Force Prediction Research Articles

Machine Learning-Driven Optimization of Micro-Textured Surfaces for Enhanced Tribological Performance: A Comparative Analysis of Predictive Models

Micro-textured surfaces show promise in improving tribological properties, but predicting their performance remains challenging due to complex relationships between surface features and frictional behavior. This study evaluates five algorithms—linear regression, decision tree, gradient boosting, support vector machine, and neural network—for their ability to predict load-carrying capacity and friction force based on texture parameters including depth, side length, surface ratio, and shape. The neural network model demonstrated superior performance, achieving the lowest MAE (24.01) and highest R-squared value (0.99) for friction force prediction. The results highlight the potential of machine learning techniques to enhance the understanding and prediction of friction-reducing micro-textures, contributing to the development of more efficient and durable tribological systems in industrial applications.

Crushing Force Prediction Method of Controlled-Release Fertilizer Based on Particle Phenotype

This study proposed a method to predict the crushing force of controlled-release fertilizer granules based on their phenotypic characteristics to prevent coating damage during production, transport, and fertilization, which could affect nutrient diffusion rates. The phenotypic features, including sphericity, particle size, and texture, of three commonly used controlled-release fertilizers were obtained using machine vision, while the crushing force was measured using a universal testing machine. A principal component analysis was applied for data reduction, and the optimal parameters for the support vector machine (SVM) were selected using particle swarm optimization (PSO) combined with k-fold cross-validation. A particle swarm optimization–support vector machine (PSO-SVM) model was then developed to predict the crushing force based on fertilizer shape features. Compared with the traditional method, the innovation of this paper is that a non-destructive prediction method is proposed, which enables high-precision predictions of the crushing force by integrating multi-dimensional phenotypic features and an intelligent optimization algorithm. Comparative tests with a random forest regression, the K-nearest neighbor, a back propagation (BP) neural network, and a long short-term memory (LSTM) neural network have demonstrated that the PSO-SVM model outperforms these methods in terms of mean absolute error, root mean square error, and correlation coefficient, underscoring its effectiveness. The proportion of predictions within the −10% to +10% error range reached 0.82, 0.82, and 0.86 for the three fertilizers, confirming the high reliability and accuracy of the PSO-SVM method for non-destructive testing.

Deposition simulations of realistic dosages in patient-specific airways with two- and four-way coupling.

Inhalers spray over 100 million drug particles into the mouth, where a significant portion of the drug may deposit. Understanding how the complex interplay between particle and solid phases influence deposition is crucial for optimising treatments. Existing modelling studies neglect any effect of particle momentum on the fluid (one-way coupling), which may cause poor prediction of forces acting on particles. In this study, we simulate a realistic number of particles (up to 160 million) in a patient-specific geometry. We study the effect of momentum transfer from particles to the fluid (two-way coupling) and particle-particle interactions (four-way coupling) on deposition. We also explore the effect of tracking groups of particles ('parcels') to lower computational cost. Upper airway deposition fraction increased from 0.33 (one-way coupled) to 0.87 with two-way coupling and 10µm particle diameter. Four-way coupling lowers upper airway deposition by approximately 10% at 100µg dosages. We use parcel modelling to study deposition of 4-20µm particles, observing significant influence of two-way coupling in each simulation. These results show that future studies should model realistic dosages for accurate prediction of deposition which may inform clinical decision-making.

Estimation of Joint Kinetics During Manual Material Handling Using Inertial Motion Capture: A Follow-Up Study.

