Composition Measurement And Error Estimation Of Cu-Ni-Zn Alloy

National measurement standard is not available to master alloy but to final production, and it is the same to Cu-Ni-Zn alloy. The available measure methods were not including all compositions of that, but also bulk of method with instrumental analysis, which needs pretty penny for instrument and high wage for operationer. Aim of dissertation is to provide method what can measure composition of Cu-Ni-Zn alloy, which is facility, convenience, innocuity, and available for normal precision, or what can minish measurement error with mathematical methods.Based on the available measure method of Copper, Nickel, and Znic, complexometry is a practicable way. With the process of design, calculated for indicator, selected and calculated to buffer solution, and calculated for titration process control, titrimetric analysis with uncertainty evaluating was found, which is a low cost method with less error, easy operation, nonhazardous for environment, and need not operate expert.Dissertation contrived to calculate measure result with more than one demarcated-data by the method of linear regression. To calculate the error and uncertainty of measured result with three methods: single-demarcated, which demarcated only with one demarcated-data;proportional-demarcated, which demarcated with more than one demarcated-data by method of linear regression through zero perforce;linear-demarcated, which demarcated with more than one demarcated-data by method of linear regression. Results show that amount of measure error with linear-demarcated, both the absolute error and the relative error, is far less than that of the others;that amount of measure error with proportional-demarcated is not distinctly minished compare with that of single-demarcated. That system error was corrected only in linear demarcate is the reason that linear demarcate better than the others.Dissertation compared measure error of two demarcated-data with four demarcated-data with linear-demarcated. Measure error of two demarcated-data is more than that of four demarcated-data. But the labor of two demarcated-data is decreased half. Measure error with linear-demarcated of two is far less than that of single-demarcated, when the consumed volume of sample located two demarcated-data, and the gap of the two demarcated-data is enough.Dissertation contrived computer assistant analysis eventually to program the programmable data process, results calculate, and uncertainty evaluate. It enhanced reliability and efficiency, simultaneity decrease manual labor greatly, because decrease manual interfere.Method is available for composition of Cu-Ni-Zn alloy, Cu: 33.33 %~ 80.00%, Ni:10.00%~33.33%, Zn:10.00%~33.33%0...