Energy Conversion and Management 86 (2014) 1091–1101

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Effects of groove shape of notch on the flow characteristics of spool valve

Yi Ye, Chen-Bo Yin ⇑, Xing-Dong Li, Wei-jin Zhou, Feng-feng Yuan
Institute of Automobile and Construction Machinery, Nanjing Tech University, Nanjing 211816, China

a r t i c l e i n f o a b s t r a c t

Article history: The grooves of notches of hydraulic spool valves are usually designed into various shapes for their desired
Received 23 February 2014 flow characteristics. The aim of this paper is to clarify the effects of the groove shape on the flow char-
Accepted 26 June 2014 acteristics through computational fluid dynamics (CFD) and experimental investigations. The RNG k–e
Available online 15 July 2014
turbulence model is used to simulate the pressure distributions of the flow fields inside three notches
with their corresponding typical structural grooves in order to analyze the changes of restricted locations
Keywords: along with the openings and, furthermore, to calculate the flow areas of the notches. The accuracy of the
Flow characteristics
employed model is demonstrated by comparing the computational results with the experimental data.
Spool valve
Notch
Additionally, the flow rate vs. pressure drop data obtained from the experiment is fitted by least square
Groove shape method. On this basis, the discharge coefficient as a function of groove geometry, flow condition, fitting
Computational fluid dynamics (CFD) coefficients and its stable value is deduced, proving to be quite consistent with the experimental result.
Thanks to the jet flow angles estimated by CFD simulation, the steady flow forces are calculated, which
show good agreement with the experimental results except for some small differences. Finally, the throt-
tling stiffness of the three notches is investigated, with that of divergent U-shape groove falls between
spheroid-shape groove and triangle-shape groove. Similar results are found for steady flow force. The
results indicate that the groove shape has significant effects on the flow characteristics (flow area, dis-
charge characteristic, jet flow angle, steady flow force and throttling stiffness) of spool valve.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction characteristics of servo-valve spool valve under both laminar and

turbulent flow by the CFD method and derived a formula for dis-
Work performance of a hydraulic system is largely determined charge coefficient and Reynolds number. Posa et al. [4] carried
by the hydraulic valve. And fluid flow control of the hydraulic valve out an analysis of the discharge coefficient and the flow force of
essentially depends on the throttling form of notches. Notches a directional valve utilizing 2D CFD method and, in addition, the
with throttling grooves are widely used in high performance fluid-body interaction had been represented by an immersed-
hydraulic valves where excellent precision and stability are boundary technique. Amirante et al. [5] performed a complete 3D
required due to the following advantages, such as diverse struc- numerical simulation to deal with the driving forces acting on an
ture, wide range of flow rate, the gradient of flow area being easy open center directional control valve. Yuan and Li [6] highlighted
to control, good stability in the case of small flow, abundant flow the important effects of two often ignored components, viscosity
control characteristics. Some scholars have launched a number of effect and non-metering momentum flux, on the steady flow force
pioneering researches to study the relationship between the valve for the purpose of achieving outstanding dynamics performances.
geometry and its flow characteristics. All the above researches have demonstrated that the CFD method
Lisowski et al. [1] proposed a logic type directional control valve has advantages to predict flow characteristics of valve and is effi-
with reduction of pressure losses over 35%, while the difference cient in capturing many partial details in flow field. Moreover, an
taken from CFD analysis and experimental research does not experimental procedure had been proposed to determine the dis-
exceed 5%. They [2] also proposed the calculation of flow forces charge coefficient of a spool valve as a function of opening and
(pressure force and viscous force) acting on the spool of solenoid pressure drop in Ref. [7]. Ji et al. [8] developed a pressure distribu-
operated directional control valve using 3D CFD modeling along tion and noise measurement test equipment of hydraulic valve
with k–e turbulence model. Pan et al. [3] analyzed the discharge with a mobile valve sleeve, and discussed the interior relations
among the throttling groove structure, pressure distribution and
noise characteristics. Furthermore, relevant works with reference
⇑ Corresponding author. Tel./fax: +86 025 58139599. to flow area, flow coefficient and steady flow force of spool valve
E-mail address: yinchenbo@njtech.edu.cn (C.-B. Yin).

http://dx.doi.org/10.1016/j.enconman.2014.06.081
0196-8904/Ó 2014 Elsevier Ltd. All rights reserved.
1092 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101

