Substantiation of Reference Method For Determining Concrete's Freeze-Thaw Resistance

. It has been analytically proved that using concrete’s rate of set ε as a measure of damage, instead of decreasing of tensile strength R, increases freeze-thaw resistance’s accuracy of estimation a lot under otherwise equal conditions by the time of freeze-thaw cycling. Also it has been experimentally shown that ratio of relative decreasing R to ε in direction, perpendicular to compression, is assumed to be independent on values R and ε for a given concrete and on the ways of achieving them during mechanical or freeze-thaw cycling. Taking this into account patented methods for estimation of concrete’s freeze-thaw resistance as per values R and ε received after freezing and thawing cycles of some specimens and their postliminary failure by linear compression was substantiated.


Introduction
Freeze-thaw resistance of concrete is an ability of water-saturated concrete specimen to maintain repeated standard thermo cycles without noticeable damage. Different types of water pressure cause concrete's freeze-thaw deterioration, such as hydraulic and osmotic pressure [1], capillary pressure [2] and other types of water influence according to existing freeze-thaw resistance theory. Decreasing strenght of construction material is associated with the processes of its water freezing (for example, rock [3]). Water gets into structure of porous bodies, separates particles and breaks coupling between them [4]. Porosity of material is determinative factor for frost resistance and durability subsequently [5,6]. So, strength of concrete could be pictured as a porosity function [7]. In order to determine the concrete mix composition, it is necessary to take into account freeze-thaw resistance. Worldwide experience offers some test methods to determine durability of concrete by freeze-thaw damage, such as Slab test [8], CDF [9], CIF-Test [10] and Cube-Test [11]. But they differ in terms of their procedures and conditions and are referenced to particular conditions. There are two different standard types of methods of determining the freeze-thaw resistance of concrete-basic [12] and reference [13] in the Russian Federation. Considerable spread of values of strength for concrete specimens (variation coefficient ρ≈17%) in constant conditions of preparing and testing determines random spread of choosing average values R in the range of where n is the volume of bath of specimens. In this case for proving the significance of relative decreasing R on 0,05…0,15 it is necessary to test more that 50 specimens and support R R R 068 . 0 /   , when confidence figure is equal 0,95. Therefore, main disadvantages of the basic method are high labor input and small operability by virtue of the fact that duration of the basic thermo cycle is equal 4,5 at least and F >> 50. One of the existing reference methods is a Dilatometric rapid method of determining the freeze-thaw resistance of concrete [13]. This method is a prototype for the method, which has been offered by me.
In this method concrete's freeze-thaw resistance is determined by the maximum relative difference of volume deformations of the tested concrete and standard specimens in accordance with tables provided in standard specification [12] taking into account concrete's type, its form and the size of specimens. However, the results from the tables provided in state standard specification are acceptable only for Portland cement concrete and slag Portland cement concrete without surface-active additives (PEAHENS), such concretes are used extremely seldom now. Now lot of new concretes are investigated, tested and used, for example, nanomodificated concrete [14], high-strength concrete [15], concrete on the basis of fine-grained dry powder mixes [16], concrete with using recycling concrete aggregates [17] etc. In order to obtain new tables long labour-consuming experiences, which imply using, basic methods are needed [18]

Project's objectives
The aims of this paper are to obtain the dependence between relative decreasing in strength by rate of strain and substantiate the new method for determination of concrete frost resistance. Using of measure of damage D by the time of freeze-thaw cycling is offered for removal of disadvantages. In accordance with experimental data [19,20] are maximal relative measurement errors of determination original value R , decreasing R after c thermo cycles or values q correspondingly. By the time of using non-destructive method [ After differentiation (2): After plugging (5) in (3): . ln The result proved that contribution of measurement error of q in F F /  is more in 8 times than contribution of measurement error of R .
It's the case, when measure of damage is a decreasing of strength. To decide this problem using of rate of strain , where l ≈ 100mmbase length, δl=l c -lthe base deformation and l cbase length after "c" thermo cycles, is offered. But for that it is necessary to determine dependence between decreasing in strength and rate of strain. So, this value is defined as

3.
To determine this dependence concrete mix was tested. Mix contained a portland cement (12%) of brand 400, sand (25%), granite crushed stone 5 … 20 mm (56%) and water (7%). 108 samples cubes with an edge of 150 mm were prepared. There points of intersection of diagonals of opposite sides were spaced far apart from each other (to 1,5 mm). Samples hardened 28 days in water at the room temperature, and then 60 days in damp sand at 18 … 26 °C. From these 108 samples, 8 samples were cycled alternately thermally and mechanically, and 100 samples were used for a frost resistance assessment by a basic method. At realization of a basic method l distance between points of each sample was measured by means of a tool microscope before tests for frost resistance or for durability. Rate of strain was counted as . Changing of distance between points at realization of a method [22] was carried out on 20±2 °C before and after cycling by means of a bracket with variable base and a measuring head of hour type (the price of division of 1 micron). Threshold loading of a watersaturated sample was determined by a way [23], registering the acoustic issue (AI) by means of the AF-15 device at cyclic loading and unloading of a sample to zero. In the first experience load of L was brought to 11 t; in the absence of acoustic issue in the course of the end of unloading value of L was increased by 5% and so until at the end of unloading there was no AE. For accepted an average of L two last cycles. The limit of long durability was found as the relation to the average area of two loaded sides. Dependence z was determined as (L 1 -L 2 )/ L 1 e , where L 1 is a maximal value L in the first loading and L 2in the second loading ( fig. 1).
Results of this experiment are given in the Table 1. Dependence z is near to linear.