You are on page 1of 10

Tamkang Journal of Science and Engineering, Vol. 11, No. 2, pp.

145-154 (2008) 145

Study of Numerical Simulation Applying to the Design


of an Orifice with High-Velocity Waterjet
Fu-Cheng Yang1*, Sheau-Wen Shiah2, Tzeng-Yuan Heh2,
Chan-Yung Jen3 and Wen-I Chang4
1
Graduate School of Defense Science, Chung Cheng Institute of Technology, National Defense University,
Taoyuan, Taiwan 335, R.O.C.
2
Department of Power Vehicle and System Engineering, Chung Cheng Institute of Technology, National Defense University,
Taoyuan, Taiwan 335, R.O.C.
3
Department of Mechanical and Marine Engineering, Naval Academy,
Kaohsiung, Taiwan 813, R.O.C.
4
Duo De Company of Limited Liability,
Taoyuan, Taiwan 325, R.O.C.

Abstract
The purpose of this paper is using numerical simulation method to discuss the influences of
high-velocity waterjet on different types of orifice which is made of diamond. Not only we aim at
establishing the model of analysis to the orifice geometry but also change the design conditions of the
orifices in order to find the optimal orifice shape. During the numerical simulation, a two-dimensional
axial-symmetric model is established first, and then some practical geometric parameters are considered
for the orifice, including flow channel with 0.5 mm in depth, 0.1 mm to 0.56 mm for the diameters of
inlet and outlet, and 0.005 mm to 0.346 mm for the neck height. The analytic results can show the
pressure and velocity distributions of high-velocity waterjets in the orifices under different conditions
and can be used for assessment of orifice geometric design. Through this study, it is expected to know
more physical insight about the effect of high-velocity waterjet on orifice geometry and further to
improve the life cycle of orifice.

Key Words: Design of Orifices, High-Velocity Waterjets, Numerical Simulation

1. Introduction process.
The steady flow problem of a viscous incompres-
There are essentially two types of waterjets: pure sible fluid was first discussed by Liepmann [3] and
waterjet (WJ) and abrasive waterjet (AWJ). WJ has been Naramisimha [4]. Vahedi Tafreshi and Pourdeyhimi [5]
moving onto the scene very slowly since 1970 and AWJ performed a series of numerical simulations using com-
a bit more rapidly since about 1980 [1]. WJ was the origi- putational fluid dynamics (CFD) code from Fluent Inc.
nal water cutting method which was used to cut corru- to solve the effects of nozzle geometry on waterjet break-
gated cardboard. Today, WJ is used to cut a variety of up at high Reynolds number problem.
soft/light materials including paper, food, fiberglass in- Zheng et al. [6] discussed the quality and cost com-
sulation, some plastics and some composites [2]. In WJ, parisons between laser cutting and waterjet cutting. They
cutting performance is obtained by a supersonic erosion finished quality analyses on cutting of stainless steel pla-
tes, aluminum plates and mild steel plates. And the cost
*Corresponding author. E-mail: g970403@ccit.edu.tw analyses are determined by the costs of equipment, con-
146 Fu-Cheng Yang et al.

