This action might not be possible to undo. Are you sure you want to continue?
pump
M. van Beijnum
Report number: WPC 2007.08
Committee:
Prof. dr. ir. J.J.H. Brouwers (chairman)
Dr. B.P.M. van Esch (supervisor)
Dr. J.G.M. Kuerten
Dr. ir. C.C.M. Rindt
Dr. ir. O.J. Teerling
Technische Universiteit Eindhoven
Department of Mechanical Engineering
Division TFE, Section Process Technology
Eindhoven, August 2007
Abstract
Multiphase pumping in the oil and gas industry is the ability to boost pressure without sepa
rating the liquid and the gas phases. This gives opportunities to process the diﬀerent phases
centrally when using multiple well sites, or processing on land for an oﬀshore well. No sep
aration of the phases and only one pipeline have to be used. For optimal performance of a
twin screw pump a small liquid fraction is necessary to seal the internal clearances of a twin
screw pump. 100 percent gas void fractions can be pumped for a short period of time when
arrangements are made to recirculate ﬂuid, to seal the clearances. This makes the internal
design of a twin screw pump an engineering challenge. To gain more insight of the ﬂow in a
twin screw pump a CFD model can be used.
The goal of this graduation report is to predict the leakage ﬂow rate in a twin screw pump
with a threedimensional model of the pump and a commercial CFD package. This goal can
be divided in three parts. The leakage ﬂow rate ﬁrst for nonrotating screws and singlephase
ﬂow, secondly for rotating screws and singlephase ﬂow, and ﬁnally considering multiphase
ﬂow with rotating screws. In this report the ﬁrst part is considered, and recommendations
are given for the other parts.
The leakage ﬂow rate is simpliﬁed in two diﬀerent cases. The ﬁrst case is ﬂow through
an annulus with inner rotating cylinder, this represents the ﬂow between the screw and
housing (liner) of a twin screw pump. The second case is ﬂow through a straightthrough
labyrinth seal. The screw thread viewed in axial direction is similar to a labyrinth seal.
Recirculation and throttling of ﬂuid in the screw cavities can be expected. For these two
cases the performance of the turbulence model is evaluated and coupled to requirements for
the dimensionless wall distance in the ﬁrst cell near the wall.
To simulate the ﬂow in a twin screw pump the ﬂow domain has to be meshed, the mesh must
have a limited number of cells to perform calculations with normal PC requirements in a rea
sonable amount of time. The number of cells in the clearance between the tips of the screw and
the liner is estimated. The number of cells with an unstructured tetrahedral mesh is too large
to perform CFD simulations. Structured hexahedral cells can be used, however these cells
have to be elongated in axial and tangential direction to reduce the number of cells. A struc
tured grid with hexahedral cells is created by layering crosssections perpendicular to the axial
direction. The cells on a crosssection are placed along gradient lines of the Laplace problem
solved for this crosssection. The Laplace problem is solved, for an unstructured triangular
mesh of the crosssection, with a mesh generator and solver of a commercial CFD package.
Gradient lines never cross each other, so a robust twodimensional grid is created. For the
next crosssection, a small displacement in axial direction, the screws are rotated slightly and
the Laplace problem is solved again. Merging the crosssections gives a threedimensional grid
i
of the screws of a twin screw pump. The quality of this threedimensional grid is examined.
The low screw pitch gives a relatively large tangential displacement compared to the axial
displacement, resulting in highly skewed cells. This reduces the applicability of this grid for
turbulent ﬂows. Elongating the screws in axial direction (higher pitch) gives a better quality
grid, however the original geometry is lost.
The leakage ﬂow through a twin screw pump has two paths, ﬁrst, leakage through the clear
ance between the screw and the liner, and secondly between the screws itself. The leakage
ﬂow through the twin screw pump is simulated for a diﬀerential pressure of up to 10 kPa per
screw thread (seal) on the threedimensional grid. For higher diﬀerential pressures the simu
lation does non converge. The simulated leakage ﬂow rate in the clearance between screw and
liner is approximately the same as the analytic laminar leakage ﬂow rate through a stationary
annulus. The relation between diﬀerential pressure and leakage ﬂow rate is determined for
low axial Reynolds numbers, and for higher axial Reynolds numbers using an elongated grid
in axial direction.
The static numerical simulation of the ﬂow in a twin screw pump show realistic ﬂow features.
The diﬀerential pressure per screw thread has to be increased to simulate real pump per
formance. The numerical model is created with the ability to add a dynamic mesh, this to
simulate the rotation of the screws. Also multiphase models can be added to predict leak
age ﬂow characteristics with liquidgas mixtures. For these extensions to the current model
recommendations are given.
ii
Nomenclature
Symbol Description [Unit]
a edge length of cell [m]
A
cl
projected area of clearance between screw and liner per
pendicular to the axial direction
[m
2
]
A
liner
area of two joined circles [m
2
]
A
screw
area of screw crosssection [m
2
]
D outer diameter screw [m]
d
h
hydraulic diameter [m]
e elongation of grid []
f
AR
aspect ratio of cell edges []
h center distance screws [m]
h
1
height of screw cavity [m]
GVF gas void fraction []
L length [m]
L screw thickness [m]
˙ m mass ﬂow rate [kg/s]
Ma Mach number []
N rotation speed [rpm]
n
ax
number of axial cells []
n
cell
number of cells in radial direction []
n
tan
number of tangential cells []
n
rev
number of screw threads (revolutions) []
n
str
number of structured cells []
n
unstr
number of unstructured cells []
R outer radius of screw [m]
r
i
inner radius of screw [m]
R
liner
radius of liner [m]
Re axial Reynolds number []
Re
ω
tangential Reynolds number []
s clearance between screw and liner [m]
p absolute pressure [Pa]
∆p pressure diﬀerence [Pa]
q grow rate []
r pitch of one screw thread [m/rev]
Q
c
l leakage ﬂow rate in clearance between screw and liner [m
3
/s]
iii
Symbol Description [Unit]
Q
l
leakage ﬂow rate of pump [m
3
/s]
Q
r
realized ﬂow rate of pump [m
3
/s]
Q
t
theoretical ﬂow rate of pump [m
3
/s]
u
τ
friction velocity [m/s]
v ﬂuid velocity [m/s]
v
ax
mean axial velocity [m/s]
V
cl
volume of clearance between screw and liner [m
3
]
V
D
screw displacement volume per revolution [m
3
/rev]
V
hex
volume of hexahedral cell [m
3
]
V
tet
volume of tetrahedral cell [m
3
]
y
+
dimensionless wall distance []
Greek symbols
Symbol Description [Unit]
θ angle [rad]
λ resistance coeﬃcient []
µ dynamic viscosity [kg/(ms)]
ν kinematic viscosity [m
2
/s]
τ shear tensor [N/m
2
]
τ
w
wall shear stress [N/m
2
]
Φ potential []
ρ density [kg/m
3
]
ω angular velocity of the screw [rad/s]
iv
Contents
Abstract i
Nomenclature iii
1 Introduction 1
1.1 Theory of screw pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Construction of screw pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Special multiphase applications . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Goal and outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Simpliﬁed leakage ﬂow in a twin screw pump 7
2.1 Reynolds number in twin screw pump . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Annulus with rotating inner cylinder . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.2 Numerical simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Labyrinth seal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3.1 Experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3.2 Numerical simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4 Leakage ﬂow in twin screw pump . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Method of threedimensional meshing 20
3.1 Number of cells in the clearance region . . . . . . . . . . . . . . . . . . . . . . 20
3.2 Dynamic mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.1 Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.2 Dynamic layering method . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.3 Remeshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.4 Arbitrary LagrangianEulerian calculations (ALE) . . . . . . . . . . . 24
3.3 Nonconformal mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4 Construction of structured grid 26
4.1 Basic structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2 Reﬁnements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.1 Dividing line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.2 Smoothing of nodes on gridline starting on the cusps . . . . . . . . . . 28
4.2.3 Nonequidistant node placement . . . . . . . . . . . . . . . . . . . . . 28
v
4.3 Grid evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3.1 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3.2 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4 Threedimensional basis structure . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5 CFD computations with structured grid 37
5.1 Numerical setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
6 Conclusions and recommendations 42
6.1 Final conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.2 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Bibliography 44
A Theoretical Screw proﬁle 45
B Screw proﬁle Houttuin 47
Acknowledgements 49
vi
Chapter 1
Introduction
Screw pumps are a special type of rotary positive displacement pump in which the ﬂow
through the pumping element is axial. The liquid is carried between screw threads on one
or more screws and is displaced axially as the screws rotate and intermesh (see ﬁgure 1.1).
In all other rotary pumps the liquid is forced circumferentially, thus giving the screw pump
its unique axial ﬂow pattern and low internal velocities a number of advantages in many
applications where liquid agitation or churning is not desired. Another property for a twin
screw pump, as opposed to centrifugal pumps is the capability of handling mixtures of liquid
and vapour. In this report only twin screw pumps are considered.
The applications of screw pumps cover a diversiﬁed range of markets including navy, marine,
and utilities fuel oil services; marine cargo; industrial oil burners; lubricating oil services;
chemical processes; petroleum and crude oil industries; power hydraulics for navy and machine
tools; and many others. The screw pump can handle liquids in a range of viscosities, from
molasses to gasoline, as well as synthetic liquids in a pressure range from 3.5 to 350 bar and
ﬂows up to 1820 m
3
/h. In this report only crude oil transportation is considered .
Because of relatively low inertia of their rotating parts, screw pumps are capable of operating
at higher speeds than other rotary or reciprocating pumps of comparable displacement. Screw
pumps, like other rotary positive displacement pumps, are selfpriming and have a delivery
ﬂow characteristic, which is essentially independent of pressure, provided there is suﬃcient
viscosity in the liquid being pumped.
Twin screw pumps are available in two conﬁgurations: single end and double end (see ﬁgure
1.2). Reference is made to [1] with respect to this chapter.
Figure 1.1: Diagrams of screw and gear elements, showing (a) axial and (b) circumferential
ﬂow
1.1 Theory of screw pumps 2
Figure 1.2: Twin screw pump with doubleend arrangement and internal timing gears
1.1 Theory of screw pumps
In screw pumps, it is the intermeshing of the threads on the screws and the close ﬁt of the
surrounding housing (liner) that create one or more sets of moving seals in a series between
the pump inlet and outlet. These sets of seals act as a labyrinth and provide the screw
with its positive pressure capability. The successive sets of seals form fully enclosed cavities
that move continuously from inlet to outlet, providing a smooth ﬂow. Because the screw
pump is a positive displacement device, it will deliver a deﬁnite quantity of liquid every
revolution of the screws. This delivery can be deﬁned in terms of displacement volume V
D
,
which is the theoretical volume displaced per revolution of the screws and is dependent only
upon the physical dimensions of the screws. This delivery can also be deﬁned in terms of
theoretical capacity or ﬂow rate Q
t
measured in cubic meters per second, which is a function
of displacement volume and speed N (rpm):
Q
t
=
1
60
V
D
N (1.1)
If no internal clearances existed, the pump’s actual delivered or net ﬂow rate Q
r
would equal
the theoretical ﬂow rate. Clearances, however, do exist with the result that whenever a pres
sure diﬀerential occurs, there will always be internal leakage from outlet to inlet. This leakage
Q
l
varies depending upon the pump type or model, the geometry of the clearance, the liquid
viscosity, the rotation speed of the screws, and the diﬀerential pressure. The delivery ﬂow rate
or net ﬂow rate is the theoretical ﬂow rate minus the leakage ﬂow rate: Q
r
= Q
t
−Q
l
. If the
diﬀerential pressure is almost zero, the leakage ﬂow rate may be neglected and Q
r
= Q
t
. The
theoretical ﬂow rate is not dependent on the diﬀerential pressure over a positive displacement
pump, see ﬁgure 1.3(a).
The theoretical ﬂow rate of any pump can readily be calculated if all essential dimensions
are known, see ﬁgure 1.3(b) for the screw dimension parameters. For any particular thread
1.2 Construction of screw pumps 3
∆p
Q
Q
t
Q
l
Q
r
N = constant
(a) Theoretic and realized volume ﬂow rate
against diﬀerential pressure for positive displace
ment pumps
r s
D
L
(b) Parameters
Figure 1.3: Twin screw pump
conﬁguration, assuming geometric similarity, the size of each cavity mentioned earlier is pro
portional to its length and crosssectional area. The thread pitch r measured in terms of
the same nominal diameter, which is used in calculating the crosssectional area, deﬁnes the
length. Therefore, the volume ﬂow rate of each cavity is proportional to the cube of this
nominal diameter and the speed of rotation N (rpm):
Q
t
= kD
3
N (1.2)
or writing it in terms of pitch,
Q
t
= k
1
· r · D
2
N (1.3)
where r = kD/k
1
with k and k
1
being constants.
Thus, for a given geometry, it can be seen that a relatively small increase in pump size
can provide a large increase in ﬂow rate. This is also true for centrifugal pumps which scale
according to Φ =
Q
ND
3
= constant. The theoretical ﬂow rate of centrifugal pumps is inﬂuenced
by the diﬀerential pressure over the pump, this is in contrast to positive displacement pumps
like screw pumps.
The leakage ﬂow rate can also be calculated, but it is usually estimated based on empirical
values obtained from extensive testing. These test data are the basis of the design parameters
used by every pump manufacturer. The leakage ﬂow rate generally varies approximately with
the square of the nominal diameter and linearly with the pressure diﬀerence. When neglecting
the eﬀect of rotation of the screws on the leakage ﬂow rate, the net ﬂow rate Q
r
can be written
as:
Q
r
= kD
3
N −k
2
∆pD
2
(1.4)
with k and k
2
empirical constants depending on the geometry and the working ﬂuid.
1.2 Construction of screw pumps
Design concepts The pressure gradient in the pump elements of all the types of screw
pumps produces various hydraulic reaction forces. The mechanical and hydraulic technique
1.2 Construction of screw pumps 4
employed for absorbing these reaction forces are one of the diﬀerences in the types of screw
pumps produced by various manufacturers. Another fundamental diﬀerence lies in the method
of engaging, or meshing, the screws and maintaining the running clearances between them.
Two basic design approaches are used:
• The timed screws approach is based on an external means for phasing the mesh of the
treads and for supporting the forces acting on the screws. In this concept, theoretically,
the threads do no come into contact with each other nor with the housing bores in which
they rotate.
• The untimed screws approach is based on the precision and accuracy of the screw forms
for the proper mesh and transmission of rotation. They utilize the housing bores as
journal bearings supporting the pumping reactions along the entire length of the screws.
Timed screw pumps require separate timing gears between the screws and separate support
bearings at each end to absorb the reaction forces and maintain the proper clearances. Un
timed screw pumps do not require gears or external bearings and thus are considerably simpler
in design.
Doubleend screw pumps The doubleend arrangement is basically two opposed, single
end pumps or pump elements of the same size with a common driving screw that has an
opposed, doublehelix design with one casing. As can be seen from ﬁgure 1.2, the ﬂuid enters
a common inlet with a split ﬂow going to the outboard ends of the two pumping elements and
is discharged from the middle or center of the pumping elements. The two pump elements are,
in eﬀect, pumps connected in parallel. For lowpressure applications, the design can pump
backwards when the direction of screw rotation is reversed. In either of these arrangements,
all axial loads on the screws are balanced, as the pressure gradients in each end are equal and
opposite. The doubleend screw pumps construction is usually limited to low and medium
pressure applications, with 28 bar being a good practical limit to be used for planning pur
poses. Doubleend pumps are generally employed where large ﬂows are required or where
highly viscous liquids are handled.
Timed design Timed screw pumps having timing gears and screw support bearings are
available in two general arrangements: internal and external. The internal version has both
the gears and the bearings located in the pumping chamber and the design is relatively simple
and compact. This version is generally restricted to the handling of clean lubricating ﬂuids,
which serve as the only lubrication for the timing gears and bearings.
The external timing arrangement is the most popular and is extensively used. It has both
the timing gears and screw support bearings located outside the pumping chamber. This
type can handle a complete range of ﬂuids, both lubricating and nonlubricating, and, with
proper materials, has good abrasion resistance. The timing gears and bearings are oilbath
lubricated from an external source. This arrangement requires the use of four stuﬃng boxes
or mechanical seals, as opposed to the internal type, which employs only one shaft seal.
The main advantage of the timed screw pump is that the timing gears transmit power to the
screws with no contact between the screw threads, thus promoting long pump life. The gears
and screws are timed at the factory to maintain the proper clearance between the screws.
The timing gears can be either spur or helical, herringbone, hardenedsteel gears with tooth
1.3 Special multiphase applications 5
proﬁles designed for eﬃcient, quiet, positive drive of the screws. Antifrictional radial bearings
are usually of the heavyduty roller type, while the trust bearings, which position the screws
axially, are either doublerow, ballthrust or sphericalroller types.
The housing can be supplied in a variety of materials, including cast iron, ductile iron, cast
steel, stainless steel, and bronze. In addition, the screw bores of the housing can be lined
with industrial hard chrome for abrasion resistance.
Since the screws are not generally in metallic contact with the housing or with one another,
they can also be supplied in a variety of materials, including cast iron, heattreated alloy steel,
stainless steel, Monel, and nitralloy. The outside of the screws can also be covered with a
variety of hard coating materials such as nickel based alloys, tungsten carbide, chrome oxide,
or ceramic.
1.3 Special multiphase applications
Screw pumps have been used with gasentrained application for many years, but recent process
changes in oilﬁeld technologies have created requirements for pumping multiphase ﬂuids,
containing more than just nominal amount of gases. In many oil well applications, the liquid
oil ﬂow eventually degenerates into all sorts of diﬃcult multiphase mixtures of oil, gas, water,
and sand. In the past, it was common for the gas to be separated and ﬂared oﬀ at the well
head with only the liquid product to be retained for further processing. If the gas is to be
processed as well, separators, compressors, and dual pipelines are required to handle the gas
phase. Therefore, a pump which can handle these diﬃcult liquids with high gas contents,
can save signiﬁcant equipment costs as well as operating costs. Under various conditions, the
well output can vary from 100 percent liquid to 100 percent gas, while maintaining the full
discharge pressure.
When pumping multiphase products with high gas void fractions, the pump must be designed
with a small pitch to provide a suﬃcient number of seals. The key to pumping multiphase
products is to ensure that some liquid is always available to seal the screw clearances and
reduce the leakage ﬂow rate. Even a small amount of recirculated liquid is suﬃcient to
provide this seal and enable the screw pump to operate with GVFs approaching 100 percent.
Depending on a number of factors, the volume of liquid required to seal and cool the screws
can be three to six percent of the total inlet volume ﬂow rate. In order to ensure that suﬃcient
liquid is available at conditions of high GVFs, a separate liquid ﬂush can be provided or a
separator type of body pump can be used. This type of body includes a special chamber
that can separate some liquid from the multiphase mixture being pumped. This liquid can
be recirculated back to the screws and mechanical seals to provide sealing and cooling liquid
at times when the product is almost all gas.
When pumping liquids, the leakage ﬂow rate through the internal clearances is proportional
to the diﬀerential pressure and inversely proportional to the viscosity (assuming laminar ﬂow
through the clearances). However, in multiphase applications, as the void fraction increases,
the leakage ﬂow rate decreases. See ﬁgure 1.4(a) for the typical pump performance when
pumping multiphase mixtures. Leakage ﬂow compresses the gas in the upstream screw cavity.
While the screw cavities move downstream, the gas void fraction reduces by the increased
pressure. Some leakage ﬂow ﬁlls the reduction in gas volume in the cavity, and a smaller
leakage ﬂow progresses to the next upstream cavity. This is clearly visible in ﬁgure 1.4(b).
When the leakage ﬂow rate is smaller, the diﬀerential pressure over a clearance is also smaller.
1.4 Goal and outline 6
The leakage ﬂow rate through the most upstream clearance is the total leakage ﬂow rate of a
twin screw pump for multiphase applications.
Q
Q = Q
t
GVF = 0.95
GVF=0
∆p
(a) Typical twin screw pump performance
curve (N=constant)
gas gas gas
liquid liquid liquid
Q
l
gas
liquid
FLOW
p
high
p
low
(b) Gas compression by the leakage ﬂow
Figure 1.4: Multiphase application of a twin screw pump
1.4 Goal and outline
The aim of the research is to predict the leakage ﬂow rate of a multiphase twin screw pump by
numerical simulations of the internal ﬂow. This research can be divided in three stages: ﬁrst,
single phase ﬂow in a nonrotating pump. Secondly, single phase ﬂow in a rotating pump and
thirdly, multiphase ﬂow in a rotating pump. This report is restricted to the ﬁrst stage and
mainly focusses on the method to generate a suitable computational grid.
In chapter 2, the characteristics of the leakage ﬂow are studied by considering two simpler
cases: the ﬂow through an annulus with rotating inner cylinder, and the ﬂow through a
straightthrough labyrinth seal. Recommendations for the grid and turbulence model are
given. In chapter 3, a method to create a threedimensional grid of the screw pump and
mesh methods for rotating screws is presented. It is used in chapter 4 to construct a three
dimensional grid of the twin screw pump. In chapter 5, the generated grid is used to simulate
the leakage ﬂow rate for nonrotating screws and singlephase ﬂow with the threedimensional
grid. Finally in chapter 6, the conclusion of the developed mesh method and simulated leakage
ﬂow rate are discussed. Also recommendations to expand the model for timedependent ﬂow
simulations and multiphase ﬂow are discussed in this chapter.
Chapter 2
Simpliﬁed leakage ﬂow in a twin
screw pump
In section 2.1 the ﬂow regime in the clearances of the screw is considered and the corresponding
Reynolds number is estimated. The leakage ﬂow in a twin screw pump is dependent on two
phenomena. First, the leakage ﬂow rate is dependent on friction in the small clearance
between the tips of the rotating screw and the stationary liner. Secondly, the leakage ﬂow
rate is dependent on the ﬂow resistance of the inlet and outlet of this clearance. These two
phenomena in the leakage area of a twin screw pump are represented by two characteristic
ﬂow cases. First, the ﬂow in an annulus with rotating inner cylinder. Secondly, ﬂow through
a stationary labyrinth seal. In section 2.2 the ﬂow in an annulus with rotating inner cylinder
is simulated for the corresponding Reynolds number of a twin screw pump. In section 2.3 the
leakage ﬂow through a stationary labyrinth seal is simulated for the corresponding Reynolds
number of a twin screw pump. For the annulus with rotating inner cylinder and the labyrinth
seal the simulated ﬂow is compared with experimental results and recommendations for grid
and turbulence model are given. In section 2.4 the simpliﬁed ﬂow from section 2.1 and 2.2
is applied for a twin screw pump to estimate the leakage ﬂow rate for a complete twin screw
pump. In section 2.5 recommendations are made for the grid and turbulence model.
2.1 Reynolds number in twin screw pump
The Reynolds number is an important parameter for the ﬂow regime. For pipe ﬂow a Reynolds
number smaller than 2100, based on hydraulic diameter and mean velocity, represents laminar
ﬂow and a higher Reynolds number represents turbulent ﬂow in general. In the clearance area
between screw and liner of a screw pump two directions of ﬂuid motion are present. First the
leakage ﬂow in axial direction and secondly the tangential motion of ﬂuid in the clearance.
The axial Reynolds number Re and the tangential Reynolds number Re
ω
are deﬁned as:
Re =
v
ax
sρ
µ
(2.1)
Re
ω
=
ρωRs
µ
(2.2)
2.1 Reynolds number in twin screw pump 8
where v
ax
is the mean axial ﬂow velocity, R the outer screw radius, ω the angular velocity
of the screw, s the clearance between screw and liner, ρ the density of the medium, and µ the
dynamic viscosity. The axial Reynolds number is normally based on the hydraulic diameter
d
h
= 2s but in equation (2.1) the notation of [11] is used. Note that with this notation a
axial Reynolds number larger than 1050 represents turbulent ﬂow in a pipe.
From these equations it is clear that the Reynolds numbers are dependent on dynamic viscosity
and density of the medium, and the clearance. The tangential Reynolds number is dependent
on the circumferential speed of the screw, while the axial Reynolds number depends on the
axial velocity, which is in turn dependent on the diﬀerential pressure over the screw. The
variables not concerning the geometry of the screw will be discussed point wise:
Viscosity A multiphase twin screw pump designed for pumping crude oil is considered. The
dynamic viscosity of crude oil varies between 1.4 · 10
−3
−20 · 10
−3
kg/(ms)
1
.
