CA2322931A1 - Multi-probe conductivity method for monitoring time-dependent processes in fresh cementitious and other dense slurry systems - Google Patents

Abstract

ABSTRACT
The invention provides a novel probe comprising multiple conductivity electrodes, which can be used to monitor various types of time-dependent processes in aqueous suspension, slurries, pastes, etc. for example, fresh cementitious materials, such as grouts, mortars and concrete. The processes which can be monitored as function of time through this novel method can be either "physical" or "chemical";
physical-type processes comprise settling, sedimentation, segregation, water migration, surface bleeding, channelling, etc.; chemical-type processes comprise dissolution or precipitation of solids, and any reactions involving the solution and solid phases present in the slurry.

Description

MULTI-PROBE CONDUCTIVITY METHOD FOR MONITORING TIME-DEPENDENT
PROCESSES IN FRESH CEMENTITIOUS AND OTHER DENSE SLURRY SYSTEMS
FIELD OF THE INVENTION
The present invention relates to the development of novel devices and methods for in-situ, non-destructive, continuous and quantitative measurement of changes occurring in aqueous-based suspensions, slurries, pastes, sludge and other colloidal systems.
BACKGROUND OF THE INVENTION
The basis and significance of the present invention is best described for systems in which both physical-type and chemical-type processes are present simultaneously, both playing an important role in the evolution of the slurry system. This is precisely the situation in fresh cementitious materials, in which, both, physical effects (i.e., migration of solid particles (cement, sand, aggregate, pigment, or other) and solution phase) and chemical effects (i.e., dissolution or hydrolysis of the reactive minerals, precipitation and growth of the hydrate products) can drastically affect the physical properties of the hardened material. The invention will therefore be described and illustrated mainly as it applies to cementitious systems; similar application to other reactive, or non-reactive slurries is obvious to anyone in the field of colloid chemistry or process engineering.
Physical Effects In cementitious systems, particularly those which are highly fluid (e.g., flowable, pumpable, or self-levelling concrete), segregation of the various solid materials may occur as a result of an upward migration of fines and sedimentation of the aggregates.
Bleeding may then result from the movement of part of the "free" water in the fresh mixtures to the surface, due to the inability of the solid constituents of the mix to hold all of the mixing water when they settle downwards (1 ). Migration of interstitial solution can lead to surface bleeding and to "channelling", the latter resulting from preferential migration paths through the cement paste. Whether these phenomena are sufficiently pronounced to be apparent on the external surfaces of the material, or not, the occurrence of segregation and bleeding results in structural heterogeneity which reduces the performance characteristics of the material, including surface finish, strength, impermeability, and durability. This can also affect the characteristics of the interface of the paste with the reinforcing elements, reducing bond strength (2-4).
In spite of the critical importance of bleeding and segregation effects, there exists no direct experimental method to evaluate such effects quantitatively. The extent of surface bleeding is usually measured by collecting the excess surface solution as a function of time after placement (ASTM-C232, CRD-C9). Bleeding can also be evaluated by determining the surface settlement (or subsidence) per unit height of concrete (5-6).
On the other hand, the degree of segregation occurring in a cementitious material is determined by analysis of the distribution of coarse aggregate in the fresh state or after the hardening. For example, the testing of the spread of a pile of concrete subjected to jolting (ASTM C124) gives an indication of the consistency of the concrete and its tendency to undergo segregation during the flow. Similarly, the susceptibility of fresh concrete towards coarse aggregate separation can be assessed by observing the material scattering following a drop over a cone from two hoppers (7). The observation of the distribution of coarse aggregate in a cored sample offers another means of determining segregation in hardened concrete (6).
Chemical Effects In chemically reacting slurries, such as cementitious systems, the chemical processes occurring in the system, also have important consequences on the physical properties of the hardened material. In cement-based materials, the dissolution/hydration of the reactive components play a crucial role in determining how the setting reactions are initiated and how the microstructure evolves to ensure development of the early-age strength of the material. These chemical-type processes, occurring simultaneously with the physical effects described above, clearly have a marked influence on the way in which the stability and the structure of the slurry evolve in time, and ultimately on the properties of the hardened material.

In the case of cement-based materials, the time-dependence of the chemical processes has been successfully monitored using electrical conductivity (8 to 15). Since the dissolution/hydration reactions involve ionic solids and soluble electrolytes, the chemical processes can be monitored continuously through measurement of the electrical conductivity of the fresh material. Typically, the electrical conductivity of a cement-based slurry (paste) will initially increase as the soluble alkali salts quickly dissolve into the mixing water; the conductivity continues to increase as the aluminate and silicate phases of the cement react with water producing calcium and hydroxyl ions. The paste conductivity will begin to decrease as the hydrate products (particularly portlandite) begins to precipitate, and will continue on decreasing as the microstructure of the hardening cement matrix develops (10, 13, 14 ) The Need for Monitoring Methods Its is therefore apparent that, in cementitious systems, there is a need for a measuring device and method which could provide real-time information on the evolution of the systems through its initial consolidation and early age behaviour. Since this consolidation and early age behaviour depend simultaneously on physical-type and chemical-type effects, the required method must be sufficiently incisive to measure the evolution of the system properties as affected by both types of effects. To comply ideally with the usual constraints of concrete applications, the method must allow non-destructive, in-situ, real-time measurement; moreover, the method must use devices which will not interfere with the processes monitored and will cause minimal disturbance when they remain in the hardened material. Since no such device or method currently exist, it is the purpose of the present invention to provide novel probe and application conditions to achieve the desired type of measurement and monitoring in fresh cement-based material, and in any aqueous slurries and colloidal system, as well .
For application in other aqueous-based slurries, the constraints may be less severe, particularly in systems which exhibit no chemical reactivity between the various components and phases present. However, the principle and application of the probe and method of the present invention are valid for all such aqueous-based slurries.

