Danesh Tafti

Prof. Danesh Tafti


Research Focus & Metrics

Academic Career Length

23Years (2004-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 November 2018 in Computers & Fluids
Verified
Journal Article
Closed access
ZCZe Cao
Ze Cao
DTDanesh Tafti
Danesh Tafti

28 February 2006 in International Journal of Heat and Fluid Flow
Verified
Journal Article
Closed access
AVAroon K. Viswanathan
Aroon K. Viswanathan
DTDanesh Tafti
Danesh Tafti

28 February 2010 in International Journal of Heat and Fluid Flow
Verified
Journal Article
Closed access
MEMohammad A. Elyyan
Mohammad A. Elyyan
DTDanesh Tafti
Danesh Tafti

13 July 2005 in Journal of Fluids Engineering
Verified
Journal Article
Closed access
PDF file
AVAroon K. Viswanathan
Aroon K. Viswanathan
DTDanesh Tafti
Danesh Tafti
Detached Eddy Simulation (DES) of a hydrodynamic and thermally developed turbulent flow is presented for a stationary duct with square ribs aligned normal to the main flow direction. The rib height to channel hydraulic diameter (e∕Dh) is 0.1, the rib pitch to rib height (P∕e) is 10 and the calculations have been carried out for a bulk Reynolds number of 20,000. DES calculations are carried out on a 963 grid, a 643 grid, and a 483 grid to study the effect of grid resolution. Based on the agreement with earlier LES computations, the 643 grid is observed to be suitable for the DES computation. DES and RANS calculations carried out on the 643 grid are compared to LES calculations on 963∕1283 grids and experimental measurements. The flow and heat transfer characteristics for the DES cases compare well with the LES results and the experiments. The average friction and the augmentation ratios are consistent with experimental results, predicting values within 10% of the measured quantities, at a cost lower than the LES calculations. RANS fails to capture some key features of the flow.

1 July 2023 in Physics of Fluids
Verified
Journal Article
Open access
PDF file
NRNeil Ashwin Raj
Neil Ashwin Raj
DTDanesh Tafti
Danesh Tafti
NMNikhil Muralidhar
Nikhil Muralidhar
Three reduced order models are evaluated in their capacity to predict the future state of an unsteady chaotic flow field. A spatially fully developed flow generated in a random packing of cylinders at a solid fraction of 0.1 and a nominal Reynolds number of 50 is investigated. For deep learning (DL), convolutional autoencoders are used to reduce the high-dimensional data to lower dimensional latent space representations of size 16, which were then used for training the temporal architectures. To predict the future states, two DL based methods, long short-term memory and temporal convolutional neural networks, are used and compared to the linear dynamic mode decomposition (DMD). The predictions are tested in their capability to predict the spatiotemporal variations of velocity and pressure, flow statistics such as root mean squared values, and the capability to predict fluid forces on the cylinders. Relative errors between 15% and 20% are evident in predicting instantaneous velocities, chiefly resulting from phase differences between predictions and ground truth. The spatial distribution of statistical second moments is predicted to be within a maximum of 5%–10% of the ground truth with mean error in the range of 1%–2%. Using the predicted fields, instantaneous fluid drag force predictions on individual particles exhibit a mean relative error within 20%, time-averaged drag force predictions to within 5%, and total drag force over all particles to within 1% of the ground truth values. It is found that overall, the non-linear DL models are more accurate than the linear DMD algorithm for the prediction of future states.

17 February 2016 in Journal of Engineering for Gas Turbines and Power
Verified
Journal Article
Closed access
PDF file
SSSukhjinder Singh
Sukhjinder Singh
DTDanesh Tafti
Danesh Tafti
Sand transport and deposition is investigated in a two-pass internal cooling ribbed geometry at near engine conditions. Large-eddy simulation (LES) calculations are performed for bulk Reynolds number of 25,000 to calculate flow field and heat transfer. Constant wall temperature boundary condition is used to investigate the effect of temperature on particle deposition. Three different wall temperatures of 950 °C, 1000 °C, and 1050 °C are considered. Particle sizes in range 5–25 μm are considered. A new deposition model which accounts for particle composition, temperature, impact velocity and angle and material properties of particle and surface is developed and applied. Calculated impingement and deposition patterns are discussed for different exposed surfaces in the two pass geometry. Other than the leading rib faces, the highest particle impingement and deposition is observed in the bend region and first quarter of the second pass. Significant deposition is observed in the two pass geometry for all three wall temperatures considered. Particle impingement and hence deposition is dominated by larger particles except in the downstream half of the bend region. In total, approximately 38%, 59%, and 67% of the injected particles deposit in the two passes, for the three wall temperatures of 950 °C, 1000 °C, and 1050 °C, respectively. While particle impingement is highest for wall temperature of 950 °C, higher deposition is observed for 1000 °C and 1050 °C cases. Deposition increases significantly with wall temperature. For 1000 °C, roughly 12% of the impacting particles deposit. For 1050 °C, approximately 23% of the particles deposit on impact. For all the three cases, the second pass experiences higher deposition compared to the first pass due to higher turbulence and direct impingement.