Research
My research focuses on the optimization of powder/process metallurgy and metal additive manufacturing through computational simulation and physical modeling. My background in thermofluids and computational fluid dynamics (CFD) helps me to understand the complex fluid flow and heat transfer phenomena involved in industrial processes. My experience in the design of power and process plants and familiarity with industrial codes allows me to design and build simplified versions of industrial processes for research purposes. The following (clickable) list summarizes my recent research work:
Metal Powder Production
I'm a paragraph. Click here to add your own text and edit me. It's easy.
Background
Water atomization (WA) of molten metals is a cost-effective industrial approach to producing metal powders for powder metallurgy and metal additive manufacturing. In WA, a stream of molten metal is atomized using high-pressure water sprays to form a metal powder. The main issue in the current operation is the large fraction of over-sized particles that must be discarded or recycled, thus decreasing production yield and increasing the energy consumption and carbon footprint.​
During my PhD studies, I used my expertise in experimentation and modeling to examine some of the poorly understood thermofluidic phenomena involved in WA, and contributed to various aspects of this field, as follows.
Fig.1 (Left) Lab-scale water atomizer designed and built at UofT
(right) Bi-Sn powders produced by the water atomizer
Lab-Scale Powder Production
A lab-scale water atomizer (Fig.1 - left) was designed and built at the University of Toronto (UofT), with flexibility for adjusting the design and operating parameters, and with special features for visualization. Bismuth-tin alloy powders (Fig.1 - right) were produced in the atomizer at various conditions and characterized using sieve analysis. Through parametric studies, correlations between the design and operating conditions of a water atomizer and the particle size distribution of the resulting powder were developed, allowing powder producers to improve their powder quality by design and process optimization. Some of the research outcomes have already been implemented at the Rio Tinto Metal Powder plant and have proven effective in increasing the production yield.
Visualization of Water Sprays
In an effort to understand the process of WA, an image feature consolidation technique (IFTC) was developed and experimentally validated to facilitate the characterization of water sprays. This technique utilizes high-speed shadow imaging using two different optical setups (a high resolution and a high depth-of-field setup) to capture images of a spray. Machine learning and image processing techniques are then employed to obtain the droplet size distribution of the spray from the two sets of shadow images. More information can be found in [1]; the graphical abstract is shown below.
Fig.2 Graphical abstract for IFCT [1]
Besides spray characterization, novel features of a water spray were identified and quantified using experiment and CFD simulation. A simple mathematical model for the momentum transfer from a water spray to a molten metal stream was developed.
The breakup of a Turbulent Flat Fan Liquid Sheet
In WA, water sprays are the main drivers of molten metal disintegration. These are high-pressure and turbulent flat fan sprays. Thus, the breakup of a flat fan liquid sheet at near industrial conditions, i.e. high pressure and turbulent, was investigated; the following are the main outcomes:
-
Through high-speed shadow imaging (Fig.3 - left), three modes of sheet breakup were identified in the central region of a turbulent sheet; one of them, bag breakup, has not been reported before.
-
It was concluded that existing flat fan atomizer models cannot precisely predict the characteristics of a turbulent flat fan spray, as each of them resolves one breakup mode only.
-
A CFD model for the simulation of the sheet breakup was developed and resolves all breakup modes (Fig.3 - right).
-
Using the CFD results, the formation of a bag was related to flow instabilities emerging from the nozzle.
More information about this work can be found in [2]
Fig.3 Breakup of a turbulent flat fan liquid sheet; (left) High-speed imaging and (right) CFD simulation) [2]
Acknowledgment
The research on water atomization was conducted at the Process Metallurgy Research Lab (PMRL) and sponsored by Rio Tinto Metal Powders. Additional financial support received from other funding agencies including NSERC. This project would have not been possible without the support of Kinnor Chattopadhyay and Markus Bussmann. Also thanks to many of my labmates at PRML for their assistance in the construction and operation of the water atomizer (I will post some pictures on this soon!). The CFD simulation was a collaborative work with Martin Heinrich and Rüdiger Schwarze from the Freiberg University of Technology.
Metal Additive Manufacturing
Background
Metal Additive Manufacturing (metal AM) is an emerging technology that uses metal powder to build net shape or near-net shape parts, and so it is intrinsically sustainable. In this novel manufacturing process, metal powders are bonded by laser, electron-beam, binders, or (recently) LED light to make high-precision parts. Regardless of the method and details of bonding the metal particles, all AM processes start with metal powder. Therefore, metal powder production and metal AM are linked together, and the quality of the powder defines the quality of the final AM part.
