Ce travail s'inscrit dans le contexte de l'évaluation et de la prévention des risques d'effets dominos suite à une explosion sur un site industriel. Il s'intéresse plus précisément à la vulnérabilité de réservoirs de stockage soumis à une onde de souffle. Le mode d'explosion étudié est celui de la détonation gazeuse. L’objectif est double. Il consiste d'une part à caractériser le chargement auquel est soumis le réservoir et d'autre part à prédire la réponse du réservoir à ce chargement. Des expérimentations ont été réalisées à échelle réduite, pour trois géométries caractéristiques de réservoirs, satisfaisant des conditions de similitudes énergétiques et mécaniques. Des essais sur maquettes de réservoirs rigides ont permis de caractériser le chargement issu d'une détonation : répartition spatio-temporelle de la surpression et de l’impulsion et détermination de coefficients de réflexion. Des essais sur maquettes métalliques déformables ont permis de déterminer le niveau de chargement nécessaire pour endommager la structure par flambage. Une approche semi-analytique du flambage dynamique basée sur une modélisation simplifiée du chargement de surpression de type Friedlander associée à un modèle de coque élastique surbaissée de Donnell et à la croissance d'imperfections géométriques jusqu'à une taille critique a été développée. Son application permet la construction des courbes critiques de flambage des réservoirs métalliques dans des diagrammes pression-impulsion. La confrontation entre les résultats expérimentaux et les prédictions numériques souligne l'intérêt et la pertinence de la démarche proposée. / This work is a contribution to the assessment and prevention of potential domino effects caused by an explosion in an industrial area. It focuses specifically on the vulnerability of storage tanks subjected to a blast wave. Within this context, we investigate the effects of gaseous detonation. The objective is twofold: on one hand, to characterize the blast loading and, on the other, to predict the structural response. Three representative atmospheric chemical tanks were selected. Experiments were conducted at small scale satisfying Hopkinson-Cranz and structural scaling laws. A first campaign was performed on rigid instrumented cylinders to quantify the loading in terms of time and space pressure distribution. A second campaign was performed on flexible cylinders to quantify the structural response in terms of dynamic buckling damage. Simplified semi-analytical models were developed to provide damage predictions. They are based on a Friedlander pressure-time history pulse description, shallow-shell (Donnell's) equations and critical shape imperfection amplification thresholds. These models are used to construct pressure-impulse buckling damage diagrams for atmospheric tanks subjected to blast loading. The comparison between experimental results and numerical predictions highlights the interest and the relevance of the proposed approach.

Internationalization of industries, international competition, and the deregulation and harmonization within the EC are some of the important factors changing the prerequisites and requirements for effective international transport systems. Thus there is a need for more articulated knowledge regarding the internationalization processes of freight transport companies. In this longitudinal study the internationalization processes of three of the largest freight transport groups in Sweden have been analyzed in terms of changes in foreign representation: agents, join ventures, sales offices, subsidiaries. In addition, the relational effects of these changes on suppliers, owners, and competitors have been analysed. The results show that internationalization patterns change over time. First there was a gradual change of single relationships and systems. As an increasing number of transport companies became internationalized, there was a shift into more leapwise changes of nets of related companies. Furthermore the patterns and effects of internationalization differed between the three cases depending on when their internationalization process started. Typical sequences of changes have been identified to occur over time. As international interconnections between transport companies has increased these sequences have induced the development of domino effects in the network of transport companies. Based on the findings in this dissertation a dynamic model of internationalization is formulated. The model synthesizes the basic driving forces derived from network, distribution and internationalization research / Diss. Handelshögsk. : Stockholm

