Tsung-Nan Tsai

Modeling and optimization of stencil printing operations: A comparison study

This paper presents a comparison study for the optimization of stencil printing operations using hybrid intelligence technique and response surface methodology (RSM). An average 60% of soldering defects are attributed to solder paste stencil printing process in surface mount assembly (SMA). The manufacturing costs decrease with increasing first-pass yield in the stencil printing process. This study compares two hybrid intelligence approaches with RSM as methods of solving the stencil printing optimization problem that involves multiple performance characteristics. The optimization process is threefold. A data set obtained from an experimental design following data preprocessing process provides an accurate data source for RSM study and training neural networks to formulate the nonlinear model of the stencil printing process with/without combining multiple performance characteristics into a single desirability value, followed by a genetic algorithm searching the trained neural networks for obtaining the optimal parameter sets. The empirical defect-per-million-opportunities (DPMO) measurements demonstrate that the two hybrid intelligence methods can provide satisfactory performance for stencil printing optimization problem.

Selection of the optimum promotion mix by integrating a fuzzy linguistic decision model with genetic algorithms

Integrated marketing communication (IMC) is an important process by which a company can influence a target market, improve the position of that companys product/service in the target market, and effectively build up its brand image. Sales promotion is an important communication channel designed to influence the customers purchasing behavior in the target market. There are many promotion tools available. Variations in business objectives and budgetary limits make it impossible for a company to employ all these promotion tools to convey sales messages to the customers. The selection of the best mix of promotion tools involves subjective information processing, instead of a numerically expressed objective decision-making process. In this research, we integrate a fuzzy linguistic decision model with a genetic algorithm (GA) to extract the optimum promotion mix of a variety of tools under satisfying expected marketing performance and budget limitations. The proposed methodology shows satisfactory results in an empirical study in terms of estimating the degree of satisfaction for achieving the business objectives, determining the optimum promotion mix, and minimizing expenditure on sales promotion activities.

A neural network-based prediction model for fine pitch stencil-printing quality in surface mount assembly

The soldering problems in surface mount assembly can represent considerable production cost increases and yield loss. About 60% of the soldering defect problems can be attributed to the solder paste stencil printing process. This paper proposes to solve a solder-paste stencil-printing quality problem by a neural network approach. Employment of a neuro-computing approach allows multiple inputs to the generation of multiple outputs. In this study, the inputs are composed of eight important factors in modeling the nonlinear behavior of the stencil-printing process for predicting deposited paste volumes. A 3^8^-^3 fractional factorial experimental design is conducted to efficiently collect structured data used for neural network training and testing. The results show that the proposed neural-network model is effective in solving a practical application.

A hybrid intelligent approach for optimizing the fine-pitch copper wire bonding process with multiple quality characteristics in IC assembly

Gold is the primary material used for wire bonding in integrated circuit (IC) assembly. Owing to the high appreciation in the price of gold, copper (Cu) wire has become an important substitute material in order to save on manufacturing costs. However, an average of 40% in yield loss during IC assembly can be attributed to improper control of the Cu wire bonding process. To assure cost savings without losing yield, and ensure cost-effective IC assembly, optimization of the parameters for the Cu wire process is critical. This work proposes a hybrid intelligent approach to derive robust parameter settings for a fine-pitch Cu wire bonding process with multiple quality characteristics. The proposed methodology utilizes grey relational analysis and an entropy measurement method to convert the multiple responses into a single synthetic performance index without involving the subjective judgment of an engineer and causing unbalanced improvements of the responses. An integrated neural network model and genetic algorithm method is then applied to acquire the optimal parameter settings. The performance of this method is evaluated experimentally and the results compared with that of the response surface methodology and original parameter settings. The results confirm the feasibility and practicality of this strategy to improve production yield and process capability during Cu wire bonding.

Modeling and Optimization of Reflow Thermal Profiling Operation: A Comparative Study

ABSTRACT In this study, a comparative study of optimizing the reflow thermal profiling parameters using a hybrid artificial intelligence and the desirability function approaches without/with combining multiple performance characteristics into a single desirability is presented. Reflow soldering is the key determinant for the improvement of the first-pass yields of electronics assembly. A reflow thermal profile is a time-temperature contour with multiple performance characteristics utilized to monitor the heating effects on a printed circuit board (PCB) and surface mount components (SMCs) in the reflow oven. The use of an inadequate reflow thermal profile may not only produce a variety of soldering failures, but can also result in the needs for considerable reworking and waste. An L18 (21*37) Taguchi experiment design is conducted to collect the thermal profiling data. A quick propagation (QP) neural network is modeled based on experimental data to formulate the nonlinear relationship between the thermal p...

