Skip to content

Week 3

Design of Experiment

To design and analyze any experiment through using statistical methods, it is important to have a perspicuous idea in advance of what is to be investigated and to start with an adequate knowledge about the variables and the possible interaction on the desired response for a better control over the experiments in which less experimental errors are encountered and consequently achieving the aim of the project with the most significant outcomes. This is achieved by pre-experimental planning which helps in recognizing the problem and selecting the proper factors, levels, range of interest, and the response variables. Consequently, the experimental design is selected correctly in order to perform the experiment efficiently and collect the data for statistical analysis, and therefore practical conclusions about the results are drawn.

Selection of the Variables of Interest

Once the problem has been defined and before starting the design of any experiment, it is essential to develop all possible ideas about the objectives of the experiments and acquire full understanding of the variables that are affecting the response and the range of interest in which the experiments are performed. Through looking at literature and performing some experiments, the following experimental ranges were: Milk composition: 65-85 vol% Temperature: 40-60 Degree Celsius

Choice of Design of Experiment (DOE)

The chosen experimental design to investigate the strength of the adhesive was two-level full factorial design with one replicate. This type of design was implemented to study the effect of both milk composition and temperature on the strength of the adhesive. For this type of study where the effect of two factors is involved, factorial design is more efficient compared to a one-factor-at-a-time design, since it considers interactions and will steer clear of misleading results. The following is a picture of the run order for our experiments

Last update: September 8, 2021