Les effets dominos ou cascade d’événements dans les industries et particulièrement dans les industries chimiques et de transformation, sont reconnus comme des scénarios d’accidents possibles depuis environ trois décennies. Ils représentent une préoccupation croissante, car ils ont le potentiel de provoquer des conséquences dévastatrices. L’effet domino, comme phénomène, est un sujet controversé lorsque son évaluation est nécessaire. L’examen de la bibliographie a démontré l’absence d’une définition commune et d’une procédure simple d’utilisation et précise pour son appréciation. C’est pourquoi l’un des objectifs de cette recherche est de formaliser les connaissances relatives aux effets dominos afin de comprendre les mécanismes de leurs occurrences. Pour ce faire nous avons étudié les paramètres à examiner pour déterminer la possibilité de cascade et être en mesure d’identifier les scénarios dominos. L’enjeu étant de permettre l’amélioration de la prévention du risque d’effet domino. L’autre objectif est donc de produire une méthode pour l’identification et l’analyse des effets dominos. Nous avons développé une méthodologie globale pour l’étude des effets dominos en chaîne initiés par des pertes de confinement. Elle permet l’identification et la hiérarchisation des chemins de propagation des accidents. Cette méthode facilite la prise de décision pour la prévention des effets dominos, tout en proposant un outil efficace et simple d’utilisation. Les résultats de l’étude sont fournis sous forme d’une hiérarchisation quantitative des équipements impliqués dans les scénarios dominos, en tenant compte des effets des conditions météorologiques et des mesures de maîtrise des risques existantes ou proposées.Cette hiérarchisation donne une idée claire des dangers que représentent les équipements par rapport aux accidents en cascade, en précisant si la dangerosité de l’équipement provient de sa capacité à initier ou à propager un effet de cascade.La méthode est basée sur une description topographique de la zone étudiée, incluant les caractéristiques de chaque équipement, et prend en compte les mesures de maîtrise des risques mises en œuvre par l’industriel. Elle repose sur deux phases principales : La première, est l’identification des chemins de propagation des accidents. Pour ce faire, la méthode d’analyse par arbre d’événements est utilisée. Les cibles potentielles sont déterminées en combinant les valeurs seuils d’escalade et les modèles de vulnérabilité (pour l’estimation de la probabilité d’endommagement). Cette première phase est implémentée sous MATLAB® et Visual Basic for Applications (VBA) afin de faciliter l’entrée des données, et l’analyse des résultats dans Microsoft Excel®. La deuxième phase est l’identification des équipements les plus dangereux vis-à-vis des effets dominos. Elle consiste à hiérarchiser les équipements impliqués dans les chemins de propagation, en fonction de leur vraisemblance à causer ou à propager un effet domino. L’algorithme qui effectue cette phase est codé sous VBA. La méthode a été conçue de façon à ce qu’elle puisse être utilisée sans qu’il soit nécessaire de s’appuyer sur les résultats des études de dangers. Néanmoins, si ces résultats sont disponibles, il est alors possible d’alléger certaines étapes de la méthode. Elle s’est révélée facile à utiliser, cela a été constaté lors de son application dans le cadre de projets et stages d’étudiants. / Domino effects or cascading events in the chemical and process industries are recognized as credible accident scenarios since three decades. They are raising a growing concern, as they have the potential to cause catastrophic consequences. Domino effect, as phenomenon, is still a controversial topic when coming to its assessment. There is still a poor agreement on the definition of domino effect and its assessment procedures. A number of different definitions and approaches are proposed in technical standards and in the scientific literature. Therefore, one of this research objectives is to formalize domino effects knowledges in order to comprehend their occurrence mechanisms. Thus, the parameters that should be looked at so as to understand the escalation possibility and in order to identify domino scenarios, were analyzed. The aim is to improve domino effect hazards prevention, through the development of a methodology for the identification and the analysis of domino effects.We developed a method for the analysis of domino accident chain caused by loss of containments. It allow the identification and prioritization of accident propagation paths. The method is user-friendly and help decision making regarding the prevention of cascading events. The final outcomes of the model are given in form of quantitative rankings of equipment involved in domino scenarios, taking into account the effect of meteorological conditions and safety barriers. The rankings give a clear idea of equipment hazard for initiating or continuing cascading events.The methodology is based on a topography of the industrial area of concern, including the characteristics of each unit and accounting for protection and mitigation barriers. It is based on two main stages. The first is the identification of accident propagation paths. For this, the event tree method is used. The possible targets are identified combining the escalation thresholds and vulnerability models (to estimate damage probability). This first stage was implemented using the MATLAB® software and Visual Basic for Applications (VBA) to enable an easy input procedure and output analysis in Microsoft Excel®.The second stage is the identification of the most dangerous equipment. It consists in prioritizing equipment involved in the propagation paths according to their likelihood to cause/propagate domino effect. The