Tangential velocity The tangential velocity is dependent on the outer radius of the screw
and the rotational speed of the screws. The maximum rotational speed typical for twin screw
pumps designed for nonlubricating liquids is 1750 rpm
2
and for lubricating liquids 2900
rpm
3
.
Axial velocity The mean axial leakage velocity for HagenPoiseuille ﬂow in a cylindrical
annulus is given in equation (2.3). This value of v
ax
may serve as a ﬁrst estimate since no
axial movement of the annulus is taken into account. Also laminar ﬂow is considered in this
estimation. When the leakage ﬂow is turbulent, the turbulent ﬂow proﬁle and the rotation of
the screw produces deviations from this estimated mean axial velocity. Reference is made to
[2] for this equation.
v
ax
=
∆ps
2
12µL
(2.3)
where L is the length of the annulus, ∆p the pressure diﬀerence, s the clearance, and µ the
dynamic viscosity.
With this mean axial velocity v
ax
and the twin screw pump dimensions the axial Reynolds
number can be estimated. The diﬀerential pressure over one seal is the total pressure build
up divided by the number of seals, considering single phase ﬂow. The total number of seals is
dependent on the number of screw threads. In ﬁgure 2.1 a schematic view and a photograph
of two screw threads are given. The cavities A and B in this ﬁgure are connected. To seal one
cavity at least 2 screw threads are necessary, but normally just over 2 screw threads are used
to ensure proper sealing. For example: 5
1
3
screw threads seal 4 cavities. Thus the typical
total pressure build up over the pump is 16 bar
2 3
at maximum. Assuming 4 seals, this
gives a diﬀerential pressure ∆p of 4 bar per seal.
The axial Reynolds number Re is estimated at 10,000 and the tangential Reynolds number
Re
ω
is 3000 for crude oil with the parameters given in table 2.1.
In this estimation of the axial Reynolds number the following phenomena are neglected:
• Axial ﬂow proﬁle is turbulent, resulting in an even higher axial velocity and Reynolds
number.
• No inlet and outﬂow resistances are considered, and the axial velocity will be lower
resulting in a lower axial Reynolds number.
1
reference to http://www.engineeringtoolbox.com
2
Houttuin 216.10 screw pump, http://houttuin.nl/contents/21610bro.pdf
3
Houttuin 249.40 screw pump, http://houttuin.nl/contents/24940bro.pdf
2.2 Annulus with rotating inner cylinder 9
A
B
(a) Schematic view (b) Photograph
Figure 2.1: Two threads of twin screw pump
Table 2.1: Parameters for Reynolds numbers of a twin screw pump
µ 1.4 · 10
−3
kg/(ms)
ρ 800 kg/m
3
N 1350 rpm
∆p 4 bar
L 22 mm
R 147 mm
s 0.25 mm
• The screw translates opposite to the leakage ﬂow, so the axial ﬂow proﬁle will be diﬀer
ent. The mean axial velocity will be lower resulting in a lower axial Reynolds number.
The axial screw velocity is given by (N/60) · r, where N is the rotational speed in rpm,
and r the screw pitch in m. The axial screw velocity is 1.35 m/s at 1350 rpm and a
screw thread of 60 mm. This is low compared to the estimated mean axial velocity of
69.3 m/s.
• The screw rotates and this leads to a tangential motion of the ﬂuid and, at high rotation
speeds, to the occurrence of Taylor vortices in the ﬂuid. This secondary ﬂow leads to an
additional resistance and a lower axial velocity and axial Reynolds number. In section
2.2 this rotation is considered.
2.2 Annulus with rotating inner cylinder
The resistance of a water ﬂow in an annulus with a rotating inner cylinder is studied experi
mentally in [11]. For the CFD simulations, these experiments will serve as a testcase which
will give recommendations for grid properties and turbulence model.
2.2 Annulus with rotating inner cylinder 10
2.2.1 Experiment
The geometry of the test setup used in [11] is given in ﬁgure 2.2. In this ﬁgure the ﬂuid
ﬂows from left to right. The pressure is measured at three locations (N
I
, N
II
and N
III
) and
also the ﬂow rate is measured. As is apparent from the ﬁgure, the pressure is measured at
some distance from the inlet to minimize inlet ﬂow eﬀects. At every measuring location four
holes of 0.4 mm in diameter and 90
o
apart in the outer cylinder are connected to minimize
measuring errors.
Anticipating a turbulent ﬂow, the pressure drop ∆p over the annulus is written as:
∆p
ρ
=
λ
L
2s
+ δ
io
v
ax
2
2
(2.4)
with λ the resistance coeﬃcient, δ
io
the resistance factor for the inlet and the outlet, L
the axial length, s the clearance between inner and outer cylinder of the annulus, v
ax
the
mean axial velocity, and ρ the density. In ﬁgure 2.3, the resistance coeﬃcient λ of various
experiments with tangential Reynolds number Re
ω
between 1000 and 20000 are given for axial
Reynolds numbers Re in the range 100 to 25000. The axial Reynolds number for leakage ﬂow
in the clearances between screw and liner is more accurately estimated with equation 2.4. A
δ
io
of 1.5 is chosen according to [8] for the resistance of the in and outlet. The axial Reynolds
number is estimated at 2230.
For ﬂow in an annulus with rotating inner cylinder, there are no inlet and outlet resistances,
and δ
io
is set to zero.
Figure 2.2: Experimental setup of an annulus with rotating inner cylinder (dimensions in
mm)
In table 2.2 the dimensions of the smallest annulus considered in the experimental study are
given, and compared with typical dimensions of a screw pump. The clearance ratio s/R of a
screw is much smaller than that of the annulus. It can be concluded that the clearance ratio
does have an eﬀect on the resistance coeﬃcient λ (see ﬁgure 2.3). Therefore, the leakage ﬂow
rate of a screw pump can not be accurately determined from the measurements. However,
this diﬀerence has no inﬂuence on recommendations for grid and turbulence model, because
the exact geometry of the annulus is used in the numerical simulations.
2.2.2 Numerical simulation
For Re = 2230 and Re
ω
= 3000 a numerical simulation of ﬂow in an annulus with rotating
inner cylinder is performed with the commercial CFDpackage Fluent. These Reynolds num
bers indicate a turbulent ﬂow. This is a steady axisymmetric problem, and is solved using
2.2 Annulus with rotating inner cylinder 11
Figure 2.3: Relation between resistance coeﬃcient λ and tangential Reynolds number Re
ω
for various
s
R
and axial Reynolds number Re
Table 2.2: Clearance ratio s/R for a twin screw pump and an annulus with rotating inner
cylinder
Radius R [mm] Clearance s [mm] Clearance ratio s/R []
Annulus 31.72 0.43 0.0136
Screw pump 147.25 0.25 0.0017
the segregated solver and implicit formulation. The segregated solver is used because in Flu
ent this solver is capable of adding multiphase models. The alternative, a coupled solver, is
not capable of adding multiphase models. The coupled solver in Fluent solves the governing
equations of continuity, momentum, and if necessary energy simultaneously as a set of equa
tions. The segregated solver solves the governing equations sequentially (segregated from one
another). The segregated solver can only be used with implicit formulation in Fluent. The
continuity and momentum equations solved by the segregated solver are given in diﬀerential
form and cylindrical coordinates as:
∂ρ
∂t
+
1
r
∂
∂r
(ρru
r
) +
1
r
∂
θ
(ρu
θ
) +
∂
∂z
(ρu
z
) = 0
ρ
∂u
r
∂t
+ u
r
∂u
r
∂r
+
u
θ
r
∂u
r
∂θ
+ u
z
∂u
r
∂z
−
u
2
θ
r
= −
∂p
∂r
−
¸
1
r
∂
∂r
(rτ
rr
) +
1
r
∂
∂θ
τ
θr
+
∂
∂z
τ
zr
−
τ
θθ
r
2.2 Annulus with rotating inner cylinder 12
ρ
∂u
θ
∂t
+ u
r
∂u
θ
∂r
+
u
θ
r
∂u
θ
∂θ
+ u
z
∂u
θ
∂z
−
u
r
u
θ
r
= −
1
r
∂p
∂θ
−
¸
1
r
2
∂
∂r
(r
2
τ
rθ
) +
1
r
∂
∂θ
τ
θθ
+
∂
∂z
τ
zθ
ρ
∂u
z
∂t
+ u
r
∂u
z
∂r
+
u
θ
r
∂u
z
∂θ
+ u
z
∂u
z
∂z
= −
∂p
∂z
−
¸
1
r
∂
∂r
(rτ
rz
) +
1
r
∂
∂θ
τ
θz
+
∂
∂z
τ
zz
(2.5)
where ρ is the density, u the ﬂuid velocity, τ the shear tensor, and p the pressure. The term
∂
∂θ
is zero for axisymmetric problems. In case of steady state calculations the
∂
∂t
term is also
zero. The equations of continuity and momentum are implemented in the SIMPLE method
(Semi Implicit Method for Pressure Linked Equations). This method calculates the ﬂow ﬁeld
with an estimate of the pressure. After this, the pressure is corrected with the solution of the
ﬂow ﬁeld to form a new estimate of the pressure. In this way iteratively a solution for the
ﬂow ﬁeld and the pressure can be calculated that satisﬁes both continuity and momentum
equations. The equations of the SIMPLE method are solved with the GaussSeidel algorithm
for an Algebraic MultiGrid (AMG). Discussing these methods falls outside the scope of this
report. In [4] and [7] these methods are discussed.
Numerical models have two approaches to model turbulent ﬂow behaviour near solid walls;
wall functions and nearwall modeling approaches. Wall functions use semiempirical formulas
to resolve the ﬂow in large cells near the wall. Nearwall modeling uses small cells near the
wall to resolve the ﬂow. The approaches each have diﬀerent requirements on the size of the
ﬁrst cell at the wall. This is expressed in the dimensionless wall distance y
+
, deﬁned as:
y
+
≡ ρu
τ
y/µ (2.6)
where y is the distance of the center of the element to the wall, µ the dynamic viscosity and u
τ
the friction velocity given by
τ
w
ρ
where τ
w
is the wallshear stress deﬁned as τ
w
= µ
∂u
∂y
y=0
.
The required y
+
for the wall function approach is 30 < y
+
< 200. A value closer to the lower
bound is most desirable. The required y
+
for nearwall modeling is of the order of 1. The
upper bound is y
+
< 4 ∼ 5. Another criterion for nearwall modeling is that there are at least
10 cells within the viscosityaﬀected boundary layer. In [4] en [6] these methods are discussed
extensively.
In the remainder of this section, numerical results will be compared with results from ex
periments as given in ﬁgure 2.3. Values of resistance coeﬃcient λ at intermediate Reynolds
numbers are determined by linear interpolation. In the simulations, two nearwall approaches
will be considered: ﬁrst wall functions will be evaluated, and secondly nearwall modeling.
Wall functions
With the k − ǫ model with standard wall functions, the value of y
+
is calculated for grids
with diﬀerent number of cells in radial direction. The value of y
+
is solution dependent: it
will vary slightly for diﬀerent turbulence models, but it gives a good idea of the dimensionless
wall distance with that grid. In table 2.3 the dimensionless wall distance is given for diﬀerent
number of cells n
cell
in radial direction. The required dimensionless wall distance 30 < y
+
<
200 is not reached with these Reynolds numbers (Re = 2230 and Re
ω
= 3000). The calculated
resistance coeﬃcient does not seem to be inﬂuenced much by the dimensionless wall distance.
To maintain some radial cells in the annulus and to keep the value of y
+
close to the required
value, n
cell
= 5 is chosen to determine the accuracy of the diﬀerent turbulence models.
2.2 Annulus with rotating inner cylinder 13
Table 2.3: Dimensionless wall distance y
+
and resistance coeﬃcient λ for diﬀerent grids of
the annulus problem determined with the standard k − ǫ model and standard wall function
(Re = 2230, Re
ω
= 3000)
n
cell
y
+
λ
num
4 22 0.0455
5 18 0.0461
6 15 0.0464
7 13 0.0466
8 11 0.0466
In table 2.4 the calculated resistance coeﬃcient λ is given for the k −ǫ and k −ω turbulence
models with various model options and wall functions. Furthermore, the diﬀerence between
the calculated and the experimental value of the resistance coeﬃcient λ is given. A struc
tured grid with equilateral edges is used, see ﬁgure 2.5(a). The axial length of the structured
elements is not considered of importance due to the axial nature of the ﬂow. Additional
numerical simulations with elongated cells in axial direction (2, 5 and 10 times) give the same
resistance coeﬃcients λ.
Table 2.4: Resistance coeﬃcient λ for various turbulence models with wall functions deter
mined for the annulus problem (n
cell
= 5, y
+
= 18, λ
exp
= 0.0579,Re = 2230, Re
ω
= 3000)
turb. model version option wall function λ
num
dev [%]
k −ǫ standard standard 0.0461 20.4
nonequilibrium 0.0477 17.6
RNG standard 0.0447 22.8
nonequilibrium 0.0463 20.0
diﬀerential standard 0.0421 27.3
nonequilibrium 0.0442 23.7
swirl standard 0.0442 23.7
nonequilibrium 0.0459 20.7
diﬀ+swirl standard 0.0419 27.6
nonequilibrium 0.0422 27.1
realizable standard 0.0452 21.9
nonequilibrium 0.0469 19.0
k −ω standard 0.0554 4.3
shear ﬂow corr. 0.0544 6.0
SST 0.0553 4.5
The k − ω turbulence models are most accurate in this situation. In ﬁgure 2.4 the y
+

dependency for the Standard and SST k −ω model is given, using grids with diﬀerent values
of n
cell
. Also the results for the standard k − ǫ turbulence model are given in this ﬁgure.
2.2 Annulus with rotating inner cylinder 14
The dependency on y
+
is obvious, over prediction for low y
+
values and under prediction for
higher y
+
values, for the k − ω turbulence model. Still, these models can be used because
the deviation is smaller than with the k − ǫ models. In [6] this dependency on y
+
was also
mentioned for k −ω turbulence models.
10 12 14 16 18 20 22 24
−25
−20
−15
−10
−5
0
5
10
y
+
D
e
v
i
a
t
i
o
n
[
%
]
n
cell
= 8
7
6
5 4
n
cell
= 8
7
6
5
4
standard k−ω
SST k−ω
standard k−ε
Figure 2.4: Relation between deviation of calculated and experimental resistance coeﬃcient
λ for diﬀerent y
+
values for the annulus problem (Re = 2230, Re
ω
= 3000, standard wall
function)
The standard k − ω model is used to calculate the resistance coeﬃcient for diﬀerent axial
and tangential Reynolds numbers. The deviation with the experimental results and the
dimensionless wall distance are given in table 2.5. There is a high deviation at y
+
= 10.5
which could be explained by the linear (laminar) law that Fluent employs for turbulent
boundary layers at approximately y
+
< 11.
Table 2.5: Resistance coeﬃcient λ for diﬀerent Re combinations with standard k − ω model
and wall functions (n
cell
= 6)
Re Re
ω
y
+
λ
num
λ
exp
dev [%]
1000 2000 10.5 0.105 0.076 38
3000 12 0.112 0.098 14
6000 17 0.135 0.152 11
2500 2000 17 0.0522 0.0457 14
3000 18 0.538 0.543 1
6000 21 0.0616 0.068 9
2.3 Labyrinth seal 15
(a) wall function (n
cell
= 5) (b) nearwall modeling (n
cell
= 15)
Figure 2.5: Partial mesh of the annulus with rotating inner cylinder for diﬀerent wall ap
proaches
Nearwall modeling
Similar to the previous section, ﬁrst the dimensionless wall distance will be calculated for
diﬀerent grids, while keeping in mind that the total number of cells should be kept as low as
possible. Three diﬀerent grids are considered, n
cell
equal to 10, 15 and 20 and a smooth grid
reﬁnement towards the solid walls (ﬁgure 2.5(b)). In table 2.6 the corresponding y
+
values of
the diﬀerent grids are given. For Re = 2230 and Re
ω
= 3000 the dimensionless wall distance
with n
cell
= 10 is too high. Another criterion states that there are at least 10 cells in the
viscosity aﬀected region [4]. This would bring the total number of cells up, so simulations are
performed to test this criterion. For n
cell
= 15 the resistance coeﬃcient is determined for the
diﬀerent turbulence models with nearwall modeling, results are given in table 2.7.
Table 2.6: Dependency of dimensionless wall distance y
+
for diﬀerent grids determined with
standard k − ω turbulence model and nearwall modeling for the annulus problem (Re =
2230, Re
ω
= 3000)
n
cell
y
+
λ
num
10 7 0.0517
15 3.2 0.0548
20 1.7 0.0632
2.3 Labyrinth seal
Leakage ﬂow in a twin screw pump has a stream path that is similar to the stream path through
a ’straightthrough labyrinth seal’. The results of numerical simulations are compared with
the experimental results to give recommendations for the grid and turbulence model.
2.3.1 Experimental setup
Airﬂow through a stationary straightthrough labyrinth seal is studied experimentally in [5].
In this study, air ﬂows through the seal with axial Reynolds numbers in a range of 300 to 7500
2.3 Labyrinth seal 16
Table 2.7: Resistance coeﬃcient λ for various turbulence models with nearwall modeling for
the annulus problem (n
cell
= 15, y
+
= 3, λ
exp
= 0.0579, Re = 2230, Re
ω
= 3000)
turb. model version option λ
num
dev. [%]
k −ǫ standard 0.0572 1.2
RNG 0.0573 1.0
diﬀerential 0.0578 0.2
swirl 0.0574 0.9
diﬀ+swirl 0.0578 0.2
realizable 0.0578 0.2
k −ω standard 0.0548 5.4
shear ﬂow corr. 0.0506 12.6
SST 0.0543 6.2
in the seal clearance. In ﬁgure 2.6 a schematic view of the labyrinth seal is given. In table
2.8 the dimensions of the labyrinth seal and the twin screw pump are given. The clearance
s and the diameter of the labyrinth seal D are similar to the screw pump, the length of the
clearance L and the pitch r are considerably smaller than for the screw pump. Inlet and
outﬂow resistances are the main sealing principle of a labyrinth seal. For a twin screw pump
the sealing between screw cavities is also established by a relatively long sealing clearance.
The total leakage ﬂow rate of the screw pump can not be estimated directly with the leakage
ﬂow rate through the labyrinth seal, because of this diﬀerence in sealing principle and the
nonrotating seal. The larger sealing length of the clearance has to be taken into account to
estimate the leakage ﬂow rate in a screw pump. However, this diﬀerence has no inﬂuence on
recommendations for grid and turbulence model.
h
1
r L
s
D
outlet
station 2 station 1
Figure 2.6: Experimental setup of straight
through labyrinth seal
Labyrinth Twin screw
seal pump
s [mm] 0.36 0.25
D [mm] 356 295
L [mm] 0.25 22
r [mm] 6 60
h
1
[mm] 6 85
Table 2.8: Dimensions of straightthrough
labyrinth seal and twin screw pump
2.3.2 Numerical simulation
A numerical simulation in Fluent is performed to calculate the mass ﬂow rate through the
labyrinth seal for Re equal to 1572 and 2195. The Mach number Ma is a measure for the
variation of density, according to Ma
2
∝ ∆p. A density variation smaller than ten percent is
2.4 Leakage ﬂow in twin screw pump 17
Table 2.9: Experimental results labyrinth seal
p
1
[kPa] p
2
[kPa] ˙ m [kg/s] ρ [kg/m
3
] Re [] Ma []
100.695 92.503 0.027303 1.1738 1572 0.20
108.793 93.806 0.038118 1.2680 2195 0.28
present here, and compressibility is neglected in the numerical simulation. Diﬀerent turbu
lence models are evaluated with the grid given in ﬁgure 2.7. This is a structured quadrilateral
grid with 5 radial cells in the seal clearance. In Fluent a steady axisymmetric problem is
solved with the segregated solver and implicit formulation using standard wall function for
incompressible ﬂow. Anticipating the results presented in chapter 3, it is concluded that the
number of cells should be kept as low as possible. Therefore, only wall functions are consid
ered for the simulation of ﬂow through a labyrinth seal. In table 2.10 the calculated mass ﬂow
rate and the deviation with the experimental mass ﬂow rate is given. The ﬂow through the
labyrinth seal results in a vortex in the seal cavity, which is clearly visible at the streamlines
of the ﬂow in the labyrinth seal, given in ﬁgure 2.8.
Figure 2.7: Structured quadrilateral grid of straightthrough labyrinth seal
Table 2.10: Mass ﬂow rate in a stationary labyrinth seal for various turbulence models using
wall functions
Re = 1572, y
+
> 11, Re = 2195, y
+
> 14,
˙ m = 0.0273 ˙ m = 0.0381
˙ m [kg/s] dev [%] ˙ m [kg/s] dev [%]
k −ǫ Standard 0.0281 2.9 0.0399 4.7
RNG 0.0294 7.7 0.0421 10.4
Realizable 0.0299 9.5 0.0428 12.3
k −ω Standard 0.0267 2.2 0.0379 0.6
SST 0.0304 11.4 0.0437 14.6
2.4 Leakage ﬂow in twin screw pump
The ﬂow through an annulus gives an idea of the leakage ﬂow rate in the clearances between
the screw and the liner of a screw pump. In this section the leakage ﬂow is estimated with
the method to calculate the mean axial ﬂow through a stationary annulus and an annulus
2.4 Leakage ﬂow in twin screw pump 18
Figure 2.8: Streamlines of the ﬂow in a straightthrough labyrinth seal
with rotating inner cylinder. The diﬀerence in clearance ratio s/R between the twin screw
pump and the annulus with rotating inner cylinder is neglected.
The leakage ﬂow rate through the clearance between screw and liner Q
cl
of a twin screw pump
is approximately given by the following equations:
Q
cl
= v
ax
A
cl
(2.7)
A
cl
=
4π −4 cos
h/2
R
Rs (2.8)
where h is the distance between the centers of the screws, s the clearance between screw and
liner, R the outer radius of the screw, and A
cl
the projected area of the clearance on a plane
perpendicular to the axial direction.
The ratio between the total leakage ﬂow rate and the theoretical pumped volume ﬂow rate
of a screw is an important pump performance parameter. The leakage ﬂow rate through the
clearance between screw and liner is part of the total leakage ﬂow, and is considered in the
ratio Q
cl
/Q
t
. For the theoretical ﬂow rate Q
t
reference is made to chapter 1. The theoretical
volume ﬂow rate is given by:
Q
t
=
1
60
V
D
N (2.9)
V
D
= (A
liner
−2 · A
screw
) · r (2.10)
where N is the rotational speed of the screws in rpm, V
D
the displacement volume for one
revolution, r the screw pitch, A
screw
the area of the crosssection of the screw, A
liner
the area
of the crosssection of the liner.
The displacement volume V
D
for the screw is determined by the dimensions of the screw and
liner. For the screw given in appendix B, the A
screw
= 37· 10
−3
m
2
, and A
liner
= 124· 10
−3
m
2
.
This gives a displacement volume V
D
of 3 · 10
−3
m
3
/rev.
The leakage ﬂow rate in a twin screw pump is estimated in three ways. First, using equa
tion (2.3) assuming laminar leakage ﬂow through a stationary seal without inlet and outlet
resistances. Secondly, using equation (2.4) for a turbulent leakage ﬂow through an annulus
with rotating inner cylinder, also without inlet and outlet resistances. Thirdly, using equation
(2.4) for a turbulent leakage ﬂow through an annulus with rotating inner cylinder, now with
a value of 1.5 for the inlet and outlet resistance δ
io
. For the annulus with rotating inner
cylinder an iterative procedure is applied, since the resistance coeﬃcient λ depends on axial
Reynolds number, and thus on the leakage ﬂow rate. In table 2.11 the leakage ﬂow rate is
given for the diﬀerent methods.
