Category Archives: Robust Parameter Experiments

Analysis of a Combined-Array Design in A Robust Parameter Experiment

This posting presents the analysis of results from the Robust Parameter Design Experiment introduced in the previous posting.  The five factors are A = CO2 pressure (bar), B = CO2 temperature oC, C = peanut moisture (% by wt), D = CO2 flow rate (liters/min), and E = peanut particle size (mm).   The factor E is the noise factor).   The purpose of the experiment is to show the effects of these factors on Solubility, S, or mg of oil removed from the peanuts.  Continue reading Analysis of a Combined-Array Design in A Robust Parameter Experiment

Estimating Interaction Effects using a Combined-Array Design in Robust Parameter Experiments

This posting presents another example illustrating the advantage of combined-array experiments in Robust Parameter Designs over the Taguchi crossed-array designs.  Kilgo (1988)  presents an example of a 25-1 fractional factorial experiment providing data to construct a model for estimating the mean response.  We modify the use of the experiment to make it relevant to a Robust Parameter Design experiment. Continue reading Estimating Interaction Effects using a Combined-Array Design in Robust Parameter Experiments

Combined Array Designs in Robust Parameter Experiments

This posting introduces the use combined arrays in Robust Parameter Designs. Combined array designs have both controllable and noise factors in the same experimental design.   The previous posting describes Taguchi Parameter Designs using crossed arrays consisting of an inner array containing the controllable factors and an outer array containing the noise factors. Continue reading Combined Array Designs in Robust Parameter Experiments

Crossed Array Design Problems in Robust Parameter Experiments

This posting describes problems in using Taguchi Parameter designs with crossed arrays in Robust Parameter Designs.  The two previous postings describe an application of Taguchi Parameter Designs to reduce plasma cutter cycle time.  That is, the Taguchi Design posting and the Taguchi Results posting.  The crossed array designs proposed by Taguchi when used with the maximum allowable factors can’t estimate the interaction effects among the controllable factors. Continue reading Crossed Array Design Problems in Robust Parameter Experiments

Plasma Cutter Cycle Time Experimental Design Results

This post describes the results and their analysis of the experimental design to reduce plasma cutter cycle time.  The experimental design is a Taguchi Parameter Design.   The previous posting describes the experimental design, and refers to the Value Stream Map Case Study posting to review the Lean Six Sigma project that produced the experimental design.   Continue reading Plasma Cutter Cycle Time Experimental Design Results

Experimental Design to Reduce Plasma Cutter Cycle Time

This posting describes the corrective action using an experimental design to reduce a machine’s cycle time.  The machine is a plasma cutting machine, and a Lean Six Sigma (LSS) team identified it as the bottleneck operation in producing electrical switchboards by an electrical manufacturer.    Continue reading Experimental Design to Reduce Plasma Cutter Cycle Time