Intelligent Generation Method of Innovative Structures Based on Topology Optimization and Deep Learning

Computer-aided design has been widely used in structural calculation and analysis, but there are still challenges in generating innovative structures intelligently. Aiming at this issue, a new method was proposed to realize the intelligent generation of innovative structures based on topology optimization and deep learning. Firstly, a large number of structural models obtained from topology optimization under different optimization parameters were extracted to produce the training set images, and the training set labels were defined as the corresponding load cases. Then, the boundary equilibrium generative adversarial networks (BEGAN) deep learning algorithm was applied to generate numerous innovative structures. Finally, the generated structures were evaluated by a series of evaluation indexes, including innovation, aesthetics, machinability, and mechanical performance. Combined with two engineering cases, the application process of the above method is described here in detail. Furthermore, the 3D reconstruction and additive manufacturing techniques were applied to manufacture the structural models. The research results showed that the proposed approach of structural generation based on topology optimization and deep learning is feasible, and can not only generate innovative structures but also optimize the material consumption and mechanical performance further.     

Virtual reality approaches for immersive design

The current application of virtual reality (VR) systems in the design process is limited mostly to design review. The reason for this limitation is the different data formats used for CAD and VR visualization. To use the benefits of VR during the design process, solutions for immersive design, the model manipulation inside the VE based on CAD data, are required. There are different approaches allowing VR systems to work as an active development platform. Three examples introduce the realization of the integration of CAD and VR software at different levels by the online coupling of complete applications or by integration of CAD core functionalities in VR systems.     

Application of digital practice to improve head movement,visual perception and activities of daily living for subacute stroke patients with unilateral spatial neglect: Preliminary results of a single-blinded,randomized controlled trial

Background:Virtual reality (VR) based digital practice is an attractive way to provide a patient engagement, motivation and adaptable environment for stroke rehabilitation. However, clinical evidence of efficacy with VR-based digital practice is very limited. In this study, we investigated the effects of VR-based digital practice program on unilateral spatial neglect (USN) rehabilitation in patients with subacute stroke.Methods:Twenty-four subacute stroke patients with USN were enrolled and randomly assigned to digital practice group (n = 12) and control group (n = 12). Patients in digital practice group received training programs with VR-based applications with leap motion environment. Control group received conventional USN specific training programs. All patients were underwent 4 week practice program (3 sessions/week, a half-hour/session). We analyzed training effects before and after training by assessing the line bisection test, Catherine Bergego Scale, modified Barthel index, Motor-Free Visual Perception Test Vertical Version (MVPT-V), and horizontal head movements (rotation degree and velocity during the VR-based applications), and compared the results between the two groups.Results:Compared to control group, digital practice group showed significantly greater improvements in the line bisection test (P = .020), and visual perceptual tasks (MVPT-V, responded more on left visual task, P = .024; correctly respond more on both left and right visual tasks, P = .024 and P = .014, respectively; and faster response time, P = .014). Additionally, horizontal head movement of rotation degree and velocity during the VR based practice in the digital practice group were significantly increased more than control group (P = .007 and P = .001, respectively).Conclusions:VR-based digital practice program might be an affordable approach for visual perception and head movement recovery for subacute stroke patients with USN.     

Enhanced Cellular Materials through Multiscale,Variable-Section Inner Designs: Mechanical Attributes and Neural Network Modeling

In the current work, the mechanical response of multiscale cellular materials with hollow variable-section inner elements is analyzed, combining experimental, numerical and machine learning techniques. At first, the effect of multiscale designs on the macroscale material attributes is quantified as a function of their inner structure. To that scope, analytical, closed-form expressions for the axial and bending inner element-scale stiffness are elaborated. The multiscale metamaterial performance is numerically probed for variable-section, multiscale honeycomb, square and re-entrant star-shaped lattice architectures. It is observed that a substantial normal, bulk and shear specific stiffness increase can be achieved, which differs depending on the upper-scale lattice pattern. Subsequently, extended mechanical datasets are created for the training of machine learning models of the metamaterial performance. Thereupon, neural network (NN) architectures and modeling parameters that can robustly capture the multiscale material response are identified. It is demonstrated that rather low-numerical-cost NN models can assess the complete set of elastic properties with substantial accuracy, providing a direct link between the underlying design parameters and the macroscale metamaterial performance. Moreover, inverse, multi-objective engineering tasks become feasible. It is shown that unified machine-learning-based representation allows for the inverse identification of the inner multiscale structural topology and base material parameters that optimally meet multiple macroscale performance objectives, coupling the NN metamaterial models with genetic algorithm-based optimization schemes.     

