The effects of high temperature heating on the gas permeability and porosity of a cementitious material

Abstract

This experimental study investigates variations in the transfer properties (permeability) and porosity, of a normalized mortar heated up-to 600 °C. New experimental techniques allow simultaneous measurements of permeability and porosity variations, under confining pressure. The main results show that when temperature raises from 105 °C to 600 °C, the permeability increases by 2 to 3 orders of magnitude, and nearly 50% in relative porosity (from 17% to 24% in absolute value). The higher the temperature, the greater the sensitivity to confinement of the permeability (and porosity), thus indicating substantial cracking and the irreversible nature of the material's behaviour. The application of confining pressure leads to irreversible crack closure and crushing of the pores. Various poro-mechanical measurements also revealed a considerable loss in stiffness for strongly heated material, providing complementary information concerning the porosity variation under loading, the amplitude of which is too high to be due to crack closure alone.

Introduction

This experimental study was designed to allow us to understand and analyse the effects of de-structuration in a cementitious material, in the present case a normal mortar, when it is subjected to high temperatures. The study focuses on two of the material's transfer properties: its gas permeability, and its porosity measured with a gas, as well as the variations in these properties following a heating phase up to 600 °C, followed by a slow phase involving cooling down to room temperature. As far as the measurements are concerned, their originality stems from the fact that we have developed techniques allowing permeability and porosity to be measured, by injecting a gas under hydrostatic loading (i.e. confining pressure). In fact, the observed deterioration occurs in several stages: decomposition of the AFm/AFt [1], [2], [3], [4], progressive decomposition of the C-S-H over a broad range of temperatures [1], [5], [6], [7], [8], and decomposition of the Portlandite at approximately 450 °C [6], [8]. The decomposition of C-S-H is accompanied by shrinkage [9], as a consequence of the loss of chemically bounded water. There are thus two distinct phases, swelling then shrinkage [9], [10], [11], which are quite distinct from the continuous swelling of granulates [12], [13]. All of these phenomena are known to cause strong micro-cracking and progressive de-structuration of the material, modifying its mechanical [6], [14], [15], [16], [17] and transfer properties [18], [19], [20], [21]. The decision to focus on transfer properties was related to their recognised importance, with respect to an evaluation of the material's durability, and more importantly their ability to very accurately reveal its state of deterioration. The use of confining pressure makes it possible to adjust the crack opening and closing mechanisms, which have a strong impact on gas permeability [22]. The measured variations in these properties as a function of loading (or strain), confining pressure in particular, have made it possible to progress towards a full understanding of the (micro-) structural deterioration produced by different forms of stress (thermal, chemical, mechanical, …). These variations also make it possible to ascertain their relationship with the material's mechanical performance. The observed interactions between the material's transfer properties and mechanical behaviour appear to be highly complex. In many real-life situations (nuclear reactors, radioactive waste storage, reservoirs, tunnels …), whereas the transfer properties of building materials are essential characteristics, the mechanical resistance of large structures is rarely endangered or defective. This is however not the case for the material's air-tightness. For this reason, we consider that defining the material's state of deterioration (rather than its damage, which has a highly “mechanical” connotation), can be achieved by directly measuring the permeability, which can be related to the material's deformation and mechanical history. Mechanical damage, in the sense that it is generally associated with cracking, will obviously have an influence on the material's permeability. However, cracking is in fact just one of the many causes of deterioration, which is also related to the degree of crack opening, such that it does not provide a sufficiently accurate description to allow correct “modelling” of variations in the material's transfer properties. The high levels of confining pressure (up to 40 MPa) applied to the material make it possible to obtain very useful information concerning its variations in porosity under loading, which appear to be too high to be accounted for by the effects of crack closure alone, thus suggesting the (initiation of) collapse of the porous network.

Many studies dealing with the impact of temperature on the decomposition of hydrates [5], [6], [8], [23], and its effect on mechanical properties [15], [24], [25], can be found in the bibliography. An exhaustive compilation is also provided in [17], [26]. On the other hand, only a small number of experimental studies can be found on the combined effects of stress and permeability and/or porosity [13], [19], [27], following intense heating. Previous laboratory studies, carried out by Chen [20], [28], on an equivalent mortar heated to only 400 °C revealed a clear reduction in the material's bulk modulus and a strongly irreversible behaviour under confining pressure. This irreversibility was mainly related to the definitive closure of cracks. Heating below 450 °C did not allow the effects of Portlandite decomposition to be observed [5]. In this new study, which is more comprehensive with respect to the transfer properties, the material was heated to 500 °C, and then to 600 °C, which made it possible to determine whether the decomposition of Portlandite has a notable influence on these properties, since this is often considered to have an obvious impact on its resistance [29]. Similarly, we introduced a small number of poro-mechanical elements by measuring the material's volumetric strain under confining pressure, and by comparing this with its variation in porosity. This aspect provides new insight into the material's deterioration, and modifications to its porous network.