Water treatment and granular flow lab

​

(a) Water treatment

​

       1. Development and testing of a LED-based spectrophotometer

        A device costing about US $ 63 for the parts was designed and fabricated at Norwich University, USA. It was used to estimate the concentrations of iron, manganese, and fluoride in water samples. The first two were estimated at Norwich University and the third at IISc. The performance of the device was comparable to that of four commercial spectrophotometers, even though its cost is lower by factors of 40 - 120. (Joint work with M.W. Prairie, S.H. Frisbie, A.H. Saksri, S. Parbat, and E.J. Mitchell.)

​

       2. Defluoridation of water

​

       Reject water from a reverse osmosis unit at Yellampalli village and also synthetic water samples were treated using adsorption. Activated alumina (AA) and a hybrid anion exchange resin with zirconia nanoparticles (HAIX-Zr) were used as the adsorbents. The performance of the adsorbents was fairly good for water containing only fluoride or a mixture of ions, and the cost of treated water was in the range Rs. 0.1/L - 0.4/L for AA and Rs. 0.2/L - 1.5/L for HAIX-Zr. However, for reject water, the cost increased significantly, as it was Rs. 1 - 1.5/L for AA and Rs. 11.5/L for HAIX-Zr. Further work is needed to examine how the cost can be reduced. (Joint work with M.V.V. Naga Samrat, J.R. Mudakavi, J. Riotte, and A. SenGupta.)

​

       3. Denitrification of water in a microbial fuel cell (MFC)

​

     It was shown for the first time that a consortium of seawater bacteria can be used to reduce nitrate in synthetic water samples in the cathode compartment of a MFC to nitrogen gas. The same consortium of bacteria were used in the anode compartment, along with acetate, peptone, and glucose. The power generated by the cell was higher when the commonly used phosphate buffer was replaced by bicarbonate buffer in the cathode compartment. Surprisingly, high concentrations of ammonium ions were found in both the compartments, but the reasons for this result are not clear. Further, water from the cathode compartment contains bacteria, and hence additional treatment is needed before it can be used for drinking. (Joint work with M.V.V. Naga Samrat, B. Ruggeri, and T. Tommasi.)

​

(b) Granular flow

​

      1. The lift force on a stationary disc immersed in a rotating granular bed

​

      The discrete element method (DEM) was used to examine this problem. The results suggest that the lift arises because of an asymmetry in the dilation (or decrease in solids fraction) and the velocity field above and below the disc. The occurrence of a lift is demonstrated visually by showing that a disc that is not constrained in the vertical direction rises as the cylinder rotates. (Joint work with B. Debnath and P. Nott.)

​

       2. Flow through channels and chutes

These flows are being studied using the discrete element method and continuum models. (Joint work with B. Debnath and V. Kumaran.)

​

(c) Thermodynamics

​

   Augmented Gibbs-Tolman model for surface tension: Gibbs modelled the interfacial region of a two-phase system by introducing a dividing surface separating the phases. He chose a particular dividing surface, called the surface of tension, which simplifies the expression for the internal energy and whose tension represents the surface tension sigma of the fluid. By considering expressions for the variation of the pressure in the transition region, Tolman (1948,1949) derived expressions for the sigma and the surface of tension. In the Gibbs-Tolman (GT) model, the unstable region in the pressure-density plane is collapsed onto a surface across which the pressure and density are discontinuous. Thus there is a physical surface of discontinuity separating the vapour and liquid phases and a surface of tension. By proposing simple expressions for the variation of the pressure in the two phases, Tolman derived expressions for sigma and the location of the surface of tension. In the augmented Gibbs-Tolman model (AGT), the same approach is used, except that the pressure variation is augmented by a term involving the spatial gradient of the density, as suggested by Rowlinson and Widom (1982). Using 4 adjustable parameters, it is found that the AGT model correlates data for 152 pure liquids and 57 liquid mixtures with an absolute average deviation in the range of 2 - 3 %. However, one of the length scales in the model turns out to be negative, for reasons that are not clear. (Joint work with S. Tumram and M.S. Ananth.)