Estudio de la redistribución tensional en la interfase frp-hormigón

El uso de materiales compuestos para el refuerzo, reparación y rehabilitación de estructuras de hormigón se ha convertido en una técnica muy utilizada en la última década. Con independencia de la técnica del refuerzo, uno de los principales condicionantes del diseño es el fallo de la adherencia entre el hormigón y el material compuesto, atribuida generalmente a las tensiones en la interfaz de estos materiales. Las propiedades mecánicas del hormigón y de los materiales compuestos son muy distintas. Los materiales compuestos comúnmente utilizados en ingeniería civil poseen alta resistencia a tracción y tienen un comportamiento elástico y lineal hasta la rotura, lo cual, en contraste con el ampliamente conocido comportamiento del hormigón, genera una clara incompatibilidad para soportar esfuerzos de forma conjunta. Esta incompatibilidad conduce a fallos relacionados con el despegue del material compuesto del sustrato de hormigón. En vigas de hormigón reforzadas a flexión o a cortante, el despegue del material compuesto es un fenómeno que frecuentemente condiciona la capacidad portante del elemento. Existen dos zonas potenciales de iniciación del despegue: los extremos y la zona entre fisuras de flexión o de flexión-cortante. En el primer caso, la experiencia a través de los últimos años ha demostrado que se puede evitar prolongando el refuerzo hasta los apoyos o mediante el empleo de algún sistema de anclaje. Sin embargo, las recomendaciones para evitar el segundo caso de despegue aún se encuentran lejos de poder prever el fallo de forma eficiente. La necesidad de medir la adherencia experimentalmente de materiales FRP adheridos al hormigón ha dado lugar a desarrollar diversos métodos por la comunidad de investigadores. De estas campañas experimentales surgieron modelos para el pronóstico de la resistencia de adherencia, longitud efectiva y relación tensión-deslizamiento. En la presente tesis se propone un ensayo de beam-test, similar al utilizado para medir la adherencia de barras de acero, para determinar las características de adherencia del FRP al variar la resistencia del hormigón y el espesor del adhesivo. A la vista de los resultados, se considera que este ensayo puede ser utilizado para investigar diferentes tipos de adhesivos y otros métodos de aplicación, dado que representa con mayor realidad el comportamiento en vigas reforzadas. Los resultados experimentales se trasladan a la comprobación del fallo por despegue en la región de fisuras de flexión o flexión cortante en vigas de hormigón presentando buena concordancia. Los resultados condujeron a la propuesta de que la limitación de la deformación constituye una alternativa simple y eficiente para prever el citado modo de fallo. Con base en las vigas analizadas, se propone una nueva expresión para el cálculo de la limitación de la deformación del laminado y se lleva a cabo una comparación entre los modelos existentes mediante un análisis estadístico para evaluar su precisión. Abstract The use of composite materials for strengthening, repairing or rehabilitating concrete structures has become more and more popular in the last ten years. Irrespective of the type of strengthening used, design is conditioned, among others, by concrete-composite bond failure, normally attributed to stresses at the interface between these two materials. The mechanical properties of concrete and composite materials are very different. Composite materials commonly used in civil engineering possess high tensile strength (both static and long term) and they are linear elastic to failure, which, in contrast to the widely known behavior of concrete, there is a clear incompatibility which leads to bond-related failures. Bond failure in the composite material in bending- or shear-strengthened beams often controls bearing capacity of the strengthened member. Debonding failure of RC beams strengthened in bending by externally-bonded composite laminates takes place either, at the end (plate end debonding) or at flexure or flexure-shear cracks (intermediate crack debonding). In the first case, the experience over the past years has shown that this can be avoided by extending laminates up to the supports or by using an anchoring system. However, recommendations for the second case are still considered far from predicting failure efficiently. The need to experimentally measure FRP bonding to concrete has induced the scientific community to develop test methods for that purpose. Experimental campaigns, in turn, have given rise to models for predicting bond strength, effective length and the stress-slip relationship. The beam-type test proposed and used in this thesis to determine the bonding characteristics of FRP at varying concrete strengths and adhesive thicknesses was similar to the test used for measuring steel reinforcement to concrete bonding conditions. In light of the findings, this test was deemed to be usable to study different types of adhesives and application methods, since it reflects the behavior of FRP in strengthened beams more accurately than the procedures presently in place. Experimental results are transferred to the verification of peeling-off at flexure or flexure-shear cracks, presenting a good general agreement. Findings led to the conclusion that the strain limitation of laminate produces accurate predictions of intermediate crack debonding. A new model for strain limitation is proposed. Finally, a comprehensive evaluation based on a statistical analysis among existing models is carried out in order to assess their accuracy.

