Scipedia: F. Salazar's Conference proceedings

A free software for dam monitoring data analysis: exploration, curation and machine learning model fitting

Fernando Salazar — Mon, 26 Jun 2023 11:35:03 +0200

Pre-processing of dam data coupled with behavior analysis through machine learning

Fernando Salazar — Thu, 07 Apr 2022 09:31:03 +0200

A predictive study framework for dams is proposed using Machine Learning (ML) algorithms programmed in R. This study is capable of estimating the
relationship between dam response and internal properties or external forces. For a better efficiency of the analysis with predictive models, a previous step
has been developed consisting of database cleaning and expansion processes. An interactive interface for each of the two phases has been developed using the Shiny package with the aim of making these processes accessible also to professionals who do not possess knowledge of laborious programming scripts.

Improving hydrological safety by rehabilitating the stilling basin of a gravity dam using numerical modelling

Fernando Salazar — Mon, 14 Mar 2022 12:15:03 +0100

La presa de Sant Ponç entró en funcionamiento en 1954. Es de gravedad de planta recta, con 59,5 m de altura sobre cimientos. Su aliviadero consta de un vertedero recto de tres vanos de 17 m cada uno regulados por compuertas. La estructura de disipación es singular e incluye un tramo de solera no hormigonado que ha sufrido erosiones. Aunque no comprometen la seguridad de la presa, se ha diseñado una solución para la mejora de su funcionamiento, que se ha basado fundamentalmente en los resultados de modelos numéricos. Los avances en las prestaciones de estas herramientas permiten analizar con detalle el comportamiento hidráulico de las alternativas planteadas en diversas situaciones de vertido, así como el de los desagües intermedios y de fondo.

Analysis of La Baells Dam monitoring data with machine learning techniques

Fernando Salazar — Mon, 14 Mar 2022 09:55:04 +0100

La Baells arch dam has a height above the foundations of 102 m. It entered service in 1976 and since then it had correct performance, which has been controlled with a very complete monitoring system and conventional interpretation techniques. This communication presents the results of a pilot study in which machine learning techniques have been applied to analyze their monitoring data. These techniques offer greater flexibility than conventional procedures for generating dam response prediction models, as well as for identifying relationships between devices of a different nature. This allows a better understanding of the behavior of the structure and greater safety control.

Combination of advanced numerical methods and machine learning for dam safety assessment

Andre Conde — Wed, 29 Apr 2020 17:36:03 +0200

The availability of machine learning techniques opens up possibilities in different fields of civil engineering. Their application in conjunction with numerical simulations overcomes the limitations in traditional approaches and pave the road for some new horizons. This communication presents several applications of such a hybrid tool in design and safety assessment of dams and hydraulic structures. They include the generation of behavior prediction models from monitoring data, the identification of behavior patterns with classification models, the analysis of the seismic response of gravity dams with heterogeneous concrete in a probabilistic framework, the investigation of the performance of arch dams, and the estimation of the discharge capacity of arched labyrinth spillways.

Mejora de la seguridad hidrológica e incremento de la capacidad de embalse de presas de fábrica mediante aliviaderos con cajeros altamente convergentes

Fernando Salazar — Tue, 21 Jan 2020 08:46:04 +0100

El empleo en aliviaderos de presas de fábrica de canales laterales para la recogida de los vertidos en la zona de estribos ha sido una solución utilizada con relativa frecuencia lo largo de la historia en casos en los que la longitud de vertido era sensiblemente superior a la anchura del valle aguas abajo de la presa. Este tipo de aliviaderos en presas existentes puede tener como objetivo el aumento de capacidad de su aliviadero o bien el aumento de los niveles máximos de explotación del embalse manteniendo la capacidad de desagüe con los resguardos exigidos. A pesar de estos antecedentes, no existen estudios metodológicos que proporcionen criterios para su diseño, ni conclusiones sobre el efecto que producen este tipo de aliviaderos en la disipación de energía en cuencos de pie de presa. Ante esta situación, se está desarrollando el proyecto de investigación aplicada CALA, cuyo objeto principal es definir criterios para el diseño de ese tipo de aliviaderos y el desarrollo de un software de cálculo que permita su predimensionamiento y optimización.

