This chapter discusses nonconventional methods for the processing and net-shape forming of entire concrete systems (i.e., cement, aggregates, and reinforcements) in order to modify structure and composition and enhance properties and performance. The first section deals with process control, which includes orchestration of the procedures needed at the point of production (e.g., mixing techniques and concrete transportation) to obtain the desired product. The second section discusses materials testing and quality assurance, including the need for improved testing procedures to permit better decision making throughout the production process and particularly at the point of placement. The third section investigates placement methods, especially if the concrete contains fibers or intrinsic reinforcements that might reduce the amount of labor needed to place rebar.
The committee concludes that the following issues should be taken into consideration when examining potential nonconventional concrete processes: (1) processing requirements that correspond to new material developments; (2) processing and constructability constraints that influence the success of a construction material; (3) scale-up possibilities to attain acceptable performance and required amounts from novel process plants for expected construction scenarios; (4) robustness and simplicity of the process controls that are needed in a production facility to produce the performance variables and strengths desired; and (5) parameters that influence the placing, finishing, and curing requirements of new concrete material (e.g., flow and workability, resistance to mix separation under a variety of conditions, tolerance of environmental conditions, and the need for and interaction with a range of typical reinforcing materials).
The production processes of conventional and nonconventional concrete involve the selection and preparation of source materials, the mixing and blending of raw materials (including cement binders) in controlled amounts, the delivery of blends either as slurries or dry mixes, and the placement of the mixes, usually as slurries, into forms or machines for final net-shaping. Every aspect of the production process must be controlled to optimize performance and properties, and higher levels of process control than are currently available will be required for nonconventional concrete technologies. Four major aspects of processing require improved control: feedstock production, hydration, concrete mixing, and fiber mixing.
Control of Cement Feedstock Production
Even apparently small variations in the dry cement feedstock (e.g., chemistry, particle size) can cause large changes in the resulting concrete. Essential is an understanding of the effects of materials variability on the course of reaction so that the quality control factors can be specified. The initial procedures must be controlled in order to provide a known base for the process activities that follow. In Denmark, for example, the cement industry aims at producing a uniform cement strength at a specified age (e.g., 7 and 28 days). This is accomplished by intensive quality assurance efforts in feedstock production and through close attention to fineness, size distribution, and grinding temperature during the grinding process. The concrete producers apply tight rheological controls of the cement and cement paste to achieve a consistent and conforming end-product (Henrichsen, 1996).
If clays are to be added to the mix, as proposed in Chapter 2, they can be added dry or as slurry. It should be noted that ASTM specifications set limits to the amount of clay that can be beneficially used in concrete. Popovics (1992) discusses some of the potential problems. The heterogeneous reactions of clay with liquids, for example, are difficult to control. The most common method is for the clay to be predispersed in a slurry, as is the case for the micro-silica additions that are used to improve the matrix. Modified clay can also be mixed with a plasticizer. A nonwetting agent could be used to improve the mixing of fine powders. These processes might benefit from the separation of the reactive part of the mixing from the rheological part of the placement.
The slump test measures the consistency or mobility of the slurry and is the main method of determining the workability of concrete before placement. In the slump test, fresh concrete is placed in an inverted open-ended steel cone that is set vertically on a steel base with the large end of the cone at the bottom. All cone dimensions and filling procedure are carefully detailed and specified in the ASTM C–143 Standard. After filling the cone and striking off the top surface of the concrete, the operator lifts the steel mold off, leaving a cone of concrete that will them slump (settle or flow) to varying degrees, depending on its consistency. The difference in height in inches between the slumped concrete and the steel cone (initial state) is the slump value, which is then used to characterize that mix. A greater slump indicates a more mobile, easier flowing mixture. Many parameters of the mixture affect the slump in a nonlinear manner, however.
Accurately measuring the water content of the aggregates in bulk is a persistent problem in the production of concrete and can cause mix deviations and placement problems. This problem is compounded by the difficulty in accounting for the various shapes found in huge quantities of aggregates in fine-tuning mix designs and blend behaviors.
The ceramics industry has developed numerous devices for the preparation and mixing of powders, such as twin-vee mixers, attritor mills, and ball mills. One novel approach to cement feedstock production would be to ball mill the powders, using the aggregate itself as the milling medium. Thus, the mixing media would not have to be separated from the powder at the conclusion, and the mixing balls would be free. A disadvantage to dry ball milling in conventional concrete is that the rough aggregate would be rounded somewhat. In a nonconventional concrete, however, this type of milling might lead to better pouring properties and less particle–particle interaction among the chunks of aggregate.