Effect of adhesive thickness and concrete strength on FRP-Concrete Bonds

The use of fiber-reinforced polymer (FRP) composites for strengthening, repairing, or rehabilitating concrete structures has become more and more popular in the last 10 years. Irrespective of the type of strengthening used, design is conditioned, among others, by concrete-composite bond failure, normally attributed to stress at the interface between these two materials. Single shear, double shear, and notched beam tests are the bond tests most commonly used by the scientific community to estimate bond strength, effective length, and the bond stress-slip relationship. The present paper discusses the effect of concrete strength and adhesive thickness on the results of beam tests, which reproduce debonding conditions around bending cracks much more accurately. The bond stress-slip relationship was analyzed in a cross section near the inner edge, where stress was observed to concentrate. The ultimate load and the bond stress-slip relationship were visibly affected by concrete strength. Adhesive thickness, in turn, was found to have no significant impact on low-strength concrete but a somewhat greater effect on higher strength materials.

Estudio de la adherencia de armaduras corroídas y su influencia en la capacidad resistente de elementos de hormigón armado

La verificación de la seguridad estructural, tanto de estructuras que permitan un cierto grado de deterioro en su dimensionado como de estructuras existentes deterioradas, necesita disponer de modelos de resistencia que tengan en cuenta los efectos del deterioro. En el caso de la corrosión de las armaduras en las estructuras de hormigón armado, la resistencia depende de múltiples factores tales como la sección del acero corroído, el diagrama tensión-deformación del acero corroído, la adherencia hormigón-acero corroído, la fisuración o desprendimiento del hormigón debido a la expansión de los productos de corrosión. En este sentido, la transferencia de las fuerzas a través de la superficie de contacto entre el hormigón y el acero, la adherencia, es uno de los aspectos más importantes a considerar y es la base del comportamiento del hormigón armado como elemento estructural. La adherencia debe asegurar el anclaje de las armaduras y transmitir la tensión tangencial que aparece en las mismas como consecuencia de la variación de las solicitaciones a lo largo de un elemento estructural. Como consecuencia de la corrosión de las armaduras, el desarrollo de la adherencia se altera y, por tanto, la transferencia de la tensión longitudinal. Esta Tesis Doctoral aborda el comportamiento en estado límite último de la adherencia en el hormigón estructural con armaduras corroídas. El objetivo principal es la obtención de un modelo suficientemente realista y fiable para la evaluación de la adherencia con armaduras corroídas en el marco de la verificación de la seguridad estructural de elementos de hormigón armado con armaduras corroídas. Para ello se ha llevado a cabo un programa experimental de ensayos tipo pull-out excéntricos, con diferentes probetas, unas sin corrosión y otras sometidas tanto a procesos de corrosión natural como a procesos de corrosión acelerada, con diferentes grados de deterioro. Este tipo de ensayo de adherencia representa de forma realista y fiable realista los esfuerzos de adherencia en la zona de anclaje. Por otra parte, para la realización de estos ensayos se ha puesto a punto, además del procedimiento de ensayo, un sistema de adquisición de datos entre los que se incluye el empleo de sensores de tipo fibra óptica con redes de Bragg embebidos en la armadura para determinar los parámetros representativos de la adherencia en el hormigón estructural con armaduras corroídas. Por otra parte, la recopilación de los datos de los estudios de adherencia con armaduras corroídas procedentes de la literatura científica, además de los resultados de la presente investigación, junto con la identificación de las variables relevantes en el comportamiento de la adherencia con armaduras sanas y corroídas ha servido para la obtención de una formulación realista y fiable para la evaluación conjunta de la adherencia con armaduras sanas y corroídas a partir de modelos de regresión múltiple. La formulación propuesta ha sido validada mediante criterios estadísticos y comparada con otras formulaciones propuestas en la literatura científica. Además se ha realizado un análisis de las variables influyentes de la formulación propuesta. También se ha obtenido un modelo numérico simple y eficiente, validado con alguno de los ensayos realizados en esta tesis, para simular la adherencia con armaduras sanas y corroídas. Finalmente, se presenta un procedimiento para realizar la evaluación de vigas deterioradas por corrosión mediante el método de los campos de tensiones que incluye la evaluación de la adherencia mediante la formulación sugerida en esta Tesis Doctoral. Las