Tratamiento y análisis inteligente de datos del comportamiento de estructuras hidráulicas

Fernando Salazar — Tue, 21 Jan 2020 08:35:02 +0100

El rendimiento de los dispositivos de monitorización ha experimentado mejoras permitiendo disponer de más información sobre el comportamiento de las estructuras. Sin embargo, las inversiones realizadas en modernización de sistemas de medida no se recuperan a menos que se complementen con aplicaciones capaces de manejar una información tan amplia y diversa.

En esta contribución, se presenta una herramienta de software para importar, explorar, limpiar y analizar datos de monitorización. Además, permite la generación de modelos de predicción basados en machine learning, así como la interpretación de la respuesta del sistema a las acciones o cargas en funcionamiento.

La metodología y la estructura general se dividen en dos fases: i) carga, depuración, completado y análisis datos, y ii) generación e interpretación de modelos predictivos basados en machine learning. El mismo modelo puede utilizarse para la detección de anomalías comparando las predicciones con el comportamiento registrado.

Optimización geométrica de aliviaderos en tecla de piano para la mejora de su capacidad hidráulica

Fernando Salazar — Tue, 21 Jan 2020 09:04:02 +0100

En esta comunicación se expone el estudio realizado para la mejora de la capacidad de desagüe de aliviaderos en tecla de piano mediante la optimización de su diseño geométrico en base a modelación numérica validada físicamente. Se muestra para un cierto diseño base de aliviadero en tecla de piano, el patrón de flujo que induce la geometría y su influencia en la capacidad de desagüe. Se observó una reconducción del flujo por los voladizos del aliviadero, lo que provoca zonas de baja velocidad y concentración de líneas de corriente, dificultando el desagüe. Una vez estudiado el patrón flujo se procedió al rediseño de la geometría del aliviadero, en base a la detección en el modelo numérico de las zonas de baja velocidad inducidas por el contorno geométrico. La nueva geometría de aliviadero optimizada se modeló numéricamente, observándose un aumento de la
capacidad de desagüe respecto a la geometría base.

Algoritmo adaptativo para la estimación de pérdidas de agua en redes basado en análisis avanzado de Caudales Mínimos Nocturnos (CMN)

Fernando Salazar — Tue, 21 Jan 2020 08:50:03 +0100

El método del Caudal Mínimo Nocturno (CMN) es un concepto ampliamente utilizado para la estimación de fugas y otros objetivos estratégicos en empresas operadoras de agua. Si bien es un procedimiento de fácil aplicabilidad, existe un gran abanico de submétodos de aplicación, pudiendo introducirse una gran incertidumbre en los cálculos si no se seleccionan correctamente los criterios a utilizar. En este estudio se presenta un algoritmo capaz de asistir en la elección del mejor enfoque para el cálculo de CMN en función de los datos disponibles, las características de cada caso y el objetivo perseguido. Para elaborar dicho algoritmo se han ejecutado tres acciones: (i) revisión exhaustiva de literatura especializada y recopilación de submétodos utilizados en casos reales a nivel global, (ii) elaboración de un algoritmo multi-criterio capaz de seleccionar la mejor opción en cada fase del cálculo y (iii) presentación de una plataforma informática, llamada WatEner, en la que se integrará progresivamente el algoritmo propuesto.

Cracking of a concrete arch dam due to seasonal temperature variations

Fernando Salazar — Wed, 10 Oct 2018 17:29:02 +0200

A finite element thermomechanical analysis of a concrete arch dam has been carried out to predict the extent of cracking due to temperature variations. A linear elastic analysis allows us to find out the areas exceeding the tensile strength of the concrete. Then, a smeared crack approach based on a non-local damage model is used to accurately predict the pattern of the damage at the dam. Additionally, joint elements are introduced in the damaged regions of the wall to analyze the evolution of the crack aperture over the year.

Advanced computational methods for dam protections against overtopping

Fernando Salazar — Wed, 10 Oct 2018 09:35:03 +0200

The paper presents an overview on the computational tools developed in CIMNE for the analysis of the behavior of rockfill dams in overtopping scenarios, as well as for the design of protections systems against overtopping. In this latter case two different applications are presented: the design tool for a wedge shaped blocks (WSB) shoulder protection and the computational models for the analysis of the hydraulic performance of highly convergent spillways. Advanced finite element techniques are combined with particle techniques and artificial neural networks in the framework of the Kratos Multiphysics for reaching the optimal solution procedure for any of the above mentioned problems. An extensive validation is carried out using experimental data provided by the SERPA group of the Technical University of Madrid.