2.5 Conclusion 19
Table 2.11: Estimated leakage ﬂow of oil through a twin screw pump (app. B) for Re
ω
=
3000, ∆p = 4 bar, N = 1350 rpm
v
ax
Re Q
t
Q
cl
Q
cl
/Q
t
[m/s] [] [m
3
/s] [m
3
/s] [%]
stationary + laminar 69.3 10,000 0.0675 0.0246 36.4
rotating + turbulent 22.3 3220 0.0675 0.0079 11.7
rotating + turbulent + inlet/outlet losses 15.7 2230 0.0675 0.0056 8.3
2.5 Conclusion
The leakage ﬂow in the clearance between screw and liner of a twin screw pump is dependent
on two phenomena. First, the leakage ﬂow rate is dependent on friction in the small clearance
between the tips of the rotating screw and the stationary liner. Secondly, the leakage ﬂow
rate is dependent on the ﬂow resistance of the inlet and outlet of this clearance. These two
phenomena in the leakage area of a twin screw pump are represented by two characteristic
ﬂow cases. First, the ﬂow in an annulus with rotating inner cylinder. Secondly, ﬂow through
a stationary labyrinth seal. The axial Reynolds number in the clearance is determined by the
ﬂow in an annulus. Anticipating the results as presented in chapter 3 it is concluded that the
number of radial cells must be limited and that nearwall modeling is not preferable.
The k − ω turbulence model performs best for the ﬂow in an annulus with rotating inner
cylinder using wall functions, and performs good for the ﬂow in an annulus using nearwall
modeling. The standard k − ω and standard k − ǫ turbulence model perform best for the
ﬂow in a labyrinth seal. In [6], the ﬂow phenomena in an internal combustion chamber were
represented by two characteristic ﬂow cases: the backward facing step and the free jet ﬂow.
Based on diﬀerent grid reﬁnements and diﬀerent turbulence models, it was concluded that
the standard k −ω model performs good, but has a grid dependency. In this study, however,
it is shown that this dependency has little eﬀect on the accuracy to represent the leakage ﬂow
rate in an annulus with rotating inner cylinder.
The k −ǫ turbulence model performs best for ﬂow in an annulus using nearwall modeling.
Nearwall modeling is not preferred, and the standard k −ω turbulence model with wall func
tions is preferred for the simulation of ﬂow in a twin screw pump.
The axial Reynolds number in the clearance between screw and liner is estimated in three
ways. First, considering laminar ﬂow in a stationary annulus with rotating inner cylinder
without inlet and outlet resistances. Secondly, considering turbulent ﬂow in an annulus with
rotating inner cylinder without inlet and outlet resistances. Thirdly, turbulent ﬂow in an
annulus with rotating inner cylinder with inlet and outlet resistances. The axial Reynolds
number reduces when rotation and inlet and outlet resistances are considered.
Chapter 3
Method of threedimensional
meshing
In chapter 2, twodimensional leakage ﬂow in screw pumps is discussed. For threedimensional
ﬂow calculations the number of cells can be very high. The calculation time is dependent on
the number of cells used to simulate the ﬂow in a screw pump. Therefore, a limited number
of cells is desirable. Furthermore the quality of the grid has to be acceptable to perform
meaningful ﬂow calculations. In the ﬁrst section of this chapter the number of cells for an
unstructured, and a structured grid is estimated assuming cells with perfect quality (based on
skewness). A timedependent simulation of ﬂow in a twin screw pump requires the screws to
rotate, thereby the cavities between the screws progress to the pump outlet. This requires a
dynamic mesh. Methods for a dynamic mesh are discussed in section 3.2. While the screws are
rotating, the inlet and outlet remain stationary. This could give problems connecting those
regions if nodes should always coincide. A nonconformal grid does not require that nodes
coincide on the combined faces. This is discussed in section 3.3. Finally, recommendations
are made to setup a timedependent simulation of a screw pump in section 3.4.
3.1 Number of cells in the clearance region
The number of cells in the clearance region of a screw pump is estimated because the axial
distance L of the clearance is high compared to the radial distance s of the clearance between
screw and liner. Therefore, the number of cells in the clearance constitutes a substantial part
of the total number of cells to describe a complete twin screw pump. First the number of cells
in the clearance is estimated in case an unstructured tetrahedral mesh is used. After this,
the number of cells is estimated in case a structured mesh is used. The approximate volume
of the clearance region (ﬁgure 3.1(a)) is given by the following equation:
V
cl
= n
rev
LA
cl
(3.1)
where n
rev
is the number of screw threads, L the (axial) screw thickness, and A
cl
the cross
sectional area of the clearance, given in equation (2.8).
In an unstructured tetrahedral mesh of high quality, all cells are truly tetrahedral with edges
of equal length. The volume of such a regular tetrahedral cell V
tet
is given by:
V
tet
=
1
12
a
3
√
2 (3.2)
3.1 Number of cells in the clearance region 21
(a) Clearance between screw and liner (b) 5 threads of the screws
Figure 3.1: Twin screw pump
where a is the edge length of the cell. The length a is determined from the required number
of cells in radial direction from screw to liner, n
cell
. The height h of a regular tetrahedral
cell is h =
a
3
√
6. In ﬁgure 3.2 the height h is given by h =
s
n
cell
. This gives the following
expression for a:
a =
3s
√
6 n
cell
(3.3)
The estimated number of unstructured cells in the clearance between screw and liner is given
by:
n
unstr
=
V
cl
V
tet
=
n
rev
LA
cl
1
12
a
3
√
2
(3.4)
With a from equation (3.3), and n
cell
= 5 according to chapter 2, the number of unstructured
cells in the clearance n
unstr
is approximately 1.5 · 10
9
cells for a twin screw pump. The global
dimensions of the twin screw pump are given in table 2.1 and in appendix B a complete
drawing of the twin screw pump is given.
s
1
2
n
cell
·
·
Figure 3.2: Triangular cells in clearance between screw and liner
The number of structured cells in the clearance between screw and liner is estimated, similar
to the unstructured case. The skewness of structured hexahedral cells is not aﬀected by the
aspect ratio of the edges of the cell f
AR
. In ﬁgure 3.3 the structured mesh is given.
The volume of a hexahedral cell is given by:
V
hex
=
a
3
f
AR
(3.5)
3.2 Dynamic mesh 22
s
2
2
·
·
n
cell
1
· 1 · · · · · m ·
a a/f
AR
a
Figure 3.3: Hexahedral cells in clearance between screw and liner with aspect ratio f
AR
The length a of the edge of a cell is determined by clearance s, the number of cells in radial
direction across the clearance, n
cell
, and the chosen value for the aspect ratio f
AR
. The length
a is then given by:
a =
s f
AR
n
cell
(3.6)
The estimated number of structured cells in the clearance between screw and liner is given
by:
n
str
=
V
cl
V
hex
=
n
rev
bA
cl
f
AR
a
3
(3.7)
With a from equation (3.6), V
cl
from equation (3.1), f
AR
= 10, and n
cell
= 5, the number of
unstructured cells in the clearance n
str
is approximately 3.3 · 10
6
cells.
3.2 Dynamic mesh
In this section the distortion of the mesh is considered. In Fluent a dynamic mesh can be
achieved by repositioning of nodes. The nodes can be repositioned by smoothing, dynamic
layering and remeshing. Another dynamic mesh method is the use of ALEcalculations, this is
not a method to reposition the nodes. ALEcalculations are a method to perform calculations
on cells with nodes moving in time, while maintaining the grid topology. Still an initial mesh
has to be generated before repositioning and ALEcalculations can be performed. For ALE
calculations, grids have to be generated with the same grid topology for every time step.
3.2.1 Smoothing
In the spring based smoothing, the edges between any two mesh nodes are idealized as
a network of interconnected springs. The initial spacings of the edges before any boundary
motion constitute the equilibrium state of the mesh. A displacement at a given boundary
node will generate a force proportional to the displacement along all the springs connected
to the node. For nontetrahedral cell zones (nontriangular in 2D), the springbased method
is recommended when the following conditions are met:
• The boundary of the cell zone moves predominantly in one direction (i.e., no excessive
anisotropic stretching or compression of the cell zone).
• The motion is predominantly normal to the boundary zone.
3.2 Dynamic mesh 23
If these conditions are not met, the resulting cells may have high skewness values, since not
all possible combinations of node pairs in nontetrahedral cells (or nontriangular in 2D) are
idealized as springs.
Laplacian smoothing is the most commonly used and the simplest mesh smoothing method.
This method adjusts the location of each mesh node to the geometric center of its neighboring
nodes. This method is computationally inexpensive but it does not guarantee an improve
ment on mesh quality, since repositioning a node by Laplacian smoothing can result in poor
quality elements. To overcome this problem, Fluent only relocates the node to the geometric
center of its neighboring nodes if and only if there is an improvement in the mesh quality
(i.e., the skewness has been improved).
3.2.2 Dynamic layering method
In prismatic (hexahedral and/or wedge) mesh zones, one can use dynamic layering to add
or remove layers of cells adjacent to a moving boundary, based on the height of the layer
adjacent to the moving surface. The dynamic mesh model in Fluent allows to specify an ideal
layer height on each moving boundary. The layer of cells adjacent to the moving boundary
is split or merged with the layer of cells next to it based on the height of the cells in layer
attached to the moving boundary.
3.2.3 Remeshing
On zones with a triangular or tetrahedral mesh, the springbased smoothing method is nor
mally used. When the boundary displacement is large compared to the local cell sizes, the cell
quality can deteriorate or the cells can become degenerate. This will invalidate the mesh (e.g.,
result in negative cell volumes) and consequently, will lead to convergence problems when the
solution is updated to the next time step. To circumvent this problem, Fluent agglomerates
cells that violate the skewness or size criteria and locally remeshes the agglomerated cells or
faces. If the new cells or faces satisfy the skewness criterion, the mesh is locally updated with
the new cells (with the solution interpolated from the old cells). Otherwise, the new cells are
discarded. Fluent includes several remeshing methods that include local remeshing, local face
remeshing (for 3D ﬂows only), face region remeshing, and 2.5D surface remeshing (for 3D
ﬂows only). The available remeshing methods in Fluent work for triangulartetrahedral zones
and mixed zones where the nontriangular/tetrahedral elements are skipped. The exception
is the 2.5D model, where the available remeshing method only work on wedges extruded from
triangular surfaces.
Using the Local remeshing method, Fluent marks cells based on cell skewness and min
imum and maximum length scales as well as an optional sizing function. Fluent evaluates
each cell and marks it for remeshing if it meets one or more of the following criteria:
• It has a skewness that is greater than a speciﬁed maximum skewness.
• It is smaller than a speciﬁed minimum length scale.
• It is larger than a speciﬁed maximum length scale.
• Its height does not meet the speciﬁed length scale (at moving face zones, e.g., above a
moving piston).
3.2 Dynamic mesh 24
Face region remeshing method
In addition to remeshing the volume mesh, Fluent also allows triangular and linear faces on a
deforming boundary to be remeshed. Fluent marks deforming boundary faces for remeshing
based on moving and deforming loops of faces. For face region remeshing, Fluent marks the
region of faces on the deforming boundaries at the moving boundary based on minimum and
maximum length scales. Once marked, Fluent remeshes the faces and the adjacent cells to
produce a very regular mesh on the deforming boundary at the moving boundary.
3.2.4 Arbitrary LagrangianEulerian calculations (ALE)
The Arbitrary LagrangianEulerian method is a method allowing timedependent CFD. It is
a mix between the Lagrangian and the Eulerian methods.
The Lagrangian method is a timedependent CFD method in which the calculation grid moves
with the local ﬂuid velocity. It has a particular advantage that advective terms eﬀectively
disappear in the ﬂow equations. It is however severely limited by tangling of the grid in
multidimensional problems, especially when vortices occur.
In the Eulerian method, the calculation grid is stationary. It obviously is not hampered by
tangling grids, but is incapable of representing moving domain boundaries.
The ALE method is a mix of both methods in that it allows the grid to move with a velocity
that is independent of the ﬂow solution. The only limitations are the need to have a valid
grid at each time step, and the restriction that the grid topology must be maintained. The
latter means that the only allowed action on the grid is a free movement of the nodes.
To discretize the NavierStokes equations in ALEformulation with a ﬁnitevolume technique,
the velocity of the boundaries of the control volumes u
b
must be known for each time step.
The boundary velocity can be calculated from the position of the nodes at subsequent time
steps separately from the NavierStokes equations. At the present, commercial CFD packages
are available that can handle ALE calculations. In equation (3.8) the NavierStokes equations
are given. Reference is made to [10] for these equations.
geometric conservation law
∂
∂t
Ω
dV
−
∂Ω
u
b
· dS = 0,
conservation of mass
∂ρ
∂t
+
∂
∂x
(ρ(u −u
b
)) = 0,
conservation of momentum
∂ρu
∂t
+
∂
∂x
(ρu(u −u
b
) −τ) +
∂p
∂x
= 0, (3.8)
where ρ is the density, u
b
the velocity vector of the cell boundary, u the ﬂuid velocity vector,
τ the shear tensor, and p the pressure.
For the twin screw pump this ALEmethod has a particular advantage. The positioning of the
nodes only has to be performed once for one screw thread. Rotation of the screws equals axial
displacement of the grid. The last layers can be removed and added to the front to complete
the grid. More screw threads are easily added due to periodicity, this is also applicable for
the initial grid of other meshes.
3.3 Nonconformal mesh 25
3.3 Nonconformal mesh
While the screws are rotating, the inlet and outlet remain stationary. Connecting those
regions could give problems if nodes should always coincide. A nonconformal grid does not
require that nodes coincide on the combined faces. In Fluent it is possible to use a grid
composed of cell zones with nonconformal boundaries. That is, the grid node locations need
not to be identical at the boundaries where two subdomains meet. Fluxes across the grid
interface are computed using the faces resulting from the intersection of the two interfaces,
not from the two interfaces separately.
3.4 Conclusion
From the estimation of the number of cells in the clearance between screw and liner, an
unstructured mesh of the clearances between screw and liner is impossible due to the large
number of needed cells. The total number of cells to describe the twin screw pump has to
be less than a few million cells, to be solvable with normal PC performance and memory.
The clearance between screw and liner can be represented with structured hexahedral cells.
The aspect ratio of the hexahedral cell has to be larger than 10 to make a model of the total
twin screw pump with a few million cells. The inlet and outlet of the pump can be connected
with a nonconformal mesh interface, or special care has to be taken to make the nodes at
the interface coincide. To develop a dynamic mesh, smoothing and remeshing can be used,
even when the topology of the mesh changes in time. ALEcalculations for a dynamic mesh,
without change in topology, are preferred above smoothing and remeshing because of the lower
demands on PC performance. ALEcalculations are performed for twin screw compressor in
[9], so this method is considered feasible for a twin screw pump.
Chapter 4
Construction of structured grid
For a CFD calculation of a twin screw pump, the need for a structured grid is twofold. First
the number of elements needs to be kept within limits and secondly the quality of the grid
needs to be well controlled. To create a CFD simulation that is solvable with normal PC
performance and within a reasonable time, the number of elements has to be smaller than a
few million. The representation of small clearances between screws and the liner can be done
using many unstructured elements, but also with a limited number of well positioned elements,
see chapter 3. Therefore, structured elements are preferred to unstructured elements.
In this chapter a method to divide a twin screw pump in structured hexahedral cells is dis
cussed. This method consists of the layering of twodimensional crosssections, perpendicular
to the axial direction, to form a threedimensional grid. In section 4.1 quadrilateral cells
will be placed on this crosssection, and in section 4.2 some reﬁnements will be applied to
this quadrilateral twodimensional grid. In section 4.3 the quality of the twodimensional
and threedimensional grid will be evaluated. In section 4.4 grid quality improvement by
placing the nodes on another crosssection is evaluated. In section 4.5 conclusions about the
applicability of the grid are discussed.
4.1 Basic structure
The algorithm described here, positions the nodes based on the solution of the Laplace equa
tion in a twodimensional section of the ﬂow domain (see ﬁgure 4.1). The Laplace solution
is obtained on an unstructured grid (ﬁgure 4.2(a)) of a crosssection normal to the axial di
rection with wellchosen boundary conditions. At this stage the boundary conditions for the
potential Φ are: Φ = 0 for the liner, Φ = −1 for the left screw, and Φ = 1 for the right screw.
After solving this problem, equipotential values of this problem are given in ﬁgure 4.2(b). A
valuable characteristic of the solution of the Laplace equation is that the equipotential lines
never cross. Furthermore, the direction of the gradient ∇Φ is perpendicular to the potential
lines. With the present boundary conditions these gradient lines are also perpendicular to the
boundaries. A grid based on equipotential and gradient lines can obtain high quality almost
everywhere, except near the bottom and top cusp. Highly distorted cells are generated near
the cusps since the potential line Φ = 0 does not end in the top and bottom cusp. This is
corrected with small adjustments in section 4.2. To expand this twodimensional grid to three
dimensions, twodimensional grids with small axial distance apart are computed as described
above. These twodimensional grids have an equal number of radial and tangential nodes.
4.2 Reﬁnements 27
Now nodes with equal index are connected to form a threedimensional mesh of hexahedral
cells.
The position of the grid nodes are calculated with the CFDprogram Comsol with Matlab
interface. The nodes are written in a mesh ﬁle suited for calculation with Fluent. Faces are
grouped in inlet, outlet, screw and liner. In [9] and [10] methods to create a structured mesh
are described. More detailed information of the screw is given in appendix B. Furthermore a
theoretical screw proﬁle without clearance is described in appendix A, to gain more insight
in the shape of the screw. In table 4.1 the characteristic dimensions of the screw are given.
T
B
y
x
Figure 4.1: Crosssection of a twin screw pump, indicating top (T) and bottom (B) cusps
(not to scale)
(a) Unstructured mesh of crosssection of a twin
screw pump
Φ = 1
Φ = −1
Φ = 0
(b) Contourlines of potential value, with boundary
conditions constant along each separate wall
Figure 4.2: Unstructured mesh and potential ﬂow solution of crosssection of twin screw pump
4.2 Reﬁnements
4.2.1 Dividing line
The equipotential line Φ = 0 does not give a satisfactory dividing line between the left and
right screw. A smooth dividing line from top to bottom cusp is created by: ﬁrstly changing the
boundary conditions, and secondly, deﬁning a line with a potential dependent on the vertical
coordinate y. The boundary condition on the liner becomes a linearly changing potential
4.2 Reﬁnements 28
Table 4.1: Dimensions of twin screw pump
Screw outer radius R 147.25 mm
Screw inner radius r
i
62.5 mm
Liner radius R
liner
147.5 mm
Center oﬀset h 210 mm
Clearance screw  liner s 0.24 −0.265 mm
Clearance screw  screw 0.29 −0.38 mm
Pitch of the screw r 60 mm
Number of threads n
rev
5
1
3
Radius rounded edges 0.5 mm
from Φ
T
to Φ
B
. The boundary conditions of the screws remain the same. The values of Φ
T
and Φ
B
are chosen iteratively so that the line with constant potential (Φ
T
or Φ
B
) departing
from the cusp edge lies between the two tangents of the liner, represented by the dashed line
in ﬁgure 4.3(a). The dividing line with a varying potential Φ
D
, so that a smooth line from
top to bottom cusp arises (see ﬁgure 4.3(b)), is deﬁned as:
y ≤ α Φ
D
= Φ
B
·
1 + cos(
y·π
α
)
2
β
α < y ≤ 1 −α Φ
D
= 0 (4.1)
y > 1 −α Φ
D
= Φ
T
·
1 + cos(
(1−y)·π
α
)
2
β
where y is the normalized vertical distance, see ﬁgure 4.1. The parameter y changes from zero
in cusp B to one in cusp T. The constants α and β are chosen to create a smooth dividing
line, here α = 0.345 and β = 0.67 are chosen.
4.2.2 Smoothing of nodes on gridline starting on the cusps
The radial gridlines starting on the cusps have an equidistant node placement, and this results
in nonorthogonal cells (see ﬁgure 4.4(a)). It suﬃces to move these nodes to a spline created
from neighboring cells on the same tangential line. This movement gives a grid near the top
cusp as given in ﬁgure 4.4(b).
4.2.3 Nonequidistant node placement
On the liner and the dividing line, nodes are placed equidistantly. From these nodes, gridlines
towards the screws emerge, which are directed along the potential gradients. These gridlines
are then partitioned equidistantly to obtain the nodes of the structured twodimensional
grid. Since the potential does not have a constant value on the liner and the dividing line, the
resulting gridlines are not orthogonal to the wall, nor are the grid cells orthogonal themselves.
However the deviation turns out to be small in practice.
4.2 Reﬁnements 29
T
(a) Potential lines near the top cusp in
a crosssection of a twin screw pump
Φ
T
Φ = 1
Φ = −1
Φ
B
(b) Dividing line between top and bottom cusps, with
linear varying boundary condition on the liner from Φ
T
to Φ
B
(not to scale)
Figure 4.3: Radial line near cusp
(a) No adjustment of radial gridline starting in
cusp
(b) Adjusted radial gridline starting in cusp
Figure 4.4: Gridlines near the top cusp in a crosssection of a twin screw pump
Gridlines, starting at points on the dividing line, towards the left and right screw have diﬀerent
lengths. With equidistant placement of the nodes, a discontinuity in cell volume exists. A
reﬁnement to the equidistant placing of the nodes is applied. The longest gradient line is
divided in elements with a linear grow rate q, see ﬁgure 4.5. To avoid a discontinuity in cell
volume for cells at the transition of dividing line and liner, nodes on gridlines starting at
points on the liner are also placed nonequidistant.
In ﬁgure 4.6 the twodimensional grid for a crosssection is given. The discontinuity in the
area of the cells, near the dividing line, is clearly visible with equidistant node placement.
Nonequidistant node placement reduces this discontinuity signiﬁcantly. The linear grow
rate q is limited, this to prevent the formation of extremely large cells at the screw surface
and discontinuities in cell volume in other parts of the domain than near the dividing line.
With nonequidistant node placement, a threedimensional grid of the twin screw pump is
generated, the surface mesh of the screws is displayed in ﬁgure 4.7.
4.3 Grid evaluation 30
equidistant
nonequidistant
dividing line left screw right screw
a aq
2
aq
Figure 4.5: Equidistant and nonequidistant node placing
(a) Equidistant node placement (b) Nonequidistant node placement
Figure 4.6: Reﬁned quadrilateral grid of a crosssection of a twin screw pump
4.3 Grid evaluation
The twodimensional and the threedimensional grid will be evaluated on skewness, and the
threedimensional grid also on volume ratio of neighboring cells. First the theory of the two
methods is discussed in section 4.3.1 and applied in section 4.3.2.
4.3.1 Theory
A normalized measure of skewness is EquiAngle Skew (Q
EAS
), which is deﬁned for an indi
vidual grid cell as:
Q
EAS
= max
θ
max
−θ
eq
180 −θ
eq
,
θ
eq
−θ
min
θ
eq
¸
(4.2)
where θ
max
and θ
min
are the maximum and minimum angles (in degrees) between the edges
of the cell, and θ
eq
is the characteristic angle corresponding to an equilateral cell of similar
form. For triangular and tetrahedral cells, θ
eq
= 60. For quadrilateral and hexahedral cells,
θ
eq
= 90. It can be applied to twodimensional as well as threedimensional grids.
By deﬁnition,
0 ≤ Q
EAS
≤ 1 (4.3)
where Q
EAS
= 0 describes an equilateral cell, and Q
EAS
= 1 a degenerated cell. Table 4.2
outlines the overall relationship between Q
EAS
and cell quality.
4.3 Grid evaluation 31
Figure 4.7: Threedimensional structured screw surface mesh for one thread of a twin screw
pump (color indicating axial coordinate)
Table 4.2: Q
EAS
vs. Mesh Quality
Q
EAS
Quality
Q
EAS
= 0 Equilateral (Perfect)
0 < Q
EAS
≤ 0.25 Excellent
0.25 < Q
EAS
≤ 0.5 Good
0.5 < Q
EAS
≤ 0.75 Fair
0.75 < Q
EAS
≤ 0.9 Poor
0.9 < Q
EAS
≤ 1.0 Very Poor
Q
EAS
= 1 Degenerate
In general, highquality meshes have an average Q
EAS
value of 0.1 for twodimensional grids,
and 0.4 for threedimensional grids.
Volume ratio VR only applies to threedimensional cells. A common method to calculate the
volume of hexahedral cells in numerical simulations is to estimate the volume of a tetrakis
hexahedron. The volume of a tetrakis hexahedron is given by equation (4.4). A tetrakis
hexahedron are 24 tetrahedrals that form a hexahedral. The numbering of the nodes is given
in ﬁgure 4.8. For derivation of this equation reference is made to [3].
