The main objective of this research is to establish durability design models for flexural concrete elements subjected to a given aggressive environment. Because corrosion of reinforcement due to chloride ingress is the most significant threat to an existing reinforced concrete structure such as a road bridge or a harbor facility which is exposed to chloride-rich environments, corrosion due to chloride ingress is emphasized in this research. The concrete beams can get corroded to different corrosion levels. Mass loss of the reinforcement is an important parameter, and it can help define the corrosion level, and this information can be used to develop a correlation between corrosion, cracking, bond strength at the steel-concrete interface, and the ultimate strength of the reinforced concrete elements. Therefore, the prediction model of reinforcement mass loss under different levels of chloride concentration needs to be established first using the principles of corrosion electrochemistry. Secondly, design development length is determined based on the mass loss prediction and bond strength equation. Thirdly, the model for the prediction of flexural capacity of reinforced concrete beam for a given design service life is established. Some examples of practical durability design are presented. Finally, a step-by-step durability design procedure is recommended for use by practicing engineers. (Abstract shortened by UMI.)

Lack of a scientific design approach for repair and rehabilitation of corroding reinforced concrete infrastructure has resulted in significant financial and social costs. This experimental program was primarily undertaken to examine the corrosion process in reinforced concrete repair, which has different characteristics as compared with corrosion in new construction. The program was designed to gain a deeper understanding of how certain restoration strategies may lead to problems of electrochemical incompatibility and result in ineffective corrosion mitigation. / Fifteen specimens, 1m by 1m by 0.2m, were cast to represent a section of a deteriorating reinforced concrete bridge deck slab. The central portion was uniquely designed to simulate the deterioration caused by corrosion activity in a bridge deck slab. After initiating corrosion using wetting and drying cycles with 15% salt solution, each specimen was subjected to a unique restoration strategy. The wetting and drying cycles continued, and a monitoring program was established to observe the corrosion activity of each specimen. / The results corroborate current research, that patch repairs can trigger the formation of a macrocell corrosion cell, or a ring of active corrosion surrounding the repaired zone. In addition, the results from the electrochemical testing revealed sharp differences in the corrosion behaviour of the different restoration strategies. However, the physical evidence of minimal corrosion for all four specimens that were demolished at the end of the testing period, reveals a discrepancy with the electrochemical testing results.

Montreal is the second largest metropolitan centre in Canada. It is exposed to significant seismic hazards and ranks second in urban seismic risk in Canada. The seismicity of Eastern Canada is controlled by intraplate earthquakes. The damages caused by the earthquakes on infrastructures depend on the quake magnitude, distance from the epicentre and the type of underlying soil deposit. The surficial deposits and landscape of Montreal have been shaped by many geological episodes. Alternate periods of glaciation and melting, deposited considerable quantities of soft soil, and were later altered by the emergence of the Champlain Sea, depositing clay and sand layers. Lastly, the influence of the St Lawrence River and its secondary channels deposited fine and coarse materials on the river banks. / Unconsolidated cohesionless soil with water leads to liquefaction phenomenon during earthquakes. Various researches have been carried out on, the nature of soil deposits in Montreal and Microzonation of the Island of Montreal. The objective of this thesis is to map liquefaction hazard for the Island of Montreal. The methodology consists of using 'the Simplified procedure', developed by Seed and Idris (1971) and updated by Youd et al., (1998), to obtain the factor of safety against liquefaction. SHAKE2000 is used to perform the analysis. A comprehensive data on the type of soil and standard penetration number was required to perform this analysis, which was provided by the City of Montreal. The resulting maps of this analysis illustrate the liquefaction hazard of the region in terms of factor of safety against liquefaction for two earthquake magnitudes, 7.5 and 5.5.

