Published on: September 5, 2024
1. Introduction
A traditional method for experimenting is One Variable at a Time (OVAT) in which one variable is varied, while all others are kept constant, and its influence is assessed on the target properties (Figure 1.1).
Design of Experiments is a method in which all variables are changed at the same time and their influence is assessed on the target properties. It proposes testing of the most informative locations in the design space which are focused towards the interpretability of the process and space filling of the design space (Figure 1.1).
In table 1.1 OVAT and DOE experiments are compared based on different characteristics.
Figure 1.1 Graphical Representation of One Variable at a Time and Design of Experiments
Table 1.1 Comparison of OVAT and DOE
A lot of gaps in design space
Better filled design space
Time to achieve optimal values
Longer (in most cases not achieved)
Faster (with the help of predictive models)
A lot (several levels are tested for each material)
Number of experiments increases fast with increasing the number of input variables
In certain region accurate in another very inaccurate
Better accuracy throughout different regions of design space
The main types of designs available in Alchemy are:
2. Screening Design
Screening Design is intended to be used for identifying significant main effects from a list of many potential varied materials.
2.1 Screening Design in Alchemy
- Automatically proposed for more than 5 varied variables (materials and/or processing steps).
- A minimum number of experiments related to the materials and/or processing steps must be tested in order to identify which input variables are actually relevant for the given properties.
- Analysis of the main effects is available, after all experiments are tested, with contribution of each material to the tested property and significance according to the Pareto principle (the 80/20 rule)1.
- After the analysis has been given in the screening analysis table users are advised to change constraints for materials and/or processing steps in accordance with the contribution of each material to the tested property.
1 The Pareto principle, also known as the 80/20 rule, is a theory maintaining that 80 percent of the output of a process or system is determined by 20 percent of the input.
2.2 Screening Design Goals
The goals of screening design are to reduce the size of design space through:
- Reducing the number of varied materials and/or processing steps - in further experiments only significant materials should be varied while insignificant materials should be kept constant.
- Narrowing the constraints range.
2.3. Subtypes of Screening Design
Subtypes of screening designs available in Alchemy are:
3. Optimal Design
Optimal Design is intended to be used for filling the design space with experimental points.
3.1. Optimal design in Alchemy
- Automatically proposed for 5 or less varied variables (materials and/or processing steps) or users can choose it after completing screening design.
- Higher number of experiments than for screening are needed.
- After all experiments are tested, models need to be trained and performance of predictive models are shown.
3.2. Optimal Design Goals
The goals of optimal design are to:
- Get high performance of the predictive models.
- Recommend experiments with optimal values for target properties.
3.3. Subtypes of Optimal Design
Subtypes of optimal design available in Alchemy:
4. Adaptive Design
Adaptive Design is intended to be used for augmenting the available dataset (historical dataset and/or designed experiments) with the experiments from the regions of design space with the highest uncertainty.
4.1. Adaptive Design in Alchemy
- Requires previous dataset (either historical and/or designed experiments).
- Number of experiments can be any number defined by the users, although sequentially adding less experiments is recommended.
- After all experiments are tested, models need to be trained and performance of predictive models are shown.
4.2. Adaptive Design Goals
The goal of adaptive design is to:
- Augment the dataset with experiments from the regions of design space with the highest uncertainty in order to explore thoroughly the design space for getting better performance of the models and more accurate predictions.