sumables, service, and maintenance. Bach et al. [7] stud- compressible fluid through an orifice by using the Ado-
ied characterization for simulating the pure WJ cleaning mians decomposition method. Almeida Medeiros et al.
process. The method reduced the costs as well as the time- [17] aimed to evaluate the parameters that should be ac-
to-market of new waterjets cleaning devices drastically. counted for the dimensional inspection of orifice plate
Liu et al. [8] presented the range of downstream dis- and roughness of the straight stretches of the tubing mea-
tances. The results indicate that a jet is characterized by an surement of natural gas, with the objective of guiding the
initial rapid decay of the axial velocity at the jet centre responsible professionals for the measurement of vol-
while the cross-sectional flow evolves towards a top-hat umes of nature gas by orifice plate, as well as the profes-
profile downstream. Borutzky et al. [9] proposed orifice sionals of calibration laboratories.
model has been used in bond graph model of a hydraulic For this purpose, CFD analysis is found to be a viable
sample circuit. Their simulation results have proved to be approach because direct measurement and visualization
accurate. Guo et al. [10] had ever numerical modeling via of pure water velocities distributions are very difficult for
Finite Element Analyses (FEA) and experimental mea- the high velocity and small dimensions involved. In this
surements via moir interferometry. FEA and experimen- paper, CFD, model simulation for high velocity waterjets
tal measurements were applied to the investigation of are established using CFD-RC ACE solver. Jet dynamic
abrasive waterjet (AWJ) drilling. They finished analyses characteristics such as the pure water velocities for the
of the waterjet contact and impact on target materials. flow downstream from a very fine nozzle are then simu-
Bunnell et al. [11] made a model numerically by solution lated under steady state, turbulence flow, no-slip wall con-
of two-phase laminar Navier-Stokes equations which for- ditions, and a range of inlet conditions and outlet para-
mulate with a homogeneous fluid model on unsteady 2D meters. The results of the CFD study are then analyzed to
and axisymmetric viscous cavitating flows. They pre- gain an insight into the jet characteristics and a fundamen-
sented increasing in the pressure drop across the orifice or tal understanding for efficient cutting with high velocity
the site density which had the tendency to increase the ex- WJ considering the variations in geometric orifice.
tent of cavitation and decrease the frequency of oscillation
about the average length of the cavitating region. 2. Mathematical Formulations
Charinrat and Sakamon [12] used commercial soft-
ware, FEMLAB, with the finite element method to solve 2.1 Fundamental Assumption
the problem which coupled heat, mass, and momentum The geometry of orifice model for waterjet cutting in
balanced equations. Garcia-Valladares [13] developed a this paper is divided into six types of two kinds and the
numerical model allowing analyses such as geometry, basic orifice configuration is shown as Figure 1. The re-
different working conditions, critical or non-critical flow
conditions, etc. K. H. Ahn and J. H. Ahn [14] studied nu-
merically on particle motion through a disk type critical
orifice placed in a 2.2 cm diameter pipe.
Recently, a powerful method which is called decom-
position methodology developed by Adomian [15] can
be provided for analytic approximations to a wide class
of nonlinear ordinary differential equations, partial dif-
ferential equations, and systems of partial differential
equations, and this method demands to be parallel to any
modern supercomputer. Bulut et al. [16] studied an an-
Figure 1. Computational domain and boundary conditions of
alytic solution of steady flow problem of a viscous in- the CFD model.
Study of Numerical Simulation Applying to the Design of an Orifice with High-Velocity Waterjet 147

levant sizes are illustrated in Table 1. Six CFD-GEOM where r is the fluid density, p0 is the pump pressure, do
models according to Table 1 were set up and each profile is the inlet diameter, and m is the dynamic viscosity.
is given from Figure 2, respectively. CFD-RC software is
used to simulate the 2D calculated field in numerical mo- Table 2. Parameters setting of orifice
dels. The physical characteristic states of waterjet cut- Simulation parameters Setting value
ting orifices as follows were in order to simplify calcula- Pump pressure 4 108 N/m2
tion. Every parameter setting value is described in details Reference pressure 101,325 N/m2
as Table 2. Internal flow field pressure 101,325 N/m2
Outlet pressure 101,325 N/m2
A. Field flows treated as steady state. Inlet velocity 1000 m/s
B. Field flows considering no-slip condition. Acceleration of gravity 9.8 m/s2
C. The fluids (water ) in the orifice are turbulence Inlet temperature 300 K
Internal flow field temperature 300 K
flows:
Outlet temperature 300 K
Fluid density 1000 kg/m3
Fluid viscosity 0.001 kg/m-s
Wall velocity u = 0, v = 0 m/s

Table 1. Dimensions of orifices model (unit: mm)


TYPE Inlet diameter (di) Outlet diameter (do) Neck height (ln)
A 0.20 0.47 0.044
B 0.23 0.47 0.201
C 0.35 0.47 0.346
D 0.10 0.30 0.005
E 0.20 0.40 0.005
F 0.36 0.56 0.005
Note. Depth of flow channel: L = 0.5 mm

Figure 2. CFD model for orifice geometries.


