Schriftenreihe der Zementindustrie, Heft 77: Assessing the autogenous shrinkage cracking propensity of concrete by means of the restrained ring test

Autogenous shrinkage is the major shrinkage component of concretes that contain much less water than would be required for complete hydration. This mainly applies to ultra-high strength concrete and, to a lesser extent, to high strength concrete. Both have particularly low water-cement ratios. The relative surplus of cement leads to an internal drying, irrespective of whether the concrete dries out to the ambient air or not. This process of so called selfdesiccation is associated with autogenous shrinkage which, if restrained, can lead to cracks, potentially impairing the in many respects outstanding durability of these kinds of concrete. Hence, to fully benefit from the advantages of high and ultra-high strength concrete, it is essential to minimize the risk of autogenous shrinkage cracking. Attempts to do so, however, require a reliable method for assessing this risk. Presently, there is no such method. Cracks are the result of relatively complex processes, in particular at early age as concrete properties change rapidly. A dependable assessment of the cracking risk requires comprehensive testing and a thorough understanding of the interacting parameters. Early age cracking in cementitious systems is not a new problem; cracking due to restrained drying shrinkage and thermal contraction has been examined at length. However, the investigation and prediction of stresses and cracks due to autogenous shrinkage brings about new challenges. The essential issue is the onset of stresses at very early age. This greatly increases the influence of creep and relaxation. Especially at stress levels close to failure this influence is highly non-linear and difficult to quantify, experimentally as well as mathematically. Another challenge is the fact that temperature strongly influences the autogenous shrinkage and, presumably, the cracking risk as well. From isothermal tests at different temperatures it appears that this influence cannot be accounted for by formulas conventionally used to describe the temperature dependency of cement hydration. The lack of clarity in this regard in part is a consequence of a series of methodological issues, most importantly the large number of different test methods and the difficulties in defining the onset of the autogenous shrinkage. The measurement of autogenous shrinkage, yet error-prone at constant temperatures, becomes particularly demanding at realistic temperature histories. The thermal deformations that inevitably superimpose the shrinkage strains are difficult to compensate for. At present there is no general agreement on how to measure the autogenous shrinkage under non-isothermal conditions. In brief, the current knowledge about the influence of creep and temperature on autogenous shrinkage, restraint stress and cracking is insufficient. Obviously the experimental methods need to be improved in order to overcome the existing deficiencies. The main aim of this study therefore is to contribute to this improvement. The experimental focus is put on tests of the autogenous shrinkage and on restrained ring tests. The common stress-strength failure criterion is used to assess the risk of cracking due to restrained autogenous shrinkage, or as it will be called herein, the 'autogenous shrinkage cracking propensity'. The strength is determined mainly by splitting tension tests. Special restrained ring tests are carried out to further examine the applicability of the chosen failure criterion. The potential of the comparably simple restrained ring test in quantifying creep is investigated. While the principal tests are conducted under quasi-isothermal conditions, the methodological analysis takes into account that tests of shrinkage and stresses under non-isothermal conditions will be required as well. Non-reinforced fine-grained ultra-high strength concrete is used as material. Steel fibers are disregarded to obtain a more undisturbed view on essential phenomena, and large aggregates are omitted to reduce the test efforts as well as to control thermal effects more easily. The autogenous shrinkage of the chosen reference concrete can be expected to be very high, making it particularly suitable. To provide for a wide range of results, concrete compositions include superabsorbent polymers, shrinkage-reducing admixture and ground-granulated blast furnace slag, representing common measures for reducing the autogenous shrinkage. The following chapters report the individual aspects and results of this study. In Chapter 2, following an introduction into the autogenous shrinkage and the primary research approaches, the major issues regarding its measurement are described. The present knowledge as to the influencing parameters is briefly depicted. The concretes used in the own investigations are presented in Chapter 3, including mechanical and some other properties. Development and validation of the new shrinkage cone method for measuring the autogenous shrinkage are comprehensively described in Chapter 4. The autogenous shrinkage of the concrete compositions investigated is given. Furthermore, the method's suitability for tests under non-isothermal conditions is outlined. The subsequent Chapter 5 summarizes the basics of stress development and cracking due to restrained autogenous shrinkage and provides explanations of the terms 'autogenous shrinkage cracking propensity' and 'very early age'. In Chapter 6 it is first explained why an available temperature-stress testing machine was not used in this study. Then the restrained ring test is analyzed as to its suitability for investigating the autogenous shrinkage cracking propensity. The methodological foundations of this method are given and the own investigations with the restrained ring test are presented. The work is summarized and conclusions are drawn in Chapter 7. The utilized literature can be found in Chapter 8 and additional data are annexed in Chapter 9.