Top Banner
A Summer Internship Report on Image Processing of 2 D Bubble Column At Submitted by: Sagar Chawla, B.Tech. III year Department of Chemical Engineering Indian Institute of Technology Gandhinagar Supervisor: Prof. Milorad Dudukovic Start Date for Internship: 11 th May, 2015 End Date for Internship: 20 th July, 2015
12
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Report

A

Summer Internship Report on

Image Processing of 2 D Bubble Column

At

Submitted by:

Sagar Chawla, B.Tech. III year

Department of Chemical Engineering

Indian Institute of Technology Gandhinagar

Supervisor: Prof. Milorad Dudukovic

Start Date for Internship: 11th May, 2015

End Date for Internship: 20th July, 2015

Page 2: Report

Abstract:-

Multiphase flow is simultaneous flow of two different phases mainly gas and liquid phase. A

bubble column reactor is an example of the application of two phase flows. The movement of

bubble is a basic subject in gas-liquid two-phase flow research. In this report changes in the

properties of bubbles due to the change in flow rate of air of 2 Dimensional flat bubble column

has been studied using image processing techniques in ‘Matlab 2015a’. Multiple Images of

532x516 pixels were taken by high speed camera Dalsa CA d6- 0512W and stored as a grey

scale images in the computer by using EPIX XCAP lite 3.8 software. Experiment was performed

on two different flow rates 10 SCPH and 20 SCPH. Twenty five images were captured for each

flow rate. Image processing was done on the best of 15 images and results were calculated for

the same on each flow rate. It was observed that interfacial area, surface area and perimeter and

of bubbles increases as gas flow rate increases.

Introduction:-

Bubble columns are intensively used as multiphase and reactors in chemical, biochemical and

petrochemical industries. They provide several advantages during operation and maintenance

such as high heat and mass transfer rates, compactness and low operating and maintenance costs.

High speed visualization is a non contact method which does not interfere with the motion of

bubbles. There is a need to understand the properties of the bubble such as its size, shape and

interfacial area since they are the critical parameters in the reactor design and control. Mass

transfer is the key phenomenon in the chemical reactions taking place in the reactor, it is

important to estimate the mass transfer coefficients for design and scale-up of these reactors. In a

bubble column reactor the variation in mass transfer is primarily due to variations in the

interfacial area. Gas-liquid interfacial area of two dimension flat bubble column reactor is

calculated as length of bubbles per unit area of image (in pixels).A number of works has been

done to study the flow characteristics of gas-liquid flow in the bubble column.

Page 3: Report

Other Methods:-

Using a bubble column operated with continuous gas injection Becker et al. (1999) studied

bubble swarm motion (bubble plume). The measurements were focused on the dynamics of

circulation flows employing Laser Doppler Anemometry (LDA). Bubble plume behavior

determined by Particle Image Velocimetry (PIV) and Particle Tracking Velocitmetry (PTV).

Both LDA and PIV are helpful to track fluid motion and to observe mixing characteristic in

bubble columns but it does not give much knowledge about bubble interaction. The dynamic gas

disengagement (DGD) technique is a very widely adopted method to study bubble groups,

bubble holdup structures and rise velocities. The principle involves tracing the drop in dispersion

height after the gas flow has been shut off. The resulting disengagement profile can be used to

separate the contributions of small and large bubbles to the total gas holdup.

Experimental setup:-

The bubble column reactor is made as a one sided open cuboid (175x 46 x 4 cm) of glass. An air

pump is used to pump air into the bubble column reactor through a needle. A flow control valve

is used to adjust the flow rate. A high-speed Dalsa CA d6-0512W camera consists of a CCD

sensor is used to capture images of the rising bubble. Its frame frequency is up to 960 fps and its

highest resolution is 512× 532. EPIX XCAP lite V3.8 software is used to store images in the

computer. Uniform lighting and a black square paper in the background is provided for the

bubble column reactor in order to get images of good quality.

