Characterization of masonry structural members in compression by modal analysis-based methods
- Autores: Miquel Llorens Sulivera
- Directores de la Tesis: Pere Roca Fabregat (dir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2011
- Idioma: inglés
- Tribunal Calificador de la Tesis: Climent Molins Borrell (presid.), Francesc Xavier Cahís Carola (secret.), Zdravko Bonev Petkov (voc.), Geert Lombaert (voc.), Peter Van Broeck (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Numerical and experimental approaches have been used for characterizing the behaviour of these ancient structures.
Nowadays, a combined approach taking into account both possibilities seems to be the best option.
This presentation deals with the characterization of compressed masonry members, using modal analysis-based techniques. For the sake of simplicity, the theoretical developments consider these members almost or completely isolated, neglecting the influence of possible surrounding existent structures.
The effectiveness of a process like the one that is presented in this dissertation depends to a great extend, on the sensitivity exhibited by the final result, when the variables of the problem, experience small changes. Therefore, to establish beforehand, which are the variables that more influence the global behaviour, and to determine which their usual range of variability is - even in a coarse way ¿ are key issues.
In the presentation, the influence of the different parameters is analyzed. The information is mainly provided by two sources: the available literature, and the numerical simulation of a huge number of cases. The two more characteristic properties of these structural elements are stressed below: .- The influence of the curvature.
This is the geometric characteristic that differentiates the studied typologies: elements without curvature, elements with curvature in a single plane and finally double curved structural members.
.- The influence of axial stresses.
A significant part of the presented developments, have been addressed to determine how the dynamic behaviour vary depending on whether the axial stresses are related to the mass of the member itself, or to external requests. The stiffening effect in the boundary conditions due to the axial forces has been considered as well.
The Young¿s modulus , the Poisson and damping ratios, the geometric imperfections or the characteristics of the chosen finite element and its application in the numerical description of the structure, are some of the parameters which influence has been studied in this previous analysis.
In the framework of the presentation, a simplified procedure for determining the axial load and the boundary conditions of under compression masonry existing structures, based on modal models has been developed.
The procedure can be extended to obtaining the value of other variables although it¿s not pos s ible to increas e their number.
Like most of the structural or assessment analyses, the procedure deals with errors. Two different kinds of errors are summarized below: .- Averaged errors.In this classification are gathered those errors which are difficult to determine and those which are avoidable or with a contemptible influence. All of them are grouped in a homogenised value. This error is introduced in the process as a percentage.
.- Singular errors.Those errors are connected to those variables that have a significant influence in the final result. The error is bounded by establishing an upper and a lower value.
By means of a numerical model, a field of possible solutions is generated. This is the so-called numerical model.
In parallel, the dynamic behaviour of the existing analyzed structure is characterized, building with the results, the modal model. The quality of the modal data is verified using a number of assessment techniques.
The procedure is based on the fact that the objective variables are inherent properties of the system analyzed; therefore, their value should be determined no matter which is the modal data, used for it. Solving the problem, simultaneously using different experimental frequencies, should give as a result a redundant system with a repeated solution.
All the paired values corresponding to the objective variables which fulfil the condition of matching the calculated frequency with the one experimentally determined are gathered in a curve.
By intersecting several of them, a single point should be determined.