148 Fu-Cheng Yang et al.

A. The fluid densities of waterjets are constants. In these equations p is the static pressure and tij is the
B. The heat transfer effect of waterjet is neglected. viscous stress tensor, where
C. All the inlet velocities of the orifices in this re-
search are 1000 m/s.
(5)

2.2 Governing Equations


The major governing equations are used from the (6)
CFD model and the boundary conditions for the simula-
tion are given as follows.
(7)
2.2.1 Continuity Equation
Conservation of mass requires that the time rate of (8)
change of mass in a control volume can be balanced by
the net mass flow into the same control volume (out-
(9)
flow-inflow). This can be expressed as:

(1) (10)

The first term on the left hand side is the time rate of
(11)
change of the density (mass per unit volume). Because
the initial condition in this paper was set steady state, the
first term can be cancelled. The second term describes (12)
the net mass flow across the boundaries of control vol-
ume and is called the convective term.
(13)
2.2.2 Momentum Equations (Navier-Stokes Equations)
The x-component of the momentum equation is found The momentum equations involve three dimension
by setting the rate of change of x-momentum of the fluid functions, but the study of numerical simulation was two
particle equal to the total force in the x-direction on the dimension calculations. So the z-axial part can be can-
element due to surface stresses plus the rate of increase celled in some equations.
of x-momentum due to sources:
2.2.3 Specification of Turbulence Quantities Using
Turbulence Intensity
(2)
The turbulence intensity, I is defined as the ratio of
the root-mean-square of the fluctuation velocity, u to the
Similar equations can be written for the y- and z- mean flow velocity, u
components of the momentum equation:
(14)
(3)
The turbulence intensity generally ranges from 1%
to 10%. That with turbulence intensity less than 1% is
(4) considered as low turbulent flow and that with turbu-
Study of Numerical Simulation Applying to the Design of an Orifice with High-Velocity Waterjet 149

lence intensities greater than 10% are considered as high generation of turbulent kinetic energy due to the mean
turbulent flows. velocity gradients, uj represents the velocities in the xj
The turbulence intensity at the core of a fully deve- coordinate directions, Cm, C e1 and C e 2 are constant, sk
loped duct flow can be estimated as: and se denote the turbulent Prandtl numbers for k and e,
respectively, and r is the constant density, m is the dy-
(15) namic viscosity. The values of the constants are taken
from [18], i.e. Cm = 0.09, C e1 = 144 . , sk = 1.0,
. , C e 2 = 192
where Re is the Reynolds number. and se = 1.3.

2.2.4 Finite Volume Method 2.3 Boundary Conditions


Most of the governing equations can be expressed in The related boundary conditions are set in this re-
the form of a generalized transport equation, Eq. (16), search shown as follows:
A. Reference pressure:
The pressure of system is established as one atmo-
(16)
sphere pressure (101,325 N/m2).
B. Boundary condition of the inlet:
rf v The velocity of inlet flow is established as 1000
where is a transient term, (rVf ) is a convection
t m/s.
term, ( Gf ) is a diffusion term, and Sf is a source C. Fluid physical characteristics:
term. The fluid is set as water whose density r = 1000 kg/
This equation is also known as the generic conserva- m and viscosity m = 0.001 kg/m-s.
3

tion equation for a quantity f. Integrating this equation D. Isolating wall boundary conditions:
over a control-volume cell, we have, Isolating wall velocities u = 0 and v = 0. The rough-
ness is established as smooth one (Roughness height =
0).
(17)
E. Initial conditions of internal fluid field:
The velocities of internal flow are established as u =
2.2.5 Transport Equations 0 and v = 0, pressure P =101,325 N/m2, temperature T =
The transport equations for turbulence energy k and 300 K.
dissipation rate e are solved and shared by the phases F. Boundary condition of the outlet:
throughout the field. The standard k - e model is chosen The outlet pressure is established as a atmosphere
in this simulation work. This model is mathematically pressure (101,325 N/m2) and temperature, T = 300 K.
given by
3. Modeling Approach
(18)
The aim of the simulation is to predict the waterjet
cutting performance of a given waterjet orifice based on
(19) a set of input parameters. In the following these parame-
ters are schematically listed in Figure 1.

rCm k 2 3.1 Modeling Geometry Configuration


where m t = is the turbulent viscosity, rp is the
e The 2D geometry which be built by the commercial
150 Fu-Cheng Yang et al.