conclusiones alcanzadas en este trabajo han permitido evaluar la adherencia con armaduras corroídas de forma realista y fiable. Asimismo, se ha podido incluir la evaluación de la adherencia en el marco de la verificación de la seguridad estructural en elementos de hormigón armado deteriorados por corrosión. ABSTRACT Structural safety verification of both structures allowing a certain degree of deterioration in design and deteriorated existing structures needs strength models that factor in the effects of deterioration. In case of corrosion of steel bars in reinforced concrete structures, the resistance depends on many things such as the remaining cross-section of the corroded reinforcement bars, the stress-strain diagrams of steel, the concrete-reinforcement bond and corrosion-induced concrete cracking or spalling. Accordingly, the force transfer through the contact surface between concrete and reinforcement, bond, is one of the most important aspects to consider and it is the basis of the structural performance of reinforced concrete. Bond must assure anchorage of reinforcement and transmit shear stresses as a consequence of the different stresses along a structural element As a consequence of corrosion, the bond development may be affected and hence the transfer of longitudinal stresses. This PhD Thesis deals with ultimate limit state bond behaviour in structural concrete with corrode steel bars. The main objective is to obtain a realistic and reliable model for the assessment of bond within the context of structural safety verifications of reinforced concrete members with corroded steel bars. In that context, an experimental programme of eccentric pull-out tests were conducted, with different specimens, ones without corrosion and others subjected to accelerated or natural corrosion with different corrosion degrees. This type of bond test reproduces in a realistic and reliable way bond stresses in the anchorage zone. Moreover, for conducting these tests it was necessary to develop both a test procedure and also a data acquisition system including the use of an embedded fibre-optic sensing system with fibre Bragg grating sensors to obtain the representative parameters of bond strength in structural concrete with corroded steel bars. Furthermore, the compilation of data from bond studies with corroded steel bars from scientific literature, including tests conducted in the present study, along with the identification of the relevant variables influencing bond behaviour for both corroded and non-corroded steel bars was used to obtain a realistic and reliable formulation for bond assessment in corroded and non-corroded steel bars by multiple linear regression analysis. The proposed formulation was validated with a number of statistical criteria and compared to other models from scientific literature. Moreover, an analysis of the influencing variables of the proposed formulation has been performed. Also, a simplified and efficient numerical model has been obtained and validated with several tests performed in this PhD Thesis for simulating the bond in corroded and non-corroded steel bars. Finally, a proposal for the assessment of corrosion-damaged beams with stress field models including bond assessment with the proposed formulation is presented. The conclusions raised in this work have allowed a realistic and reliable bond assessment in corroded steel bars. Furthermore, bond assessment has been included within the context of structural safety verifications in corrosion-damaged reinforced concrete elements.

Influence of acrylic adhesive viscosity and surface roughness on the properties of adhesive joint

In adhesion, the wetting process depends on three fundamental factors: the surface topography of the adherend, the viscosity of the adhesive, and the surface energy of both. The aim of this paper is to study the influence of viscosity and surface roughness on the wetting and their effect on the bond strength. For this purpose, an acrylic adhesive with different viscosities was synthesized and some properties, such as viscosity and surface tension, were studied before adhesive curing took place. Furthermore, the contact angle and the lap-shear strength were analyzed using aluminum adherends with two different roughnesses. Scanning electron microscopy was used to determine the effect of the viscosity and the roughness on the joint interface. The results showed that the adhesive exhibits an optimal value of viscosity. Below this value, at low viscosities, the low neoprene content produces poor bond strength due to the reduced toughness of the adhesive. Additionally, it also produces a high shrinkage during curing, which leads to the apparition of residual stresses that weakens the interfacial strength. However, once the optimum value, an increase in the viscosity produces a negative effect on the joint strength as a result of an important decrease in the wettability.