CAD data workflow toward the thermal simulation and visualization in virtual reality

This paper presents a CAD transfer method to visualize thermal simulation results in a virtual environment in order to optimize building design determined by energy efficiency. As the goal is to present a more immersive thermal simulation and to project the calculation results in projective displays particularly in cave automatic virtual environment, the main idea of the method is to provide a workflow so that the thermal result can be visualized and simulated in the virtual environment (VE). The paper shows the experiment results conducted in an immersion room by some respondents. An evaluation of the method has demonstrated that it is possible to conduct a specific data workflow in order to represent building performance data and particularly thermal simulation results in virtual reality (VR). With this method, the data flow that starts from the design process is completely and accurately channeled to the VR system. CAD data in 3D mock-up models composed of basic geometry are usually designed and transferred to other tools particularly if intended for VR. However, CAD data needed for the VE is not simple, it needs to be structured in a workflow method. Challenges in data exchange and interoperability among design, simulation and VR tools used have become an important issue to investigate how the workflow works.     

Evaluation of virtual reality bronchoscopy as a learning and assessment tool

BACKGROUND: Conventional training in bronchoscopy involves a trainee performing on a real patient under supervision. This method of training is not only expensive, but there is also potential for increased patient discomfort. Simulators permit the acquisition of necessary technical skills required for the procedure. Virtual reality (VR) has been an integral part of training in aviation, and the application of this technology in medical training needs to be evaluated. OBJECTIVE: This study was conducted to evaluate the efficacy of a VR bronchoscopy simulator as a learning and assessment tool. METHODS: The bronchoscopic simulator (HT Medical Systems, Maryland, USA) is a VR computer programme. The simulator has the ability to assess competence by a set of parameters, which formed the data for the study. Nine novices without previous bronchoscopic experience formed the study group (group 1). Nine experienced bronchoscopists having performed between 200 and 1000 bronchoscopies formed the other group (group 2). We assessed the efficacy of the system as a learning tool by studying whether there was a significant difference between the first and subsequent sessions of the subjects from group 1 and by comparing the performance of the two groups. Statistical analysis was done using the Mann-Whitney U test and the Wilcoxon signed ranks test. RESULTS: There was a significant difference in performance between the first attempt of group 1 and the performance of the experts in terms of percentage of segments visualised and number of wall collisions and the economy of performance. Among the subjects from group 1, there was a significant improvement in percentage of segments visualised by the third attempt (p = 0.04), in the economy of performance by the sixth attempt (0.008) and in the number of wall collisions by the sixth attempt (0.024). When each attempt of the novices was compared with the performance of group 2, the significance in the difference of the percentage of segments studied (p = 0.09) and the economy of performance disappeared by the third attempt (0.06), while the difference in the number of wall collisions disappeared by the fifth attempt (p = 0.06). CONCLUSIONS: This study has been able to establish the face, construct and content validity of the simulator and the potential for it to be an effective training tool.     

A comparative evaluation of human interaction for design and assembly of 3D CAD models in desktop and immersive environments

Computer aided design (CAD) systems have become today the basic tools used to design and develop products in the industry. In current CAD software most of the editing commands are issued with the aid of widgets and alphanumeric data input devices, while research community is proposing the use of virtual reality environments for CAD modelling. This paper presents an experimental study which compares the performance and usability of a multimodal immersive VR (virtual reality)-CAD system with a traditional CAD system. A comparative analysis was done for the modelling and the assembling process of 3D models. The results obtained from this investigation have shown that, in spite of the variety of interface devices in the virtual environment which provide a natural interaction to the user, the modelling time is about the same compared with a traditional desktop interface. The assembling time, however, is shown to be much smaller for multimodal system. Furthermore, the multimodal interface poses a higher physical stress factor, the hand movement distance being on average 1.6–2.3 times greater than the desktop interface for modelling process and assembling process, respectively. A post-experiment questionnaire shows that the multimodal system produce a great satisfaction for users in modelling and assembly processes.     