with notches were experimentally investigated in other studies of the axial direction are set to be equal (3 mm). To reduce the impact
Ji et al. [9–11]. Some other notable works aimed at flow character- of radial imbalance force on the experimental results, two throt-
istics include studies on flow test of ball valve by He et al. [12], tling grooves are symmetrically processed in the circumferential
dynamic analysis of a two-stage solenoid valve flow by Ye et al. direction on the spool.
[13], flow structure inside a spool type pressure regulating valve
by Chattopadhyay et al. [14] and visualization experiment of flow 2.2. Modeling and simulation
patterns and cavitation phenomena by Chern et al. [15]. In general,
these researches were focused on global performances of valves. Control of fluid flow by notches is accomplished through regu-
Unfortunately, the influences of valve geometry on flow character- lating the opening so as to change the flow area. Therefore, deter-
istics were not highlighted. mining the relationship between the opening and flow area is the
Some investigations proposed functional structures achieved foundation of carrying out the flow characteristics investigation.
through modifying or changing the geometry of valve to improve Generally, several cross-sections in the notch contribute to the
flow force balance, anti-cavitation ability, flow capacity, etc. Amir- throttling effect, resulting in complex flow structure and pressure
ante et al. [16] numerically studied the effect of optimized groove distribution of the fluid field inside the notch. For single-stage
and compensation profile on reducing the flow force of a propor- throttling groove in the previous studies [9,10,23,24], throttling
tional valve, and also provided the corresponding experimental effect was usually attributed to axial, radial or minimum cross-sec-
verification. Borghi et al. [17] presented three different valve port tion across throttling edge (defined as throttling cross-section
connections which were used to provide compensation to station- thereof). But in fact, as the opening changes, the throttling cross-
ary axial flow forces of a reference spool valve, and the compensa- section varies dynamically. According to the Bernoulli law, flow
tion effect was verified by experiment. Many other innovative velocity will rise sharply along with a rapid pressure drop when
designs were also presented, for instance, arc notches in spool land fluid flow through the throttling cross-section. In other words, a
and sleeve [18], additional parallel and compensatory channels in a rapid pressure drop only happens in the throttling cross-section
solenoid operated directional control valve [2], tapered structure of the notch. Thus, in order to obtain the accurate location of the
on spool rod [19] and rotated channel in non-metering port [20]. throttling cross-section, CFD numerical simulation method is
And positive effects of the novel structures in the above articles applied to analyze the pressure distribution inside the notch.
were also shown. The grid model is established using half part of the flow field
However, limited work [21–23] has been carried out to study inside the valve for it is an axis symmetrical structure. As shown
the influence of groove shape on flow characteristics although in Fig. 2 (take spheroid-shape groove in X = 0.6 mm as an example),
groove shape, to a great extent, may determine the observable it is divided by the application of tetrahedral mesh which contains
behavior of valve. Moreover, the effects of groove shape on flow about 450,000 cells. Grid models for triangle-shape groove and
characteristics still lack systematic research. divergent U-shape groove cases contain about 435,000 and
By investigating the effects of groove shape on flow characteris- 455,000 cells, respectively. What’s more, it can be seen from
tics of spool valve from multiple perspectives, this study fills up the Fig. 2 that the grid gradually becomes intensive in the process of
gap of previous studies by proposing three types of throttling getting close to the wall and the inlet and outlet boundaries of
grooves with different structural features, namely spheroid-shape the notch. The RNG k–e turbulence model is selected, and boundary
groove, triangle-shape groove and divergent U-shape groove. Fol- conditions of inlet and outlet pressures are imposed. In addition,
lowing a brief description of groove structures, the pressure distri- solution accuracy is set to be 105, and hydraulic oil with viscosity
butions of the flow fields inside the notches are numerically of 46 cSt and density of 896 kg/m3 is chosen as the medium. Grid-
simulated so as to analyze the changes of restricted locations along independent test of the model is carried out with various grid sizes
with the increasing spool opening and to calculate the equivalent to make sure the obtained numerical solutions are independent on
notch area. Then, a least square method is utilized to fit the test grid density. Five positions (X = 0.6, 1.4, 2.0, 2.8, 3.1 mm) during
data of flow rate vs. pressure drop. The discharge coefficient as a opening process are selected to be analyzed.
function of groove geometry, flow condition, fitting coefficients
and its stable value is deduced, and stable values of the discharge 2.3. Pressure distribution
coefficient are obtained and discussed. After that, the steady flow
force is calculated as the jet flow angle is estimated by CFD numer- The pressure distributions in the symmetrical surface of the
ical simulation. Finally, the throttling stiffness of the notch is also notches at the corresponding positions are shown in Fig. 3. The
investigated. The results provide effective guidance for the design pressure drop through the spheroid-shape groove is concentrated
of high-performance hydraulic valve. on cross-sections A2 and A3 within the entire range of the spool
stroke. In other words, the throttling effect of the spheroid-shape
2. CFD analysis groove is provided by cross-sections A2 and A3 together, showing
two-level throttling characteristics. In particular, with the increase
2.1. Groove structures of the opening, the proportion of the pressure drop changes grad-
ually from cross-section A2 to A3.
The three types of throttling grooves with typical structural fea- The triangle-shape groove is a typical divergent structure,
tures are shown in Fig. 1. Fig. 1a is a spheroid-shape groove and whose pressure distribution characteristics are relatively simple.
Fig. 1b is a triangle-shape groove. Fig. 1c is a divergent U-shape In the triangle-shape groove, cross-sections A1 and A2 do not have
groove with a relatively complex structure, which could be consid- throttling function. The pressure drop is consistently centralized in
ered as being overlaid in parallel by a U-shape part and a divergent cross-section Amin.
one. In Fig. 1, X is the spool opening; A1 and A2 are the axial and The divergent U-shape groove has a more complex behavior of
radial cross-sections respectively, and both cross the throttling the pressure distribution than the other two types. Migration of
edge. A3 is the cross-section which crosses both the throttling edge throttling cross-sections exists in this type as shown in Fig. 3. At
and the lowest point of the groove. Amin is the smallest cross-sec- small openings, pressure drop is consistently centralized in cross-
tion across the throttling edge and which is usually perpendicular sections A1 and A2. In addition, even though cross-section A2 bears
to the bottom. To demonstrate the effects of the structural features more pressure drop, two-level throttling characteristics dominate
on the flow characteristics, the lengths of the throttling grooves in in the pressure distribution. With the increasing opening, the
 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101 1093

