Liquid storage tanks are used inmany fields such as water, oil, and gas industries. Many cases of tanks damageshave been observed due to past earthquakes. In order to sure the safety ofthese storage systems and avoid fires, explosions, and environmental pollution, the seismic behavior of these systemsshould be taken into consideration. Generally, the seismic response of astructure has flexible base may differ from the same structure supported on therigid base. This difference occurs due toeffects of the soil-structure interaction (SSI). Soil-structure and fluid-structure interaction are important for tanks in the estimation of the dynamic behavior of elevated tanks which are subjected to dynamicloads.

Therefore, it is important to consider the effects of interactioncorrectly for elevated tanks subjected to dynamic loads like earthquakes. Soil-structure interaction influences many aspects in the design of a structure likesafety, serviceability, and cost. In addition, the soil-structure interactioninfluences on the response of the structure during seismic events. Soilinfluences should be taken intoconsideration for different soil properties as substructure method to considersoil effects 1. Generally, elevated tanks are assumed to be fixed at thebase, hence, the attention is focused on the dynamic behavior of the fluid. Onthe other hand, the effects of soil on the dynamicbehavior of elevated tanks are also quite important. Many studies have beenmade to investigate the effects of soil-structure and fluid-structureinteraction.

Haroun and Ellaithy (1985) developed a model (elevated rigid tank)considering liquid sloshing modes. Resheidat and Sunna (1986) modeled arectangular elevated tank and studied its dynamic behavior during earthquakesconsidering soil-structure interaction. Resheidat and Sunna did not take intoconsideration the sloshing effects on the seismic behavior of the elevatedtank. Haroun and Temraz (1992) studied models consist of two-dimensional X-bracedelevated tanks supported on isolated footings to determine the effects ofdynamic interaction between the tank and the supporting soil foundation. Theyneglected the sloshing effects as well. It can be seen a few studies have beenmade to understand the effects of interactions on elevated tanks.

In thisstudy, a finite element elevated reinforced concrete tank is modeled and analyzed considering soil-structureinteraction. Fluid-structure interaction is out the scope therefore, it is notconsidered in this review. The structural data of elevated tank are describedin Table II. The elevated tank models were put on three different cases; soft,medium, and hard soil. Soil-structure interaction is represented by two typesof models. The first model is equivalentspring model. There are different methods to evaluate the dynamic stiffness andthe effective input motion of a foundation.

Thefirst method is modifying the fixed base of a structural system (Veletsosand Meek 1974). Springs and dashpots represent the elastic and viscoelasticproperties, respectively. The elastic and viscoelastic properties depend on thefrequency of excitation. The secondmethod is the direct method. This method consists of a finite element and boundary element methods or a mixture of them in the time or frequency domain(Wolf and Song 1996a, 1996b, Wolf 2003). Radiation damping occurs in an unbounded soil consists of a homogeneoushalf-space (Wolf 2002), which changes the unbonded soil function into a complexfunction. The third method is the substructure method.

This method considersthe frequency dependent or independent dynamic stiffness and the damping of thesoil foundation system. A cone model proposed by Wolf and Meek (1992,1993)represents an example of substructure method to calculate the dynamic stiffnessand the effective input motion of a foundation. In this method the structure isassumed to be supported on springs.

These springs represent the vertical,horizontal, rocking, and torsion stiffness of the soil. The stiffness of thespring is governed by the modulus of grade reaction of soil. The most commonmodel is the Winkler’s model.

Springs are assumed independent hence, the effectof the externally applied load becomes localized to the subgrade only to the point of applied loads. The raft is modeled asa thick plate and elastic modulus is provided to raft material. Sap2000software is used to model the water tank and provide springs in differentdirections at the base. The static stiffness values of rigid circularfoundation are provided in form of spring as given in Table III, G: shearmodulus, r0: radius of circular foundation, v: Poisson ratio.

Second model is equivalentelastic solid model. In this model, soil mass is modeled as solid elements,each solid element has eight nods with three degrees of freedom. Size ofelements should be between 1-1.5. Different type of soil properties is appliedon solid elements, especially modulus of elasticity and Poisson ratio. The width and depth odsoil is taken as 5 and 3 times of the water tank base dimension, respectively. Atthe base of soil model fixed supports are provided and at periphery vertical rollersare provided.

Both models are analyzed and discussed as well as results are comparedby using Sap2000 structural software. Inconclusion, it can be seen from the results that method of modeling depends on soilconfigurations. It means, if there is hard soil under the structure, it’s usefulto use spring model instead of solid element model because solid element model consumeslots of time for modeling and analysis. But, for medium or soft soils, it’s favorableto use solid element model.