Browsing by Subject "Process control"

Now showing 1 - 15 of 15

Advanced process control and optimal sampling in semiconductor manufacturing
(2008-08) Lee, Hyung Joo, 1979-; Edgar, Thomas F.
Semiconductor manufacturing is characterized by a dynamic, varying environment and the technology to produce integrated circuits is always shifting in response to the demand for faster and new products, and the time between the development of a new profitable method of manufacturing and its transfer to tangible production is very short. The semiconductor industry has adopted the use of advanced process control (APC), namely a set of automated methodologies to reach desired process goals in operating individual process steps. That is because the ultimate motivation for APC is improved device yield and a typical semiconductor manufacturing process can have several hundred unit processes, any of which could be a yield limiter if a given unit procedure is out of control. APC uses information about the materials to be processed, metrology data, and the desired output results to choose which model and control plan to employ. The current focus of APC for semiconductor manufacturers is run-to-run control. Many metrology applications have become key enablers for the conventionally labeled “value-added” processing steps in lithography and etch and are now integral parts of these processes. The economic advantage of effective metrology applications increases with the difficulty of the manufacturing process. Frequent measurement facilitates products reaching its target but it increases the cost and cycle time. If lots of measurements are skipped, the product quality does not be guaranteed due to process error from uncompensated drift and step disturbance. Thus, it is necessary to optimize the sampling plan in order to quickly identify the sources of prediction errors and decrease the metrology cost and cycle time. The goal of this research intend to understand the relationship between metrology and advanced process control (APC) in semiconductor manufacturing and develop an enhanced sampling strategy in order to maximize the value of metrology and control for critical wafer features.
Advanced tabulation techniques for faster dynamic simulation, state estimation and flowsheet optimization
(2009-08) Abrol, Sidharth; Edgar, Thomas F.
Large-scale processes that are modeled using differential algebraic equations based on mass and energy balance calculations at times require excessive computation time to simulate. Depending on the complexity of the model, these simulations may require many iterations to converge and in some cases they may not converge at all. Application of a storage and retrieval technique, named in situ adaptive tabulation or ISAT is proposed for faster convergence of process simulation models. Comparison with neural networks is performed, and better performance using ISAT for extrapolation is shown. In particular, the requirement of real-time dynamic simulation is discussed for operating training simulators (OTS). Integration of ISAT to a process simulator (CHEMCAD®) using the input-output data only is shown. A regression technique based on partial least squares (PLS) is suggested to approximate the sensitivity without accessing the first-principles model. Different record distribution strategies to build an ISAT database are proposed and better performance using the suggested techniques is shown for different case studies. A modified ISAT algorithm (mISAT) is described to improve the retrieval rate, and its performance is compared with the original approach in a case study. State estimation is a key requirement of many process control and monitoring strategies. Different nonlinear state estimation techniques studied in the past are discussed with their relative advantages/disadvantages. A robust state estimation technique like moving horizon estimation (MHE) has a trade-off between accuracy of state estimates and the computational cost. Implementation of MHE based ISAT is shown for faster state estimation, with an accuracy same as that of MHE. Flowsheet optimization aims to optimize an objective or cost function by changing various independent process variables, subject to design and model constraints. Depending on the nonlinearity of the process units, an optimization routine can make a number of calls for flowsheet (simulation) convergence, thereby making the computation time prohibitive. Storage and retrieval of the simulation trajectories can speed-up process optimization, which is shown using a CHEMCAD® flowsheet. Online integration of an ISAT database to solve the simulation problem along with an outer-loop consisting of the optimization routine is shown using the sequential-modular approach.
Control-friendly scheduling algorithms for multi-tool, multi-product manufacturing systems
(2011-12) Bregenzer, Brent Constant; Qin, Joe; Hasenbein, John J.; Edgar, Thomas F.; Hwang, Gyeong S.; Kutanoglu, Erhan; Bonnecaze, Roger T.