The absence of a concrete slab results in flexible pavement systems that are supported only by the underlying granular base layer. However, local deterioration of the base material supporting flexible pavements may develop during the service life of the pavement structure. This can result in unexpected surface displacements and premature distressing of the pavement. This research program quantifies the deformation rate of a flexible pavement system subject to a weakened base course layer within the confines of geogrid-reinforced and non-reinforced pavement. A series of simplified weakened zones of cylindrical and ellipsoidal shapes are introduced within the base layer. These zones are then subjected to gradual reduction in strength and are increased in volume to represent the growth of the weakened section. The pavement response to an applied tire load is analyzed using three-dimensional finite element analysis. The analysis is conducted for a varying base layer thicknesses and alternate geogrid locations. The mechanisms of base layer weakening, the development of the numerical model and a summary of the calculated pavement performance are presented. The results of the models indicate that when used appropriately, the presence of geogrid reinforcements aids in reducing the surface settlement of road pavements while increasing the load carrying of the pavement. Furthermore, variations in the base layer thickness of a flexible pavement system have a significant effect on the pavement performance as does the presence of weakened zones within the base layer. / L'absence d'une dalle de béton oblige les systèmes de chaussées flexibles d'être entièrement soutenus par la base granulaire souterraine. Mais la détérioration localisée du matériel granulaire peut se développer durant la durée de vie de la chaussée. Par conséquent, des tassements ainsi que la détresse prématurée peuvent apparaître. Ce projet de recherche examine le processus de détérioration d'une chaussée flexible suite à l'affaiblissement du matériel comprenant la couche de base. Ce processus est étudié pour des chaussées renforcées et non-renforcées par une géogrille. La formation de zone de sols faiblissant en forme cylindrique et ellipsoïdaux est introduite dans la base granulaire. Ces zones sont alors soumises à un affaiblissement progressif et une augmentation en volume pour simuler la croissance de la section affaiblie. À l'aide d'une modélisation numérique tridimensionnelle, le tassement qui se développe suite à l'application d'une charge de roue est évalué. Cette analyse est réalisée en variant l'épaisseur de la couche de base ainsi que le placement de la géogrille. Les mécanismes de l'affaiblissement de la base granulaire, le développement du modèle numérique et un résumé des résultats sont présentés. Les résultats de cette recherche montrent que l'introduction d'une couche renforcée par une géogrille diminue les tassements et augmente la capacité portante de la chaussée. Cependant, ces résultats sont limités seulement à l'installation convenable de la géogrille. De plus, la variation de l'épaisseur de la base granulaire a un effet significatif sur la performance d'une chaussée flexible ainsi que le processus d'affaiblissement du matériel qui comprend la couche de base.

This thesis deals with the assessment of seismic accelerations in buildings and the seismic analysis of components installed on building rooftops, with special focus on operational telecommunition towers during and after earthquake shaking. / First, acceleration data recorded during the 1999 Chi Chi earthquake from 11 instrumented buildings located in Taiwan were studied. Fundamental building periods were extracted by system identification and compared to those evaluated according to the equations proposed in the 2005 edition of the National Building Code of Canada (NBCC). Next, rooftop acceleration spectra and time histories were evaluated using 3-D finite element building models; three models of instrumented buildings were calibrated using accelerograms from the Chi Chi earthquake and the fourth model is a building located in downtown Montreal. The building models were subjected to 44 historical strong motion accelerograms and 30 synthetic accelerograms compatible with the target uniform hazard spectra specified in NBCC 2005 for Montreal. Based on both the experimental and numerical results, a maximum rooftop acceleration amplification of 4 is proposed for low/medium rise buildings and 3 for flexible high-rise buildings (T > 1.7 s). / In the second stage, a simplified method for the prediction of seismic shear forces and overturning moments at the base of self-supporting steel lattice telecommunication towers mounted on building rooftops is presented. The proposed method involves the estimation of four parameters: the rooftop seismic acceleration, the mass distribution profile of the tower along its height, the maximum acceleration amplification at the tower top, and the fundamental sway mode shape of the tower on a rigid base. The method was validated by means of numerical results of nine generated building-tower combinations composed of three towers assumed to be mounted on three of the building models studied in the first stage of the research. The building-tower combinations were subjected to the same sets of earthquake records used for the prediction of accelerations. It was found the proposed method yields conservative results in all the cases analyzed. / In addition, the empirical component force amplification factor for telecommunication towers as proposed in the NBCC 2005 was compared to the factors evaluated for the towers of the 16 building-tower combinations. Improved component force amplification factors based on rational analysis are proposed.

