In this study, the possibility of using a layered silicate-reinforced polylactic acid (PLA) in additive manufacturing applications was investigated. In particular, the aim of this work was to study the influence of printing temperature in the 3D printing process of PLA/clay nanocomposites. For this reason, two PLA grades (4032D and 2003D, D-isomer content 1.5 and 4, respectively) were melt-compounded by a twin screw extruder with a layered silicate (Cloisite 30B) at 4 wt %. Then, PLA and PLA/clay feedstock filaments (diameter 1.75 mm) were produced using a single screw extruder. Dog-bone and prismatic specimens were 3D printed using the FDM technique at three different temperatures, which were progressively increased from melting temperature (185–200–215 °C for PLA 4032D and 165–180–195 °C for PLA 2003D). PLA and PLA/clay specimens were characterized using thermogravimetric analysis (TGA), dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC), and tensile tests. Moreover, the morphology of the 3D printed specimens was investigated using optical microscopy and contact angle measurements. The different polymer matrix and the resulting nanocomposite morphology strongly influenced 3D printed specimen properties. DMA on PLA/clay filaments reported an increase in storage modulus both at ambient temperature and above the glass transition temperature in comparison to neat PLA filaments. Furthermore, the presence of nanoclay increased thermal stability, as demonstrated by TGA, and acted as a nucleating agent, as observed from the DSC measurements. Finally, for 3D printed samples, when increasing printing temperature, a different behavior was observed for the two PLA grades and their nanocomposites. In particular, 3D printed nanocomposite samples exhibited higher elastic modulus than neat PLA specimens, but for PLA 4032D+C30B, elastic modulus increased at increasing printing temperature while for PLA 2003D+C30B slightly decreased. Such different behavior can be explained considering the different polymer macromolecular structure and the different nanocomposite morphology (exfoliated in PLA 4032D matrix and intercalated in PLA 2003D matrix).

The cushion spring plays an important role in an automotive dry-clutch system. It strongly influences the clutch torque transmission from the engine to the driveline through its non-linear load–deflection curve. Therefore, knowledge of the cushion spring compression behaviour is crucial to improve the gearshift performance in an automated manual transmission. Furthermore, the cushion spring compression behaviour is influenced by the temperature because of the frictional heat generation of the clutch facings with the flywheel and the pressure plate surfaces during the engagement phase. In this paper an analysis of the load–deflection curve, taking into account the thermal load to which it is subjected, of a typical passenger car cushion spring is proposed. Six temperatures, in addition to room temperature, were analysed to investigate how the cushion spring load–deflection curve depends on the temperature.

A custom HW design of a Fully Convolutional Neural Network (FCN) is presented in this paper to implement an embeddable Human Posture Recognition (HPR) system capable of very high accuracy both for laying and sitting posture recognition. The FCN exploits a new base-2 quantization scheme for weight and binarized activations to meet the optimal trade-off between low power dissipation, a very reduced set of instantiated physical resources and state-of-the-art accuracy to classify human postures. By using a limited number of pressure sensors only, the optimized HW implementation allows keeping the computation close to the data sources according to the edge computing paradigm and enables the design of embedded HP systems. The FCN can be simply reconfigured to be used for laying and sitting posture recognition. Tested on a public dataset for in-bed posture classification, the proposed FCN obtains a mean accuracy value of 96.77% to recognize 17 different postures, while a small custom dataset has been used for training and testing for sitting posture recognition, where the FCN achieves 98.88% accuracy to recognize eight positions. The FCN has been prototyped on a Xilinx Artix 7 FPGA where it exhibits a dynamic power dissipation lower than 11 $\mathrm{m}\mathrm{W}$ and 7 $\mathrm{m}\mathrm{W}$ for laying and sitting posture recognition, respectively, and a maximum operation frequency of 47.64 $\mathrm{M}\mathrm{Hz}$ and 26.6 $\mathrm{M}\mathrm{Hz}$, corresponding to an Output Data Rate (ODR) of the sensors of 16.50 $\mathrm{k}\mathrm{Hz}$ and 9.13 $\mathrm{k}\mathrm{Hz}$, respectively. Furthermore, synthesis results with a CMOS 130 $\mathrm{n}\mathrm{m}$ technology have been reported, to give an estimation about the possibility of an in-sensor circuital implementation.