The research objective of this Faculty Early Career Development (CAREER) Program project is to advance our understanding of material removal and surface smoothening mechanisms for creating components with nanometric form accuracy and sub nanometric surface finishes. This will be achieved by applying frequency response analysis techniques, as used by the machine tool industry, to precision polishing to experimentally measure and define the relationship between polishing tools, process vibrations and dynamics, temperature, and the final polished workpiece quality. Unique elements of the research include 1) measuring the dynamic moduli of polishing pitch used on polishing tools through frequency sweep and impulse excitation testing, 2) prediction of the tool's dynamic behavior by finite element modeling, 3) quantification and isolation of process vibrations through instrumentation of a polishing machine, 4) modification of process vibrations through tool design considerations, 5) isolation of vibration threshold values affecting surface finish and material removal rates and 6) collaboration with industrial partners to further validate the process assessment, monitoring and tuning methodology. Research deliverables, centered on advanced understanding of process fundamentals, will include a novel quantitative method of process assessment and a validated, deterministic algorithm for optimized polishing tool design. The objective of the educational program is to provide enhanced material on precision manufacturing, give graduate students exposure to industrial considerations, involve undergraduate students in the laboratory and provide K-12 students with a hands-on engineering experience through existing programs at the University of North Carolina at Charlotte.
Successful completion of this research will ensure better quality polished surfaces and a higher degree of process automation. This will positively impact the costs associated with fabricating optical components for laser fusion systems, lithography systems, lasers for military and medical applications, and X-ray optics. The integrated circuit manufacturing industry can apply the knowledge to their chemical mechanical planarization processes. Integration of the results into current educational programs, combined with additional outreach efforts, will help improve the quality and number of engineering students available to US companies in a competitive manufacturing environment.
The main success of this project was uncovering a scientific based understanding of how process vibrations affect material removal rates and the surface quality obtainable during the polishing of high quality, fused silica (glass) optical components. The polishing processes typically used in these applications are ‘artisan’ based and rely extensively on skilled technicians. By providing fundamentally based process knowledge we have opened up several avenues by which technicians can deterministically monitor and fine tune their processes to optimize efficiency. While the research focused on polishing processes utilizing pitch (a hard viscoelastic material) coated metal tools, the research is applicable to other polishing processes. A summary of the key results include; 1) process vibrations existing during polishing have amplitudes in the nanometer range (frequency bandwidth 10 to 16 kHz), 2) the amplitudes of the vibrational content will vary from machine to machine, with process parameters and tooling construction, 3) up to certain limits, material removal rates increase linearly with increased vibrational input power, 4) low cost methods (<$100) of attenuating (addition of a passive damping layer to the tool) and amplifying (addition of small battery operated unbalanced motors to the workpiece holder) the vibration input power on actual polishing machines was demonstrated with the expected impact on the material removal rates (Figure 1), 5) an analytical model incorporating both contact mechanics and fluid flow was generated to explain how process vibrations influence the material removal mechanisms, 6) dynamic testing and high resolution imaging of different types of polishing pitch used on the tools indicates that variations in the surface topography of the tool (see Figure 2) affects material removal rates more than pitch dynamic properties. Intellectual merit: Process vibrations during high end polishing processes had not previously been measured. Vibrations measured in the 10 to 16 kHz bandwidth were found to have maximum amplitudes in the nano-meter range. Subsequent testing revealed that even these low level vibrations can affect the material removal rate. (Note: vibrational assisted polishing processes employ micron scale vibrational amplitudes.) A linear relationship was found between the vibrational power and the removal rate whereby higher power inputs produced higher removal. Analysis of the power spectral densities of the surfaces produced under differing vibrational conditions, spectral analysis of the process vibrations, and process kinematics strongly suggest that lower frequency vibrations (<500 Hz) have more of an effect on the surface quality than the higher frequency vibrations. A fundamentally based model which includes both contact mechanics and thin film fluid flow between two surfaces, one of which is not smooth, was proposed to explain how the vibrations affect the material removal rate. The fluid shear and pressure gradients within the thin film play a significant role in the vibrational driven removal mechanism. The model can also accommodate different types of polishing pitch commonly used on polishing tools. Experimental verification of the model predictions was provided (Figure 3). In depth analysis of the different types of pitch used in polishing included high resolution scanning electron microscope measurements (SEM) and atomic force measurements (AFM) of the surface along with dynamic impact and frequency sweep tests on bulk samples. Analysis of the information obtained from these measurements provides new insights into how the pitch selection affects the material removal rates. While different types of pitch can have the same hardness values (a typical workshop floor metric) they can have differing dynamic stiffness values, however it appears that the surface topography, i.e. how the abrasive particles embed themselves into the tool surface, is more significant with respect to material removal rates. Broader impacts: The project supported four graduate students (1 PhD, 1 MSc, and two part time) during their studies. Three undergraduate students and approximately two dozen high school students were given insights into precision manufacturing and metrology. Eight papers were produced and ten oral presentations were given to a wide range of audiences. Project findings and modeling efforts are of interest to others such as those involved in IC manufacture and those interested in thin film fluid flow theory. The grant has help establish effective working collaborations with other faculty members within the department and these collaborations will continue past this project’s end date.
