publications | Alberto Ciampaglia

2024

Review on machine learning techniques for the assessment of the fatigue response of additively manufactured metal parts

Alessio Centola, Andrea Tridello, Alberto Ciampaglia, and 2 more authors

Aug 2024

Abs

The present review paper addresses the increasing interest in the application of machine learning (ML) algorithms in the assessment of the fatigue response of additively manufactured (AM) metal alloys. This review aims to systematically collect, categorize, and analyze relevant research papers in this domain. The most commonly used ML algorithms are presented, discussing their specific relevance to the fatigue modeling of AM metal alloys. A detailed analysis of the most relevant input features used in the literature to predict the main parameters related to the fatigue response is provided. Each work has been analyzed to highlight its strengths and peculiarities, thereby offering insights into novel methodologies and approaches for addressing critical challenges within this field. Particular attention is dedicated to the role of defects and the related size-effect, as they strongly influence the fatigue response. In conclusion, this review not only synthesizes existing knowledge but also offers forward-looking recommendations for future research directions, providing a valuable resource for researchers in the domain of ML-assisted fatigue assessment for AM metal alloys.
A Statistical Mesoscale Approach to Model the Size Effect on the Tensile Strength of Notched Woven Composites

Andrea Ferrarese, Carlo Boursier Niutta, Alberto Ciampaglia, and 1 more author

Applied Sciences (Switzerland), Apr 2024

Abs

The scaling of the strength of composite parts with part size is referred to as the size effect. In the presence of notches, stress concentration affects a portion of material that increases with the notch size. Furthermore, in woven composites, the notch and tow size can be comparable, thus demanding a mesoscale approach to properly capture the stress intensification. In this paper, a probabilistic mesoscale method to model the size effect in notched woven composites is presented. First, the stress distribution is estimated with a finite element model, calibrated on experimental Digital Image Correlation data. The FE model simulates the mesoscale heterogeneity of the woven reinforced material and replicates the local stress intensification at the tow level. Then, a three-parameter Weibull-based statistical model is introduced to model the probability of failure from the calculated stress distribution and the volume of the part. An equivalent stress is used to capture the relevant fiber and matrix failure modes and the maximum value within the specimen volume is the random variable of the model. The method is applied to open-hole tension tests of a woven twill carbon fiber–epoxy composite. Two specimen widths and three width-to-diameter ratios, from 3 to 12, are considered. Specimen width produced an observable size effect, whereas the variation of hole size in the range considered did not. The statistical model is found to accurately describe the experimental observations, efficiently replicating an inverse size effect, regardless of hole size, while wider specimens lead to a lower probability of failure.
Design of a Lightweight Origami Composite Crash Box: Experimental and Numerical Study on the Absorbed Energy in Frontal Impacts

Alberto Ciampaglia, Luca Patruno, and Raffaele Ciardiello

Journal of Composites Science, Jun 2024

Abs

Origami-shaped composite structures are currently being explored for their ability to absorb energy in a progressive and controlled manner. In vehicle passive safety applications, this prevents the occurrence of peak forces that could potentially cause injuries to vehicle passengers. The work presents the design of a carbon fiber-reinforced polymer (CFRP) crash box for a Formula Student race car, using a numerical model validated by experimental tests. An initial characterization of the material is conducted according to the standards. Following, six origami samples are manufactured and subjected to crash tests to gather accurate experimental data. The numerical model is validated on the tests and used for the design of the race car’s impact attenuator. The designed crash box meets the Formula Student requirements while reducing the total mass by 14% and the maximum deceleration of 21% compared with the previous design. The study confirms the potential use of origami structures to improve crashworthiness while reducing vehicle weight.

2023

Machine learning methods to predict the fatigue life of selectively laser melted Ti6Al4V components

Alessio Centola, Alberto Ciampaglia, Andrea Tridello, and 1 more author

Jun 2023

Abs

The aim of the present paper is to predict the fatigue life of Selectively Laser Melted (SLMed) Ti6Al4V components via the process parameters, the thermal treatments, the surface treatments and the stress amplitude, adopting machine learning techniques to reduce the cost of further fatigue testing, and to deliver better predictive fatigue designs. The studies resulted in reliable algorithms capable of predicting trustful fatigue curves. The methods have been trained with experimental data available in the literature and validated on testing sets to assess the extrapolation limits and to compare the different methods. The behavior of the networks has also been mapped by varying one SLM process parameter at the time, highlighting how each one affects the life. © 2023 The Authors. Fatigue & Fracture of Engineering Materials & Structures published by John Wiley & Sons Ltd.
Data driven method for predicting the effect of process parameters on the fatigue response of additive manufactured AlSi10Mg parts