Page 4: Report

Results:

(1) At 10 SCPH

(i) Interfacial area = 0.022

(ii) Distribution of perimeter of bubbles

Number of bubbles0 100 200 300 400 500 600 700 800 900

Pe

rime

ter

of b

ubbl

es (

in p

ixel

s)

0

50

100

150

200

250

300

350

400

450

500Number of bubbles vs Perimeter of bubbles(in pixles)

Page 5: Report

(2) At 20 SCPH.

(i) Interfacial area at 20 = 0.031

(ii) Distribution of perimeter of bubbles

Number of bubbles0 100 200 300 400 500 600 700 800 900 1000 1100

Pe

rime

ter

of b

ubbl

es (

in p

ixle

s)

0

100

200

300

400

500

600

700

800

900

1000Number of bubbles vs Perimeter of bubbles (in pixles)

Page 6: Report

Discussion & Conclusion:-

The effect of the superficial gas velocity on the bubble size is studied in a bubble column reactor

using image processing techniques. The morphological operations of background subtraction and

contrast of image on the captured images followed by the binary image conversion and some

other operation result in a convenient approach to measure the properties of bubbles in bubble

column reactor. Perimeter and surface area of bubbles increases as gas flow rate increases thus.

Length of bubbles and bubbles per image is increases as flow rate increases thus interfacial area

also increases and enhanced the mass transfer rate, because interfacial area is calculated as a ratio

of average sum of perimeter of bubbles of 15 images to the area of image of 532x512 Pixel.

Experiment was performed on 10 SCPH and 20 SCPH, because bubbles clusters increases too

much as gas flow rate further increase above 20 SCPH. It cannot be separated effectively during

image processing. Binary image considers all connected object as a one object thus bubbles

clusters leads too much error in the calculation of bubble’s properties.

Image at 10SCPH Image at 20 SCPH Image at 30 SCPH

Page 7: Report

Appendix:-

High speed camera specification

Model Dalsa CA d6-0512W

Sensor CCD array up to 532 x 512 pixels,

Recording rate(fps) 4 outputs at 25 MHz: frame rates to 955 or 262

frames/sec

Lens mounts M6Z1212-3S

2/3" 12.5-75mm f1.2 6X Manual Zoom,

Manual Iris (C-Mount).

Steps for capturing images:-

1. First, camera is connected to the computer and turned on the camera.

2. Open the EPIX xcap lite 3.8 software.

3. Clicked on ‘Live’ option in software to see the live picture on computer.

4. Focus of high-speed camera is adjusted until the image is clear.

5. Inlet of water is turned on and filled the column upto 100 cm.

6. Turned on the light and the air control valve to make bubbles rising from the needle.

7. Region of 7x7 squares (60 cm from bottom) is selected as a research object and continuous

images are captured of rising bubbles at two different flow rates. (EPIX XCAP lite software

captures images in grey scale).

8. All the images are saved in single folder.

Page 8: Report

Steps for Image processing

1. Read image file from folder in ‘Matlab 2015a’ software.

2. Uneven background distribution is removed and adjusted the contrast of image.

3. Grey scale image is converted into RGB scale image.

4. Open the ‘Color Thresholder’ app and converted RGB scale image is loaded.

5. RGB image is changed into Binary image by choosing appropriate threshold. Threshold can set

by moving slider of threshold level.

6. Binary image is loaded in the workspace.

7. Holes of the Binary image are filled by ‘imfill’ method.

8. Smaller bubbles (Those bubbles whose area are less than 30 pixel2) of binary image are removed

by ‘bwareaopen’ method.

9. Open the ‘Image Region Analyzer’ app and modified Binary Image is loaded.

10. Properties of Bubbles are calculated by ‘Image Region Analyzer’ app.

11. Array of properties are created by ‘Image Region Analyzer’ app and loaded in workspace (Image

Region analyzer app create Structure array of properties).

12. Structure array is converted into cell array followed by matrix array.

13. Sum of perimeter is calculated.

14. Interfacial area is calculated as a ratio of average sum of perimeters of 15 images to the area of

image (532x512).

Page 9: Report