Sérgio de Souza Custódio Filho

The knowledge of the forces acting on a structure enables, among many other factors, assessments of whether the component’s useful life is compromised by the current machine condition. In many cases, a direct measurement of those forces becomes unfeasible, and an inverse problem must be solved. Among the solutions developed, machine learning techniques have stood out as powerful predictive tools increasingly applied to engineering problem-solving. This study evaluates the ability of different machine learning models to identify parameters of multi-force systems from accelerometer measurements. The models were assessed according to their prediction potential based on correlation coefficient (R²), mean relative error (MRE), and processing time. A computational numerical model using the finite element method was generated and validated by vibration measurements performed using accelerometers in the laboratory. A robust database created by the response surface methodology in conjunction with Design of Experiment (DOE) was used for the evaluation of the ability of machine learning models to predict the position, frequency, magnitude, and number of forces acting on a structure. Among the six machine learning models evaluated, k-NN was able to predict with a 0.013% error, and Random Forests showed a maximum error of 0.2%. The innovation of this study lies in the application of the proposed method for identifying parameters of multi-force systems.

Leisure activities are known to be especially important for the health of people with disabilities. In Belém, PA, an Amazonian city in Brazil, a nonprofitable organization has promoted leisure ridings in bicycles for those people in Utinga State Park, a large green area for physical and leisure activities. The handcrafted bikes have a sidecar attached for users with disabilities which are ridden by trained volunteers. Since such bikes have been empirically manufactured, they require some minor improvements in safety, comfort, and handling, and verification of structural strength. Therefore, ergonomic, modal, and forced vibration analyses assessed the user’s comfort and safety and a structural analysis with the use of strain gauges evaluated the bicycle’s structural strength. Initially, a numerical modal analysis was performed using the finite element method, and the modal model obtained was validated by an experimental modal analysis employing shaker excitation. ISO-2631-based evaluations of forced vibration and human body comfort were conducted regarding whole-body vibration in vehicles and mechanical equipment. Vibration measurements at the position of the rider and sidecar occupant were obtained during rides on the bicycle and, according to the results, in general, when subjected to loads, the bicycle showed low stress levels far from the yield stress of the material, promoting an excellent safety factor in relation to its structural integrity. The modal, comfort, and forced vibration analyses revealed a mode of vibration in the sidecar that caused discomfort to the back of the users. Ergonomics analysis pointed out changes in the handlebars, the bicycle seat, the coupling between the sidecar and the bike, and the dimensions of the sidecar will provide greater comfort and safety. This paper presents and discusses the proposed modifications to both bicycle and sidecar.

The quantification of torques and moments of inertia of horizontal axis hydrokinetic turbine driveline is important to precisely predict the dynamic behavior of the complete system. Initially, this paper presents different models used in the literature for describing torques and moments of inertia of turbine components. A dynamic model of a small hydrokinetic turbine using belt transmission is developed. The model uses the Blade Element Momentum (BEM) for determining the power coefficient of the turbine rotor and considers the inertial effects and dissipative torques of whole system using some of the torque models and moments of inertia described previously. The results for the dynamical behavior of the turbine model are compared with experimental data, showing good agreement. In order to numerically analyze a more efficient drivetrain, the belt transmission is replaced by a planetary gearbox in the dynamic model, and the new results are also assessed. As a result, it was found that with planetary gears, a more compact transmission can be used, reducing the inertial effects, which brings a better efficiency in the starting of the machine and shortening the transient regime time.

Several dynamic projects and fault diagnosis of mechanical structures require the knowledge of the acting external forces. However, the measurement of such forces is often difficult or even impossible; in such cases, an inverse problem must be solved. This paper proposes a force identification method that uses the response surface methodology (RSM) based on central composite design (CCD) in conjunction with a random forest regression algorithm. The procedure initially required the finite element modal model of the forced structure. Harmonic analyses were then performed with varied parameters of forces, and RSM generated a dataset containing the values of amplitude, frequency, location of forces, and vibration acceleration at several points of the structure. The dataset was used for training and testing a random forest regression model for the prediction of any location, amplitude, and frequency of the force to be identified with information on only the vibration acquisition at certain points of the structure. Numerical results showed excellent accuracy in identifying the force applied to the structure.