TP-7D-1
Worcester Control Valves
Concepts of Vacuum Testing,
Leakage and Pressure Ratings,
Cavitation and Flashing
Seating materials compatible with
almost any media. Temperatures
to 500°F, pressures to 500 psi.
Bottom entry stem for safety.
Simple 1/4 turn handle operation
with visual indication of flow.
Quick disconnect or XBO
pipe ends.
20 Ra standard interior finish.
Available electropolished to
10 Ra finish.
Seat/body seals eliminate potential leak
paths and contamination trap areas.
Parallel bored ball maximizes flow
rates, minimizes pressure drops,
eliminates solids build up.
Three-piece design facilitates installation, allows
interchangeability of ends, and is easily maintained.
316L construction has superior
corrosion resistance and allows
welding and prevents rouging.
Introduction
TABLE OF CONTENTS
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .1
Design of Valves for Vacuum . . . . . . . . . . . .2
Leakage Ratings . . . . . . . . . . . . . . . . . . . . . .5
This paper was originally entitled “Vacuum Testing”, but has been
expanded upon to include the subjects of cavitation and flashing,
leakage rate misconceptions, and various pressure terminology
clarifications.
Therefore, the intent of this paper is to consolidate a lot of this
fragmented information that is scattered throughout various
brochures and technical papers, and to expand on the subject matter
in more detail. Hopefully, the information presented will make things
clear. In addition to vacuum testing, the three areas to be discussed
in more detail are:
Pressure Ratings . . . . . . . . . . . . . . . . . . . . . .7
1. Leakage Ratings
Valve Body Cavity Pressure Relief . . . . . . . . .7
2. Pressure Ratings, Cavity Pressure
Cavitation and Flashing . . . . . . . . . . . . . . . . .8
3. Cavitation and Flashing
Flow Control Division
Worcester Control Valves
DESIGN OF VALVES FOR VACUUM
The term vacuum refers to a given space filled with a fluid at any
pressure below atmospheric. The fluid (gas) most frequently
encountered in vacuum applications is air.
There are two principle ways to designate a vacuum:
1. In terms of absolute pressure where the measurement is made from
zero pressure — A unit conversion chart for absolute pressure
equivalents is listed below. A more complete table is attached.
2. In terms of differential where the measurement is made from
atmospheric pressure — The vacuum is stated as the difference
between absolute pressure and standard atmospheric pressure.
This measurement is always followed by the word “vacuum”.
Typically, this is stated in inches such as; 28 inches of vacuum
which is equivalent of 29.92 minus 28 or 1.92 inches of mercury in
absolute pressure units.
CONVERSION CHART PRESSURE EQUIVALENTS
Atmospheres
1
PSIA
Feet of
Inches of
Millimeters of
(Torr X 1000)
Water
Mercury
Mercury (Torr)
Microns
14.7
33.93
29.92
760
760,000
0.068
1
2.31
2.04
51.7
51,700
0.029
0.43
1
0.88
22.4
22,400
0.33
0.49
1.13
1
25.4
25,400
0.00132
0.0193
0.0446
0.0394
1
1,000
0.00000132
0.0000193
0.0000446
0.0000394
0.0010
1
DEGREES OF VACUUM
Condition
Pressure Range (Torr)* Microns
Since it is not possible, in practice, to remove all matter from a
chamber or vessel, degrees of vacuum have been established to
indicate different pressure ranges. These pressure ranges are usually
expressed in terms of “Torr”, which is the international standard term
replacing the designation millimeters of mercury and its abbreviation
mm of Hg. Listed below are the generally accepted degrees of
vacuum.
Low Vacuum
760 to 25 Torr
1. Atmospheric pressure (sea level, 0°C) = one atmosphere.
10-3
Torr
760,000 to 25,000
Medium Vacuum
25 to
High Vacuum
10-3 to 10-6 Torr
1 to 0.001
25,000 to 1
Very High Vacuum
10-6 to 10-9 Torr
0.001 to 0.000001
Ultra High Vacuum
10-9 Torr and Below
0.000001 and Below
*10-3 Torr is actually 1/1000ths Torr and it could be expressed as .001 Torr.
