Search

International audience; Structural health monitoring (SHM) of civil structures often is limited due to changing environmental conditions, as those changes affect the structural dynamical properties in a similar way like damages can do. In this article, an approach for damage detection under changing temperatures is presented and applied to a beam structure. The used stochastic subspace-based algorithm relies on a reference null space estimate, which is confronted to data from the testing state in a residual function. For damage detection the residual is evaluated by means of statistical hypothesis tests. Changes of the system due to temperature effects are handled with a model interpolation approach from linear parameter varying system theory. From vibration data measured in the undamaged state at some few reference temperatures, a model of the dynamic system valid for the current testing temperature is interpolated. The reference null space and the covariance matrix for the hypothesis test is computed from this interpolated model. This approach has been developed recently and was validated in an academic test case on simulations of a mass-spring-damper. In this paper, the approach is validated experimentally on a beam structure under varying temperature conditions in a climate chamber. Compared to other approaches, the interpolation approach leads to significantly less false positive alarms in the reference state when the structure is exposed to different temperatures, while faults can still be detected reliably.

International audience; In the context of detecting changes in structural systems, multiple vibration-based damage detection methods have been proposed and successfully applied to both mechanical and civil structures over the past years. One of the popular schemes is based on a robust subspace-based residual and enjoys favorable statistical and computational properties, like invariance to changes in the excitation covariance and numerical stability. This paper presents an alternative Gaussian residual that is based on the difference of normalized Hankel matrices between reference and damaged states, which can be easily computed. The statistical properties of the residual are reported and used for efficient hypothesis testing. Its robustness to excitation changes is shown. The proposed scheme is evaluated in numerical simulations, validating its robustness, and tested on real data sets from a full scale bridge.

International audience; Temperature variation can be a nuisance that perturbs vibration based structural health monitoring (SHM) approaches for civil engineering structures. In this paper, temperature affected vibration data is evaluated within a stochastic damage detection framework, which relies on a null space based residual. Besides two existing temperature rejection approaches-building a reference state from an averaging method or a piecewise method-a new approach is proposed, using model interpolation. In this approach, a general reference model is obtained from data in the reference state at several known reference temperatures. Then, for a particular tested temperature, a local reference model is derived from the general reference model. Thus, a well fitting reference null space for the formulation of a residual is available when new data is tested for damage detection at an arbitrary temperature. Particular attention is paid to the computation of the residual covariance, taking into account the uncertainty related to the null space matrix estimate. This improves the test performance, contrary to prior methods, for local and global damages, resulting in a higher probability of detection (PoD) for the new interpolation approach compared to previous approaches.

International audience; Detecting small and local damages on structures based on ambient vibrations is a major challenge in structural health monitoring. However, being able to identify the minimum damage is essential for quantifying the effectiveness of the instrumentation and for defining the limitations of low-frequency vibration monitoring in general. This paper shows how subspace-based methods could be used by engineers to predict the minimum damage that can be detected. The method employs a Gaussian subspace-based residual vector as a damage-sensitive criterion and evaluates its deviation from zero mean through two different statistical hypothesis tests, a parametric version and a non-parametric one. A sensitivity analysis is carried out to parametrize the deviation from the nominal state, and link it to physical parameters in a finite element model through the Fisher information matrix. This link can also be used to predict the minimum detectable damage, e.g. by prescribing a minimum probability of detection based on code-based reliability concepts. Ultimately, the developed theory is verified by means of a numerical example.