A bi-directional interface for improved interaction with engineering models in virtual reality design reviews

Leveraging virtual reality (VR) technology to enhance engineering design reviews has been an area of significant interest for researchers since the advent of modern VR. The ability to interact meaningfully with 3D computer-aided engineering models in these VR design reviews is an important, though often neglected, capability due to the difficulty of performing data translation between native computer-aided design (CAD) data and VR compatible file formats. A bi-directional interface was developed between a VR design review environment and a commercial CAD package that streamlines the data translation process. Transmitting both geometric data and selected metadata from the CAD system enabled the development of enhanced model interaction tools in a VR design review application. User experiments were performed that compared the enhanced tools developed to a baseline toolset. Participants success using these toolsets was measured as they performed tasks related to design understanding and decision making, such as counting the number of gears in a gearbox or evaluating the feasibility of a proposed design change in a four-cylinder engine. The analysis of the data from these experiments showed a statistically significant improvement in participants ability to understand the geometry of the model correctly, confidently, and quickly, as well as in participants ability to correctly and confidently understand the implications of a proposed design change when using the Enhanced Toolset. We conclude that the bi-directional interface concept developed in this work can be extended to enable advanced interaction with a diversity of engineering data in VR.     

Modeling and Composition Design of Low-Alloy Steel’s Mechanical Properties Based on Neural Networks and Genetic Algorithms

Accurately improving the mechanical properties of low-alloy steel by changing the alloying elements and heat treatment processes is of interest. There is a mutual relationship between the mechanical properties and process components, and the mechanism for this relationship is complicated. The forward selection-deep neural network and genetic algorithm (FS-DNN&GA) composition design model constructed in this paper is a combination of a neural network and genetic algorithm, where the model trained by the neural network is transferred to the genetic algorithm. The FS-DNN&GA model is trained with the American Society of Metals (ASM) Alloy Center Database to design the composition and heat treatment process of alloy steel. First, with the forward selection (FS) method, influencing factors—C, Si, Mn, Cr, quenching temperature, and tempering temperature—are screened and recombined to be the input of different mechanical performance prediction models. Second, the forward selection-deep neural network (FS-DNN) mechanical prediction model is constructed to analyze the FS-DNN model through experimental data to best predict the mechanical performance. Finally, the FS-DNN trained model is brought into the genetic algorithm to construct the FS-DNN&GA model, and the FS-DNN&GA model outputs the corresponding chemical composition and process when the mechanical performance increases or decreases. The experimental results show that the FS-DNN model has high accuracy in predicting the mechanical properties of 50 furnaces of low-alloy steel. The tensile strength mean absolute error (MAE) is 11.7 MPa, and the yield strength MAE is 13.46 MPa. According to the chemical composition and heat treatment process designed by the FS-DNN&GA model, five furnaces of Alloy1–Alloy5 low-alloy steel were smelted, and tensile tests were performed on these five low-alloy steels. The results show that the mechanical properties of the designed alloy steel are completely within the design range, providing useful guidance for the future development of new alloy steel.     

Virtual reality as a learning tool for improving infection control procedures

BackgroundHand hygiene and donning personal protective equipment (PPE) are essential techniques for infection control; however, low compliance is an issue. The effectiveness of virtual reality (VR) in learning infection control procedures is unknown.MethodsTo verify the effectiveness of VR, medical students were categorized into VR or lecture groups (n=21 each). Each group was given the same curricular content; one group received the training through VR learning using a fully-immersive 360-degree video and the other was conventional lecture-style learning. Before and after the training, they were evaluated for the implementation of hand hygiene and PPE using an Objective Structured Clinical Examination method. Post-test questionnaires were administered.ResultsThe scores for hand hygiene, donning PPE, and the total score increased after learning in both groups. There was no difference between the pre-test total scores of the two groups (7 [5-9] vs 6 [5-7.5], P=.352); however, the VR group had significantly higher post-test total scores than the lecture group (12 [9.5-12] vs 9 [8-12], P=.024). More students in the VR group responded that they enjoyed the training and would like to use the same learning method next time.ConclusionsVR can be a useful tool for learning and practicing appropriate infection control procedures.     

