1558-1748 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JSEN.2018.2797086, IEEE Sensors Journal > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1 Abstract—Curvature sensors based on polymer optical fibers (POFs) present some advantages over the conventional technologies for joint angle assessment such as compactness, electromagnetic field immunity and multiplexing capabilities. However, the polymer is a viscoelastic material, which does not have a constant response with stress or strain. In order to understand and model this effect, this paper presents the dynamic characterization of a polymer optical fiber. The effects of temperature, frequency and loads on the fiber are analyzed for obtaining the influence of these parameters on the polymer dynamic Young modulus and time constant. Results show that a temperature on the range between 24°C and 45°C does not lead to considerable variations on the sensor output. Moreover, it is possible to estimate the storage modulus and loss factor from the frequency and temperature. The polymer time constant is defined on creep recovery experiments. Since the viscoelastic parameters are evaluated in different conditions of temperature, frequency and load, a model for the stress behavior of the fiber is proposed. Such model leads to a root mean squared error between the modelled and measured results more than 15 times lower than the one obtained with the model for bending stress without account the POF viscoelastic behavior. Index Terms—Polymer optical fiber, Curvature sensor, Viscoelasticity, Dynamic mechanical analysis. I. INTRODUCTION ONVENTIONAL technologies for human gait assessment include camera-based systems, inertial measurement units (IMUs), electrogoniometry, encoders, among others. Although camera-based systems analysis offer the most reliable measurements, hardware and software complexity contribute to its higher cost and poor portability [1]. Electrogoniometry employs potentiometers attached to exoskeletal linkages, which are bulky and intrusive [2]. Although flexible This research is financed by FCT through the fellowship SFRH/BPD/109458/2015, program UID/EEA/50008/2013 by the National Funds through the Fundação para a Ciência e a Tecnologia / Ministério da Educação e Ciência, and the European Regional Development Fund under the PT2020 Partnership Agreement. A. G. Leal-Junior, A. Frizera and M. J. Pontes acknowledge CAPES (88887.095626/2015-01) and FAPES (72982608). A. Frizera and M. J. Pontes acknowledge CNPq for the research productivity fellowships 304192/2016-3 and 310310/2015-6, respectively. A. G. Leal-Junior (arnaldo.leal@aluno.ufes.br), A. Frizera (frizera@ieee.org) and M. J. Pontes (mjpontes@ele.ufes.br) are with the Graduate Program of Electrical Engineering of Federal University of Espirito Santo, Vitória, Brazil C. Marques is with Instituto de Telecomunicações and Physics Department & I3N, University of Aveiro, Portugal (carlos.marques@ua.pt) goniometers adapt better to body parts and are not sensitive to misalignments due to movement of polycentric joints, they have to be carefully attached to the skin due to its sensitivity and it is a costly technology [3]. Furthermore, encoders need mechanical supports precisely assembled due its sensibility to misalignments and result in a less compact system. IMUs overcome most of the disadvantages associated with previously mentioned instruments. However, they are sensitive to electromagnetic interferences and need frequent calibration [4]. Facing the limitations of conventional technologies, optical fiber sensors are employed on the measurement of human kinematics because they are compact, lightweight, allow multiplexing systems, and present immunity to electromagnetic interference [5]. Polymer optical fibers (POFs) have more resistance to impact loads and vibrations, and higher strain limits than silica optical fibers [6]. These advantages make POFs preferable for direct measurement of angles due to the possibility of applying larger strains [6]. Furthermore, it can also be applied for antibodies [7-8] and glucose level detection [9-10], accelerometers [11], humidity sensors [12]. There are different types of POFs, which can be microstrutured [13], small-core step index [14] and large core diameter [5]. Nevertheless, POFs with higher core diameter enable the application of low precision connectors, which generally results on low cost systems [15] and can be applied on sensors based on bending loss [16] or fluorescence-based biosensors [9-10]. Furthermore, there are some different materials for POFs as well, which can be made of polycarbonate that can resist higher temperature and stress [17], TOPAS [18] and Zeonex [19] that are insensitive to humidity variations [20-21]. However, polymethyl methacrylate (PMMA) is the most employed material for POFs fabrication and is widely available commercially [17]. For this reason, the PMMA POF is employed in this work. The operation principle of some curvature sensors based on POF is the variation of the output power when the fiber is under curvature. In order to enhance the sensor sensitivity, a lateral section is made on the fiber to create a region known as the sensitive zone [22]. If the fiber with the sensitive zone on the convex side is under curvature, there will be a variation of the reflections in the convex and concave sides, which leads to output signal attenuation in concave bending and an increase of the signal in convex bending [22]. There is also the variation of the transmission mode with the coupling between higher and lower modes, which increases the surface Dynamic Mechanical Analysis on a PolyMethyl Methacrylate (PMMA) Polymer Optical Fiber Arnaldo G. Leal-Junior, Carlos Marques, Anselmo Frizera, Member, IEEE, and Maria José Pontes C