We have recently published an updated version of the Design of Experiments (DOE) App for OriginPro version 2024.
The two major improvements in this update are:
- We have introduced “available design tables” in Factorial Design and Response Surface Design. The available runs with different number of factors and fractions are displayed to help user select an appropriate design. For 2-Level Factorial Design and Central Composite Design, you can choose a design with default settings, and then tune the settings.
- Design parameters can now be saved and loaded. In Power and Sample Size and Create Design tools, the Dialog Theme tab provides two buttons, Save and Load, with which you can save current parameters as an xml file, and later import it into the dialog.
We now look at a couple of examples that make use of the improvements mentioned above.
Example of 2-Level Factorial Design
We want to conduct a series of experiments to study the effects of two detergent A and B, and the effects of temperature, concentration and period of time on dirt removal (measured in mg).
Open Factorial Design and choose 2-Level Factorial(default generators). On Factors tab, set Number of Factors to 4. Enter factor names “detergent”, “temperature”, “concentration” and “time”. Select factor types (Text for detergent, and Numeric for the rest). Then Enter low and high values for each factor(detergent:A/B, temperature:30/50, concentration:0.05/0.1, time:15/30).
Click Display Available Designs. The table summarizes available runs for number of factors and fraction combinations. For 4 factors, IV resolution (8 runs) and full resolution (16 runs) are available . Because of limited experimental resources, we select IV resolution.
Go to Settings tab and click on the 1st row in the table to choose 1/2 fraction with 8 runs. You can also add more blocks or replicates to the design by changing the value of Number of Blocks and Number of replicates. The number of total runs will be updated simultaneously to help you evaluate the size of experiment.
To save current parameters, go to Dialog Theme tab and click Save button. Choose a location to save as an xml file. You can load it later by clicking Load->Open.
Click OK to output the design table.
Example of Central Composite Design
Response surface design can be used to estimate curvature in the model. It is usually performed after significant factors are identified by factorial design experiment.
A chemical engineer is interested in determining the operating conditions that maximize the yield of a process. Four factors have been selected by previous experiments: temperature, time, power and ethanol percentage.
Open Central Composite Design. On Factors tab, set Number of Continuous Factors to 4. Enter factor names “temperature”, “time”, “power ” and “ethanol “. Then Enter low and high values for each factor (temperature:25/75, time:180/240, power :15/45, ethanol :60/80).
Click Display Available Designs. The table displays available central composite and Box-Behnken designs.
Go to Settings and click on the 2nd row in the table to select a full factorial design with 2 blocks. The design is blocked into a cube block and an axial block. By default, there are 6 center points in total (4 in cube block, 2 in axial block). To reduce the number of center points, enable Customize Number of Center Points and set Cube block to 2 and axial block to 1.
Click OK to output design tables. The Blocks column split the design into two blocks: 1 represents cube block and 2 represents axial block. The Point Type column includes 1, 0, and -1, representing cube point, center point and axial point respectively.