Control of the Hydration Process
As stated in Chapter 2, the sol-gel transition of the conventional cement-matrix has advantages and disadvantages for processing. The
advantage is that although there is some change in the consistency of concrete, particularly due to such factors as absorption of water by the aggregate, loss of water by evaporation, etc., the rheology is relatively insensitive to the extent of the hydration reaction until the material is very near the gel point, making the material physically insensitive from the time of mixing to the time of placement. When viewed as a chemical process, however, the sol-gel transition is a disadvantage because the degree of hydration at the time of placement is unknown and uncontrolled. Part of the problem is that the hydration process has several phases: the saturation of the soluble components of the cement powder with water, the development of a hydration layer on the surface of the cement grains as they dissolve, and the development of the gel phase that “precipitates” from the supersaturated region at the water-hydration layer interface. There are several approaches that are incorporated into current practice or at least partly known in the laboratory that could provide the needed amount of control:
On-site mixing of the concrete would permit closer monitoring and control of the time of the hydration reaction.
Electrochemical or electrorheological sensors could be used on-site to monitor the extent of reaction so that the material could always be poured at some well-defined point in the reaction process.
Viscosity modifiers could be added on-site to make the working characteristics of the concrete more uniform and more independent of the extent of hydration.
Uncertainties in the hydration reaction rate could also be reduced by using such nonconventional techniques as:
Presaturating the process water with certain soluble components before adding it to the cement powder.
Pretreating the powder itself, perhaps with steam, to ensure more uniform initiation of hydration (caution would be needed to prevent excess hydration).
Adding small amounts of nucleating agents for the crystalline phases, such as ettringite, to eliminate such uncertainties as the induction period and the critical supersaturation levels for nucleation (prehydrated cement has been used in some laboratory studies as a nucleating agent to accelerate reaction).
Controlling temperature either physically by insulating the mixing equipment or chemically by adding some type of material that undergoes an exothermic or endothermic reaction (see Chapter 2).
Concrete Mixing Control
The main instrument used at the fabrication plant for the batching of concrete components is the drum mixer. Drum mixers rotate at variable speeds and contain interior vanes to blend the ingredients by shearing and mixing actions. The drums on the trucks that deliver concrete to the construction site also contain these features. Although large-scale fabrication plant and small-scale truck drum mixers have been used for years, much remains unknown about their operation and effectiveness.
Two major matters of concern in mixing are the development of uniform slurries or mixes and the prevention of agglomerations of particles. Little is known about the proper particle size and the breakdown of particle groups for optimum mixing of the cement feedstock or about the changes in the mix-water content that fluidizes the mixture and eventually reacts chemically with the cement. High-range water reducers may be used to disperse the cement particles and prevent flocculation. Smaller cement particles hydrate more uniformly and can improve the resulting concrete, but agglomeration problems may occur if the particle size is too fine. Agglomeration may also occur when cement is used that has been stored too long, although this problem is well known by batch-plant technicians and is usually averted by proper quality assurance procedures.
In the polymer and ceramics industries, control of reproducible mixing is one of the most difficult yet important processing variables. It should not be taken for granted in concretes either. Hence, resolution of the difficulties of blending materials would greatly improve the entire control process. Research is required to obtain more detailed knowledge about what occurs during mixing and whether the rolling drum is the best way to mix concrete. Nonconventional concrete may require higher intensity mixing and improved mixing processes to ensure better products.
Continuous processes generally yield more consistent products than batch processes. If both dry and wet mixing could be performed as one continuous process, the results could be continuously
monitored and the input stream modified to produce a more consistent product. Such a process would be ideal for the application of statistical process control techniques, especially if it could be done on-site, and would obviate the need to find other uses for off-specification products.
The techniques and experience of other industries that must measure and mix ingredients to get a consistent product in large quantities (e.g., ceramics and paint producers) should be examined to determine their applicability and relevance to the concrete industry. Indeed, some ceramics techniques are already being used in the concrete industry. For example, the high-intensity mixing used for shearing agglomerates in the ceramics industry is currently being used to prepare the concrete for nuclear facility construction in Europe (Henrichsen, 1996). The rheological properties of cement pastes and concretes have been studied extensively in the laboratory, primarily the parameters of viscosity and yield stress. Roy and Asaga (1979), Chappuis (1991), and Yang and Jennings (1993) showed the effect of shear mixing rate and time on the yield stress of cement paste and concrete; the yield stress was markedly decreased by high shear rates.
