Manash J Konwar

In this project work, proper emphasis is given to experimental, numerical as well as simulation work to understand the operation of a solar flat plate collector and proper correlation is made among all these data to get a proper validity with real life situations. The study involves the temperature and pressure variation study of fluid along the length of pipe which helps us to calculate the overall thermal coefficient for the specific setup and also to determine the Nusselt number of the flow… Show more

In this project work, proper emphasis is given to experimental, numerical as well as simulation work to understand the operation of a solar flat plate collector and proper correlation is made among all these data to get a proper validity with real life situations. The study involves the temperature and pressure variation study of fluid along the length of pipe which helps us to calculate the overall thermal coefficient for the specific setup and also to determine the Nusselt number of the flow in the pipes. Moreover, a detailed parametric study is also done by varying the air gap thickness between the glazing surface and absorber so as to optimize the heat loss by minimizing it for the proper constraint. The solar flat plate collector is simulated for different types of fluid medium both inside the pipes and in the air gap medium and at last a comparative study is done to determine the best setup so as to minimize the heat loss from the collector.
Further, in this project use of vortex generators such as baffles or ribs are introduced in the prepared setup and simulations are done so as to understand the variation of results between solar flat plate collectors without rib setup and that with it. The rib shape that is implemented is having a rectangular cross section and is located at the top of the absorber surface. The analytical calculations are carried out in MATLAB, followed by simulations using ANSYS Fluent and proper validation being done for experimental results and simulation study
Show less

Natural convection heat transfer or buoyancy driven flow is used in large variety of engineering applications especially in electronic cooling since it is low price, simple and quiet solution .As compared to a bare plate, a finned surface increases the heat transfer area. However, with the fins the flow rate could be reduced. Hence, if not properly designed it is possible that no improvement is achieved in terms of overall heat transfer. Therefore, only if the fins are properly designed, they… Show more

Natural convection heat transfer or buoyancy driven flow is used in large variety of engineering applications especially in electronic cooling since it is low price, simple and quiet solution .As compared to a bare plate, a finned surface increases the heat transfer area. However, with the fins the flow rate could be reduced. Hence, if not properly designed it is possible that no improvement is achieved in terms of overall heat transfer. Therefore, only if the fins are properly designed, they are very attractive for these applications
In present study numerical study is investigated on inverted trapezoid fin for natural convection flow instead of conventional rectangular fins because of its more heat transfer capacity. Heat flux varying from 3-20W/mm2 is applied at base plate of heat sink at an ambient temperature of 25°C. The simulation work is matched with the experimental results obtained from literature for using standard boundary conditions. The fin length based on inversion angle, fin height, fin spacing, fin thickness are varied and optimal configuration of inverted trapezoidal fins for maximum heat transfer will be found.. The main objective is to increase Heat transfer coefficient by varying geometrical parameters thus increasing the overall performance of heat sink.
Show less

In this project work, a 3D mathematical model with single and double glazed covers are prepared along with the introduction of TMS (thin metal sheet) and fins which are rectangular in shape. Energy and Exergy Analysis is performed on the solar flat plate collector with TMS and fins and is found to be much more effective in performance than most of the solar flat plat setups presently available in the market. Moreover, the influence of effective parameters such as depth, length, fin shape and… Show more

In this project work, a 3D mathematical model with single and double glazed covers are prepared along with the introduction of TMS (thin metal sheet) and fins which are rectangular in shape. Energy and Exergy Analysis is performed on the solar flat plate collector with TMS and fins and is found to be much more effective in performance than most of the solar flat plat setups presently available in the market. Moreover, the influence of effective parameters such as depth, length, fin shape and Reynolds number is also presented in the paper. In this study, the flow patterns are studied for fluid flow through the pipes at forced conditions i.e. with the aid of pumps instead of normal thermo-siphonic action. The addition of forced condition allows more effective heat transfer between the outside fluid and the forced fluid through pipes. The study helps to understand the flow parameters which vary often with change in flow velocities throughout the pipes.
Moreover, in this project use of vortex generators and heat enhancement setups such as fins and TMS (thin metal sheet) are used so as to increase the heat transfer rate between the interacting fluids. The basic geometrical dimensions for the 3D mathematical model is decided based on basic design calculations for the solar flat plat collector and CAD model is made in Ansys Design Modeller. The prepared model is simulated in ANSYS Fluent 14.5 with the use of all the required boundary and input conditions. A parametric study is done so as to understand the effect of fin spacing, air gap thickness and fluid mediums on the heat transfer rate. At last a comparative study is carried on between models having fins and TMS with another similar model but without fins and TMS so as to predict the effect in overall FPC efficiency.
Show less