Fig. 1. Geometric structures of: (a) the spheroid-shape groove; (b) the triangle-shape groove and (c) the divergent U-shape groove.

cross-section A2 and inclined cross-section A3 caused by simplified

treatment are both verified to be less than 4%, demonstrating its
feasibility. MATLAB is used to develop the flow area calculation
procedure of the notch. When the flow area curves are drawn, A/
Amax and X/Xmax are respectively set as ordinate and abscissa as
shown in Fig. 4. In addition, A1, A2, A3 are the area sizes of A1, A2,
A3, respectively.
As shown in Fig. 4a, Aeq is the minimum and should be regarded
as the notch flow area value of the spheroid-shape groove during
the entire opening process. In the first half of the spool stroke,
Aeq is almost equal to A2 for A2 is much smaller than A3. However,
A2 and A3 contribute more equally to Aeq during the second half of
the spool stroke as the ratio of A3 to A2 gradually decreases to 1. For
the triangle-shape groove (Fig. 4b), the ratio of A1 to A2 remains
constant. In this case, the notch flow area is set to the value of
Aeq, which is also equal to Amin. With regard to the divergent U-
shape groove, there is an intersection point in the spool opening
at X = 2.0 mm which is approximately 0.667 in term of X/Xmax from
Fig. 4c. When X is smaller than 2.0 mm, the notch flow area equals
Aeq. When X is larger than 2.0 mm, it changes to the minimum
throttling cross-sectional area Amin.

Fig. 2. Grid model for the spheroid-shape groove case. 3. Experimental setup
throttling effect of cross-section A2 declines gradually. Meanwhile,
A sketch of the notch flow characteristics experimental setup is
the pressure loss at cross-section A1 migrates to cross-section Amin.
presented in Fig. 5. The inlet and outlet pressures of the notch are
The migration process is still characterized by the two-level throt-
respectively adjusted by a variable pump (range 0–150 l/min, class
tling. Later in the remaining opening process, as the area of cross-
0.2 absolute) and a counterbalance valve. The maximum pressure
section A2 continually increases, the throttling effect of cross-sec-
of the system is set to be 30 MPa by a safety valve. The mechanical
tion A2 decreases until it disappears. At the same time, single-level
oscillation of the testing apparatus is effectively suppressed by the
throttling feature appears and Amin becomes the centralized cross-
hose and filter, and so is the interference caused by the pulsation of
section for the total pressure drop.
the fluid flow. The movement of the valve spool is adjusted by a
ball screw which is driven by a stepping motor. Two pressure sen-
2.4. Flow area analysis sors (range 0–60 MPa, class 0.2 absolute) are arranged on the valve
chamber. The rate of flow through the tested notch is measured by
Currently, in the study of flow area of the notch with throttling a flow meter (range 0–150 l/min, class 0.1 absolute). The axial force
groove, an effective approach is applied by replacing multiple acting on the spool is measured by a force sensor (range 0–150 N,
throttling cross-sections with an equivalent throttling cross-sec- class 0.1 absolute). A compensation method [25] is used to test the
tion. In present paper, the minimum value between Amin and Aeq steady flow force. Also, hydraulic oil with viscosity of 46 cSt and
is taken as the notch flow area where Amin and Aeq are the area sizes density of 896 kg/m3 is selected as the working fluid in the test.
of Amin and the equivalent throttling cross-section, respectively. Considering that the test result is greatly influenced by the form
Furthermore, Aeq is set to the value of the equivalent area of A2 error of the spool structure (such as the circular bead of the groove)
and A3 in series in the spheroid-shape groove and A1 and A2 in ser- when the spool opening is at the limits, the opening range of
ies in the other two types according to the analysis of pressure dis- 0.4 mm to 2.8 mm is set as the scope of study. What’s more, three
tribution in section 2.3. feature positions of X = 0.6, 1.4, 2.2 mm are selected to be emphat-
In fact, the top surface of the throttling groove is a circular sur- ically analyzed within the range.
face. The boundary lines of the enclosed surface are space curves.
Therefore, obtaining the specific analytical solution of the flow area 4. Results and discussion
by applying numerical integration method or solving nonlinear
function becomes really complicated [24]. As such, the circular sur- 4.1. Discharge characteristics
face is simplified to a plane when the flow area is calculated. While
the spheroid-shape groove has the maximum circumferential span MATLAB is used to deal with the obtained data from the test.
among the three throttling grooves, the area errors of its radial The curves of the flow rate vs. pressure drop when the openings
1094 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101

Fig. 3. Pressure distributions inside the notches at positions: (a) X = 0.6 mm; (b) X = 1.4 mm; (c) X = 2.0 mm; (d) X = 2.8 mm and (e) X = 3.1 mm.

of the three notches are at the feature positions are shown in Fig. 6. followed by the divergent U-shape groove and then the triangle-
It is worth noting that the scale of the horizontal axis of the flow shape groove under the same pressure drop. In addition, the corre-
rate is different in different plots. From Fig. 6, it is obvious that sponding curves of CFD simulation results are also provided in
the spheroid-shape groove has the strongest flow capacity, Fig. 6. A comparison between the experimental and the CFD
 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101 1095

Fig. 5. Sketch of the experimental setup: (1) variable pump; (2) counterbalance
valve; (3) safety valve; (4) hose; (5) filter; (6) stepping motor; (7) ball screw; (8)
valve spool; (9) pressure sensor; (10) flow meter; (11) slid platform and (12)
pedestal.