The fabrication of semiconductor devices is a highly competitive and capital intensive industry. Due to the high costs of building wafer fabrication facilities (fabs), it is expected that products should be made efficiently with respect to both time and material, and that expensive unit operations (tools) should be utilized as much as possible. The process flow is characterized by frequent machine failures, drifting tool states, parallel processing, and reentrant flows. In addition, the competitive nature of the industry requires products to be made quickly and within tight tolerances. All of these factors conspire to make both the scheduling of product flow through the system and the control of product quality metrics extremely difficult. Up to now, much research has been done on the two problems separately, but until recently, interactions between the two systems, which can sometimes be detrimental to one another, have mostly been ignored. The research contained here seeks to tackle the scheduling problem by utilizing objectives based on control system parameters in order that the two systems might behave in a more beneficial manner. A non-threaded control system is used that models the multi-tool, multi-product process in a state space form, and estimates the states using a Kalman filter. Additionally, the process flow is modeled by a discrete event simulation. The two systems are then merged to give a representation of the overall system. Two control system matrices, the estimate error covariance matrix from the Kalman filter and a square form of the system observability matrix called the information matrix, are used to generate several control-based scheduling algorithms. These methods are then tested against more tradition approaches from the scheduling literature to determine their effectiveness on both the basis of how well they maintain the outputs near their targets and how well they minimize the cycle time of the products in the system. The two metrics are viewed simultaneously through use of Pareto plots and merits of the various scheduling methods are judged on the basis of Pareto optimality for several test cases.
Decision-making frameworks for practical industrial applications in optimal process design and control
(2021-08-16) Costandy, Joseph Gamal Nessim; Baldea, Michael; Edgar, Thomas F.; Beaman, Joseph; Bonnecaze, Roger; Rochelle, Gary
While economics are the driving force behind many of the decisions made by industrial stakeholders, the methodologies employed to make high-level decisions often utilize heuristics that may not be quantitatively optimal. In this dissertation, I develop optimization-based frameworks that enable quantitatively driven high-level decision-making for two problems of practical industrial significance. In the first part of the dissertation, I address the problem of deciding the operating mode (batch or continuous-flow) of a chemical process, while taking into account the fundamental differences in the natures of the two operating modes (such as the batch advantage of utilizing reactors for the manufacture of multiple products, or the batch disadvantage of reactor cleanup between campaigns), the size and cost of the respective reactor units, and the potential use of reactor networks to optimize performance. I develop a first-principles-based non-dimensionalization algorithm that unifies the model for all reactor types and chemical systems from the two operating modes which enables direct performance comparisons between reactors of the two operating modes. In addition, I introduce a novel discretization method, the orthogonal collocation on finite elements for reactors (OCFERE), that allows the consideration of networks of reactors of either of the two operating modes, and I unify the description of the economics of the two operating modes. This results in a framework that encompasses the solution of a single optimization problem to make the decision about operating mode and find the optimal reactor network design. In the second part of the dissertation, I address the problem of quantifying the monetary value of improvements in process control. While methods have been developed for quantifying the value of control in the case of predominantly steady-state processes, there has been no attempt to quantify the monetary value of control for predominantly transient processes. I first review the problem, and highlight the relationship between optimal scheduling and process control for transient processes. Then, I utilize the general framework of integrated scheduling and control to develop novel performance functions that enable the quantification of the monetary value of control from a scheduling perspective for a predominantly transient process. I posit that the transition time between one product and the next in a production sequence can be used as a performance metric over which the value of control can be quantified.
Development of a feedforward laser control system for improving component consistency in Selective Laser Sintering
(2019-06-18) Phillips, Timothy Bryce; Beaman, Joseph J.; Milner, Thomas; Crawford, Richard; Seepersad, Carolyn; Fish, Scott
Selective Laser Sintering makes up a significant portion of the polymer additive manufacturing market and is often the process of choice for structurally significant polymer components. With its expanding market, especially among end-use components, comes a growing need for improving reproducibility. Components built using Selective Laser Sintering display a large range among their mechanical properties and it has been shown that the thermal history of the building process has a strong influence over these variations. Temperature fluctuations of just a few degrees can mean the difference between scrapped parts or those with excellent mechanical and dimensional properties. This dissertation will introduce a novel method of reducing temperature and mechanical variations among parts. Physical simulations and empirical measurements of laser-polymer interaction are evaluated and used to guide development of an advanced laser power controller. The feedforward control system developed uses thermal imagery and dynamic surrogate modeling to systematically modulate laser energy impinging on the polymer surface to homogenize post-sintering temperatures. Results from thermal and mechanical tests will be presented, showing the laser control system is capable of reducing standard deviations by up to 57% for post-sintering temperature and 45% for ultimate flexural strength.