Musculoskeletal models based on inertial motion capture (IMC) and ground reaction force (GRF) prediction hold great potential for field-based risk assessment of manual material handling (MMH). However, previous evaluations have identified inaccuracies in the methodology's estimation of spinal forces, while the accuracy of other key outcome variables is currently unclear. This study evaluated knee, shoulder, and L5-S1 joint reaction forces (JRFs) derived from a musculoskeletal model based on inertial motion capture and GRF prediction against a model based on simultaneously collected optical motion capture (OMC) and force plate measurements. Data from 19 healthy subjects performing lifts with various horizontal locations (HLs), deposit heights (DHs), and asymmetry angles (AAs) were analyzed, and the consistency and absolute agreement of the model estimates statistically compared. Despite varying levels of agreement across tasks and variables, considerable absolute differences were identified for the L5-S1 axial compression (AC) (root-mean-square error (RMSE) = 63.0-94.2%BW) and anteroposterior (AP) shear forces (RMSE = 40.9-80.6%BW) as well as the bilateral knee JRFs (RMSE = 78.9-117%BW). Glenohumeral JRFs and vertical GRFs exhibited the highest overall consistency (r = 0.33-0.91, median 0.78) and absolute agreement (RMSE = 7.63-34.9%BW), while the L5-S1 axial compression forces also showed decent consistency (r = 0.04-0.89, median 0.80). The findings generally align with prior evaluations, indicating persistent challenges with the accuracy of key outcome variables. While the modeling framework shows promise, further development of the methodology is encouraged to enhance its applicability in ergonomic evaluations.

Real-time prediction of structural responses in deep-water jacket platforms using Ada-STLNet: A comprehensive analysis with prototype monitoring data

Real-time prediction of the mechanical behaviors based on prototype monitoring data is crucial for ensuring the integrity and safety of deep-water jacket platforms. As complex nonlinear systems, these platforms’ response exhibit varying temporal trends, dynamic spatial correlations, and load dependencies. This makes accurate prediction of future axial force and bending moment responses a significant challenge. In this paper, a novel deep learning method called Adaptive Spatio-Temporal Load Network (Ada-STLNet) is proposed aiming to address the issue of multi-node axial force and bending moment response prediction of deep-water jacket platform. Our model utilizes an enhanced graph convolution (EGCN) and self-attention mechanism for efficient spatial correlation modeling in multi-node response, LSTM for long sequence temporal feature capture. It integrates spatial and temporal, along with load-aware modules to capture intricate patterns in structural responses. Additionally, a multi-gated coupling mechanism with two independent gating modules is designed to adaptively represent the complex dependencies of structural responses, ensuring model stability and robustness. Extensive experiments conducted on prototype structural monitoring data from a deep-water jacket platform in the South China Sea demonstrate that Ada-STLNet outperforms six traditional models, including HA, SVR, KNN, MLP, LSTM, and GRU, across various prediction steps. The proposed model exhibits superior accuracy, particularly in short-term predictions, achieving up to 62.80% improvement in MAE and 48.11% in RMSE for axial force predictions compared to the second-best model. Ablation studies confirm the effectiveness of each component within Ada-STLNet. This research highlights the potential of Ada-STLNet for reliable and accurate prediction of structural responses, contributing significantly to the field of marine engineering.

Comparative study and experimental validation of cutting forces in high-speed ball-end milling for AL 7075-T6 and AL 6061-T6 alloys

This study examines the importance of precise cutting force prediction during high-speed ball-end milling of the AL 7075-T6 and AL 6061-T6 aluminum alloys, which are widely used in the aerospace and automotive industries, because to their strength and fatigue resistance. Predicting cutting forces is critical for optimizing milling conditions and creating high-quality surfaces. We have used a mechanistic model to estimate cutting forces based on experimental data. To generate the more precise predictions, the developed model takes into account tool geometry, cutting settings and material properties. We have measure the cutting forces on a CNC SOMAB DIAM 850 milling machine by using a Kistler 9257A dynamometer. We have done simulations with MATLAB software. The results reveal that the mechanistic model accurately predicts both alloys. ALU 7075-T6 produces higher cutting forces because to its increased strength. This comparative research demonstrates that the mechanistic approach is useful in forecasting cutting forces in high-speed ball-end milling of aluminum alloys, providing useful insights for improving machining operations.

