Search

PurposeIn the field of heritage science, especially applied to buildings and artefacts made by organic hygroscopic materials, analyzing the microclimate has always been of extreme importance. In particular, in many cases, the knowledge of the outdoor/indoor microclimate may support the decision process in conservation and preservation matters of historic buildings. This knowledge is often gained by implementing long and time-consuming monitoring campaigns that allow collecting atmospheric and climatic data.Design/methodology/approachSometimes the collected time series may be corrupted, incomplete and/or subjected to the sensors' errors because of the remoteness of the historic building location, the natural aging of the sensor or the lack of a continuous check of the data downloading process. For this reason, in this work, an innovative approach about reconstructing the indoor microclimate into heritage buildings, just knowing the outdoor one, is proposed. This methodology is based on using machine learning tools known as variational auto encoders (VAEs), that are able to reconstruct time series and/or to fill data gaps.FindingsThe proposed approach is implemented using data collected in Ringebu Stave Church, a Norwegian medieval wooden heritage building. Reconstructing a realistic time series, for the vast majority of the year period, of the natural internal climate of the Church has been successfully implemented.Originality/valueThe novelty of this work is discussed in the framework of the existing literature. The work explores the potentials of machine learning tools compared to traditional ones, providing a method that is able to reliably fill missing data in time series.

In this paper, a passive structural health monitoring (SHM) method capable of detecting the presence of damage in carbon fibre/epoxy composite plates is developed. The method requires the measurement of strains from the considered structure, which are used to set up, train, and test artificial neural networks (ANNs). At the end of the training phase, the networks find correlations between the given strains, which represent the ‘fingerprint’ of the structure under investigation. Changes in the distribution of these strains is captured by assessing differences in the previously identified strain correlations. If any cause generates damage that alters the strain distribution, this is considered as a reason for further detailed structural inspection. The novelty of the strain algorithm comes from its independence from both the choice of material and the loading condition. It does not require the prior knowledge of material properties based on stress-strain relationships and, as the strain correlations represent the structure and its mechanical behaviour, they are valid for the full range of operating loads. An implementation of such approach is herein presented based on the usage of a distributed optical fibre sensor that allows to obtain strain measurement with an incredibly high resolution.

PurposeClimate-induced damage is a pressing problem for the preservation of cultural properties. Their physical deterioration is often the cumulative effect of different environmental hazards of variable intensity. Among these, fluctuations of temperature and relative humidity may cause nonrecoverable physical changes in building envelopes and artifacts made of hygroscopic materials, such as wood. Microclimatic fluctuations may be caused by several factors, including the presence of many visitors within the historical building. Within this framework, the current work is focused on detecting events taking place in two Norwegian stave churches, by identifying the fluctuations in temperature and relative humidity caused by the presence of people attending the public events.Design/methodology/approachThe identification of such fluctuations and, so, of the presence of people within the churches has been carried out through three different methods. The first is an unsupervised clustering algorithm here termed “density peak,” the second is a supervised deep learning model based on a standard convolutional neural network (CNN) and the third is a novel ad hoc engineering feature approach “unexpected mixing ratio (UMR) peak.”FindingsWhile the first two methods may have some instabilities (in terms of precision, recall and normal mutual information [NMI]), the last one shows a promising performance in the detection of microclimatic fluctuations induced by the presence of visitors.Originality/valueThe novelty of this work stands in using both well-established and in-house ad hoc machine learning algorithms in the field of heritage science, proving that these smart approaches could be of extreme usefulness and could lead to quick data analyses, if used properly.

A model describing the fatigue behavior of composite materials under variable loadings is presented. The approach is based on a two-parameter model fulfilling the principal responses’ features of composite materials subjected to constant amplitude cyclic loadings. It is shown that when coupled with the equivalent residual strength assumption the model can describe the strength degradation of composite materials subjected to variable amplitude loadings without simplifications or substantial reductions of the loading spectrum. Compared to the linear Miner’s rule, the results highlight the importance of the sequence of loads and the approach discriminates the ascending, descending and random patterns. It is shown that for the case under study the ordered descending (high to low) sequence showed the worst scenario in terms of damage accumulation extent, while the random spectrum resulted in an intermediate response between ordered ascending and descending sequences.

In the present work, a model describing the fatigue behaviour of composite materials under spectrum loading is presented. The approach is based on a two-parameter model describing the strength degradation kinetics for fibre-reinforced materials subjected to cyclic loadings, both with constant and variable amplitudes. The point of strength of the presented approach is that it allows to consider cyclic loadings as they are, without simplifications or reductions. In particular, the analytical background of the model is based on the statements that govern the composites behaviour and focused on the issues related to this class of materials when subjected to fatigue. Then, the final step is the definition of a damage accumulation rule that goes over Miner’s rule unreliability when applied to fibre-reinforced materials. As a matter of fact, it allows to take into account the effects of different loading histories, underlining the importance of the sequence of loads application in the framework of the presented approach.