Dynamic modeling of post-combustion amine scrubbing for process control strategy development
(2016-05) Walters, Matthew Scott; Rochelle, Gary T.; Edgar, Thomas F.; Baldea, Michael; Akella, Maruthi R; Chen, Eric
Intensified process designs with advanced solvents have been proposed to decrease both capital and operating costs of post-combustion carbon capture with amine scrubbing. These advanced flowsheets create process control challenges because process variables are designed to operate near constraints and the degrees of freedom are increased due to heat recovery. Additionally, amine scrubbing is tightly integrated with the upstream power plant and downstream enhanced oil recovery (EOR) facility. This work simulated an amine scrubbing plant that uses an intercooled absorber and advanced flash stripper configuration with aqueous piperazine to capture CO2 from a 550 MWe coal-fired power plant. The objective of this research was to develop a process control strategy that resulted in favorable closed-loop dynamics and near-optimal conditions in response to disturbances and off-design operation. Two models were created for dynamic simulation of the amine scrubbing system: a medium-order model of an intercooled absorber column and a low-order model of the entire plant. The purpose of the medium-order model was to accurately predict the absorber temperature profile in order to identify a column temperature that can be controlled by manipulating the solvent circulation rate to maintain a constant liquid to gas ratio. The low-order model, which was shown to sufficiently represent dynamic process behavior through validation with pilot plant data, was used to develop a plantwide control strategy. A regulatory control layer was implemented and tested with bounding cases that represent either electricity generation requirements, CO2 emission regulations, or EOR constraints dominating the control strategy. Satisfying the operational and economic objectives of one system component was found to result in unfavorable dynamic performance for the remainder of the system. Self-optimizing control variables were identified for the energy recovery flowrates of the advanced flash stripper that maintained good energy performance in off-design conditions. Regulatory control alone could not satisfactorily achieve the set point for CO2 removal rate from the flue gas. A supervisory model predictive controller was developed that manipulates the set point for the stripper pressure controller in order to control removal. The straightforward single-input, single-output constrained linear model predictive controller exhibited a significant improvement compared to PI control alone.
Dynamic modeling, optimization, and control of monoethanolamine scrubbing for CO2 capture
(2012-08) Ziaii Fashami, Sepideh; Rochelle, Gary T.; Edgar, Thomas F.; Seibert, A F.; Masada, Glenn Y.; Freeman, Benny D.
This work seeks to develop optimal dynamic and control strategies to operate post combustion CO2 capture in response to various dynamic operational scenarios. For this purpose, a rigorous dynamic model of absorption/stripping process using monothanolamine was created and then combined with a simplified steady state model of power cycle steam turbines and a multi-stage variable speed compressor in Aspen Custom Modeler. The dynamic characteristics and interactions were investigated for the plant using 30% wt monoethanolamine (MEA) to remove 90% of CO2 in the flue gas coming from a 100 MW coal-fired power plant. Two load reduction scenarios were simulated: power plant load reduction and reboiler load reduction. An ACM® optimization tool was implemented to minimize total lost work at the final steady state condition by adjusting compressor speed and solvent circulation rate. Stripper pressure was allowed to vary. Compressor surge limit, run off condition in rich and lean pumps, and maximum allowable compressor speed were found as constraints influencing the operation at reduced loads. A variable speed compressor is advantageous during partial load operations because of its flexibility for handling compressor surge and allowing the stripper and reboiler to run at optimal conditions. Optimization at low load levels demonstrated that the most energy efficient strategy to control compressor surge is gas recycling which is commonly applied by an anti-surge control system installed on compressors. Trade offs were found between initial capital cost and optimal operation with minimal energy use for large load reduction. The examples are, designing the stripper in a way that can tolerate the pressure two times larger than normal operating pressure, over sizing the pumps and over designing the compressor speed. A plant-wide control procedure was used to design an effective multi-loop control system. Five control configurations were simulated and compared in response to large load variations and foaming in the stripper and the absorber. The most successful control structure was controlling solvent rate, reboiler temperature, and stripper pressure by liquid valve, steam valve, and compressor speed respectively. With the investigated disturbances and employing this control scheme, development of an advanced multivariable control system is not required. This scheme is able to bring the plant to the targeted set points in about 6 minutes for such a system designed initially with 11 min total liquid holdup time.Frequency analysis used for evaluation of lean and rich tanks on the dynamic performances has shown that increasing the holdup time is not always helpful to damp the oscillations and rejecting the disturbances. It means there exists an optimum initial residence time in the tanks. Based on the results, a 5-minute holdup can be a reasonable number to fulfill the targets.