Fig.4 3D printing of optimization cubes using low alloy steel powder
Technique: Laser Powder Bed Fusion (LPBF), aka Selective Laser Melting (SLM)
Optimization of Metal 3D printing
In addition to the powder quality, the operating conditions of a metal 3D printer directly affect the quality of the final part. Each metal AM process involves several operating parameters. For instance, in laser powder bed fusion (LPBF), the main parameters are laser power, scanning speed, hatch spacing, and layer thickness. These parameters need to be adjusted/optimized in accordance with the expectations of the final part and the characteristics of the powder being used. A common approach to optimizing the operating parameters of a 3D printer is the parametric study.
In a recent collaborative project, we studied the effect of LPBF operating parameters on the density and strength of the final 3D printed parts. Four (in-house produced) low-alloy steel powders were used, and several optimization cubes were additively manufactured at various laser powers, hatch spacings, and scan speeds (Fig.4). The density of each cube was measured by the Archimedes method, and the best three candidates were further examined for internal pores or defects using a nano-CT scanner. Another set of parts were 3D printed at the operating conditions of the highest density cubes and went through the tensile test. The part with the highest strength reflected the optimized operating parameters.
​
Acknowledgment
The research on LPBF parameters optimization was conducted at the Process Metallurgy Research Lab (PMRL) In a collaboration with researchers from the Institute for Iron and Steel at the Freiberg University of Technology in Germany. The 3D printing was done in the Multiscale Additive Manufacturing Laboratory (MSAM) at the University of Waterloo.
Process Metallurgy
Background
Process metallurgy involves extracting metals from their ores, and to some extent shaping them. An important process in this field is the continuous casting of molten metals (especially steel) into semifinished shapes like billets or slabs. In a continuous casting mold, argon gas is injected into the molten steel through a submerged entry nozzle (SEN) to deter nozzle clogging with solid inclusions. It also affects the flow pattern in the nozzle, and subsequently in the mold, and improves the removal of inclusions from the molten steel to a slag layer at the top of the mold.
Fig.5 Physical modeling of argon gas injection into a continuous casting mold
using water-air experiments
Argon gas injection in a continuous casting mold
A common approach to study steel-making and -shaping processes is water modeling. In a recent collaboration with a team of process metallurgists at PMRL, we characterized the formation and size distribution of argon bubbles injected into molten steel in a continuous casting mold. We built a 1/8 scale water-air model and applied the optical imaging technique that was developed earlier [1] to capture the multi-range bubble size distribution (Fig.5). This research contributes to the improvement of steel quality by improving inclusion removal and reducing bubble-caused defects.
​
Acknowledgment
The research on argon injection was conducted at the Process Metallurgy Research Lab (PMRL), the main collaborator was Amiy Srivastava.
Thermofluids and CFD
Background
Thermofluids is a branch of science that encompasses several intersecting fields including heat transfer, thermodynamics, and fluid mechanics, and applies to many processes that involve transport phenomena including mass, momentum, and energy transfer in fluids. Physical modeling and Computational Fluid Dynamics (CFD) are the most common methods to study thermofluidic processes and are often used side by side to complete each other. Much of my research includes experimentation and simulation of such processes; what comes next are a few examples.
Droplets Impingement
In water atomization of molten metals, multiple sprays of water impinge onto the surface of a molten metal stream. This is a rather complex phenomenon and may be simulated in steps. In this study, we took the first step towards modeling the impingement of two water sprays; two broken jets of water were simulated using OpenFOAM software and compared with the experimental videography (Fig.6).
Acknowledgment
This is a collaborative work with Martin Heinrich from the Freiberg University of Technology.
Fig.6 Impingement of multiple water droplets (a simplified version of spray impingement); (left) a schematic of water atomization of molten metals, (middle) experiment, and (right) CFD simulation
Rapid solidification (quenching) of a molten metal droplet
To control the atomization of molten metals and to improve the quality of metal powders, one needs to investigate the simultaneous deformation and solidification of a molten metal droplet. A fundamental study was performed by dropping a small droplet of liquid metal into a bath of water. High-speed shadowgraphy was used to visualize the droplet impingement and deformation, and the process was simulated using CFD (Fig.7).
​
Acknowledgment
Thanks to Ziqi Tang (from PMRL) and Sebastian Borrmann (from the Freiberg University of Technology in Germany) for their help on this project.
Fig.7 A liquid metal droplet falling into a bath of water; (left) experiment, (middle) CFD simulation, and (right) the final solidified shape.