FLOW-FORCE COMPENSATION IN A HYDRAULIC VALVE

Jan Lugowski
Purdue University
West Lafayette, IN, USA
lugowskj@purdue.edu

ABSTRACT
Flow-reaction forces acting in hydraulic valves have been studied for many decades. Despite this, they are difficult
to account for due to the complexities of the jet flow. This paper focuses only on the reduction, also referred to as
compensation, of the flow force as applied to a valve spool featuring a profile of a turbine bucket. Fluid power
textbooks explain the compensation taking place on such a profile by applying Newton’s laws of motion to the profile
and deliver an equation for the magnitude and the direction of the flow force. This paper shows that both the magnitude
and the direction of the compensating flow force are incorrect if calculated from the textbook equation. A corrected
analysis of the dynamic forces is presented that are in agreement with earlier experiments by this author. It follows
that the compensating flow force should be calculated from the static-pressure imbalance on the spool profile. That is,
not Newton’s but Pascal’s law should be applied to calculate the compensating flow force.

INTRODUCTION
A hydraulic valve often controls a fluid flow through it by restricting its flow area. This restriction results in a jet
downstream of the orifice and the accompanying force, known as flow induced force, hydraulic reaction force, or
Bernoulli force [1]. The flow force is a design problem if it is acting on a moving part of the valve and is large enough
to affect the operation of the valve. Pilot-operated valves, featuring large actuating force, have been a practical solution
to the flow-force problem. But even for servo systems, it is important to know the axial forces necessary to operate
the valve piston [2]. The need to understand the flow force is even greater if the valve is to be operated directly, without
the pilot stage.

In their paper that first explained and accounted for the flow force, Lee and Blackburn [2] offered a design solution
to reduce, or compensate, the flow force by shaping the valve piston, or spool, like a turbine bucket. They also provided
a formula to compensate the magnitude of the compensated flow force. The formula and the theory behind it, based
on Newton mechanics, has been later reprinted and expanded in textbooks, such as by Blackburn et al [3], Merritt [1],
and Guillon [4].

To the best of this author’s knowledge, no publication by other researchers can be found that has put the textbook
theory in question. To be fair, the geometry of a turbine-bucket profile is more complex than that of a rocket engine
or a water-turbine bucket. No literature could be found on how to calculate the flow reaction force on a turbine bucket
of a valve spool. The profile on the spool consists of two adjacent turbine buckets. The fluid mechanics textbooks
limit their analysis to only one turbine bucket.

This author attempted to compensate the flow force by shaping the spool as a turbine bucket and found
experimentally that the flow force is reduced more if the entry angle of the profile is smaller (between 25° and 30°),
much less than the suggested 69°. The experiments with various profile angles also revealed that the exit angle of the
turbine-bucket profile did not contribute to the reduction of the flow force, contrary to the textbook suggestions. More
experiments followed that showed that the compensating force acts on the spool profile upstream from the orifice, on
the opposite end of its supposed location [5].

This paper explains how the flow force should be calculated by following Newton mechanics and attempts to
clarify where and how errors have been made by the authors of the textbook theory. The errors are very difficult to
recognize due to the simplicity of the momentum theory and the convincing explanation by its authors. Also, some
textbooks mention the complexities of the jet flow and admit that the actual flow forces may differ from the calculated

1
ones. It also does not help to see a problem when one deals with a well-established textbook theory. Still, the errors
are big, rendering the theory useless and misleading. The actual forces on the turbine bucket would act in the opposite
directions and have quite different magnitude.

CONTROL VOLUME ANALYSIS OF JET FLOW

In fluid mechanics, control volume (CV) is an arbitrary volume through which fluid flows. Control volume
method is used to make the analysis of the flow more convenient, allowing focusing attention on a volume in space
[6]. Figure 1 shows a jet impinging against a wall with velocity v. If the flow rate is Q and the fluid density is ρ, then
the active force Fy can be calculated as shown from Newton’s laws of motion. The reaction force would have equal
value but opposite direction. In this paper, only active forces will be considered, just as they are in the literature on
flow force in hydraulic valves [1-4]. If the jet is perpendicular to the wall, there is no jet force acting parallel to the
wall, if friction forces are disregarded.

Fy= ρQv Fx= 0

CV

x
y

Fig. 1 Control volume applied to calculate jet force Fy

Even if the jet is not perpendicular to the wall, as in Fig. 2, there will be no active force Fx, parallel to the wall.
This is due to the negligibly small exit velocity v of the jet out of the control volume CV.

Fx= 0
Fy= ρQv cos Θ

CV

x Θ
y

Fig. 2 Force Fy from jet impinging on wall at angle Θ

When the jet is perpendicular to the wall and is deflected by 180° after impinging on the wall, while maintaining
its velocity v, the force Fy acting on the wall is twice as large, Fig. 3. There are two accompanying forces Fx that cancel
each other, acting to push the wall profile apart.