Plaster reinforcement with fibers by mineral wool obtained from the recycle of construction and demolition waste

The use of mineral wool is becoming more widespread due to increased acoustic and thermal demands of Spanish Technical Building Code. This increase affects both in rehabilitation and new construction projects. Therefore, waste generation of this type of insulating material is having more importance. The main objective of this research is to study the possibility of recycling fiber obtained from mineral wool of the C&DW as an alternative material to chopped glass fibers that are currently used as reinforcing elements in the prefabricated plaster. To achieve this objective, series are made of plaster E-35 additived with rock wool residue and glass wool residue at different rates of addition. These series are repeated by changing the additive by E fiberglass (length of 25mm) to make a comparative analysis with respect to the series additived with mineral wool waste. All the series are subjected to the test to determine Shore C surface hardness and mechanical testing to determine the compressive and flexural strength. From the results it can be concluded that: with rock wool residue, increases Shore C hardness up to 15% with respect to the glass fiber and 9% with respect to the glass wool, with a percentage of addition 2%. With rock wool residue, weight is decreased by 5% with respect to the glass fiber and 4% with respect to the glass wool waste, with an addition percentage of 4%. For an addition rate of 4%, results in the flexural strength test with fiberglass are 85% higher than those obtained with glass wool residue. However, for a percentage of 1% addition, the results obtained with glass wool residue are 35% higher than those obtained with fiberglass. For an addition rate of 3% results in the compressive strength test with fiberglass are 54% lower than those obtained with rock wool waste and 70% lower than those obtained with glass wool waste. Comparing the two mineral wools, it can be concluded that up to 3% of the addition, the glass wool series results obtained are 10% higher than those additived with rock wool. However, higher percentages of addition show that the results obtained with rock wool are 35% higher than those obtained with glass wool. The general conclusion is that the series additived with mineral wool from C&DW show better results in tests than the ones used nowadays as plaster reinforcement.

Evaluación del comportamiento mecánico de estructuras de hormigón armado utilizadas como elementos termoactivos

El objetivo de este trabajo es conocer las posibles modificaciones que puede producir en el comportamiento de las estructuras de hormigón armado (EHA) el hecho de que sean utilizadas como estructuras termoactivas, ya sea como intercambiadores en contacto con el terreno, o como sistema de distribución de calor utilizando la inercia térmica de los elementos de hormigón del edificio, basándose en el uso de energías renovables. Las EHA termoactivas se caracterizan por la incorporación en su interior de tubos de polietileno por los que circulan fluidos a temperaturas medias, que pueden incidir en el comportamiento mecánico de los elementos estructurales debido a dos efectos fundamentales: el incremento de temperatura que se produce en el interior de la EHA y la perturbación provocada por la incorporación de los tubos de polietileno. Con este fin, se ha realizado una campaña experimental de probetas de hormigón, estudiando los dos efectos por separado, por un lado se ha evaluado el comportamiento de probetas de hormigón tipo H-25 y tipo H-30 sometidas a cuatro temperaturas diferentes: 20ºC, 40ºC, 70ºC y 100ºC, ensayando la resistencia a compresión y la resistencia a anclaje/adherencia mediante ensayo “pull-out”; y, por otro lado, se ha evaluado el comportamiento de probetas de hormigón tipo H-25 y tipo H-30, elaboradas con dos tipos de molde (cilíndrico y cúbico), en las que se ha colocado tubos de polietileno en su interior en distintas posiciones, ensayando su resistencia a compresión. Los resultados de los ensayos han puesto de manifiesto que aunque se produce una disminución en la resistencia a compresión, y a arrancamiento, del hormigón, al ser sometido a aumentos de temperatura, esta disminución de la propiedades mecánicas es inferior al 20% al no superar esta tecnología los 70ºC; y respecto a la variación de la resistencia a compresión de probetas cilíndricas y cúbicas, debidas a la incorporación de los tubos de polietileno, se observa que si la posición de los mismos es paralela a la dirección de la carga tampoco se ven comprometidas las propiedad mecánicas del hormigón en valores superiores al 20%. ABSTRACT The aim of this project is to study the effects of using concrete structures as thermo-active constructions, either as energy foundations or other kind of thermo-active ground structures, or as a thermally activated building structure utilizing its own thermal mass conductivity and storage capacity to heat and cool buildings, based on renewable or “free” energy sources. The pipes, filled with a heat carrier fluid, that are embedded into the building´s concrete elements may bring on two different adverse effects on concrete structures. In one hand, the consequence of thermal variations and, on the other hand, because of the fact that the pipes are inside of the concrete mortar and in direct contact with the reinforcing steel bars. For this reason, different types of specimens and testing procedures have been proposed to discuss the effects of temperature (20º, 40ºC, 70ºC y 100ºC) on the performance of two different hardened concrete: H-25 and H-30, and the effects of having the pipes embedded in different positions inside of specimens made of two types of concrete, H-25 and H-30, and with two kind of cast, cylindrical and cubical. The experimental program includes the use of compressive strength test and also pull-out test, in order to investigate the interfacial adhesion quality and interfacial properties between steel bar and concrete. The results of the mechanical test showed that the increase of temperature in hardened concrete specimens lower than 70ºC, and the introduction of embedded pipes placed in parallel to the load, in cylindrical or cubic specimens, does not jeopardize the mechanical properties of concrete with strength decreases higher than 20%.