Dp ¼ a1 Q 2 þ a2 Q ð1Þ
The values of coefficients a1, a2 are presented in Table 1. All the
values of the fitting regression coefficients R are higher than 0.994,
indicating the fitted curves of the flow rate vs. pressure drop are in
good agreement with the experimental data (Fig. 6).
The fluid flowing through the throttling groove can be described
by the orifice flow formula:
sffiffiffiffiffiffiffiffiffi
2Dp
Q ¼ Cd A ð2Þ
q
where Cd is the discharge coefficient, A is the notch flow area, q is
the density of the fluid.
As shown in Eq. (2), the discharge characteristic of the notch is
determined by parameters Cd, A, q and Dp. Cd and A are inherent
properties of the notch, which are only related to the throttling
groove shape. The notch flow area A reflects the flow capacity of
the notch. Greater A means stronger flow capacity. Cd reflects the
resistance characteristic of the notch. The greater Cd is, the smaller
the flow resistance of the notch is. When the pressure difference
between the inlet and outlet of the flow field is certain, the rate
of flow through the notch becomes larger as A and Cd increase.
Eq. (2) can be rearranged as
Dp q
¼ ð3Þ
Q 2 2C 2d A2
And likewise, the ratio of Dp to Q2 can be expressed from Eq. (1)
as follows
Dp a2
¼ a1 þ ð4Þ
Q2 Q
Since the fluid medium is selected and the groove shape as well
Fig. 4. Plot of the flow area vs. opening for: (a) the spheroid-shape groove; (b) the
triangle-shape groove and (c) the divergent U-shape groove. as the opening is certain, the corresponding parameters q, A can be
determined. And a1, a2 are known from Table 1. Combining Eqs. (3)
and (4), the discharge coefficient Cd as a function of Q can be
simulation results clearly indicates that, in each case the depen- expressed as
dence of the flow rate on the pressure drop is not exactly the same rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 q
yet very close. The differences between the test results and their Cd ¼ ð5Þ
A 2ða1 þ a2 =Q Þ
corresponding simulation values do not exceed 7%.
The distribution of the flow rate vs. pressure drop data con- When increasing flow rate Q continually to infinity, the dis-
forms to a quadratic curve with its vertex being exactly the original charge coefficient will gradually approach a stable value (Cdst):
point. Therefore, curve fitting of the data obtained from the exper- rffiffiffiffiffiffiffiffi
1 q
iment is conducted by the least square method. The fitting relation C dst ¼ lim C d ¼ ð6Þ
Q!1 A 2a1
of the pressure drop Dp and the flow rate Q can be expressed as:
1096 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101

Table 1
Polynomial coefficients for data interpolation.

Notch structure Opening a1 (MPa/(l/ a2 (MPa/(l/

(mm) min)2) min))
Spheroid-shape groove 0.6 2.573e2 7.299e03
1.4 2.647e3 7.397e04
2.2 9.170e4 1.700e03
Triangle-shape groove 0.6 2.083e + 1 8.606e01
1.4 7.605e1 9.459e02
2.2 1.347e1 2.654e02
Divergent U-shape 0.6 1.168e1 3.962e03
groove 1.4 2.103e2 5.121e03
2.2 1.058e2 2.785e03

Table 2
Stable value of the discharge coefficient.

Notch structure Opening (mm) Cdst

Spheroid-shape groove 0.6 0.747
1.4 0.682
2.2 0.620
Triangle-shape groove 0.6 0.720
1.4 0.692
2.2 0.666
Divergent U-shape groove 0.6 0.651
1.4 0.646
2.2 0.652

The relation between the discharge coefficient and the flow rate
is described by Eq. (7). However, it cannot tell the change of the
discharge coefficient with the flow condition. Therefore, Reynolds
number (Re) of the fluid flowing through the notch is introduced
to characterize the flow condition and defined as:
mqDh
Re ¼ ð8Þ
l
where m is the flow velocity, which can be expressed as m = Q/2Amin,
and l is the dynamic viscosity of the hydraulic oil. Dh is the hydrau-
lic diameter, which is defined as four times the ratio of wet cross-
sectional area to wet perimeter length. In this study, the flow
regime of the flow field inside the notch is determined by the max-
imum flow velocity which occurs in the cross-section with the min-
imum area. Therefore, to calculate the Reynolds number the values
of the wet cross-sectional area and wet perimeter length are got in
cross-section Amin. Thus, Dh is expressed as
4Amin
Dh ¼ ð9Þ
C
where C is the perimeter length of Amin. Combining Eqs. (8) and (9),
the relation between the flow rate and the Reynolds number can be
obtained:
lCRe
Q¼ ð10Þ
Fig. 6. Curves of the flow rate vs. pressure drop of: (a) the spheroid-shape groove; 2q
(b) the triangle-shape groove and (c) the divergent U-shape groove.
Substitute Eq. (10) into Eq. (7), and after arranging we have
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
C d ¼ C dst ð11Þ
In this case, the fluid flow through the notch is fully turbulent. 1 þ 2qa2 =a1 lCRe
The stable values of the discharge coefficient at the feature posi-
Therefore, the discharge coefficient is a function of groove
tions of the three notches are calculated using Eq. (6) and the
geometry, flow condition, fitting coefficients and its stable value.
results are shown in Table 2.
Substitute the values of Cdst, a1, a2, l, q and C into Eq. (11), plots
Put Cdst into Eq. (5), it becomes:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi of the dependences of the discharge coefficients on the Reynolds
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi number are obtained for different notches at the feature openings
1 q 1
Cd ¼ ¼ C dst ð7Þ as shown in Fig. 7. Re0.5 is set as the abscissa and its scale is differ-
A 2a1 ð1 þ a2 =a1 QÞ 1 þ a2 =a1 Q
ent in different plots. After data processing of the test results using
 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101 1097

