Abstract
•ACI 209 and B3 models showed good shrinkage prediction in specimens cured at 28°C and 50% R.H.•Best predictions at 50°C and 5% R.H. were by GL2000 in standard specimens and CEB-FIP in plain prisms.•Effect of size on shrinkage strains increases as curing temperature increases.•ACI 209 and Sakata models show good prediction of specimen size effects at 28°C and 50% R.H.•Inaccurate prediction is shown by assuming constant size effects for different drying conditions.
There are several models developed to predict shrinkage strains in concrete. These models are adopted by various codes or suggested by prominent researchers. The results of experimental measurements of shrinkage strains in plain concrete are compared to the theoretical shrinkage strains predicted using five available models which are: ACI 209, CEB-FIP, B3, Sakata, and GL2000. Measurements were taken on standard specimens (50×50×300mm) and plain prisms (80×150×500mm) subjected to moderate curing condition (28°C and 50% R.H.) and a harsh curing condition (50°C and 5% R.H.). ACI 209 and B3 models were found to have good prediction of experimental shrinkage strains in specimens cured in humidity room while GL2000 and Sakata models showed poor approximations. The best approximation of experimental shrinkage in standard specimens that are cured in temperature room are shown by GL2000 model while CEB-FIP model shows the best prediction in plain concrete prisms that are cured in temperature room. The effect of the considered specimen sizes on shrinkage strains is moderate when cured in humidity room while the effect is noticeable in specimens cured in temperature room. ACI 209 and Sakata models show good quantifications in specimens cured in humidity room and the predicted effect of specimen size in temperature room is far from the experimental effects for all models. All models assume constant and similar size effects on shrinkage for all type of drying conditions which cause inaccurate prediction.