Perfil antropométrico y respuesta psico-fisiológica en escalada deportiva en roca: Diferencias entre modalidades

En la literatura se ha descrito el perfil antropométrico y la respuesta psicofisiológica en escalada deportiva en roca, pero hasta la fecha, no se habían analizado las diferencias existentes entre sus principales modalidades. El objetivo de la presente tesis fue describir las características antropométricas del escalador de competición y comprobar la existencia de diferencias entre los participantes de distintas modalidades, así como analizar la respuesta psico-fisiológica durante la ejecución de un búlder y una vía, además de evaluar las diferencias entre su realización a vista o tras un ensayo. Para ello, efectuamos dos estudios diferentes: en el primero participaron voluntariamente 61 hombres y 18 mujeres, participantes en cuatro pruebas del circuito nacional de competición de escalada durante el año 2009, tres de ellas de la modalidad de dificultad a vista y una de búlder. Se realizaron mediciones antropométricas, prueba de fuerza de prensión manual antes y después de competir, y se cumplimentó un cuestionario donde se evaluaba la percepción del esfuerzo y la experiencia deportiva. En el segundo estudio, 23 escaladores, 15 hombres y 8 mujeres, divididos en tres grupos en función de su nivel de rendimiento, realizaron de manera voluntaria distintas pruebas durante tres días separados entre sí al menos 48 horas. El primer día rellenaron un cuestionario sobre su experiencia deportiva y nivel de rendimiento, fueron pesados, tallados y sometidos a un escáner de cuerpo completo en densitómetro con objeto de medir la composición corporal. El segundo día realizaron previo calentamiento, un búlder a vista y, tras un descanso de 15 minutos, escalaron una vía a vista acorde con su nivel. El tercer día, después de calentar y disponer de 20 minutos para ensayarlo, repitieron la escalada del búlder. Tras un descanso de 15 minutos y 20 minutos de ensayo, realizaron un segundo intento a la vía. Se registraron los valores en la respuesta cardiorrespiratoria, se obtuvieron muestras de lactato en sangre del lóbulo de la oreja y se realizaron pruebas de fuerza de prensión manual antes y después de la escalada. También se pasó un cuestionario para medir la ansiedad y autoconfianza así como el esfuerzo percibido. Los resultados no mostraron diferencias antropométricas significativas entre los participantes en competiciones de búlder y los que participaron en competiciones de escalada de dificultad a vista. Se dieron diferencias en la pérdida de fuerza antes y después de escalar entre dichos participantes. Las mujeres obtuvieron menor fuerza de prensión manual que los hombres pero la misma pérdida de fuerza entre el instante antes de competir y el posterior. La respuesta fisiológica durante la ejecución del búlder fue menor que la obtenida durante la ejecución de la vía. Hubo pérdida de fuerza de prensión manual entre el instante anterior y el posterior a ejecutar la vía, pero no al hacer el búlder. Sin embargo, no se dieron diferencias en la ansiedad y la autoestima provocada por ambas modalidades, por lo que deducimos que la ejecución de un búlder y una vía presentan una respuesta fisiológica distinta. Proponemos que la respuesta está relacionada, sobre todo, con las variables de ejecución, de tal manera que a mayor distancia y/o tiempo recorrido en la escalada, mayor será la contribución anaeróbica al esfuerzo y la fatiga manifestada como pérdida de fuerza que, en el caso del búlder, fue mínima o inexistente. En el segundo intento, tras un ensayo de 20 minutos en el búlder, se consiguió mejorar el rendimiento respecto al primer intento, que se manifestó con un aumento en la distancia recorrida. Sin embargo, en la vía no se dieron diferencias entre ambos intentos, ni en la ejecución, ni en la respuesta fisiológica, ni en la ansiedad, ni siquiera en la fuerza de prensión manual. ABSTRACT It has been described in the literature the anthropometric profile and psychophysiological response in rock climbing, but so far not been analyzed differences between its main modalities. The aim of this thesis was to describe the anthropometric characteristics of the climber competitor and check for differences between participants of different modalities and to analyze the psycho-physiological response during the execution of a boulder and a route also to assess differences between on sight and redpoint attempts. We made two different studies: in the first 61 men and 18 women who attended four competitions of national climbing circuit in 2009, three of them on-sight difficulty competitions and a boulder competition participated voluntarily. Anthropometric measurements, a hand grip strength test before and after competing were registered for each climber, and a questionnaire which assessed perception of effort and the climbing experience was fulfilled. In the second study, various tests were conducted on 23 volunteer climbers, 15 men and 8 women, during three days separated for, at least, 48 hours of resting, divided into three groups according to their performance. The first day, climbers completed a questionnaire on their experience and performance level. It was recorded weight, height and they underwent a full body scan densitometer in order to measure body composition. The second day, after previous warming-up, they climbed a boulder on sight and, after a break of 15 minutes, climbed a route on-sight according to their level. The third day, after warming-up and have 20 minutes to try it, they repeated the bouldering climbing. After a break of 15 minutes and 20 minutes of essaying, they made a second attempt at the route. Values in the cardiorespiratory response were recorded, blood lactate samples were obtained from earlobe, and hand grip strength was tested before and after the climb. They also filled a questionnaire to measure anxiety and self-confidence and perceived exertion. The results showed no significant anthropometric differences between participants in bouldering competitions and participants in competitions on-sight difficulty climbing. There were found differences in strength loss before and after climbing between those participants. Women had less hand grip strength than men but the same loss of strength between the records carried out before and after competing. The physiological response recorded for boulder climbing was lower than the obtained for the route. There was loss of hand grip strength between the time before and after running the route but not for bouldering. However, there were no differences in anxiety and self-esteem caused by both modalities, so we conclude that the implementation of a boulder and a route have different physiological responses. We think that this response is mainly related to performance variables, as a greater distance and/or travel time on the climb, the higher the anaerobic contribution to the effort and fatigue as manifested by loss of strength in the case of the boulder was minimal or nonexistent. In the second attempt after 20 minutes in the boulder better performance was achieved on the first attempt, which was manifested by an increment of climbing distance. However, there were the differences in the route between the two attempts, either in execution or in the physiological response, or anxiety, or even in hand grip strength.