K-Means Clustering in Fingerprint-Based Configuration Selection for Fitting Interatomic Potentials.

In this study, we present a method for selecting an arbitrary number of distinct configurations from a larger data set by applying k-means clustering to atomistic configuration fingerprints based on the CrystalNN model and radial distribution function (RDF). This approach improves the accuracy of fitting classical molecular dynamics interatomic potentials to density functional theory (DFT) data for both energies and forces while requiring fewer configurations than random selection. We demonstrate this improvement by fitting an embedded-atom method (EAM) potential for titanium, using various configurational sizes from an initial set of 1800 configurations. The k-means clustering consistently achieves better precision and lower standard deviations for a smaller number of configurations than random selection. The results also suggest that only about 30 configurations are sufficient to obtain an EAM model that describes well the full set of 1800 configurations in terms of energies and forces. Additionally, t-distributed stochastic neighbor embedding (t-SNE) method was used to reduce the configuration fingerprints into 2D space, and it revealed an overlap between two configuration subsets with and without Ti vacancy, indicating similar atomic environments. This similarity is captured by k-means clustering but not by random selection. Furthermore, when the overlapping configurations with vacancies were excluded from the k-means algorithm and used only as a test set, their energy and force predictions showed similar precision to those when they were included. This indicates that the overlapping configurations in the 2D t-SNE space indeed imply potential information redundancy among the atomistic configurations.

Automatic Feature Selection for Atom-Centered Neural Network Potentials Using a Gradient Boosting Decision Algorithm.

Atom-centered neural network (ANN) potentials have shown high accuracy and computational efficiency in modeling atomic systems. A crucial step in developing reliable ANN potentials is the proper selection of atom-centered symmetry functions (ACSFs), also known as atomic features, to describe atomic environments. Inappropriate selection of ACSFs can lead to poor-quality ANN potentials. Here, we propose a gradient boosting decision tree (GBDT)-based framework for the automatic selection of optimal ACSFs. This framework takes uniformly distributed sets of ACSFs as input and evaluates their relative importance. The ACSFs with high average importance scores are selected and used to train an ANN potential. We applied this method to the Ge system, resulting in an ANN potential with root-mean-square errors (RMSE) of 10.2 meV/atom for energy and 84.8 meV/Å for force predictions, utilizing only 18 ACSFs to achieve a balance between accuracy and computational efficiency. The framework is validated using the grid searching method, demonstrating that ACSFs selected with our framework are in the optimal region. Furthermore, we also compared our method with commonly used feature selection algorithms. The results show that our algorithm outperforms the others in terms of effectiveness and accuracy. This study highlights the significance of the ACSF parameter effect on the ANN performance and presents a promising method for automatic ACSF selection, facilitating the development of machine learning potentials.

Investigating in vivo force and work production of rat medial gastrocnemius at varying locomotor speeds using a muscle avatar.

Traditional work loop studies, that use sinusoidal length trajectories with constant frequencies, lack the complexities of in vivo muscle mechanics observed in modern studies. This study refines methodology of the 'avatar' method (a modified work loop) to infer in vivo muscle mechanics using ex vivo experiments with mouse extensor digitorum longus (EDL) muscles. The 'avatar' method involves using EDL muscles to replicate in vivo time-varying force, as demonstrated by previous studies focusing on guinea fowl lateral gastrocnemius (LG). The present study extends this method by using in vivo length trajectories and electromyographic activity from rat medial gastrocnemius (MG) during various gaits on a treadmill. Methodological enhancements from previous work, including adjusted stimulation protocols and systematic variation of starting length, improved predictions of in vivo time-varying force production (R2=0.80-0.96). The study confirms there is a significant influence of length, stimulation and their interaction on work loop variables (peak force, length at peak force, highest and average shortening velocity, and maximum and minimum active velocity), highlighting the importance of these interactions when muscles produce in vivo forces. We also investigated the limitations of traditional work loops in capturing muscle dynamics in legged locomotion (R2=0.01-0.71). While in vivo length trajectories enhanced force prediction, accurately predicting work per cycle remained challenging. Overall, the study emphasizes the utility of the 'avatar' method in elucidating dynamic muscle mechanics and highlights areas for further investigation to refine its application in understanding in vivo muscle function.

