Dimensional Stability of Grout-Type Materials Used As Connections for Prefabricated Bridge Elements

Chapter 3. EXPERIMENTAL PROGRAM

INTRODUCTION

A broad-scope research project on the performance of field-cast grout-type materials and their use in PBE connections is being conducted at the Federal Highway Administration (FHWA) Turner-Fairbank Highway Research Center. Graybeal et al. reported on an extensive grout-type materials characterization study.⁽¹⁷⁾ One of the outcomes of that research was the wide range of grout performance that can be obtained as well as the propensity of the materials to undergo volumetric deformations. The purpose of the current report is to expand upon this work and present one of the focus points of the aforementioned research study: evaluation of dimensional stability of grout-type materials that may be used in PBE connections. This chapter describes the grout-type materials selected and testing procedures used for material and dimensional stability characterization. The approach followed in this research to provide IC in cement-based grouts is also discussed.

MATERIALS AND MIXING PROCEDURES

A total of 11 grout-type materials of different nature and manufacturer were used in this study. Among them there were conventional grouts, repair materials, and a UHPC that could potentially be used as a grout in connections between prefabricated concrete elements. UHPC is a class of cementitious material designed to exhibit exceptional mechanical and durability properties.^(29-31) All these materials provide two of the main properties required for grout-type materials: high flowability and high early-age strengths. The grout category, nomenclature used, and cost range are listed in table 1. Material categories were chosen based on past performance, applicability to onsite construction processes, and suitable published properties.[1] The grout is normally supplied in a bag containing the solid fraction (e.g., cementitious materials, additives, and fine aggregates) that is mixed with a certain amount of water following the recommendations of each of the grout's manufacturer, with the exception of the epoxy-based grouts, which are mixed with a resin and a hardener in the amounts also recommended by the manufacturer. The mixing details are summarized in table 2.

^aCost range estimated from cost purchasing values corresponding to the year 2013.

1 yd³ = 0.765 m³

^aSpecific gravity of the solids fraction measured using a gas (He) pycnometer.

^bResin component = 6 lb (2.7 kg); hardener component = 1.1 lb (0.5 kg).

^cResin component = 5.2 lb (2.4 kg); hardener component = 1.6 lb (0.7 kg).

^dMixing time based on time needed for the materials to achieve a temperature of 90 °F (32 °C). Time ranges from 10 to 15 min.

^eRefers to water-to-binder ratio (w/b) as formulation is known.

The mixing amounts are based on the amount of solid that is contained in one bag (or bucket, when applicable). When evaluating grout-type materials, it is difficult to talk in terms of w/c or w/b because the reactive fraction of the solid is unknown. Instead, w/s is typically used. As observed, most of the grout-type materials had similar low w/s for high early-age strength development (0.16 to 0.19), with the exception of two of them (F1 and M1), which had an even lower w/s (0.08). U3 had a reported formulation, thus the value of 0.18 refers to the actual w/b. U3 also included the addition of chemical admixtures and steel fibers. The supplier of E1 and E2 recommends two formulations: standard and high flow. In this study, the high flow formulation was used. The specific gravity values of all the materials ranged from 2.59 to 3.16, inferring that the solid fraction is a mix of cementitious materials and fine aggregates. The high specific gravity value of the G2 grout was due to the presence of metallic particles in the solid fraction. Mixing times were typically 5 min (based on ASTM C1107), except for rapid set materials (G5, F1, M1), for which the mixing times were reduced according to the manufacturers' recommendations.⁽¹⁾ M2 and U3 had longer mixing times (about 10 to 15 min) also based on their manufacturers' recommendations. All grouts were prepared in either a mortar or concrete mixer (depending on batch size) except for F1 and M2, which were mixed in buckets with a drill and paddle (see figure 4). For more information about these materials, refer to
appendix A.

Figure 4. Photo. Mixing M2 grout using a drill and a paddle.

FRESH PROPERTIES

Initial Workability

Since the materials selected in this study were intended to be used in the same type of application (i.e., as connections between prefabricated concrete elements), the comparative criterion chosen was to have similar fresh properties in terms of initial workability. This was done using the consistency definitions described in ASTM C1107, where the consistency is classified in three categories (plastic, flowable, and fluid) based on the flow measured in accordance with ASTM C1437.^(1,32) In this study, plastic consistency was chosen, which corresponds to a flow increase between 100 and 125 percent of the original base diameter of the mold used in the flow test. (Although earlier in the report it was stated that grouts need to have high flowability, the plastic consistency described in ASTM C1107 is considered to be sufficient for placing and pumping purposes.) The flow of the grout was measured on the standard flow table after five drops in 3s, as shown in figure 5. Then, the diameter was measured along the four lines marked on the table, and the average was recorded. This measurement was taken 7 and 15 min after mixing.

