Design of Experiments offers four (4) experiment types, the selection of which depends on your personal objective.
One variable calculates the effect of a single variable of your choosing on part quality criteria that you select. This is the fastest of the DOE experiments and is a good option if you are interested in the effect of only one variable. The results of this experiment are written in the Analysis Log. XY plots of the data can be accessed in the Results section of the Study Tasks pane.
Variable Influences calculates the relative influence of each of the variables that you are interested in on the part quality criteria you select. The variables are ranked according to their relative impact, with those variables having the most significance given a higher percentage than those that have less. Since the results are strictly percent rankings, they are written to the Analysis Log; no plots are drawn.
Variable Influences is the recommended experiment to run if you are unsure which variables to monitor. The DOE solver launches an optimized set of analyses to determine this ranking, and from the results you can decide if you need to monitor all the selected variables, or a subset of them. You can select as many input variables as you like for this experiment.
If you know which variables to monitor, select Variable Responses to determine their effect on each quality criterion that you are interested in. In this experiment, a larger set of experiments is run than for Variable Influences, to test, extensively, various combinations of the input variables. As a result, this experiment takes more time than the previous two experiments. You can select as many input variables as you like for this experiment.
The results of this experiment are written in the Analysis Log, and 2D/3D response surface plots of the data can be accessed in the Results section of the Study Tasks pane. The results can then be examined graphically to determine the optimum conditions.
Variable Influences then Responses uses the Taguchi method to determine which variables have the most influence on specific quality criteria, and then runs extensive factorial experiments on the most significant input variables to determine how they impact part quality. This option should be selected if you are interested in the effect on part quality of the various input variables, but are unsure which input variables have the most significant effect. The minimum number of input variables for this experiment is three (3). If you leave the default as three, but select more than three, all the selected variables will be ranked using the Taguchi method, then the three variables with the greatest influence will be used for the Responses experiments.
The rankings of the various input variables are written to the Analysis Log first, followed by the responses data. 2D/3D response surface plots of the data can be accessed in the Results section of the Study Tasks pane. The results can then be examined graphically to determine the optimum conditions.