PRIOR ART
There exist few references regarding the use of electrical conductivity in studies on cement-based materials and other aqueous col loidal systems (8 to 15 and 17 to 21 ).
In fresh cementitious systems, typically, the variations in electrical conductivity have been used to monitor time-dependent changes in the composition of the solution phase. During the last decade, there has been a growing interest in the development and use of electrical response techniques in cement and concrete research (9).
The electrical measurement technique has already proven successful in studies of ion exchange resins, soils, ion exchange membranes and polyelectrolytes (15, 16).
Previous work in the area of cement technology has focussed on determining the relationship between the evolution in the solution conductivity and the hydration processes which lead to the precipitation of portlandite (calcium hydroxide) in the slurry, the latter effect associated with the beginning of the setting (stiffening and hardening) processes in the cement paste.
In hardened cement materials, conductivity measurements are often used for evaluating the mobility of ions through the matrix, for example conductivity of the hardened concrete due to mobility of ions between two electrodes with a given potential (ASTM, C 1202). The results are related to the permeability of the material and hence to its durability. Recently, it has been reported that rapid estimation of water-cementitious ratio and chloride ion diffusivity in hardened and plastic concrete could be made by using electrical resistivity measurement, whereby, the concentration of the various electrolytes in the interstitial water of the cement paste is dependent on the initial water content that affects the conductivity (12).
References are there to the use of electrical conductivity in monitoring or processing various other types of aqueous slurries and colloidal systems.
Salient features of some of them are briefly mentioned below.
A new in-situ technique has been developed for hexavalent chromium removal from sand by imposing of constant electrical potential gradient across the soil matrix through graphite cathode and iron anode (17). In another reference, laboratory experiments and mathematical modelling have been used to study the changes in the flows of ions and pore liquid during the process. These flows were directly related to the removal of charged and uncharged contaminants by electromigration and electro-osmosis (18). The use of a low -frequency square wave alternating current was made in studying resistivity characteristics of compacted clayas. this method avoids difficulties due to electrode polarization and reduces capacitative and lead inductive effects to 5 minimum (19). Effects of water content, orientation of particles, electrolyte concentrations, type of electrolytes, have been studied for the electrial response characteristics of soil-water structure. Furthermore, effects of temperatures, and nature of surfaces have also been reported regarding soil-liquid system (20).
Electrocoagulation of bio-organic impurities in waste waters from biochemical processes has been reported as an important industrial purification process (21 ).
A brief account of patents of interest to the present invention are given, as fol lows:
- Patent No:4, 176,038 (Nov. 27, 1979, USA). The patent pertains to, "Water purification method and apparatus". The process comprises of passing the liquid between spaced electrode plates in the presence of a fluidized bed of conductive particles. The liquid suspension is subjected to an alternating electrical field applied across the electrodes through conductive particles of the said bed. Under such a system, suspending forces of the solids are rapidly and efficiently broken. The agglomerated solids may then be separated from the liquid.
- Patent 5572123 ( Nov. 5, 1996, USA). This patent deals with an apparatus and a method for on-line inspection of electrically conductive food products using liquid electrolyte. Changes in electrical conductivity are measured on-line to correlate the quality of the food product in the production.
- Patent 548949 ( Feb. 6, 1996, USA). This patent deals with, "High accuracy calibration-free electrical parameter measurements using differential measurement with respect to immersion depth". The apparatus is used to measure electrical parameters of a medium, such as electrical conductivity and dielectric constant, between a pair of electrodes. By obtaining a differential conductance measurement with respect to the immersion depth, the effects of fringe conductance are eliminated from the measurement.
- Patent 5458747 (Oct. 17, 1995, USA). This patent deals with, "In-situ bio-electrokinetic remediation of contaminated soils containing hazardous mixed wastes". It involves the basic cation and anion movements towards respective energized electrodes. The pore fluid is moved from the anode area and collects in the cathode area and may pool at the soil surface. The technique is dependent on the mobility of the ions and the conductivity of the medium that plays an important role.
SUMMARY OF THE INVENTION
The present invention relates to the development of novel devices and methods for in-situ, non-destructive, continuous and quantitative measurement of changes occurring in aqueous-based suspensions, slurries, pastes, sludge and other colloidal systems. The devices consist of multiple electrode (multi-electrode) conductivity probes which can measure, as function of time, variations of electrical conductivity at different positions (vertically or horizontally) in the colloidal system. The localised changes in the electrical conductivity in the slurry, which result from physical processes or chemical reactions occurring in this slurry, yield rate parameters for such processes as sedimentation, settling, bleeding, de-watering, etc. When used in reactive colloidal systems, such as fresh cement-based materials and hydraulic binders, the device and measurement method yield a quantitative measure of bleeding and segregation effects, as well as data on the beginning of the hardening reaction and hardening rate at early age (e.g., 0-72 hrs).
The principle of the method is as follows: any change in composition (solids or solution) occurring in a volume element of an electrically conducting slurry, will result in a change of electrical conductivity of this volume element. Hence, simultaneous measurement of the electrical conductivity at numerous point locations in a conducting slurry can be used to follow the evolution of physical or chemical processes in this slurry.
One important application of the "Multi-probe Conductivity Method" is for monitoring the evolution, as function of time, of fresh cementitious systems, grouts, mortars and concrete. In such systems, the Method allows continuous-real time monitoring of the following processes:
1. The migration of liquids and solids in the fresh material, leading to sedimentation, segregation and bleeding 2. Rate of hydration of the reactive materials in the cementious system and the onset of the setting (hardening) process 3. The rate of hardening and strength development at early age, typically, 1 to 4 days.
The Method of this invention is non destructive and can be applied "in-situ"
for monitoring the time-dependent evolution of virtually any aqueous slurry, for example, waste water, mine tailings, industrial waste effluents or sludge and industrial process slurries.
DETAILED DESCRIPTION OF THE INVENTION
Principles The method devised for in-situ, non-destructive, continuous and quantitative measurement of changes occurring in aqueous-based colloidal systems is based on changes in the electrical conductivity which occur as a result of composition variations in a given volume element of a system. For example, the electrical conductivity measured at different depths in an aqueous slurry and as function of time will reflect both, the physical and chemical processes occurring in this slurry. Physical processes, such as sedimentation or settling, will result in increasing the solids content of the slurry (volume or weight fraction) as function of depth; the increased solids content will be reflected by a proportional decrease in the electrical conductivity. Chemical processes, such as dissolution or precipitation of ionic solids, also induce changes in the electrical conductivity; if such a process occurs homogeneously throughout the solution phase, its time dependence will be reflected in the evolution of the average conductivity measured over the entire sample. This novel method can be used in any aqueous-based colloidal system, typically, suspensions, emulsions, slurries, pastes, sludge, or other. It was found particularly useful in monitoring the time evolution of fresh cement-based systems, such as grouts, mortars and concrete, wherein both physical-type and chemical-type processes occur and have an important impact on the final properties of the hardened materials. The potential of the novel multi-electrode conductivity measurements in fresh cementitious systems is readily visualised from the following qualitative description of the phenomena occurring in fresh-cement-based systems.
The evolution with time of the average conductivity of a fresh portland cement paste is illustrated schematically in Fig. 1a. The conductivity value measured in the first instants following the mixing of the cement and water is low (point A in Fig.
1 a), but increases rapidly (point B) due to the rapid dissolution of alkali sulphates, increasing the concentration of the Na+, K+, and S04-2 ions, and the early surface hydration of the most reactive mineral phases in the cement, generating Ca+2 and OH- ions.
During the dormant period (B to C), the conductivity increases slowly as the hydration reactions proceed, further increasing the Ca+2 and OH- solution concentration into supersaturation (point C). When the conditions for Ca(OH)2 precipitation become favourable, a sharp decrease in the Ca+2 and OH- concentration will take place, initiating a rapid decrease in conductivity which signals the onset of setting. As setting and hardening occur, the conductivity decrease continuously (C-D) due to the development of the pore structure which greatly limits the mobility of electrolytes in the interstitial solution (C-D).
The electrical conductivity measurement (single point or average) performed on a fresh cement paste as function of time can thus yield key information on the hydration reactions and on the setting process. The use of conductivity measurements in monitoring the evolution of the solution phase and the setting process in cement-based system had already been discussed by Vernet ( 13, 14) and others (8 to 10), as noted in the "Prior Art" section.
The novel multi-electrode probe of the present invention allows the measurement of the electrical conductivity of the fresh cement-based material at different positions in the material volume examined. Such measurements performed simultaneously (in rapid sequence) as function of time yield results such as illustrated schematically in Fig. 1 b (further detailed in the "RESULTS" section). The latter shows qualitative trends which may be observed in the conductivity measurements performed with three pairs of electrodes placed at different depths in the sample (for example, top, middle and bottom). In this case, the shape of each conductivity-vs-time curve is dependent first, on chemical processes as described above, and second, on physical processes, which may also occur simultaneously in the material, typically sedimentation, segregation and bleeding. In the hypothetical situation considered in Fig 1 a, the "Top" curve exhibits high excess conductivity, relative to the other two curves (or relative to the average conductivity curve); the reverse is seen with the bottom electrode pair, which shows initial decrease in conductivity, and significantly lower conductivity compared with the other two pairs of electrodes over the entire measurement period. These conductivity trends reflects a concomitant migration of the conducting solution phase upwards (increasing the conductivity in the upper portion of the sample) and a sedimentation of some of the solids (decreasing the conductivity in the lower part of the sample volume). Hence, simultaneous measurement of conductivity at several points in a fresh cement-based material (e.g., at different heights in a cylinder), can measure variations in the uniformity of the material resulting from bleeding and segregation effects.
The novelty of the present invention lies in its ability to monitor simultaneously chemical and physical processes, with a unique probe allowing, quantitative, in-situ, non-destructive, continuous measurement and real-time analysis of the data, which enable rapid observation of improper system behaviour.