V =
1
12
det [(
−→
x
7
−
−→
x
1
) + (
−→
x
6
−
−→
x
0
), (
−→
x
7
−
−→
x
2
), (
−→
x
3
−
−→
x
0
)] +
1
12
det [(
−→
x
6
−
−→
x
0
), (
−→
x
7
−
−→
x
2
) + (
−→
x
5
−
−→
x
0
), (
−→
x
7
−
−→
x
4
)] +
1
12
det [(
−→
x
7
−
−→
x
1
), (
−→
x
5
−
−→
x
0
), (
−→
x
7
−
−→
x
4
) + (
−→
x
3
−
−→
x
0
)] (4.4)
4.3 Grid evaluation 32
Figure 4.8: Tetrakis hexahedron with facecentered vertices (subdivision of front and back
faces not drawn)
The volume ratio VR is deﬁned as the maximum ratio of the volume of an cell with the volume
of adjacent cells normalized between 0 and 1, according to:
VR =
min
i
min
V
i
V
,
V
V
i
(4.5)
where i are the neighboring cells. A volume ratio of 1 represents cells with equal volume.
4.3.2 Evaluation
In ﬁgure 4.9 the values for Q
EAS
are given for a twodimensional grid of a crosssection and for
one slice of threedimensional cells. Mean value for the skewness in the twodimensional and
threedimensional grids is 0.12 and 0.43 respectively. These mean values for Q
EAS
normally
represent a good quality grid, but from the distributions of Q
EAS
(ﬁgure 4.9(d)) it is clearly
visible that there are a large number of highly skewed cells in the threedimensional grid. So,
the grid is not as good as the average value of Q
EAS
wants us to believe. Highly skewed cells
in the twodimensional grid are on radial lines starting near the cusps. Further reﬁnements
of the twodimensional meshing method can improve the mesh. Near the ﬂanges of the screw
also highly skewed cells are created in the threedimensional mesh. The skewness near the
ﬂanges cannot be improved by changing the grid of a crosssection. The number of skewed
cells near the ﬂanges is much larger than near the cusp, so no further reﬁnements are made to
the twodimensional grid of a crosssection. The eﬀect of elongation in axial direction on the
skewness of the threedimensional grid is evaluated. Elongation of the screw pump in axial
direction improves the cell skewness. In table 4.3 the mean skewness Q
EAS
for elongated
structured threedimensional grids of a twin screw pump is given. In ﬁgure 5.2 the skewness
for a 2.5 and a 5 times elongated grid is given. The overall skewness improves, however some
highly skewed cells remain present near the highly curved part of the screw (red area in ﬁgure
4.9(d)).
The volume ratio VR is low at the dividing line near the cusps of the twin screw pump, see
ﬁgure 4.11. Changing the twodimensional grid of a crosssection can improve the volume
ratio. The grow rate q of cells on radial lines (section 4.2.3) is limited to ensure a high volume
ratio in most part of the mesh, but is responsible for this discontinuity in volume at the
dividing line.
4.4 Threedimensional basis structure 33
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
(a) Skewness of twodimensional grid of a cross
section
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
(b) Skewness of one layer of threedimensional cells
projected on crosssection
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
500
1000
1500
2000
2500
3000
3500
n
u
m
b
e
r
o
f
c
e
l
l
s
skewness Q
EAS
(c) Distribution of skewness for the twodimensional
grid of a crosssection
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
500
1000
1500
n
u
m
b
e
r
o
f
c
e
l
l
s
skewness Q
EAS
(d) Distribution of skewness for one layer of three
dimensional cells
Figure 4.9: Skewness Q
EAS
4.4 Threedimensional basis structure
The quality of the twodimensional grid is not the limiting factor. The expansion to three
dimensions gives skewed cells. This favours the idea of solving the potential problem in three
dimensions, and using the equipotential surfaces and gradient lines as a basis for a three
dimensional grid. Gradient lines depart perpendicular to the wall, so grid skewness should be
improved. To solve the threedimensional potential problem, ﬁrst a threedimensional grid of
the screw is needed. A potential problem is known to pose less stringent demands to the grid
quality. Therefore, the skewed threedimensional structured grid can be used. Secondly, the
improvement of the grid has to be evaluated. To determine the quality of the grid, a cross
section in axial direction of the twin screw pump is made and the potential problem is solved
in two dimensions with the following boundary conditions: liner Φ = 0 and screw Φ = 1. In
ﬁgure 4.12(a) gradient lines are drawn. These lines are not ideal, since many lines accumulate
in the center of the cavity. By changing the boundary conditions this can be prevented.
Dividing the gradient lines equidistantly gives quadrilateral cells, see ﬁgure 4.12(b). In ﬁgure
4.12(c) the distribution of skewness Q
EAS
is given for this grid. The skewness for this two
4.5 Conclusion 34
Table 4.3: Mean skewness Q
EAS
for elongated grids in axial direction
Elongation e mean skewness Q
EAS
1 0.43
2 0.35
2.5 0.33
5 0.26
10 0.21
dimensional section in axial direction is similar to the threedimensional structured grid of the
twin screw pump already described. Placing nodes on basis of the threedimensional potential
problem therefore does not improve the quality of the mesh much, and is not employed.
4.5 Conclusion
A method to describe a twin screw pump with structured hexahedral cells is developed. Two
dimensional structured grids of crosssections perpendicular to the axial direction are layered
to a structured threedimensional grid. The twodimensional grid on a crosssection is placed
along gradient lines of the Laplace problem. The Laplace problem is solved on an unstructured
grid of the crosssection with suitable Dirichlet boundary conditions. Some adjustments are
made to improve the quality of the twodimensional grid. All twodimensional grids have the
same topology, this is required to combine all twodimensional grids to a threedimensional
grid. The quality of the threedimensional grid, in terms of skewness, deteriorates due to the
high surface curvature of the geometry, compared to the quality of the twodimensional grid.
Placing the gridlines along the equipotential surfaces and gradient lines of a threedimensional
Laplace problem gives a skewness similar to the threedimensional grid created with the
layering method, so this is no improvement of the grid.
4.5 Conclusion 35
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
(a) Skewness for one layer of threedimensional cells
of a 2.5 times elongated grid in axial direction
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
(b) Skewness for one layer of threedimensional cells
of a 5 times elongated grid in axial direction
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
500
1000
1500
n
u
m
b
e
r
o
f
c
e
l
l
s
skewness Q
EAS
(c) Distribution of skewness for one layer of three
dimensional cells of a 2.5 times elongated grid in axial
direction
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
200
400
600
800
1000
1200
1400
1600
1800
n
u
m
b
e
r
o
f
c
e
l
l
s
skewness Q
EAS
(d) Distribution of skewness for one layer of three
dimensional cells of a 5 times elongated grid in axial
direction
Figure 4.10: Skewness Q
EAS
of an elongated twin screw pump in axial direction
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
(a) Volume ratio VR of one layer projected on cross
section
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0
500
1000
1500
2000
2500
3000
3500
4000
n
u
m
b
e
r
o
f
c
e
l
l
s
Volume Ratio VR
(b) Distribution of volume ratio
Figure 4.11: Volume ratio VR
4.5 Conclusion 36
(a) Gradient lines of
Laplace problem with
boundary conditions
constant along each wall
(b) Structured quadrilat
eral grid for an axial cross
section of a screw
(c) Distribution of skewness for an ax
ial crosssection of a screw
Figure 4.12: Crosssection in axial direction of one screw thread
Chapter 5
CFD computations with structured
grid
5.1 Numerical setup
With the structured grid generated in chapter 4, threedimensional ﬂow calculations are
performed. The grid consists of one screw thread of the twin screw pump and each screw
has 10 radial cells, 300 tangential cells, and 180 axial cells (see ﬁgure 4.7). Thus the total
number of hexahedral cells is approximately 1 million. The leakage ﬂow rate through periodic
stationary screws is calculated for a ﬁxed pressure drop over the screw thread. The ﬂow is
simulated with the steady segregated solver and implicit formulation, turbulence is modeled
with the standard k − ω model with standard wall function. For the initial condition the
velocity and the turbulent kinetic energy are set to zero in the complete domain, and the
speciﬁc dissipation rate of the turbulence to one. Water is taken as the leakage medium. Also
calculations on an elongated twin screw pump are performed to evaluate the eﬀect of screw
geometrie and skewness of the grid.
5.2 Results
There are three cases tested. First, the leakage ﬂow through a twin screw pump is simulated
at low axial Reynolds numbers. The axial Reynolds number varies from 19 to 381 in the
clearances between screw and liner. Secondly, the leakage ﬂow is simulated for an elongated
twin screw pump in axial direction at approximately equal Reynolds number in the clearance
between screw and liner. Thirdly, for a 5 times elongated grid of a twin screw pump the
leakage ﬂow is simulated for axial Reynolds numbers up to 1514.
The total leakage ﬂow rate is compared with the leakage ﬂow rate through the clearances
between screw and liner. In ﬁgure 5.1(a) the clearance between the screw and liner is given
by region 2. Also the mean axial velocity through the clearance between screw and liner is
compared with the analytical equation for laminar ﬂow in an annulus with and without inlet
and outlet resistances, given in equation (2.4).
The low quality of the grid results in non converging solutions with the standard parameters.
By lowering the underrelaxation factors and taking a low diﬀerential pressure per screw
thread, a converged leakage ﬂow rate is calculated. A underrelaxation factor reduces the
change of the variable for the next iteration.
5.2 Results 38
1
2
B
B
A
A
(a) Regions of leakage: 1 between screws and 2 between screw and
liner. Crosssection AA given in ﬁgure 5.2(a) (not to scale)
(b) Crosssection B
B
Figure 5.1: Crosssections of twin screw pump
In the ﬁrst case of the threedimensional ﬂow simulations the pressure drop is set from 300
to 9600 Pa over one screw thread. For water the diﬀerential pressure should be approx
imately 130 kPa over one screw thread for realistic ﬂow and a axial Reynolds number of
2230. At a higher diﬀerential pressure than 9600 Pa over one screw thread the solution is
nonconverging, even with low underrelaxation factors. The used pressure diﬀerence per
seal is low, resulting in a low axial leakage velocity, and a low dimensionless wall distance.
The eﬀect of the dimensionless wall distance on the turbulence model is discussed in chap
ter 2 and should be greater that 11 for wall functions. The dimensionless wall distance in
the clearance is lower than the required value, it varies from 0.5 to 3 for the diﬀerent axial
Reynolds numbers. No accurate solution of the numerical simulation can be expected, but
some tendencies are visible. In table 5.1 the simulated leakage ﬂow rate for a stationary
twin screw pump is given for diﬀerent diﬀerential pressures over the screw. The leakage
percentage of ﬂuid through the clearance between the screw and the liner compared to the
total leakage ﬂow rate ˙ m
2
/ ˙ m, is higher at a higher diﬀerential pressures over the screw. The
simulated mean axial velocity in the clearance between screw and liner is compared to the
ﬂow rate in an annulus from equation (2.4) with and without in and outlet resistances. For a
nonrotating stationary annulus, the resistance coeﬃcient is λ = 48Re
−1
[11] for laminar ﬂow.
In the second case the grid is elongated, the elongation of the grid changes the dimensions
of the two leakage regions (ﬁgure 5.1(a)). First region 1, the clearance between the screws
gets larger and the ﬂow resistance reduces. Elongation in axial direction reduces the ﬂow
resistance from region R to S in ﬁgure 5.1(b). Secondly region 2, the length of the clearance
between screw and liner gets larger and thus becomes a larger ﬂow resistance. Elongation in
axial direction increases the length of the clearance between region O and P in ﬁgure 5.2(a).
In table 5.2 the leakage ﬂow rate is given for elongated screws with a laminar leakage ﬂow
rate in the clearances between screw and liner.
In the third case a 5 times elongated grid in axial direction is used and the diﬀerential
pressure over the screw is varied. The axial Reynolds number in the clearance between
5.3 Conclusion 39
(a) Crosssection AA in axial direction of two threads (black: screw/liner, white:
ﬂuid)
A
B
ri R
(b) Crosssection
CC (not to
scale)
Figure 5.2: Crosssections of twin screw pump
screw and liner varies from 20 to 1500. There is no converged solution found for higher
axial Reynolds numbers. In table 5.3 the simulated leakage ﬂow rate through the twin screw
pump is given. The mean axial velocity in the clearance v
ax
is given for the simulation and
calculated with equation (2.4), with a value of 1.5 for in and outlet resistance δ
io
. In [11], the
resistance coeﬃcient for a stationary annulus is given as: λ = 48Re
−1
for laminar ﬂow, and
λ = 0.26Re
−0.24
for turbulent ﬂow. The mean axial velocity of the numerical simulation and
equation (2.4) are similar for laminar ﬂow up to a Reynolds number of approximately 500.
At higher axial Reynolds numbers, the simulated and calculated mean axial velocity deviate
from each other. The dimensionless wall distance for an axial Reynolds number of 1514 on a
5 times elongated grid is given in ﬁgure 5.3. The value of the dimensionless wall distance y
+
in the clearance and in the cavity are approximately 8 and 1000 respectively.
Table 5.1: Water leakage ﬂow rate for threedimensional stationary twin screw pump simu
lations with various pressure diﬀerences per screw thread, v
ax
in clearance between screw
and liner
v
ax
[m/s]
∆p [Pa] ˙ m [kg/s] eq. (2.4) δ
io
= 0 eq. (2.4) δ
io
= 1.5 numeric Re ˙ m
2
/ ˙ m [%]
300 0.7 0.072 0.072 0.074 19 4.0
600 0.9 0.144 0.142 0.145 37 5.7
1200 1.4 0.29 0.28 0.277 70 7.0
2400 2.1 0.58 0.54 0.51 129 8.6
4800 3.2 1.15 1.0 0.90 227 10.8
9600 4.8 2.3 1.8 1.51 381 11.1
5.3 Conclusion
The quality of the used grid is low, lowering underrelaxation factors only results in converged
solutions for a low diﬀerential pressure over the twin screw pump. This low diﬀerential pres
sure results in a low leakage velocity and the dimensionless wall distance dependent on this
velocity is too low to calculate an accurate solution with the k − ω turbulence model with
wall functions. From the simulations some conclusions can be draw. First: elongation of the
5.3 Conclusion 40
Table 5.2: Water leakage ﬂow rate for threedimensional stationary twin screw pump simula
tions with 5000 Pa/m pressure diﬀerence for elongated grid in axial direction (Re ≈ 19)
e ˙ m [kg/s] v
ax
[m/s] ˙ m
2
/ ˙ m [%]
1 0.7 0.0740 4.0
2 1.6 0.0762 1.7
2.5 2.1 0.0762 1.3
5 4.7 0.0745 0.6
10 8.0 0.0731 0.3
Table 5.3: Water leakage ﬂow rate for threedimensional stationary twin screw pump simu
lations for a 5 times elongated grid in axial direction, v
ax
according to equation (2.4) with
in and outlet resistance for laminar and turbulent deﬁnition of λ
v
ax
[m/s]
∆p [kPa] ˙ m [kg/s] laminar turbulent numeric Re ˙ m
2
/ ˙ m [%]
1.5 4.7 0.072 0.38 0.0745 19 0.6
3 6.6 0.14 0.57 0.15 38 0.8
6 9.4 0.29 0.84 0.29 73 1.1
12 12.9 0.57 1.2 0.57 144 1.6
24 18.3 1.1 1.8 1.08 273 2.1
48 26.2 2.2 2.7 1.9 482 2.6
96 37.6 4.1 4.0 2.9 740 2.8
192 52.4 7.3 5.8 4.0 1005 2.7
288 62.5 10.2 7.3 5.0 1262 2.8
384 75 12.7 8.6 6.0 1514 2.8
screws in axial direction results in a larger leakage ﬂow for the same diﬀerential pressure over
the screw thread. The leakage ﬂow rate in the clearance between screw and liner reduces,
however the leakage ﬂow rate between the screws increases, even more than the decrease in
the other region. Secondly: increasing the diﬀerential pressure per screw thread changes the
leakage ﬂow ratio between the two regions.
The leakage ﬂow rate through the clearances between the screw and liner has a larger portion
of the total leakage ﬂow rate at a higher diﬀerential pressures for laminar ﬂow. This can
be expected from analytical equations. The relation between diﬀerential pressure and axial
velocity is given in equation (2.4) for ﬂow in an annulus. The diﬀerential pressure is approx
imately linearly proportional to the axial leakage velocity for laminar ﬂow. For an expansion
in the diameter of a pipe the relation between diﬀerential pressure and axial velocity is given
in [8] and is ∆p = K
1
2
ρv
2
, where K is a constant for turbulent ﬂow. In the numerical simu
lations the value of K also remains approximately constant, even for laminar ﬂow. In a twin
screw pump the leakage ﬂow between the screw and liner is similar to ﬂow in an annulus,
however the leakage ﬂow between the screws is similar to a expansion in the diameter of a
pipe. This expansion between the screws is most clearly visible in the crosssection given in
5.3 Conclusion 41
Z
Y
X
Position (m)
Yplus
Wall
0.3 0.25 0.2 0.15 0.1 0.05 0
1.00e+04
1.00e+03
1.00e+02
1.00e+01
1.00e+00
clearance
cavity
screw_cl
screw_ca
liner_y+
Figure 5.3: Dimensionless wall distance for e = 5 and pressure diﬀerence of 384 kPa over one
thread for a twin screw pump (Re in the clearance is 1514)
ﬁgure 5.2(b) between region A and B, and also in ﬁgure 5.1(b) between region R and S.
At higher axial Reynolds numbers the ratio between leakage ﬂow rate through the clearances
and the total leakage ﬂow rate is found constant in the numerical simulations. This indicates
a constant resistance coeﬃcient λ for axial Reynolds numbers above a certain number or a
negligible eﬀect of sealing length compared to in and outlet resistances.
Chapter 6
Conclusions and recommendations
6.1 Final conclusion
To model the ﬂow in a twin screw pump a grid of the screw pump has to be made. This grid
has to describe the geometrie of the pump well. Also the turbulence model poses requirements
on the dimensionless wall distance of the ﬁrst cell at solid walls. To obtain a solution for the
ﬂow in a twin screw pump with normal PC performance and in reasonable time, the number
of cells must be limited. A grid of the twin screw pump is made and the leakage ﬂow through
the twin screw pump is simulated for a low diﬀerential pressure over the pump.
The demands of the turbulence model are evaluated by comparison with the results of two
experiments. First, the ﬂow through an annulus with rotating inner cylinder is modeled. Near
wall modeling and wall functions are considered for the diﬀerent variants of the k−ǫ and k−ω
turbulence model. Secondly, the ﬂow through a stationary straightthrough labyrinth seal is
simulated. Here only wall functions are considered for the diﬀerent turbulence models. The
standard k − ω model gives the most accurate results when simulating with wall functions.
Wall functions are needed to keep the total number of cells limited. The dimensionless wall
distance y
+
has to be larger than 11 to give accurate results with wall functions.
A mesh method is constructed that generates a structured hexahedral grid by layering of
crosssections perpendicular to the axial direction. The quality of the twodimensional grid
of a crosssection is good, however the threedimensional grid has poor quality. The three
dimensional cells have a high skewness in some regions.
Between the screw and the liner structured cells with elongated sides in axial and tangential
direction have to be used to keep the total number of cells limited. Numerical simulations
proved that this elongation has no inﬂuence on the ﬂow through an annulus with inner rotating
cylinder. Gradients of the solution in axial and tangential direction are low compared to the
radial direction.
From the simulations some conclusions can be drawn. First: elongation of the screws in axial
direction results in a large leakage ﬂow for the same diﬀerential pressure per screw thread.
The leakage ﬂow rate in the clearance between screw and liner reduces, however the leak
age ﬂow rate between the screws increases, even more than the decrease in the other region.
Secondly: increasing the diﬀerential pressure per screw thread changes the leakage ﬂow ra
tio between the two regions for the laminar ﬂow regime. The leakage ﬂow rate through the
clearances between the screw and liner has a larger portion of the total leakage ﬂow rate
6.2 Recommendations 43
at higher diﬀerential pressures. Besides the numerical result this is also expected from the
analytical equations for the relation between diﬀerential pressure and axial ﬂuid velocity in
the two regions. At higher axial Reynolds numbers the leakage ﬂow ratio between the two
regions is approximately constant.
The static twin screw pump simulation shows realistic ﬂow features at a low diﬀerential pres
sure over the pump. Expanding the simulation to a diﬀerential pressure present in industrial
screw pumps gives a better understanding of the ﬂow in a twin screw pump. The static sim
ulations can be extended to a dynamic simulation with multiphase ﬂow to obtain a complete
model of the twin screw pump.
6.2 Recommendations
• The constructed threedimensional grid has highly skewed cells, the grid quality has to
be improved to allow accurate CFD simulations. On the existing grid small improve
ments can be made but it remains questionable if the quality will improve enough to
perform accurate CFD simulations. There is another method of meshing the twin screw
pump, while fulﬁlling the demands set in this report. This method consists of a hybrid
grid, meaning a mixed grid of hexahedral and tetrahedral cells. The clearance between
the screw and liner can be meshed with a structured hexahedral mesh and the interior
with unstructured tetrahedral cells. The nodes on the interface either have to coincide,
or a nonconformal interface (section 3.3) can be used.
• For comparison of twin screw pump simulations with experiments, measurements of the
ﬂow in the twin screw pump are needed.
• There are some small improvements possible on the construction of the crosssectional
mesh. Like: automatic determination of Φ
T
, Φ
B
, correctly describe the roundoﬀ of the
screw tips, and curvature dependent distribution of nodes at the dividing line. This
mesh can be used, for example, for a twin screw pump with a larger screw pitch.
• The multiphase modeling packages in Fluent should be tested on validation cases, to
determine which model is the most accurate for ﬂow in a twin screw pump.
• To create a timedependent simulation, use can be made of smoothing and remeshing
in Fluent. Smoothing and remeshing can deteriorate the mesh. When the mesh dete
riorates beyond a certain limit, a new mesh can be used and the old solution can be
interpolated on the new grid. This method has proven its applicability on a completely
unstructured mesh of one screw thread of a twin screw pump with large clearances of 6
mm at consulting company Bunova Development BV.
• Validate behaviour of turbulence models for laminar ﬂow in the clearance between screw
and liner. There is laminar ﬂow present when more viscous oil is pumped. A turbulence
model is needed because in a dynamic simulation there are regions with turbulent ﬂow.
• The ﬁrst cells near the wall in a cavity are too large, resulting in a high dimensionless
wall distance. So, adjustment of the grid in the cavities is necessary.
Bibliography
[1] Karassik et al. Pump handbook, pages 3.99–3.121. Mc GrawHill, third edition, 2001.
[2] W. Matek et al. Roloﬀ/Matek Machineonderdelen. Academic service, third edition, 2000.
[3] J. Grandy. Eﬃcient computation of volume of hexahedral cells.
http://www.osti.gov/bridge/servlets/purl/6327934p2OLa/webviewable/ Report
number: UCRLID128886.
[4] Fluent Inc. Fluent 6.2 user’s guide.
[5] B.V.S.S.S. Prasad and V. Sethu Manavalan. Computational and experimental investiga
tions of straight  through labyrinth seals. ”ASME Paper 97GT326”, 1997.
[6] J.J.M. Smits. Modeling of a ﬂuid ﬂow in an internal combustion engine. Graduation
report TU/e, Report number WVT 2006.22.
[7] A. Uasghiri. Een multigrid versnelling van de simple meth
ode voor incompressible stromingen. Graduation report TU Delft
(http://ta.twi.tudelft.nl/nw/users/vuik/numanal/uasghiri.html).
[8] Bart van Esch and Erik van Kemenade. Procestechnische constructies 1  4b660. Lecture
notes, March 2005.
[9] J. Vande Voorde and J. Vierendeels. ALE calculations of ﬂow through rotary positive
displacement machines. ”Proceedings of FEDSM2005”, (FEDSM200577353), 2005.
[10] J. Vande Voorde, J. Vierendeels, and E. Dick. Development of a laplacianbased mesh
generator for ALEcalculations in rotary voluemtric pumps and compressors. ”J. Com
put. Methods Appl. Mech. Engrg.”, (193):4401–4415, 2004.
[11] Y. Yamada. Resistance of ﬂow through an annulus with an inner rotating cylinder. ”Bull.
JSME”, 5(1):302–310, 1962.
Appendix A
Theoretical Screw proﬁle
The screw proﬁle is described with a crosssection normal to the axial direction of the twin
screw pump, see ﬁgure A.1. In this ﬁgure proﬁle A and B are visible.
