A Simplified PIV-based Method for Flame Velocity
Distribution Measurement
Rui Yin, Lijun Xu, Xiaolu Li*
School of Instrumentation and Opto-Electronic Engineering, Beihang University, Beijing 100191, China.
Key Laboratory of Precision Opto-mechatronics Technology of Ministry of Education, Beihang University, Beijing
100191, China
E-mail: xiaoluli@buaa.edu.cn
Abstract—In the paper, a simplified method based on particle
image velocimetry (PIV) for flame velocity distribution
measurement is proposed. A high speed camera was first used to
capture consecutive images, and multilevel thresholding
segmentation based on Otsu’s method was used to enhance the
image contrast. Then the velocity distribution is calculated by
means of DFT. Experiments were carried out to validate the
feasibility of the proposed method by using a methane/air
premixed burner. Calculation results show that the flame velocity
distribution can be completely obtained by the aid of multilevel
thresholding segmentation. The experimental results show that
the flame velocity distribution agrees well with the flame
structure and the motion feature. The proposed method can be
further applied in monitoring of industrial combustion process.
Keywords—PIV; flame; velocity distribution; multilevel
thresholding segmentation; Otsu
I. INTRODUCTION
The velocity distribution of flame is an important
parameter reflecting the combustion characteristic. Velocity
distribution demonstrates the transmission of energy and
spread of material motion more directly compared to other
parameters, such as temperature or concentration. However,
the measurement of velocity distribution in flame flow field is
a tough challenge because of high temperature, intense
radiation and other awful factors.
Generally, Pitot tube, hot-wire anemometer and Laser
Doppler Velocimetry (LDV) are used as traditional detection
method for velocity measurement. Pitot tube and hot-wire
anemometer are of low accuracy and slow response. LDV gets
rid of the above shortcomings, but it is complex, expensive
and difficult to be installed. By using these methods, only the
velocity of a single point rather than the complete flame
velocity distribution can be measured. Since 1970s, Particle
Image Velocimetry (PIV) has been widely used in the
measurement of velocity distribution. A series of continuous
images can be captured in a short time by high speed camera
after tracer particles are poured into the flow field. The
instantaneous velocity field distribution is obtained by
calculating the displacements of tracer particles. The PIV
method has advantages of non-contact measurement, high
time resolution and high spatial resolution. Furthermore, the
tracer particles have negligible effect on the fluid flow if they
are properly chosen [1]. The PIV method can be used not only
in gas-liquid flow field, but also in the combustion flow field.
Thus, it is an important technical mean for the monitoring of
combustion process.
Properties of turbulent premixed flames, including burning
velocities, stretch rates, and surface densities were analyzed
using PIV method by S.A. Filatyev in [2]. The flames of flat
burner at atmospheric and low pressures were examined using
PIV method by A.A. Konnov in [3]. Compared with the
traditional methods, a two-dimensional velocity distribution
can be obtained by the PIV method. However, the common
PIV is expensive and complex. The method need to be
simplified to be widely used. In recent years, the PIV
application with a camera was carried out on material surface
deformation measurement. D.J. White developed a
deformation measurement system based on PIV and used it in
geotechnical testing [4]. B. Pan discussed the methodologies
of the PIV technique for material surface deformation
measurement [5]. Only a camera was used in these researches.
The measurement process is greatly simplified. Therefore, it is
feasible to measure flame velocity distribution with a camera.
In this paper, the main contribution of the simplified
method based on PIV for flame velocity distribution
measurement is concluded as follows.
(1) Flame images were captured by a high speed camera
and then processed by multilevel thresholding
segmentation based on Otsu’s method.
(2) The velocity distribution is calculated by discrete
Fourier transform (DFT) approach according to the
time interval of two images and the relationship
between world coordinate system and image
coordinate system.
(3) Experiments were carried out to validate the proposed
method. Experimental results show that velocity
distribution agrees well with the flame structure and
the motion feature.
II. M
ETHOD
Each image captured by the camera can be regarded as a
matrix. The pixel gray level is the matrix’s value. Therefore,
by calculating the correlation of two matrixes, the flame flow
field velocity distribution can be obtained. The discrete cross
correlation function is used to calculate the correlation [6],
which is
978-1-4799-8633-0/15/$31.00 ©2015 IEEE