Novel Multiple Electrode (multi-electrode) Probes The configuration and construction of the novel electrical conductivity probes for the purpose of the present invention can be performed in a variety of ways.
Two types of multi-electrode probes are described below as non-limiting examples: a reusable-type 5 probe for laboratory investigations and a sacrificial probe for laboratory or field studies in which the probes remain in the hardened material.
Multiple Electrode Reusable Probe A first conductivity probe was constructed for laboratory investigations on fresh 10 grout or mortar samples, following the schematic illustration of Fig.2a.
The electrodes consist of rectangular stainless steel (304) plates measuring 2.0 cm x 3.0 cm, and 0.15 cm thick; five such plates are mounted on the face of a vertical PVC post (3.0 cm wide x 30 cm high x 1.0 cm thick) equally spaced at 5.0 cm intervals. The electrical leads to each electrode are grouped on the back face of the PVC support post and isolated from the slurry by a second PVC plate; the latter has a vertical groove which conveniently houses the lead wires and is glued onto the PVC post. A second identical electrode assembly is mounted facing the first at a distance of 6.0 cm.
The complete probe unit is mounted vertically into a transparent PVC cylinder having a 10 cm diameter and a height of 35 cm. A level marking at 30 cm from the bottom of the cylinder indicates the position of the top surface of the cementitious materials to be investigated. The complete conductivity cell thus comprises five equally-spaced pairs of electrodes, which are permanently fixed at the following depths from the top surface of the cementitious material: 7, 12, 17, 22 and 27 cm. As noted above, this type of probe was intended for systematic investigations of bleeding-segregation effects in grouts and mortars; hence, it was only suitable for measurements during the dormant period.
Multiple electrode sacrificial probes A second type of probe was designed to perform, again, the measurement of conductivity as function of time and of depth in the fresh cementitious samples, but over extended periods into the setting and hardening phases. This type of measurement ' 11 requires probes, that will remain in the hardened material; these probes were thus designed for optimum simplicity of fabrication and use, as well as minimum cost.
In the probe design reproduced in Fig. 2b, the electrodes consist of stainless steel machine screws (0.5 cm dia. x 3 cm length) fixed in a PVC post. The latter is made by cutting a PVC pipe (2.5 cm dia.) longitudinally into two half cylinder channels. Five electrodes are fixed onto one of these PVC channels, equally spaced at 6.0 cm intervals;
the lead wires are housed in the channel which, in the end, is filled with an epoxy resin. With this type of probe, conductivity measurements are taken between each pair of electrodes in succession: 1-2, 2-3, etc.; the readings thus pertain to the average conductivity of a 6-cm high narrow column, as opposed to a 3 x 5 cm rectangular element in the reusable probe.
Electrical Circuitry for Conductivity Measurement The electrical conductivity of cementitious system is obtained from measurements of resistivity between inert metal electrodes using an AC
voltage, preferably a square wave signal. The electrical circuitry used for these measurements is illustrated schematically in Fig. 3. The electrical resistivity between the various pairs of electrodes embedded in the cementitious material is measured sequentially using a low voltage, 1 kHz square-wave signal; in these conditions, electrode polarization and related artifacts are minimized. A computer-controlled switching device enables automated measurements on all electrode pairs in a predetermined sequence. A
complete measurement cycle corresponding to the time required to record consecutive readings for five pairs of electrodes (r.m.s. current and voltage) requires 20 sec. The computer management of the system, as well as the data acquisition and analysis, and the calculation of stability index values (as described below) allow real-time evaluation of changes in the properties of the fresh cementitious material.
EXAMPLES
Typical results obtained using the novel conductivity probe and methods of the present invention are presented below for three types of fresh cement-based materials grout, mortar and concrete. These examples considered are typical, or representative, of a wide range of cement-based systems, which constitute and important class of colloidal systems for application of the present invention. The following paragraphs first describe the mix compositions of the systems in the selected examples, together with the common experimental methods used in the corresponding investigation.
Materials A Type 10 Canadian portland cement (CSA3-A5-M83, similar to ASTM C 150 Type I cement) was used for all the grouts, mortar and flowable underwater concrete (C1 and C2 mixtures). A ternary cement containing approximately 20% Class F
fly ash and 6% silica fume was used for the self-consolidating concrete (SCC) mixtures (mixtures C3 and C4).
Continuously graded crushed limestone aggregate with a nominal particle size of 14 and 10 mm were used for the C1 and C2 mixtures and C3 and C4 mixtures, respectively. A well-graded siliceous sand with a fineness modulus of 2.5 was employed for the mortar and concrete mixtures described in this paper, except for the M4 mortar where a slightly finer sand was used. The bulk specific gravities of the coarse and fine aggregates were 2.68 and 2.64, respectively, and their absorption values were 1.3 and 1.5%, respectively.
A sulfonated naphthalene-based HRWR with 42% solid content was used throughout this study; a dry powdered sodium gluconate was used as a set retarder. A
powdered welan gum was employed in the SCC mixtures to enhance stability.
Throughout this paper, the concentrations of the HRWR and set retarder are expressed in terms of solid content by mass of cementitious materials, while that of the viscosity agent (welan gum) by mass of water.
Mixture Preparation All mixtures were prepared at similar material temperatures of 20 + 3oC. The grout mixtures were prepared in 3-L batches using a Hobart mixer operating at a relatively low speed. The set retarder was consistently pre-blended with the cement, then introduced to the water in the mixer. The grout then received 3 minutes of mixing.
In the case of mortar, a Hobart mixer was also used. The batching sequence consisted ~ 13 of adding the cement to the water with the HRWR diluted in the latter. The mode of sand introduction and mixing protocol were consistent with ASTM C 109 recommendations.
The concrete was mixed in an open pan mixer. The sand and coarse aggregate were homogenized, then the cement, the water, the HRWR, and finally the welan gum were introduced. Once all mixture constituents were added, the concrete was mixed for 3 minutes. Following a 3-minute rest, the mixing was resumed for 2 additional minutes Conductivity Probe Calibration A simple calibration procedure was developed which allows the calculation of an effective cell constant for each electrode pair (with either type of probe). This is performed by immersing the probe in a standard solution of NaCI or KCI of known conductivity and measuring sequentially the electrical resistivity of every electrode pairs as configurated for measurements on cementitious systems. The resistivity values were obtained by averaging five current/voltage readings under a square-wave at 1.0 kHz excitation. The readings were taken with increasing voltage in the range of 0 to 10 V.
The constant currendvoltage slope observed ensures that the circuit is mainly resistive during the period of observation. This calibration procedure enables correction for variations among electrode pairs due to unavoidable geometrical variations in probe construction.
Conductivity Measurement Procedures Reusable probes Immediately following completion of the mixing process of the grout, mortar, or concrete, the fresh material is poured into the test jar (300 mm effective height), and the measurement sequence is initiated, taking, as noted above, five consecutive readings for each electrode pair, then passing on to the next electrode pair until all five electrodes have been scanned. This is repeated at 2-minute intervals, typically up to four hours, then the material removed from the cell before setting occurs.