orifice flow formula directly, we can see that the calculated results
are in good agreement with the experimental ones. It is worth not-
ing that in all samples the discharge coefficients first increase rap-
idly with regard to Reynolds number in the laminar flow region,
and then gradually achieve the stable values around the transi-
tional zones which range from 10 to 30 in terms of Re0.5. As the
throttling groove is narrow, the fluid flowing through the notch
is quite fast and seriously constrained by its wall, leading to a con-
stant change of flow direction. In this case, laminar flow is easy to
be damaged as viscous forces between liquid particles weaken.
Consequently, inertia force of the liquid plays a dominant role,
resulting in a transitional zone in advance. With the increasing
opening, the discharge coefficient of the spheroid-shape groove
reaches the stable value Cdst more laggingly, while the triangle-
shape groove and the divergent U-shape groove show no obvious
change.
The relation between the stable value of the discharge coeffi-
cient and the opening is significant for designing hydraulic valves
with diverse flow control functions. Within the opening range of
0.4–2.8 mm, the discharge characteristic test is performed every
0.2 mm. The pressure difference between the inlet and the outlet
of the notch is set large enough to obtain the stable value of the
discharge coefficient in each test.
Substitute the test data and the notch flow area into Eq. (2), we
obtain the discharge coefficient distribution with the opening as
shown in Fig. 8.
From the trend of the curve, the impact of the opening on the
discharge coefficient could be summarized as follows:

(1) The discharge coefficient of the spheroid-shape groove has

an approximately negative linear relationship with the spool
opening. At small openings, A3  A2, radial cross-section A2
plays the main throttling role. And the flow resistance of
the throttling groove is small, meaning a relatively large
Cd. With the increase of the opening, A2 increases rapidly.
The area ratio of A3/A2 gradually decreases to 1 as the two-
level throttling feature becomes obvious and Cd decreases
continuously.
(2) When the spool opening X < 0.8 mm, the discharge coeffi-
cient of the triangle-shape groove increases quickly from
0.639 to 0.71 as a result of clogging which often occurs in
fluid flow through the notch under the condition of small
flow area and large pressure difference. On the contrary, it
keeps decreasing when X > 1 mm. This phenomenon is
mainly attributed to the occurrence of cavitation in the par-
tial low-pressure zones of the throttling cross-section, which
has been observed inside a similar notch with V-shape
groove by visualization experiment [21]. The occurrence of
cavitation decreases the actual flow area of the notch. The
influence becomes more significant as the opening increases
under a constant pressure drop which leads to a continuous
decline of the discharge coefficient.
(3) As for the divergent U-shape groove, when X < 1 mm, the
discharge coefficient drops quickly from 0.683 to 0.587. As
X > 1 mm, the discharge coefficient increases gradually.
When X = 1.8 mm, Cd increases to 0.648 and remains almost
unchanged thereafter. According to the flow area character-
istics, in the initial opening process, the pressure drop is
mostly concentrated in the cross-section A2. And the flow Fig. 7. Plots of Cd vs. Re0.5 for: (a) the spheroid-shape groove; (b) the triangle-shape
resistance of the throttling groove is mainly caused by A2, groove and (c) the divergent U-shape groove.
showing single-level throttling characteristic. With the
increase of the opening, A2 increases drastically. At approxi- value. Thereafter, A2 continues to increase drastically. As
mately X = 1 mm, the total pressure drop distributes equally the difference of A1 and A2 becomes larger, the two-level
in the cross-sections A1 and A2 as A1/A2 approaches 1, result- throttling characteristic weakens. When the opening
ing in the maximum throttling effect during the entire open- exceeds 2 mm, the throttling effect is solely caused by Amin,
ing process. As a consequence, Cd decreases to its lowest showing single-level throttling characteristic.
1098 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101

Fig. 9. Schematic diagram of fluid flow through the notch.