A. Ciampaglia, A. Tridello, D.S. Paolino, and 1 more author

Jun 2023

Abs

The fatigue response of Additive Manufacturing (AM) components is driven by manufacturing defects - whose size mainly depends on process parameters - and by the resulting microstructure - mainly affected by heat treatments and process parameters. In the paper, Machine Learning (ML) algorithms are applied to estimate the fatigue response from AM process parameters and heat treatment properties. Feed-forward neural networks (FFNN) and physics-informed neural network (PINN) algorithms are designed and validated on literature datasets of AM AlSi10Mg alloy, proving the effectiveness of physics-based ML approaches in predicting the fatigue response of AM parts. Leveraging PINN interpretability, the authors analyse the relationship between process parameters and fatigue response. © 2023 Elsevier Ltd
Physics based data driven method for the crashworthiness design of origami composite tubes

Alberto Ciampaglia, Dario Fiumarella, Carlo Boursier Niutta, and 2 more authors

Jun 2023

All Open Access, Hybrid Gold Open Access

Abs

A novel method based on a physics informed data driven model is developed to design an origami composite crash tube. The structure consists of two axially stacked basic components, called modules. Each module presents lower and upper square sections with an octagonal section in the middle. The parameters of the octagonal cross-section and the height of each module are optimized to maximize the energy absorption of the tube when subjected to an axial impact. In contrast to standard surrogate modelling techniques, whose accuracy only depends on the amount of available data, a Physics-informed Neural Network (PINN) scheme is adopted to correlate the crushing response of the single modules to that of the whole origami tube, constraining the data driven method to physically consistent predictions. The PINN is first trained on the results obtained with an experimentally validated Finite Element model and then used to optimize the structure. Results show that the PINN can accurately predict the crushing response of the origami tube, while consistently reducing the computational effort required to explore the whole design domain. Also, the comparison with a standard Feed Forward Neural Network (FFNN) shows that the PINN scheme leads to more accurate results. © 2023, The Author(s).
CompStruct

Data driven statistical method for the multiscale characterization and modelling of fiber reinforced composites

A. Ciampaglia

Jun 2023

Abs

Multiscale analysis of composite laminates allows for predicting the mechanical response of these materials avoiding cumbersome experimental campaigns. The matrix and fibre material properties and the size of the Representative Volume Element (RVE) are the main parameters affecting the accuracy of multiscale models. This paper proposes a statistical inverse method to calibrate micromechanical material parameters from macroscale experiments and 3D reconstruction. First, glass fiber reinforced epoxy laminates have been analysed with Computer Tomography (CT), then, the material 3D microstructure has been reconstructed and fibre, matrix, and voids were segmented. Tensile tests have been performed on the composite specimen, measuring the surface strains with a Digital Image Correlation (DIC) system. The reconstructed volume, converted to a voxel mesh, has been used to compute the homogenized response of composite by Fast Fourier Transform (FFT) analysis. By comparing the marginal distribution of homogenized material stiffness extracted from DIC data of tensile tests, with the conditioned distribution computed by varying the FFT model parameter, a Stochastic Volume Element (SVE) is finally calibrated. A probabilistic multiscale model based on the SVE that propagates the uncertainty from the microscale to the structure level is presented. © 2023 Elsevier Ltd
Multifunctional material design for strain sensing: Carbon black effect on mechanical and electrical properties of polyamides

Alberto Ciampaglia, Raffaele Ciardiello, Federico Cesano, and 2 more authors

Composite Structures, Jan 2023

Abs

This work investigates the coupled effect on mechanical and electrical properties of carbon black dispersed in polyamide 6 and 6.6 matrices. The CB content is varied between 15 % and 25 % wt. The elastic modulus increases of 12 % by increasing the CB concentration. Results reveal the capability of CB to functionalize thermoplastic polymers by activating conductive networks, as obtained over the percolation threshold of 13 % wt. The conductivity sensitivity to the mechanical strain is analyzed within direct (DC) and alternate current (AC) in the range of 0–100 kHz. Composite with lower CB concentrations exhibited a linear increase of the Gauge Factor (GF) with the frequency (from 2 to 4 at 20 %wt), while at 25 % wt of CB, the GF is 17 in DC regime and linearly decreases toward 6, at 100 kHz. A novel model for the estimation of the material Gauge Factor (GF) variation with the applied electric frequency is proposed.
Experimental and Numerical Investigation of the Mesoscale Size Effect in Notched Woven Composites