Control of Fiber Mixing
The placement of discontinuous reinforcing fibers in concrete presents several processing issues. First, special processes are needed to disperse the fibers, especially if fiber content is to exceed one volume-percent. Second, the addition of fibers to the mix causes changes in both the workability and the rate of strength gain, often because other additives are needed to allow the fibers to co-mingle successfully (see Chapter 3). Third, assuring the quality of a mix full of chopped fibers is difficult, and process controls to avoid clumps of fiber sticking together are critical. Hence, the use of reinforcing fibers in the matrix must be given careful consideration in the rheometric procedures for nonconventional concrete technologies.
There are precedents for using a steel mesh and cement for some articles, and the use of chopped fiber in plastic boat building is well known. After shearing of the fiber bundle is done successfully, the chopped fiber is sprayed and combined with resins to create fiberglass hulls. An analogy in the construction industry would be the use of fiber-reinforced gunite that is sprayed onto surfaces for special
building applications (Hannaut, 1978). Technology has also been developed for the extrusion or disbursement of fiber in a slurry as it is moved or poured, which is then dewatered by vacuum on a moving belt and formed into thin sheets that can be layered together, much like plywood, to achieve thicknesses up to 30 millimeters (Majumdar and Laws, 1991).
Woven fibers may be preferred in some instances because they can be seen and thus placed accurately before being infiltrated with a cement slurry. There are methods for using three-dimensional textiles, such as geotextiles, that form a continuous-space net for reinforcement. All of these technologies will demand accurate rheometrics to ensure proper placement and to achieve the expected properties and performance.
MATERIALS TESTING AND QUALITY ASSURANCE
Quality assurance is a key activity for any materials use, including concrete. It is important to be able to perform realistic and routine tests to provide specific information about the quality of a material, especially when current experience is superseded by the requirements of a new material. Thus, quality assurance will be critical when considering the delivery of a nonconventional concrete from a process facility to the precasting yard or construction site. The current tests performed on conventional concrete are listed in Appendix C. The critical problems that lend themselves to basic research will require identification, however. Standards organizations will also need to be involved in expediting the protocols for new materials.
The emphasis of quality assurance procedures should be on the use of nonconventional test programs that: (1) are seamless throughout the entire process; (2) improve the availability and reliability of the data collected during the production and construction process for use by all parties; (3) ensure the delivery and use of the proper raw concrete or concrete elements; and (4) allow the implementation of model-based design for the concrete system within an MSE-systems approach.
For example, the predominant method of determining the workability of concrete before placement is the slump test. Workability is the property that determines a concrete slurry's capacity to be placed, consolidated, and finished without segregation of its components. Although the slump test may assure a certain level of quality, the
variables and parameters that influence the test results are so numerous and difficult to quantify that it takes experience and insight to determine why a material failed a test and how to correct the problem. Water is often added to obtain a better result, which is frequently the worst option available. The complex process dependency included in the slump measurement is a fruitful topic for further research, and the parameters controlling slump, including time between batching and placement, need to be clearly identified and understood. Some research on theology is being conducted in the U.S. and France. For example, Banfill (1993) summarized the relation between slump and yield stress as measured by a more complex rheometer for concrete. The yield stress for concrete is about two orders of magnitude greater than that of cement paste.
Even with further improvement, however, the slump test will not provide sufficient information to allow the formulation of optimum decisions or the implementation of model-based design within an MSE-systems approach. As stated above, process control and test methods should be performed continuously so that (1) variations from desired performance can be promptly identified and rectified, and (2) new products can be designed and manufactured faster and cheaper with higher quality. Research is urgently needed to develop advanced test methods and sensors that better describe the dynamics of concrete's development in a continuous fashion.
Electrochemical and electroacoustical methods could be developed to measure the extent of chemical reaction and the progress of gelation. High-frequency electroacoustic measurements could probe the relatively short mechanical relaxation times and might be able to detect the onset of gelation before it becomes visually apparent. Electrical conductivity and dielectric loss measurements could also be used to monitor the extent of chemical reactions. There are also numerous nonintrusive sensing technologies (e.g., neutron scattering to measure water content and the degree of binding of water [Livingston et al., 1995]) that might be adapted to cement processing.