MICROSCOPIE+CARE+IKL:TCV:ABO:LBU:FGA:LRE:ACV:PVE; International audience; Catalytic combustion as a process used for removal of hydrocarbons from automotive gas exhausts or for energy production has been widely implemented on supported PGM (Platinum Group Metals) based catalysts1. Since PGMs are very costly and rare, there is a strong need for an equally effective and less expensive catalyst. The electrochemical promotion of catalysis (EPOC), is a promising concept to in-operando boost catalytic processes in a reversible and controlled manner2. The aim of this study was to develop Ag-based electrochemical catalysts for low temperature propylene deep oxidation. EPOC of propylene combustion has been carried out in the literature but mainly on Pt catalytic films1,3, while this phenomenon was attributed to the modification in the propylene chemisorption3.Nanostructured electrochemical catalysts were prepared by screen-printing and reactive Physical Vapor Deposition (PVD) method. Thickness and porosity of Ag coatings were modified by changing the deposition parameters (duration and pressure for PVD, nature of the ink and calcination temperature for screen-printing) to optimize the catalytic properties. Catalytic and electrocatalytic tests have been carried out in a quartz reactor1 which operated under continuous flowing conditions at atmospheric pressure. The catalytic activity was monitored in a temperature range of 100 to 400oC under lean-burn conditions, as encountered in Diesel exhausts. The most active Ag films were also evaluated under closed circuit conditions (± 2V) in order to measure the effect of polarization between the Ag working electrode and an Au reference electrode. Both electrodes were exposed to the same atmosphere in a single chamber configuration.Values of Faradaic efficiencies in the range of 300 were obtained while the conversion could be tailored from 14 to 21% (Fig. 1). Negative current applications lead to the decrease of the CO2 production while positive current application corresponds to a pronounced increase of the catalytic performance. Upon positive current applications, the rate enhancement ratio increases with the intensity of the current. This indicates that the coverage of promoting ionic species (Oδ-) increases with the current, then producing more weakly bonded oxygen species coming from the gas phase4.This study reports, for the first time, that the catalytic activity for propylene of Ag coatings deposited onto YSZ can be tailored by current applications in a non-Faradaic manner. The predominant impact of current applications is to modify the reactivity of oxygen present on the Ag surface. Positive current applications increase the propylene conversion by producing more reactive oxygen species. This beneficial effect is more pronounced in an oxidizing atmosphere, where the oxygen coverage on Ag is high. This demonstrates that EPOC can enhance catalytic properties of Ag coatings for the abatement of propylene in air.References 1.P. Vernoux et al., J. Catal., 208 (2002) 412-421.2.C.G. Vayenas, Electrochemical Activation of Catalysis: Promotion, Electrochemical Promotion, and Metal-Support Interactions, Springer, 2001.3.A. Kaloyannis et al., J. Catal., 182 (1999) 37-47.4.I. Kalaitzidou et al., Materials Today: Proceedings, 5, 27345 (2018).Acknowledgments: This study was performed in the “EPOX” project, funded by the French National Research Agency (ANR), ANR-2015-CE07-0026.

MICROSCOPIE+CARE+TCV:IKL:ABO:LBU:FGA:LRE:ACV:PVE; National audience; Des tests catalytiques et electrocatalytiques ont été menés sur des membranes denses conductrices ionique pour la combustion du propène en utilisant une couche catalytique d'argent synthétysée via PVD ou sérigraphie déposée sur une membrane dense de Zircone stabilisée avec 8% d'yttrium (YSZ), un conducteur ionique de O2-. Une adsorption compétitive entre le propène et l'oxygène a été observée à 300°C. Les performances catalyques des films d'argents ont pu être modifiées via une polarisation et ce, de manière non-faradique (effet EPOC) à 300°C. L'impact de la polarisation depend principalement du taux de recouvrement du propène sur l'argent : pour des taux de recouvrement faible comme dans des conditions d'oxydation, des effets de promotions ont été observés lors de polarisation positive alors que pour un fort taux de recouvrement, obtenu dans des conditions stoechiomètrique, seulement des effets d'inhibition ont pu être observé lors de polarisations négatives.

ISMA 2018, 28th Conference on Noise and Vibration Engineering, Leuven, BELGIQUE, 17-/09/2018 - 19/09/2018; International audience; Modal validation is an integral part of any modal analysis scheme. Recent advances in efficient variance computation of modal parameter estimates from the output-only subspace-based identification algorithms made the computation of variance feasible for the modal parameters, such as natural frequencies, damping ratios and mode shapes. One of the other practical modal indicators is Modal Assurance Criterion (MAC), for which uncertainty computation scheme is missing. This paper builds on the previous results using the propagation of the measurement uncertainties to estimates of MAC. The sensitivity of the MAC with respect to output covariances is derived using a first order perturbations and the uncertainties are propagated using the Delta method. The influence of the underlying mode shape scaling on both the uncertainty of mode shapes and MAC is investigated. The results are presented in context of operational modal analysis (OMA) of a spring mass system.

International audience; This study is intended to the evaluation and improvement of instrumentation solutions for long term monitoring of new generation of transport infrastructures. A test site has been instrumented with thermocouples and an infrared thermographic system coupled with the monitoring of environmental parameters. A spatial reconstruction method is proposed. Measured data acquired on site and post-processed are analyzed through time. Finally, obtained results are commented and perspectives are proposed.