Figure 5. Photo. Flow table test as per ASTM C1437, including accessories to run the test (left) and grout spread (right).⁽³²⁾

Air Content and Unit Weight

The air content and unit weight of the mixtures were measured in the fresh state using the method described in ASTM C231 and ASTM C138, respectively.^(33,34) The fresh material was poured in the measuring bowl of the air meter apparatus in one single layer (due to the high flowability of the material) and flushed with the top edge. In some cases (e.g., epoxy grouts), external vibration was needed to further consolidate the material in the bowl. Immediately after this process, the mass of the bowl and material was taken, and the previously measured mass of the bowl was subtracted. The result was divided by the known volume of the bowl, resulting in the unit weight of the material. Then, the air meter apparatus was assembled, and the air content was measured according to ASTM C231. The testing apparatus is shown in figure 6.

.

Figure 6. Photo. Air content and unit weight.

Set Time

The set time of the mixtures was measured according to ASTM C403.⁽³⁵⁾ The test was based on measuring the pressure force needed to force a set of standard flat-headed needles to penetrate 1 inch (25.4 mm) into the material being tested, as shown in figure 7. The material was placed in a 6-inch (152-mm)-diameter by 6-inch (152-mm)-height cylinder and stored in a controlled environmental room at 73.4 ±1.8 °F (23 ±1 °C) and a relative humidity (RH) of 50 ±5percent. Readings were taken periodically after placing the material until a pressure of 4,000 psi (27.6 MPa) was reached, which indicates the time of final set, whereas a pressure of 500 psi (3.45MPa) indicated the initial setting time. Two samples were used for this test. The data are plotted as pressure versus time after mix initiation. An example is shown in figure 8.

Figure 7. Photo. Loading apparatus and penetration needles.

1 psi = 0.007 MPa.

Figure 8. Graph. Typical setting time data plot indicating initial and final set times.

HARDENED PROPERTIES

Compressive Strength

For the evaluation of compressive strength, three 2-inch (51-mm) cube specimens (like those described in ASTM C109, see figure 9) were prepared according to ASTM C1107.^(36,1) The cubes were tested at several ages: 4 and 8 h and 1, 3, 7, and 28 d. The cube specimens were kept in their molds for 24 h, at which time they were demolded and sealed within plastic bags until the age of testing, unless the testing age was within the first 24 h, in which case, the specimens were tested immediately after being demolded.

Figure 9. Photo. Compressive strength cube specimens (left) and cube specimen after test via ASTM C109 (right).⁽³⁶⁾

CHEMICAL PROPERTIES

Heat of Hydration

Numerous properties of cementitious materials are controlled by their initial hydration rate, volume change being one of them. One convenient way to measure the hydration reaction (rate and degree of hydration) is using isothermal calorimetry. An isothermal calorimeter was used in accordance with ASTM C1679, pictured in figure 10.⁽³⁷⁾ Approximately 0.25 oz (7 g) of an externally mixed material was weighed and placed in a glass ampoule, which was then capped and placed into the isothermal calorimeter about 10 min after mix initiation. The cumulative heat of hydration was measured during the first 7 d after mixing. It is important to mention that for some of the fast-setting grouts (typically the repair materials), it is difficult to detect the first heat release because it happens a few minutes after the solid-water contact.

Figure 10. Photo. Isothermal calorimeter and sample specimens via ASTM C1679.⁽³⁷⁾

Chemical Shrinkage

This test was performed in this study for IC design purposes. (Further details on IC mix design are provided later in the report.) The primary objective of any chemical shrinkage test is to quantify the change in volume that occurs as a result of hydration reactions. According to Le Chatelier, the volume of the hydration products in cementitious systems is lower than that of the initial reactants.⁽³⁸⁾ The chemical shrinkage is generally quantified by measuring the amount of water that is absorbed by a saturated sample, as described in ASTM C1608.⁽²⁷⁾ A thin layer of a fresh sample was placed in a glass vial, and the vial was filled with water. A rubber stopper with an inserted capillary tube was tightly placed in the vial. As the rubber stopper was inserted, the water level in the capillary tube rose. The test setup is shown in figure 11. This sample setup was placed on a horizontal surface in a controlled environmental room at 73.4 ±1.8 °F (23 ±1 °C) and an RH of 50 ±5 percent. Three replicate samples were prepared. As hydration occurred, the water level in the capillary tube decreased. The volume decrease corresponds to the volume of chemical shrinkage and, thus, to the extent and rate of reaction. This is similar to the way that the extent and rate of reaction is captured using an isothermal calorimeter by measuring the heat release instead of the volume change.