Predicting Mechanical Properties of High-Performance Fiber-Reinforced Cementitious Composites by Integrating Micromechanics and Machine Learning

Current development of high-performance fiber-reinforced cementitious composites (HPFRCC) mainly relies on intensive experiments. The main purpose of this study is to develop a machine learning method for effective and efficient discovery and development of HPFRCC. Specifically, this research develops machine learning models to predict the mechanical properties of HPFRCC through innovative incorporation of micromechanics, aiming to increase the prediction accuracy and generalization performance by enriching and improving the datasets through data cleaning, principal component analysis (PCA), and K-fold cross-validation. This study considers a total of 14 different mix design variables and predicts the ductility of HPFRCC for the first time, in addition to the compressive and tensile strengths. Different types of machine learning methods are investigated and compared, including artificial neural network (ANN), support vector regression (SVR), classification and regression tree (CART), and extreme gradient boosting tree (XGBoost). The results show that the developed machine learning models can reasonably predict the concerned mechanical properties and can be applied to perform parametric studies for the effects of different mix design variables on the mechanical properties. This study is expected to greatly promote efficient discovery and development of HPFRCC.     

A Critical Review of the Design,Manufacture, and Evaluation of Bone Joint Replacements for Bone Repair

With the change of people’s living habits, bone trauma has become a common clinical disease. A large number of bone joint replacements is performed every year around the world. Bone joint replacement is a major approach for restoring the functionalities of human joints caused by bone traumas or some chronic bone diseases. However, the current bone joint replacement products still cannot meet the increasing demands and there is still room to increase the performance of the current products. The structural design of the implant is crucial because the performance of the implant relies heavily on its geometry and microarchitecture. Bionic design learning from the natural structure is widely used. With the progress of technology, machine learning can be used to optimize the structure of bone implants, which may become the focus of research in the future. In addition, the optimization of the microstructure of bone implants also has an important impact on its performance. The widely used design algorithm for the optimization of bone joint replacements is reviewed in the present study. Regarding the manufacturing of the implant, the emerging additive manufacturing technique provides more room for the design of complex microstructures. The additive manufacturing technique has enabled the production of bone joint replacements with more complex internal structures, which makes the design process more convenient. Numerical modeling plays an important role in the evaluation of the performance of an implant. For example, theoretical and numerical analysis can be carried out by establishing a musculoskeletal model to prepare for the practical use of bone implants. Besides, the in vitro and in vivo testing can provide mechanical properties of bone implants that are more in line with the implant recipient’s situation. In the present study, the progress of the design, manufacture, and evaluation of the orthopedic implant, especially the joint replacement, is critically reviewed.     

Quality Quantification and Control via Novel Self-Growing Process-Quality Model of Parts Fabricated by LPBF Process

Laser Powder Bed Fusion (LPBF) presents a more extensive allowable design complexity and manufacturability compared with the traditional manufacturing processes by depositing materials in a layer-wised manner. However, the process variability in the LPBF process induces quality uncertainty and inconsistency. Specifically, the mechanical properties, e.g., tensile strength, are hard to be predicted and controlled in the LPBF process. Much research has recently been reported exploring the qualitative influence of single/two process parameters on tensile strength. In fact, mechanical properties are comprehensively affected by multiple correlated process parameters with unclear and complex interactions. Thus, the study on the quantitative process-quality model of the metal LPBF process is urgently needed to provide an enough-strength component via the metal LPBF process. Recent progress in artificial intelligence (AI) and machine learning (ML) provides new insight into quality prediction in terms of computational accuracy and speed. However, the predictive model quality through the traditional AL/ML is heavily determined by the training data size, and the experimental analysis can be expansive on LPBF. This paper explores the comprehensive effect of the tensile strength of 316L stainless-steel parts on LPBF and proposes a valid quantitative predictive model through a novel self-growing machine-learning framework. The self-growing framework can autonomously expand and classify the growing dataset to provide a high-accuracy prediction with fewer input data. To verify this predictive model of tensile strength, specimens manufactured by the LPBF process with different group process parameters (laser power, scanning speed, and hatch spacing) are collected. The experimental results validate the predicted tensile strengths within a less than 3% deviation.     