B
B
A
A
Figure A.1: Crosssection twin screw pump
The proﬁle A is an involute proﬁle (x and y) and described by the following equations:
s = αr (A.1)
x
c
= r cos α (A.2)
y
c
= r sin α (A.3)
x = x
c
+ s sin α (A.4)
y = y
c
−s cos α (A.5)
where r ≤ r
i
, with r
i
, the screw inner diameter. x
c
and y
c
are construction points on the
circle with radius r. In ﬁgure A.1, r is equal to r
i
. The other variables of proﬁle A are deﬁned
in Figure A.2.
The proﬁle B is described in polar coordinates θ and R by equations (A.6:A.10). The variables
of proﬁle B are deﬁned in Figure A.3.
a = r
o
cos α (A.6)
46
α
α
r
s
x
c
, y
c
x, y
Proﬁle A
Figure A.2: Variables for proﬁle A
h = r
o
sin α (A.7)
b = y −a (A.8)
R =
b
2
+ h
2
(A.9)
θ = π + α + tan
−1
(h/b) (A.10)
α
h
a
b
θ
ro
y
R
Figure A.3: Variables for proﬁle B
Appendix B
Screw proﬁle Houttuin
48
Acknowledgements
First of all I would like to thank my supervisor Bart van Esch for his help and guidance
in completing my graduation project. Furthermore I like to thank the consulting company
Bunova Development BV in Zwolle for creating this graduation report. Especially I would
like to thank Timco Visser and Jo`an Teerling for answering practical questions regarding the
numerical model. Finally I would like to thank my roommates and my colleague graduate
students for the pleasant and stimulating working atmosphere.
Maarten van Beijnum
Abstract
Multiphase pumping in the oil and gas industry is the ability to boost pressure without separating the liquid and the gas phases. This gives opportunities to process the diﬀerent phases centrally when using multiple well sites, or processing on land for an oﬀshore well. No separation of the phases and only one pipeline have to be used. For optimal performance of a twin screw pump a small liquid fraction is necessary to seal the internal clearances of a twin screw pump. 100 percent gas void fractions can be pumped for a short period of time when arrangements are made to recirculate ﬂuid, to seal the clearances. This makes the internal design of a twin screw pump an engineering challenge. To gain more insight of the ﬂow in a twin screw pump a CFD model can be used. The goal of this graduation report is to predict the leakage ﬂow rate in a twin screw pump with a threedimensional model of the pump and a commercial CFD package. This goal can be divided in three parts. The leakage ﬂow rate ﬁrst for nonrotating screws and singlephase ﬂow, secondly for rotating screws and singlephase ﬂow, and ﬁnally considering multiphase ﬂow with rotating screws. In this report the ﬁrst part is considered, and recommendations are given for the other parts. The leakage ﬂow rate is simpliﬁed in two diﬀerent cases. The ﬁrst case is ﬂow through an annulus with inner rotating cylinder, this represents the ﬂow between the screw and housing (liner) of a twin screw pump. The second case is ﬂow through a straightthrough labyrinth seal. The screw thread viewed in axial direction is similar to a labyrinth seal. Recirculation and throttling of ﬂuid in the screw cavities can be expected. For these two cases the performance of the turbulence model is evaluated and coupled to requirements for the dimensionless wall distance in the ﬁrst cell near the wall. To simulate the ﬂow in a twin screw pump the ﬂow domain has to be meshed, the mesh must have a limited number of cells to perform calculations with normal PC requirements in a reasonable amount of time. The number of cells in the clearance between the tips of the screw and the liner is estimated. The number of cells with an unstructured tetrahedral mesh is too large to perform CFD simulations. Structured hexahedral cells can be used, however these cells have to be elongated in axial and tangential direction to reduce the number of cells. A structured grid with hexahedral cells is created by layering crosssections perpendicular to the axial direction. The cells on a crosssection are placed along gradient lines of the Laplace problem solved for this crosssection. The Laplace problem is solved, for an unstructured triangular mesh of the crosssection, with a mesh generator and solver of a commercial CFD package. Gradient lines never cross each other, so a robust twodimensional grid is created. For the next crosssection, a small displacement in axial direction, the screws are rotated slightly and the Laplace problem is solved again. Merging the crosssections gives a threedimensional grid i
of the screws of a twin screw pump. The quality of this threedimensional grid is examined. The low screw pitch gives a relatively large tangential displacement compared to the axial displacement, resulting in highly skewed cells. This reduces the applicability of this grid for turbulent ﬂows. Elongating the screws in axial direction (higher pitch) gives a better quality grid, however the original geometry is lost. The leakage ﬂow through a twin screw pump has two paths, ﬁrst, leakage through the clearance between the screw and the liner, and secondly between the screws itself. The leakage ﬂow through the twin screw pump is simulated for a diﬀerential pressure of up to 10 kP a per screw thread (seal) on the threedimensional grid. For higher diﬀerential pressures the simulation does non converge. The simulated leakage ﬂow rate in the clearance between screw and liner is approximately the same as the analytic laminar leakage ﬂow rate through a stationary annulus. The relation between diﬀerential pressure and leakage ﬂow rate is determined for low axial Reynolds numbers, and for higher axial Reynolds numbers using an elongated grid in axial direction. The static numerical simulation of the ﬂow in a twin screw pump show realistic ﬂow features. The diﬀerential pressure per screw thread has to be increased to simulate real pump performance. The numerical model is created with the ability to add a dynamic mesh, this to simulate the rotation of the screws. Also multiphase models can be added to predict leakage ﬂow characteristics with liquidgas mixtures. For these extensions to the current model recommendations are given.
ii
Nomenclature
Symbol a Acl Aliner Ascrew D dh e fAR h h1 GVF L L m ˙ Ma N nax ncell ntan nrev nstr nunstr R ri Rliner Re Reω s p ∆p q r Qc l Description edge length of cell projected area of clearance between screw and liner perpendicular to the axial direction area of two joined circles area of screw crosssection outer diameter screw hydraulic diameter elongation of grid aspect ratio of cell edges center distance screws height of screw cavity gas void fraction length screw thickness mass ﬂow rate Mach number rotation speed number of axial cells number of cells in radial direction number of tangential cells number of screw threads (revolutions) number of structured cells number of unstructured cells outer radius of screw inner radius of screw radius of liner axial Reynolds number tangential Reynolds number clearance between screw and liner absolute pressure pressure diﬀerence grow rate pitch of one screw thread leakage ﬂow rate in clearance between screw and liner iii [Unit] [m] [m2 ] [m2 ] [m2 ] [m] [m] [] [] [m] [m] [] [m] [m] [kg/s] [] [rpm] [] [] [] [] [] [] [m] [m] [m] [] [] [m] [P a] [P a] [] [m/rev] [m3 /s]
Symbol Ql Qr Qt uτ v vax Vcl VD Vhex Vtet y+
Description leakage ﬂow rate of pump realized ﬂow rate of pump theoretical ﬂow rate of pump friction velocity ﬂuid velocity mean axial velocity volume of clearance between screw and liner screw displacement volume per revolution volume of hexahedral cell volume of tetrahedral cell dimensionless wall distance
[Unit] [m3 /s] [m3 /s] [m3 /s] [m/s] [m/s] [m/s] [m3 ] [m3 /rev] [m3 ] [m3 ] []
Greek symbols
Symbol θ λ µ ν τ τw Φ ρ ω Description angle resistance coeﬃcient dynamic viscosity kinematic viscosity shear tensor wall shear stress potential density angular velocity of the screw [Unit] [rad] [] [kg/(ms)] [m2 /s] [N/m2 ] [N/m2 ] [] [kg/m3 ] [rad/s]
iv
. . . . . .1 Reynolds number in twin screw pump 2. . . . . .3 Nonequidistant node placement . . . . . .2 Dynamic layering method . . . . . .2 Construction of screw pumps . . . .4 Arbitrary LagrangianEulerian calculations (ALE) 3. . . . . . . . .1 Basic structure . . . . . . . . . .Contents Abstract Nomenclature 1 Introduction 1. . . . .3 Nonconformal mesh . . . . . . . . . . . . . . . . . . .3 Remeshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. . 4 Construction of structured grid 4. . . . . . . . . . . . . pump . . . . . . . . . . . . . . . . . i iii 1 2 3 5 6 7 7 9 10 10 15 15 16 17 19 20 20 22 22 23 23 24 25 25 26 26 27 27 28 28 . . . . . . . . . . .2 Smoothing of nodes on gridline starting on the cusps 4. . . . . 3 Method of threedimensional meshing 3. . . . . . . . . . . . . . .2. . . . . . . . . . . .4 Goal and outline . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . .2 Annulus with rotating inner cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . v . . . . . . .1 Experimental setup . . . . . . . . . . . . .1 Dividing line . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . .1 Number of cells in the clearance region . . . . 3. . . . . . . . . . . . . . . . . . . 2. . . . . . . . .2. . . . . . . . . . . . . . . 2. . . . 3. . . . . . . . . . . . .1 Theory of screw pumps . . .2. . . . . 3. . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . . . . . . . . . . . . . .2 Numerical simulation . . . . . . . . . . . . . . . . . . . . . . . 4. . . . 2 Simpliﬁed leakage ﬂow in a twin screw 2. . . . . . . . . . . . 2. . 3. . . . . . . . . . . . . . . . . . .2 Reﬁnements . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Dynamic mesh . . . . . . . 4. . . . . . . . . . . . . . . .4 Conclusion . . . . 1. . . . .2. . . . . . . . .5 Conclusion . . . . . .3 Labyrinth seal . . . . . .3 Special multiphase applications 1. . . . . . . . . . 2. . . . . . . .2 Numerical simulation . .3. . . . 3. . . . . . 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2.1 Smoothing . . . . . 2. . . . . . . . .1 Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . .4 Leakage ﬂow in twin screw pump . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . 30 30 32 33 34 37 37 37 39 42 42 43 44 45 47 49 5 CFD computations with 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . Bibliography A Theoretical Screw proﬁle B Screw proﬁle Houttuin Acknowledgements vi . . . . . . . . . . . . . . . . 5. . . . . . . . . . . . . . . grid . . . . . . . . . . . . . 6. . . . .2 Results . . . . . . . . . . . 4. . . . . . . . . . . . 5. . . . . . . . . . . . .3 4. . . . . . . . . . .1 Final conclusion . . . . . . . . 4. . . . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Recommendations . . . . . . . . . . . .2 Evaluation . . . . . .3. structured . . . . . . . . . . . . . . Threedimensional basis structure Conclusion . . . . . . . . . . . . . . . .4 4. . . . . . . . . .5 Grid evaluation . . . . .1 Numerical setup .3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Theory . . . . . . . . . . . . . . . . . . . . . 6 Conclusions and recommendations 6. . . . . . . . . .
petroleum and crude oil industries. thus giving the screw pump its unique axial ﬂow pattern and low internal velocities a number of advantages in many applications where liquid agitation or churning is not desired. In this report only crude oil transportation is considered . and utilities fuel oil services. industrial oil burners. The applications of screw pumps cover a diversiﬁed range of markets including navy. lubricating oil services. marine. chemical processes. which is essentially independent of pressure.1). In all other rotary pumps the liquid is forced circumferentially. Twin screw pumps are available in two conﬁgurations: single end and double end (see ﬁgure 1. showing (a) axial and (b) circumferential ﬂow . and many others. Figure 1. screw pumps are capable of operating at higher speeds than other rotary or reciprocating pumps of comparable displacement. marine cargo. Because of relatively low inertia of their rotating parts. In this report only twin screw pumps are considered.1: Diagrams of screw and gear elements. The screw pump can handle liquids in a range of viscosities. from molasses to gasoline. The liquid is carried between screw threads on one or more screws and is displaced axially as the screws rotate and intermesh (see ﬁgure 1. Another property for a twin screw pump.5 to 350 bar and ﬂows up to 1820 m3 /h. Reference is made to [1] with respect to this chapter. as well as synthetic liquids in a pressure range from 3. as opposed to centrifugal pumps is the capability of handling mixtures of liquid and vapour.2). power hydraulics for navy and machine tools.Chapter 1 Introduction Screw pumps are a special type of rotary positive displacement pump in which the ﬂow through the pumping element is axial. provided there is suﬃcient viscosity in the liquid being pumped. Screw pumps. are selfpriming and have a delivery ﬂow characteristic. like other rotary positive displacement pumps.
the rotation speed of the screws. Clearances.1 Theory of screw pumps 2 Figure 1. the liquid viscosity. it is the intermeshing of the threads on the screws and the close ﬁt of the surrounding housing (liner) that create one or more sets of moving seals in a series between the pump inlet and outlet. If the diﬀerential pressure is almost zero. This leakage Ql varies depending upon the pump type or model.3(b) for the screw dimension parameters. which is a function of displacement volume and speed N (rpm): Qt = 1 VD N 60 (1. the leakage ﬂow rate may be neglected and Qr = Qt . it will deliver a deﬁnite quantity of liquid every revolution of the screws. providing a smooth ﬂow. This delivery can also be deﬁned in terms of theoretical capacity or ﬂow rate Qt measured in cubic meters per second. and the diﬀerential pressure. the pump’s actual delivered or net ﬂow rate Qr would equal the theoretical ﬂow rate. see ﬁgure 1.2: Twin screw pump with doubleend arrangement and internal timing gears 1. For any particular thread . see ﬁgure 1. there will always be internal leakage from outlet to inlet. The theoretical ﬂow rate of any pump can readily be calculated if all essential dimensions are known.3(a). The successive sets of seals form fully enclosed cavities that move continuously from inlet to outlet. These sets of seals act as a labyrinth and provide the screw with its positive pressure capability. This delivery can be deﬁned in terms of displacement volume VD . do exist with the result that whenever a pressure diﬀerential occurs.1. however. The delivery ﬂow rate or net ﬂow rate is the theoretical ﬂow rate minus the leakage ﬂow rate: Qr = Qt − Ql .1) If no internal clearances existed. The theoretical ﬂow rate is not dependent on the diﬀerential pressure over a positive displacement pump. the geometry of the clearance. Because the screw pump is a positive displacement device.1 Theory of screw pumps In screw pumps. which is the theoretical volume displaced per revolution of the screws and is dependent only upon the physical dimensions of the screws.
These test data are the basis of the design parameters used by every pump manufacturer. The thread pitch r measured in terms of the same nominal diameter.3) (1.4) (1. the net ﬂow rate Qr can be written as: Qr = kD3 N − k2 ∆pD2 with k and k2 empirical constants depending on the geometry and the working ﬂuid. When neglecting the eﬀect of rotation of the screws on the leakage ﬂow rate. This is also true for centrifugal pumps which scale according to Φ = NQ 3 = constant. Therefore.3: Twin screw pump conﬁguration. it can be seen that a relatively small increase in pump size can provide a large increase in ﬂow rate. the volume ﬂow rate of each cavity is proportional to the cube of this nominal diameter and the speed of rotation N (rpm): Qt = kD3 N or writing it in terms of pitch. Thus. this is in contrast to positive displacement pumps like screw pumps. deﬁnes the length. but it is usually estimated based on empirical values obtained from extensive testing. for a given geometry. (1. The mechanical and hydraulic technique .2 Construction of screw pumps Design concepts The pressure gradient in the pump elements of all the types of screw pumps produces various hydraulic reaction forces. which is used in calculating the crosssectional area. the size of each cavity mentioned earlier is proportional to its length and crosssectional area. Qt = k1 · r · D2 N where r = kD/k1 with k and k1 being constants. The leakage ﬂow rate can also be calculated.1. assuming geometric similarity.2 Construction of screw pumps 3 s r L N = constant Q Qt D Qr Ql ∆p (a) Theoretic and realized volume ﬂow rate against diﬀerential pressure for positive displacement pumps (b) Parameters Figure 1. The theoretical ﬂow rate of centrifugal pumps is inﬂuenced D by the diﬀerential pressure over the pump. The leakage ﬂow rate generally varies approximately with the square of the nominal diameter and linearly with the pressure diﬀerence.2) 1.
This version is generally restricted to the handling of clean lubricating ﬂuids. or meshing. In this concept. both lubricating and nonlubricating. The gears and screws are timed at the factory to maintain the proper clearance between the screws.2 Construction of screw pumps 4 employed for absorbing these reaction forces are one of the diﬀerences in the types of screw pumps produced by various manufacturers. The timing gears and bearings are oilbathlubricated from an external source. The two pump elements are. the ﬂuid enters a common inlet with a split ﬂow going to the outboard ends of the two pumping elements and is discharged from the middle or center of the pumping elements. Untimed screw pumps do not require gears or external bearings and thus are considerably simpler in design. and. singleend pumps or pump elements of the same size with a common driving screw that has an opposed. pumps connected in parallel. the threads do no come into contact with each other nor with the housing bores in which they rotate. the screws and maintaining the running clearances between them. all axial loads on the screws are balanced. with proper materials. doublehelix design with one casing. as opposed to the internal type. thus promoting long pump life. Two basic design approaches are used: • The timed screws approach is based on an external means for phasing the mesh of the treads and for supporting the forces acting on the screws. Doubleend screw pumps The doubleend arrangement is basically two opposed. This type can handle a complete range of ﬂuids. The main advantage of the timed screw pump is that the timing gears transmit power to the screws with no contact between the screw threads. • The untimed screws approach is based on the precision and accuracy of the screw forms for the proper mesh and transmission of rotation. This arrangement requires the use of four stuﬃng boxes or mechanical seals.1.2. has good abrasion resistance. Doubleend pumps are generally employed where large ﬂows are required or where highly viscous liquids are handled. which employs only one shaft seal. Another fundamental diﬀerence lies in the method of engaging. herringbone. Timed screw pumps require separate timing gears between the screws and separate support bearings at each end to absorb the reaction forces and maintain the proper clearances.and medium pressure applications. It has both the timing gears and screw support bearings located outside the pumping chamber. The doubleend screw pumps construction is usually limited to low. As can be seen from ﬁgure 1. The external timing arrangement is the most popular and is extensively used. as the pressure gradients in each end are equal and opposite. The internal version has both the gears and the bearings located in the pumping chamber and the design is relatively simple and compact. Timed design Timed screw pumps having timing gears and screw support bearings are available in two general arrangements: internal and external. which serve as the only lubrication for the timing gears and bearings. They utilize the housing bores as journal bearings supporting the pumping reactions along the entire length of the screws. hardenedsteel gears with tooth . In either of these arrangements. theoretically. For lowpressure applications. the design can pump backwards when the direction of screw rotation is reversed. with 28 bar being a good practical limit to be used for planning purposes. in eﬀect. The timing gears can be either spur or helical.
separators. and nitralloy. and bronze. while the trust bearings. are either doublerow. can save signiﬁcant equipment costs as well as operating costs. they can also be supplied in a variety of materials. quiet. In the past. Since the screws are not generally in metallic contact with the housing or with one another. Leakage ﬂow compresses the gas in the upstream screw cavity. While the screw cavities move downstream. including cast iron. as the void fraction increases. and dual pipelines are required to handle the gas phase. a separate liquid ﬂush can be provided or a separator type of body pump can be used. the screw bores of the housing can be lined with industrial hard chrome for abrasion resistance. including cast iron. When pumping multiphase products with high gas void fractions. it was common for the gas to be separated and ﬂared oﬀ at the well head with only the liquid product to be retained for further processing. the diﬀerential pressure over a clearance is also smaller. or ceramic. the well output can vary from 100 percent liquid to 100 percent gas. Depending on a number of factors. The key to pumping multiphase products is to ensure that some liquid is always available to seal the screw clearances and reduce the leakage ﬂow rate. Antifrictional radial bearings are usually of the heavyduty roller type. If the gas is to be processed as well. When pumping liquids. the leakage ﬂow rate decreases. This type of body includes a special chamber that can separate some liquid from the multiphase mixture being pumped. heattreated alloy steel. This liquid can be recirculated back to the screws and mechanical seals to provide sealing and cooling liquid at times when the product is almost all gas. in multiphase applications. . 1. the leakage ﬂow rate through the internal clearances is proportional to the diﬀerential pressure and inversely proportional to the viscosity (assuming laminar ﬂow through the clearances). the pump must be designed with a small pitch to provide a suﬃcient number of seals. and a smaller leakage ﬂow progresses to the next upstream cavity. containing more than just nominal amount of gases. Monel. the volume of liquid required to seal and cool the screws can be three to six percent of the total inlet volume ﬂow rate.1. When the leakage ﬂow rate is smaller. chrome oxide. See ﬁgure 1. positive drive of the screws. but recent process changes in oilﬁeld technologies have created requirements for pumping multiphase ﬂuids.3 Special multiphase applications Screw pumps have been used with gasentrained application for many years. gas. This is clearly visible in ﬁgure 1. and sand. ballthrust or sphericalroller types. The housing can be supplied in a variety of materials. tungsten carbide. However. In addition.4(a) for the typical pump performance when pumping multiphase mixtures. the gas void fraction reduces by the increased pressure. which position the screws axially. Even a small amount of recirculated liquid is suﬃcient to provide this seal and enable the screw pump to operate with GVFs approaching 100 percent. stainless steel. In many oil well applications. compressors. the liquid oil ﬂow eventually degenerates into all sorts of diﬃcult multiphase mixtures of oil. ductile iron. stainless steel. cast steel.4(b). Under various conditions. water.3 Special multiphase applications 5 proﬁles designed for eﬃcient. The outside of the screws can also be covered with a variety of hard coating materials such as nickel based alloys. In order to ensure that suﬃcient liquid is available at conditions of high GVFs. while maintaining the full discharge pressure. Therefore. Some leakage ﬂow ﬁlls the reduction in gas volume in the cavity. a pump which can handle these diﬃcult liquids with high gas contents.
Finally in chapter 6. In chapter 5. a method to create a threedimensional grid of the screw pump and mesh methods for rotating screws is presented. single phase ﬂow in a nonrotating pump. the characteristics of the leakage ﬂow are studied by considering two simpler cases: the ﬂow through an annulus with rotating inner cylinder. This report is restricted to the ﬁrst stage and mainly focusses on the method to generate a suitable computational grid. It is used in chapter 4 to construct a threedimensional grid of the twin screw pump.1. and the ﬂow through a straightthrough labyrinth seal. Recommendations for the grid and turbulence model are given. FLOW Q Q = Qt GVF = 0. In chapter 2. single phase ﬂow in a rotating pump and thirdly.4 Goal and outline The aim of the research is to predict the leakage ﬂow rate of a multiphase twin screw pump by numerical simulations of the internal ﬂow. In chapter 3.4 Goal and outline 6 The leakage ﬂow rate through the most upstream clearance is the total leakage ﬂow rate of a twin screw pump for multiphase applications. the conclusion of the developed mesh method and simulated leakage ﬂow rate are discussed.95 plow phigh GVF=0 gas gas gas gas liquid liquid liquid Ql liquid ∆p (a) Typical twin screw pump performance curve (N=constant) (b) Gas compression by the leakage ﬂow Figure 1. . the generated grid is used to simulate the leakage ﬂow rate for nonrotating screws and singlephase ﬂow with the threedimensional grid. This research can be divided in three stages: ﬁrst. Also recommendations to expand the model for timedependent ﬂow simulations and multiphase ﬂow are discussed in this chapter. Secondly.4: Multiphase application of a twin screw pump 1. multiphase ﬂow in a rotating pump.
In section 2. The axial Reynolds number Re and the tangential Reynolds number Reω are deﬁned as: vax sρ µ ρωRs µ Re = Reω = (2. Secondly. First.1) (2.3 the leakage ﬂow through a stationary labyrinth seal is simulated for the corresponding Reynolds number of a twin screw pump. The leakage ﬂow in a twin screw pump is dependent on two phenomena. These two phenomena in the leakage area of a twin screw pump are represented by two characteristic ﬂow cases. the ﬂow in an annulus with rotating inner cylinder.2 the ﬂow in an annulus with rotating inner cylinder is simulated for the corresponding Reynolds number of a twin screw pump.2) . In the clearance area between screw and liner of a screw pump two directions of ﬂuid motion are present. based on hydraulic diameter and mean velocity. Secondly. the leakage ﬂow rate is dependent on friction in the small clearance between the tips of the rotating screw and the stationary liner. In section 2.1 Reynolds number in twin screw pump The Reynolds number is an important parameter for the ﬂow regime.2 is applied for a twin screw pump to estimate the leakage ﬂow rate for a complete twin screw pump. 2. In section 2. In section 2. represents laminar ﬂow and a higher Reynolds number represents turbulent ﬂow in general. For pipe ﬂow a Reynolds number smaller than 2100. the leakage ﬂow rate is dependent on the ﬂow resistance of the inlet and outlet of this clearance.1 the ﬂow regime in the clearances of the screw is considered and the corresponding Reynolds number is estimated. ﬂow through a stationary labyrinth seal.5 recommendations are made for the grid and turbulence model.Chapter 2 Simpliﬁed leakage ﬂow in a twin screw pump In section 2. For the annulus with rotating inner cylinder and the labyrinth seal the simulated ﬂow is compared with experimental results and recommendations for grid and turbulence model are given.4 the simpliﬁed ﬂow from section 2. First. First the leakage ﬂow in axial direction and secondly the tangential motion of ﬂuid in the clearance.1 and 2.