Figure 11. Photo. Chemical shrinkage testing via ASTM C1608.⁽²⁷⁾

SHRINKAGE PERFORMANCE

Shrinkage Performance Requirements According to ASTM C1107

As previously mentioned, the ASTM C1107 standard specification covers packaged dry hydraulic cement non-shrink grouts.⁽¹⁾ One of the performance requirements stated in this specification is the allowed volume change that the grouts can undergo. Volume changes are measured in terms of height change of a cylindrical specimen through two other ASTM test methods, ASTM C827 Standard Test Method for Change in Height at Early Ages of Cylindrical Specimens of Cementitious Mixtures and ASTM C1090 Standard Test Method for Measuring Changes in Height of Cylindrical Specimens of Hydraulic-Cement Grout, for early-age (fresh) and hardened height changes, respectively.^(39,40) The following sections describe these two test methods.

ASTM C827

The height change of a 3-inch (76-mm)-diameter by 6-inch (152-mm)-tall cylindrical specimen was measured in accordance with ASTM C827.⁽³⁹⁾ However, a modification of the ASTM C827 test method was made where a non-contact laser was placed above the specimen and used to measure the vertical distance from the laser to the indicator ball placed on the top surface of the specimen (see figure 12). This approach provided more simplicity in the execution of the test, rather than using a projector lamp, magnifying lens, and indicator charts as described in ASTM C827 (see figure 13). The laser approach has been compared with the original setup, and similar results were obtained.⁽⁴¹⁾ The measured vertical distance corresponds to the increase or decrease in height (expansion or shrinkage) of the material laterally confined in the cylindrical mold from the time of molding to when the mixture becomes hard (i.e., final set).

Figure 12. Photo. Modified ASTM C827 setup (left) and change length in a hardened specimen with ASTM C1090 (right).^(39,40)

ASTM C1090

Two cylindrical specimens with the same dimensions as those used for the ASTM C827 test method were prepared in which the change in height was measured at four points on the top surface of the specimens using a micrometer in accordance with ASTM C1090 at 1, 3, 7, 14, and 28 d (see figure 12).⁽⁴⁰⁾ An initial four-reading measurement was taken right after placing a glass plate on top of the fresh sample surface. The glass plate was removed from the top of the test specimen after 24 h. After removal, the thickness of the glass plate was measured with a caliper at the points of contact between the glass plate and the micrometer. This thickness was then added to the four-reading initial measurement.

Figure 13. Illustration. The apparatus for early change in height adapted from ASTM C827.⁽³⁹⁾

In both test methods, there is always a certain degree of friction between the specimen's sides and the inner surface of the metallic mold. The degree of restraint varies with the mixture viscosity and degree of hardening. Though not recommended by the ASTM standards, and in order to provide the lowest friction possible, an acetate sheet was used in between the test specimen and the mold. The height change results in both test methods are expressed in terms of percentage increase or decrease of the original specimen height.

Autogenous and Drying Deformations

Autogenous deformation was assessed using an automated version of the sealed corrugated tubes test, as described in ASTM C1698.⁽³⁾ Three replicate specimens were evaluated concurrently. The tubes were placed over supports provided with spring-loaded linear variable differential transformers at each end that were connected to a data acquisition system. The displacement (converted to strain) was measured every 5 min for 7 d (see figure 14). The autogenous deformations were zeroed at the final time of set measured as described in ASTM C403.⁽³⁵⁾ To guarantee isothermal conditions, the specimens were kept in an environmental room at 73.4 ±1.8 °F (23 °C ±1 °C) and an RH of 50 percent ±5 percent. The mass of the samples was taken at the beginning and after 7 d of testing to confirm that the specimens were properly sealed. For later age measurements, four 1 by 1 by 12 inch (25 by 25 by 305 mm) prismatic specimens were prepared in accordance with ASTM C157, in which all four faces were sealed with two layers of aluminum tape after removal from the molds at 24 h.⁽²⁾ Similar samples were prepared without aluminum tape for shrinkage assessment in drying conditions (see figure 14). The specimens were kept in the same environmental room as the corrugated tubes. Length change measurements, as well as mass measurements, were taken every week for the first month and once a month for the next 6 mo.

