotech Trade Show and Exhibition in Fraunhofer, Germany in June, 2001.]

There are two magnetorheological (MR) fluids currently in widespread industrial use. One composition consists of cerium oxide in an aqueous suspension of magnetic carbonyl iron (CI) powder, and it has been found appropriate for almost all soft and hard optical glasses and low expansion glass-ceramics. The second composition uses nanodiamond powder as the polishing abrasive, and it is better suited to calcium fluoride, IR glasses, hard single crystals like silicon and sapphire, and very hard polycrystalline ceramics.

Advances have recently been made toward understanding the mechanism of removal with MRF, based in part on the hardness of the CI particles, the tribochemical interaction of cerium oxide or other abrasives with the work piece surface, and the type of slurry. We have found indirect evidence that, upon coming into contact with the work piece surface in the region of high magnetic field, the converging ribbon of MR fluid separates into layers. The top layer seems to be composed of the carrier liquid and the polishing abrasives, while the bottom layer consists of compacted chains of magnetic particles attached to the rotating wheel. In this condition, fluid flow in the polishing interface for the two commercially available MR fluids described above is optimal for promoting a high removal rate and smooth surfaces.

We maintain an active research program to address various issues associated with MRF. We concentrate on how MRF may be improved through experiments to better understand the relationship between MR fluid composition and the following: rheology in and out of a magnetic field, stability, material removal, the polishing process, and the quality of the finished part. Materials of interest are defined to be anything with relevance to research and industrial applications, and are not limited to "optical" materials. Specific areas of research are currently identified as follows:

• Colloid science-based strategy for MR fluid formulation, including magnetic particle surface modification, surfactant additives as coatings, 3D-gel formation, and dispersion/blending protocols;
• MR fluid formulation for different materials to be polished, including magnetic abrasive selection, polishing abrasive selection, carrier fluid selection, and stabilizer selection;
• MR fluid property measurement, including structure within the ribbon, sedimentation stability, chemical stability, base viscosity, and shear stress increment;
• MR fluid performance evaluation, including material removal rate and lowering the rms surface micro-roughness to below 0.5 nm rms.

Schematic depiction of MRF circulation system. The low viscosity MR fluid is pumped through a shaping nozzle onto a vertical, rotating wheel. At the apex of the wheel, the fluid stiffens into a ribbon ,under the influence of a dc magnetic field. The workpiece is placed into the ribbon and forms a converging gap. Material is removed by the shearing action of the nonmagnetic abrasives in the MR fluid. Shaping and smoothing are accomplished simultaneously as the rotating workpiece is moved through the ribbon under computer control.

We are interested in addressing issues that may arise from the use of MRF machines in industrial settings. User observations can offer additional insights toward our mission to improve the MRF process. To this end, we perform experiments to study user-related issues brought to our attention.

H. Romanofsky monitoring the progress of removal with MRF on a 266 mm diameter, fused silica parabola intended for the OMEGA EP laser at LLE.	Close-up of the Q22-Y after being set up by Ed Fess (LLE ME Department) to polish a 300 mm diameter, convex fused siica asphere for the OMEGA laser system.

Recent PhD Dissertations:

Chunlin Miao, "Frictional Forces in Material Removal for Glasses and Ceramics using Magnetorheological Finishing", Materials Science Program, University of Rochester, Rochester, NY, Dec. 2009.

Jessica E. DeGroote, "Surface Interactions Between Nanodiamonds and Glass in Magnetorheological Finishing (MRF)," The Institute of Optics, University of Rochester, Rochester, NY, June 2007.

Shai N. Shafrir, "Surface Finish and Subsurface Damage in Polycrystalline Optical Materials," Materials Science Program, University of Rochester, Rochester, NY, June 2007.

Recent Papers:

S. N. Shafrir, H. J. Romanofsky, C. Miao, M. Wang, R. Shen, H. Yang, J. C. Lambropoulos and S. D. Jacobs, "Zirconia-coated carbonyl-iron particle-based magnetorheological fluid for polishing optical glasses and ceramics", Applied Optics 48, 6797-6810 (2009)

R. Shen, S. Shafrir, C. Miao, M. Wang, J. Lambropoulos, S. Jacobs and H. Yang, "Synthesis and Corrosion Study of Zirconia Coated Carbonyl Iron Particles", J. Colloid Interface Sci. (2009), doi:10.1016/j.jcis.2009.09.033

C. Miao, S. N. Shafrir, J. C. Lambropoulos, J. Mici, and Stephen D. Jacobs, "Shear stress in magnetorheological finishing for glasses", Appl. Opt. 48, 2585-2594 (2009).

S.N. Shafrir, J. C. Lambropoulos, and S. D. Jacobs “Subsurface damage (SSD) and microstructure development in precision microground hard ceramics using MRF spots”, Appl. Opt. 46, pp. 5500-5515 (2007).

J. E. DeGroote, A. E. Marino, J. P. Wilson, A. L. Bishop J. C. Lambropoulos and S. D. Jacobs, “Removal rate model for magnetorheological finishing (MRF) of glass,” Appl Opt. 46, pp. 7927-7941 (2007).

S. N. Shafrir, J.C. Lambropoulos, S.D. Jacobs, 2007, “Toward magnetorheological finishing of magnetic materials,” J. of Manuf. Sci. and Eng., Vol. 129 5, pp.961-964 (2007).