The modelling of contaminant transport in porous media is an important topic to geosciences and geo-environmental engineering. An accurate assessment of the spatial and temporal distribution of a contaminant is an important step in the environmental decision-making process. Contaminant transport in porous media usually involves complex non-linear processes that result from the interaction of the migrating chemical species with the geological medium. The study of practical problems in contaminant transport therefore usually requires the development of computational procedures that can accurately examine the non-linear coupling processes involved. However, the computational modelling of the advection-dominated transport process is particularly sensitive to situations where the concentration profiles can exhibit high gradients and/or discontinuities. This thesis focuses on the development of an accurate computational methodology that can examine the contaminant transport problem in porous media where the advective process dominates. / The development of the computational method for the advection-dominated transport problem is based on a Fourier analysis on stabilized semi-discrete Eulerian finite element methods for the advection equation. The Fourier analysis shows that under the Courant number condition of Cr=1, certain stabilized finite element scheme can give an oscillation-free and non-diffusive solution for the advection equation. Based on this observation, a time-adaptive scheme is developed for the accurate solution of the one-dimensional advection-dominated transport problem with the transient flow velocity. The time-adaptive scheme is validated with an experimental modelling of the advection-dominated transport problem involving the migration of a chemical solution in a porous column. A colour visualization-based image processing method is developed in the experimental modelling to quantitatively determinate the chemical concentration on the porous column in a non-invasive way. A mesh-refining adaptive scheme is developed for the optimal solution of the multi-dimensional advective transport problem with a time- and space-dependent flow field. Such mesh-refining adaptive procedure is quantitative in the sense that the size of the refined mesh is determined by the Courant number criterion. Finally, the thesis also presents a brief study of a numerical model that is capable to capture coupling Hydro-Mechanical-Chemical processes during the advection-dominated transport of a contaminant in a porous medium.

The inadequate and inaccurate application of corrosion monitoring techniques for testing reinforced concrete structures had led to considerable early deterioration of these facilities and has resulted in high restoration and replacement costs. The prospect and diligent use of readily available early damage detection tools for use in regular maintenance of these structures could permit timely interventions for restoration and upkeep. / This research program was aimed at determining the effectiveness of selected nondestructive testing methods as reliable indicators of early onset corrosion in steel-reinforced concrete. The methodology analyzed the onset of corrosion in a set of steel reinforcing rebars at a specific depth from the concrete surface, and used the resulting data as a predictor of corrosion activity in other reinforcing bars at different depths. / Two series of tests were conducted: the first series included nine individually reinforced concrete samples with varying cover thicknesses; the second series consisted of nine reinforcing steel bars, distributed equally in concrete specimens at different depths from the concrete cover. All specimens were subjected to accelerated corrosion using methods developed at McGill University to reproduce corrosion of steel rebars in a naturally aggressive environment as closely as possible. / The research results concur with the findings of other research programs, affirming that these monitoring tools are adequate predictors of corrosion. Nonetheless, they lack precision and are unable to track the time to corrosion initiation in structural concrete elements.