Wind energy is the most sustainable and clean fuel source of energy, it doesn't pollute the air like power plants that rely on combustion of fossil fuels, such as coal or natural gas. It has experienced a rapid growth from last decade and has achieved quite attention from both government and private sector. Wind turbines like horizontal axis wind turbine (HAWT) and vertical axis wind turbine (VAWT) have been the principle medium for harnessing wind energy due to their high power coefficients by… Show more

Wind energy is the most sustainable and clean fuel source of energy, it doesn't pollute the air like power plants that rely on combustion of fossil fuels, such as coal or natural gas. It has experienced a rapid growth from last decade and has achieved quite attention from both government and private sector. Wind turbines like horizontal axis wind turbine (HAWT) and vertical axis wind turbine (VAWT) have been the principle medium for harnessing wind energy due to their high power coefficients by converting kinetic wind energy in to electrical power.
For onshore conditions HAWT has a great efficiency in converting wind energy to power but in offsore conditions due to high wake and high structural cost it is not preferred. Thus for onshore conditions VAWT are used being less structural cost, insensitivity towards wind direction and simple drive train.
Thus in present work computational fluid dynamics (CFD) is applied on VAWT using ANSYS Fluent for increasing its offshore power coefficient. The 3-D numerical modeling using ANSYS Design Modular is performed on turbine blades and parameters like blade thickness ratio, rotor diameter, span length, pitch angle and tilting angle are varied for optimization and further enhancement in performance of overall turbine. The simulation results are validated with experimental data by comparing power coefficient.
Show less

The flow of an inviscid gas is governed by the Euler Equations. Additionally the equation of state also relates the different properties by some functional relationships.
The time evolution of shock tube problem can be described by solving the Euler equations which leads to three characteristics, describing the propagation speed of the various regions of the system namely the rarefaction wave, the contact discontinuity and the shock discontinuity. If this is solved numerically, one can… Show more

The flow of an inviscid gas is governed by the Euler Equations. Additionally the equation of state also relates the different properties by some functional relationships.
The time evolution of shock tube problem can be described by solving the Euler equations which leads to three characteristics, describing the propagation speed of the various regions of the system namely the rarefaction wave, the contact discontinuity and the shock discontinuity. If this is solved numerically, one can test against the analytical solution, and get information how well a Matlab code captures and resolves shocks and contact discontinuities and reproduce the correct density profile of the rarefaction wave.
In this project work, some Computational Fluid Dynamics (CFD) techniques have been used to compute the variations in different parameters like pressure, density etc. by solving the Euler Equations for the shock tube problem. The governing equations are discretized on a Finite Volume framework. Van Leer’s first order and second order flux vector splitting method have been used to compute the inviscid flux terms. It is planned to study the effect of the order of the scheme on the accuracy of the solution. For higher order schemes the effect of the limiter type on the performance of the scheme shall also be highlighted. The performance of the scheme shall be judged by its ability to resolve shocks, expansion fans and contact discontinuities present in the shock tube problem.
Show less

In this project, a CAD model of the bicycle is made and based on different type of contact regions and critical regions, it was meshed accordingly and loaded for different loading situations using ANSYS. The resuts obtained show the total deformation and also modal analysis was carried on it.

In this mini project, a code was made in Matlab for any 3D frame having 6 DOF at nodes with 3 translational displacements and 3 rotational motions of each node. The main challenge was to plot the deformed shape of the model under different loading conditions such as self weight, model hits a bump or a speed breaker. The deformed shape was plotted with the knowledge of shape functions for both translational and rotational DOFs.