SSCI-VIDE+CARE+JGR:SGI:LRE:AGF; National audience; Preparación de catalizadores Diesel NOx Trap por via coloidalJ. M. García-Vargasa,*, S. Gila, L. Retailleaua, L. Veyreb, C. Thieuleuxb, A. Giroir-Fendleraa Université Claude Bernard Lyon 1, CNRS, IRCELYON, 2 avenue Albert Einstein, Villeurbanne, F-69622, France.b , ICL, C2P2 UMR 5265, CPE Lyon, 43 Bd du 11 Novembre 1918, F-69616 Villeurbanne, France.*Autor principal: jesus.vargas@ircelyon.univ-lyon1.fr1. IntroducciónHoy en día, la legislación relativa a la emisión de contaminantes por parte de vehículos a motor ha llevado al empleo de diferentes sistemas catalíticos para tratar los gases que emiten, consiguiendo así cumplir con los requerimientos en cuanto a niveles máximos de emisión. Dichos límites han disminuido de forma continuada en los últimos años, convirtiendo la reducción de las emisiones de óxidos de nitrógeno (NOx) por parte de los vehículos diésel en uno de los principales desafíos tecnológicos. Se han planteado diferentes tecnologías para resolver esta problemática, siendo una de estas el uso de catalizadores Diesel NOx Trap (DNT) que permiten almacenar NOx en forma de nitratos durante el funcionamiento normal del motor y posteriormente reducirlos a N2 mediante la inyección durante cortos periodo de tiempo de un agente reductor. Estos catalizadores han mostrado buen rendimiento pero baja durabilidad1. En el presente trabajo se muestra la aplicación de un nuevo método de preparación de catalizadores DNT Pt/Pd/Rh por vía coloidal, procedimiento que se ha utilizado con éxito en la obtención de partículas metálicas muy dispersas, con tamaños de entre 1 y 2 nm, al preparar catalizadores de Ni2 y Pt3.2. ExperimentalDos catalizadores Pt/Pd/Rh fueron preparados con diferente carga de metal mediante la impregnación de coloides metal-Si, mezclando complejos organometálicos de Pt, Pd y Rh en THF con octilsilano, que fue utilizado como agente estabilizante, en atmósfera de H2. Un oxido mixto BaO/CeO2/Al2O3-La2O3 fue seleccionado como soporte. Más detalles acerca del procedimiento utilizado pueden ser encontrados en la literatura2. Los experimentos catalíticos fueron realizados con 10 mg de catalizador usando SiC como inerte para conseguir un GHSV = 52.000 h-1, con un caudal total de 10 l h-1. La composición del alimento fue 14% O2, 4% H2O, 1500 ppm CO, 430 ppm C3H6, 100 ppm NO, usando He como balance. Los catalizadores preparados fueron caracterizados mediante análisis ICP, BET, DRX, TPR, TEM y TPD-NO2. Así mismo fue analizada la actividad de los catalizadores después de envejecimiento, tratando los materiales a 1050°C durante 10 h en atmosfera de N2 con un 10% de H2O. Además de los catalizadores preparados, se testó en las mismas condiciones un catalizador comercial usado como referencia.Tabla 1. Composición de los catalizadores y resultados TPD-NO2.Catalizador %Ce %Al %La %Ba %Mg %Pt %Pd %Rh NOx desorbido (mol/g)Pt/Pd/Rh 4/2/129.1128.511.641.08245 ppm0.490.200.15306 NO2/54 NOPt/Pd/Rh 14/3/129.1128.511.641.08245 ppm0.36719 ppm358 ppm354 NO2/35 NOReferencia com.40.8317.841.130.860.760.650.300.16600 NO2/204 NOCatalysis for a moresustainable worldOviedo, June 26 – 28, 2017 O/P 3. Resultados y discusiónLa composición de los catalizadores preparados y del catalizador de referencia aparece en la Tabla 1. Los análisis DRX no mostraron la presencia de picos de difracción correspondientes a los metales nobles, lo que indica una buena dispersión que fue confirmada mediante análisis TEM, con tamaños de partícula en torno a 1 nm. Experimentos TPD de NO2 fueron realizados, mostrando la Tabla 1 la cantidad de NOx adsorbidos por los catalizadores frescos y el de referencia, siendo la capacidad de adsorción de este catalizador mayor. Comparando los dos catalizadores preparados, el catalizador con una mayor proporción de Pt (Pt/Pd/Rh 14/3/1) fue capaz de adsorber más NO2, apareciendo los picos de desorción a temperaturas mayores, lo que implica una mayor fortaleza de dicha adsorción. La actividad catalítica fue evaluada en términos de conversión y velocidad de reacción. El catalizador preparado con la menor proporción de Pt (Pt/Pd/Rh 4/2/1) fue el más activo tanto en la combustión de C3H6 como en la de CO, mientras que el otro catalizador preparado mostró una actividad similar a la del catalizador de referencia. En cuanto a la conversión de NO (Figura 1), el catalizador que presentó mayor conversión fue el preparado con mayor proporción de Pt, siendo los dos catalizadores preparados más activos que el de referencia. La resistencia al envejecimiento fue también evaluada en términos de actividad catalítica, siendo el catalizador Pt/Pd/Rh 14/3/1 el menos afectado, con valores de conversión de C3H6 y CO cercanos a los del catalizador fresco.Figura 1. Conversión de NO.4. ConclusionesLa proporción de metales nobles ha mostrado una clara influencia en la actividad y propiedades de catalizadores Diesel NOx Trap preparados por vía coloidal, método eficaz a la hora de obtener catalizadores con gran dispersión de metal. El catalizador preparado con una menor proporción de Pt mostro la mayor actividad en la combustión de CO y C3H6, incluso al compararlo con un catalizador comercial. Sin embargo, la presencia de una mayor proporción de Pt favorece la resistencia a la desactivación hidrotermal.Referencias1. E. Rohart, V. Bellière-Baca, K. Yokota, V. Harlé, C. Pitois, Top. Catal. 42 (2007) 71-75.2. D. Baudouin, K.C. Szeto, P. Laurent, … , C. Thieuleux, J. Am. Chem. Soc. 134 (2012) 20624-20627.3. P. Laurent, D. Baudouin, B. Fenet, L. Veyre, S. Donet, C. Coperet, C. Thieuleux, New J. Chem. 38 (2014) 5952-5956