A new evolution on the wedge-shaped block for overtopping protection of embankment dams: the ACUÑA block

Fernando Salazar — Wed, 10 Oct 2018 09:38:01 +0200

The article summarizes the research studies performed by the Research Group on Dam Safety (SERPA) of the Technical University of Madrid and the International Center for Numerical Methods in Engineering (CIMNE) in collaboration with the company PREHORQUISA. Such studies aim to deepen on the theoretical and practical understanding of wedge shaped blocks (WSB) technology. This research was funded by the Spanish Ministry of Economy and Competitiveness through the research projects called ACUÑA (IPT-2011-0997-020000) and DIABLO (RTC-2014-2081-5). One of the projects goals was to develop a new model of WSB looking for improving the performance of the existing ones. This research led to the new model of WSB called ACUÑA, proprietary in Spain since May 2017 (ES2595852). The paper presents a comparison between the behaviour of the new block with one of the existing models, specifically Armorwedge^TM. Such comparison has been made using physical and numerical modelling, studying the hydrodynamic pressures on the block and the leakage flow through the joints between blocks and the aeration vents

Static and Seismic Analysis of the Janneh Arch-gravity Dam

Fernando Salazar — Thu, 31 May 2018 09:28:01 +0200

The static and seismic analysis of Janneh arch-gravity dam (157 m) is carried out by considering a combination of self-weight, hydrostatic, uplift and seismic loads. Linear and nonlinear analyses are performed for both static and seismic cases. Nonlinear behavior is studied by means of joint elements in the contact between the rock foundation and the dam following a bilinear cohesive law. The hydrodynamic effects derived to seismic loads are also considered. Hydrodynamic Westergaard’s approach is applied in the pseudo-static analysis, and the Westergaard’s generalized added mass is used for time-history case.

A systematic assessment of the influence of geometry and materials properties on the performance of arch dams

Fernando Salazar — Tue, 03 Apr 2018 09:41:01 +0200

Arch dams have different properties that play a relevant role in their behavior, although it is not clear to what degree or in what sense. There is some consensus regarding the relevance of certain factors such as length at crown, height, base and crest thickness, or Young modulus of dam and foundation. However, others such as the shape of arcs and cantilevers, which are correlated and whose effect is more difficult to consider, can also be influential. In this work, a systematic study of the response of arch dams in front of the common loading scenarios has been carried out, taking into account the usual range of variation of their properties. In total, 39 input variables related to geometry, material strength and thermal load were considered. Ranges of variation for each of these parameters have been defined according to the usual design criteria and 3,000 different geometries – together with the corresponding FEM models have been generated with random values of these parameters. The resulting displacements and stresses have been used to fit prediction models based on a machine learning technique named ‘random forests’ that give an estimate of the dam response. The interpretation of these models can be associated with the relative importance of the characteristics of arch dams on each of the behavior variables.

Computer-aided design and analysis of arch dams

Fernando Salazar — Tue, 03 Apr 2018 13:03:01 +0200

Double curvature arch dams feature geometrical complexity with a significant amount of parameters involved. Different criteria exist to assist in the design task, from simplified geometrical approaches to optimization procedures. However, most of them present a lack of flexibility and are not integrated in computer-aided design tools. In this contribution, an interactive and flexible software tool is presented to support the complete design process: geometrical definition, FEM model generation (including the mesh, the loads and the boundary conditions) and thermo-mechanical analysis. The design can be performed with different levels of detail to adapt to the information available in each stage of the project. The tool allows defining the shape of the reference cylinder, the excavation depth and slope along the foundation, the crown cantilever thickness and curvature, the shape and location of horizontal arcs; all these steps were described in former contributions. Here, special attention is paid to the introduction of additional features such as joints, spillways, abutments of varying shape and outlet works. All steps have been defined with a high degree of flexibility in the design process. The tool is integrated with the pre and post process software GiD, which allows taking advantage of its functionalities, such as mesh generation and results analysis. It is also coupled with a specific application for thermo-mechanical analysis of dams, developed in Kratos Multiphysics – a framework for building parallel multi-disciplinary simulation software. The whole design process can be followed in a unique environment, because the structural response of preliminary designs can be computed and the results considered to refine the dam geometry.