Thesis Title: In-Situ Process Monitoring and Quality Improvement for Nanomanufacturing

Advisors: Dr. Jianjun (Jan) Shi

Committee members:

Dr. Ben Wang 

Dr. Chuck Zhang 

Dr. Kamran Paynabar 

Dr. Judy Jin (University of Michigan) 

Date and Time: Tuesday, April 17, 2018, 10:00 AM

Location: ISyE Groseclose 403

Abstract:

Carbon Nanotubes (CNTs) buckypaper is an important multifunctional platform material with great potential for creating lightweight, high-performance materials for various applications due to its superior mechanical and electrical characteristics. One of the critical roadblocks to scale-up production of high-quality buckypaper is the monitoring and control of buckypaper made from the continuous nanomanufacturing process. There is a pressing need to address the fundamental research issues, develop transformative manufacturing technologies and propose systematic methodologies to realize in situ process monitoring and quality improvement for high-quality buckypaper.

On the basis of these initiatives, the dissertation focuses on developing systematic methodologies for effective system modeling, processing monitoring and quality control in the continuous nanomanufacturing process. This dissertation starts from an introduction to illustrate the motivation, research objectives, state-of-the-art and organization structure of the dissertation.

In Chapter 2, a generalized wavelet shrinkage is proposed to realize real-time denoising and signal enhancement for nanomanufacturing process. Raman spectroscopy is an attractive in-line quality characterization and quantification tool because of its non-destructive nature, fast data acquisition speed, and ability to provide detailed material information. However, there is signal-dependent noise buried in the Raman spectra, which reduces the signal-to-noise ratio (S/N ratio) and affects the accuracy, efficiency, and sensitivity for Raman spectrum-based quality control approaches. Based on the validated signal-noise relationship, a novel generalized wavelet shrinkage approach is introduced to remove noise in wavelet coefficients by applying individual adaptive wavelet thresholds. The effectiveness of this method is demonstrated using both simulation and case study.

In Chapter 3, a penalized mixed-effects decomposition (PMD) is proposed to solve the multichannel profile detection problem in nanomanufacturing. The proposed PMD exploits a regularized high-dimensional regression with linear constraints to decompose the profiles into four parts: fixed effects, normal effects, defective effects, and signal-dependent noise. Finally, the separated fixed effects coefficients, the normal effects coefficients, and the defective effects coefficients can be used to monitor fabrication consistency, within-sample uniformity, and defect information, respectively. Using surrogated data analysis and case study, we evaluated the performance of the proposed method, and demonstrated a better detection power with less computational time.

In Chapter 4, we focus on quality inspection from massive high-dimensional Raman mapping data with mixed-effects. The existing tensor decomposition methods cannot separate mixed effects, and existing mixed-effects model can only handle vector/matrix data instead of multi-array data. We propose a novel tensor mixed effects (TME) model to analyze massive Raman mapping data with complex structure. The proposed TME model can (i) separate fixed effects and random effects in a tensor domain; (ii) exploit the correlations along different dimensions; and (iii) realize efficient parameter estimation by a proposed double Flip-Flop algorithm. Properties of the TME model, existence and identifiability of parameter estimation are investigated. The numerical analysis demonstrates the efficiency and accuracy of the parameter estimation. Convergence and asymptotic properties are discussed in the simulation and surrogate data analysis. The Real case study shows an application of the TME model in quantifying the influence of alignment on CNTs buckypaper.

In summary, the dissertation contributes to the area of System Informatics and Control (SIAC) to develop systematic in-situ process monitoring and quality improvement methodologies based on massive high-dimensional data. The methodologies can also be applied to other advanced manufacturing systems.