Conventional methods of measuring concrete workability such as the slump test provide a broad an indication of the amount of work required to compact the concrete mixture. With the advent of more fluid concretes (pumpable concrete, self-levelling concrete), it was necessary to measure the flow properties of concrete. A two parameter approach, based on rheological principles, was proposed by Tattersall1.
Rheology measurements on concrete indicate that it is reasonable to approximate the concrete flow behaviour using a Bingham model1,2,9. The Bingham model, along with other common rheological models, are shown in Fig 1. Shear yield stress (y-axis intercept), τo, indirectly measures inter-particle friction, while the viscosity µ (slope of the line) depends on the rheology of the paste and the volume fraction of aggregates. The shear yield stress for SCC is very low (0 – 60 Pa as compared to a couple of hundred Pa for normal concrete) and the plastic viscosity is highly variable, ranging from 20 to 100 Pa.s3. A few more models have also been proposed to account for the non-linearity in the flow behaviour, such as the Herschel-Bulkley model4. Other types of flow behaviour are also possible, as indicated in Fig 1. ‘Shear thinning’ means that the rate of increase of shear stress slows down with increasing shear strain rate; the opposite is true for ‘shear thickening’ behaviour. On the other hand, pseudoplastic behaviour, exhibited by some VMAs such as water-soluble high molecular weight polysaccharides (Welan gum and Diutan gum), means a drop in shear stress at high shear rates. Thixotropy defined as the property exhibited by certain gels of becoming fluid when stirred or shaken and returning to the semisolid state upon standing5 would also be equivalent to the pseduoplastic behaviour.
Designing rheometers for concrete using standard arrangements such as parallel plate geometry or coaxial cylinders is not easy because of size effects related to coarse aggregates. In spite of this limitation, a few concrete rheometers have been designed and successfully used by researchers6,7,3. The ConTec BML viscometer developed in Iceland3 is a coaxial cylinder rheometer suitable for both dilute suspensions of cement paste and for stiff concrete mixes. The BTRHEOM has also produced reliable results with concrete mixtures6. Generally, it is observed that rheometers give reliable results when the yield values of concrete (and also the ratio of yield stress to plastic viscosity) are low.
(a) Bingham model
(b) Other rheological models
Fig 1. Common rheological models for fluids
It might take a long time for the use of rheometers to become widespread in concrete- related applications. However, available data shows that measurements from rheometers can provide reliable information about the nature of self-compacting concrete. In particular, it can help in understanding the complexity of the cement-superplasticizer-VMA system. For example, in a rheometric study of SCC equivalent mortars, Ghezal and Khayat8 were able to show that the efficiency of VMA depends on the class of superplasticizer used (that is, whether it was a copolymer, polysulphonate, or polycarboxylate). Rheological characteristics such as shear thinning or pseudoplasticity of the system could also be measured using the rheometer. Apart from resolving the compatibility issues, rheometer measurements are also useful in evaluating the effect of other factors, such as type and amount of filler, other admixtures (such as air entrainers, accelerators, etc.) on the flow of self-compacting concrete.
The application of rheology to mixture proportioning of concrete has been attempted by some researchers3,9. Wallevik3 proposes that the influence of mix constituents on rheology of the resulting concrete should be studied using a rheometer. This data can further be used to adjust the mixture proportions to produce the desired degree of flow and segregation resistance in concrete. For example, as shown in Fig 2, compared to a reference concrete, a stiffer concrete could be produced by increasing the yield stress, while a wetter concrete could be produced by decreasing both yield stress and plastic viscosity. Moreover, silica fume would initially reduce the viscosity (since the fine particulate content would increase the fluidity of the paste), while higher dosage would cause increase in the yield stress as well as viscosity. Based on rheological studies Domone10 and Wallevik3 were also able to suggest the desired rheological properties for various types of concrete, as shown in Fig 3.
(a) Qualitative description of concrete rheology
(b) Effect of air, water, and additives on rheological parameters
Fig 2 Concrete rheology and the effect of additives 10,3
Note: ‘Ref’ indicates reference concrete; the X-axis refers to plastic viscosity, and not to the shear strain rate as in Figure 1
Fig 3 Ranges of rheological
properties for normal concrete and SCC10,3
Saak et al.11 proposed a method based on rheological principles to control segregation both in static and dynamic conditions. In static condition, the heavier aggregate particle tends to settle, and this tendency can be controlled ensuring a minimum yield stress, τo, for the paste. Dynamic control of the settling particles can be achieved by minimising its terminal velocity, which depends on the plastic viscosity µ. To make these results widely applicable, the rheological parameters τo and µ were normalised by dividing by Δρ (the difference in densities of aggregate and paste), and ‘self-flow’ zones for SCC were delineated on a graphical plot (Fig 4).
Saak et al.’s recommendation is based on the generalization
that there is not much variation in SCC mixtures produced around the world,
in respect of paste volume fraction and aggregate distribution. Paste
rheology is all that requires to be adjusted and this can be done using
the ‘self-flow zone’ concept. Fig 4 shows three zones, as
marked by Saak et al.11, one each for plain cement paste, cement paste
with silica fume, and cement paste containing silica fume and a cellulose
derivative. Only a few experiments will be required to arrive at a suitable
paste composition that will fall in one of the self-flow zones.
Fig 4 Rheological self-flow zones11
Dynamic control of segregation of SCC can also be done by means of an effective VMA. Certain VMAs, such as water-soluble polysaccharides (Welan gum or Diutan gum), exhibit a pseudoplastic rheological behaviour12. As explained in the previous section, such materials show a decrease in shear stress at high shear rates. In other words, when the concrete is flowing, the viscosity of water with such VMAs will be low, causing virtually no hindrance to the flow of the concrete; when the concrete comes to rest (this is when there is maximum danger of segregation), the viscosity of the mixture due to the nature of VMA shoots up, which helps in keeping the mixture stable.
This site is best viewed in 1024 x 768 resoulution