A methodology to measure the interface shear strength by means of the fiber push-in test

A methodology is presented to measure the fiber/matrix interface shear strength in composites. The strategy is based on performing a fiber push-in test at the central fiber of highly-packed fiber clusters with hexagonal symmetry which are often found in unidirectional composites with a high volume fraction of fibers. The mechanics of this test was analyzed in detail by means of three-dimensional finite element simulations. In particular, the influence of different parameters (interface shear strength, toughness and friction as well as fiber longitudinal elastic modulus and curing stresses) on the critical load at the onset of debonding was established. From the results of the numerical simulations, a simple relationship between the critical load and the interface shear strength is proposed. The methodology was validated in an unidirectional C/epoxy composite and the advantages and limitations of the proposed methodology are indicated.

Bonding shear strength in timber and GFRP glued with epoxy adhesives

The bonding quality of epoxy glued timber and glass fibre reinforced polymers (GFRP) was evaluated by means of compression loading shear test. Three timber species (Radiata pine, Laricio pine and Oak) and two kinds of GFRP (plates and rods made with polyester resin reinforced with mat and roving glass fibre) were glued and tested using three epoxy formulations. The increase in shear strength with age after the setting of epoxy formulations and the effect of surface roughness on timber and GRP gluing (the planing of the surface of timber and the previous sanding of GRP) were studied. It can be concluded that the mechanical properties of these products make them suitable for use in the reinforcement of deteriorated timber structures, and that a rough timber surface is preferable to a planed one, while the previous sanding of GRP surfaces is not advantageous.