Reliability Analysis And Design Methods Of Engineering Structures In Seismic Combination

Earthquake is one of the world’s three major scientific challenges. Earthquake hashappened frequently in recent years, while China is an earthquake-prone country. Thedesign aspects of structures need to find a balance that is based on probability betweensafety and economy. The method of approximate probability limit state designexpressed by partial factors is recognized internationally as the main one in reliabilitydesigns of engineering structures. However, it has been uncovered that the method islack of enough flexibility in practice. Yet, checking formula for strength in seismiccombination is expressed in this way. Once the formula needs to be carried on reliabilityanalysis, problems appear. Because the combination involves quasi-permanent value ofvariable actions, and combination of actions based on random variables has shortage intheory. Besides, once design working life differs from design reference period, themaximum distribution of seismic action has not been solved. This paper puts emphasison probabilistic combination of actions. A more reasonable reliability analysis model isestablished. The default of reliability control precision is uncovered by checking thepresent formula. By associating design value method, a flexible, practical and owninghigh precision formula is proposed to be used for checking for strength in seismiccombination. The research contents are as follows:1) The reliability analysis method of combination formula involvedquasi-permanent value of variable actions is analyzed, aiming at probability-basedPartial Factor Method in seismic combination. Ascertaining the maximum distributionof seismic action is short of theory in the case of design working life differing fromdesign reference period. 2) Performance function is established by means of combination of actions basedon stochastic process. Quasi-permanent sequence value of variable action correspondingto exceeded rates is proposed by associating stochastic process probabilistic model withthe maximum distribution of variable action in design reference period and sequencevalue distribution theory. Stochastic process probabilistic model of seismic action isestablished, and the maximum distribution of seismic action in any design working lifeis deduced by the achievement of probabilistic seismic hazard analysis and the statisticof seismic action in design reference period. Reliability analysis and design methods inseismic combination can be provided with theoretical knowledge.3) By choosing reinforced concrete beam having tension steel only, checkingformula for strength on seismic combination is carried on reliability analysis by meansof combination of actions based on stochastic process. Because absolute value ofreliability index is a small number under most effect proportion, the advice that targetreliability index is calculated by the average of failure probability is proposed. Becausereliability index may be a negative number, it will lead an unreasonable phenomenonthat uncertain factor of action performs a resistance’s role. In this case, the uncertainfactor of action should not be taken into account. Instability of reliability index underdifferent proportions of action effect indicates the shortage of present design methodabout reliability control precision. Reliability indexes of one single formula reducedramatically with the increase of design working life, which indicates that designworking life plays an important role in determining the values of actions.4) Design values are achieved by associating combination of actions based onstochastic process with design value method based on probability. In order to increasepracticality, the design values are expressed in the form of multiplying partial factor bycharacteristic value. A shortcut and effective method of optimizing sensitivity factors isproposed. Finally, a flexible, practical and owning high precision formula is establishedto be used for checking for strength in seismic combination. The variability of basicvariables and proportion of action effects are main controlling factors for the value ofstandardized sensitivity factors. The optimization that is based on50design workinglife and a suitable target reliability index can be used for different design working livesand different target reliability indexes, and the errors are all within the allowable range. Whether the uncertainty factor of action effects is considered or not, it has a greatimpact on the values of sensitivity factors, but has little effect on the final resistancevalue.