Dominic Method

Reference cited in this section 1.S, Pugh, Total Design: Integrated Methods for Successfiul Product Development, Addison-Wesley, 1991 Decision Matrices Materials Selection David L, Bourell, The University of Texas at Austin The Dominic Method Description. The Dominic method (Ref 6) is used to evaluate altematives using five deseriptors: excellent, good, fai, poor, and unacceptable. Further, eriteria are rated qualitatively as being high priority, moderate priority, or low priority ‘The Dominic method, like the Pugh method, is completely qualitative, but it allows for more complexity in the selection process. Method. A maiix of altematives and criteria is constructed as described for the Pugh method, but no datum is selected. Altematives are labeled using capital letters (’ c.). Criteria are listed in descending order of importance to ,, moderate priority, or low priority (or, alternatively, “erucial,” "important," and "desirable". The matrix is completed by ranking each criterion on an absolute basis as excellent, good, fair, poor, or unacceptable for each alternative. the design and are then gathered into three categories: high pri ‘Once the matrix is complete, the results are tabulated using a Dominic method overall evaluation table (DMOET), Table 3. The rows are the ratings for alternatives and the columns are ratings for criteria. The blanks are filled in with the letters representing altematives. For example, if alternative B had three “excellent” ratings in high-priority criteria, three B's would be placed in the blank cell for "High" and "Excellent." Once the DMOET is complete, altematives are given an overall rating. All letters for an altemative must be above an overall rating line to obtain a given rating. The rating lines are shown on the DMOET. To obtain an "excellent" overall rating, the alternative must be rated as excellent in all high- priority criteria, good or excellent in all moderate-priorty criteria, and fai or better in the low-priorty criteria, ‘The results ‘ofthe Dominic method are analyzed as described for the Pugh method. Table 3 Dominic method overall evaluation table (OMOET) Individual Criteria porty valuations High Moderate Low ‘Unacceptable (Overall evaluation: all above heavy solid ine, excellent: all above heavy dashed line, good; all above light solid line, fair; al above light dotted line, poor Example 2: Use of the Dominic Method for Mate: Thermal Conductor in Late Concept Design. | Selection of a Loaded‘The objective in this example was to select a mass-produced, non-heat-treatable cylindrical sheath material to conduct heat efficiently near room temperature and to withstand a static load. A heat transfer application required that a sheet of ‘material be bent around a heat transfer medium. The sheet, whose thickness was defined by the necessary heat transfer conditions, also had to withstand a static compressive load and be resistant to denting during handling. Manufacturing issues were tooling cost and springback. Table 4 lists the materials that were initially considered and their properties. The criteria are listed in Table 5. High- priority criteria were the amount of heat transferred (thermal conductivity, high was desirable), tooling costs (bend force index, low was desirable), and springback (index, low was desirable). Medium-priorty criteria were the time to steady- state Conduction (thermal diffusivity, high was desirable) and the ability to carry a static load (stati load index, high was desirable). The low-priority criterion was the ability to resist denting during handling (hardness, high was desirable), The completed DMOET is shown as Table 6. Table 4 Properties of candidate materials for the Dominic example Properties A B 6 D E 1020 steel 304 sinless 5052-34 C11000-HO4-—_ 22000-1101 (ASTM A. Stel aluminum ETP copper commercial ou, (ASTM. A666, (ASTM B (ASTM B bronze rade B) rade D) 209) 152) (ASTM B36) ‘Thickness (9, m, 0.107 07 0407 0.208 0.208 ‘Thermal conductivity, Blu © fh +°F + 27 94 80 24 109 c ‘Thermal diffusiviey, (0 909 270 3750 6730) 3810 Blastic modulus (E), 10° psi 20 28 10s 7 7 Yield strength ('S), ksi 30 uo %6 40 3” Hardness, HB m 240 os 50 2 3210 11,770 10.580 8200 75390 Bend force index" th 38 1260) 410 1680 1350 Springback index", 10° in! 367 608 us 106 (a) ‘Staticoad-carying capability is proportional to YS(0), and lage numbers are desirable. Bending force is proportional to YS(n". Extent of springhnack is proportional to YS), Small numbers fr bth indices denote desirable manuatrabiiy‘Table 5 Criteria and ratings for candidate materials for the Dominic example teria Priority A. B © D E 1020 steel 304 stainless steel $052-H34——C11000-HOs 22000-8101 (ASTMAGII, (ASTM. A 666, aluminum ETP. copper commercial bronze rade B) grade D) (ASTMB209) (ASTM B 152) (ASTMB 36) Heat transfer High Good air Excellent Excellent Excellent ‘Tooling costs High Excellent Good Fie Good Good Springback High Excellent. Fair Excellent Excellent Exeelent Time to steady'state Moderate Good Fale Excellent Excellent Excelent Static load Moderate Good Excellent Excellent Excellent Excellent Dent resistance Low Good Excellent Pair Fair Fair ic method overall evaluation table (DMOET) for the loaded thermal conductor (Criteria Moderate High Overall evaluation: all above heavy solid tine, excellent; all above heavy dashed line, good: all above tight solid line, fair; ll above light dotted Tine, poor Analysis. No material received an overall “excellent” rating. Low-carbon stcel, copper, and bronze had "good" ratings Steel had the highest rating, but steel is limited by its heat transfer characteristics, a difficult parameter to improve on. ‘Aluminum and copper are limited by the high tooling cost associated with their thickness. Because the design stage was late conceptual, it was considered reasonable to continue with low-carbon steel, copper, and bronze into configuration design. Reference cited in this section 6. A.B. Howe, P.R. Cohen, J.R. Dixon, and M.K. Simmons, Dominic: A Domain-Independent Program for Mechanical Design, The International Journal of Artificial Intelligence in Engineering, Vol | (No. 1), 1986