Andrea Ferrarese, Carlo Boursier Niutta, Alberto Ciampaglia, and 3 more authors

Applied Sciences (Switzerland), Apr 2023

Abs

The uniaxial strength of 2 × 2 twill carbon fiber-epoxy composite with circular open holes is governed by the stress concentration induced by the notch, as well as by the variation of the material response with the stress localization. In this study, these concurrent phenomena were studied using an experimental-numerical approach, considering the effect of the composite heterogeneity. An innovative simplified mesoscale model using shell elements to replicate the woven pattern was developed. The material properties of the model were identified with a data-driven optimization scheme that minimizes the difference between the experimental full-field strain, measured with Digital Image Correlation, and the one predicted by the model itself. The elastic properties identified for the material at tow level are analogous to that of a unidirectional composite, with longitudinal and transverse moduli of around 102 GPa and 15 GPa, respectively. The calibrated model was used to analyze the stress concentration and localization induced by circular notches with increasing diameter. A size effect describing the increase in the carbon fiber tow strength following the increase in stress localization was observed. The mesoscale material strength was found to increase by about 2.5 times in the presence of the notch, from 2131 to 3922 MPa in fiber direction, partially balancing the stress intensification effect of the hole that leads to an approximately three-fold stress increase in the material.
Experimental and Numerical Investigation of the Mesoscale Size Effect in Notched Woven Composites

Andrea Ferrarese, Carlo Boursier Niutta, Alberto Ciampaglia, and 3 more authors

Applied Sciences (Switzerland), Apr 2023

Abs

The uniaxial strength of 2 × 2 twill carbon fiber-epoxy composite with circular open holes is governed by the stress concentration induced by the notch, as well as by the variation of the material response with the stress localization. In this study, these concurrent phenomena were studied using an experimental-numerical approach, considering the effect of the composite heterogeneity. An innovative simplified mesoscale model using shell elements to replicate the woven pattern was developed. The material properties of the model were identified with a data-driven optimization scheme that minimizes the difference between the experimental full-field strain, measured with Digital Image Correlation, and the one predicted by the model itself. The elastic properties identified for the material at tow level are analogous to that of a unidirectional composite, with longitudinal and transverse moduli of around 102 GPa and 15 GPa, respectively. The calibrated model was used to analyze the stress concentration and localization induced by circular notches with increasing diameter. A size effect describing the increase in the carbon fiber tow strength following the increase in stress localization was observed. The mesoscale material strength was found to increase by about 2.5 times in the presence of the notch, from 2131 to 3922 MPa in fiber direction, partially balancing the stress intensification effect of the hole that leads to an approximately three-fold stress increase in the material.
Assessment of the Critical Defect in Additive Manufacturing Components through Machine Learning Algorithms

Andrea Tridello, Alberto Ciampaglia, Filippo Berto, and 1 more author

Applied Sciences (Switzerland), Apr 2023

Abs

The design against fatigue failures of Additively Manufactured (AM) components is a fundamental research topic for industries and universities. The fatigue response of AM parts is driven by manufacturing defects, which contribute to the experimental scatter and are strongly dependent on the process parameters, making the design process rather complex. The most effective design procedure would involve the assessment of the defect population and the defect size distribution directly from the process parameters. However, the number of process parameters is wide and the assessment of a direct relationship between them and the defect population would require an unfeasible number of expensive experimental tests. These multivariate problems can be effectively managed by Machine Learning (ML) algorithms. In this paper, two ML algorithms for assessing the most critical defect in parts produced by means of the Selective Laser Melting (SLM) process are developed. The probability of a defect with a specific size and the location and scale parameters of the statistical distribution of the defect size, assumed to follow a Largest Extreme Value Distribution, are estimated directly from the SLM process parameters. Both approaches have been validated using literature data obtained by testing the AlSi10Mg and the Ti6Al4V alloy, proving their effectiveness and predicting capability.
Data-driven method to assess the influence of process parameters on the fatigue response of additively manufactured Ti6Al4V

Alberto Ciampaglia, Andrea Tridello, Filippo Berto, and 1 more author

In Procedia Structural Integrity, Apr 2023

Abs

The fatigue behavior of Additive Manufacturing (AM) parts is influenced by manufacturing defects, whose dimensions are primarily determined by the parameters of the AM process, which, in turn, also affect the resulting microstructure, together with heat treatments. This study employs Machine Learning (ML) techniques to forecast the fatigue response of AM parts from the AM process variables and the heat treatment characteristics. Feed-forward neural networks (FFNN) and physics-informed neural network (PINN) models are formulated and verified employing published datasets on AM Ti6Al4V alloy. The results demonstrate that physics-based ML approaches are effective in forecasting the fatigue response of AM components.