10-6 Torr could be expressed as .000001 Torr, etc.
2. 1 Atmosphere = 760 millimeters of Mercury (mm Hg).
3. 1 Atmosphere = 29.92 inches of Mercury = 14.7 psia.
4. 1 Torr = 1mm Hg = .0013 atmosphere (roughly
1/1000th of an atmosphere).
5. 1 Micron Hg = .001 millimeters of Mercury (.001 Torr).
6. 1 Micron Hg = .0000013 atmosphere (roughly
one millionth of an atmosphere).
7. Hg is the symbol for Mercury.
8. mm. = millimeters
9. Typical Leakage Unit = 1 standard cc per second = 1 cubic
centimeter of gas at atmospheric pressure (often shortened to
cc/sec). All leak rates are related to a specific pressure (or level of
vacuum).
10. 1 Torr = 1 millimeter of mercury.
2
TP-7D-1
Flow Control Division
Worcester Control Valves
VALVING
In order to produce a good vacuum valve, several design
considerations should be met. Some of the more important ones are
listed below:
1 mm of Hg = 1 Torr
1. All valve materials must be non-porous. If material is porous, the
time required to evacuate a system may be drastically increased.
This delay is due to the necessity of evacuating all the minute pores
of the material.
1 Torr = mm of Hg
1 Micron = .001 mm (10-3 mm)
The 316L clean-mizer valves are electro polished to reduce surface
area cavities.
2. Materials should be used, insofar as possible, which have the least
outgassing** properties. Outgassing is caused by improper
cleaning and/or inherent characteristics of the material.
TFE Seats, stem seals and TFE or silicon body seals
provide maximum performance.
3. Special handling should be used to prevent grease, oil or
even fingerprints which contribute to the problem of
outgassing.
Standard service valves
are used in the vacuum
range down to 20 microns
Vacuum cleaning to Flowserve specifications and
assembly in our “white rooms” with proper packaging
and protection.
4. A valve must completely seal to prevent in-leakage.
Sealing surfaces must have an especially good finish.
.1 mm of Hg (10-1)
5. In a ball valve, the area between the ball and the body
cavity must be accessible to the evacuation system when
the valve is in the open position.
.05 mm of Hg
20 Microns (.02 mm)
of Hg, or 10-2 mm
The hole in the ball below the stem allows pumping out
of the cavity and reduces time to reach
proper levels of vacuum.
.0000001 mm of Hg
10-7mm or 10-5 microns
absolute vaccum
**Outgassing is a molecular flow of gas
from the surface and/or the interior of
the material. All materials outgas but
the rate varies with the temperature,
cleanliness and material composition.
6. In some systems, a vented ball (V-3) is required.
The ratings of a valve for vacuum are established by
pumping the valve to a specific level of vacuum (altitude)
and then measuring leak rates for the product. Examples of
standards used for the Worcester Ball Valves are:
A) Standard Three Piece Products:
20 micron (2x10-2 Torr)
1x10-8 SccHe/sec or less
B) Valves Vacuum Cleaned, Teflon Parts, Vacuum Tested:
1 micron (1x10-3 Torr)
1x10-8 SccHe/sec or less
C) Clean Valves (XBO) Teflon Seats, Special Cleaning,
Inspection and Testing:
.01 micron (1x10-5 Torr)
1x10-8 SccHe/sec or less
O.D.
I.D.
Port Dia.
WORCESTER HIGH VACUUM VALVE
XBO Clean Valve
Hole in ball
Electropolished
Teflon Seats, Seals
Body Seals - TFE or Silicon
Special Inspection
Vacuum Cleaning
White Room Assembly
Vacuum Certification
Helium Mass Spectrometer Tested
Special Packaging
To give you an example of these leak rates:
3 oz. of fluid leaked over 5 years is equal to 1x10-4 Scc/sec.
