Cost analysis can be completed at different levels of detail. We will consider three categories of cost analysis (Meeker and Thornton 1996). A Level 1 Estimate is a first impression by a knowledgeable engineer of what a part, assembly, or system would cost, based on prior experience. This would involve less than 10 minutes of work for a system of 50 parts and is generally accurate to within 20%.
A Level 2 Analysis is an estimate based on a breakdown of costs and itemized cost data based on prior experience with similar products, budgetary estimates, vendor quotes, and expert opinion and experience. This would involve a day of work for a system of 50 parts and is generally accurate to within 5%. The methods in this chapter are generally Level 2 analyses.
A Level 3 Cost Accounting applies a detailed costing of every part, accomplished by using material cost estimation data bases and time/motion studies. A high degree of accuracy is achieved by comparisons to industry standards and requests for vendor quotes. This would involve a week of work for a system of 50 parts and is generally accurate to within 1 %. Many major manufacturers have cost analysis teams that provide their product development teams with Level 3 cost analyses of their competitors' products (Tsiantar).
We now present cost models of different production processes. First, we provide a very simplistic model of part costs by providing tables of different parts. Then, we provide a simple assembly cost model. These can be used to cost account in a crude way as a Levell cost estimate.
Basic DFM Part Cost Method: Analogy Approach
For piece part manufacturing, we now present different cost models for different piece part production processes. Table 14.2 lists typical material costs, Table 14.3 lists typical processing labor rates, Table 14.4 lists typical processing times, and Figures 14.15-14.19 list typical tooling costs, all in 1998 dollars. The data in these tables can be used in association with the previous formulas. That is, for any particular part, one can compare the part on material, processing, and tooling to those listed in the tables and use an equivalent value.
To determine an estimate for plastic injection -molded parts, a cost per part can be gathered by estimating of the material, processing, and tooling costs. The material costs can be determined as before. The processing time can be determined using Table 14.4, and a labor rate applied from Table 14.3. To estimate the tonnage machine, a reasonable estimate is a 100- or 200-ton machine, or for large manufacturers, larger 500-ton machines. An equation to determine an effective number of cavities for a mold compared to the press tonnage is:
where A is the projected area of the part along the mold parting line and σ is a material factor (tons/cm2 ) and is given approximately by Table 14.5.
The tooling cost can be determined using Figure 14.15 and Tables 14.5 and 14.6, in conjunction with Eq. (14.5) above.
Die cast parts:
To determine an estimate for die cast parts, a cost per part can be gathered by estimating the material, processing, and tooling costs. The material costs can be determined as before. The processing cost can be determined using Tables 14.3 and 14.4. The tooling cost can be determined using Figure 14.17.
Powdered-metal parts: To determine an estimate for powdered metal parts, a cost per part can be gathered by estimating of the material, processing, and tooling costs. The material costs can be determined as before. The tooling cost can be determined using Figure 14.19. The processing cost is a combination of the green mold filling, an oven sintering operation, and then subsequent finishing operations. Generally, the process times are dominated by the firing times, given by Table 14.7. The total processing time is given by Table 14.4.
Cardboard packaging material costs: Before a product ships, it must be put into retail packaging. This can become surprisingly expensive, especially when considering the assembly costs of inserting the product into the packaging. A formula for the cardboard shipping container packaging itself, with no printing on the packaging, is
where V is the envelope volume of the product, N is the number of products per container, and C is in 1998 dollars (adapted from Santandreu) .
More accurate cost accounting:This chapter has presented some simple formulas for estimating costs of some manufactured components. These are very approximate; more detailed models are available to obtain a more accurate analysis, such as Boothroyd and Dewhurst's DFM suite of analysis tools, Galorath's SEER-DFM, (Galorath), Cognition Systems' cost estimating tools, and a wide range of university developed software tools.
Beyond the selected materials and processes, there are a large set of other anufacturing processes and materials. Each can be cost-estimated in a similar way, by first determining the cost drivers of the process and then establishing cost coefficients against these drivers. Cost estimating is a well established profession. Every bid submitted on any job has a professional cost estimate developed with cost analysis tools, and professionals strive to improve these tools to submit good cost estimates. Many job shops and industry groups are willing to share these tools.
We now turnto estimating the assembly cost of a product design.
Basic Assembly Method: Adapted Xerox Producibility Index
We present here a basic method for determining an indented manufacturing assembly cost model. That is, an account of the costs built up in the product hierarchically by subassembly down to individual parts. The method is a combination of assembly trees developed by Ishii (1994), and an adapted version of the Xerox Producibility Analysis (XPI), originally developed at the Xerox Corporation (Lewis; Waterbury), and an experimentally determined time equivalent labor factor against the XPI score.
Assembly trees: The first step in establishing an assembly cost model is to establish the assembly sequence hierarchy. Using a tree diagram with the final product as the trunk and each attached part as a leaf node, one should diagram what is attached to what. Exploded view diagrams are very helpful in constructing the assembly tree. On the assembly tree, one can characterize each assembly step by indicating fixturing needs (the symbol "F"), reorientation (the symbol "R"), and insertion directions (straight and rotational arrows). This information feeds directly into assembly analysis worksheets. As an ex ample, Figure 14.20 depicts an exploded view and assembly tree for the Krups Coffee Mill (Chapter 7).
Xerox producibility analysis: To estimate assembly difficulty, one very simple method is to use a variation on the Xerox Producibility Index (XPI) as an assembly difficulty indicator. The subsequent material is adapted from the original method developed at the Xerox Corporation (Lewis; Waterbury). It is a very simplified design for assembly method requiring simple tabulation of efficiency values.
V. CHAPTER SUMMARY AND "GOLDEN NUGGETS"
Design for manufacture and assembly is the most effective methodology to reduce product cost. Golden nuggets that should be retained from this chapter include:
The basic techniques to improve a design are mostly a collection of common-sense rules.
One can determine the most effective approach for redesign through stacked-up cost analysis.
Modularize to minimize part count, design for top-down insertion with alignment features.
Think thoroughly and simplify the fabrication difficulty of each feature on every part.