Furthermore, the electrical and dielectric properties of Portland cement and other cementitious materials are important diagnostic properties during curing and hardening. Gorur et al. (1982) were among the first to use microwave equipment (9 GHz) to monitor the hydration process continuously, and they found a nearly exponential decay in the complex dielectric constant as hydration took place (Figure 1-17). McCarter and Curan (1984) and Perez and Roy (1984)
attempted to interpret cement's electrical behavior in terms of various conduction and polarization mechanisms. Perez-Pena (1986) and Perez-Pena et al. (1989) showed that the function of the electrical conductivity or relative dielectric permittivity determined at low frequencies (approximately 1 KHz) had a complementary relation to the heat produced during hydration. Electrical conductivity diminished as the heat released increased. This finding enabled a convenient nondestructive means of evaluating the hydration and hardening process. It also made it possible to assess the effects of various chemical admixtures and additives on the hydration process. At the other extreme in the hardened materials, the relative permittivity and dielectric loss diminished with increase in frequency and were found to be very low in the MHz region for low w/c pastes, especially in the microwave range up to 200 MHz (Perez-Pena et al. 1986, 1989). Other work has resulted in the use of microwave heating to rapidly determine the water content of concrete (Naik and Ramme, 1987).
Extensive research is currently being conducted in other fields to develop sensors to measure chemical and physical parameters for extreme environments (NRC, 1995a, 1995b; see Appendix B). This work should be examined to determine its applicability and relevance to concrete.
Batch mixing and concrete placement are usually separated by both space and time. There is time between the mixing of the constituents until the plastic concrete is placed in the forms or machines to take on final net-shape. The stiffness of the concrete is increasing during this interval, and the workability of the mix is decreasing. These changes must be taken into account when carrying out the activities (e.g., consolidation and finishing) that take place after the constituents of a concrete are combined. The time period prior to setting might be lengthened (e.g., to allow for long-distance delivery) or shortened (e.g., to accelerate strength gain). Thus, the placement methods are vitally related to the workability of new concrete and its changes with time.
The physical changes in concrete are due to gel formation that causes the plastic concrete to be somewhat thixotropic (Figure 1–17). Colloidal silicate pastes, which exhibit high thixotropy, show big changes in their shear rates, depending on conditions. Thixotropy is
a difficult phenomenon to take into account and needs to be better understood. Thus far, researchers have been unable to obtain uniform results from sol-gel experiments in the laboratory, where control of the experiment is optimum. The ability to probe and continuously determine free- and bound-water concentrations in the microstructure (e.g., neutron scattering or microwave methods; see above) may prove useful for handling workability issues as well as for predicting the final properties of the material.
Better tests are also needed to indicate performance during placement, especially in the area of advanced rheometrics. Comparable tests exist in other areas that might be adaptable to concrete:
Shear viscosity testing is routinely used in the paint industry.
Centrifuge methods are used to determine the particle size distribution of insoluble particles for analytical measurements of suspensions.
Other viscometric measures are available that might be considered for the rheometrics of concrete—for instance, suspensions could be added to modify flowability.
For conventional concrete, the slurry is delivered to the forms or placing machines, where it is poured or pumped around steel reinforcement or prestress tendons that are already in place. All of the aforementioned process controls will have a new measure of complexity if the reinforcement is included in the concrete itself in the form of small fibers. Such reinforcement is already a trend in conventional concrete. In Canada, for example, bridge decks are being reinforced with fiber, thereby eliminating corrosion of the rebar and consequent deterioration of the deck.
A potential nonconventional placement method would be the on-site production of nonconventional concrete by combining individually stable material streams, one of which would be an activator liquid that would initiate the gelation process. Such a method would eliminate the time and distance variations typically found in conventional concrete.
As stated above, there is difficulty in making measurements during the concrete production process that will allow critical control in achieving the desired performance specifications. Once the trial mix has been designed in the laboratory and tested full-scale, the process is regarded as largely fixed and is controlled mainly for obvious deviations. The development and advent of new rheometrics will not
only enhance the delivery, placement, consolidation, and finishing of concrete, it may also promise a new measure of control to meet target properties and performance specifications. Applying more advanced rheometrics raises the new problem of analyzing a host of parameters, however, all of which vary simultaneously. Thus, advanced statistical approaches will probably be required to deal with this new problem (see Chapter 5).