2021

CompStruct

Impact response of an origami-shaped composite crash box: Experimental analysis and numerical optimization

A. Ciampaglia, D. Fiumarella, C. Boursier Niutta, and 2 more authors

Apr 2021

Abs

In this work, an investigation of the crush response of a simplified CFRP origami crash box subjected to axial impact is proposed. Crash boxes are thin-walled structural components of the vehicles designed to absorb energy during impact events at low-medium velocity. In particular, the crash boxes must guarantee a progressive and controlled energy absorption, avoiding peak of force (and thus acceleration) that can lead to passenger injury. In recent years, crash boxes made of carbon fibre reinforced polymer (CFRP) have found application in the automotive sector. However, their brittle failure mode leads to an irregular crushing trend characterized by peak of force. Thereafter, the crushing behaviour of the composite material structures can be improved by modifying their geometrical parameters. Among the most promising solutions, the origami structure is increasingly considered for crash boxes. The origami crash box here considered consists of four axially stacked basic structures. Each basic structure is composed of four trapezoidal faces and four triangular faces. The upper cross section is squared, whereas the lower cross section has an octagonal shape. The structural behaviour of the origami component was investigated according to different sizes of the triangular faces. The numerical models were simulated with the finite element commercial code LS-Dyna in its explicit formulation. The optimal shape of the origami structure in terms of maximum energy absorption and limited force peak was defined in LS-OPT environment. The objective function of the shape optimization algorithm was set to maximize the energy absorption, while limiting the peak of force. The optimal shape defined presented larger sizes in the top basic structures than in the bottom parts, resulting in more inclined faces. The result suggested that more inclined faces in the top part can guarantee a fracture-triggering effect in the crash box, which ensured a smaller peak force. © 2020 Elsevier Ltd
Artificial Intelligence for Damage Detection in Automotive Composite Parts: A Use Case

Alberto Ciampaglia, Antonio Mastropietro, Alessandro De Gregorio, and 3 more authors

SAE Technical Papers, Apr 2021

Abs

The detection and evaluation of damage in composite materials components is one of the main concerns for automotive engineers. It is acknowledged that defects appeared in the manufacturing stage or due to the impact and/or fatigue loads can develop along the vehicle riding. To avoid an unexpected failure of structural components, engineers ask for cheap methodologies assessing the health state of composite parts by means of continuous monitoring. Non Destructive Technique (NDT) for the damage assessment of composite structures are nowadays common and accurate, but an on-line monitoring requires properties as low cost, small size and low power that do not belong to common NDT. The presence of a damage in composite materials, either due to fatigue cycling or low-energy impact, leads to progressive degradation of elastic moduli and strengths. Since there is a well-known relationship between the elastic modulus reduction and the amount of damage, the stiffness degradation can be used for the scope of detecting the position and the amount of damage that has taken place. Relying on these concepts, a novel strain-based damage sensing procedure is here proposed, that can identify damages in composite structures by processing strain measures from a distributed sensors network. To achieve this result a combined Machine Learning pipeline, composed by Principal Component Analysis (PCA) and One Class Support Vector Machine (OC-SVM) is proposed. First, PCA learns a linear transformation on the undamaged measurements to reduce the data dimensionality; secondly, OC-SVM trained to detect anomalies in the projected components. A cross-validation procedure is used to find the optimal pipeline configuration. The methodology is virtually tested on a carbon fiber suspension. The results suggest dropping the first components of the PCA to feed the classifier. In addition, results show the capability of the algorithm to detect anomalies in the component strain response.
Design and analysis of automotive lightweight materials suspension based on finite element analysis

Alberto Ciampaglia, A. Santini, and G. Belingardi

Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, May 2021

Abs

In this paper, an integrated method for the design of suspension components with lightweight materials is proposed. Starting from the well-established design flow for chassis components, the paper analyse the manufacturing and design aspect which are mutual dependent in composite structures, providing technical guidance. The proposed design methodology is applied to an innovative suspension archetype patented by Magneti Marelli, F.L.E.C.S. (Flexible Link Elevated Compliance Suspension). First, an analysis of available lightweight material is carried out to define which one best fits the design requirements in terms of density, stiffness, cost and manufacturability. Second, structural simulations, based on finite element analysis, are performed to investigate the influence of the thickness and material on the elasto-kinematic performance of the suspension. According to the stress distribution, an optimization of the shape and the stacking sequence of the component is performed. The results show that 8.5 mm carbon fibre reinforced epoxy is the best solution for the blades, and 5 mm aluminium alloy for the reinforcement tube wall thickness. Third, with respect to the final design, an adhesive interface is analysed and designed, using the Goland-Reissner model.