A large portion of bridges in North America was built 50 to 75 years ago, designed with a similar service lifespan. As a result, many existing bridges require full inspections, repairs, retrofitting or decommissioning. With limited resources available, it is crucial that maintenance strategies and rehabilitations are planned accordingly. To do so, designers, engineers and owners must be informed on the timeframe of the degradation process. Deterioration of reinforced concrete bridges is strongly associated with corrosion of steel caused by the application of deicing salts. In Montreal, the severity and duration of the winter period is a major concern. Motorists require safe roads to commute, thus large quantities of salt are spread each winter season to achieve bare pavement. The chlorides from the deicing salt enter the concrete and destroy the passive oxide layer protecting the reinforcing steel, making the steel vulnerable to corrosion. In this work, the "TransChlor" computer software was used to simulate chloride ion ingress into concrete cover. A climatic model for Montreal weather was used in the simulations to accurately portray real life environmental conditions. Cumulative probability distributions for corrosion initiation with respect to time and embedment depths were produced for mist, direct, and splash exposure conditions. / Une grande partie des ponts en Amérique du Nord a été construite il y a 50 à 75 ans, et conçus avec une durée de vie de service similaire. En conséquence, de nombreux ponts existants requièrent des inspections complètes, des réparations, lune modernisation ou un déclassement. Avec les ressources limitées disponibles, il est essentiel que des stratégies de maintenance et de réhabilitation soient planifiées. Pour ce faire, les concepteurs, les ingénieurs et les propriétaires doivent être informés sur les processus de dégradation. La détérioration des ponts en béton armé est fortement associée à la corrosion de l'acier par l'application de sels de déglaçage. À Montréal, la sévérité et la durée de la période hivernale est un problème majeur. Les automobilistes ont besoin des routes sûres et, une grande quantité de sel est utilisée chaque hiver pour déglacer les routes. Les chlorures provenant du sel de déglaçage pénètrent le béton et détruisent la couche d'oxyde passive qui protège l'acier d'armature, et rend l'acier plus vulnérable à la corrosion. Dans ce travail, le logiciel "TransChlor" a été utilisé pour simuler la pénétration des ions chlorures dans la couverture de béton. Un modèle climatique de Montréal pour la météorologie a été utilisé dans les simulations pour représenter exactement les conditions réelles de l'environnement. Les distributions de probabilité cumulative du temps requis pour l'initiation de la corrosion ont été obtenus en fonction de la couverture de béton pour des conditions d'exposition de brouillard, d'exposition directe, et par éclaboussures.

This thesis documents an investigation of the use of the roof diaphragm flexibility in the seismic design and analysis procedure of single-storey steel buildings designed otherwise in accordance with the provisions of the 2010 NBCC and the 2009 CSA S16. The design approach considers the members of the vertical bracing system as the ductile fuse elements in the seismic force resisting system (SFRS), whereas the diaphragm remains elastic. An alternative design approach was also examined in which the steel deck roof diaphragm acts as a ductile fuse element in the SFRS; at present this procedure is not permitted by the NBCC or CSA S16. The investigation was reliant on a complementary three phase test program in which nineteen large-scale roof diaphragm specimens were dynamically excited with a sequence of increasing amplitude loading protocols. The first part of the study comprised the development of a deep horizontal plane truss numerical model using the OpenSees software platform to reproduce the dynamic characteristics as well as the elastic and inelastic response of the nineteen test specimens. The predicted fundamental period of vibration, the elastic response and the inelastic hysteretic response were compared with the test results and the models were calibrated accordingly. In the second part of the study, the detailed design and non-linear time history dynamic analyses of representative medium size and large size single-storey steel buildings were carried out. The intent was to evaluate the overall behaviour of four structural systems whose design was tailored to either rely on the flexibility of the diaphragm or to allow the roof decking / connections to deform inelastically. OpenSees building models were developed by integrating a non-linear brace model with the non-linear diaphragm model. Dynamic analyses were performed on the designed buildings using the corresponding OpenSees building model and responses were evaluated under a suite of design level earthquake signals. The study illustrated that the analytically predicted fundamental period of vibration which includes the influence of the roof deck diaphragm could be used in the design of such single-storey steel buildings. This finding