For creep analysis of large structural systems, the linear aging integral-type creep law needs to be converted to a rate-type form which consists of a system of first-order linear differential equations with age-dependent coefficients. The system may be visualized by the Kelvin chain model with agedependent elastic moduli and viscosities. In the existing formulation, the independent variable is actual time. It is shown that, by using as the independent variable a certain reduced time which increases with time at a gradually declining rate, one reduces the number of differential equations needed to describe creep within the given time range, thereby making numerical structural analysis more efficient. An algorithm for identifying the material parameters from given creep data, based on minimization of a sum of squared deviations, is also presented. Thermodynamic restrictions on the material coefficients are analyzed. Finally, the capability of closely approximating available creep data is demonstrated.
INTRODUCTIONFor numerical creep analysis of large structural systems, particularly finite element analysis, the rate-type form of a linear aging creep law is far more efficient than the integral-type form based directly on the principle of superposition. Instead of storing during step-by-step computation the values of stresses or strains in all previous steps, it suffices to store only the current values of several internal variables. The number of arithmetic operations also gets significantly reduced (3,5,6,10,11,19).The existing rate-type formulation of the aging creep law of concrete consists of a system of first or second-order linear differential equations with the actual time as an independent variable. This paper examines the question whether the rate-type formulation can be made more efficient by means of introducing as the independent variable a certain monotonically increasing function of time, called the reduced time.Linear Aging Creep Law Based on Principle of Superposition.-For stresses less than about one-half of the strength, the creep law of concrete may be approximately considered as linear, i.e., obeying the principle of superposition. The linearity assumption greatly simplifies structural analysis, and is generally used in practice. According to the principle of superposition, the uniaxial creep law of concrete may be written in the form Vol. 110, No.3, March, 1984. ©ASCE, ISSN 0733-9399/84/0003-0329/$01.00. Paper No. 18621. 329 in which CT, E = uniaxial stress and strain; EO = prescribed stress-independent strain representing shrinkage and thermal dilatation; and J (t, It may now be checked by substitution into Eqs: ~ and 1 that the ~ol lowing linear differential equations are always satisfied by the precedmg expression for ElL:dy~ dYIL CIL(t) dYIL Expressing the derivative d'YlL/dYIL from Eq. 5, we may verify that functions "I IL always satisfy the differential equationVariables EILand "I IL are related to E by the differential equations
. (8 )It can be further shown that int...