It is important to note that the parts used in a product are divided into standard or off-the-shelf parts and nonstandard or designed parts. Examples of standard parts are bearings, shafts, pulleys, belts, screws, nuts, snap rings, etc. These parts come with the manufacturers specified tolerances. The designer does not have any latitude in changing these limits. Nonstandard or designed parts are custom made for the product. Hence, the designer can specify wider or narrower limits based on the functionality requirements. There are two types of tolerance analysis: extreme case tolerance analysis and statistical analysis.
Extreme case analyses are further subdivided into two categories: best-case analysis and worst-case analysis.
• Best-case analysis describes a situation in which the parts fall on that side of the tolerance (positive or negative) in which there is no chance of interference in the assembly of these parts.
• Worst-case analysis is the study of a situation wherein the parts produced are assembled as per the worst case. The probability of interference is certain or unity.
The extreme case analysis method is the most widely used method for tolerance analysis. Most designs are analyzed using this concept and have worked successfully. The method is simple to apply and consists of designing the parts to nominal dimensions and then assigning tolerances in such a way that if tolerances accumulate in one direction or the other, the assembly continues to meet the functional requirements of the design. This method, though ensuring that all parts will always be able to be assembled correctly, has a built-in waste mechanism. Designs can be overly conservative, leading to high product costs by assigning tighter tolerance zones. By using statistical analysis based on six sigma, a more reasonable understanding of the design specifications and how parts will be assembled will be demonstrated in the next section.