leads to the recommendation to revise the expression given in 2010 NBCC for the fundamental period of vibration as well as for the period limitation. Further, compared to the different structural systems, the buildings designed with EBF structural system were found most promising in terms of the relative capacity force on the steel deck diaphragm and the building response. The study also found that the diaphragms in the EBF and CBF structural systems could be designed for the force corresponding to the seismic base shear with RdRo = 2, if it controls the design. Moreover, significant shear strength degradation and concentration of inelastic demand were observed in the diaphragm at the edge of the buildings when the steel decks were parallel to the loading direction and the diaphragm was designed as a ductile fuse element. This illustrates that the value of 2.0 that was assumed for the seismic force reduction parameter Rd may not be appropriate in the design of such buildings. Similar strength degradation and concentration of inelastic demand in the diaphragm were observed in the buildings with a Type CC structural system, which shows that the diaphragm may need to be designed corresponding to the elastic seismic force. / La présente thèse porte sur une recherche sur l'utilisation de la flexibilité du diaphragme de toit dans la conception et l'analyse parasismiques des bâtiments d'un seul étage en acier conçus selon les dispositions parasismiques des normes de construction CNBC 2010 et CSA S16-09. L'approche de conception consiste à considérer les diagonales de contreventement faisant partie du système de résistance aux forces sismiques (SRFS) comme les éléments ductiles, alors que le comportement du diaphragme de toit demeure dans le domaine élastique. Une approche différente a aussi été examinée selon laquelle le diaphragme de toit en acier agit comme un élément ductile dans le SFRS, approche qui n'est pas autorisée dans les codes CNBC et CSA S16 présentement en vigueur. L'étude est tributaire d'un programme d'essais complémentaires en trois phases durant lequel dix-neuf spécimens de diaphragme de toit à grande échelle ont été soumis à des essais dynamiques selon un protocole de chargement à amplitude variable. La première partie de l'étude a porté sur l'élaboration avec le logiciel OpenSees d'un modèle numérique de diaphragme de toit composé d'un système de treillis afin de reproduire les caractéristiques dynamiques de même que les comportements élastique et inélastique des dix neuf spécimens. Les prédictions de la période fondamentale de vibration, du comportement élastique et de la réponse sous sollicitation inélastique cyclique ont été comparées aux résultats des essais au laboratoire, et les modèles ont été ajustés en conséquence. Dans le seconde partie du programme d'essais, la conception de différents bâtiments à un étage de taille moyenne et de taille grande, ainsi que l'analyse non-linéaire de ceux-ci, a été complétée. L'objectif était d'évaluer le comportement global de quatre systèmes structuraux dont la conception avait été adaptée pour prendre en compte la flexibilité du diaphragme de toit ou permettre les déformations inélastiques des connecteurs du tablier métallique. Des modèles des bâtiments ont été développés avec le logiciel OpenSees, en intégrant un modèle non linéaire des diagonales et le modèle non linéaire du diaphragme. Des analyses dynamiques des bâtiments ainsi conçus ont été réalisées avec le logiciel OpenSees et leur comportement a été évalué sous un ensemble de mouvements de sol sismique d'amplitude correspondant au niveau sismique de conception. L'étude à démontré que la période qui inclus l'influence du diaphragme peut être utilisée dans la conception d'un bâtiment à un étage en acier avec ce type de construction. Cette découverte mène à la recommandation de réviser l'expression du CNBC 2010 pour la période fondamentale du bâtiment ainsi que la limite empirique sur celle-ci. Les bâtiments construits avec un système de contreventements de type excentrique sont les plus prometteurs au niveau de la capacité relative du diaphragme en acier et la comportement du bâtiment. L'étude a aussi démontré que les diaphragmes qui sont unis avec un système de contreventements concentriques ou excentriques peuvent êtres conçus pour la force qui correspond au cisaillement calculé avec RdRo = 2, si celui-ci contrôle la conception du diaphragme. Il faut aussi noter qu'une dégradation significative de la capacité en cisaillement et une concentration de la demande élastique à été observée aux côtés des bâtiments quand la tôle est installée parallèle à la direction de la charge et quand le diaphragme est conçu comme l'élément sacrifiant. Ceci illustre le fait que la valeur de 2.0 assumé pour la ductilité du système (Rd) n'est pas nécessairement appropriée pour la conception de ce genre de bâtiments. Cette même concentration de la demande aux côtes et dégradation du système a aussi été observée dans les bâtiments conçus avec un système latéral de type 'construction conventionnelle' ce qui veut dire que le diaphragme devrait sans doute être construit pour la force sismique élastique.