~ 14 Sacrificial probes The measurement procedure followed with the sacrificial probes was identical to that described above for the reusable probes. Since sacrificial probes were designed to remain permanently in the cementitious material sample, they enabled measurements over longer periods, beyond the consolidation and into the hardening stages;
in the examples given below, the conductivity measurements were recorded for periods of up to 72 hours.
Procedures for Measuring Bleeding and Segregation in Concrete Surface bleeding was determined using standard 150 x 300 mm cylinders. The external bleeding was determined by measuring the amount of bleed water at the top of the column. The cumulative bleeding volume was monitored at fixed intervals of and then of 40 minutes until reaching steady state conditions, corresponding approximately to the beginning of hardening. Except during the bleeding measurements, the cylinders were covered to prevent evaporation.
The segregation of the hardened concrete was evaluated by vertically sawing the 150 x 300 mm cylinders used for the bleeding test to determine the variations in the volume fraction of the aggregate as a function of height. The concentration of coarse aggregate (> 5 mm) was determined through image analysis of the sawed section which yields the surface area of the aggregate on the section, relative to the total surface of the section. The segregation coefficient was estimated using a sum of squares approach by expressing the deviation of coarse aggregate distribution from a weighted average value (6).
Procedures for Data Analysis The conductivity data collected as function of time and depth in the sample (from the surface) is examined in three distinct time domains:
- the dormant period: to calculate a "stability index";
- the setting region: to identify an effective setting time - the hardening period: to determine the rate of hardening Stability Index The primary purpose of the method presented here is to derive a quantitative measure of bleeding- segregation effects in a fresh material, using the conductivity as a local probe to monitor the development of heterogeneity in the material.
Taking the 5 five conductivity readings at different levels at a given time, the vertical heterogeneity in the sample is adequately reflected through the standard deviation of the conductivity values (aA). The latter should be normalized to the initial value of the average conductivity measured over the five electrodes (Aav) to allow for variations in conductivity values between different systems due to differences in their chemical 10 compositions. An apparent stability index (Is) can then be defined simply as Is - ( 1 - aA / Aav ) Eq. 1 A highly stable system will exhibit a unit Is value; a system with a strong tendency towards bleeding and segregation will show Is < 1 . This definition of 15 homogeneity is only valid for the plastic stage of the material. Upon setting and hardening, large changes in the average conductivity values will induce Is variations which are not related to heterogeneity of the material.
Apparent Setting Time As noted by other workers previously (9 to 14), the onset of paste setting in a cement-based material is generally accompanied by a decrease in the conductivity of the paste due to the precipitation of portlandite (Ca(OH2)). Since the conductivity of the paste usually increases up to the setting point (neglecting segregation effects), the maximum value in the conductivity vs. time curve would appear a good indicator of the initial set time.
Rate of Hardening During the hardening period, the strong decrease in the electrical conductivity of cementitious systems is largely due to progressive blocking of the ionic transport channels as the microstructure of the material develops. The rate of drop in the conductivity values between the setting point (measured as maximum conductivity) and a fixed curing period should provide a reasonable estimate of the rate of strength development in the material. Therefore, Ot1 values were obtained as [ llset -Wset +
Ot) ] and compared to the compressive strength developed over the same period (Ot).
The compressive strength was determined on 50 mm cube samples maintained at room temperature and de-moulded immediately before strength testing.
RESULTS
Example 1 : Conductivity Measurements on Grouts with Reusable Multi-electrode Probes Seven grout mixtures, labelled G1 to G7 were prepared with various compositions reported Table 1.
Highly fluid grout with Water/ Cement (W/C) = 1.0 The type of results obtained from multiple electrode (reusable probes) conductivity measurements is illustrated in Fig. 4 with a highly fluid cement grout (W/C = 1 ). The conductivity readings are given from t = 15 minutes, to about t = 4 hours. The elapsed time here refers to the elapsed period following the introduction of water to cement. At t = 0, all electrode pairs must yield the same conductivity value since the suspension is still in a homogenous state. The first set of readings of conductivity however correspond to 15 minutes after the introduction of the sample in the test mould. The time-dependence of the conductivity values at the top (7 cm) and at the bottom (27 cm) of the 30 cm grout column clearly indicates important bleeding-segregation effects during the first two hours. The electrodes located near the bottom of the sample show conductivity decreasing with time, which is consistent with a decrease of the volume fraction of the conducting solution phase due to sedimentation of cement particles. On the other hand, the electrodes near the surface show a rapid conductivity increase, corresponding to an increase in the volume fraction of solution in that region (i.e., due to bleeding), and to an increasing concentration of electrolytes in the solution as hydration reactions proceed. Electrodes located at intermediate heights reflect the combined influence of segregation-bleeding and electrolyte solubilization on the bulk conductivity throughout the sample. Because of the opposite effects of these phenomena on bulk conductivity, the readings at intermediate heights (e.g., corresponding to the 12 cm and 17 cm electrode pairs) exhibit a maximum as function of time (these maxima are not related to the setting phenomenon which occurs much later). The position of the maximum on the time scale is related to the migration rate of solids and solution in the grout; it thus provides direct indication of the kinetics of the bleeding-segregation phenomena as it occurs.
The conductivity data as function of time in Fig. 4 are presented as function of the position of the electrodes in the sample, at selected times (15, 75, 135 and 255 minutes) in Fig. 5. At the beginning, the conductivity values measured are comparable throughout the sample depth, and this is reflected by a near-vertical A vs.
height curve.
The latter is increasingly curved at longer times reflecting the evolution in the sample heterogeneity. This representation illustrates the significance of the stability index calculated according to Eq. 1 and reported as function of time in Fig. 6. For each selected time, the mean conductivity value is determined; the standard deviation is then calculated and normalized to the value at t = 15 minutes, yielding a measure of the heterogeneity of the sample. In the present case, the initial stability index value is greater than 0.97, but it decreases rapidly with time in the first 90 minutes, indicating a highly unstable system.
Reproducibility of method on grouts with W/C=0.75 The repeatability of the method is illustrated in Fig. 7 which shows the variation in conductivity with time for the top (7 cm) and bottom (27 cm) pairs of electrodes, as observed in three distinct experiments with cement grouts G2, G3 and G4, all with W/C
of 0.75. The corresponding variations in the calculated stability index (Is) are illustrated in Fig. 8. As expected, the Is values for these grouts are higher than that observed for the W/C of 1 grout Previous example), but significant bleeding-segregation is still apparent during the first hour of consolidation. As may be seen from the consistency in the behaviour of the Is values, the repeatability of the measurement is quite satisfactory.
The coefficient of variations calculated from the conductivity data at each electrode pair over the entire duration of the measurement for the three grouts was less than 5%.

