Design of

Experiments

Case histories

The Design of Experiments (DoE) is a system of statistical methods that guide the definition of an experimental plan.

Thanks to DoE, it is possible to identify the small group of highly informative experimental trials, capturing the variability of the phenomenon under exam. It allows to obtain the answers to the initial questions while remaining within the established time and budget constraints.

The fields we have applied its skills in DoE range from chemical and industrial to the medical field.

Our Solution

Challenge

Every industry has its own specificities and each Research and Development department has to deal with quite different types of problems and verifications. They all have one thing in common: collecting data about all possible combinations of the characteristics under investigation, in order to verify them with statistical significance, is untenable for any laboratory. Identifying the approach to be followed for the design of the experiment and the selection of experimental tests to execute, requires expertise both in the related field as well as in statistical techniques.  

What we do

Once analysed the case together with the client, our DoE team establishes what characteristics to consider when studying the phenomenon of interest. Therefore, among the many techniques available, our experts select the solution that best suits the specific needs of the experiment. 

The choice of the methodology for the specific experimental design depends, indeed, on the number of factors to manipulate during the experiment, the amount of different declinations each factor can take, and the type of factor. As an example: in problems involving molecular mixtures, it is necessary to manage the constraints dictated by the formulation.

A good experimental design saves time and money, as the trials to be carried out are small in number but very informative. We accompany our customers through the process of innovation and research, making the path to new products quicker and cheaper.

Organic plastic field

A company producing bioplastics was researching new bio-formulations to produce eco-compatible products with the same elasticity properties, strength and resistance of traditional plastics.

Two experimental designs have been realised for this project:

  • a first design for mixtures with constraints on 8 components and 1 linear type constraint on 4 components;
  • a second design for mixtures with constraints on 9 components and 2 linear type constraints.

The objective of the DoE is to study the properties of interest in the product, including elastic modulus, elongation at break, yield strength and MFI.

Research project on semiconductors

As part of a research project on sheaths and semiconductors for the development of an innovative material, we were commissioned to design 12 experimentations.

Three of them concerned the design of low-voltage electrical cables, studying in particular the sheath and the semiconductor through experimental designs for blends with 9 components. Others focused on the design of the cable, with experimentation on 5 factors of interest, which was carried out by optimising a full factorial design.

The second tranche of experimental projects was aimed at studying flame retardant nanofillers dispersed in functionalised polyolefins and polymer bases, by modifying the concentration of 7 components of the blend.

Upon completion of the project, our team followed the entire analysis phase of the experimental results to support the evaluation of the best solution among those examined.

Research project on Optical fibres

As part of a research project aimed at evaluating the ease of handling innovative geometries for fibre-optic cable assembly, we were tasked with the development of three experimental designs and the analysis of the results that emerged from the planned tests.

The designs took into account the thicknesses of the various sheath layers and those of the fibres, as well as the method of assembly: the data collected during the experimentation were initially explored by means of principal component analysis to identify any links between the modified factors.

Afterwards, a linear regression model was adapted to assess the link between the various factors under study and the number of fibres that could be inserted within the cable, with the aim of maximising the number of fibres without decreasing their manipulability.

Paper industry

A major paper industry requested the creation of experimental designs for the study of factors believed to be linked to the generation of vibrations on machines during the production of non-woven fabric.

The aim of the project was to identify the risky combinations to avoid and, consequently, optimise the production process by eliminating production defects and related waste.

Our team structured and supervised the execution of the experimental designs defined, to ensure their correct implementation and, as a consequence, the correctness of the research results.

Research project on bio-sensors

As part of a research project on patients with neuromuscular diseases at different stages, an experimental design was requested to test the use of sensors aiming to objectify the assessment of the extension and flexion ability of the dominant limb during the session with the physiotherapist.

ur team was in charge of analysing the collected data and integrated the results obtained into a dedicated platform.

The platform, created custom by our team, supported physiotherapists and technicians in calibrating the sensors and processing the data sent wirelessly.

Want to know more about these projects?

Drop us a line!

Contact form

Thank you for your message