As the area of the outlet is considerably larger than that of the

inlet, the outlet velocity v2 is much smaller than the inlet velocity
v1. Furthermore, the downstream flow is constrained by the spool
wall with the outlet flow angle h2 being close to 90°. As a conse-
quence, the outlet efflux force, namely the second term on the right
Fig. 8. Discharge coefficient distribution with the opening. side of Eq. (14) can be ignored. According to the literature [18,20],
the viscous force can be neglected because it is too small compar-
Based on the above analysis, we can see that the three throttling ing with the inlet efflux force (the first term on the right side of Eq.
grooves can be applied to the hydraulic system for different pur- (14)), i.e., Fbody = 0. It’s worth noting that the inlet velocity v1 is not
poses according to their corresponding discharge characteristics. the jet flow velocity at the notch, but the velocity on the inlet sur-
At small openings, the discharge coefficient of the spheroid-shape face (i.e. A2) of the control volume. By substituting Eq. (2) into Eq.
groove is great, which is good at quickly building up the working (14), steady flow force can be further expressed as:
pressure for the applied hydraulic system. The discharge coeffi-
2C 2d A2 cot h1
cient of the triangle-shape groove decreases gently in relation to Ff ¼ Dp ð15Þ
the opening after its initial rise, which is beneficial to fine tuning A2
of the flow rate and smooth running of the hydraulic system. As
to the divergent U-shape groove, the discharge coefficient is big Table 3
Jet flow angles at variable openings.
when the opening is small and followed by a rapid decline, which
is advantage to avoiding the vibration occurred during startup pro- Opening Jet flow angle (°)
cess of the hydraulic system. (mm)
Spheroid-shape Triangle-shape Divergent U-shape
groove groove groove
4.2. Steady flow force 0.4 78.5 44.6 67.0
0.6 74.8 45.2 66.4
Flow force is one of the key factors that determine the pros or 0.8 72.0 44.2 66.1
cons of the hydraulic control valve performance. It not only affects 1.0 70.3 43.9 65.8
1.2 69.1 43.8 65.3
the operating force on the spool, but also causes the self-excited 1.4 67.9 42.5 64.9
vibration of the components of the hydraulic system. Accurate cal- 1.6 67.2 43.4 64.1
culation of the flow force is the prerequisite for control accuracy 1.8 66.6 44.4 63.4
and force balance of the spool. As shown in Fig. 9, the control vol- 2.0 66.2 44.5 63.0
2.2 65.7 44.2 62.7
ume under flow force is the region surrounded by spool wall, notch
2.4 65.3 43.9 62.5
horizontal surface and valve body underside. According to the 2.6 64.9 44.5 62.2
momentum conservation law, the axial force balance equation 2.8 64.6 43.6 61.1
for the control volume can be expressed in vector form (positive
to the right) as follows:
~
F land þ ~
F body þ ~ v1  ~
F rod þ qQ ð~ v2Þ ¼ 0 ð12Þ
~
F land is pressure force vector subjected by the spool lands; ~ F body is
viscous force vector subjected by the valve body; ~ F rod is viscous
force vector subjected by the spool rod; ~ v 1 and ~ v 2 are inlet velocity
vector and outlet velocity vector, respectively. The direction of ~ F body
and ~ F rod are toward the left as the fluid flow to the right, which can
be judged easily from Fig. 9.
The steady flow force ~ F f that the spool experiences from fluid
flow is equal to the sum of pressure forces ~ F land0 acting on the spool
lands and viscous force ~ F rod0 acting on the spool rod, which is given
in vector form as:
~
Ff ¼ ~
F land0 þ ~
F rod0 ð13Þ
~
F land and ~
F land0 are equal but in the opposite direction, the same as
F rod and ~
~ F rod0 . The above two equations lead to the following scalar
form:
F f ¼ qQ v 1 cos h1  qQ v 2 cosh2  F body ð14Þ Fig. 10. Steady flow force at variable openings.
 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101 1099

When the opening is certain, Cd, A, A2, h1 are fixed values. in the control volume cannot be observed in the experiment, so
Assuming coefficient Ks = 2Cd2A2coth1/A2 is constant, the steady CFD simulation is used to determine the jet flow angle h1. Table
flow force Ff is proportional to pressure drop Dp. Flow structure 3 summarizes the jet flow angles of the three throttling grooves

Fig. 11. Contours of pressure and velocity distributions of the divergent U-shape groove at opening: (a) X = 0.6 mm; (b) X = 1.4 mm and (c) X = 2.2 mm.
1100 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101

