Search

The integration of advanced manufacturing processes with ground-breaking Artificial Intelligence methods continue to provide unprecedented opportunities towards modern cyber-physical manufacturing processes, known as smart manufacturing or Industry 4.0. However, the “smartness” level of such approaches closely depends on the degree to which the implemented predictive models can handle uncertainties and production data shifts in the factory over time. In the case of change in a manufacturing process configuration with no sufficient new data, conventional Machine Learning (ML) models often tend to perform poorly. In this article, a transfer learning (TL) framework is proposed to tackle the aforementioned issue in modeling smart manufacturing. Namely, the proposed TL framework is able to adapt to probable shifts in the production process design and deliver accurate predictions without the need to re-train the model. Armed with sequential unfreezing and early stopping methods, the model demonstrated the ability to avoid catastrophic forgetting in the presence of severely limited data. Through the exemplified industry-focused case study on autoclave composite processing, the model yielded a drastic (88%) improvement in the generalization accuracy compared to the conventional learning, while reducing the computational and temporal cost by 56%.

Manufacture and characterization of a novel, environmentally friendly stone composite based on acrylonitrile butadiene styrene (ABS) and recycled natural granite stone particles is introduced. The proposed composite can be considered as a byproduct of natural stone processing, and potentially lead to a solution to the associated waste management problem. Results indicated that the composite’s tensile strength and fracture property generally deteriorated with increasing the stone particle content, especially beyond 50%. Stiffness and thermal conductivity, however, continually increased with the addition of the granite powder, which would suggest multi-functional applications of this material.

In the present investigations, Zinc oxide (ZnO) and ZnO-ZrO2 composite nanoparticles were synthesized by ultrasonic assisted wet chemical method to investigate their structural, optical and humidity sensing properties. The synthesized nanoparticles were characterized by the techniques like X ray diffraction (XRD), UV–Vis absorption spectroscopy, Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS). XRD and EDS were employed to confirm the phase formation and phase purity respectively. SEM micrographs showed that morphology of the parent compound ZnO is considerably changed with varying concentration of ZrO2. The optical absorption spectra showed that optical absorption of ZnO decreases with increase in ZrO2 content in the composite. The observed band gap values for ZnO and ZnO-ZrO2 composites were higher as compared to the bulk sample. The humidity sensing performance was substantiated for all the samples and the result of effect of concentration of ZrO2 in ZnO-ZrO2 composites on sensitivity, response and recovery time are discussed in detail.

Tuning the piezoresistive behavior of conductive polymeric composites requires an in-depth understanding of the structure-property relationships and deformation mechanisms at the polymer/conductive filler interface. In this paper, the unknown sources of nonlinearity (as frequently observed for specific material systems) are identified for the flexible and stretchable chopped carbon fiber (CCF)/polydimethylsiloxane (PDMS) conductive yarns. It is found that under cyclic loading the resistance initially reduces (negative piezoresistivity) up to a specific transition strain, beyond which positive piezoresistivity is observed until the onset of unloading. Apart from the opposing effects of fiber separation and Poisson's effect, the results suggest that such nonlinearity stems from carbon fiber (CF) buckling induced by CCF/PDMS interfacial friction upon unloading. Nevertheless, a low percolation threshold (1.41 wt%) was attained along with high sensitivity (gauge factors as high as ∼60 in cyclic loading) and stretchability (up to 25%). These characteristics make the CCF/PDMS sensors suitable for 3D-printable inks and stain sensing applications. For the latter, the sensors were integrated with a picture frame shear test setup to monitor the strain along the yarns in fiberglass/polypropylene fabrics subjected to Picture Frame test. Also, the capability of the sensors was demonstrated while used as wearable devices for detecting human joint motion.

Picture frame test (PFT) is frequently employed for characterizing the shear behavior of woven fabrics. In the present study, the main sources of possible imperfections in the PFT, arising from the operator error during sample installation, to the fixture misalignment and the inherent non-uniformities within the fabric, are reviewed and modeled using the kinematic as well as continuum-based approaches. Upon new understandings from these models, a new test methodology was designed where the shearing frame boundary condition is inscribed on the fabric sample itself during sample preparation; in lieu of fabricating and installing a metallic picture frame. The new frameless picture frame (FPF) test was proven to effectively mitigate the imperfections and provide a simple operation and better control and uniformity of the fabric installation and deformation. Contrary to the conventional PFT, the wrinkling behavior and normalized force response in the FPF was in agreement with the conventional bias extension test, with no close-to-arm fiber bending, while showing a superior test repeatability at both loading and unloading stages. Finally, a mechanically compatible, stretchable sensor was developed and integrated into the fabric samples to monitor and verify the induced local deformation along the yarns under different test methods.