International audience; For Operational Modal Analysis (OMA), the vibration response of a structure from ambient and unknown ex-citation is measured and used to estimate the modal parameters. For OMA with eXogenous inputs (OMAX), some of the inputs are known in addition, which are considered as realizations of a stochastic process. When identifying the modal parameters from noisy measurement data, the information on their uncertainty is most relevant. Previously, a method for variance estimation has been developed for the output-only case with covariance-driven subspace identification. In this paper, a recent extension of this method for the in-put/output covariance-driven subspace algorithm is discussed. The resulting variance expressions are easy to evaluate and computationally tractable when using an efficient implementation. Based on Monte Carlo simulations, the quality of identification and the accuracy of variance estimation are evaluated. It is shown how the input information leads to better identification results and lower uncertainties.

SSCI-VIDE+CARE+JGR:SGI:LRE:AGF; International audience; 1. IntroductionPresently, catalysts are widely employed in the abatement of pollutants from automobiles, being the three-way catalyst (TWC) the most common system used in the gasoline powered vehicles, due to its capacity to achieve the elimination of CO, HC and NO simultaneously. Pt/Rh containing catalysts dominated the automobile exhaust converter market for many years. However, Pd is more abundant and cheap, and has therefore drawn much attention. In addition, the Pd catalysts developed demonstrated significantly improved light-off and NOx conversion activities compared to the commercial Pt/Rh TWCs [1]. Generally, the particle size has a great effect in the catalytic activity of Pd in the TWC, showing smaller particles a higher activity, especially for the oxidation of CO and HC; probably due to a higher presence of Pd (I) species [2]. In this work, we apply a novel colloidal procedure in the catalysts preparation of a Pd/Rh TWC, procedure that has proven to be successful in order to obtain very small metal particles (1-2 nm) when preparing Ni [3] or Pt [4] catalysts.2. ExperimentalPd, Rh and four Pd/Rh catalysts with different impregnation order were prepared by impregnation of metal-Si colloids, mixing Pd and/or Rh organometallic complexes in THF with octylsilane as a stabilizing agent, in the presence of H2; using a standard TWC washcoat. More details about the procedure followed can be found elsewhere [3]. After impregnation, the catalyst was stabilized at 650 °C for 2 h under a flow of N2 with 10% of H2O. Light off tests were carried out using 40 mg of catalyst and SiC as inert in order to reach a GHSV = 52.000h-1 with a total flow of 10 l h-1. The feed gas composition was 4% CO2, 3% H2O, 0.86 % O2, 3000 ppm C3H6, 2300 ppm H2, 1500 ppm NO, 700 ppm CO, using He as balance. 3. Results and discussionMetal particle size was analyzed using the TEM technique, showing all the catalysts metal particles sizes around 1 nm both for Pd and Rh. No differences in particle size were observed depending in the impregnation order or comparing monometallic with bimetallic catalysts. The actual metal content was analyzed by ICP, Table 1.Table 1: Main characterization and catalytic results.Catalyst%Pd%RhT50 (°C) C3H6-CO2T50 (°C) CO-CO2T50 (°C) NO-N2Pd/Washcoat0.51 - 313184-454304Rh/Washcoat- 0.06 297228-412313Pd/Rh/Washcoat0.70 0.05 296173-434299Rh/Pd/Washcoat0.76 0.06 304177-451299Pd-Rh/Washcoat0.67 0.09 301177-447306Pd/Washcoat+Rh/Washcoat0.27 0.05 310183-457314Great differences were observed depending on the metal, as can be observed in Figure 1. At 100 ºC a decrease of the concentration of H2 and NO for the Pd catalyst was obtained, probably due to