TP-7D-1
3
Flow Control Division
Worcester Control Valves
VACUUM
In vacuum technology, a given space filled with gas at pressures
below atmospheric pressure. Since it is not possible, in practice, to
remove all matter from a space, degrees of vacuum (for example,
high vacuum, ultrahigh vacuum) have been arbitrarily distinguished
according to the pressure ranges characteristic of various techniques,
pumping equipment, or laws of gas flow.
To illustrate how little the leakage of a valve at 1x10-5 Torr is, we can
use the term “Mean Free Path”.
MEAN FREE PATH
The average distance a molecule will travel before it is likely to collide
with another molecule. This distance varies with pressure (molecule
density) as shown:
760 Torr
6x10-6 cm
(6 millionths of an inch)
1 Torr
5x10-3
(2 thousandths of an inch)
10-3
5 cm
(about 2 inches)
5x106 cm
(about 30 miles)
Torr
10-9 Torr
cm
Valves that are used above our ratings require special metal seats and
seals that can be baked at high temperatures to drive off the gasses.
These are usually produced by specialty companies that make high
vacuum equipment.
LEAK TESTING OF VALVES
Leaks are measured in units of gas flow per time. To specify the
quantity, the pressure, as well as the volume, must be recorded. This
pressure is generally referred to as Torr, and standard cubic
centimeters per second for the volume. Leaks are measured by the
rate by which the gas or fluid flows across a closed valve at a specific
temperature and pressure. Flowserve’s standard method for
measuring performance of the valve is to pressurize it with 80 psi,
immerse it under water and look for visible bubbles. This gives you a
threshold 1x10-4 standard cc’s per second. A more definitive method
is using our helium mass spectrometer. Here the valve is pressurized
with helium gas, a vacuum is pulled on the product and a calibrated
leak standard is used to measure the performance of the valve at a
specific altitude. Our machine is capable of working at 1x10-8 torrs
and leak rates of 1x10-8 Scc/he. Independent laboratories can go to
higher levels but this exceeds the vacuum capability of our Teflon,
stainless steel valves.
Common leak detection methods are as follows:
Acoustical 10-3 standard cc’s per second; pressure decay or bubble
testing 10-4; vacuum pumps, Halogen leak detectors, 1x10-7 standard
cc’s per second and the Helium mass spectrometer or radioisotope
measurements, 1x10-11 standard cc’s.
There is no specific level of performance which will meet all process
conditions. There are several questions that must be asked — What
type of leakage will damage the product or cause bacteria, moisture
or fugitive emissions? How long can the valve be exposed to the
process conditions without leakage through or externally and how
long must the product last in the intended service?
Most chemical plants or rough vacuum services work at 20 microns. In
space chambers or very critical laboratory environments, 10-8 standard
cc’s of leakage is the acceptable criteria. The more stringent the leak
rates, or difficult the specifications, the more expensive the valves.
Flowserve is in the unique position to be able to supply a wide range
of products for most industrial and scientific vacuum applications. If
you have specific requirements please contact the factory so we can
resolve what products to offer.
PRESSURE EQUIVALENTS
ConvertConvert
To
Convert
To
Convert
From
From
kg per
square cm
lb. per
square in.
atm.
meters of
Hg
in. of
Hg
meters of
water
in. of
water
ft. of
water
kg per
square cm
1
14.22
0.9678
0.7355
28.96
10.01
394.05
32.84
lb. per
square in.
0.07031
1
0.06804
0.05171
2.036
0.7037
27.7
2.309
atm.
1.0332
14.696
1
0.76
29.92
10.34
407.14
33.93
meters of
Hg
1.3596
19.34
1.316
1
39.37
13.61
535.7
44.64
in. of
Hg
0.03453
0.4912
0.03342
0.0254
1
0.3456
13.61
1.134
meters of
water
0.09991
1.421
0.0967
0.07349
2.893
1
39.37
3.281
in. of
water
0.002538
0.0361
0.002456
0.001867
0.07349
0.0254
1
0.0833
ft. of
water
0.03045
0.4332
0.02947
0.0224
0.8819
0.3048
12
1
1 ounce/sq. inch = 0.0625 lb/sq. inch
4
TP-7D-1
Flow Control Division
Worcester Control Valves
LEAKAGE RATINGS
Many misconceptions have been propagated regarding the application
and definition of the various valve seat leakage classes. These classes
were established by the American National Standards Institute and
sponsored by Fluid Controls Institute (ANSI B16.104 & FCI 70-2).