The influence of coronary artery disease on quality of life after mechanical valve replacement

BACKGROUND AND AIM OF THE STUDY: Coronary artery disease (CAD) is known to impact negatively on long-term survival following valve replacement (VR). However, its influence on quality of life (QOL) remains undefined in patients with mechanical VR. METHODS: A total of 318 consecutive patients undergoing VR with the St. Jude Medical (SJM) mechanical valve were matched for age and gender with 318 patients who had VR (SJM valve) and coronary artery bypass grafting (VR+CABG). The VR group comprised 197 men and 121 women; the VR+CABG group also comprised 197 men and 121 women. The mean age of all patients was 66.0 +/- 8.0 years (range: 40-87 years). The Short Form-36 (SF-36) health survey was administered to all survivors at follow up examination. RESULTS: Operative mortality was comparable between groups (4.7% for VR, 7.5% for VR+CABG; p = 0.186). Hospital complications were also similar, except for reoperation for bleeding (p = 0.049). The mean follow up was 6.0 years for VR patients and 4.7 years for VR+CABG patients. Actuarial survival was significantly better in VR patients than VR+CABG patients (79.4 +/- 2.4% versus 75.0 +/- 2.7% at five years; 58.6 +/- 4.3% versus 47.5 +/- 4.5% at 10 years; p = 0.018). The equality of survival distribution was significantly different (p = 0.008). Multivariate analysis identified CABG as a predictor of late mortality (p = 0.003) but not of late QOL. QOL was similar on the eight health scales and physical health (44.5 +/- 10.3 versus 45.5 +/- 10.7) and mental health (52.4 +/- 9.8 versus 52.5 +/- 10.1) summary components, respectively. Age (p = 0.004), time from surgery to SF-36 administration (p = 0.007) and gender (p = 0.029), but not CABG, were significantly associated with QOL as assessed by the SF-36. CONCLUSION: CAD is a predictor of late mortality after mechanical VR. However, provided CABG is performed concomitantly with VR, the patient's longterm QOL appears to return to that expected for the general population.     

Design of an innovative assembly process of a modular train in virtual environment

In this paper we present an innovative assembly cycle of railway vehicles that can improve the manufacturing process. The study was carry out using virtual reality (VR) technologies in the VR Laboratory of the Competence Regional Centre for the qualification of transportation systems set up by Campania Regional Authority. Using the developed simulation environment we were able to evaluate different workplaces layout configurations of the train assembly cycle. The best workplace layout configuration was detected in order to minimize the lead time in the production line and optimize the automation level and human component for each workplace.     

User-centered design of a virtual reality exhibit for archaeological museums

Nowadays, the adoption of virtual reality (VR) exhibits is increasingly common both in large and small museums because of their capability to enhance the communication of the cultural contents and to provide an engaging and fun experience to its visitors. The paper describes a user-centered design (UCD) approach for the development of a VR exhibit for the interactive exploitation of archaeological artefacts. In particular, this approach has been carried out for the development of a virtual exhibit hosted at the “Museum of the Bruttians and the Sea” of Cetraro (Italy). The main goal was to enrich the museum with a playful and educational VR exhibit able to make the visitors enjoy an immersive and attractive experience, allowing them to observe 3D archaeological artefacts in their original context of finding. The paper deals with several technical issues commonly related to the design of virtual museum exhibits that rely on off-the-shelf technologies. The proposed solutions, based on an UCD approach, can be efficiently adopted as guidelines for the development of similar VR exhibits, especially when very low budget and little free space are unavoidable design requirements.     