Application of method in grouts with varying W/C and with set retarding admixture As further test of the method, several other grouts were examined, typically, as function of W/C and in the presence of admixtures. The results for three grout mixtures G5, G6, and G7 (W/C = 1.0, 0.8, and 0.4) in the presence of a set-retarding admixture (0.1 % sodium gluconate) are presented in Fig. 9. The results again show variations in conductivity at the top and bottom electrode pairs; the corresponding stability index for these systems are illustrated in Fig. 10. In some of these systems, there appears to be rapid initial segregation, as indicated by significant differences in the first conductivity readings taken 15 minutes after mixing. As observed earlier with other grouts, the stability index is rather high initially and decreases with time in a way related to the W/C values.
The initial results obtained with the multiple electrode conductivity probe (reusable type) demonstrate the usefulness of the method for monitoring variations in the homogeneity of a cement paste with time.
Example 2. Conductivity Measurements on Mortars with Sacrificial Multi-electrode Probes This Example is given to illustrated the applicability of the novel multi-electrode probe and for monitoring cement-based mortars over extended periods typically 24 to 96 hours. In such case, the conductivity-vs-time data can provide information, initially on the variations in homogeneity of the system, and later on its behaviour during the setting and hardening period.
A total of six mortar mixtures (labelled M1 to M6) were prepared with a ratio of cementitious materials to sand of 1:1.5 and W/CM of 0.4 and compositions as given in Table 1 Except for the M1 mortar, the mixtures were highly fluid with approximate spread diameter values (ASTM C 143) of 31 + 1 cm. The conductivity vs. time curves observed with the top and bottom electrode pairs are illustrated in Figs. 11-14. The general features of these curves are similar, and their shape is related to the schematic illustration presented in Fig. 1a; in accordance with the latter, the mortar conductivity may be examined in three distinct periods: consolidation, setting and hardening.