This paper discusses the effect of loading rate on the static strength and fatigue behavior of glass/polyester and graphite/epoxy composites. It is shown that the rate-sensitivity of the static strength and the fatigue data can be analytically taken into account in the framework of a two-parameter “residual strength” model, inherently fulfilling the Strength-Life Equal-Rank Assumption (SLERA). Based on the static strength's loading rate sensitivity, a method to use the fatigue life data obtained at any fixed frequencies to predict the fatigue life under different frequencies is proposed. The method's reliability is verified by predicting different data sets of CA fatigue life and residual strength data obtained on glass/polyester composites under different loading rate conditions. The responses of angle-ply carbon/epoxy laminates to arbitrary two-block cyclic loadings in ascending and descending frequency sequences revealed the method's capability to discriminate between the applied loading sequences and the potential to develop a numerical routine accounting for the cycle-by-cycle variability during realistic loading conditions.

The static strength of carbon/epoxy laminates with three different stacking sequence has been analyzed at different strain rates under three point flexural tests. The materials under study were low cost, low Tg carbon/epoxy laminates with nominal thickness of 1 mm obtained stacking four fabric 0/90 balanced prepregs. The Weibull distribution function was used to compare the data on statistical basis.Quasi–isotropic, cross-ply and angle ply laminates were tested and the strain dependence of static strength was accounted for in the framework of a phenomenological model for fatigue with the objective test the model reliability in predicting the fatigue life and residual strength under variable spectrum with block loadings characterized by different loading frequencies.Preliminary results show that the phenomenological approach can be suitably applied in a rather simple manner solving one of the most intricate tasks within the field of fatigue of composite materials. The research activity represents the last step to build up a software that under variable amplitude loadings comprehensively accounts simultaneously and cycle-by-cycle for the change in the loading ratio, R=σmin/σmax, the maximum applied stress, σmax, and the frequency of loading, f. Details of the approach and the physics behind it are briefly presented with a series of preliminary results.

This paper provides a literature review of one of the most notable models relevant for modelling the fatigue behavior of composite materials. The review first introduced the importance of modelling fatigue in respect of the residual strength degradation of the studied materials, before classifying fatigue models between two categories: the phenomenological models and the mechanistic ones. Among the fatigue modelling approaches, one of them belonging to the category of the “residual strength models” is considered and deepened. In particular, it is analyzed from its genesis to the latest outcomes, passing through the different goals gained during the years.

The paper describes a system that it is able of monitoring the health state of a composite structure in real time. The hardware of the system consists of a wire of strain sensors connected to a control unit. The software of the system elaborates the strain data and in real time is able to detect the presence of an eventual damage of the structures monitored with the strain sensors. The algorithm requires as input only the strains of the monitored structured measured on real time, i.e. those strains coming from the deformations of the composite structure due to the working loads. The health monitoring system does not require any additional device to interrogate the structure as often used in the literature, instead it is based on a self-learning procedure. The strain data acquired when the structure is healthy are used to set up the correlations between the strain in different positions of structure by means of neural network. Once the correlations between the strains in different position have been set up, these correlations act as a fingerprint of the healthy structure. In case of damage the correlation between the strains in the position of the structure near the damage will change due to the change of the stiffness of the structure caused by the damage. The developed software is able to recognize the change of the transfer function between the strains and consequently is able to detect the damage.

In this paper the results coming from modelling the fatigue life and residual strength of fiber reinforced composites subjected to variable amplitude (VA) loadings are reported. The approach is based on a recently developed two-parameter residual strength model where the damage accumulation rule relies on the “equivalent strength approach” that allows considering variable loading sequences without any simplifying assumption.The development of a numerical procedure is illustrated by a series of simulations following High-Low (H-L), Low-High (L-H) and random sequences of loadings. The analytical predictions reveal that the sequence and interaction of events has a strong influence both on the rate of damage accumulation and fatigue life.

A method is proposed to predict the fatigue life and residual strength of fiber reinforced composites subjected to variable amplitude (VA) loadings. Based on a recently developed two-parameter residual strength model, the damage accumulation rule is based on the “equivalent strength approach” allowing to account for the variable loading without any simplifying assumption. The cycle-by-cycle effects can be easily recovered on the basis of strength degradation kinetics and the approach reliability is checked on the basis of a series of constant amplitude (CA) literature data.The approach allows correcting the imperfections of Miner’s rule and its unreliability when applied to fiber composites and to build on a solid foundation a numerical procedure for fatigue calculations under spectrum loadings. High-Low (H-L) and Low-High (L-H) three-block sequences of loadings have been simulated highlighting the basic concepts of the proposed damage rule. The analytical predictions reveal that the sequence and interaction of events has a strong influence both on the rate of damage accumulation and fatigue life.