3) where L is the length of the annulus. When the leakage ﬂow is turbulent. To seal one cavity at least 2 screw threads are necessary. s the clearance. In ﬁgure 2. s the clearance between screw and liner. The maximum rotational speed typical for twin screw pumps designed for nonlubricating liquids is 1750 rpm 2 and for lubricating liquids 2900 rpm 3 . In this estimation of the axial Reynolds number the following phenomena are neglected: • Axial ﬂow proﬁle is turbulent. The axial Reynolds number is normally based on the hydraulic diameter dh = 2s but in equation (2.pdf . From these equations it is clear that the Reynolds numbers are dependent on dynamic viscosity and density of the medium. Also laminar ﬂow is considered in this estimation.10 screw pump.com Houttuin 216. The cavities A and B in this ﬁgure are connected. Thus the typical 3 total pressure build up over the pump is 16 bar 2 3 at maximum.000 and the tangential Reynolds number Reω is 3000 for crude oil with the parameters given in table 2. The dynamic viscosity of crude oil varies between 1.3). while the axial Reynolds number depends on the axial velocity. ρ the density of the medium.2. R the outer screw radius. ω the angular velocity of the screw. Reference is made to [2] for this equation.nl/contents/24940bro. Tangential velocity The tangential velocity is dependent on the outer radius of the screw and the rotational speed of the screws. considering single phase ﬂow. http://houttuin. • No inlet and outﬂow resistances are considered. resulting in an even higher axial velocity and Reynolds number. and the axial velocity will be lower resulting in a lower axial Reynolds number. ∆p the pressure diﬀerence. The tangential Reynolds number is dependent on the circumferential speed of the screw.1. vax = ∆ps2 12µL (2.1 Reynolds number in twin screw pump 8 where vax is the mean axial ﬂow velocity. which is in turn dependent on the diﬀerential pressure over the screw.1) the notation of [11] is used.nl/contents/21610bro.pdf 3 Houttuin 249. The diﬀerential pressure over one seal is the total pressure build up divided by the number of seals. Axial velocity The mean axial leakage velocity for HagenPoiseuille ﬂow in a cylindrical annulus is given in equation (2. Note that with this notation a axial Reynolds number larger than 1050 represents turbulent ﬂow in a pipe. and µ the dynamic viscosity. and µ the dynamic viscosity. but normally just over 2 screw threads are used to ensure proper sealing. the turbulent ﬂow proﬁle and the rotation of the screw produces deviations from this estimated mean axial velocity. this gives a diﬀerential pressure ∆p of 4 bar per seal. For example: 5 1 screw threads seal 4 cavities. 1 2 reference to http://www. This value of vax may serve as a ﬁrst estimate since no axial movement of the annulus is taken into account.engineeringtoolbox. The axial Reynolds number Re is estimated at 10.40 screw pump. http://houttuin. The total number of seals is dependent on the number of screw threads. Assuming 4 seals.4 · 10−3 − 20 · 10−3 kg/(ms) 1 . The variables not concerning the geometry of the screw will be discussed point wise: Viscosity A multiphase twin screw pump designed for pumping crude oil is considered. With this mean axial velocity vax and the twin screw pump dimensions the axial Reynolds number can be estimated.1 a schematic view and a photograph of two screw threads are given. and the clearance.
For the CFD simulations. The mean axial velocity will be lower resulting in a lower axial Reynolds number.2 Annulus with rotating inner cylinder The resistance of a water ﬂow in an annulus with a rotating inner cylinder is studied experimentally in [11]. . 2. In section 2.1: Parameters for Reynolds numbers of a twin screw pump µ ρ N ∆p L R s 1.3 m/s.1: Two threads of twin screw pump Table 2. and r the screw pitch in m.35 m/s at 1350 rpm and a screw thread of 60 mm. so the axial ﬂow proﬁle will be diﬀerent.2. • The screw rotates and this leads to a tangential motion of the ﬂuid and.2 Annulus with rotating inner cylinder 9 A B (a) Schematic view (b) Photograph Figure 2.4 · 10−3 kg/(ms) 800 kg/m3 1350 rpm 4 bar 22 mm 147 mm 0.2 this rotation is considered. This is low compared to the estimated mean axial velocity of 69.25 mm • The screw translates opposite to the leakage ﬂow. these experiments will serve as a testcase which will give recommendations for grid properties and turbulence model. The axial screw velocity is 1. The axial screw velocity is given by (N/60) · r. This secondary ﬂow leads to an additional resistance and a lower axial velocity and axial Reynolds number. at high rotation speeds. where N is the rotational speed in rpm. to the occurrence of Taylor vortices in the ﬂuid.
s the clearance between inner and outer cylinder of the annulus. NII and NIII ) and also the ﬂow rate is measured. and ρ the density. and δio is set to zero.2 the dimensions of the smallest annulus considered in the experimental study are given. The pressure is measured at three locations (NI .2. Anticipating a turbulent ﬂow. This is a steady axisymmetric problem.2 Numerical simulation For Re = 2230 and Reω = 3000 a numerical simulation of ﬂow in an annulus with rotating inner cylinder is performed with the commercial CFDpackage Fluent. A δio of 1.4.3. At every measuring location four holes of 0. there are no inlet and outlet resistances. Therefore.4 mm in diameter and 90o apart in the outer cylinder are connected to minimize measuring errors. vax the mean axial velocity.1 Experiment The geometry of the test setup used in [11] is given in ﬁgure 2. The clearance ratio s/R of a screw is much smaller than that of the annulus. 2.2. the pressure is measured at some distance from the inlet to minimize inlet ﬂow eﬀects. because the exact geometry of the annulus is used in the numerical simulations. The axial Reynolds number for leakage ﬂow in the clearances between screw and liner is more accurately estimated with equation 2. δio the resistance factor for the inlet and the outlet.3). the leakage ﬂow rate of a screw pump can not be accurately determined from the measurements. L the axial length.2.and outlet. Figure 2. In ﬁgure 2.2 Annulus with rotating inner cylinder 10 2. this diﬀerence has no inﬂuence on recommendations for grid and turbulence model. These Reynolds numbers indicate a turbulent ﬂow. the resistance coeﬃcient λ of various experiments with tangential Reynolds number Reω between 1000 and 20000 are given for axial Reynolds numbers Re in the range 100 to 25000. As is apparent from the ﬁgure. the pressure drop ∆p over the annulus is written as: ∆p L = λ + δio ρ 2s vax 2 2 (2.2. and compared with typical dimensions of a screw pump.5 is chosen according to [8] for the resistance of the in. It can be concluded that the clearance ratio does have an eﬀect on the resistance coeﬃcient λ (see ﬁgure 2. The axial Reynolds number is estimated at 2230.4) with λ the resistance coeﬃcient. and is solved using . For ﬂow in an annulus with rotating inner cylinder. In this ﬁgure the ﬂuid ﬂows from left to right.2: Experimental setup of an annulus with rotating inner cylinder (dimensions in mm) In table 2. However.
The continuity and momentum equations solved by the segregated solver are given in diﬀerential form and cylindrical coordinates as: 1∂ ∂ ∂ρ 1 ∂ + (ρrur ) + (ρuθ ) + (ρuz ) = 0 ∂t r ∂r rθ ∂z ∂ur uθ ∂ur u2 ∂p 1 ∂ 1 ∂ ∂ τθθ ∂ur ∂ur ρ + ur + + uz − θ =− − (rτrr ) + τθr + τzr − ∂t ∂r r ∂θ ∂z r ∂r r ∂r r ∂θ ∂z r .72 147. The segregated solver solves the governing equations sequentially (segregated from one another).0136 0. momentum. The segregated solver is used because in Fluent this solver is capable of adding multiphase models.43 0. The coupled solver in Fluent solves the governing equations of continuity.2.0017 Annulus Screw pump the segregated solver and implicit formulation.25 Clearance ratio s/R [] 0.3: Relation between resistance coeﬃcient λ and tangential Reynolds number Reω s for various R and axial Reynolds number Re Table 2.2: Clearance ratio s/R for a twin screw pump and an annulus with rotating inner cylinder Radius R [mm] 31. is not capable of adding multiphase models.25 Clearance s [mm] 0. The alternative. a coupled solver. and if necessary energy simultaneously as a set of equations.2 Annulus with rotating inner cylinder 11 Figure 2. The segregated solver can only be used with implicit formulation in Fluent.
This is expressed in the dimensionless wall distance y + . two nearwall approaches will be considered: ﬁrst wall functions will be evaluated. The required y + for nearwall modeling is of the order of 1. In [4] and [7] these methods are discussed. The required dimensionless wall distance 30 < y + < 200 is not reached with these Reynolds numbers (Re = 2230 and Reω = 3000). Wall functions use semiempirical formulas to resolve the ﬂow in large cells near the wall. In [4] en [6] these methods are discussed extensively. y=0 . The approaches each have diﬀerent requirements on the size of the ﬁrst cell at the wall. µ the dynamic viscosity and uτ w . Values of resistance coeﬃcient λ at intermediate Reynolds numbers are determined by linear interpolation. The term ∂ ∂ ∂θ is zero for axisymmetric problems.2. The calculated resistance coeﬃcient does not seem to be inﬂuenced much by the dimensionless wall distance. The value of y + is solution dependent: it will vary slightly for diﬀerent turbulence models. ncell = 5 is chosen to determine the accuracy of the diﬀerent turbulence models.6) where y is the distance of the center of the element to the wall. The upper bound is y + < 4 ∼ 5. In the remainder of this section. wall functions and nearwall modeling approaches. τ the shear tensor. The equations of the SIMPLE method are solved with the GaussSeidel algorithm for an Algebraic MultiGrid (AMG). In table 2.3. Another criterion for nearwall modeling is that there are at least 10 cells within the viscosityaﬀected boundary layer. Nearwall modeling uses small cells near the wall to resolve the ﬂow. Numerical models have two approaches to model turbulent ﬂow behaviour near solid walls. After this. In the simulations. In case of steady state calculations the ∂t term is also zero. the value of y + is calculated for grids with diﬀerent number of cells in radial direction. A value closer to the lower bound is most desirable.3 the dimensionless wall distance is given for diﬀerent number of cells ncell in radial direction. u the ﬂuid velocity.2 Annulus with rotating inner cylinder 12 ρ ρ ∂uθ 1 ∂ ∂uθ uθ ∂uθ ∂uθ ur uθ 1 ∂ ∂ 1 ∂p + ur + + uz − − 2 (r2 τrθ ) + τθθ + τzθ =− ∂t ∂r r ∂θ ∂z r r ∂θ r ∂r r ∂θ ∂z ∂uz uθ ∂uz ∂p 1 ∂ 1 ∂ ∂ ∂uz ∂uz (2. In this way iteratively a solution for the ﬂow ﬁeld and the pressure can be calculated that satisﬁes both continuity and momentum equations. The equations of continuity and momentum are implemented in the SIMPLE method (Semi Implicit Method for Pressure Linked Equations). This method calculates the ﬂow ﬁeld with an estimate of the pressure. numerical results will be compared with results from experiments as given in ﬁgure 2. and p the pressure. deﬁned as: y + ≡ ρuτ y/µ (2. and secondly nearwall modeling. Wall functions With the k − ǫ model with standard wall functions. To maintain some radial cells in the annulus and to keep the value of y + close to the required value. the pressure is corrected with the solution of the ﬂow ﬁeld to form a new estimate of the pressure. Discussing these methods falls outside the scope of this report.5) + ur + + uz =− − (rτrz ) + τθz + τzz ∂t ∂r r ∂θ ∂z ∂z r ∂r r ∂θ ∂z where ρ is the density. the friction velocity given by τρ where τw is the wallshear stress deﬁned as τw = µ ∂u ∂y The required y + for the wall function approach is 30 < y + < 200. but it gives a good idea of the dimensionless wall distance with that grid.
0452 0.6 27.2.0461 0.0447 0.0419 0.0 27.3: Dimensionless wall distance y + and resistance coeﬃcient λ for diﬀerent grids of the annulus problem determined with the standard k − ǫ model and standard wall function (Re = 2230.7 27.0553 dev [%] 20.0554 0.0422 0.0 4.0469 0.4 17. Reω = 3000) turb.0466 In table 2.0579. The axial length of the structured elements is not considered of importance due to the axial nature of the ﬂow.0442 0. y + = 18.7 20.5(a).0421 0. model k−ǫ version standard RNG diﬀerential swirl diﬀ+swirl realizable k−ω standard shear ﬂow corr.4: Resistance coeﬃcient λ for various turbulence models with wall functions determined for the annulus problem (ncell = 5.2 Annulus with rotating inner cylinder 13 Table 2.5 The k − ω turbulence models are most accurate in this situation.0442 0. In ﬁgure 2.4 the calculated resistance coeﬃcient λ is given for the k − ǫ and k − ω turbulence models with various model options and wall functions.8 20. the diﬀerence between the calculated and the experimental value of the resistance coeﬃcient λ is given.7 23.4 the y + dependency for the Standard and SST k − ω model is given. Also the results for the standard k − ǫ turbulence model are given in this ﬁgure.0544 0. see ﬁgure 2. SST option wall function standard nonequilibrium standard nonequilibrium standard nonequilibrium standard nonequilibrium standard nonequilibrium standard nonequilibrium λnum 0.6 22. Furthermore. Reω = 3000) ncell 4 5 6 7 8 y+ 22 18 15 13 11 λnum 0. 5 and 10 times) give the same resistance coeﬃcients λ.0455 0.3 23.3 6.1 21. λexp = 0. .0477 0.0459 0.9 19.Re = 2230.0 4.0463 0. using grids with diﬀerent values of ncell . Table 2. A structured grid with equilateral edges is used.0461 0.0466 0. Additional numerical simulations with elongated cells in axial direction (2.0464 0.
4: Relation between deviation of calculated and experimental resistance coeﬃcient λ for diﬀerent y + values for the annulus problem (Re = 2230.5: Resistance coeﬃcient λ for diﬀerent Re combinations with standard k − ω model and wall functions (ncell = 6) Re 1000 Reω 2000 3000 6000 2000 3000 6000 y+ 10.0457 0.538 0. There is a high deviation at y + = 10. over prediction for low y + values and under prediction for higher y + values.112 0. Still. In [6] this dependency on y + was also mentioned for k − ω turbulence models. for the k − ω turbulence model. 10 n 5 cell =8 7 6 standard k−ω SST k−ω standard k−ε 0 5 Deviation [%] −5 4 −10 −15 n −20 cell =8 7 6 5 4 −25 10 12 14 16 y + 18 20 22 24 Figure 2. these models can be used because the deviation is smaller than with the k − ǫ models.135 0.0522 0. Table 2.105 0. Reω = 3000.5 12 17 17 18 21 λnum 0.2 Annulus with rotating inner cylinder 14 The dependency on y + is obvious.5 which could be explained by the linear (laminar) law that Fluent employs for turbulent boundary layers at approximately y + < 11.543 0.068 dev [%] 38 14 11 14 1 9 2500 .098 0. The deviation with the experimental results and the dimensionless wall distance are given in table 2. standard wall function) The standard k − ω model is used to calculate the resistance coeﬃcient for diﬀerent axial and tangential Reynolds numbers.2.152 0.5.0616 λexp 0.076 0.
2.6: Dependency of dimensionless wall distance y + for diﬀerent grids determined with standard k − ω turbulence model and nearwall modeling for the annulus problem (Re = 2230. 15 and 20 and a smooth grid reﬁnement towards the solid walls (ﬁgure 2. For ncell = 15 the resistance coeﬃcient is determined for the diﬀerent turbulence models with nearwall modeling. This would bring the total number of cells up. while keeping in mind that the total number of cells should be kept as low as possible.6 the corresponding y + values of the diﬀerent grids are given. so simulations are performed to test this criterion.5(b)).0632 2. ﬁrst the dimensionless wall distance will be calculated for diﬀerent grids. air ﬂows through the seal with axial Reynolds numbers in a range of 300 to 7500 . The results of numerical simulations are compared with the experimental results to give recommendations for the grid and turbulence model. Table 2. In this study.3.7.2 1. For Re = 2230 and Reω = 3000 the dimensionless wall distance with ncell = 10 is too high.3 Labyrinth seal Leakage ﬂow in a twin screw pump has a stream path that is similar to the stream path through a ’straightthrough labyrinth seal’. 2.7 λnum 0. Reω = 3000) ncell 10 15 20 y+ 7 3.0517 0.1 Experimental setup Airﬂow through a stationary straightthrough labyrinth seal is studied experimentally in [5]. In table 2. ncell equal to 10.3 Labyrinth seal 15 (a) wall function (ncell = 5) (b) nearwall modeling (ncell = 15) Figure 2. results are given in table 2.5: Partial mesh of the annulus with rotating inner cylinder for diﬀerent wall approaches Nearwall modeling Similar to the previous section.0548 0. Three diﬀerent grids are considered. Another criterion states that there are at least 10 cells in the viscosity aﬀected region [4].
[%] 1.6: Experimental setup of straightthrough labyrinth seal Table 2.9 0.2 5. The Mach number Ma is a measure for the variation of density. The larger sealing length of the clearance has to be taken into account to estimate the leakage ﬂow rate in a screw pump.3.6 a schematic view of the labyrinth seal is given.2 diﬀerential swirl diﬀ+swirl k−ω realizable standard shear ﬂow corr.0578 0. In table 2.0572 0. The total leakage ﬂow rate of the screw pump can not be estimated directly with the leakage ﬂow rate through the labyrinth seal. Reω = 3000) turb.25 6 6 Twin screw pump 0. Re = 2230.8 the dimensions of the labyrinth seal and the twin screw pump are given.7: Resistance coeﬃcient λ for various turbulence models with nearwall modeling for the annulus problem (ncell = 15.6 6.8: Dimensions of straightthrough labyrinth seal and twin screw pump 2.25 295 22 60 85 Figure 2.0 0. For a twin screw pump the sealing between screw cavities is also established by a relatively long sealing clearance.0578 0.2 Numerical simulation A numerical simulation in Fluent is performed to calculate the mass ﬂow rate through the labyrinth seal for Re equal to 1572 and 2195. according to Ma2 ∝ ∆p.0506 0. In ﬁgure 2.0578 0. However. Inlet and outﬂow resistances are the main sealing principle of a labyrinth seal.2 1.0574 0. this diﬀerence has no inﬂuence on recommendations for grid and turbulence model. the length of the clearance L and the pitch r are considerably smaller than for the screw pump.0573 0. model k−ǫ version standard RNG option λnum 0. station 1 L r station 2 outlet s h1 D s D L r h1 [mm] [mm] [mm] [mm] [mm] Labyrinth seal 0.36 356 0. The clearance s and the diameter of the labyrinth seal D are similar to the screw pump.3 Labyrinth seal 16 Table 2.0543 dev. because of this diﬀerence in sealing principle and the nonrotating seal.2 0. λexp = 0. y + = 3.2 0.0579. A density variation smaller than ten percent is . SST in the seal clearance.4 12.0548 0.2.
9 0.0281 2.695 108. In table 2.9: Experimental results labyrinth seal p1 [kP a] 100.0437 14.1738 1. The ﬂow through the labyrinth seal results in a vortex in the seal cavity. and compressibility is neglected in the numerical simulation.027303 0.28 present here.0421 10.2 0. m = 0.6 0. Therefore.4 Leakage ﬂow in twin screw pump The ﬂow through an annulus gives an idea of the leakage ﬂow rate in the clearances between the screw and the liner of a screw pump. only wall functions are considered for the simulation of ﬂow through a labyrinth seal.6 k−ǫ k−ω Standard RNG Realizable Standard SST 2.10 the calculated mass ﬂow rate and the deviation with the experimental mass ﬂow rate is given.2.0379 0.0294 7. y + > 14. Anticipating the results presented in chapter 3.7 0.8.7: Structured quadrilateral grid of straightthrough labyrinth seal Table 2. In this section the leakage ﬂow is estimated with the method to calculate the mean axial ﬂow through a stationary annulus and an annulus . it is concluded that the number of cells should be kept as low as possible. y + > 11.7 0.806 m [kg/s] ˙ 0. m = 0.3 0. which is clearly visible at the streamlines of the ﬂow in the labyrinth seal. In Fluent a steady axisymmetric problem is solved with the segregated solver and implicit formulation using standard wall function for incompressible ﬂow. Figure 2.0304 11.038118 ρ [kg/m3 ] 1.503 93.7.0299 9.0381 ˙ m [kg/s] dev [%] ˙ 0. given in ﬁgure 2.0273 ˙ m [kg/s] dev [%] ˙ 0. Diﬀerent turbulence models are evaluated with the grid given in ﬁgure 2. This is a structured quadrilateral grid with 5 radial cells in the seal clearance.10: Mass ﬂow rate in a stationary labyrinth seal for various turbulence models using wall functions Re = 1572.20 0.4 Leakage ﬂow in twin screw pump 17 Table 2.2680 Re [] 1572 2195 Ma [] 0.0399 4.0267 2.5 0.793 p2 [kP a] 92.4 0.4 Re = 2195.0428 12.
5 for the inlet and outlet resistance δio .9) (2. The displacement volume VD for the screw is determined by the dimensions of the screw and liner.7) (2. First. since the resistance coeﬃcient λ depends on axial Reynolds number. using equation (2. the Ascrew = 37 · 10−3 m2 . and is considered in the ratio Qcl /Qt .2. now with a value of 1. . using equation (2. VD the displacement volume for one revolution. r the screw pitch. Secondly. In table 2. Ascrew the area of the crosssection of the screw. R the outer radius of the screw.8: Streamlines of the ﬂow in a straightthrough labyrinth seal with rotating inner cylinder. For the theoretical ﬂow rate Qt reference is made to chapter 1.8) where h is the distance between the centers of the screws. The leakage ﬂow rate in a twin screw pump is estimated in three ways.11 the leakage ﬂow rate is given for the diﬀerent methods.4) for a turbulent leakage ﬂow through an annulus with rotating inner cylinder. also without inlet and outlet resistances.4 Leakage ﬂow in twin screw pump 18 Figure 2. The diﬀerence in clearance ratio s/R between the twin screw pump and the annulus with rotating inner cylinder is neglected. The theoretical volume ﬂow rate is given by: 1 VD N 60 = (Aliner − 2 · Ascrew ) · r Qt = VD (2. The leakage ﬂow rate through the clearance between screw and liner is part of the total leakage ﬂow. The leakage ﬂow rate through the clearance between screw and liner Qcl of a twin screw pump is approximately given by the following equations: Qcl = Acl = vax Acl 4π − 4 cos h/2 R Rs (2. The ratio between the total leakage ﬂow rate and the theoretical pumped volume ﬂow rate of a screw is an important pump performance parameter. For the screw given in appendix B. Aliner the area of the crosssection of the liner. and Aliner = 124 · 10−3 m2 . For the annulus with rotating inner cylinder an iterative procedure is applied. This gives a displacement volume VD of 3 · 10−3 m3 /rev. Thirdly. s the clearance between screw and liner. and thus on the leakage ﬂow rate.10) where N is the rotational speed of the screws in rpm. using equation (2.4) for a turbulent leakage ﬂow through an annulus with rotating inner cylinder. and Acl the projected area of the clearance on a plane perpendicular to the axial direction.3) assuming laminar leakage ﬂow through a stationary seal without inlet and outlet resistances.
it was concluded that the standard k − ω model performs good. Nearwall modeling is not preferred.0056 Qcl /Qt [%] 36. Anticipating the results as presented in chapter 3 it is concluded that the number of radial cells must be limited and that nearwall modeling is not preferable. These two phenomena in the leakage area of a twin screw pump are represented by two characteristic ﬂow cases.3 22. B) for Reω = 3000. Thirdly.0675 0.7 8. First. Secondly. . The standard k − ω and standard k − ǫ turbulence model perform best for the ﬂow in a labyrinth seal.0246 0. Secondly. and performs good for the ﬂow in an annulus using nearwall modeling.0079 0. the leakage ﬂow rate is dependent on the ﬂow resistance of the inlet and outlet of this clearance.0675 Qcl [m3 /s] 0.0675 0. considering turbulent ﬂow in an annulus with rotating inner cylinder without inlet and outlet resistances.5 Conclusion The leakage ﬂow in the clearance between screw and liner of a twin screw pump is dependent on two phenomena. In [6].3 stationary + laminar rotating + turbulent rotating + turbulent + inlet/outlet losses 2. the ﬂow phenomena in an internal combustion chamber were represented by two characteristic ﬂow cases: the backward facing step and the free jet ﬂow. The k − ω turbulence model performs best for the ﬂow in an annulus with rotating inner cylinder using wall functions. First. but has a grid dependency. The k − ǫ turbulence model performs best for ﬂow in an annulus using nearwall modeling.3 15. however. ﬂow through a stationary labyrinth seal. The axial Reynolds number in the clearance between screw and liner is estimated in three ways. In this study.5 Conclusion 19 Table 2. N = 1350 rpm vax [m/s] 69.4 11. considering laminar ﬂow in a stationary annulus with rotating inner cylinder without inlet and outlet resistances. turbulent ﬂow in an annulus with rotating inner cylinder with inlet and outlet resistances. Based on diﬀerent grid reﬁnements and diﬀerent turbulence models. the ﬂow in an annulus with rotating inner cylinder.000 3220 2230 Qt [m3 /s] 0. the leakage ﬂow rate is dependent on friction in the small clearance between the tips of the rotating screw and the stationary liner. and the standard k − ω turbulence model with wall functions is preferred for the simulation of ﬂow in a twin screw pump.7 Re [] 10. ∆p = 4 bar. First. it is shown that this dependency has little eﬀect on the accuracy to represent the leakage ﬂow rate in an annulus with rotating inner cylinder.11: Estimated leakage ﬂow of oil through a twin screw pump (app. The axial Reynolds number in the clearance is determined by the ﬂow in an annulus. The axial Reynolds number reduces when rotation and inlet and outlet resistances are considered.2. Secondly.