Electrical parameter control for semiconductor manufacturing
(2007-12) Schoene, Clare Butler, 1979-; Qin, S. Joe; Edgar, Thomas F.
The semiconductor industry is highly competitive environment where modest improvements in the manufacturing process can translate to significant cost savings. An area where improvements can be realized is reducing the number of wafers that fail to meet their electrical specifications. Wafers that fail to meet electrical specifications are scrapped, which negatively impacts yield and increases manufacturing costs. Most of the existing semiconductor process control research has focused on controlling individual steps during the manufacturing process via run-to-run control, but almost no work has looked at directly controlling device electrical characteristics. Since meeting electrical specifications is so critical to reducing scrap a fab-wide electrical parameter control scheme is proposed to directly control electrical parameter values. The goal of the controller is reducing the variation in the electrical parameters. The control algorithm uses a model to predict electrical parameter values after each processing step. Based on this prediction the decision to make a control move is made. If a control move is necessary, optimal adjustments for the subsequent processing steps are determined. The process model is continually updated so that it reflects the current process. A simple implementation using a least squares model is first proposed. Simulations and an industrial case study demonstrate the potential improvements that can be achieved with the algorithm and the limitations of the simple implementation are discussed. A partial least squares modeling and control algorithm combined with missing data algorithms are proposed as enhancements to the electrical parameter control algorithm to address many of the issues faced when implementing such a control strategy in real manufacturing environments. The enhancements take the input variable correlations into account when making control moves and utilize the correlation structure to make better model predictions. Simulations are performed to determine the effectiveness of the enhancements. A cost function formulation and a Bayesian based alternative are also presented and evaluated. The cost function implementation uses a different method to determine the optimal set points for the subsequent processing steps than the other implementations use. Simulations are used to compare the cost function formulation with the other methods presented. The Bayesian implementation addresses the stochastic nature of the manufacturing process by dealing with the probabilities of events occurring. A simulation of the Bayesian algorithm is preformed and the algorithms limitations are discussed.
Equation-oriented modeling, simulation, and optimization of integrated and intensified process and energy systems
(2016-12) Pattison, Richard C.; Baldea, Michael; Edgar, Thomas F.; Rochelle, Gary T; Bonnecaze, Roger T; Biros, George
Process intensification, defined as unconventional design and/or operation of processes that results in substantial performance improvements, represents a promising route toward reducing capital and operating expenses in the chemical/petrochemical process industry, while simultaneously achieving improved safety and environmental performance. In this dissertation, intensification is approached from three different angles: reactor design and control, process flowsheet design and optimization, and production scheduling and control. In the first part of the dissertation, three novel concepts for improving the controllability of intensified microchannel reactors are introduced. The first concept is a latent energy storage-based temperature controller, where a phase change material is confined within the walls of an autothermal reactor to improve local temperature control. The second concept is a segmented catalyst layer which modulates the rate of heat generation and consumption along the length of an autothermal reactor. Finally, the third concept is a thermally actuated valve, which uses small-scale bimetallic strips to modulate flow in a microchannel reactor in response to temperature changes. The second part of the dissertation introduces a novel framework for equation-oriented flowsheet modeling, simulation and optimization. The framework consists of a pseudo-transient reformulation of the steady-state material and energy balance equations of process unit operations as differential-algebraic equation (DAE) systems that are statically equivalent to the original model. I show that these pseudo-transient models improve the convergence properties of equation-oriented process flowsheet simulations by expanding the convergence basin in comparison to conventional steady state equation-oriented simulators. A library of pseudo-transient unit operation models is developed, and several case studies are presented. Models for more complex unit operations such as a pseudo-transient multistream heat exchanger and a dividing-wall distillation column are later introduced, and can easily be included in the flowsheet optimization framework. In the final part of the dissertation, a paradigm for calculating the optimal production schedule in a fast changing market situation is introduced. This is accomplished by including a model of the dynamics of a process and its control system into production scheduling calculations. The scheduling-relevant dynamic models are constructed to be of lower order than a detailed dynamic process model, while capturing the closed-loop behavior of a set of scheduling-relevant variables. Additionally, a method is given for carrying out these production scheduling calculations online and in "closed scheduling loop,"' i.e., recalculating scheduling decisions upon the advent of scheduling-relevant process or market events. An air separation unit operating in a demand response scenario is used as a representative case study.