The development of engineered wood products (EWPs), such as glued laminated timber, has led to the production of structural members with increased strength and spanning capabilities compared with solid sawn lumber and timber, and similar to what one could attain with commonly sized steel and reinforced concrete members. An important reason for the underutilisation of wood for heavy construction is the limitation of the CSA O86 Wood Design Standard especially as it applies to the design of fastenings for highly loaded connections. This thesis addresses the connection design of glulam and sawn wood using structural self-tapping/self-drilling screws from Europe.The CSA–O86 Standard provides formulas for the design of both lag and wood screws in Canada. Recently in Europe structural screws have been developed which combine the advantages of both lag and wood screws. These screws have high load carrying capacities and withdrawal strengths and are also self-tapping/self-drilling, hence in most cases do not need lead holes. However CSA O86 has no specific design provisions for these screws. To assess their viability for use in Canada, two test programs were carried out; the first on the withdrawal resistance of the screws and the second on the performance of inclined screws in joist-to-header connections. Douglas fir Larch(20f-E), Spruce Pine (20f-E) and Nordic Lam (24f-1.9E) glulam in conjunction with a variety of 6, 8, 10 and 12mm diameter European structural screws were used for the withdrawal tests, while the joist to header connections were made of No.2 white pine timber connected by double threaded 8.2mm WT-T screws from SFS intec. In all 1960 withdrawal test were carried out. The test setup and procedure was modelled after ASTM D1761.The joist to header test set-up, involving 14 tests was modelled after ASTM D7147. The main aim of the withdrawal test program was to recommend a generic equation for use in the design of these screw connections with Canadian glulam. In the process the effects of Canadian wood density, depth of penetration, screw diameter and lead holes on the withdrawal resistance of the screws were assessed. The aim of the joist to header connection test program was to compare the performance of dry specimen with that of the same connection in wet timber.The test results demonstrated that the withdrawal strength per unit length increases with denser wood, except 6mm and 8mm in Nordic Lam glulam. The depth of penetration affects the withdrawal strength, where for larger screws an increase of slightly more than double was obtained for 12d penetration compared to 6d.The orientation of the glulam, that is either top or side, was insignificant other than the effect it had on the scatter of the strength results. The use of lead holes was shown for both the 8mm and 10mm not to influence the withdrawal strength. However the lead holes improved the ease of installation for the larger screws, especially in the dense glulam.The tests results were compared with the predicted characteristic and average withdrawal resistance values, which were calculated using formulas found in timber codes around the world, namely CSA O86 (Canada), NDS (USA), Eurocode 5 (Europe), DIN 1052 (German) and from other researchers including Frese and Blaβ (GER), Pirnbacher and Schickhofer (AUT) and McLain (USA). All methods resulted in a reasonable prediction of the withdrawal resistance except for the CSA O86 formula for lag screws which was very conservative. The McLain formula for lag screws provided the closest prediction of the test result, but bearing in mind the variability of wood, all the other methods could be considered as acceptable except for the CSA O86 lag screw equation. Regarding the joist-to-header cross screw connections, the dry tests were measured to have a 35% increase in resistance compared with the wet specimens; furthermore, the Kevarinmäki formula provided the most accurate prediction of the resistance. / Le