An algorithm to compute the parameters of a Generalized Kelvin Chain model to represent aging creep of concrete

Resumo

The numerical modelling of viscoelastic behavior of large concrete structures can be optimized by representing the compliance function in terms of series of exponentials, usually termed Dirichlet series, Prony series, or Generalized Kelvin Chain mechanical model. The viscoelastic behavior of concrete is usually described
in terms of a compliance function obtained from classical creep tests or design codes, being, thus, necessary to devise a workflow to derive the set of retardation times and moduli that completely describe the correspondent Generalized Kelvin Chain model. Direct fitting of the Generalized Kelvin Chain model to a given compliance function is usually not feasible, as the problem becomes ill-posed if the chain contains a reasonable number of units, which is necessary for correct representing the viscoelastic behavior. The present work presents an open-
source Python algorithm to compute the parameters of a Generalized Kelvin Chain model associated to a set of aging compliance functions. The use of the algorithm is exemplified by computing the Generalized Kelvin Chain model associated to compliance functions obtained from Eurocode 2 and fib Model Code 2010, which are common viscoelastic models used in engineering practice.