algorithm that performs this phase was coded in VBA.The method was designed so as it can be used without the need to rely on the results of safety reports. However, if such results are available, it is possible to lighten some steps of the method. It revealed easy to apply, this was confirmed through projects and student internships. / Gli effetti domino, in cui un primo incidente causa in cascata altri scenari incidentali, sono tragli scenari incidentali più severi che avvengono nell’industria chimica. Nonostante l’attenzioneche anche la normativa dedica a tali scenari, la valutazione dell’effetto domino è un soggettocontroverso. L’analisi della letteratura tecnica e scientifica ha mostrato l’assenza di unadefinizione comune di « effetto domino » e di una semplice procedura per l’identificazione ditali scenari. È per tale motivo che uno degli obiettivi di questo lavoro di ricerca è diformalizzare le conoscenze relative agli effetti domino al fine di meglio comprendere imeccanismi che possono provocarli. A tal proposito sono stati studiati i parametri necessariper determinare la possibilità dell’insorgere di cascate di eventi e per essere in grado diidentificare i possibili scenari incidentali dovuti ad effetto domino. L’obiettivo finale del lavoroè stato di sviluppare un metodo per l’identificazione e l’analisi quantitativa della propagazionedi incidenti primari nell’ambito di scenari dovuti ad effetto domino.E’ stata sviluppata una metodologia generale per l’analisi degli effetti domino causati daperdite di confinamento. Tale metodologia permette l’identificazione e la classificazione deipercorsi di propagazione degli incidenti. Tale metodo facilita inoltre la prevenzione deglieffetti domino, proponendo uno strumento efficace e semplice da utilizzare.I risultati di questo studio sono forniti in forma di una classificazione delle apparecchiaturecoinvolte in scenari dovuti ad effetto domino, tenendo conto degli effetti delle condizionimeteorologiche e delle misure esistenti per la gestione del rischio. Tale classificazione fornisceanche un chiara idea dei pericoli rappresentati dalle singole apparecchiature nel caso diincidenti in cascata, in quanto precisando se la pericolosità delle attrezzature proviene dallaloro capacità di innescare o propagare un reazione a catena.Il metodo è basato su una descrizione topografica del sito studiato, che comprende anche lecaratteristiche di ogni attrezzatura, che tiene conto delle misure di gestione dei rischi e dellebarriere di sicurezza presenti, basato su due fasi principali. La prima è l’identificazione deipercorsi di propagazione degli incidenti. A tale scopo è stato utilizzato un metodo basatoVIsull’albero degli eventi. I potenziali bersagli vengono determinati combinando i valori di sogliaper la propagazione degli eventi ed i modelli di vulnerabilità delle apparecchiature. Questaprima fase è implementata in MATLAB® e Visual Basic for Applications (VBA) in modo dafacilitare la gestione dei dati e l’analisi dei risultati in Microsoft Excel®.La seconda fase è l’identificazione delle apparecchiature più pericolose per gli effetti domino.Tale fase consiste nel classificare le apparecchiature coinvolte nei percorsi di propagazione infunzione della loro capacità di causare o propagare un effetto domino. L’algoritmo dedicato inquesta fase è eseguito su VBA.I risultati ottenuti anche nell’applicazione ad un caso di studio hanno evidenziato le potenzialitàdel metodo, che rappresenta un significativo progresso nell’analisi quantitativa dell’effetto domino.

While previous research shows that civil wars can spread to neighboring states, we do not know why certain neighbors are more at risk than others. To address this research gap, this dissertation proposes a contagion process approach that can identify the most likely targets of contagion effects from an ongoing conflict. Using data with global coverage, theoretical expectations about why and where civil wars would have contagion effects, are examined in a series of statistical analyses. Paper I argues and empirically supports that a country is more susceptible to contagion effects when it is characterized by ethnic polarization, where few ethnic groups form a delicate balance. Paper II argues and provides evidence that the involvement in conflict by an ethnic group in one country increases the likelihood of ethnic conflict erupting in a neighboring country that shares the same ethnic group. Paper III suggests and finds support that the arrival and long-term hosting of refugees from states in civil conflict make host states more likely to experience civil conflict. Paper IV examines the common notion that the granting of autonomy or independence to separatist groups may spur other ethnic groups to violently pursue similar demands, starting off a domino effect. Using new global data on such territorial concessions, the analysis does not support this version of the “domino theory,” which is popular among policy-makers. In sum, this dissertation contributes by demonstrating the usefulness of the contagion process approach. It offers a more comprehensive view of contagion among neighbors, and as such is able to specify arguments and intuitions in previous research.