Predicting success before full scale manufacturing is the cornerstone of every high-tech industry. Many factors should be considered when designing a product. Design of Experiments (DOE) has been a major, conventional method for such purpose. Moreover, nowadays, Machine Learning (ML) provides fast processing, real-time predictions with a continuous quality improvement for complex industrial design processes. ML with big data is known to handle multi-dimensional and multi-variate data in dynamic industrial systems. However, the ability of the ML methods against limited training datasets has been addressed scarcely. In the present study, a Taguchi DOE approach combined with two selected common ML approaches, the Support Vector Regression (SVR) and Artificial Neural Network (ANN), have been assessed against producing a robust prediction tool for mechanical properties of carbon fibers with a limited training dataset. To this end, empirical models based on three rendering parameters of the stabilization reaction process of oxidized Polyacrylonitrile (PAN) fibers (OPF) have been created. These rendering parameters were obtained according to the explanation of whole Fourier transform infrared attenuated total reflectance (FT-IR ATR) spectra. ML revealed a very promising agreement between the trained empirical models and experimental data. Namely, in terms of the fibers Young’s modulus, the results showed that the SVR empirical model with the average error of less than ±2.4% had a better performance than ANN-LMA (Levenberg –Marquardt algorithm) with the average error of less than ±2.7%. However, in terms of the fibers tensile strength, it was concluded that ANN-LMA empirical model with the average error of less than ±3.7% had slightly surpassed the SVR with the average error of less than ±4.1% in this case study.

Construction of residential buildings using off-site methods, in particular modular construction, is receiving considerable attention. However, the sustainability performance of modular buildings has rarely been investigated through a life cycle perspective. In this paper, a life cycle sustainability performance assessment framework is developed for modular buildings and its application is examined. In the first part of the paper, suitable life cycle sustainability performance criteria (SPCs) for modular buildings were developed and ranked. In this regard, potential SPCs were identified through a comprehensive literature review and expert interviews. These SPCs were then evaluated by construction experts through two questionnaire surveys against three evaluation criteria: applicability, data availability, and data accuracy. The evaluation criteria's weights were determined through a group decision making process using the Analytic Hierarchy Process (AHP) multi-criteria decision analysis (MCDA) method. Consequently, the experts' feedback was analyzed with the help of the Elimination and Choice Translating Reality (ELECTRE) outranking MCDA method and all the SPCs were ranked within their associated sustainability categories (i.e., environmental, economic, and social). In the second part of the paper, application of the proposed framework has been discussed and validated through a case study of a modular building in British Columbia, Canada. Sustainability performance of modular buildings in the proposed framework were assessed by developing aggregated sustainability indices for the selected SPCs and comparing them with corresponding benchmarks. In this regard, appropriate sustainability performance indicators (SPIs) under each selected SPC have been developed and calculated. Consequently, through an aggregation process, sustainability indices are developed using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) MCDA method. In this paper, the environmental life cycle performance of the case study building has been benchmarked and recommendations have been made for performance improvements. This research is deemed useful for the construction practitioners since it provides a methodical framework for life cycle sustainability performance assessment of modular buildings and assists with the selection of sustainable methods of construction.

Textile preforms have become materials of choice in numerous modern industries, partly due to their superior conformability onto complex 3D mould shapes. Maximum formability of this category of composite reinforcements, however, is still limited by defects such as wrinkling, which remains a challenging issue for composite designers during optimization of thermo-stamping operations. The aim of this article is to gain a deeper understanding of the effect of blank holding boundary condition on the extent of wrinkling as well as other local defects such as tow slippage and yarn jamming, and thereby to introduce a passive defect mitigation approach via geometrical modification of the blanks. To verify the applicability of the approach, a series of hemisphere forming experiments under unmodified and modified forming boundary conditions have been performed and compared on both single and multiple ply lay-ups, using a comingled polypropylene/E-glass thermoplastic plain weave.

This paper reports on the damage mechanisms of 3D circular braided composite tubes under transverse impact, applied by a Split Hopkinson Pressure Bar (SHPB) system. X-ray computed tomography (XCT) was employed to obtain 3D images of the 3D braided composite tubes after the impact test. Effects of impact velocity, braiding angle and the number of braiding layers on the damage mechanisms were researched, using the XCT data. For low impact velocity, it was found that cracks nucleate in weak areas and propagate along the path of least resistance between yarns, which give rise to matrix cracking and debonding between yarns. For high impact velocities, cracks develop along a relatively straight path with fiber breakage under high shear and tensile stresses. The orientation of yarns in the tubes showed a great effect on crack propagation.

An analytical-experimental study of the shear wrinkling behavior of plain woven composite preforms under bias extension test (BET) and picture frame test (PFT) is presented. The intentionally induced tension in the PFT has been regarded in a large portion of the literature as the source of delay in fabric wrinkling initiation as compared to the BET. Through this study, however, it is demonstrated that shear within yarns – known as intra-yarn shear – could be another cause of the delayed wrinkling in the PFT, even when the tension level in yarns is kept at zero. To better explain this hypothesis, an analytical model has been developed and the meso-level nature of wrinkle formation in the BET and PFT is compared. Analytical equations are provided to predict both the fabric locking and wrinkling onsets under these characterization tests, and verified experimentally on a carbon fiber plain woven fabric.