Evaluation and extension of threaded control for high-mix semiconductor manufacturing
(2010-12) Patwardhan, Ninad Narendra; Flake, Robert H.; Edgar, Thomas F.
In the recent years threaded run-to-run (RtR) control algorithms have experienced drawbacks under certain circumstances, one such trait is when applied to high-mix of products such as in Application Specific Integrated Circuits (ASIC) foundries. The variations in the process are a function of the product being manufactured as well as the tool being used. The presence of semiconductor layers increases the number of times the lithography process must be repeated. Successive layers having different patterns must be exposed using different reticles/masks in order to maximize tool utilizations. The objectives of this research are to develop a set of methodologies for evaluation and extension of threaded control applied to overlay. This project defines methods to quantify the efficacy of threaded controls, finds the drawbacks of threaded control under production of high mix of semiconductors and suggests extensions and alternatives to improve threaded control. To evaluate the performance of threaded control, extensive simulations were performed in MATLAB. The effects of noise, disturbances, sampling and delays on the control and estimation performance of threaded controller were studied through these simulations. Based on the results obtained, several ideas to extend threaded control by reducing overall number of threads, by improving thread definitions and combination have been introduced. A unique idea of sampling the measurements dynamically based on the estimation accuracy is also presented. Future work includes implementing the extensions to threaded control suggested in this work in real production data and comparing the results without the use of those methods. Future work also includes building new alternatives to threaded control.
Low-cost Machine Vision Monitoring of the SLS Process
(1997) Gibson, Ian; Ming, Ling Wai
During the building of a part using SLS, it is common practice to adjust the temperature parameters. It is important to control these parameters because ifthey are too high then part breakout is difficult. Ifthey are too low then parts have poor material properties. One method of controlling these parameters is by observation through the process chamber window. Any adjustment can be determined by examining the colour ofthe cross-section in process. By using a machine vision system to determine colour variation, it is possible to calculate temperature or laser power adjustments necessary to maintain consistent part quality.
Modeling and optimization of process systems for unconventional technologies and feedstocks
(2020-04-14) Tsay, Calvin; Baldea, Michael; Biros, George; Edgar, Thomas F; Rochelle, Gary T
In the present era, the petrochemical/chemical process industries must adapt to unconventional feedstocks and energy sources, in order to keep pace with increased competition, regulatory pressure, and changing markets. However, developing processes compatible with these changes requires deviating from traditional and accepted process design and operation paradigms. This dissertation addresses fundamental challenges related to this transition from three angles: incorporation of custom (and detailed) models into process design, integration of variable operation with process design, and optimization of transient process operations. The first part of the dissertation introduces a framework for modeling, simulation, and optimization of process flowsheets incorporating highly detailed physical models of important and complex process units, termed “multi-resolution flowsheets”. The framework relies on pseudo-transient continuation as a numerical method and allows for the robust optimization of large-scale process models. Several case studies demonstrate the method, including process flowsheets featuring both intensified (e.g., dividing-wall distillation column, multistream heat exchanger) and unconventional (e.g., quenched reactor, packed column for carbon capture) process units. Furthermore, these results reveal significant benefits of considering the added level of detail at the design stage. Finally, an avenue is presented to accelerate the convergence of the pseudo-transient method, which is especially important for the large-scale models considered. In the second part of the dissertation, the focus shifts to process design optimization for variable operation, or optimization under uncertainty. Here, I present a method for process design that considers the effect of uncertain physical parameters (assumed to follow continuous probability distributions), using a formulation that exploits the semi-infinite nature of dynamic optimization. I compare the method to traditional “scenario-based” approaches using both theoretical analyses and multiple case studies. In addition to demonstrating the effectiveness of the proposed method, these case studies also emphasize the importance of considering several practically relevant uncertainties during process design. The final part of the dissertation examines explicit consideration of process dynamics for operational optimization. First, I examine periodic (dynamically intensified) processes, which operate at a cyclic steady state. I present a pseudo-transient method for robust optimization of fully discretized dynamic process models, and I present an approach for implementing cyclic conditions based on their fundamental relation to material/energy recycle loops. Lastly, I propose a framework for optimal production scheduling in fast changing market situations. Towards this end, I show how data-driven dynamic models can represent the behavior of a set of scheduling-relevant (physical or latent) variables. A method is also given for executing scheduling calculations using these models, and the framework is demonstrated by considering the demand response operation of both simulated and real-world air separation units.
