An experimental investigation into superposition of creep strains for concrete.

Abstract

This experimental investigation was performed to examine the validity of the hypothesis of superposition of creep strains by measuring and comparing the time dependent strains of concrete prisms subjected to increasing and decreasing loads, constant loads, and late loads (for unsealed concrete specimens which were dried before loading and sealed concrete specimens kept in a moist room). Three series of specimens were made; one was moist cured for four days, then sealed and the other two series of specimens were moist cured for three and four days respectively, then transferred to a 50% relative humidity room. The conclusions must be read noting that the use of sixteen prism specimens limits the number of specimens available at second and third load events. Use of a single specimen for either shrinkage or compliance measurement is not desirable. For increasing loads under sealed conditions, it is noticeable that at the second load cycle superposition closely approximates the experimental results. There is a 30% underestimation for the third load cycle. For creep recovery of sealed specimens superposition, considering the calculations involved approximates the experimental results at both 20 days and 87 days. Two series of specimens exposed to drying were prepared for this phase of the experimental investigation. For drying before loading the results indicate that superposition underestimates the measured strains by 5--10% for increasing loads, experimental strains are greater than calculated strains. Creep recovery was overestimated by 10--12%. It would appear that the application of the superposition rule is valid for basic creep but is less valid for drying creep. Therefore an alternative hypothesis is necessary to predict with more precision the creep strains for concrete subjected to increasing or decreasing loads for drying before loading and sealed conditions. The influence of compressive load shows very little effect on weight loss compared to specimens without load. These results would indicate that the moisture loss depends only on relative humidity and is not influenced by loading i.e. the seepage theory does not hold.