We have arrived at a concept that is simple, robust and easy to manufacture and maintain. We have a clear idea of the power splitter which is our main innovation. The design of ECU which requires extensive knowledge of electronics could not be materialized completely. However, we have arrived at a basic algorithm that allows ECU to control the drive mechanism of vehicle and also have managed to build up some circuits for the control of the whole transmission.
We have managed to fully design… Show more

We have arrived at a concept that is simple, robust and easy to manufacture and maintain. We have a clear idea of the power splitter which is our main innovation. The design of ECU which requires extensive knowledge of electronics could not be materialized completely. However, we have arrived at a basic algorithm that allows ECU to control the drive mechanism of vehicle and also have managed to build up some circuits for the control of the whole transmission.
We have managed to fully design the innovative power splitter keeping in mind all the technical aspects of different components that we have incorporated in our proposed design. Based on the design, CAD model is designed in CREO and analysis on different components are carried on in ANSYS. Show less

Thermal analysis of a double acting single cylinder engine using ethanol and diesel as fuel. The basic mechanics of the system was done which included the calculation of the forces and moments that would be accommodated due to high vibrations in the crankshaft and also the different calculations regarding the shape and size of the combustion chamber was done. Moreover, the cooling system for the engine was modified and a new innovative modelling of the piston and piston rod was made which was… Show more

Thermal analysis of a double acting single cylinder engine using ethanol and diesel as fuel. The basic mechanics of the system was done which included the calculation of the forces and moments that would be accommodated due to high vibrations in the crankshaft and also the different calculations regarding the shape and size of the combustion chamber was done. Moreover, the cooling system for the engine was modified and a new innovative modelling of the piston and piston rod was made which was based on the high temperature zones that were produced in the CC because of dual combustion processes i.e. two power strokes in two rotations of the crankshaft. The project is further to be carried and the CFD of the CC needs to be done along with the study of the spray patterns. Show less

The project involved was divided into two sections throughout the year - the first half included the study and analysis of flat plate and different prototypes b/l generation in a wind tunnel with dual motor input and butterfly valves to control the speed and amount of air intake into the wind tunnel. The second half included the study of basic building shapes like L, T, I and other such shapes for optimizing them against the coastal wind. The prototype to be build had to be fit with numerous… Show more

The project involved was divided into two sections throughout the year - the first half included the study and analysis of flat plate and different prototypes b/l generation in a wind tunnel with dual motor input and butterfly valves to control the speed and amount of air intake into the wind tunnel. The second half included the study of basic building shapes like L, T, I and other such shapes for optimizing them against the coastal wind. The prototype to be build had to be fit with numerous pressure tappings for measuring the static pressure at different locations. This was done on an sampled prototype and for making the optimum shape restriction came on the advancement of the project as the size of wind tunnel was made constraint due to only 5%blockage allowance of test section. Thus, the project can be further carried out with all the experimental data remaining intact. Show less

Project: Mortgage Domain

Description:
CCAR Project is related to Mortgage based Domain and is involved with generating FRY reports by maintaining proper Database of high end data from multiple sources for better monitoring of loan deliverables to an individual/organization.

Responsibilities:
• Understand the Entire Mismo Schema that has been prepared with relates the multiple sources
• Generate legitimate bulk Data for testing the SQL database for data quality… Show more

Project: Mortgage Domain

Description:
CCAR Project is related to Mortgage based Domain and is involved with generating FRY reports by maintaining proper Database of high end data from multiple sources for better monitoring of loan deliverables to an individual/organization.

Responsibilities:
• Understand the Entire Mismo Schema that has been prepared with relates the multiple sources
• Generate legitimate bulk Data for testing the SQL database for data quality evaluation.
• Involved in Unit testing of SQL database for the code that has been scripted in Python for Data Generation.
• Analyze system specifications and translate system requirements to task.
• Responsible for Requirement Analysis, Coding and Integration with SQL database so as to test it for bulk data.