Stability of fresh mortars of varying compositions The conductivity data illustrated in Fig. 11 for a non-superplasticized mortar with low fluidity level (spread diameter of 11.5 cm) is typical of a very stable mortar, with little difference in the conductivity curves for top vs. bottom electrode pairs. The stability index values derived from these measurements in the first 12 hours of testing is close to unity, as shown in Fig. 15 for the M1 mix.
Mortars of much greater fluidity (approximate spread of 31 cm) containing 8%
silica fume, no set retarder, and a HRWR yielded conductivity curves which point to significant bleeding-segregation, even at very early times, as well as delayed setting; this can be seen from the data for mortar M2 in Fig. 12. This behaviour is exacerbated upon addition of a set retarding admixture (0.05% and 0.10% sodium gluconate) for the same mortar composition, as shown in Fig. 13. The occurrence of bleeding-segregation effects is very pronounced in the earliest measurements recorded; the stability index values over time for this mortar is also significantly reduced, as shown in Fig. 15. The M4 mortar was prepared with a sand that is finer than the sand employed for the other mixtures; this explains the higher apparent stability of this mortar relative to M3.
The conductivity data presented in Fig. 14 illustrates the behaviour of the M5 mortar made with 20% Class F fly ash and 6% silica fume replacement, and a high level of retarding admixture (0.10% of sodium gluconate). In spite of an excessive retardation, the conductivity curves indicate a reduced level of bleeding-segregation compared to the M3 and M4 mixtures reflecting the high stability of the binder combination.
As with the multi-electrode conductivity measurements on grouts, the conductivity data obtained on fresh mortars clearly shows the usefulness of the method in monitoring variations in the homogeneity of the system, resulting from bleeding and segregation effects.
Apparent setting times from multi-electrode conductivity measurements As expected from prior art on the conductivity of fresh cement-based systems (8 to 15), the shape of the conductivity-vs.-time curve observed in Figs. 11-14 initially exhibit increasing conductivity with time during the lag phase; later, a sharp decrease in the conductivity occurs due to precipitation of portlandite, and the resulting maximum signals the beginning of the setting of the cement paste. The actual setting times (initial) of the mortar mixes could not be measured through the standard techniques, but the time at which the maxima is observed with the different mortars appears consistent with the mix design of these mortars. For example, in the case of the M2, M3, and 5 mortars made with the same composition and with increasing dosages of set retarder (0, 0.05, and 0.10%), the time periods corresponding to the peak before observing a sharp reduction in conductivity are 5.5, 24, and 35.5 hours, respectively. These data confirm the usefulness of the multi-probe conductivity method for monitoring chemical reactions occurring in the cement-based system and leading to setting and hardening of 10 the material.
Hardening rates The rate of drop in the conductivity of a setting cement-based system should be qualitatively related to the rate of strength development (in a way similar to the heat evolution which reflects maturity and is related to strength). In attempts to relate the 15 conductivity variations measured and the strength development in the mortars, the following simple procedure was adopted. A mortar hardening rate was calculated using the compressive strength, (f'~ ) observed at the end of the sharp conductivity decline, and in some cases at intermediate points of the descending curve, taking the compressive strength as nil at the maximum in the conductivity curves. The data used 20 for this comparison are collected in Table 2. The slopes Of'~ / ~Aav (where DAav is the average conductivity value over the entire sample height) obtained for the mortars examined in this part of the study are plotted in Fig. 16. A
strength/conductivity correlation with a coefficient of approximately 0.9 is observed, which, given the extreme range of mortar behaviors, is satisfactory confirmation of the validity of the approach.
The multi-electrode conductivity method is thus seen to provide useful quantitative information on the rate of hardening (rate of strength development) of the mortars, an important parameter of these systems in application.

Example 3 : Measurements on Fresh Concrete with Sacrificial Multi-electrode probe The applicability of the multi-electrode conductivity approach to monitor the stability of plastic concrete was also demonstrated in concrete mixtures described in Tables 1 and 3. The mixture compositions were adjusted to allow comparisons at constant workability between C1 and C2, and C3 and C4 corresponding to highly flowable concrete for underwater and SCC, respectively. The C1 and C2 mixtures were prepared with Type 10 cement and differed by their W/C values (0.41 and 0.55).
Similarly, the C3 and C4 mixtures had similar compositions, except for W/CM of 0.42 and 0.48 and the concentrations of chemical admixtures.
The external bleeding (ASTM C-232) characteristics of these materials were also measured and the results are shown in Fig. 17. As expected, the mixtures with greater W/C exhibited higher bleeding. Comparing the behaviour of C1 and C2, the rate of surface bleeding is found to be highest with the C2 concrete, a superplasticized concrete having a rather high W/C of 0.55. With the SCC mixtures C3 and C4, the external bleeding is much lower, and it occurs significantly later (i.e., 6 to 10 hours).
The lowest bleeding rate is found with C4, an SCC containing welan gum as a thickening agent, but a lower W/CM than the C3 concrete.
The segregation of the coarse aggregate is illustrated in Fig. 18 which shows the distribution of the coarse aggregate along the height of the cylinders. The values indicated on the graphs refer to the segregation coefficients. The relative segregation behaviors of these materials follows the same pattern as the bleeding rates:
the greater the bleeding, the greater the tendency of the coarse aggregate to segregate towards the lower part of the fresh concrete sample. The segregation coefficients of the C1 and C2 mixtures were 4.3 and 6%, respectively, and those for the C3 and C4 mixtures 1.9 and 3.1 %.
The conductivity data measured with these fresh concrete samples yields curves similar to those observed with mortars and the stability indices obtained from the conductivity data are reproduced in Fig. 19. The trend observed in the latter follows roughly the trends reported above within the bleeding and segregation data.
However, ' 22 the detailed variations in stability index values with time probably reflect limits of the equipment used in the present study.
In particular, because of the relatively small size of the electrodes, and the short spacing between them, the presence of 10- or 14-mm nominal size aggregate may have a considerable influence on the bulk values measured; upon segregation, the motion of such large particles in the electrode gap leads to substantial variations in the volume fraction of the paste within the measuring volume, thus inducing large fluctuations in the measured conductivity. Application of the method to a broad range of concrete types will, therefore, need to consider changes in electrode geometry to alleviate such difficulties.
Table 1 - Composition of the examined cement-based materials Mixture W/CM CementitiousChemical Test Consistency materials admixtures duration Grout mixtures G1 1.0 Non 240 mm h G2, G3, 0.75 Non 240 mm i G5 1.0 0.1 % Na 150 mm bfg I uconate G6 0.8 0.1 % Na 150 mm ; '' g,,.

