Multi-level Model Calibration for the Assessment of Historical Masonry Structures

Informations projet

MultiCAMS

N° de convention de subvention: 744400

DOI

10.3030/744400

Projet clôturé

Date de signature de la CE 30 Mars 2017

Date de début 1 Septembre 2017

Date de fin 31 Août 2019

Financé au titre de

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Coût total

€ 183 454,80

Contribution de l’UE

€ 183 454,80

Coordonné par

IMPERIAL COLLEGE OF SCIENCE TECHNOLOGY AND MEDICINE
United Kingdom

Periodic Reporting for period 1 - MultiCAMS (Multi-level Model Calibration for the Assessment of Historical Masonry Structures)

Période du rapport: 2017-09-01 au 2019-08-31

Résumé du contexte et des objectifs généraux du projet

Masonry structures constitute an important part of the built environment and architectural heritage in seismic areas. A large number of these old structures showed inadequate performance and suffered substantial damage under past earthquakes. Realistic numerical models are required for accurate response predictions under dynamic loading and for addressing the implementation of effective strengthening solutions. Numerical macromodels used in the professional practice masonry are generally not able to capture the complexity of masonry response. On the other hand, detailed mesoscale approaches developed in the scientific community are more accurate but require computational resources which are beyond what is available to ordinary practitioners. An innovative approach retaining the low computational cost of macro-models and the calibration simplicity of mesoscale models is thus desirable.
The topic of structural assessment of historical heritage is very timely, since it responds to the societal demands for resilient structures, and recent catastrophic seismic events in Europe have highlighted the necessity of urgent assessment and strengthening of historical buildings. Eurocode 8, Part 3 (informative) and the international standard ISO 13822 provide only general principles for the assessment of existing structures, the latter requiring that “Structural performance shall be analysed using models that reliably represent the actions on the structure, the behaviour of the structure, and the resistance of its components”. At the moment, however, the use of accurate dynamic analysis is precluded for the assessment of historical masonry heritage, leading to large approximations and low accuracy in the assessment of their structural behaviour under earthquake. This means possibly unconservative predictions which may jeopardize human lives in many European urban areas, as confirmed by recent catastrophic events.
The MultiCAMS project has developed an advanced computational strategy including numerical modelling and innovative model calibration for the realistic assessment of historical unreinforced masonry structures. The proposed assessment approach is based on the use of nonlinear FE models with different levels of sophistication including detailed mesoscale models, where masonry units and mortar joints are modelled separately, and homogeneous models assuming masonry as a continuum material. An innovative model calibration strategy utilising inverse analysis techniques has been developed to capitalise on the efficiency of the novel simplified mechanical model and the accuracy of existing detailed mesoscale descriptions whose material properties can be obtained in simple and non-invasive in-situ physical tests. The main outcome of MultiCAMS is thus a comprehensive and accurate methodology for the seismic assessment of historical masonry buildings.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

The project has developed through the following steps:
• Review of the mechanical behaviour of masonry structures under earthquake and state-of-art methodologies for modelling at different scales of representation, including micro-, meso-, macro- and multi-scale. This review has also covered methodologies and drawbacks in the material property estimation for the modelling strategies at different scales of representation.
• Selection of a case study involving extensive experimental activities regarding masonry at material (small samples), component (walls) and structural (building) levels. This case study has been then modelled by means of the mesoscale strategy developed at Imperial College and its accuracy and efficiency properly assessed.
• Development of a plastic-damage material model for masonry modelled as a homogeneous material, corresponding to a macro-scale representation, which is in general suitable for dynamic analysis of masonry buildings using standard computational resources. This material model, initially isotropic, has been then extended to orthotropic behaviour, which is representative of masonry in case of regular bond.
• A multi-level calibration methodology has been developed, where simple material tests are used to calibrate a mesoscale numerical model, which in turn is used to represent suitable virtual tests. The output of these virtual tests is then used to identify model parameters of the proposed macroscale material. This procedure makes use of Genetic Algorithms optimisation to achieve the minimum discrepancy in terms of stress power along the loading history. The accuracy of the material identification is assessed considering multiple virtual tests as validation. The overall procedure has been then validated against the experimental results of the case study, and remarkable levels of accuracy in terms of maximum strength, initial stiffness, cracking patterns and hysteretic energy have been realised.
• Additional calibration methodologies based on dynamic data have been developed during the secondment in collaboration with Prof. Gattulli at Sapienza University of Rome. The proposed two-step methodology is based on the estimation of structural modal properties in undamaged and damaged conditions after the occurrence of an earthquake and the identification of a suitable material model which predicts the observed modifications of such modal properties.
• The study results have been published/submitted for publication in journals and presented at various international conferences and seminars.

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

A number of micro- and macro-models have been proposed in the past decades, but realistic predictions can only be achieved when model material parameters are correctly calibrated. The calibration of mesoscale material parameters is relatively simple, as it is based on non-invasive material tests on small specimens. On the other hand, the calibration of macro-model material parameters is problematic, as it should be based upon intrusive in-situ tests carried out on large masonry portions representative of the average masonry “material”. From the point of view of efficiency, macro-models are suitable for use in accurate dynamic simulations of masonry structures while mesoscale models require resources which are beyond what is ordinarily available to practitioners. Multiscale approaches have been proposed in previous research, but while they are very effective in generating macro-scale material models with the same accuracy of underlying micro-models, requiring at the same time the calibration of material properties of constituents as the latter models, they do not show significant improvement in computational efficiency, hence they are restricted to the analysis of very small structures without significant impact on the professional sector.
An innovative approach retaining the low computational cost of macro-models and the calibration simplicity of mesoscale models has been developed in MultiCAMS project and is expected to constitute an important step towards the widespread use of realistic nonlinear analysis for the assessment and the safeguard of historical heritage. The strategy could potentially also be extended to different fields where the need for balancing efficiency and accuracy at different scales is needed, i.e. environmental engineering, biomedical sciences, advanced material design.

Prediction of yield surface by means of homogenised macroscale model and comparison with experiments

Prediction of earthquake effects by mesoscale model and comparison with experimental crack pattern

Outline of the multi-level model calibration

Prediction of earthquake effects by macroscale model and comparison with experimental crack pattern