at different openings. The jet flow angle variation range of the

spheroid-shape groove reaches 14°, with great changes at the ini-
tial openings. In the intermediate positions, it is very close to the
theoretical value of 69° derived by Von Mises. The jet flow angle
of the triangle-shape groove is significantly affected by the flow-
direction guide of the bottom slope and independent of the open-
ing. Throughout the opening, the jet flow angle is small, and main-
tains at about 44°. The jet flow angle of the divergent U-shape
groove decreases roughly linearly from 67° to 61.1°. It is notewor-
thy that for the steady flow force experiment, the pressure differ-
ence between the inlet and the outlet is set to 20 MPa for each
test, with the error range of ±0.5 MPa, and the same pressure dif-
ference is applied for the corresponding calculation.
Fig. 10 shows the comparison between the calculated and
experimental values of the steady flow force at different openings.
It can be seen that theoretical calculations can fully reflect varia-
tion trends of the experimental steady flow forces. The steady flow
forces of the triangle-shape groove and the divergent U-shape
groove increase roughly linearly with the increasing opening. At
very small openings, the steady flow force of the triangle-shape
groove appears to be negative, tending to open the valve. This
appearance may be because of the occurrence of complex vortex
and reflux in the exit section of the control volume, which makes
the axial exit momentum non-ignorable or even larger compared
to the axial entry momentum. At the same time the axial entry
momentum is very tiny as the discharge coefficient and the notch
flow area are both at a low level. Similar phenomenon was also
observed in the experimental investigation of spool valve with U-
shape grooves at small openings [22].
At the same opening, the test values are slightly smaller than
the calculated values for all the three cases. The differences gradu-
ally increase with the increasing openings. A hypothesis has been
made for the calculation of the steady flow force that the pressure
difference between the inlet and the outlet of the flow field is con-
sidered wholly concentrated on the throttling section. This hypoth-
esis actually presents a larger pressure drop through the throttling
section in the calculation than the test, leading to the differences
between the calculated values and the test values. Fig. 11 provides
the velocity and pressure distributions in the symmetrical plane at
opening X = 0.6, 1.4, 2.2 mm of the divergent U-shape groove under
the same pressure drop of 5 MPa. At X = 0.6 mm, the flow velocity
attains its maximum value inside the groove region, which means
the pressure drop is mostly restricted to the throttling section as
shown in Fig. 11a. With the increase of the opening, the region of
the maximum velocity migrates downstream of the control vol-
ume. Correspondingly, the pressure losses dissipated at non-throt-
tling sections become more and more significant with reference to
the main pressure drop through the throttling groove as can be
seen in Fig. 11b and c. Similar results have been obtained for the
other two types (data not shown).
In addition, with the increase of the opening, the outlet area
reduces and the outlet flow angle gradually deviates from 90°due
to diversion by the spool shoulder, both causing a slight increase
of the outlet efflux force. For these reasons, the differences
between the calculated values and the corresponding test values
of the steady flow forces gradually enlarge with the increasing
opening. The same reasons apply to the spheroid-shape groove as
well as the triangle-shape groove. When the notch is close to fully
opened status, the flow resistance of the throttling section virtually
disappears. And the test values of the steady flow force increase
suddenly while the flow rate increases rapidly (Fig. 10).

4.3. Throttling stiffness

Fig. 12. Plot of throttling stiffness with variable pressure drops and openings for:
When the discharge of the notch extremely depends on the (a) the spheroid-shape groove; (b) the triangle-shape groove and (c) the divergent
external load, minor change of pressure difference will still cause U-shape groove.
 Y. Ye et al. / Energy Conversion and Management 86 (2014) 1091–1101 1101

severe discharge oscillation, leading to instability of the flow con- Acknowledgments

