9.7.3. DOE Method

Figure 9.81: Method tab in DOE properties

Figure 9.82: Description of Method properties in DOE

Parameter	Symbol	Description	Dimension (Range)
1. DOE Type	N/A	Use to set the method of DOE as shown in Figure 9.83: DOE Methods. The DOE Type and Level determine the number of trials and DOE table.	N/A
2. Level	N/A	Use to set the level which is related to the sample number of each design parameter.	N/A (Integer)
3. Run	N/A	Use to show or set the trial number. When DOE type is set to User Defined, the trial number can be defined by user.	N/A (Integer>0)
4. DOE Table	N/A	Use to show the set of design parameters for each trial or change the value of design parameter of the specific trial.	N/A (Real)
5. Check Box	N/A	Use to enable or disable a specific trail.	N/A
6. Design Parameter	N/A	Use to show the name of design parameter.	N/A
7. DOE and Result…	N/A	Use to define the file name and path for Simulation Output Files. The DFDOE file is generated during the batch simulation including other simulation result files. The name of simulation result is determined as follows. "use defined name" + "_Case" + number	N/A
8. Append data and…	N/A	Use to append the date and time to the result file name. For more information, see Simulation Configuration Parameters.	N/A
9. Create result files…	N/A	Use automatically to create the DFG, DFR and DFM files for the Motion Postprocessor. For more information about the files, see Simulation Output Files.	N/A
10. Create model	N/A	Use to create the models for all trials.	N/A
11. Run	N/A	Use to run the DOE simulation. First the models are generated for all trials. And then, stand alone solver is carried out for the trials.	N/A
12.	N/A	Use to export the files of DFG, XML, and DFS or open Simulation Configuration Parameters.	N/A
13. OK/Cancel	N/A	Use to close the DOE simulation dialog with saving or without saving the parameters.	N/A

Figure 9.83: DOE Methods

DOE Type

Supported Level

Description

1. Parameter Study

2, 3, 4, 5

Parameter study is the single factor design and considers all levels for each design parameters without an interaction between design parameters. The number of trial can be determined with the number of design parameters and level as follows. If the level is odd, DOE Methods

and if even, DOE Methods

,, where

and

are the level and number of design parameters.

2. Full Factorial

2, 3

Full factorial design considers all combinations of design parameters and their variances. The number of trials can be determined with the number of design parameters and level as follows. DOE Methods

, where

and

are the level and number of design parameters.

3. Plackett Burman

2

Plackett Burman is a two-level fractional factorial design, proposed by R. L. Plackett and J. P. Burman in 1946, where only a few specifically chosen trials are performed to investigate just the main effects. The number of trials can be determined from the number of design parameters as follows: DOE Methods

, where

is the number of design parameters and the function of "INT" takes the integer value (dropping the decimal for a whole number value) of the term expressed as a fraction.

4. Orthogonal Array

2, 3

Orthogonal arrays are one of highly fractional designs, used to estimate main effects using only a few experimental trials. This design supports two and three levels but is also available to investigate main effects for certain mixed level trials where do not have the same number of levels. The trial number of two level is equal to that of Placket Burman and the trial number of three level is determined with one of 9, 18, 27, 36, 54, and 81. The trial number is dependent on the number of design parameters as follows.

Number of Design Parameter	Number of Trial
2 ~ 4	9
5 ~ 7	18
8 ~ 13	27
14 ~ 22	36
23 ~ 26	54
27 ~ 40	81

The maximum numbers of design parameter are 31 in two level and 40 in three level, respectively.

5. User Define

N/A

"User Define" supports the user defined DOE table by setting the number of "Run".