Study On Multiple Hazards Design Theory For Bridge Under Earthquakes And Heavy Truck Loads

Currently, many (American) bridge specifications, including LoadResistance Factor Design for bridge, are mainly based on single extreme load,and potential damage and losses may happen when the bridge suffers otherextreme loads during its design period due to the increasing varieties andfrequencies of extreme loads. Therefore, more studies and design onmulti-hazard subject for bridge should be placed on. This paper presents amethodology of multi-hazard limited states bridge design based on reliabilitytheory under earthquake load and heavy truck load, which is on the fashion ofLRFD bridge design and also as a part of MH-LRFD. Here, multiple hazardsincludes natural hazards (such as earthquake, scour, hurricane, tsunami etc.),man-made hazards (attack of terrorism, fire), and cascading hazards afterearthquake are incorporated, too.To meet the objectives of this study, authoritative achievements ofearthquake and (heavy) truck load (effects) probabilistic models are collected,which are modified and used here through parameters analyzing. Afterassumptions and analysis, columns of bridge substructure (usually small ormoderate) are the main load bearing members considering earthquake andtruck load. The equation of column resistance calculation is introduced afterthen.Comparing with other methods of reliability calculation, the level ofmethod used here is discussed. Onward, methods of reliability of bridgesystem are presented, and the slowness of which are analyzed. In order tosimplify the calculations, some assumptions, beams of superstructure couplingfirstly and then combining with column in some kind of failure mode, aresummarized. Result of the example calculation shows that the reliability indexof the bridge system is close to that of the column, which means that thecalculation of bridge system reliability will be convenient.Three analytical models, Ferry Borges–Castanheta model, Turkstra's ruleand Wen's load coincidence method, are reviewed and the pros and cons ofthem are analyzed. Then the rational parameters of Ferry Borges–Castanhetamodel are discussed in order to compare with the new model, a modified model of load combination, which is presented based on FerryBorges–Castanheta model, Turkstra's rule and Ghosn's model. The new model,which is based on a conditional probability model, considers respectively thatwhen earthquake occurs, trucks on a bridge combining with earthquake andearthquake alone to bridge and then the condition of earthquake can becombined together. One of the advantages of this model is that making thecombination more clearly, and reducing the mixture of sampling space, whichcan be concluded from example results.Based on the new model, partial failure probabilities, which are thepartial contributions to the whole failure, are brought about. The concept ofpartial failure can not only reflect the intensity of loads, also present theconception of combination optimization. Load factors of earthquake andheavy truck load are finally obtained after modifying the calculating methods,parameters analysis and increasing confined conditions.A typical bridge model is calculated to find the impacts of earthquakeparameter and combined parameter to load factors. Final load factors aresuggested through calculating and analyzing those in San Francisco, Seattleand New York.Finally, pushover analysis with SAP2000software is carried on to checkout the requirements of AASHTO LRFD in San Francisco, Seattle and NewYork. From the results, it is concluded that the suggested factors in this papermeets the requirements and also infers that earthquake design of bridge incurrent LRFD is conservative. The factor of earthquake can be properlyreduced and truck load should be considered, but not vague statements inLRFD.