An examination of the relationship between creep and microstructural change in hardened cement paste
MetadataShow full item record
AbstractThis thesis concerns an examination of the relationship between creep and microstructural change in cement paste. The primary purpose was to examine the feasibility of this relationship as a basis for the development of a general equation for creep prediction. Experiments were performed in which miniature, saturated hardened cement paste specimens of two ages (28 days and 2.5 years) were loaded in compression for 6 days and subjected to various temperature cycles. A cycle comprised 2 days at elevated temperature (311, 328, 348, or 365K); room temperature (293K) served as a base line before and after a cycle. In some cases a temperature cycle was completed before loading; in others, a cycle was performed during the loading period and/or the specimen was loaded during a temperature cycle. Nitrogen sorption and polysilicate tests served to monitor changes in the specimen structure before, during and after the load testing. Identically treated but unloaded companion specimens were also tested. For the most part, the microstructure test results supported the concept that cement paste tends towards a more stable state (structure) under the influences of temperature and stress. In both young and mature specimens temperature pretreatment produces large increases in polysilicate and the mean pore size and significant decreases in surface area. These trends are consistent with a stability increase. Stress increases the polysilicate content in young specimens, while it produces an unusual de-polymerisation effect in mature specimens. The state of the microstructure and microstructural change have important influences on creep behaviour. Generally, the creep results and the correlation with microstructural tests indicated that specimen stability at the time of loading is an important influence on subsequent creep behaviour. Three components of creep, recoverable, irrecoverable and transitional thermal creep, were examined and found to be highly dependent upon specimen history. Unusual creep behaviour (especially in mature specimens) coupled with the observations that stress and temperature alter the microstructure, lead to the conclusion that microstructural change (stabilisation) occurring during the loading period also has important effects on creep. It appears that the interrelation between stability at the time of loading and stabilisation during load can affect creep in complicated ways. A rate theory analysis was undertaken to evaluate the potential value of a stabilisation-creep relationship. It was discovered that simple rate theory must be refined to account for the highly variable cement paste structure. The use of an exhaustion theory produced a fundamental, constant-temperature creep equation of the form e=C₁⸱1n(1+C₂⸱t); C₁ and C₂ are basic parameters dependent upon general aspects of the microstructure and the environmental variables, temperature and stress. The equation accurately predicted the creep behaviour of several examinations. The apparent fundamental relationship between creep and microstructural change, and the production of a fundamental and accurate creep equation based on this relationship suggests that more work along similar lines can produce a general equation for creep.
Bibliography: p. 354-371.