2
 Fx= ρQv – ρQv = 0
v

CV

Fy= 2 ρQv

v x
y

Fig. 3 Force Fy from jet reflected from the wall

If we split the control volume into two adjacent ones CV1 and CV2, shown separated in Fig. 4 to help with the
analysis of the forces, we can see why the total force Fy is two times larger compared with Fig. 1. Force Fy1 is generated
by the jet entering CV1, and force Fy2 is generated by the jet exiting CV2. Since both forces act pushing the wall to the
left, the total force Fy is twice as large. The vertical force Fx1 is created by the jet exiting CV1 and pushes the wall
profile down. The other vertical force Fx2 is created by the jet entering CV2 and pushes the wall profile up. The net
vertical force Fx is zero as both forces cancel out. Still, their effect is clear as they are pushing the wall profile apart.
This is important to notice as the textbook theory ignored those forces. Ignoring them means that both vertical forces
are zero and do not push the profile apart. This may not be important for many flow cases, but leads to a serious error
if applied to a turbine-bucket profile on a spool in a hydraulic valve.

CV2

v Fy2 = ρQv
Fx2 = ρQv
Fx1 = ρQv
v

Fy1 = ρQv
v
x
y
CV1

Fig. 4 Splitting control volume helps analyze forces by jet reflected from the wall

If the entry and exit angles are not 90°, as for a turbine-bucket profile, both the horizontal forces Fy1 and Fy2
will be smaller, as shown in Fig. 5. Again, CV1 and CV2 are adjacent to each other and shown apart only to visualize
the forces. Now there are two vertical forces Fx acting on each CV. On CV1, force Fx11 is reduced because the jet angle
is less than 90° and is produced by the jet entering CV1. The opposite vertical force Fx12 is produced by the jest exiting
CV1 and has full strength because it is parallel to the vertical x axis. We assume that the jet maintains its velocity v
along the profile, just like the textbook theory does. The net force Fx1 acting on profile within CV1 is

3
 Fx1= Fx11 - Fx12 = ρQv (cos Θ1 – 1)
(1)

Force Fx1 acts vertically on the entry profile to push it down, as the force is negative. If we canceled both forces
Fx12 and Fx21 because they act in the opposite directions, the net force Fx1 would be positive (ρQv cos Θ1) and had a
different value as well than that per Eq. (1).

Similarly, there are two vertical forces Fx acting on CV2, force Fx22 is reduced because the jet angle is less than
90° and is produced by the jet exiting CV2. The opposite vertical force Fx21 is produced by the jest entering CV2 and
has full strength because it is parallel to the vertical x axis. The net force Fx2 acting on profile within CV2 is

Fx2= Fx21 - Fx22 = ρQv (1 – cos Θ2)

(2)

Force Fx2 acts vertically on the exit profile to push it up, as the force is always positive. Again, if we canceled
both forces Fx12 and Fx21 because they act in the opposite directions, the net force Fx2 would be negative (- ρQv cos
Θ2) and have a different value as well than that per Eq. (2).

The axes in the figures have been rotated by 90° on purpose to help visualize all the forces acting on a wall, and
how to properly apply Newton’s laws of motion to calculate them. The coordinate system will be rotated back when
discussing a spool profile in a hydraulic valve.

Notice that the two horizontal forces Fy add up while the vertical forces Fx subtract from each other. There are
four vertical forces Fx and two of them are of identical magnitude but acting in opposite directions. Those two forces,
Fx12 and Fx21, are only visible when we split the control volume into CV1 and CV2. Fluid power and fluid mechanics
textbooks do not discuss a turbine-bucket profile that is bent more than 90°. If they did, they would need to apply two
control volumes to properly account for all the vertical forces. Failing to do so would result in calculating forces acting
in the opposite directions, and of wrong magnitude.

Θ2
v

CV2
Fy2 = ρQv sin Θ2

Fx22 = ρQv cos Θ2

v
Fx21 = ρQv
Fx12 = ρQv
v
Fx11 = ρQv cos Θ1

Fy1 = ρQv sin Θ1

v
CV1
Θ1 x
y

Fig. 5 Forces by jet reflected from a turbine-bucket profile

4
 Both vertical forces Fx1 and Fx2 are shown in Fig. 6 depicting the turbine-bucket profile, with jet entering it at
angle Θ1 and exiting it at angle Θ2. The net vertical force Fx acting on the turbine-bucket profile is

Fx= Fx1 + Fx2 = ρQv [(cos Θ1 – 1) + (1 – cos Θ2)]

(3)

Now it is easier to see why one would be tempted to cancel both vertical forces Fx12 and Fx21: Equation (3) could
be simplified without changing the magnitude and sign of the net axial force Fx:

Fx= Fx1 + Fx2 = ρQv (cos Θ1 - cos Θ2)

(4)

At first look, this approach seems reasonable. But doing this has serious consequences if Eq. (4) were applied to
select entry and exit angles Θ1 and Θ2 of the turbine-bucket profile. Despite this, the textbooks provide Eq. (4) as
correct. And, following the incorrect Eq. (4), they advise to design the profile featuring a large entry angle Θ1 to reduce
the positive force acting upwards to close the valve. But the actual force Fx1 is negative and acts downwards on the
profile. Per Eq. (1) a larger angle Θ1 produces a larger, not smaller, force Fx1. The bigger problem, though, is that the
correct force acts downwards, not upwards.