Figure 14. Photo. ASTM C1698 tubes setup (top) and sealed and drying ASTM C157 specimens (bottom).^(3,2)

IC APPROACH

Introduction

Given the fact that cement-based grouts commonly exhibit shrinkage, this research also included additional tests focused on IC, which has become more popular during the last several years within the concrete community.^(18-20) As previously mentioned, there are several materials available for providing IC, including prewetted LWAs, SAPs, and prewetted wood fibers. In this study, prewetted LWAs were used.

Non-shrink cementitious grouts are often pre-packaged and can be extended using small aggregate for volumetrically large pours. IC can be thought of as an extension of the grouts using LWA rather than normal weight aggregate. The primary reason for using IC is to reduce shrinkage, especially during the first days when the tensile strength of the material is still low. In addition, this technology might be helpful in improving curing conditions in some locations where conventional (i.e., external) curing is difficult or impossible to implement, as well as in providing some robustness to the surface preparation (in terms of moisture content) of the precast (or existing) concrete elements since prewetted LWA also may serve as additional reservoirs if water is drawn from the grout into the substrate.

Mixture Proportioning with IC

The fine LWA used in this study consisted of rotary kiln expanded shale with a specific (dry) gravity of 1.56 and a 24-h water absorption of 16.95 percent by dry mass. The initial idea was to use the equation that Bentz et al. formulated based on the chemical shrinkage occurring in the sample (see figure 3).⁽²⁶⁾ However, as previously mentioned, C_f, CS, and α_max (as shown in figure 3) cannot be estimated unless the reactive content of the solid is known. Petrography was used to facilitate estimation of the reactive content. This technique uses a polarized light microscope to differentiate between crystalline and amorphous materials. In addition, cementitious materials have sufficient differences in their raw feeds, burning temperatures, mineral phases, and microstructure that it is possible to differentiate them, thus identifying particles such as cement, fly ash, and slag.

Two of the cementitious grouts were engaged within the IC portion of the study, namely G2 and G4. The petrographic analysis demonstrated that the cementitious contents were approximately 35 and 30 percent by mass, respectively. With this information and the known water contents, the theoretical w/b for each of the grouts could be calculated, resulting in w/b equaling 0.56 and 0.49 for G2 and G4, respectively. Consequently, α_max in figure 3 could be considered to be 1 because the w/b in both cases was above 0.42.⁽⁴²⁾

Likewise, chemical shrinkage of the two grouts was measured according to ASTM C1608, and the results were normalized by the amount of reactive material of each of the grouts.⁽²⁷⁾ The infinite chemical shrinkage could be then evaluated by plotting the chemical shrinkage over the inverse of the time, resulting in values of 1.99 and 2.15 fl oz/lb (0.13 and 0.14 mL water/g) binder for G2 and G4, respectively. (See chapter 4 of this report.) These values are high compared with plain cement (0.98 fl oz/lb (0.064 mL water/g), which is an indication of the presence of other cementitious materials that typically have higher values of chemical shrinkage.⁽³⁸⁾ Note that while petrographic images showed fly ash and slag particles, scanning electron microscope (SEM) images confirmed the presence of silica fume in both grouts. Figure 15 provides petrographic and SEM images that show fly ash, slag, and silica fume particles.

Figure 15. Photo. Images taken in G4 dry samples: cross-polarized light thin section micrograph with yellow arrows showing non-reactive sand particles (top left); plane-polarized light of the same thin section micrograph with red and green arrows showing cement and fly-ash particles, respectively (top right); SEM image with orange arrow showing a silica fume particle (bottom left); and energy-dispersive X-ray spectroscopic analysis confirming the siliceous nature of the silica fume particle (bottom right).

An appropriate IC design is crucial from the viewpoint of durability. In an overdosed system, if some LWAs remain filled with water, it may have detrimental effects from a freeze/thaw perspective. In this study, prewetted LWA were added to the base grout formulation at a mass calculated using figure 3 for each of the grouts. It is recognized that this addition of LWA will change the paste content of the formulation with respect to the total volume, thus having an influence on the shrinkage performance. However, it is determined to be a reasonable path given the lack of knowledge of the proprietary grout constituents. Table 3 shows the mixture proportions of G2 and G4 including IC.

Note: Mass of the solids corresponds to the mass of one bag.
^aTotal water including w/s mixing water and water that is absorbed by LWA.

1 lb = 0.45 kg.