It is important to review the different classes periodically as defined
by the ANSI standard. In addition, it is worthwhile to study the
applicability of each class to industry. Discovering how the results
from seat leakage tests can or cannot be extrapolated to real working
conditions also provides needed operating guidance.
Please note that the ANSI seat leakage standard was established to
serve as a means whereby users and manufacturers of control valves
could gage the ability of a valve to shut off or control seat leakage. The
material discussed here is for example only. Seat leakage tests should
be conducted with reference to the latest ANSI standards. The actual
allowable seat leakage for a valve is dependent upon the manufacturers
published sizing information for that particular valve and its geometry.
The allowable seat leakage values used within this article are only
representative of general valve classifications and should not be used
as a standard.
Control valve seat leakage is commonly referred to as either ANSI
Class I, II, III, IV, V, or VI. Each of these standards has a unique test
procedure and an allowable leakage rate under the conditions set
forth in the test procedure.
Class VI seat leakage requires that the testing media be air or nitrogen
gas and be between 50°F and 120°F (10–52°C). The pressure
differential of the testing media is allowed to be the lower of the
maximum rated differential across the valve or 50 psi (3.5 bar). Most
specifications would normally have much higher differential pressures
upon closure of the valve. It should be noted that most valves are
harder to seal at low pressure and testing is safer.
Although this standard very well establishes the extent a valve will
control leakage with a low pressure differential and near ambient
temperatures, it is difficult and inadvisable to try extrapolating this
information to higher pressure differentials or temperatures. This test
should, however, detect any major flaws in the seating surfaces of the
valve plug or seat ring, or errors in valve assembly.
CLASS V
The Class V standard mainly applies to metal-seat, single-port,
single-seated, unbalanced or balanced control valves. Unlike any
other class, Class V allowable leakage is dependent not only on the
minimum orifice diameter of the valve, but also upon the maximum
differential pressure to be present in actual working conditions while
the valve is closed.
Allowable leakage for Class V valves is defined as 5x10-4 milliliters
per minute of water per inch of orifice diameter per psi differential
(5x10-12 m3 per second of water per millimeter of orifice diameter per
bar differential).
CLASS IV
CLASS VI
The first class, Class VI, frequently is misinterpreted as “bubbletight”. Actually, a certain amount of leakage is allowed here.
Coincidentally, this leakage often is measured by the number of
bubbles of air that escape per minute within the established test
guidelines. This class normally is associated with resilient seated
valves, but can and has been applied to other seating surfaces.
The most frequently specified leakage class probably is Class IV. It is
applicable to unbalanced, single-seat control valves and to balanced
valves and metal-to-metal seats. This standard allows a testing media of
either air or water at between 50°F and 125°F (10–52°C). The
differential pressure is allowed to vary between 45 and 60 psi (3–4 bar).
VACUUM SCALE
Atmospheres
Pounds Per
Square Inch
Inches of
Mercury
Torr (mm. of
Mercury)
psi
vacuum
psi
absolute
in. Hg
vacuum
in. Hg
absolute
Torr
vacuum
Torr
absolute
microns
vacuum
microns
absolute
1.0
0
14.7
0
29.9213
0
760
0
760,000
1x10-3
14.68
0.0147
29.89
.0299
759.24
.760
759,240
760
2.64x10-5
14.6956
.000388
29.9205
.000790
759.98
0.02
759,980
20
1.32x10-8
14.6959
1.9x10-7
29.9212
3.9x10-7
759.99
.00001
759,999
.01
0
14.7
0
29.9213
0
760
0
760,000
0
TP-7D-1
Pressure
Microns of
Mercury
Use standard
ball or wafer
sphere valves
Vacuum
zone
Use high
vacuum
ball valves
Consult
home
office
5
Flow Control Division
Worcester Control Valves
Maximum allowable leakage is defined as 0.01% of the rated valve
capacity. (Note: the rated valve capacity is the maximum flow that the
valve can pass when opened to its maximum rated travel and allowed
to flow under identical conditions as specified within the test
procedure. It does not mean that the Cv of the valve while shut will be
0.01% of the maximum Cv of the valve.)