the concurrence of the H2 SCR reaction, while in the case of the Rh catalyst no catalytic activity was observed until 200 °C. This different behaviour is probably due to the higher H2 adsorption capacity of Pd compared to Rh, what increases the reactivity of this molecule. At about 150 °C it can be observed a decrease in the O2 and CO concentration also for the Pd catalyst, while the NO concentration rises to its initial value. This is probably due to the beginning of the oxidation of CO towards CO2 and also to the oxidation of H2 towards H2O, being this reaction competitive with the H2 SCR. In the case of the Rh catalyst, the oxidation of both CO and H2 did not start until reaching a temperature of 200-220 °C, showing therefore a lower activity in these oxidative reactions. In addition, two different peaks were obtained in the CO conversion for the Pd catalyst, while only one peak with a slight shoulder was observed for the Rh one, what yields a bigger range of temperature in which there are conversion of CO for the first catalyst. This is due to the higher activity of Pd in the CO oxidation at low temperature, and also to the higher activity of Rh in the reforming of C3H6 at medium temperatures (350-450 °C), what yields CO and H2 as products, as can be observed in the increase of the concentration of these compounds at such temperatures.Figure 1: Concentration (dark colors) and Conversion (light colors) vs Temp.: a) Pd/Washcoat b) Rh/Washcoat.Moreover, obtained results showed that the impregnation order had an important influence on the behaviour of the catalysts, showing the catalyst where Rh was impregnated in the last place the conversion and concentration profiles more similar to the Rh catalyst. In Table 1 it can be observed the temperature needed to reach 50% of conversion in the C3H6-CO2, CO-CO2 and NO-N2 reactions for each one of the catalysts prepared. Pd/Rh and Rh/Pd showed the better catalytic behavior, needing lower temperatures to reach the 50% of conversion of NO towards N2 and also the highest temperature range to reach at least a 50% of conversion of CO. In addition, Pd/Rh was the catalyst with the lowest T50 in the conversion of C3H6 to CO2.4. ConclusionsThis new colloidal procedure has shown to be useful in order to obtain Pd and Rh particles of size around 1 nm in a TWC catalyst. A higher activity at low temperature in the H2 SCR and CO oxidation was observed for the Pd/Washcoat catalyst compared with the Rh/Washcoat catalyst. Catalyst prepared impregnating in the first place Rh and after that Pd showed the higher catalytic activity in the three main reactions in which are involved the TWC.

International audience; Since the past decade, infrared thermography coupled with inverse models based on 1d thermal quadrupoles have shown their usefulness in civil engineering by first showing their ability to assess the quantitative non destructive testing of concrete repaired by bonded CFRP plate over a wide area (i.e. repaired or reinforced concrete beams). On the other hand early implementations of long terms monitoring methods based on such approach have given their first results over a whole bridge deck. The experimental method, allow us to have the apparent surface temperature field evolution with time for a wide area divided in pixels. Knowing this specificity, the procedure aims to apply an independent model to each pixel in order to retrieve physical properties map. Such treatment can have a high computational cost. In this paper we propose various improvement of our procedure based on GPGPU paradigm in order to shorten the computational time. This study will detail an experimental procedure able to assess the long term thermal monitoring of a bridge deck over days and to draw properties maps of the inner structure.