Performance monitoring of run-to-run control systems used in semiconductor manufacturing
(2008-08) Prabhu, Amogh V., 1983-; Edgar, Thomas F.
Monitoring and diagnosis of the control system, though widely used in the chemical processing industry, is currently lacking in the semiconductor manufacturing industry. This work provides methods for performance assessment of the most commonly used control system in this industry, namely, run-to-run process control. First, an iterative solution method for the calculation of best achievable performance of the widely used run-to-run Exponentially Weighted Moving Average (EWMA) controller is derived. A normalized performance index is then defined based on the best achievable performance. The effect of model mismatch in the process gain and disturbance model parameter, delays, bias changes and nonlinearity in the process is then studied. The utility of the method under manufacturing conditions is tested by analyzing three processes from the semiconductor industry. Missing measurements due to delay are estimated using the disturbance model for the process. A minimum norm estimation method coupled with Tikhonov regularization is developed. Simulations are then carried out to investigate disturbance model mismatch, gain mismatch and different sampling rates. Next, the forward and backward Kalman filter are applied to obtain the missing values and compared with previous examples. Manufacturing data from three processes is then analyzed for different sampling rates. Existing methods are compared with a new method for state estimation in high-mix manufacturing. The new method is based on a random walk model for the context states. This approach is also combined with the recursive equations of the Kalman filter. The method is applied to an industrial exposure process by extending the random walk model into an integrated moving average model and weights used to give preference to the context that is more frequent. Finally, a performance metric is derived for PID controllers, when they are used to control nonlinear processes. Techniques to identify nonlinearity in a process are introduced and polynomial NARX models are proposed to represent a nonlinear process. A performance monitoring technique used for MIMO processes is then applied. Finally, the method is applied to an EWMA control case used before, a P/PI control case from literature and two cases from the semiconductor industry.
Some applications of statistical method to the solution of factory management problems, with special reference to Texas
(1949) Broom, H. N. (Halsey N.); Not available
The stability and performance of the EWMA and double-EWMA run-to-run controllers with metrology delay
(2004) Good, Richard Paul; Qin, S. Joe
Because of the ever-increasing demands on product quality, feedback con- trol has become a necessary enabling component in the manufacture of modern semiconductor devices. The nature of semiconductor manufacturing is such that measurements of device quality characteristics are not available during the processing of the product. Measurements are not made until after the product is processed and necessary changes to tool setting can only be made to subsequent production runs. This control scheme, termed run-to-run control, has become the cornerstone of process control in the semiconductor manufacturing industry. In addition to the ever-increasing demands on product quality, the semi- conductor manufacturing industry continues to see stringent growth in throughput requirements. Because of the demands on production throughput, it is rarely possi- ble to perform quality measurements on a batch of wafers before processing begins on the following batch of wafers. The delay between product manufacturing and product metrology coupled with inaccurate process models can lead to process in- stabilities and deterioration in controller performance. This dissertation investigates the robust stability requirements of processes controlled with EWMA and double- EWMA run-to-run controllers with delays between processing and metrology. In addition, the effects of model mismatch and metrology delay on the closed-loop performance of the EWMA and double-EWMA run-to-run controllers are derived by extending the robust stability methodology. Finally, these robust performance requirements are used to find the optimal tuning parameters for the double-EWMA controller. These tuning parameters allow for the largest model uncertainty while guaranteeing a predetermined minimum closed-loop transient performance.