Θ2
v
CV

Fx2= ρQv (1 – cos Θ2)

Fy= ρQv (sin Θ1 + sin Θ2)

` `

Fx1= ρQv (cos Θ1 – 1)

v
x
Θ1 y

Fig. 6 Forces by jet reflected from a turbine-bucket profile

Figure 7 shows the same turbine-bucket profile from Fig. 6 but oriented horizontally, as it is customary in the
fluid power literature. The x-axis is parallel to the spool axis. The radial force Fy cancels itself on the spool
circumference and has no effect on the axial force Fx.

5
 Fx2= ρQv (1 – cos Θ2) Fx1= ρQv (cos Θ1 – 1)
Fy= ρQv (sin Θ1 + sin Θ2)

v v

Θ2 Θ1

` `
x
CV y

Fig. 7 Forces on a spool featuring a turbine-bucket profile are pushing the profile apart

Figure 8 shows only axial forces Fx with the incorrect textbook formula, see also Eq. 4, to calculate them. It is
worth mentioning that the textbook formula indicates that the closing, positive, flow force Fx1 is produced on the entry
profile while the opening, negative, flow force Fx2 is produced on the exit profile. This advice is not correct and leads
to confusion if applied to design a spool profile. The biggest problem with the textbook theory is that it ignores the
spreading of the jet due to turbulence and the accompanying loss of jet velocity v. For this reason exit profile cannot
produce the compensating flow force due to the small jet velocity. Even if the jet did maintain its velocity, this exit
profile would produce a positive flow force, not a compensating one. While it is true that the turbine-bucket profile
compensates the flow force, the compensation does not take place on the exit profile located downstream from the
valve orifice but on a very small area of the entry profile located upstream from the valve orifice [5]. This is simply
due to the uncompensated high static pressure pushing on the spool to open the valve. Thus, the textbook formula per
Eq. (4) is even more misleading as the negative compensating force should be calculated by applying Pascal’s, not
Newton’s, law.

Fx2= –ρQv cos Θ2

Fx1= ρQv cos Θ1

v v

Θ2 Θ1
` `

x
CV y

Fig. 8 Forces on a spool featuring a turbine-bucket profile per fluid power textbooks [1-4]

SUMMARY
The analysis of the forces acting on the spool shaped like a turbine bucket points to where the authors of the
textbook theory made errors in their attempt to provide a formula for calculating the flow force on such a profile. The
first error is due to canceling of the two opposite forces which are invisible if one only considers one control volume.
This resulted in wrong directions of the flow forces, and wrong magnitudes. It helps little that the sum of the two
wrong forces is the same as the sum of the correct four forces. As the exit profile produces positive, not the
compensating, flow force, the textbook theory wrongly puts a value on the exit profile to deliver the compensating
force.

6
 The presented analysis agrees with earlier experimental data by this author. The compensating force is produced
on the entry profile upstream from the orifice. The authors of the textbook theory made another mistake by
disregarding the static-pressure balance on the spool profile and focused only on the dynamic forces. Since the pressure
at the inlet of the valve can be very high, even a small area on the spool profile can produce a significant compensating
flow force,

NOMENCLATURE
ρ Mass density of fluid, kg m-3
Θ, Jet angle, deg.
Θ1 Entry jet angle, deg.
Θ2 Exit jet angle, deg.
F Flow force, N
Fx Axial flow force, N
Fx1 Axial flow force to close the valve (positive), N
Fx2 Axial flow force to open the valve (negative, compensating), N
Fy Radial flow force, N
Q Flow rate, m3s-1
v Velocity of jet at vena contracta, m s-1

ACKNOWLEDGMENTS
The author is grateful for support by Purdue University’s Maha Fluid Power Educational Advisory Board and by
Parker Hannifin Corporation, Hydraulic Valve Division.

REFERENCES
[1] Merritt, H.E., 1967, Hydraulic Control Systems, Wiley, New York, pp. 101-108.
[2] Lee, S.-Y., Blackburn, J.F., 1952, “Contributions to Hydraulic Control 1 Steady-State Axial forces on Control-
Valve Pistons,” Trans. ASME, August, pp. 1005-1011.
[3] Blackburn, J. F., Shearer, J. L., and Reethof, G., 1960, Fluid Power Control, Technology Press of M.I.T./Wiley,
New York, pp. 304-313 and 362-363.
[4] Guillon, M., 1969, Hydraulic Servo Systems: Analysis and Design, Plenum Press, New York, pp. 107-121.
[5] Lugowski, J., 1993, “Experimental Investigations on the Origin of Flow Forces in Hydraulic Piston Valves,”
Proceedings of the International Conference on Fluid Power—the Future for Hydraulics,” N. Way, ed., Brugge,
Belgium, Mechanical Engineering Publications Ltd., London, UK, pp. 233-244.
[6] Fox, R.W., Pritchard, P.J., and McDonald, A.T., 2011, Fox and McDonald’s Introduction to Fluid Mechanics,
Hoboken, NJ, Wiley, pp. 7 and 361-365.