Abstract
A recent systematic experimental characterisation of technological thin films, based on elaborated design of experiments as well as probe calibration and correction procedures, allowed for the first time the determination of nanoscale friction under the concurrent influence of several process parameters, comprising normal forces, sliding velocities, and temperature, thus providing an indication of the intricate correlations induced by their interactions and mutual effects. This created the preconditions to undertake in this work an effort to model friction in the nanometric domain with the goal of overcoming the limitations of currently available models in ascertaining the effects of the physicochemical processes and phenomena involved in nanoscale contacts. Due to the stochastic nature of nanoscale friction and the relatively sparse available experimental data, meta-modelling tools fail, however, at predicting the factual behaviour. Based on the acquired experimental data, data mining, incorporating various state-of-the-art machine learning (ML) numerical regression algorithms, is therefore used. The results of the numerical analyses are assessed on an unseen test dataset via a comparative statistical validation. It is therefore shown that the black box ML methods provide effective predictions of the studied correlations with rather good accuracy levels, but the intrinsic nature of such algorithms prevents their usage in most practical applications. Genetic programming-based artificial intelligence (AI) methods are consequently finally used. Despite the marked complexity of the analysed phenomena and the inherent dispersion of the measurements, the developed AI-based symbolic regression models allow attaining an excellent predictive performance with the respective prediction accuracy, depending on the sample type, between 72% and 91%, allowing also to attain an extremely simple functional description of the multidimensional dependence of nanoscale friction on the studied variable process parameters. An effective tool for nanoscale friction prediction, adaptive control purposes, and further scientific and technological nanotribological analyses is thus obtained.
Highlights
Tribology, with its marked influence on manufacturing processes, energy consumption, environmental impacts, aerospace technologies or, more recently, the microand the nanotechnologies, especially the study of the fundamental physio-chemical aspects determining the origin of friction at the nanoscales, is a propulsive and evolving field of study [1, 2]
Based on the performance metrics of the various used artificial intelligence (AI)-based genetic programming (GP)-symbolic regression (SR) models, it can be concluded that the multi-gene genetic programming (MG GP) model trained with pooled data shows the best predictive performances, with high achieved R2 values and a relatively compact model expression’s length and depth
MG GP models are selected as the best individuals from a population of 5,000 models from each training run, which corresponds to a 10-fold cross validation repeated 10 times for the 50 genes used in the multi-gene model
Summary
With its marked influence on manufacturing processes, energy consumption, environmental impacts, aerospace technologies or, more recently, the microand the nanotechnologies, especially the study of the fundamental physio-chemical aspects determining the origin of friction at the nanoscales, is a propulsive and evolving field of study [1, 2]. Friction in single asperity contacts was introduced in a recent study [3] This procedure is based on an elaborated design-of-experiments (DoE) methodology, conducted by using centroidal Voronoi tessellation (CVT) sampling [4], as well as on a carefully conceived characterisation of the stiffness of the used scanning probe microscopy (SPM) probes and of the influence of tip wear and adhesion on the obtained results. By using in a first instance firstorder statistical analyses, based on Pearson’s product-moment correlations (PPMC) [5], important insights into the general trends of the dependence of nanoscale friction on the multiple studied process parameters were obtained, confirming that their interactions and mutual effects must be investigated at the structural atomic level to be fully appreciated. The obtained simple functional description of the dependence of nanoscale friction on the studied process parameters is an effective tool for nanoscale friction prediction, apt to be used in practical applications, while offering fundamental insights into the tribological behaviour at the nanometric scales with multidimensional influential parameters
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
