The design of FUMD specimens requires the use of unconventional stacking sequences, obtained from a set of solutions to the equations describing some of the desired constraints on the specimens, called quasi-trivial (QT) quasi-homogeneous solutions. While QT solutions have enormous potential for laminate design, their understanding and availability are still limited. Hence, the first project activity was the development of a novel formal framework to describe and manipulate QT solutions. This enabled to reveal their mathematical nature in a new way, and to discover properties that were not known before. The insight and the tools developed allowed us to devise a novel and efficient strategy to obtain QT solutions, which exploits a recursive algorithmic procedure. Hence, we were able to create a huge QT database, that will be soon made open access.
With the newly established foundations on QT solutions, the second project activity was the creation of a database of sequences for FUMD delamination specimens. Firstly, an in-depth analysis was conducted to formalise, better than in the past, the design routes to obtain FUMD specimens from QT quasi-homogeneous solutions. This was instrumental to devise and develop routines to find, extract, and (when needed) process those solutions in the QT database that can be used to obtain sequences for FUMD specimens. As a result of this, we created the desired, comprehensive database of sequences for FUMD specimens.
Once the database of sequences for FUMD specimens had been created, we investigated in-depth aspects of practical interest for testing purposes. Firstly, we explored if and how the sequences in the database would enable testing of different types of delamination interfaces (different orientations between embedding layers, different orientations with respect to the specimen length direction, among other aspects), and established precisely which test interfaces would be possible for any given total number of plies in the specimen. We then investigated also the possibility to obtain stiff enough specimens by including a sufficient number of 0º layers.
Next, in view of the large number of feasible FUMD specimens, instruments for an informed selection of the most appropriate configurations for testing are needed. For this purpose, we developed analytical expressions allowing to quickly analyse large sets of FUMD configuration based on their main design features and visualize important properties, such as the expected thermal residual stresses, the tendency to suffer finite-width effect, and others. These tools enable a fast preliminary screening of configurations for testing purposes. Additionally, we also developed finite element models capable to provide much more detailed information about specific configurations, at the expense of a longer time needed to analyse multiple configurations. While the tools developed have already proven useful in the selection of FUMD configurations for testing, further developments are ongoing, and it is expected that they will be perfected in the coming future.
Finally, multiple experimental activities to validate the approach developed and to obtain a large experimental dataset will be carried out. Some of these activities are already underway, while others are being prepared currently.
