This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Figure 36.1 A simple example of robust design. (a) Location of two mounting holes ona sheet-metal bracket where the deviation keeping the top surface of the bracket frombeing perfectly horizontal is ±α. (b) New location of holes where the deviation (keepingthe top surface of the bracket from being perfectly horizontal) is reduced to ±α/2.
Figure 36.2 (a) Objective-function value distribution of color density for televisionsets. (b) Taguchi loss function showing the average replacement cost per unit tocorrect quality problems. Source: Courtesy of G. Taguchi.
Figure 36.3 (a) A histogram ofthe number of shafts measuredand their respective diameters.This type of curve is called afrequency distribution. (b) Anormal distribution curveindicating areas within eachrange of standard deviation.Note: The greater the range, thehigher is the percentage of partsthat fall within it.
Figure 36.5 Control chartsused in statistical qualitycontrol. The processshown is in good statisticalcontrol because all pointsfall within the lower andupper control limits. In thisillustration, the ample sizeis 5, and the number ofsamples is 1.5.
Figure 36.6 Control charts. (a) Processnegins to become out of control because ofsuch factors as tool wear (drift). The tool ischanged and the process is then in statisticalcontrol. (b) Process parameters are not setproperly; thus, all parts are around the uppercontrol limit (shift in mean). (c) Processbecomes out of control because of factorssuch as a change in the properties of theincoming material (shift in mean).
Figure 36.7 Illustration of processes that are (a) unstable or out of control and (b)stable or in control. Note in sketch (b) that all distributions have lower standarddeviations and have means closer to the desired value. Source: After K. Crow.
Operating-Characteristics Curve used inAcceptance Sampling
Figure 36.8 A typical operating-characteristics curve used in acceptancesampling. The higher the percentage of defective parts, the lower is theprobability of acceptance by the consumer. There are several methods ofobtaining these curves.
Figure 36.11 Three methods of radiographic inspection: (a) conventional radiography,(b) digital radiography, and (c) computed tomography. Source: Courtesy of AdvancedMaterials and Processes, November 1990. ASM International.