International audience; This paper presents a method based on the use of Bayesian modal parameter recursive estimation basedon a particular Kalman filter algorithm with decoupled distributions for mass and stiffness. ParticularKalman filtering is a combination of two widely used Bayesian estimation methods working together:the particle filter (also called sequential Monte Carlo samplings) and the Kalman filter. Usual systemidentification techniques for civil and mechanical structures assume the availability of large set of dataderived from a stationary quasi steady structure. On the opposite, several scenarios involve time varyingstructures. For example, due to interaction with aerodynamics in aeronautics, some critical parametermay have to be monitored, for instability monitoring (leading possibly to flutter) of in flight data due tofuel consumption and speed change. This relates to the monitoring of time varying structural parameterssuch as frequencies and damping ratios. The main idea of a particular Kalman filter is to considerstochastic particles evolving in the parameter space. For each particle, a corresponding linear state isrecursively estimated by applying a Kalman filter to the mechanical system, whose modal parametersare driven by the evolution of this time-varying particle. The weight of each particle is computed fromthe likelihood of the parameter sample it represents and its corresponding state. This result in a bank ofadaptive coupled Kalman filters combined with their particle filter. However, the system parametrizationis relatively large. In order to provide fast and convincing results for large time varying structure, suchas an airplane, the execution time of the method has to be improved. In particular, the particle evolutionscan be run in parallel, Within the Cloud2sm project, A Quadro k6000 card of 3072 cores clocked to 3GB/s has been used. This paper will show a GPGPU implementation of the particular Kalman filter andthe first results will be discussed.

International audience; The Stochastic Dynamic Damage Locating Vector (SDDLV) method is an output-only damage localiza-tion method based on both a Finite Element (FE) model of the structure and modal parameters estimated from output-only measurements in the damage and reference states of the system. A vector is obtained in the null space of the changes in the transfer matrix computed in both states and then applied as a load vector to the model. The damage localization is related to this stress where it is close to zero. In previous works an important theoretical limitation was that the number of modes used in the computation of the transfer function could not be higher than the number of sensors located on the structure. It would be nonetheless desirable not to discard information from the identification procedure. In this paper, the SDDLV method has been extended with a joint statistical approach for multiple mode sets, overcoming this restriction on the number of modes. The new approach is validated in a numerical application, where the outcomes for multiple mode sets are compared with a single mode set. From these results, it can be seen that the success rate of finding the correct damage localization is increased when using multiple mode sets instead of a single mode set.

SSCI-VIDE+CARE+PAN:JGR:LRE:SGI:AGF; International audience; None

SSCI-VIDE+CARE+PAN:JGR:LRE:SGI:AGF; National audience; None

SSCI-VIDE+CARE+ZSI:SGI:LRE:CDS:AGF; International audience; None

International audience; Operational modal analysis is an important step in many methods for vibration-based structural health monitoring. These methods provide the modal parameters (frequencies, damping ratios and mode shapes) of the structure and can be used for monitoring over time. For a continuous monitoring the excitation of a structure is usually ambient, thus unknown and assumed to be noise. Hence, all estimates from the vibration measurements are realizations of random variables with inherent uncertainty due to unknown excitation, measurement noise and finite data length. Estimating the standard deviation of the modal parameters on the same dataset offers significant information on the accuracy and reliability of the modal parameter estimates. However, computational and memory usage of such algorithms are heavy even on standard PC systems in Matlab, where reasonable computational power is provided. In this paper, we examine an implementation of the covariance-driven stochastic subspace identification on the wireless sensor platform PEGASE, where computational power and memory are limited. Special care is taken for computational efficiency and low memory usage for an on-board implementation, where all numerical operations are optimized. The approach is validated from an engineering point of view in all its steps, using simulations and field data from a highway road sign structure.

SSCI-VIDE+CARE+SGI:JGR:LRE:AGF; International audience; A series of Fe-loaded zeolites catalysts were prepared by impregnation and in-situ hydrothermal synthesis, and the influence of the preparation method and the support used (ZSM-5, Beta and LTL) on the SCR-NH3 activity was studied. These catalysts after aged were characterized by N2 adsorption, ICP, RMN, XRD, XPS and TEM analysis. Iron species on the surface were assigned to the Fe3+ ions and the small amount of Fe2+ in the case of the hydrothermal catalysts, whereas the impregnated catalysts presented the formation of Fe3O4 oxides, which resulted in difference in SCR activity. The hydrothermal catalysts presented higher activity, N2 selectivity and lower N2O formation than the impregnated one. This could be a consequence of the partial pores blockage of the zeolite and the Fe3O4 oxides formation, when the impregnated method was used. Moreover, the framework structure and/or pore geometry of the used support was crucial for determining the SCR activity.