I uconate G7 0.4 0.1 % Na 150 min luconate Mortar xtures mi M1 0.40 Non 24 hrs 11.5 cm spread diameter M2 0.40 8% SF 0.6% HRWR 72 hrs 31.5 cm spread d iameter M3 0.40 8% SF 0.6% HRWR 96 hrs 32.5 cm 0.05% Na spread luconate diameter M4 0.40 8% SF 0.6% HRWR 92 hrs 30 cm spread 0.1 % Na diameter I uconate Mixture W/CM CementitiousChemical Test Consistency materials admixtures duration M5 0.40 6% SF + 0.6% HRWR 96 hrs 31.5 cm 20% fly 0.1 % Na spread ash luconate diameter M6 0.40 30% fly 0.6% HRWR 72 hrs 32 cm spread ash 0.1 % Na diameter luconate Concrete mixtures C1 0.41 0.29% HRWR 10 hrs Slump 220 mm C2 0.55 0.15% HRWR 10 hrs Slump 220 mm C3 0.42 6% SF + 0.50% HRWR 10 hrs 650 mm 20% fly 0.16% welan slump flow ash um C4 0.48 6% SF + 0.39% 10 hrs 650 mm 20% fly HWRW slump flow ash 0.14% welan um Table 2 - Variations in conductivity vs. strength gain of 1:1.5 mortar mixtures Test Binder W/CM Admixtures0t ~f'~ 0A (mS/cm) No. (hrs)(Mpa) M1 T a 10 0.40 Non 19 21.4 3.67 M2 8% SF 0.40 0.6% 66 40.5 5.88 HRWR

M3 8% SF 0.40 0.6% 24 32.3 4.35 HRWR 73 45.5 5.01 0.05% Na I uconate M5 6% SF 0.40 0.6% 24 12.5 3.17 +

20% FA HRWR 48 26.8 3. 82 0.1 % N
a I uconate M6 30% FA 0.40 0.6% 58 34.0 5.17 HRWR

0.1 % Na I uconate Table 3 - Mixture proportioning evaluated concrete Materials Fluid Self-consolidating concrete concrete Cement 400 400 520 520 (kg/m3) Type 10 Type 10 Ternary Ternary Cement type cement cement Water (kg/m3)164 220 218 250 Sand (kg/m3) 705 705 710 700 Coarse 1040 1040 890 875 aggregate (14 mm (14 mm (10 mm (10 mm (kg/m3) MSA) MSA) MSA) MSA) HRWR (I/m3) 2.3 1.2 5.2 4.0 Welan gum 0 0 0.35 0.35 (kg/m3) Water reducer0 0 0.75 0.75 (I/m3) Slump (mm) 220 220 - -Slump flow - - 650 650 (mm) CONCLUSIONS
The results of the measurements reported in the foregoing examples demonstrate the novelty and usefulness of multi-electrode conductivity measurements in cementitious materials, specifically:
- The novel multi-electrode probe and measuring method present non destructive tools to monitor in-situ and continuously the stability (stability index) of all types of cementitious systems.
- The AC (square wave at low frequency) electrical conductivity measured at different levels in a plastic cement-based material can reliably detect variations in the homogeneity of the material resulting from bleeding and segregation effects.
- The sensitivity of the method is adequate for monitoring bleeding segregation, in-situ and in a non disruptive manner, in plastic grout, mortar, and concrete.

- The conductivity data can be used to define a quantitative measure of the inhomogeneity in a fresh cement-based material in terms of a "stability index"
(Is) for the system.
- Using disposable probes, the setting and hardening rate can also be 5 monitored.
- The results also point to a valuable relationship between the strength gain and the change in the mean conductivity value during the hardening period.
- Within sensitivity limits, mainly set by the geometry of the electrode-probe assembly, the method provides valuable data on the behaviour of plastic cement-based 10 materials.
- The method can be exploited either in the development of mixture formulations with appropriate bleeding characteristics, or for in-situ monitoring of concrete consolidation.
- Given that cement-based material are complex colloidal systems in which 15 numerous physical and chemical processes occur simultaneously, the multi-electrode conductivity probe and method is most certainly applicable in less reactive colloidal systems, typically suspensions, slurries, sludge or pastes.

REFERENCES
1. Neville, A. M.,"Properties of Concrete," Fourth Ed., Longman Group Ltd., 1995.
2. Hoshino, M., "Difference of the W/C Ratio, Porosity and Microscopical Aspect between the Upper Boundary Paste and the Lower Boundary Paste of the Aggregate in Concrete," Materials and Structures, V. 21, No. 125, Sept. 1988, pp.
3 3 6-340.
3. Khayat, K.H., "Use of Viscosity-Modifying Admixture to Reduce Top-Bar Effect of Anchored Bars Cast with Fluid Concrete," ACI Materials Journal, V. 95, No. 2, 1998, pp. 158-167.
4. Petrov, N., "Investigation of In-Situ Properties of Self-Consolidating Concrete:
Influence on Interface with Reinforcement and on Corrosion," Masters Thesis, UniversitE de Sherbrooke, Qc, 1997, 190 p.
5. Powers, T.C., "The Bleeding of Portland Cement Paste, Mortar, and Concrete,"
Portland Cement Assoc. Bul. No. 2, Chicago, July 1939.
6. Khayat, K.H., Guizani, Z., "Use of Viscosity-Modifying Admixtures to Enhance Stability of Fluid Concrete," ACI Materials Journal, V. 94, No. 4, 1977, pp.

340.
7. Ritchie, A.G.B., "Stability of Fresh Concrete Mixes," Journal of Construction Division, ASCE, Proc. V. 92, No. C01, 1966, pp. 17-36.

8. Nonat, A., Mutin, J.C., " From Hydration to Setting," Proceedings of the International RILEM Workshop on Hydration and Setting Cements, 1991.

9. McCarter, W. J., Brousseu, B., " The A.C. Response of Hydrated Cement Paste,"
Cement and Concrete Research, V.20, 1990, pp. 891 - 900.

10. McCarter, W. J., Curran, P. N., " The Electrical Response Characteristics of Setting Cement Paste," Magazine of Cement Research, V. 30, No. 126, March 1984, pp.
42-49.