Furthermore the quality of the grid has to be acceptable to perform meaningful ﬂow calculations. Therefore. recommendations are made to setup a timedependent simulation of a screw pump in section 3.Chapter 3 Method of threedimensional meshing In chapter 2. 3.4. the number of cells in the clearance constitutes a substantial part of the total number of cells to describe a complete twin screw pump. and Acl the crosssectional area of the clearance.3. In the ﬁrst section of this chapter the number of cells for an unstructured. First the number of cells in the clearance is estimated in case an unstructured tetrahedral mesh is used.1 Number of cells in the clearance region The number of cells in the clearance region of a screw pump is estimated because the axial distance L of the clearance is high compared to the radial distance s of the clearance between screw and liner. This is discussed in section 3. all cells are truly tetrahedral with edges of equal length. Therefore. The volume of such a regular tetrahedral cell Vtet is given by: 1 √ Vtet = a3 2 (3. In an unstructured tetrahedral mesh of high quality. the inlet and outlet remain stationary. given in equation (2. This requires a dynamic mesh. A nonconformal grid does not require that nodes coincide on the combined faces. This could give problems connecting those regions if nodes should always coincide. The approximate volume of the clearance region (ﬁgure 3. L the (axial) screw thickness. a limited number of cells is desirable.1) where nrev is the number of screw threads. The calculation time is dependent on the number of cells used to simulate the ﬂow in a screw pump. For threedimensional ﬂow calculations the number of cells can be very high. Finally. A timedependent simulation of ﬂow in a twin screw pump requires the screws to rotate.2) 12 .8). After this. thereby the cavities between the screws progress to the pump outlet. While the screws are rotating.1(a)) is given by the following equation: Vcl = nrev LAcl (3. and a structured grid is estimated assuming cells with perfect quality (based on skewness). Methods for a dynamic mesh are discussed in section 3. the number of cells is estimated in case a structured mesh is used.2. twodimensional leakage ﬂow in screw pumps is discussed.
3.3). The volume of a hexahedral cell is given by: Vhex = a3 fAR (3.1: Twin screw pump where a is the edge length of the cell.1 Number of cells in the clearance region 21 (a) Clearance between screw and liner (b) 5 threads of the screws Figure 3.4) With a from equation (3. The length a is determined from the required number of cells in radial direction from screw to liner.2: Triangular cells in clearance between screw and liner The number of structured cells in the clearance between screw and liner is estimated. The height h of a regular tetrahedral √ s cell is h = a 6. the number of unstructured cells in the clearance nunstr is approximately 1.3 the structured mesh is given. The skewness of structured hexahedral cells is not aﬀected by the aspect ratio of the edges of the cell fAR .3) The estimated number of unstructured cells in the clearance between screw and liner is given by: nunstr = Vcl nrev LAcl = 1 3√ Vtet 2 12 a (3.2 the height h is given by h = ncell . ncell . s 1 2 · · ncell Figure 3.1 and in appendix B a complete drawing of the twin screw pump is given. and ncell = 5 according to chapter 2. In ﬁgure 3. In ﬁgure 3. The global dimensions of the twin screw pump are given in table 2. This gives the following 3 expression for a: a= √ 3s 6 ncell (3.5 · 109 cells for a twin screw pump. similar to the unstructured case.5) .
The initial spacings of the edges before any boundary motion constitute the equilibrium state of the mesh. A displacement at a given boundary node will generate a force proportional to the displacement along all the springs connected to the node. grids have to be generated with the same grid topology for every time step. For nontetrahedral cell zones (nontriangular in 2D).1).1 Smoothing In the spring based smoothing.3: Hexahedral cells in clearance between screw and liner with aspect ratio fAR The length a of the edge of a cell is determined by clearance s. The nodes can be repositioned by smoothing. For ALEcalculations. dynamic layering and remeshing. while maintaining the grid topology. the number of cells in radial direction across the clearance.. • The motion is predominantly normal to the boundary zone. and ncell = 5. the number of unstructured cells in the clearance nstr is approximately 3. .e.2. 3. the springbased method is recommended when the following conditions are met: • The boundary of the cell zone moves predominantly in one direction (i.6) The estimated number of structured cells in the clearance between screw and liner is given by: nstr = Vcl nrev bAcl fAR = Vhex a3 (3. and the chosen value for the aspect ratio fAR . the edges between any two mesh nodes are idealized as a network of interconnected springs.2 Dynamic mesh 1 2 s m 1 2 · · ncell a/fAR a a 22 · · · · · · · Figure 3. fAR = 10. no excessive anisotropic stretching or compression of the cell zone). Still an initial mesh has to be generated before repositioning and ALEcalculations can be performed. The length a is then given by: a= s fAR ncell (3.6). 3. ALEcalculations are a method to perform calculations on cells with nodes moving in time. ncell .7) With a from equation (3. In Fluent a dynamic mesh can be achieved by repositioning of nodes. this is not a method to reposition the nodes. Vcl from equation (3.3 · 106 cells. Another dynamic mesh method is the use of ALEcalculations.2 Dynamic mesh In this section the distortion of the mesh is considered.3.
2 Dynamic layering method In prismatic (hexahedral and/or wedge) mesh zones.5D surface remeshing (for 3D ﬂows only). This method adjusts the location of each mesh node to the geometric center of its neighboring nodes. When the boundary displacement is large compared to the local cell sizes.e.3 Remeshing On zones with a triangular or tetrahedral mesh. Fluent evaluates each cell and marks it for remeshing if it meets one or more of the following criteria: • It has a skewness that is greater than a speciﬁed maximum skewness.. 3. local face remeshing (for 3D ﬂows only). since repositioning a node by Laplacian smoothing can result in poor quality elements. Laplacian smoothing is the most commonly used and the simplest mesh smoothing method. The dynamic mesh model in Fluent allows to specify an ideal layer height on each moving boundary.2 Dynamic mesh 23 If these conditions are not met. the skewness has been improved). one can use dynamic layering to add or remove layers of cells adjacent to a moving boundary. Fluent only relocates the node to the geometric center of its neighboring nodes if and only if there is an improvement in the mesh quality (i.5D model. the resulting cells may have high skewness values. the springbased smoothing method is normally used. the new cells are discarded. Using the Local remeshing method.3. This method is computationally inexpensive but it does not guarantee an improvement on mesh quality.g.2. the cell quality can deteriorate or the cells can become degenerate. e. Fluent marks cells based on cell skewness and minimum and maximum length scales as well as an optional sizing function. where the available remeshing method only work on wedges extruded from triangular surfaces... The exception is the 2. since not all possible combinations of node pairs in nontetrahedral cells (or nontriangular in 2D) are idealized as springs. • Its height does not meet the speciﬁed length scale (at moving face zones. based on the height of the layer adjacent to the moving surface. If the new cells or faces satisfy the skewness criterion. The available remeshing methods in Fluent work for triangulartetrahedral zones and mixed zones where the nontriangular/tetrahedral elements are skipped. 3. and 2. face region remeshing. • It is larger than a speciﬁed maximum length scale. will lead to convergence problems when the solution is updated to the next time step. above a moving piston). Fluent includes several remeshing methods that include local remeshing.g. To overcome this problem. To circumvent this problem. the mesh is locally updated with the new cells (with the solution interpolated from the old cells). The layer of cells adjacent to the moving boundary is split or merged with the layer of cells next to it based on the height of the cells in layer attached to the moving boundary. This will invalidate the mesh (e. • It is smaller than a speciﬁed minimum length scale. . result in negative cell volumes) and consequently. Otherwise.2. Fluent agglomerates cells that violate the skewness or size criteria and locally remeshes the agglomerated cells or faces.
Reference is made to [10] for these equations. The Lagrangian method is a timedependent CFD method in which the calculation grid moves with the local ﬂuid velocity.2 Dynamic mesh 24 Face region remeshing method In addition to remeshing the volume mesh. especially when vortices occur. the velocity of the boundaries of the control volumes ub must be known for each time step. For face region remeshing.3. ∂t ∂x ∂x geometric conservation law conservation of mass conservation of momentum (3. It has a particular advantage that advective terms eﬀectively disappear in the ﬂow equations. commercial CFD packages are available that can handle ALE calculations. Fluent marks the region of faces on the deforming boundaries at the moving boundary based on minimum and maximum length scales. Rotation of the screws equals axial displacement of the grid. Fluent marks deforming boundary faces for remeshing based on moving and deforming loops of faces.2. . The boundary velocity can be calculated from the position of the nodes at subsequent time steps separately from the NavierStokes equations.8) where ρ is the density. More screw threads are easily added due to periodicity. It is however severely limited by tangling of the grid in multidimensional problems. The only limitations are the need to have a valid grid at each time step. Once marked. In equation (3. In the Eulerian method. and p the pressure. but is incapable of representing moving domain boundaries. the calculation grid is stationary. The positioning of the nodes only has to be performed once for one screw thread. The last layers can be removed and added to the front to complete the grid. The ALE method is a mix of both methods in that it allows the grid to move with a velocity that is independent of the ﬂow solution.4 Arbitrary LagrangianEulerian calculations (ALE) The Arbitrary LagrangianEulerian method is a method allowing timedependent CFD. u the ﬂuid velocity vector. and the restriction that the grid topology must be maintained. 3. For the twin screw pump this ALEmethod has a particular advantage. At the present. this is also applicable for the initial grid of other meshes.8) the NavierStokes equations are given. It is a mix between the Lagrangian and the Eulerian methods. It obviously is not hampered by tangling grids. To discretize the NavierStokes equations in ALEformulation with a ﬁnitevolume technique. ∂t Ω ∂Ω ∂ ∂ρ + (ρ(u − ub )) = 0. Fluent remeshes the faces and the adjacent cells to produce a very regular mesh on the deforming boundary at the moving boundary. Fluent also allows triangular and linear faces on a deforming boundary to be remeshed. ∂ dV − ub · dS = 0. The latter means that the only allowed action on the grid is a free movement of the nodes. ∂t ∂x ∂ρu ∂ ∂p + (ρu(u − ub ) − τ ) + = 0. τ the shear tensor. ub the velocity vector of the cell boundary.
3 Nonconformal mesh 25 3. The inlet and outlet of the pump can be connected with a nonconformal mesh interface. ALEcalculations are performed for twin screw compressor in [9].3 Nonconformal mesh While the screws are rotating. In Fluent it is possible to use a grid composed of cell zones with nonconformal boundaries. without change in topology. the grid node locations need not to be identical at the boundaries where two subdomains meet. The total number of cells to describe the twin screw pump has to be less than a few million cells. That is. To develop a dynamic mesh. the inlet and outlet remain stationary.3. an unstructured mesh of the clearances between screw and liner is impossible due to the large number of needed cells. smoothing and remeshing can be used. not from the two interfaces separately. . Fluxes across the grid interface are computed using the faces resulting from the intersection of the two interfaces. so this method is considered feasible for a twin screw pump. or special care has to be taken to make the nodes at the interface coincide. even when the topology of the mesh changes in time. are preferred above smoothing and remeshing because of the lower demands on PC performance. ALEcalculations for a dynamic mesh. A nonconformal grid does not require that nodes coincide on the combined faces.4 Conclusion From the estimation of the number of cells in the clearance between screw and liner. 3. The clearance between screw and liner can be represented with structured hexahedral cells. The aspect ratio of the hexahedral cell has to be larger than 10 to make a model of the total twin screw pump with a few million cells. Connecting those regions could give problems if nodes should always coincide. to be solvable with normal PC performance and memory.
1 quadrilateral cells will be placed on this crosssection.4 grid quality improvement by placing the nodes on another crosssection is evaluated.1 Basic structure The algorithm described here. With the present boundary conditions these gradient lines are also perpendicular to the boundaries. to form a threedimensional grid.Chapter 4 Construction of structured grid For a CFD calculation of a twin screw pump. the need for a structured grid is twofold. structured elements are preferred to unstructured elements.2. Therefore. After solving this problem. the direction of the gradient ∇Φ is perpendicular to the potential lines. At this stage the boundary conditions for the potential Φ are: Φ = 0 for the liner. Φ = −1 for the left screw. A valuable characteristic of the solution of the Laplace equation is that the equipotential lines never cross. This is corrected with small adjustments in section 4. In section 4. and in section 4. see chapter 3. and Φ = 1 for the right screw. positions the nodes based on the solution of the Laplace equation in a twodimensional section of the ﬂow domain (see ﬁgure 4. . In section 4. twodimensional grids with small axial distance apart are computed as described above. equipotential values of this problem are given in ﬁgure 4. except near the bottom and top cusp. To create a CFD simulation that is solvable with normal PC performance and within a reasonable time. First the number of elements needs to be kept within limits and secondly the quality of the grid needs to be well controlled. In section 4. The Laplace solution is obtained on an unstructured grid (ﬁgure 4. the number of elements has to be smaller than a few million. In this chapter a method to divide a twin screw pump in structured hexahedral cells is discussed. A grid based on equipotential and gradient lines can obtain high quality almost everywhere. Highly distorted cells are generated near the cusps since the potential line Φ = 0 does not end in the top and bottom cusp. These twodimensional grids have an equal number of radial and tangential nodes.2 some reﬁnements will be applied to this quadrilateral twodimensional grid. perpendicular to the axial direction.5 conclusions about the applicability of the grid are discussed.2(a)) of a crosssection normal to the axial direction with wellchosen boundary conditions. Furthermore.3 the quality of the twodimensional and threedimensional grid will be evaluated. In section 4. This method consists of the layering of twodimensional crosssections. To expand this twodimensional grid to three dimensions. but also with a limited number of well positioned elements. The representation of small clearances between screws and the liner can be done using many unstructured elements. 4.1).2(b).
T y x B Figure 4. with boundary conditions constant along each separate wall Figure 4. The nodes are written in a mesh ﬁle suited for calculation with Fluent.1 Reﬁnements Dividing line The equipotential line Φ = 0 does not give a satisfactory dividing line between the left and right screw. In table 4.2 Reﬁnements 27 Now nodes with equal index are connected to form a threedimensional mesh of hexahedral cells. screw and liner. A smooth dividing line from top to bottom cusp is created by: ﬁrstly changing the boundary conditions.2 4. deﬁning a line with a potential dependent on the vertical coordinate y.2: Unstructured mesh and potential ﬂow solution of crosssection of twin screw pump 4. The boundary condition on the liner becomes a linearly changing potential . Faces are grouped in inlet. outlet.1: Crosssection of a twin screw pump. indicating top (T) and bottom (B) cusps (not to scale) Φ = −1 Φ=1 Φ=0 (a) Unstructured mesh of crosssection of a twin screw pump (b) Contourlines of potential value. More detailed information of the screw is given in appendix B. Furthermore a theoretical screw proﬁle without clearance is described in appendix A.1 the characteristic dimensions of the screw are given.4. to gain more insight in the shape of the screw. In [9] and [10] methods to create a structured mesh are described. and secondly.2. The position of the grid nodes are calculated with the CFDprogram Comsol with Matlab interface.
2 Reﬁnements 28 Table 4.25 mm 62. the resulting gridlines are not orthogonal to the wall.24 − 0.3(b)).67 are chosen.5 mm 147. The dividing line with a varying potential ΦD .345 and β = 0. nodes are placed equidistantly. The constants α and β are chosen to create a smooth dividing line.5 mm 210 mm 0.5 mm from ΦT to ΦB . The parameter y changes from zero in cusp B to one in cusp T. .liner Clearance screw .2.4(a)). However the deviation turns out to be small in practice. 4. and this results in nonorthogonal cells (see ﬁgure 4. nor are the grid cells orthogonal themselves.2. These gridlines are then partitioned equidistantly to obtain the nodes of the structured twodimensional grid.3(a).screw Pitch of the screw Number of threads Radius rounded edges R ri Rliner h s r nrev 147. Since the potential does not have a constant value on the liner and the dividing line. see ﬁgure 4. This movement gives a grid near the top cusp as given in ﬁgure 4.1. represented by the dashed line in ﬁgure 4.4(b). From these nodes.4. which are directed along the potential gradients. It suﬃces to move these nodes to a spline created from neighboring cells on the same tangential line. so that a smooth line from top to bottom cusp arises (see ﬁgure 4. is deﬁned as: 1 + cos( y·π ) α 2 β y ≤ α ΦD = ΦB · α < y ≤ 1 − α ΦD = 0 y > 1 − α ΦD = ΦT · (4.265 mm 0.1) 1 + cos( (1−y)·π ) α 2 β where y is the normalized vertical distance. gridlines towards the screws emerge. The boundary conditions of the screws remain the same.38 mm 60 mm 1 53 0. 4.1: Dimensions of twin screw pump Screw outer radius Screw inner radius Liner radius Center oﬀset Clearance screw .3 Nonequidistant node placement On the liner and the dividing line. The values of ΦT and ΦB are chosen iteratively so that the line with constant potential (ΦT or ΦB ) departing from the cusp edge lies between the two tangents of the liner.2 Smoothing of nodes on gridline starting on the cusps The radial gridlines starting on the cusps have an equidistant node placement.29 − 0. here α = 0.
the surface mesh of the screws is displayed in ﬁgure 4. To avoid a discontinuity in cell volume for cells at the transition of dividing line and liner. The linear grow rate q is limited. Nonequidistant node placement reduces this discontinuity signiﬁcantly. towards the left and right screw have diﬀerent lengths. with linear varying boundary condition on the liner from ΦT to ΦB (not to scale) Figure 4. starting at points on the dividing line. The longest gradient line is divided in elements with a linear grow rate q.4: Gridlines near the top cusp in a crosssection of a twin screw pump Gridlines. In ﬁgure 4.3: Radial line near cusp (a) No adjustment of radial gridline starting in cusp (b) Adjusted radial gridline starting in cusp Figure 4.2 Reﬁnements 29 ΦT Φ = −1 Φ=1 T ΦB (a) Potential lines near the top cusp in a crosssection of a twin screw pump (b) Dividing line between top and bottom cusps. A reﬁnement to the equidistant placing of the nodes is applied. a threedimensional grid of the twin screw pump is generated. is clearly visible with equidistant node placement. With nonequidistant node placement. near the dividing line. see ﬁgure 4.5. this to prevent the formation of extremely large cells at the screw surface and discontinuities in cell volume in other parts of the domain than near the dividing line. The discontinuity in the area of the cells.7. a discontinuity in cell volume exists. nodes on gridlines starting at points on the liner are also placed nonequidistant.4. .6 the twodimensional grid for a crosssection is given. With equidistant placement of the nodes.
2) where θmax and θmin are the maximum and minimum angles (in degrees) between the edges of the cell. By deﬁnition.3.3.1 Theory A normalized measure of skewness is EquiAngle Skew (QEAS ). First the theory of the two methods is discussed in section 4.5: Equidistant and nonequidistant node placing (a) Equidistant node placement (b) Nonequidistant node placement Figure 4.6: Reﬁned quadrilateral grid of a crosssection of a twin screw pump 4.2. θeq = 60. It can be applied to twodimensional as well as threedimensional grids.3 Grid evaluation 30 equidistant nonequidistant left screw a dividing line aq aq 2 right screw Figure 4.3) where QEAS = 0 describes an equilateral cell. Table 4. . For triangular and tetrahedral cells.2 outlines the overall relationship between QEAS and cell quality. 4. and θeq is the characteristic angle corresponding to an equilateral cell of similar form. and the threedimensional grid also on volume ratio of neighboring cells. and QEAS = 1 a degenerated cell.3. For quadrilateral and hexahedral cells. θeq = 90.1 and applied in section 4. 180 − θeq θeq (4. 0 ≤ QEAS ≤ 1 (4. which is deﬁned for an individual grid cell as: QEAS = max θmax − θeq θeq − θmin .3 Grid evaluation The twodimensional and the threedimensional grid will be evaluated on skewness.4.
A tetrakis hexahedron are 24 tetrahedrals that form a hexahedral.4) .3 Grid evaluation 31 Figure 4. For derivation of this equation reference is made to [3].4 for threedimensional grids.75 < QEAS ≤ 0. Volume ratio VR only applies to threedimensional cells.9 < QEAS ≤ 1.9 0.8.7: Threedimensional structured screw surface mesh for one thread of a twin screw pump (color indicating axial coordinate) Table 4.0 QEAS = 1 Quality Equilateral (Perfect) Excellent Good Fair Poor Very Poor Degenerate In general. (− 7 − − 2 ) + (− 5 − − 0 ). (− 7 − − 4 ) + (− 3 − − 0 )] x x x x x x x x 12 V = (4.4).5 < QEAS ≤ 0. (− 7 − − 4 )] + x x x x x x x x 12 1 → → → → → → → → det [(− 7 − − 1 ). The volume of a tetrakis hexahedron is given by equation (4. 1 → → → → → → → → det [(− 7 − − 1 ) + (− 6 − − 0 ). Mesh Quality QEAS QEAS = 0 0 < QEAS ≤ 0. (− 3 − − 0 )] + x x x x x x x x 12 1 → → → → → → → → det [(− 6 − − 0 ). highquality meshes have an average QEAS value of 0.25 < QEAS ≤ 0. (− 7 − − 2 ). The numbering of the nodes is given in ﬁgure 4.1 for twodimensional grids. A common method to calculate the volume of hexahedral cells in numerical simulations is to estimate the volume of a tetrakis hexahedron. (− 5 − − 0 ).75 0.25 0. and 0.2: QEAS vs.5 0.4.