software, CFD-GEOM, will be used for the CFD model- city is 1000 m/s, the acceleration of gravity as 9.8 m/s2
ing. Because the effects in variety of orifice configura- and the density of fluid as 1000 kg/m3 were considered.
tions will be studied in the paper, the orifice geometries The CFD results for outlet velocities distributions are
have been set up as shown from Figure 2 to Figure 7. In obtained and representative plots are shown in Figure 3.
this study, the six types of orifice configurations were 100 sets of velocity data at 6250, 7500, and 8750 grid
used to simulate the distributions of velocity field and points were taken in this study. The Root Mean Square
pressure field. The dimensions of orifice are shown in (RMS) is 36.35% between 6250 with 7500 grid points,
Table 1. and it is 5.39% between 7500 with 8750 grid points. The
To investigate the accuracy, stability, and conver- pressure distributions at orifice centerline are also very
gence properties of the CFD model, a number of initial close such as Figure 4. Based on the basis of exactness
computations are performed by using second-order con- and demand for saving the time of calculating, this re-
vective discretion scheme on rectangular grids with both search adopts 7500 grid points.
uniform and non-uniforms grid spacing in the Cartesian
coordinate. 4.2 Numerical Simulation Verification
Numerical simulation verification in this study is
3.2 Kinematic Data and Simulation compared with [19] in pressure and velocity parameters.
Kinetic and liquid parameters for turbulence models First the pressure parameter is used to prove the velocity
in a waterjet cutting are shown in Table 2, which is also
used for the set-up. Using the geometrical, kinematic
data, the solver of simulation, CFD-ACE, computes the
matrix of the results on the basis of user-definable dis-
crete time increments. The data generated by the solver
consist of a matrix representing the altered surfaces of
the virtual work pieces respectively.

3.3 Post Processing


For post-processing and visualization of the result
matrix, the CFD-view was used to see the profiles of ve-
locity distributions directly. The other commercial soft-
Figure 3. Flow velocity distributions at outlet of orifices for
ware, Tecplot was used in order to do analyses in detail. 6250, 7500, and 8750 grid points.

4. Numerical Methods and Verify Studies

4.1 Testing for Grid System


The Finite Volume Method (FVM) was adopted in
this research to do numerical calculation. Based on the
TYPE-A orifice, the neck length of orifice was set up 26,
51 and 76 points for grid points tests. At this moment,
there are 6250, 7500 and 8750 grid points in the two-
dimensional models respectively. The calculation con-
ditions are all in turbulence flow field. The reference
pressure, initial pressure and outlet pressure in the model Figure 4. Pressure distributions at the centerline of orifices
system are all a atmosphere pressure. When inlet velo- for 6250, 7500 and 8750 grid points.
Study of Numerical Simulation Applying to the Design of an Orifice with High-Velocity Waterjet 151

of orifice flow field. Then, 30 sets of velocity values are duced most obviously and suddenly. So TYPE-C is supe-
got when inlet pressure of TYPE-A orifice is 4 108 rior to TYPE-A and TYPE-B for waterjet cutting of sys-
N/m2. The mean velocity is equal to 833.57 m/s after tem which is at a high velocity obviously.
CFD calculation. If comparing this result with the velo-
city (853.44 m/s) in Figure 5 [19], the difference ratio is
only 2.33%. Next, the velocity parameter is used to
prove the pressure of orifice flow field. The pressure is
551,604,800 N/m2 when the velocity is 1036.32 m/s in
Figure 10. The same velocity at inlet of TYPE-A orifice
was set and 30 sets of pressure values were gotten. The
average pressure is equal to 5.55108 N/m2. The mini-
mum difference ratio reaches 0.56%. The results show
the current CFD simulation model is correct.