An interactive design approach for nuclear fusion purposes: remote handling system for FAST divertor

The paper deals with the design issues concerning the remote maintenance of divertors in fusion advanced studies torus (FAST), a satellite tokamak acting as a test bed for the study and the develop of innovative technologies oriented to ITER and DEMO programs, pilot examples of the feasibility of energy production from nuclear fusion on the Earth. FAST remote handling (RH) solutions are provided according to an “interactive design review” philosophy based on virtual prototyping techniques. Assuming an ITER configuration as start point, it foresees an iterative process of design review, carried out in virtual reality (VR) environment and oriented to obtain a sort of best solution from the RH point of view. Any iteration includes the analysis of the current solution and the proposal of new and alternative ones, based on the requirements fulfillment and the improvement of critical points highlighted. In such a way, and this is the main novelty introduced by the paper, the interactive design review in a VR collaborative environment becomes the tool able to put in cooperation and in positive competition various and different competences, required by a multidisciplinary problem as the realization of nuclear fusion machine, in order to reach a shared solution. A first preliminary FAST RH solution is hereinafter presented, accompanied by the design of a compatible support system, due to the strict relationship between the divertor maintenance and the support configuration. The work was carried out via the collaboration of the “Divertor Test Platform 2” (DTP2) team, in charge of ITER divertor RH tests and located in VTT’s Labs of Tampere (Finland), and the IDEAinVR team of CREATE Consortium, with competence in interactive design and VR simulations and located in the Virtual Reality Lab of University of Naples Federico II (Italy).     

Efflux of adenosine and total adenylate catabolites during alterations of the cellular energy state. An NMR study of continuous and discontinuous ischemia

Reperfusion after continuous or discontinuous ischemia has a bearing on clinical interventions. An important question is the washout of metabolites after periods of diminished energy state of the myocardial cell. We therefore set out to determine the washout of adenosine and its metabolites after periods of ischemia in an experimental set-up which allowed non-destructive monitoring of the cellular energy state and cytosolic pH over consecutive time intervals. Isolated rat hearts were perfused with hemoglobin-free saline in a nuclear magnetic resonance spectrometer equipped for³¹P NMR spectroscopy of phosphorus-containing metabolites, which could be measured over 3-min time blocks. The response of the heart when subjected to 18 min of continuous ischemia and subsequent reperfusion was compared with that when subjected to three 6-min periods of ischemia separated by 3-min periods of reperfusion. The mechanical performance of the hearts, oxygen consumption and efflux of adenosine and its metabolites were measured. The consecutive ischemic periods produced no evidence of preconditioning as judged from the cellular energy state, although the mechanical recovery was better than after continuous ischemia. During the repetitive ischemia/reperfusion protocol the efflux of adenosine was smaller, although the efflux of combined adenylate catabolites did not differ from that after continuous ischemia. The results do not support the view of adenosine being a major effector in the phenomenon of preconditioning.     

Design of Experiment on Concrete Mechanical Properties Prediction: A Critical Review

Concrete mix design and the determination of concrete performance are not merely engineering studies, but also mathematical and statistical endeavors. The study of concrete mechanical properties involves a myriad of factors, including, but not limited to, the amount of each constituent material and its proportion, the type and dosage of chemical additives, and the inclusion of different waste materials. The number of factors and combinations make it difficult, or outright impossible, to formulate an expression of concrete performance through sheer experimentation. Hence, design of experiment has become a part of studies, involving concrete with material addition or replacement. This paper reviewed common design of experimental methods, implemented by past studies, which looked into the analysis of concrete performance. Several analysis methods were employed to optimize data collection and data analysis, such as analysis of variance (ANOVA), regression, Taguchi method, Response Surface Methodology, and Artificial Neural Network. It can be concluded that the use of statistical analysis is helpful for concrete material research, and all the reviewed designs of experimental methods are helpful in simplifying the work and saving time, while providing accurate prediction of concrete mechanical performance.