Precision Orthodontic Force Simulation Using Nodal Displacement-Based Archwire Loading Approach.

Precision in force simulationis critical for forecasting tooth movement and optimizing orthodontic treatment strategies. While traditional techniques have provided valuable insights, there remains a need for improved methodologies that can seamlessly integrate with fixed orthodontic practices. This study aims to refine orthodontic force simulation techniques by integrating a nodal displacement approach within finite element analysis, specifically designed to enhance prediction accuracy in tooth movement and optimize orthodontic treatment planning. Three-dimensional patient-specific models of the Tooth, Periodontal Ligament, and Bone Complex (TPBC) of five volunteers were created, along with models of brackets and wires. The simulation involved an initial step of estimating node displacements to align the archwire with the brackets, followed by a subsequent step to attain the required tooth movement and determine the orthodontic force. Experimental validation of the simulation results was performed using an orthodontic force tester (OFT). Utilizing the nodal displacement approach, the simulation successfully positioned the archwire onto the brackets. When benchmarked against the OFT, 80% of the simulated force directions exhibited angular discrepancies of less than 5°. Additionally, the absolute differences in force magnitude reached 20.06 cN, and in moments, up to 71.76 cN mm. The relative differences were as high as 9.55% for force and 13.83% for moments. These findings represent an improvement of up to 10.45% in force accuracy and 8.87% in moment accuracy compared to median values reported in most recent literature. In this research, a nodal displacement methodology was employed to simulate orthodontic forces with precision across the dental arch. The results demonstrate the approache's potential to enhance the accuracy of force prediction in orthodontic treatment planning, thereby advancing our understanding of orthodontic biomechanics.

Broaching Digital Twin to Predict Forces, Local Overloads, and Surface Topography Irregularities.

Broaching is a key manufacturing process that directly influences the surface integrity of critical components, impacting their functional performance in sectors such as aeronautics, automotive, and energy. Such components are subjected to severe conditions, including high thermomechanical loads, fatigue, and corrosion. For this reason, the development of predictive models is essential for determining the optimal tool design and machining conditions to ensure proper in-service performance. This study, therefore, presents a broaching digital twin based on hybrid modelling, which combines analytical, numerical, and empirical approaches to provide rapid and accurate predictions of the forces per tooth, local overloads, and surface topography irregularities. The digital twin was validated with a critical industrial case study involving fir-tree broaching of turbine discs made of forged and age-hardened Inconel 718. The accuracy of the digital twin was demonstrated by the results: the average error in force predictions was below 10%, and the model effectively identified the most critical teeth and zones prone to failure. It also predicted surface topography irregularities with an error of less than 15%. Interestingly, the relationship between surface topography irregularities and surface residual stress variations across the machined surface was observed experimentally for the first time.

External Validation of Accelerometry-Based Mechanical Loading Prediction Equations

Accurately predicting physical activity-associated mechanical loading is crucial for developing and monitoring exercise interventions that improve bone health. While accelerometer-based prediction equations offer a promising solution, their external validity across different populations and activity contexts remains unclear. This study aimed to validate existing mechanical loading prediction equations by applying them to a sample and testing conditions distinct from the original validation studies. A convenience sample of 49 adults performed walking, running, and jumping activities on a force plate while wearing accelerometers at their hip, lower back, and ankle. Peak ground reaction force (pGRF) and peak loading rate (pLR) predictions were assessed for accuracy. Substantial variability in prediction accuracy was found, with pLR showing the highest errors. These findings highlight the need to improve prediction models to account for individual biomechanical differences, sensor placement, and high-impact activities. Such refinements are essential for ensuring the models’ reliability in real-world applications, particularly in clinical and biomechanical research contexts, where accurate assessments of mechanical loading are critical for designing rehabilitation programs, injury prevention strategies, and optimizing bone health interventions.