Environment: Python 2.7, Anaconda 2.7, Spyder, SQL developer, MS Excel
 
Show less

Project: Marketing Domain

Description:
Shoppers rely on Home Depot’s product authority to find and buy the latest products and to get timely solutions to their home improvement needs. From installing a new ceiling fan to remodeling an entire kitchen, with the click of a mouse or tap of the screen, customers expect the correct results to their queries – quickly. Speed, accuracy and delivering a frictionless customer experience are essential.
Search relevancy is an implicit measure… Show more

Project: Marketing Domain

Description:
Shoppers rely on Home Depot’s product authority to find and buy the latest products and to get timely solutions to their home improvement needs. From installing a new ceiling fan to remodeling an entire kitchen, with the click of a mouse or tap of the screen, customers expect the correct results to their queries – quickly. Speed, accuracy and delivering a frictionless customer experience are essential.
Search relevancy is an implicit measure Home Depot uses to gauge how quickly they can get customers to the right products. Currently, human raters evaluate the impact of potential changes to their search algorithms, which is a slow and subjective process. By removing or minimizing human input in search relevance evaluation, Home Depot hopes to increase the number of iterations their team can perform on the current search algorithms.

Responsibilities:
• Handled huge amount of Data including train and test datasets.
• Involved in Preprocessing: Data Cleaning, Data Visualization , Solver Side: Implementing a unique Google Spell Checker Algorithm, RAKE for important keyword extraction and Post processing: Result Extraction in CSV format
• Identification of Bugs in the entire code and rectifying them.
• Generate an overall code that would generate search relevance according to the user input and give accurate results.
• Research on Semantic Comparison between different words and also exploring results from dependent features algorithms.
• Researched on SVM and Random Forest
• Responsible for Requirement Analysis, Coding and Integration with code review.

Environment: Python 2.7, MS Excel, Anaconda 2.7, Spyder.

Show less

Project: Multiple Domains

Description:
This Project was involved with creating a tool that can be utilized by an individual or any organization in predicting the right features that brings about large changes in the overall results. The features were at random and can be from any domain (e.g. political, social, etc.) irrespective of the data being textual or numerical in nature.

Responsibilities:
• Using Feature Reduction Technique in this case Backward Elimination Method… Show more

Project: Multiple Domains

Description:
This Project was involved with creating a tool that can be utilized by an individual or any organization in predicting the right features that brings about large changes in the overall results. The features were at random and can be from any domain (e.g. political, social, etc.) irrespective of the data being textual or numerical in nature.

Responsibilities:
• Using Feature Reduction Technique in this case Backward Elimination Method and Leave One Out simultaneously to precisely select the top most features.
• Segregating the Numerical data and String data in any dataset to individually perform unique operations on both of them.
• Using Keyword Extraction Algorithm such as RAKE to reduce the huge textual data into important and more relevant keywords.
• Integrating the developed Python code and Flask Framework for storing the processed dataset in My SQL
• Integrating the developed Python code with MySQL for dataset storage and also understanding the UI part utilizing Node JS technology.

Environment: Python 2.7, Anaconda 2.7, Spyder, MS Excel, My SQL, Flask, Node JS

Show less

Project: Finance Domain

Description:
This case study deals with handling customers who can be provided with automatic loan mails from Springleaf Financial Services by predicting the requirement of the customers based on analyzing different features. The main objective is to sort out customers that are more likely to respond to the automatic loan mails based on results obtained from huge datasets with numerous features.

Responsibilities:
• Extraction of unwanted features… Show more

Project: Finance Domain

Description:
This case study deals with handling customers who can be provided with automatic loan mails from Springleaf Financial Services by predicting the requirement of the customers based on analyzing different features. The main objective is to sort out customers that are more likely to respond to the automatic loan mails based on results obtained from huge datasets with numerous features.

Responsibilities:
• Extraction of unwanted features from the large datasets and eliminating them.
• Removing the blank features by negative values so as to make it analytically flexible and data split being performed on the large dataset to simplify the approach.
• Performed Dummy coding using R
• Performing Correlation Analysis on the train dataset to list out the most correlated features.
• Data Visualization was done on dataset and library packages such as ggplot, Maps, histogram, etc. were used to better visualize different parameters such as Feature Frequency, Global presence, Unique in a specific feature, etc.
• Performing different algorithms like xgboost to train the large dataset and also performing cross validation. An accuracy of almost 86 percentage was found using xgboost.
• Further the data is being dealt with to sort out features which are very much important in predicting the results with high accuracy and precision.
• The prediction is extended also to different algorithms like random forest, SVM, logistic regression and a comparative study is to be made based on accuracy, sensitivity and ROC curve extraction.

