Time filter

Source Type

Arrasate / Mondragón, Spain

Gomez Gonzalez A.,University of Southern California | Rodriguez J.,University of Southern California | Sagartzazu X.,Ikerlan S. Coop. | Schuhmacher A.,Bruel and Kjaer Vibro | Isasa I.,Elevator Innovation Center
Proceedings of ISMA 2010 - International Conference on Noise and Vibration Engineering, including USD 2010 | Year: 2010

In this work we analyze the transmission paths of noise and vibration through the multiple coherence method. The non-stationariness of the signals in our applications motivates the introduction of the time variable. The analysis is performed using Fast Fourier transform-based techniques and auto-regressive modeling. These methods will be compared in terms of computational expenses and reliability. We illustrate their behavior computing the power spectral density and the coherence of synthetic and experimental stationary and nonstationary signals. Source

Sarrado C.,University of Girona | Turon A.,University of Girona | Renart J.,University of Girona | Urresti I.,Ikerlan S. Coop.
Composites Part A: Applied Science and Manufacturing | Year: 2012

The use of cohesive elements to simulate delamination growth involves modeling the inelastic region existing ahead of the crack tip. Recent numerical and experimental findings indicate that the mixed-mode ratio varies at each material point within the inelastic region ahead of the crack tip during crack propagation, even for those specimens whose mixed-mode ratio is expected to be constant. Although the local variation of the mode mixity may adversely affect the predicted numerical results, most existing formulations do not take it into account. In this work, the mode-decomposed J-integral is implemented as a finite element post-processing tool to obtain the strain energy release rates and the mixed-mode ratio of the inelastic region as a whole, allowing the assessment of crack propagation in terms of energy dissipation and mixed-mode ratio computation. Different cohesive elements are assessed with this method. © 2012 Elsevier Ltd. All rights reserved. Source

Salgado O.,Ikerlan S. Coop. | Altuzarra O.,University of the Basque Country | Viadero F.,University of Cantabria | Hernandez A.,University of the Basque Country
Engineering Computations (Swansea, Wales) | Year: 2010

Purpose - The purpose of this paper is to provide a general approach to compute, determine, and characterize the connectivity of the end-effector of a robotic manipulator of arbitrary architecture, in any of the postures that it can reach. Design/methodology/approach - The types of motion of this link, i.e. translational, screw motions, combinations thereof, and self-motions, are first defined and determined, simplifying the understanding of the instantaneous behaviour of the manipulator, aided by the definition of an alternative input basis. Findings - The characterization provided by this paper simplifies the understanding of the instantaneous behaviour of the manipulator. The mobility of the end-effector is completely characterized by the principal screws of its motion, which can be obtained from a generalized eigenproblem. In the process, alternative demonstrations of well-known properties of the principal screws are provided. Research limitations/implications - The approach presented is focused on the kinetostatic analysis of manipulators, and therefore, subjected to rigid body assumption. Practical implications - This paper proposes effective approaches for engineering analysis of robotic manipulators. Originality/value - This approach is based on a pure theoretical kinematic analysis that can characterize computationally the motion that the end-effector of an industrial robot of general morphology (i.e. serial, parallel, hybrid manipulators, complex mechanisms, redundant or nonredundantly actuated). Also, being implemented on a general-purpose software for the kinematic analysis of mechanisms, it provides visual information of the motion capabilities of the manipulator, highly valuable on its design stages. © Emerald Group Publishing Limited. Source

Agency: Cordis | Branch: FP7 | Program: CP | Phase: ICT-2007.3.4 | Award Amount: 2.79M | Year: 2008

The objective of the GENESYS project is to develop a cross-domain reference architecture for embedded systems that can be instantiated for different application domains to meet the requirements and constraints documented in the ARTEMIS strategic research agenda. These requirements are composability, networking, security, robustness, diagnosis, integrated resource management and evolvability. The reference architecture will address common issues, such as complexity management, separation of communication and computation, support for different levels of quality of service, security, model-based design, heterogeneity of subsystems, legacy integration, optimal power usage, and diagnosis. It will provide domain-independent services that can be customized to the needs of a particular application domain. Domain-specific platform services will be converged such that components from different application domains can be consistently integrated while preserving relevant properties. The project will result in a conceptualization of the cross-domain architecture, a specification of cross-domain core services and optional services for the selected application domains, and four exploratory prototypes that will demonstrate and help to evaluate the feasibility of selected central architectural concepts in the different application domains. The expected impact of GENESYS is a reduction of development costs and a speed-up of the time-to-market of embedded system development in different application domains, thereby strengthening the European presence in the competitive world-wide embedded systems markets. The support for cross-domain reuse of components is of particular relevance to innovative European SMEs that plan to enter those markets with high-tech products under increasing time pressure. The consortium consists of many major European embedded system suppliers and OEMs encompassing a broad range of application domains, supported by leading research and academic organizations.

Zugasti E.,Ikerlan S. Coop. | Gomez Gonzalez A.,University of Santiago de Compostela | Anduaga J.,Ikerlan S. Coop. | Arregui M.A.,Ikerlan S. Coop. | Martinez F.,Ikerlan S. Coop.
Smart Materials and Structures | Year: 2012

In this paper we present the application of two damage detection methods to a laboratory tower. The first method is based on subspace identification. The second one is based on AutoRegressive modeling of the signals involved. Both methods are tested in a tower demonstrator simulating a wind turbine. They are able to correctly detect damage in the structure that is simulated by loosening some of the bolts in the joints. The results show that the first method is computationally more efficient, but the results are more stable with the second method. © 2012 IOP Publishing Ltd. Source

Discover hidden collaborations