11. McCarter, W. J., " Monitoring the Influence of Water and Ionic Ingress on cover zone subjected to Repeated Absorption," ASTM Cement, Concrete, and Aggregate, V. 18, No., 1, June 1996, pp. 55-63.

12. MacDonald, K. A., Northwood, D. O., " Rapid Estimation of Water-Cementitious Ratio and Chloride Ion Diffusion in Hardened and Plastic concrete by Resistivity Measurement," Editor, Mohammad Khan, Water-Cement Ratio and Other Durability parameters, Techniques for Determination, Publication SP-191, ACI
International , year 2000.

13. Vernet, C., " Hydration Kinetics and Mechanical Evolution of Concrete during the First Days: Study of the Hardening Mechanism, " Proc. 9th Int. Congress of Chemistry of Cement, New-Delhi, 1992, pp. 511-517.

14. Vernet, C., Nowaryta, G., " Condctometric Test for Cement-Admixture Systems,"
Proc. 9th Int. Congress of Chemistry of Cement, New-Delhi, 1992, pp. 627-633.

15. Jolicoeur, C., Simard, M.-A., "Chemical Admixture-Cement Interactions:
Phenomenology and Physico-Chemical Concepts," Cement and Concrete Composites, V. 20, 1998, pp. 87-101.

16. Taylor, M. A., Arulanandan, K., "Relationships Between Electrical and Physical Properties of Cement Pastes," Cement and Concrete Research, V. 4, 1974, pp.
881-897.

17. Haran, B. S., Popov, B. N., Zeng, G., and White, R.E., " Development of a New Electrokinetic Technique for Decontamination of Hexavalent Chromium from Low surface charged soils," Environmental Progress, Vol. 15, No, 3, Fall 1996, pp. 166-172.

18. Dzenitis, J. M., " Soil chemistry effects and flow prediction in Electroremediation of Soil," Environmental Science and Technology, 1997, vol. 31, No. 4, pp. 1191-1197.

' 28 19. McCarter, W. J., " The electrical resistivity characteristics of compacted clays,"
Geotechnique, Vol. 34, No. 2, 1984, pp. 263-267.

20. Mitchell, J. K., and Arulandan, K., " Electrical Dispersion in Relation to Soil Structure," Journal of Soil Mechanics and Foundation Engineering, Proceedings of ASCE, Vol. 94, No. SM2, Mar. 1968, pp. 447-471.

21. Matveenko, A. P., Strizhev, E. F., Volkova, A. N., Ivanova, L. V., and Yakovlev, V.
I., " Electrocoagulation of Bio-organic Impurities in Wastewaters from Biochemical Processes," Journal. Applied Chemistry, USSR, Vol. 54, 1991, pp.
2282.

Claims (14)

1. A method for in-situ, non-destructive, continuous and quantitative measurement of changes occurring in aqueous-based suspensions, slurries, pastes, sludge and other colloidal systems, said method being based on variations of electrical conductivity measured at multiple locations in the colloidal system and as function of time.

2. The method of Claim 1 where changes in said colloidal systems result from either, physical or chemical processes, anyone of which is accompanied by a variation in the bulk electrical conductivity of the system. A typical physical process is, for example, a change in the homogeneity of the colloidal system due to sedimentation of solids, or migration of solution; a typical chemical process is, for example, precipitation of an insoluble phase from reaction between two or more components of the system.

3. The method of Claim 1 wherein variations in electrical conductivity are measured simultaneously, or in rapid sequence, at different points, or volume elements, in the colloidal system, using a novel multiple electrode (multi-electrode) probe unit; the latter is introduced in the system in a fixed position, temporarily or permanently, depending on the intended use.

4. The method of Claim 3 wherein variations in the heterogeneity of an aqueous colloidal system resulting from physical processes (sedimentation, settling, bleeding, etc.) can be monitored through measurement of electrical conductivity of the colloidal system as function of depth and time.

5. The method of Claim 3 wherein variations in the heterogeneity of an aqueous colloidal system resulting from chemical processes (dissolution, reaction, precipitation, etc.) can be monitored through variations in the average electrical conductivity over the entire volume of the colloidal system.

6. The method of Claim 3 where said multi-electrode probe consist of two or more pairs of metal electrodes fixed onto a supporting mount, with spacing between the two electrodes in a pair ranging typically between 5 and 20 cm, and the spacing between the different electrode pairs is typically also between 5 and 20 cm.

7. The method of Claim 3 where said electrode pairs of are activated with a low voltage electrical signal, typically in the range of 0.2 to 20 Volts, said electrical signal being from an AC source, having a frequency preferably in the range 100 to 10 000Hz.

8. The method of Claim 3 where said multi-electrode probe are activated in sequence (using standard multiplexing devices), each for a period of time sufficient to allow an accurate measurement of the electrical conductivity (or inversely the resistivity) in the volume element between the electrodes in the pair. The data collected as function of time for each electrode pair, typically at a fixed time interval of one minute, are logged into specific computer files for later analysis.

9. The method of Claim 3 where said multi-electrode probe is used to monitor the evolution in time of the properties of fresh cementitious materials derived from portland cement , or any other hydraulic binder.

10. The method of Claim 3 where said multi-electrode probe enables quantitative observations on changes in the homogeneity resulting, typically, from settling of solid particles, segregation of coarse aggregate, upward migration of water and surface bleeding.

11. The method of Claim 3 where said multi-electrode probe used for the continuous monitoring of fresh cementitious materials allows the detection of the onset of the hardening reaction and the rate of development of the mechanical resistance during the early hardening period, the latter being typically.

12. The method of Claim 3 where the data obtained as function of time for each electrode pair of the multi-electrode probe are analysed to yield rate parameters, or functions, describing the time evolution of the colloidal system.

13. The method of Claim 3 where the data obtained as function of time in fresh cementitious using the multi-electrode probe, typically in the time interval (0-72hrs), are analysed to yield, typically, the bleeding rate and magnitude, the segregation rate and magnitude, the setting rate and the early hardening rate.

14. The method of Claim 1 where said multi-electrode probe is designed to optimal application requirements by varying the following design parameters :
number of electrode pairs, spacing between electrodes in a pair and between pairs of electrodes, electrode materials, electrode shape, electrode mount.