Improved random forest for titanium alloy milling force prediction based on finite element-driven

Improved random forest for titanium alloy milling force prediction based on finite element-driven

Inductive graph‐based long short‐term memory network for the prediction of nonlinear floor responses and member forces of steel buildings subjected to orthogonal horizontal ground motions

Inductive graph‐based long short‐term memory network for the prediction of nonlinear floor responses and member forces of steel buildings subjected to orthogonal horizontal ground motions

Measurement of intelligent computing via Levenberg Marquardt algorithm (LMA) for accurate prediction of fluid forces in a transient non-Newtonian thermal flow

Predicting precise results for the quantities of interest in time-dependent Computational Fluid Dynamics (CFD) simulations requires a significant investment of computational resources and time. To get around these issues, CFD simulations have been joined with Artificial Neural Networks (ANN). An optimally configured artificial neural network (ANN) is given the training and validation data sets produced by computational fluid dynamics (CFD). The flow around a cylinder, which is a well-known benchmark problem for incompressible flows, has been taken into consideration by the hybrid-CFD system. The mathematical model is based on the non-stationary Navier-Stokes and energy equations with viscosity. The basic architecture of the ANN model consists of 10 hidden layers, three output levels, and five input layers. Fast second-order LMA, a top-tier approach, was used to train the network. Both the Mean Square Error (MSE) and the coefficient of determination (R) provide statistical evidence that the ANN projected values for the drag and lift coefficients and average Nusselt number obtained from the finite element analysis are accurate. This analysis suggests that ANNs have the potential to significantly cut down on the amount of time and energy needed to run time-dependent simulations.

Application of the Wave Equation to the Dynamic Capacity of Grouped Piles

Two instrumented pile groups, including one group of seven H-Piles (HP 14x102 piles) and one group of seven Pipe Piles (457 mm diameter, closed-ended Pipe Piles) were driven at a site in Northeastern Arkansas. Stress waves in the piles were monitored during pile driving utilizing 350-ohm resistance strain gauges. The recorded stress waves were compared to estimated stress waves from wave equation analyses performed using GRLWEAP (2010). The pile forces for each monitored depth were compared to predicted forces from GRLWEAP (2010), Case Pile Wave Analysis Program (CAPWAP), and Pile Driving Analysis (PDA). The predicted stress waves were in good agreement with the observed stress waves (regression coefficient values greater than 0.8); however, there was considerable error regarding the prediction of maximum forces, particularly at the pile toe. The pre-construction wave equation predicted forces were within 50 percent (H-Piles) and 30 percent (Pipe Piles), respectively. However, at the pile toe, the errors

Data-assisted non-intrusive model reduction for forced nonlinear finite elements models

AbstractSpectral submanifolds (SSMs) have emerged as accurate and predictive model reduction tools for dynamical systems defined either by equations or data sets. While finite-elements (FE) models belong to the equation-based class of problems, their implementations in commercial solvers do not generally provide information on the nonlinearities required for the analytical construction of SSMs. Here, we overcome this limitation by developing a data-driven construction of SSM-reduced models from a small number of unforced FE simulations. We then use these models to predict the forced response of the FE model without performing any costly forced simulation. This approach yields accurate forced response predictions even in the presence of internal resonances or quasi-periodic forcing, as we illustrate on several FE models. Our examples range from simple structures, such as beams and shells, to more complex geometries, such as a micro-resonator model containing more than a million degrees of freedom. In the latter case, our algorithm predicts accurate forced response curves in a small fraction of the time it takes to verify just a few points on those curves by simulating the full forced-response.