Environment: R, R Studio, MS Excel

Show less

Project: Survival Prediction

Description:
This case study deals with predicting the survival rate of people based on different categorical features such as age, gender, class, fare and others. The entire dataset is trained in such a way that the prediction model turns to be accurate enough so that it can properly identify the people with respect to random parameters that are mentioned in the test dataset.

Responsibilities:
• The first classification made was based on gender… Show more

Project: Survival Prediction

Description:
This case study deals with predicting the survival rate of people based on different categorical features such as age, gender, class, fare and others. The entire dataset is trained in such a way that the prediction model turns to be accurate enough so that it can properly identify the people with respect to random parameters that are mentioned in the test dataset.

Responsibilities:
• The first classification made was based on gender class model classification in R as the gender becomes one of the most important deciding factors in a titanic problem. The total recall obtained was 0.78.
• Decision trees were used to segregate the train datasets based on classes, the prediction was then submitted to the test dataset to enhance the algorithm and the recall was found to be almost 0.785.
• Introducing Feature engineering further enhanced the recall to 0.795. This smoothened the gender classification indirectly and resulted in better decision making while segregating the train datasets and thus predicted better results.
• The results were tried to further improved by using Random Forest algorithm however the submission showed a downfall of the results, thus the maximum score obtained was 0.795. To use Random Forest, we need much better re filtering the train dataset.

Environment: R, R Studio, MS Excel


Show less

The dynamic changes of boundary condition in running engine are investigated using modal analysis performed using ANSYS in which vibration characteristics are studied. Initially buckling analysis was done to validate the model and also to determine the required boundary conditions for buckling load.
Using the same boundary conditions and by analytically designing the connecting rod according to specific engine specifications of TVS Apache 150 model, buckling analysis was carried on the… Show more

The dynamic changes of boundary condition in running engine are investigated using modal analysis performed using ANSYS in which vibration characteristics are studied. Initially buckling analysis was done to validate the model and also to determine the required boundary conditions for buckling load.
Using the same boundary conditions and by analytically designing the connecting rod according to specific engine specifications of TVS Apache 150 model, buckling analysis was carried on the proposed model which was followed by modal analysis in WORKBENCH.

On the proposed model, parametric study is carried out to know the effect of input parameter such as shank fillet radius and shank thickness both near the big end as well as the small end side off the connecting rod. Based on the parametric study, an optimized value for shank fillet radius is obtained which is equal to 20 mm keeping other dimensions of the connecting rod fixed. It is found that there is considerable reduction of maximum value of Von Mises stress as well as deformations obtained for the modal analysis performed on the modified model of the connecting rod. The model is further validated by using MATLAB coding by considering it to be a 1D frame element both for side as well as front rear buckling conditions.
Show less

Thermal Shock Test deals with successive thermal loads being applied on both 2D as well as 3D asymmetric FEA models of pressure plate and flywheel assembly of a clutch. It is a non linear transient thermo-mechanical analysis.The project enables the proper understanding of the thermal effects as well as mechanical deformities that are observed due to frictional heat generation in between the contact surfaces of the clutch and also to understand the different critical regions of the pressure… Show more

Thermal Shock Test deals with successive thermal loads being applied on both 2D as well as 3D asymmetric FEA models of pressure plate and flywheel assembly of a clutch. It is a non linear transient thermo-mechanical analysis.The project enables the proper understanding of the thermal effects as well as mechanical deformities that are observed due to frictional heat generation in between the contact surfaces of the clutch and also to understand the different critical regions of the pressure plate which are to be considered during design phase so as to make the pressure plate more resistive to both thermal and mechanical loads that are being applied. Moreover, the main importance is given to three crucial factors for designing the pressure plate and clutch assembly which are as follows - stress ratio, yield strength and temperature. So, correlation is also to be made among these criterias for optimizing the design. Show less