5. Results and Discussions


Figure 6. The outlet velocity distributions of TYPE-A, TYPE-
B, and TYPE-C orifices.
TYPE-A, TYPE-B, and TYPE-C orifices have obvi-
ous necks, according to inlet velocity distributions in
Figure 6, where the maximum velocities are 806.42 m/s,
973.39 m/s, and 938.53 m/s respectively, and the mini-
mum velocities are 107.38 m/s, 618.46 m/s, and 736.82
m/s respectively. To sum up, the velocities of TYPE-C
orifice are generally faster than these of other two types,
but the distributions are not centralized. Figure 7 and
Figure 8 show the distributions of fluid velocities and
pressures at the Y-axis of orifice centre. To observe the
slope in the picture, we can find that if the distance far
away the outlet is farther, the velocity and pressure re-
Figure 7. The centerline velocity distributions of TYPE-A,
ductions in TYPE-C orifice are very smooth along Y- TYPE-B, and TYPE-C orifices.
axis direction. The velocities in TYPE-A orifice are re-

Figure 8. The centerline pressure distributions of TYPE-A,


Figure 5. The velocity of waterjet vs. pressure [19]. TYPE-B, and TYPE-C orifices.
152 Fu-Cheng Yang et al.

TYPE-D, TYPE-E, and TYPE-F orifices only take smooth along Y-axis direction can be found, but the ve-
the extremely small ratios which is the depth divided by locity and pressure reductions in TYPE-D and TYPE-E
the neck height. Figure 9 shows the outlet velocity and are obviously. The outlet pressures of these three types
pressure distributions, where the maximum velocities of orifice are almost the same. This calculated results
of flow are 437.39 m/s, 922.99 m/s, and 748.49 m/s re- and the assumed boundary conditions of outlet in the
spectively, and the minimum velocities are 207.17 m/s, system are matched.
335.51 m/s, and 434.75 m/s respectively. The fluid ve-
locity of TYPE-D orifice is the slowest, but it is com- 6. Conclusions and Future Works
paratively centralized to distribute. The fluid velocity
of TYPE-F is the fastest, but distributes in the surface In current study, a well-developed numerical method
relatively widely. Figure 10 and Figure 11 show the dis- is applied to solve fluid velocity and pressure distribu-
tributions of fluid velocities and pressures at the Y-axis tions for different configurations of orifices with high-
of orifice centre. To observe the slope in the picture, if velocity waterjet. From the above simulation results, a
the distance far away the outlet is farther, the velocity number of important conclusions:
and pressure reductions in TYPE-F orifice are very (1) Slight volume of the high velocity waterjets orifice
in two-dimensional model and grid system which
are used to design and analyze were set up by
CFD-GEOM.
(2) To synthesize the results of calculation of the above
six types of orifice, the waterjet of the TYPE-C or-
ifice is the best in velocity, and waterjet of the
TYPE-D orifice is the best in intensity which due
to its smallest inlet and outlet area. Above-men-
tioned analyses of configurations design can offer
high velocity waterjet orifices references in usage
actually.
(3) Design and analysis models for flow and pressure
fields of waterjet orifice configurations were set
Figure 9. The outlet velocity distributions of TYPE-D, TYPE-
E, and TYPE-F orifices. up. And the best choice in waterjet application

Figure 10. The centerline velocity distributions of TYPE-D, Figure 11. The centerline pressure distributions of TYPE-D,
TYPE-E, and TYPE-F orifices. TYPE-E, and TYPE-F orifices.
Study of Numerical Simulation Applying to the Design of an Orifice with High-Velocity Waterjet 153