A novel machine learning framework for impact force prediction of foam-filled multi-layer lattice composite structures

Numerical simulations can provide valuable insights for the optimization of design and operational management; however, they are often impractical and computationally intensive. Machine learning methods are appealing to these problems due to their sufficient efficiency and accuracy. In this study, a novel framework for predicting the impact responses of foam-filled multi-layer lattice composite structures (FMLCSs) was proposed by combining the accurate finite element (FE) analyses, surrogate models, fast Fourier transform (FFT) method, and inverse FFT (IFFT) method. Firstly, reliable FM models were established to simulate the crashworthiness of the five FMLCSs under impact loading, including an analysis of energy transformation. Subsequently, surrogate models, namely radial basis function (RBF), polynomial response surface (PRS), Kriging (KRG), and back propagation neural network (BPNN), combined with methods of FFT and IFFT, were employed to predict the impact force-time series of the FMLCSs. More than 1000 frequency points were employed for each type of FMLCS, and all the R-square (R2) values of the established surrogate models exceeded 0.95, indicating that the proposed framework accurately predicted the impact duration and impact responses in the frequency domain. In addition, parameter sensitivity analysis revealed that a high peak impact force was accompanied by a short impact duration. Moreover, increasing the lattice-web height resulted in a significant increase in the impact duration.

Prediction of feed force with machine learning algorithms in boring of AISI P20 plastic mold steel

Feed force plays a critical role in machining performance, including efficiency, hole quality, energy consumption, and tool life. While feed force is typically measured using dynamometers, these devices are often impractical in real-world experiments due to their complexity and cost. In particular, boring operations, which are essential for enlarging holes within tight tolerances, depend on precise feed force control to maintain high-quality hole outcomes. Achieving this precision requires reliable pre-process feed force estimation. However, traditional analytical methods for calculating feed force in boring operations often fall short, largely due to factors such as in-hole dynamics, the inclination of the rake surface, and the slenderness of the boring bar. This study aims to apply single and hybrid machine learning methods to accurately model and predict feed force in the boring process. Unlike previous research, which focused on feature selection, we propose the use of all possible combinations of input feature subsets to determine the optimal input features. The results indicate that feed force can be predicted effectively using these optimized feature combinations. Such accurate feed force prediction is crucial for effective process optimization, material selection, and process planning.

Enhanced Drag Force Estimation in Automotive Design: A Surrogate Model Leveraging Limited Full-Order Model Drag Data and Comprehensive Physical Field Integration

In this paper, a novel surrogate model for shape-parametrized vehicle drag force prediction is proposed. It is assumed that only a limited dataset of high-fidelity CFD results is available, typically less than ten high-fidelity CFD solutions for different shape samples. The idea is to take advantage not only of the drag coefficients but also physical fields such as velocity, pressure, and kinetic energy evaluated on a cutting plane in the wake of the vehicle and perpendicular to the road. This additional “augmented” information provides a more accurate and robust prediction of the drag force compared to a standard surface response methodology. As a first step, an original reparametrization of the shape based on combination coefficients of shape principal components is proposed, leading to a low-dimensional representation of the shape space. The second step consists in determining principal components of the x-direction momentum flux through a cutting plane behind the car. The final step is to find the mapping between the reduced shape description and the momentum flux formula to achieve an accurate drag estimation. The resulting surrogate model is a space-parameter separated representation with shape principal component coefficients and spatial modes dedicated to drag-force evaluation. The algorithm can deal with shapes of variable mesh by using an optimal transport procedure that interpolates the fields on a shared reference mesh. The Machine Learning algorithm is challenged on a car concept with a three-dimensional shape design space. With only two well-chosen samples, the numerical algorithm is able to return a drag surrogate model with reasonable uniform error over the validation dataset. An incremental learning approach involving additional high-fidelity computations is also proposed. The leading algorithm is shown to improve the model accuracy. The study also shows the sensitivity of the results with respect to the initial experimental design. As feedback, we discuss and suggest what appear to be the correct choices of experimental designs for the best results.