Manufacturing Task Process Characterization Utilizing Response Surface Methodology

Manufacturing Task Process Characterization Utilizing Response Surface Methodology

Janet H. Sanders (East Carolina University, USA) and Silvanus J. Udoka (North Carolina A&T State University, USA)
Copyright: © 2012 |Pages: 16
DOI: 10.4018/jgc.2012070105
OnDemand PDF Download:
No Current Special Offers


To meet today’s business culture of rapid deployment of new products and processes, engineering and manufacturing personnel must utilize efficient means for process development. This paper discusses a novel approach to characterize a task driven manufacturing process. The approach utilized Response Surface Methodology (RSM) to investigate, identify, and prioritize the key process drivers and subsequently develop quantifiable methods for setting the operating levels for the process drivers to determine if the current levels of these key process drivers result in a process response value that is near optimum. The approach identifies the improved response region, generates a mathematical model of the process and specifies an operating window that would yield consistent results for each of the process drivers. A High Strength Fiber Splicing process was used to demonstrate this approach. This study led to the identification of the region that improved the process yield from 65% to 85%.
Article Preview


To meet the increasingly demanding mandates of rapid deployment of new products and processes, engineering and manufacturing personnel must utilize efficient means for process development. Response Surface Methodology (RSM) uses an assortment of mathematical and statistical techniques that explore the relationship between several input variables and one or more response variables (Box & Wilson, 1951; Hicks, 1982; Myers & Montgomery, 2002; Gryna, Chua, & Defeo, 2007; Allen, 2010). RSM is a sequential technique that is completed in three phases - Phase Zero, Phase One, and Phase Two (Myers & Montgomery, 2002). Phase Zero involves the initial screening experiment(s) designed to investigate the process variables, identify the key ones and, eliminate the unimportant ones. In Phase One, the primary objective is to determine if the current levels of the independent or key process variables result in a value of the response that is near optimum. Phase Two seeks to develop a mathematical model of the process that will accurately approximate the true response function within a small region around the optimum.

In this study, RSM was used to characterize a High Strength Fiber Splicing (HSFS) process. As shown in Figure 1, the HSFS process is partially automated with a moderately intense manual component.

Figure 1.

High strength fiber splicing process diagram


In Step 1 the automated equipment is set to specified settings. In Step 2, the worker obtains approximately one meter of coated optical fiber. In Step 3 the outer jacket and buffer tube are manually stripped from one end of the fiber to allow access to the fiber coating (Figure 2). Then the stripped fiber is placed into a fiber holder in Step 4. The fiber holder is a fixture used to hold the manually stripped fiber securely and to align both ends of the fiber on each piece of equipment as it proceeds through the HSFS processes. A mechanical fiber stripper strips the optical fiber's protective coating down to the fiber cladding in Step 5. This piece of equipment uses a combination of heat and metal blades to carefully strip the glass from the fiber. In Step 6 an ultrasonic cleaner removes the stripped glass and other contaminates from the fiber. This process applies ultrasonic cleaning action in a cleaning solution to achieve the required level of cleaning. In Step 7, a mechanical fiber cleaver cuts the fiber to the specified angle in preparation for splicing. The cleaver uses a diamond-tipped ultrasonic blade set at a specified tension to precisely cut the fiber. Following the cleaving, the fiber is placed on the fiber splicer in Step 8. The electric arc-fusion fiber splicer uses metallic electrodes to supply an electric charge to splice and fuse ends of fiber together. In Step 9, if both ends of fiber are not loaded onto the splicer, Steps 4 to 7 are repeated for the other end of the fiber. The two ends of fiber are spliced in Step 11. After the ends are successfully spliced, the fiber is loaded onto the proof tester (Step 12) and the tensile strength is determined (Step 13).

Figure 2.

Diagram of fiber optic cable (ARC Electronics, 2006)


Prior to embarking on this study, the yield for the 200 thousand pounds per square inch (kpsi) tensile strength specification for the HSFS process was approximately 65%. Due to the significant costs associated with the yield losses, the management at the company was seeking significant improvement in the yield.

Complete Article List

Search this Journal:
Open Access Articles: Forthcoming
Volume 10: 1 Issue (2019)
Volume 9: 2 Issues (2018)
Volume 8: 2 Issues (2017)
Volume 7: 1 Issue (2016)
Volume 6: 2 Issues (2015)
Volume 5: 2 Issues (2014)
Volume 4: 2 Issues (2013)
Volume 3: 2 Issues (2012)
Volume 2: 2 Issues (2011)
Volume 1: 2 Issues (2010)
View Complete Journal Contents Listing