Development of a soldering quality classifier system using a hybrid data mining approach

Soldering failures lead to considerable manufacturing costs in the electronics assembly industry. Soldering problems can be caused by improper parameter settings during paste stencil printing, component placement, the solder reflow process or combinations thereof in surface mount assembly (SMA). Data mining has emerged as one of the most dynamic fields in processing large manufacturing databases and process knowledge extraction. In this study, the integration of a probabilistic network of the SMA line and a hybrid data mining approach is employed to identify soldering defect patterns, classify soldering quality, and predict new instances according to significant process inputs. The hybrid data mining approach uses a two-stage clustering method that utilizes the self-organizing map (SOM) to derive the preliminary number of clusters and their centroids from the statistical process control (SPC) database, followed by the use of K-means to precisely classify instances into definite classes of soldering quality. The See5 induction system is then applied to induce the decision tree and ruleset to elucidate associations among the defect patterns, process parameters, and assembly yield. Finally, visual C++ programming codes are implemented for both production rule retrieval and graphical user interface establishment. The effectiveness of the proposed classifier is illustrated through a real-world application to resolve practical manufacturing problems.

Thermal parameters optimization of a reflow soldering profile in printed circuit board assembly: A comparative study

This paper presents a comparative study for optimizing the thermal parameters of the reflow soldering process using traditional and artificial intelligence (AI) approaches. High yields in the reflow soldering process are essential to a profitable printed circuit board (PCB) assembly operation. A reflow thermal profile is a time-temperature graph which is used to properly control the thermal mass and heat distribution to form robust solder joints between the PCB and electronics components during reflow soldering. An inhomogeneous temperature distribution for a reflow thermal profile can cause various soldering defects, which can jeopardize product reliability and lead to significant productivity loss. In the multi-objective optimization problem, three alternative optimization methods are discussed and compared: response surface methodology (RSM), nonlinear programming (NLP), and a hybrid AI technique. A dataset was gathered using a 3^8^-^4 experimental design for the development of meta-models through response surface quadratic modeling. In the first method, RSM is used to acquire the optimal heating parameters, while in the second method NLP is used to derive a global solution based on the meta-models. The back-propagation neural network (BPN) is used in the third method to formulate the nonlinear relationship between the heating inputs and responses. A genetic algorithm (GA) is then used to elicit the optimal heating parameters from the established BPN model. The evaluation results show that all three methods provide satisfactory soldering performance in terms of the process capability, sigma level, and process window indices (PWIs). Particularly, the hybrid AI approach provides superior nonlinear formulation capability and optimization performance.

Modeling and implementation of a neurofuzzy system for surface mount assembly defect prediction and control

A high-speed surface mount assembly can reduce both production cost and time; however, it could allow an enormous number of boards to be built before a problem is detected. Therefore, early detection and assessment of a surface mount assembly problem is critical for cost-effective manufacturing. This paper proposes a neurofuzzy system for surface mount assembly defect prediction and control. Hybrid data from both in-process quality control database and from a fractional factorial experimental design are collected for neurofuzzy learning and modeling. Customized programming codes are generated for rule retrieval and for graphical user interface modeling. The proposed system is successfully implemented at a surface mount assembly plant, ll significantly improves plant throughput by the downtime reduction that is a result of a better defect prediction and control.

Robust parameter design for the micro-BGA stencil printing process using a fuzzy logic-based Taguchi method

Display Omitted The stencil printing parameters for a micro-BGA package were robustly optimized.Three optimization methods were compared to find the best stencil printing parameter set.The three methods provide satisfactory performance compared to the mass production.The proposed fuzzy-based Taguchi method outperforms the other two methodologies.Taguchi-fuzzy method improves the outputs of paste volume and centroid by 23.27% and 27.47%. Solder paste is the main soldering material used to form strong solder joints between printed circuit boards (PCB) and surface mount devices in the surface mount assembly (SMA). On average 60% of end-of-line soldering defects can be attributed to inadequate performance of solder paste stencil printing. Recently, lead-free solder paste has been adopted by electronics manufacturers in compliance with the RoHS directive. However, soldering defects in the ball grid array (BGA) packages used in lead-free SMA have become more prevalent and are difficult to detect. In this study, a fuzzy logic-based Taguchi method is proposed to optimize the fine-pitch stencil printing process with multiple quality characteristics for the micro ball grid array (micro-BGA) packages using a lead-free solder paste. A structured data set is first collected from an L18 (21?37) fractional factorial design experiment, followed by multi-response optimizations and analysis of variance (ANOVA) for identifying significant factors. The optimization performance gained by the proposed fuzzy logic-based Taguchi method is compared with the results of other two hybrid methods including a combination of neural networks and genetic algorithms, and the integration of the response surface methodology with a desirability function. The confirmation experiments show that the proposed fuzzy logic-based Taguchi method outperforms the other two methods in terms of the signal-to-noise ratios and process capability index.

A neuro‐computing approach to the thermal profile control of the second‐side reflow process in surface mount assembly

Purpose – A neural‐network‐based predictive model is proposed to model the second‐side thermal profile reflow process in surface mount assembly with a view to facilitating the oven set‐up procedure and improving production yield.Design/methodology/approach – This study performs a 38−4 fractional factorial experimental twice to collect the thermal‐profile data from a second‐side board. The first experiment has components on the second side only, while the second experiment also has additional components on the primary side. A back‐propagation neural network (BPN) is then used to model the relationship between control variables and thermal‐profile measures.Findings – Empirical results illustrate the efficiency and effectiveness of the proposed BPN in solving the second‐side thermal‐profile prediction and control problem.Originality/value – There is no study dedicated to the investigation of the second‐side thermal‐profile variance with and without the presence of primary‐side components. The study suggests ...