Generally, the vibrations in the turning process can be classified into three categories: free vibration, forced vibration and self-excited vibration [14]. Certain combinations of machining parameters generate process instabilities that produce a high level of vibrations which in turn results in a decrease in accuracy, inferior surface finish, reduced tool life time and lower metal removal rate [15]. Typically, vibrations are analysed using sensors such as accelerometers and dynamometers [16]. For instance, Thomas and Beauchamp [13] analysed the vibrations to evaluate the influence of different cutting variables (cutting speed, depth of cut and feed rate), tool variables (tool nose radius and tool length) and workpiece length on the tool vibration generated in turning by means of a statistical analysis based on 288 tests. Another noteworthy study was conducted by Stoic et al. [17]. In the study, surface roughness is identified as a consequence of both cutting disturbances and the tool/workpiece non-uniform loading distribution. Authors stated that force measurements are preferred for low frequencies and acceleration measurements at high frequencies. Lee et al. [14] recognised the different behaviour in terms of vibrations of two different materials: S45C steel and 6061 aluminium when turned, presenting a vibration absorber method to diminish the vibrations generated. Upadhyay et al. [18] developed a prediction model using cutting parameters and vibrations as input to predict surface roughness. They developed models as a function of the vibration amplitude (radial, axial and tangential directions), obtaining an average percentage error of 4.11% and maximum percentage error of 6.42%. Hessainia et al. [19] studied the surface roughness in hard turning, evaluating the influence of the cutting parameters and vibrations measured in the main and radial cutting force directions. In the study, a strong relation between surface roughness and cutting parameters, mainly feed rate, was found, while vibrations were found to be non-significant in the Analysis of Variance performed. Abouelatta and Mádl [20] predicted the surface roughness based on cutting parameters and tool vibrations, stating that the prediction is better when the model includes tool vibrations.