trol. This feature can be characterized by throttling stiffness, which
reflects the ability to resist the load change of the notch to main- The authors gratefully acknowledge the support provided by
tain flow stability. It is defined as the first derivative of pressure National Natural Science Foundation of China No. 50875122,
drop with respect to flow rate. The throttling stiffness (T) can be Jiangsu Province Science and Technology Support Program No.
derived from the orifice flow formula (Eq. (2)) as: BE2011187 and Program Granted for Scientific Innovation
Research of College Graduate in Jiangsu Province No.
pffiffiffiffiffiffi CXZZ13_0432.
dDp 2q 0:5
T¼ ¼ Dp ð16Þ
dQ Cd A
References
From the analysis in section 2.2, CdA varies with spool opening
X. Assuming CdA = K(X), the throttling stiffness T can be written as a [1] Lisowski E, Rajda J. CFD analysis of pressure loss during flow by hydraulic
directional control valve constructed from logic valves. Energy Convers
function of X, Dp: Manage 2013;65:285–91. http://dx.doi.org/10.1016/j.enconman.2012.08.015.
[2] Lisowski E, Czyzycki W, Rajda J. Three dimensional CFD analysis and
experimental test of flow force acting on the spool of solenoid operated
pffiffiffiffiffiffi Dp0:5 directional control valve. Energy Convers Manage 2013;70:220–9. http://
TðX; DpÞ ¼ 2q ð17Þ dx.doi.org/10.1016/j.enconman.2013.02.016.
KðXÞ
[3] Pan XD, Wang GL, Lu ZS. Flow field simulation and a flow model of servo-valve
By substituting the test data of pressure drop and opening of the spool valve orifice. Energy Convers Manage 2011;52:3249–56. http://
dx.doi.org/10.1016/j.enconman.2011.05.010.
three notches into Eq. (17), the values of the throttling stiffness are [4] Posa A, Oresta P, Lippolis A. Analysis of a directional hydraulic valve by a Direct
obtained and plot in Fig. 12 with X and Dp being the independent Numerical Simulation using an immersed-boundary method. Energy Convers
variables. Surface fitting is also carried out for visual analysis. Manage 2013;65:497–506. http://dx.doi.org/10.1016/
j.enconman.2012.07.012.
It can be seen from Fig. 12 that the triangle-shape groove has [5] Amirante R, Del Vescovo G, Lippolis A. Evaluation of the flow forces on an open
the maximum stiffness under the same conditions, followed by centre directional control valve by means of a computational fluid dynamic
the divergent U-shape groove and then the spheroid-shape groove. analysis. Energy Convers Manage 2006;47:1748–60. http://dx.doi.org/
10.1016/j.enconman.2005.10.005.
Nevertheless, they have almost the same variation trend with the
[6] Yuan QH, Li PY. Using steady flow force for unstable valve design: modeling
openings and pressure drops. In each case, the greater the pressure and experiments. ASME J Dyn Syst Meas Control 2005;127:451–62.
drop Dp is, the greater the stiffness T is. In addition, a relatively [7] Viall EN, Zhang Q. Determining the discharge coefficient of a spool valve. In:
Proceedings of the American control conference, Chicago; June 2000. p. 3600–
high change rate of the stiffness appears when the pressure drop
04.
is small. As a result, in order to ensure that the notches possess suf- [8] Ji H, Fu X, Yang HY. Measurement on pressure distribution of non-circular
ficient stiffness, generally the minimum pressure drop is applied. opening spool valve. Chin J Mech Eng 2004;40(4):99–102.
When the inlet and outlet pressures of notch are the same, the [9] Ji H, Wang DS, Ryu SH, Fu X. Flow control characteristic of the orifice in spool
valve with notches. Trans Chin Soc Agricult Mach 2009;40(1):198–202.
stiffness decreases rapidly as the spool opening becomes larger. [10] Ji H, Fu X, Yang HY. Study on steady flow force of non-circular opening spool
When the spool is close to fully opened status, the stiffness is near valve. Chin J Mech Eng 2003;39(6):13–7.
saturation and the pressure change will have weak impact on the [11] Ryu SH, Ochiai M, Ueno K, Fu X, Ji H. Analysis of flow force in valve with
notches on spool. In: Proceedings of the sixth international conference on fluid
stiffness. Among the three notches, the triangle-shape groove has power transmission and control, Tallinn; September 2005. p. 435–38.
the maximum stiffness at small openings, and most likely to have [12] He XH, Huang GQ, Yang YS, Li ZY. Experimental research on the flow
stable discharge. characteristics of ball valve orifice. Chin J Mech Eng 2004;40(8):30–3.
[13] Ye QF, Chen JP. Dynamic analysis of a pilot-operated two-stage solenoid valve
used in pneumatic system. Simul Model Pract Theory 2009;17:794–816.
[14] Chattopadhyay H, Kundu A, Saha BK, Gangopadhyay T. Analysis of flow
structure inside a spool type pressure regulating valve. Energy Convers
5. Conclusions Manage 2012;53:196–204. http://dx.doi.org/10.1016/
j.enconman.2011.08.021.
[15] Chern MJ, Wang CC, Ma Ch. Performance test and flow visualization of ball
CFD simulation and experimental investigation of the flow valve. Energy Convers Manage 2007;31:505–12. http://dx.doi.org/10.1016/
characteristics of three different notches with their corresponding j.expthermflusci.2006.04.019.
typical structural grooves are presented. The results show that the [16] Amirante R, Moscatelli PG, Catalano LA. Evaluation of the flow forces on a
direct (single stage) proportional valve by means of a computational fluid
groove shape has significant effects on the flow area, the discharge
dynamic analysis. Energy Convers Manage 2007;48:942–53. http://dx.doi.org/
characteristics, the jet flow angle, the steady flow force and the 10.1016/j.enconman.2006.08.024.
throttling stiffness of the spool valve. The study is starting from [17] Borghi M, Milani M, Paoluzzi R. Stationary axial flow force analysis on
the calculations of the notch flow area based on the analysis of compensated spool valve. Int J Fluid Power 2000;1:17–25.
[18] Yang YS, Semini C, Tsagarakis NG, Caldwell DG, Zhu YQ. Water hydraulics – a
pressure distributions of the flow fields inside the notches. Differ- novel design of spool-type valves for enhanced dynamic performance. In:
ent features reflected in the changes of the restricted locations Proceeding of the 2008 IEEE/ASME international conference on advanced
inside the three notches along with the spool stroke are high- intelligent mechatronics, Xi’an; July 2008. p. 1308–14.
[19] Miller R, Fujii Y, McCallum J, Strumolo G, Tobler W, Pritts C. CFD simulation of
lighted. The stable values of the discharge coefficients are derived steady-state flow forces on spool-type hydraulic valves. SAE Technical Paper
from the quadratic polynomial fitting of the experimental data of Series; 1999. p. 295–307.
pressure drop vs. flow rate. The transitional zones of the fluid flow [20] Krishnaswamy K, Li PY. On using unstable electrohydraulic valves for control.
ASME J Dyn Syst Meas Control 2002;124:183–90.
through the notches range from 10 to 30 in terms of Re0.5, which [21] Du XW, Zou J, Fu X, Ji H, Yang HY. Effect of throttling grooves structure on
show up earlier than the classic pipe flow. The steady flow forces cavitation noise. J Zhejiang Univ (Eng Sci) 2007;41(3):456–65.
with respect to the opening are calculated and are in good agree- [22] Fang WM, Cheng LL, Fu X, Zhou H, Yang HY. Investigation on steady-state flow
force of spool valve with U-grooves. J Zhejiang Univ (Eng Sci)
ment with the experimental results except for some small differ- 2010;44(3):576–80.
ences. Among the three types, the triangle-shape groove has the [23] Borghi M, Milani M, Paoluzzi R. Influence of notches shape and number of
maximum throttling stiffness, followed by the divergent U-shape notches on the metering characteristics of hydraulic spool valves. Int J Fluid
Power 2005;3:5–18.
groove and then the spheroid-shape groove. The above results
[24] Chen JS, Liu XH, Yuan WR, Wang TJ, Zhao F. Dynamic characteristics of typical
demonstrate that the flow characteristics of the spool valve depend hydraulic notches. J Southwest Jiaotong Univ 2012;47(2):325–32.
strongly on the groove shape and provide effective guidance for the [25] Zhou S, Xu B, Yang HY. Flow force compensation of high speed on/off valve.
design of hydraulic valve. Chin J Mech Eng 2006;42:5–8.