V Vi (4. see ﬁgure 4.9 the values for QEAS are given for a twodimensional grid of a crosssection and for one slice of threedimensional cells.2. The overall skewness improves.3) is limited to ensure a high volume ratio in most part of the mesh. The skewness near the ﬂanges cannot be improved by changing the grid of a crosssection.3.43 respectively.5) where i are the neighboring cells. . These mean values for QEAS normally represent a good quality grid.5 and a 5 times elongated grid is given.2 Evaluation In ﬁgure 4. A volume ratio of 1 represents cells with equal volume. Near the ﬂanges of the screw also highly skewed cells are created in the threedimensional mesh.2 the skewness for a 2. according to: VR = min i min Vi V . In table 4.4. The volume ratio VR is low at the dividing line near the cusps of the twin screw pump. Elongation of the screw pump in axial direction improves the cell skewness. So.11. so no further reﬁnements are made to the twodimensional grid of a crosssection.3 Grid evaluation 32 Figure 4.8: Tetrakis hexahedron with facecentered vertices (subdivision of front and back faces not drawn) The volume ratio VR is deﬁned as the maximum ratio of the volume of an cell with the volume of adjacent cells normalized between 0 and 1.3 the mean skewness QEAS for elongated structured threedimensional grids of a twin screw pump is given. The number of skewed cells near the ﬂanges is much larger than near the cusp. The grow rate q of cells on radial lines (section 4.9(d)). The eﬀect of elongation in axial direction on the skewness of the threedimensional grid is evaluated.9(d)) it is clearly visible that there are a large number of highly skewed cells in the threedimensional grid. but is responsible for this discontinuity in volume at the dividing line. Highly skewed cells in the twodimensional grid are on radial lines starting near the cusps. Changing the twodimensional grid of a crosssection can improve the volume ratio. Further reﬁnements of the twodimensional meshing method can improve the mesh. 4. Mean value for the skewness in the twodimensional and threedimensional grids is 0. the grid is not as good as the average value of QEAS wants us to believe. however some highly skewed cells remain present near the highly curved part of the screw (red area in ﬁgure 4. In ﬁgure 5.12 and 0. but from the distributions of QEAS (ﬁgure 4.
5 0.4 Threedimensional basis structure The quality of the twodimensional grid is not the limiting factor. Gradient lines depart perpendicular to the wall.9 1 0 0.6 skewness Q 0. so grid skewness should be improved. In ﬁgure 4.2 0.8 0.5 0.8 0.9 0.8 0. A potential problem is known to pose less stringent demands to the grid quality. ﬁrst a threedimensional grid of the screw is needed. In ﬁgure 4.5 0. Therefore. the skewed threedimensional structured grid can be used.(b) Skewness of one layer of threedimensional cells section projected on crosssection 3500 1500 3000 2500 number of cells number of cells 1000 2000 1500 500 1000 500 0 0.3 0.12(a) gradient lines are drawn.4 0.3 0. To solve the threedimensional potential problem.6 0. To determine the quality of the grid.9 0. By changing the boundary conditions this can be prevented. the improvement of the grid has to be evaluated.4. This favours the idea of solving the potential problem in three dimensions.4 Threedimensional basis structure 33 0.4 0.8 0. and using the equipotential surfaces and gradient lines as a basis for a threedimensional grid. Secondly. a crosssection in axial direction of the twin screw pump is made and the potential problem is solved in two dimensions with the following boundary conditions: liner Φ = 0 and screw Φ = 1.2 0.1 0. The expansion to three dimensions gives skewed cells.9 1 EAS (c) Distribution of skewness for the twodimensional (d) Distribution of skewness for one layer of threegrid of a crosssection dimensional cells Figure 4.1 0.9: Skewness QEAS 4.1 (a) Skewness of twodimensional grid of a cross.3 0.7 0. since many lines accumulate in the center of the cavity.1 0. see ﬁgure 4.12(b).6 skewness Q 0.4 0. These lines are not ideal.2 0.7 0.12(c) the distribution of skewness QEAS is given for this grid.6 0.2 0.7 0.5 0.7 0. Dividing the gradient lines equidistantly gives quadrilateral cells. The skewness for this two .4 EAS 0.3 0.
5 5 10 mean skewness QEAS 0.33 0. The Laplace problem is solved on an unstructured grid of the crosssection with suitable Dirichlet boundary conditions.21 dimensional section in axial direction is similar to the threedimensional structured grid of the twin screw pump already described. Twodimensional structured grids of crosssections perpendicular to the axial direction are layered to a structured threedimensional grid. Some adjustments are made to improve the quality of the twodimensional grid. in terms of skewness. 4. so this is no improvement of the grid. The twodimensional grid on a crosssection is placed along gradient lines of the Laplace problem.5 Conclusion 34 Table 4. Placing the gridlines along the equipotential surfaces and gradient lines of a threedimensional Laplace problem gives a skewness similar to the threedimensional grid created with the layering method.26 0.5 Conclusion A method to describe a twin screw pump with structured hexahedral cells is developed.35 0.43 0. The quality of the threedimensional grid.3: Mean skewness QEAS for elongated grids in axial direction Elongation e 1 2 2. All twodimensional grids have the same topology. . deteriorates due to the high surface curvature of the geometry. this is required to combine all twodimensional grids to a threedimensional grid. Placing nodes on basis of the threedimensional potential problem therefore does not improve the quality of the mesh much. compared to the quality of the twodimensional grid. and is not employed.4.
6 0.4 0.2 0.11: Volume ratio VR .8 3000 0.5 times elongated grid in axial direction of a 5 times elongated grid in axial direction 1500 1800 1600 1400 number of cells number of cells 0.4 EAS 0.10: Skewness QEAS of an elongated twin screw pump in axial direction 4000 0.4 0.5 0.6 0.1 0.3 0.6 0.7 0.7 0.1 0.2 0.6 skewness Q 0.4 Volume Ratio VR 0.2 0.8 0.3 0.8 0.9 0.9 0.4 0.8 0.5 0.4 0.3 0.8 0.9 0.2 0.6 0.5 0.7 0.2 0.5 Conclusion 35 0.9 1 1000 1200 1000 800 600 400 200 500 0 0 0.1 1000 500 0 2500 2000 1500 3500 1 0.7 number of cells 0.1 0.8 0.5 times elongated grid in axial dimensional cells of a 5 times elongated grid in axial direction direction Figure 4.5 0.(d) Distribution of skewness for one layer of threedimensional cells of a 2.5 0.9 0.7 0.3 0.7 0.1 (a) Volume ratio VR of one layer projected on crosssection (b) Distribution of volume ratio Figure 4.5 0.2 0.6 skewness Q 0.4.9 1 EAS (c) Distribution of skewness for one layer of three.3 0.3 0.1 (a) Skewness for one layer of threedimensional cells (b) Skewness for one layer of threedimensional cells of a 2.
5 Conclusion 36 (a) Gradient lines of Laplace problem with boundary conditions constant along each wall (b) Structured quadrilateral grid for an axial crosssection of a screw (c) Distribution of skewness for an axial crosssection of a screw Figure 4.12: Crosssection in axial direction of one screw thread .4.
A underrelaxation factor reduces the change of the variable for the next iteration. Water is taken as the leakage medium. Thus the total number of hexahedral cells is approximately 1 million.Chapter 5 CFD computations with structured grid 5.4). By lowering the underrelaxation factors and taking a low diﬀerential pressure per screw thread. The low quality of the grid results in non converging solutions with the standard parameters. The grid consists of one screw thread of the twin screw pump and each screw has 10 radial cells. 5. The axial Reynolds number varies from 19 to 381 in the clearances between screw and liner. a converged leakage ﬂow rate is calculated.2 Results There are three cases tested. First. For the initial condition the velocity and the turbulent kinetic energy are set to zero in the complete domain. threedimensional ﬂow calculations are performed. The leakage ﬂow rate through periodic stationary screws is calculated for a ﬁxed pressure drop over the screw thread.1(a) the clearance between the screw and liner is given by region 2. for a 5 times elongated grid of a twin screw pump the leakage ﬂow is simulated for axial Reynolds numbers up to 1514. The total leakage ﬂow rate is compared with the leakage ﬂow rate through the clearances between screw and liner. 300 tangential cells.1 Numerical setup With the structured grid generated in chapter 4. Also the mean axial velocity through the clearance between screw and liner is compared with the analytical equation for laminar ﬂow in an annulus with and without inlet and outlet resistances. and 180 axial cells (see ﬁgure 4. turbulence is modeled with the standard k − ω model with standard wall function. . The ﬂow is simulated with the steady segregated solver and implicit formulation.7). the leakage ﬂow is simulated for an elongated twin screw pump in axial direction at approximately equal Reynolds number in the clearance between screw and liner. Secondly. and the speciﬁc dissipation rate of the turbulence to one. Thirdly. the leakage ﬂow through a twin screw pump is simulated at low axial Reynolds numbers. Also calculations on an elongated twin screw pump are performed to evaluate the eﬀect of screw geometrie and skewness of the grid. In ﬁgure 5. given in equation (2.
1(a)). and a low dimensionless wall distance. Elongation in axial direction increases the length of the clearance between region O and P in ﬁgure 5.5 to 3 for the diﬀerent axial Reynolds numbers. the clearance between the screws gets larger and the ﬂow resistance reduces.1: Crosssections of twin screw pump In the ﬁrst case of the threedimensional ﬂow simulations the pressure drop is set from 300 to 9600 P a over one screw thread. The ˙ ˙ simulated mean axial velocity in the clearance between screw and liner is compared to the ﬂow rate in an annulus from equation (2. Elongation in axial direction reduces the ﬂow resistance from region R to S in ﬁgure 5.2 the leakage ﬂow rate is given for elongated screws with a laminar leakage ﬂow rate in the clearances between screw and liner. In the second case the grid is elongated.2 Results B 38 A 2 1 A B (a) Regions of leakage: 1 between screws and 2 between screw and liner.5. The eﬀect of the dimensionless wall distance on the turbulence model is discussed in chapter 2 and should be greater that 11 for wall functions. In table 5. The axial Reynolds number in the clearance between . The dimensionless wall distance in the clearance is lower than the required value.2(a). resulting in a low axial leakage velocity. it varies from 0.2(a) (not to scale) (b) Crosssection BB Figure 5.and outlet resistances. the resistance coeﬃcient is λ = 48Re−1 [11] for laminar ﬂow.1(b). In table 5. Secondly region 2.4) with and without in. but some tendencies are visible. No accurate solution of the numerical simulation can be expected. For a nonrotating stationary annulus. The leakage percentage of ﬂuid through the clearance between the screw and the liner compared to the total leakage ﬂow rate m2 /m. the length of the clearance between screw and liner gets larger and thus becomes a larger ﬂow resistance. For water the diﬀerential pressure should be approximately 130 kP a over one screw thread for realistic ﬂow and a axial Reynolds number of 2230. even with low underrelaxation factors. Crosssection AA given in ﬁgure 5. In the third case a 5 times elongated grid in axial direction is used and the diﬀerential pressure over the screw is varied.1 the simulated leakage ﬂow rate for a stationary twin screw pump is given for diﬀerent diﬀerential pressures over the screw. is higher at a higher diﬀerential pressures over the screw. the elongation of the grid changes the dimensions of the two leakage regions (ﬁgure 5. The used pressure diﬀerence per seal is low. At a higher diﬀerential pressure than 9600 P a over one screw thread the solution is nonconverging. First region 1.
7 7.5.58 1.4) are similar for laminar ﬂow up to a Reynolds number of approximately 500.3 the simulated leakage ﬂow rate through the twin screw pump is given.3 Conclusion The quality of the used grid is low.2 4.1 3.1 5. and λ = 0. (2. lowering underrelaxation factors only results in converged solutions for a low diﬀerential pressure over the twin screw pump. with a value of 1.90 1. In [11]. In table 5. vax in clearance between screw and liner ∆p [P a] 300 600 1200 2400 4800 9600 m [kg/s] ˙ 0.8 vax [m/s] eq.and outlet resistance δio .072 0. white: ﬂuid) (b) Crosssection CC (not to scale) Figure 5. The mean axial velocity in the clearance vax is given for the simulation and calculated with equation (2. (2.54 1.277 0.8 numeric 0.7 0.072 0.9 1.29 0.144 0.5 for in.0 1. the simulated and calculated mean axial velocity deviate from each other. The dimensionless wall distance for an axial Reynolds number of 1514 on a 5 times elongated grid is given in ﬁgure 5.5 0.145 0. First: elongation of the . From the simulations some conclusions can be draw.4) δio = 1.4) δio = 0 0.3 Conclusion 39 A ri R B (a) Crosssection AA in axial direction of two threads (black: screw/liner. At higher axial Reynolds numbers.6 10.51 0. The mean axial velocity of the numerical simulation and equation (2.24 for turbulent ﬂow.0 8.074 0.15 2.1: Water leakage ﬂow rate for threedimensional stationary twin screw pump simulations with various pressure diﬀerences per screw thread.4).142 0.8 11. There is no converged solution found for higher axial Reynolds numbers.2: Crosssections of twin screw pump screw and liner varies from 20 to 1500.26Re−0.3. Table 5. The value of the dimensionless wall distance y + in the clearance and in the cavity are approximately 8 and 1000 respectively. the resistance coeﬃcient for a stationary annulus is given as: λ = 48Re−1 for laminar ﬂow.51 Re 19 37 70 129 227 381 m2 / m [%] ˙ ˙ 4.28 0.3 eq.0 5.4 2. This low diﬀerential pressure results in a low leakage velocity and the dimensionless wall distance dependent on this velocity is too low to calculate an accurate solution with the k − ω turbulence model with wall functions.
14 0.8 2.1 7.8 screws in axial direction results in a larger leakage ﬂow for the same diﬀerential pressure over the screw thread.8 2.08 1.7 8.3 Conclusion 40 Table 5.0762 0.7 1.and outlet resistance for laminar and turbulent deﬁnition of λ ∆p [kP a] 1.9 4.5 3 6 12 24 48 96 192 288 384 m [kg/s] ˙ 4.9 2. In a twin screw pump the leakage ﬂow between the screw and liner is similar to ﬂow in an annulus.0 vax [m/s] 0. This expansion between the screws is most clearly visible in the crosssection given in .6 numeric 0.2 4.3 Table 5. In the numerical simu2 lations the value of K also remains approximately constant.0762 0. vax according to equation (2. The relation between diﬀerential pressure and axial velocity is given in equation (2.15 0.9 18.0740 0.8 2. For an expansion in the diameter of a pipe the relation between diﬀerential pressure and axial velocity is given in [8] and is ∆p = K 1 ρv 2 .5.2 37.8 1.7 vax [m/s] turbulent 0.7 2.0745 0.1 2.3 26. even for laminar ﬂow.57 1. where K is a constant for turbulent ﬂow.6 0.1 2.4) with in.57 0. however the leakage ﬂow rate between the screws increases.0 5. The leakage ﬂow rate through the clearances between the screw and liner has a larger portion of the total leakage ﬂow rate at a higher diﬀerential pressures for laminar ﬂow. The diﬀerential pressure is approximately linearly proportional to the axial leakage velocity for laminar ﬂow.0 Re 19 38 73 144 273 482 740 1005 1262 1514 m2 / m [%] ˙ ˙ 0.3: Water leakage ﬂow rate for threedimensional stationary twin screw pump simulations for a 5 times elongated grid in axial direction.7 1.6 2.29 0.0 1.7 4.6 2. Secondly: increasing the diﬀerential pressure per screw thread changes the leakage ﬂow ratio between the two regions.6 9.84 1.29 0. even more than the decrease in the other region. however the leakage ﬂow between the screws is similar to a expansion in the diameter of a pipe.1 4.6 2.2 1. The leakage ﬂow rate in the clearance between screw and liner reduces.072 0.8 7.1 1.4 62.0 5.38 0.4) for ﬂow in an annulus.4 12.3 10.2: Water leakage ﬂow rate for threedimensional stationary twin screw pump simulations with 5000 P a/m pressure diﬀerence for elongated grid in axial direction (Re ≈ 19) e 1 2 2.0745 0.5 75 laminar 0.6 0.6 52.0 6.57 1.2 12. This can be expected from analytical equations.5 5 10 m [kg/s] ˙ 0.7 6.3 0.3 8.0731 m2 / m [%] ˙ ˙ 4.
2(b) between region A and B.3: Dimensionless wall distance for e = 5 and pressure diﬀerence of 384 kP a over one thread for a twin screw pump (Re in the clearance is 1514) ﬁgure 5.00e+01 clearance 1.3 X Z Position (m) Y Figure 5.00e+04 1.00e+00 0 0.15 0.25 0.1 0.00e+03 cavity Wall 1.00e+02 Yplus 1. This indicates a constant resistance coeﬃcient λ for axial Reynolds numbers above a certain number or a negligible eﬀect of sealing length compared to in.2 0.5. At higher axial Reynolds numbers the ratio between leakage ﬂow rate through the clearances and the total leakage ﬂow rate is found constant in the numerical simulations.1(b) between region R and S.05 0. and also in ﬁgure 5.and outlet resistances.3 Conclusion 41 liner_y+ screw_ca screw_cl 1. .
however the threedimensional grid has poor quality. Here only wall functions are considered for the diﬀerent turbulence models. This grid has to describe the geometrie of the pump well. however the leakage ﬂow rate between the screws increases. Numerical simulations proved that this elongation has no inﬂuence on the ﬂow through an annulus with inner rotating cylinder. The leakage ﬂow rate in the clearance between screw and liner reduces. The quality of the twodimensional grid of a crosssection is good. Wall functions are needed to keep the total number of cells limited. The demands of the turbulence model are evaluated by comparison with the results of two experiments. Secondly. Secondly: increasing the diﬀerential pressure per screw thread changes the leakage ﬂow ratio between the two regions for the laminar ﬂow regime.1 Final conclusion To model the ﬂow in a twin screw pump a grid of the screw pump has to be made. The leakage ﬂow rate through the clearances between the screw and liner has a larger portion of the total leakage ﬂow rate . Also the turbulence model poses requirements on the dimensionless wall distance of the ﬁrst cell at solid walls. To obtain a solution for the ﬂow in a twin screw pump with normal PC performance and in reasonable time. First. Between the screw and the liner structured cells with elongated sides in axial and tangential direction have to be used to keep the total number of cells limited. even more than the decrease in the other region.Chapter 6 Conclusions and recommendations 6. Gradients of the solution in axial and tangential direction are low compared to the radial direction. the number of cells must be limited. The standard k − ω model gives the most accurate results when simulating with wall functions. The threedimensional cells have a high skewness in some regions. the ﬂow through a stationary straightthrough labyrinth seal is simulated. First: elongation of the screws in axial direction results in a large leakage ﬂow for the same diﬀerential pressure per screw thread. the ﬂow through an annulus with rotating inner cylinder is modeled. A grid of the twin screw pump is made and the leakage ﬂow through the twin screw pump is simulated for a low diﬀerential pressure over the pump. From the simulations some conclusions can be drawn. A mesh method is constructed that generates a structured hexahedral grid by layering of crosssections perpendicular to the axial direction. The dimensionless wall distance y + has to be larger than 11 to give accurate results with wall functions. Nearwall modeling and wall functions are considered for the diﬀerent variants of the k −ǫ and k −ω turbulence model.
meaning a mixed grid of hexahedral and tetrahedral cells. and curvature dependent distribution of nodes at the dividing line. adjustment of the grid in the cavities is necessary. This method consists of a hybrid grid. • There are some small improvements possible on the construction of the crosssectional mesh. or a nonconformal interface (section 3.6. while fulﬁlling the demands set in this report. the grid quality has to be improved to allow accurate CFD simulations. On the existing grid small improvements can be made but it remains questionable if the quality will improve enough to perform accurate CFD simulations. for example. • For comparison of twin screw pump simulations with experiments. The clearance between the screw and liner can be meshed with a structured hexahedral mesh and the interior with unstructured tetrahedral cells. This mesh can be used. So. Expanding the simulation to a diﬀerential pressure present in industrial screw pumps gives a better understanding of the ﬂow in a twin screw pump. This method has proven its applicability on a completely unstructured mesh of one screw thread of a twin screw pump with large clearances of 6 mm at consulting company Bunova Development BV. • The ﬁrst cells near the wall in a cavity are too large. A turbulence model is needed because in a dynamic simulation there are regions with turbulent ﬂow. Smoothing and remeshing can deteriorate the mesh. to determine which model is the most accurate for ﬂow in a twin screw pump. There is another method of meshing the twin screw pump. resulting in a high dimensionless wall distance. The nodes on the interface either have to coincide. • Validate behaviour of turbulence models for laminar ﬂow in the clearance between screw and liner. • The multiphase modeling packages in Fluent should be tested on validation cases. measurements of the ﬂow in the twin screw pump are needed. Like: automatic determination of ΦT . use can be made of smoothing and remeshing in Fluent. When the mesh deteriorates beyond a certain limit. correctly describe the roundoﬀ of the screw tips. 6. for a twin screw pump with a larger screw pitch. The static twin screw pump simulation shows realistic ﬂow features at a low diﬀerential pressure over the pump. Besides the numerical result this is also expected from the analytical equations for the relation between diﬀerential pressure and axial ﬂuid velocity in the two regions. At higher axial Reynolds numbers the leakage ﬂow ratio between the two regions is approximately constant. a new mesh can be used and the old solution can be interpolated on the new grid. ΦB .2 Recommendations 43 at higher diﬀerential pressures.3) can be used.2 Recommendations • The constructed threedimensional grid has highly skewed cells. • To create a timedependent simulation. . The static simulations can be extended to a dynamic simulation with multiphase ﬂow to obtain a complete model of the twin screw pump. There is laminar ﬂow present when more viscous oil is pumped.
[7] A.Bibliography [1] Karassik et al. Vierendeels. Graduation report TU/e. [4] Fluent Inc. [9] J. [3] J. and E. Graduation report TU (http://ta.through labyrinth seals. Eﬃcient computation of volume of hexahedral cells. Procestechnische constructies 1 . 2005. Engrg. Dick. (FEDSM200577353). Matek et al.twi. http://www. Fluent 6.S. Roloﬀ/Matek Machineonderdelen. Mech. third edition. Uasghiri.2 user’s guide. Report number WVT 2006. Smits. Yamada. Academic service. [2] W. ”Proceedings of FEDSM2005”. Een multigrid versnelling van de simple ode voor incompressible stromingen.html). Vierendeels. Lecture notes. . ”ASME Paper 97GT326”. ”Bull. 2001. pages 3. [6] J. Methods Appl. Prasad and V.gov/bridge/servlets/purl/6327934p2OLa/webviewable/ Report number: UCRLID128886. JSME”. 1962.J. Mc GrawHill. [11] Y.nl/nw/users/vuik/numanal/uasghiri. third edition.tudelft. [5] B. Comput. 5(1):302–310.99–3. ALE calculations of ﬂow through rotary positive displacement machines.V. Development of a laplacianbased mesh generator for ALEcalculations in rotary voluemtric pumps and compressors. Computational and experimental investigations of straight . methDelft [8] Bart van Esch and Erik van Kemenade.22. Vande Voorde and J.121.4b660.S. [10] J. Resistance of ﬂow through an annulus with an inner rotating cylinder. 1997. 2004.”. Grandy. March 2005. (193):4401–4415.M.osti. Vande Voorde. ”J. Pump handbook. Sethu Manavalan.S. 2000. Modeling of a ﬂuid ﬂow in an internal combustion engine. J.
see ﬁgure A.1: Crosssection twin screw pump The proﬁle A is an involute proﬁle (x and y) and described by the following equations: s = αr xc = r cos α yc = r sin α x = xc + s sin α y = yc − s cos α (A. In this ﬁgure proﬁle A and B are visible.4) (A. with ri . In ﬁgure A.3. The variables of proﬁle B are deﬁned in Figure A.6:A. The proﬁle B is described in polar coordinates θ and R by equations (A.5) where r ≤ ri .Appendix A Theoretical Screw proﬁle The screw proﬁle is described with a crosssection normal to the axial direction of the twin screw pump. r is equal to ri .1. xc and yc are construction points on the circle with radius r. B A B A Figure A.10).6) .1. the screw inner diameter. a = ro cos α (A. The other variables of proﬁle A are deﬁned in Figure A.2.2) (A.1) (A.3) (A.
9) (h/b) (A. y α Figure A.8) (A.10) θ = π + α + tan ro h α R b θ a y Figure A.3: Variables for proﬁle B . yc α r s Proﬁle A x.2: Variables for proﬁle A h = ro sin α R = b = y−a b2 + h2 −1 (A.7) (A.46 x c .
Appendix B Screw proﬁle Houttuin .
48 .
Maarten van Beijnum .Acknowledgements First of all I would like to thank my supervisor Bart van Esch for his help and guidance in completing my graduation project. Furthermore I like to thank the consulting company Bunova Development BV in Zwolle for creating this graduation report. Especially I would like to thank Timco Visser and Jo`n Teerling for answering practical questions regarding the a numerical model. Finally I would like to thank my roommates and my colleague graduate students for the pleasant and stimulating working atmosphere.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.