développement de produits de bois d'ingénierie a entrainé la production d'éléments de structure de la résistance et de portée supérieures, et similaires à ce qu'il pourrait être atteint avec de l'acier de taille commune et des éléments porteurs en béton armé. Une raison importante pour la sous-utilisation de bois pour la construction lourde est la limitation des règles de calcul des charpentes en bois CSA O86, en particulier lorsqu'elles s'appliquent au design d'assemblages soumis à de charges importantes. Cette thèse étudie la conception des assemblages en bois lamellé-collé et bois scié utilisant des vis auto perçantes européennes.La norme CSA O86 fournit des formules pour la conception de tire fond et des vis à bois au Canada. En Europe, des vis qui présentent à la fois les avantages de tire fond et des vis à bois ont été récemment développées. Ces vis ont de résistance latérale et à l'arrachement élevée, et sont auto-perçantes. Toutefois, la norme CSA O86 ne contient aucune disposition spécifique pour la conception de ces vis. Pour évaluer la viabilité de leur utilisation au Canada, deux séries de test ont été réalisées: la première sur la résistance à l'arrachement des vis et la seconde sur la performance de vis inclinées dans les connexions entre solive et poutre. Du lamellé-collé Douglas-Mélèze (20f-E), Épinette-Pin (20f-E) et Nordic Lam (24f-1.9E) ont été utilisés avec des variétés de vis européennes de 6, 8, 10 et 12 mm de diamètre pour les essais d'arrachement, tandis que les connections solive-poutre étaient faites de pin blanc no 2 relié par des vis WT-T à double filetage 8.2mm. Au total, 1960 tests d'arrachement ont été effectués d'après ASTM D1761 tandis que 14 du test solive-poutre mis en place ont été effectués d'après ASTM D7147. L'objectif principal du programme de tests d'arrachement était de recommander une équation générique à utiliser dans la conception de ces connexions de vis avec du lamellé-collé canadien. Les effets de la densité du bois, de la profondeur et du diamètre de la vis et des trous pilotes sur la résistance à l'arrachement des vis ont été évalués. Test de connexion solive-poutre était de comparer la performance de l'échantillon sec avec celle de la même connexion en bois humide.Les résultats montrent que la résistance à l'arrachement augmente avec la densité du bois, à l'exception de vis 6 et 8 mm dans le lamellé-collé Nordic Lam. La résistance à l'arrachement est influence par la profondeur de pénétration et pour les plus grandes vis, il a été un peu plus que doublée pour une pénétration 12d par rapport à 6d. L'orientation du lamellé-collé, qui peut être transversale ou latérale, n'a pas d'importance autre que son effet sur la dispersion des valeurs de résistance. Il a été montré que trous pilotes n'influence pas la résistance à l'arrachement pour les diamètres 8 mm et 10 mm mais elles améliorent la facilité d'installation pour les vis plus grandes, en le bois lamellé-collé dense.Les résultats des tests ont été comparés avec les valeurs caractéristiques prédites par calculs selon des formules trouvées dans cette codes ; CSA O86 (Canada), NDS(USA), l'Eurocode 5 (Europe),DIN 1052(Allemagne) et de chercheurs, dont Frese et Blaβ, Pirnbacher et Schickhofer, et McLain. Toutes les méthodes ont conduit à une prédiction raisonnable de la résistance à l'arrachement, à l'exception de la formule de CSA O86 pour les vis de compression, qui est très conservatrice. La formule de McLain pour les vis à compression donne la prédiction la plus proche des résultats des essais, mais compte tenu de la variabilité du bois, toutes les autres méthodes peuvent être considérés comme acceptables, à l'exception de celle de la CSA O86 pour les tire fond. En ce qui concerne les connexions solive-poutre par des vis inclinées,les essais à sec montrent une augmentation de 35% de la résistance par rapport aux spécimens humides;en outre la formule de Kevarinmäki donne la prédiction de résistance la plus précise.