from six types of orifices calculation results was Yeo, S., Quality and Cost Comparisons between La-
indicated. ser and Waterjet Cutting, Journal of Materials Pro-
(4) Combining the velocity and pressure distributions cessing Technology, Vol. 62, pp. 294-298 (1996).
in the six types of orifice, the type-C is the best [7] Bach, Fr. W., Louis, H., Versemann, R. and Schenk,
choice if the high cutting thickness is demanded A., Pure WJ Cleaning Process Characterization: Part
and the type-D is the best choice under accurate II-Process Simulation, Proceeding of the 8 th Interna-
cutting request. tional Conference on Management of Innovative Tech-
(5) We can combine the part which is above the orifice nologies, MIT2005, pp. 15-21 (2005).
with the part which is below the orifice to let it [8] Liu, H., Wang, J., Kelson, N. and Brown, R. J., A
match the actual modeling of waterjet cutting, and Study of Abrasive Waterjet Characteristics by CFD
investigate the Fluid-Structure Interaction pro- Simulation, Journal of Materials Processing Tech-
blems of orifice, etc. further. We believe that the nologies, Vol. 153-154, pp. 488-493 (2004).
applications of waterjets will be beneficial to some [9] Borutzky, W., Barnard, B. and Thoma, J., An Orifice
extent. Flow Model for Laminar and Turbulent Conditions,
Simulation Modelling Practice and Theory, Vol. 10,
Acknowledgement pp. 141-152 (2002).
[10] Guo, Z., Ramulu, M. and Jenkins, M. G., Analysis of
The authors wish to express their sincere thanks to the Waterjet Contact/Impact on Target Material, Op-
the Duo De Company of Limited Liability for the finan- tics and Lasers in Engineering, Vol. 33, pp. 121-139
cial support rendered to undertake this research work. (2000).
[11] Bunnell, R. A., Heister, S. D., Yen, C. and Collicott, S.
References H., Catitating Injector Flows: Validation of Nmme-
rical Models and Simulations of Pressure Atomizers,
[1] Kahlman, L., Karlsson, S., Gralsson, R. and Nilson, C. Atomization and Sprays, Vol. 9, pp.445-465 (1999).
G., Wear and Machining of Engineering Ceramics by [12] Charinrat, S. and Sakamon, D., Numerical Simula-
Abrasive Waterjets, American Ceramic Society Bul- tion of Flow and Mixing Behavior of Impinging St-
letin, Vol. 72, pp. 93-98 (1993). reams of Shear-Thinning Fluids, Ocean Engineering,
[2] Kulekci, M. K., Processes and Apparatus Develop- Vol. 61, pp. 4884-4892 (2006).
ments in Industrial Waterjet Applications, Interna- [13] Garca-Valladares, O., Numerical Simulation of
tional Journal of Machine Tools and Manufacture, Trans-Critical Carbon Dioxide (R744) Flow through
Vol. 42, pp. 1297-1306 (2002). Short Tube Orifices, Applied Thermal Engineering,
[3] Liepmann, H. W., Gas Kinetics and Gas Dynamics of Vol. 26, pp. 144-151 (2005).
Orifice Flow, Journal of Fluid Mechanics, Vol. 10, [14] Ahn, K. H. and Ahn, J. H., Numerical Simulation of
pp. 65-79 (1961). Particle Motion through a Critical Orifice, Journal of
[4] Narasimha, R., Orifice Flow at High Knudsen Num- Aerosol Science, Vol. 29, pp. S833-S834 (1998).
bers, Journal of Fluid Mechanics, Vol. 10, pp. [15] Adomian, G., A Review of the Decomposition Met-
371-384 (1961). hod in Applied Mathematics, Journal of Mathemati-
[5] Vahedi Tafreshi, H. and Pourdeyhimi, B., The Effects cal Analysis and Applications, Vol. 135, pp. 501-544
of Nozzle Geometry on Waterjet Breakup at Hight (1988).
Reynolds Numbers, Experiments in Fluids, Vol. 35, [16] Bulut, H., Ergt, M., Asil, V. and Bokor, R. H., Nu-
pp. 364-371 (2003). merical Solution of a Viscous Incompressible Flow
[6] Zheng, H. Y., Han, Z. Z., Chen, Z. D., Chen, W. L. and Problem through an Orifice by Adomian Decomposi-
154 Fu-Cheng Yang et al.

tion Method, Applied Mathematics and Computation, [18] Launder, B. E. and Spalding, D. B., Lectures in Mathe-
Vol. 153, pp. 733-741 (2004). matical Models of Turbulence, Academic Press, Lon-
[17] Almeida Medeiros, A. K., Lima, J. F., Medeiros, G. G., don (1972).
Silva Junior, N. F., Felipe, R. N. B. and Santos Felipe, [19] John, H. O., Abrasive Jet Mechanics, March 8, 2005
R. C. T, Parameters for Dimensional Inspection of thefabricator.com (2005).
Orifice Plates and Roughness of the Straight Stretches
of the Tubing, Brazilian Archives of Biology and Manuscript Received: Apr. 11, 2007
Technology, Vol. 49, pp. 1-8 (2006). Accepted: Aug. 29, 2007

You might also like