Because Class V requires that the valve be tested with a pressure
differential equal to the maximum closed pressure differential, this
class is specified in applications where high-pressure drops occur
across the seat. This can help to ensure that the valve and actuator
are configured adequately to resist against seat erosion, such as wire
drawing due to high shutoff pressure differentials.
As with Class VI, this standard does not allow for extrapolation to the
actual working conditions to be encountered by the valve. Class IV
simply provides a baseline measurement of the valves’ ability to shut
off under low pressure differentials and ambient temperature.
The second misconception involves a major factor in the ability of a
control valve to shut off. This is in the ability of the control valve
actuator to turn a ball, drive the plug, disk or sealing mechanism to
the closed position to overcome friction or provide additional seat
loading even during changing pressure conditions. Only the Class V
standard requires that the valve be tested under any actual conditions,
which can provide valuable insight as to how the valve and actuator
combination will react after installation.
CLASS III
Class III uses the identical test procedure as Class IV. However, the
allowable leakage is 0.1% of the rated valve capacity rather than the
0.01% allowed for Class IV. Class III often is used as a higher-thannormal standard for double ported, double-seated, or pressurebalanced (globe) valves that do not use an uninterrupted pressurebalancing seal such as an O-ring.
CLASS II
The identical test procedure is used in Class II as in Classes III and IV,
but allows for up to 1.0% of the rated valve capacity as seat leakage.
As with Class III, this class normally is used for double-ported,
double-seated valves and pressure-balanced valves without an
uninterrupted pressure-balanced seal (globe valves).
CLASS I
The Class I standard allows for modification in Class II, III, or IV when
the design of the valve to be tested has the same design intent as the
basic class. However, through agreement between the buyer and
supplier, the actual test is waived.
Misconceptions
Three misconceptions frequently are promoted. The first is that Class
VI allows less leakage than Class V. This statement is not always true,
although it often is true. Depending upon the size of the orifice
diameter and actual maximum pressure differential, Class V can be
more stringent than Class VI and in many cases much more stringent.
With other leakage classes, the customer must depend upon the
supplier to provide an actuator that will actually deliver sufficient force
for an allowable shutoff at the installed flow conditions.
More Misconceptions
Another frequent misconception about the established leakage classes
is that since Class II allows for 10 times more leakage than Class III and
Class III allows for 10 times more leakage than Class IV, then Class V
and Class VI must allow for approximately 10 and 100 times more
leakage (respectively) than Class IV. This concept is completely false.
The final misconception is that the allowable leakage for a ball valve
of a particular size (Classes I, II, III, and IV) will be equivalent for the
same size of globe or butterfly valve. Because each valve has a
different Cv and flow recovery, the maximum flow capacity for all
types of valves will not be equal.
The allowable leakage in Classes I through IV is a percentage of the
rated valve capacity. Hence, the allowable seat leakage for different
styles of valves will be different. Class V and Class VI leakages are
dependent upon the minimum orifice or port diameter of the valve.
Further, ball valves typically have larger port diameters than butterfly
valves. A butterfly valve has a larger port diameter than a globe.
Therefore, the allowable leakage will be greatest for a ball, followed by
butterfly valve and a globe valve, respectively.
Testing Specifications: Leakage Rates
6
Duration of
Test (sec.)
Valve Type
Test Pressure (psig)
Allowable Leakage
Resilient Seats
Standard
Vacuum
Oxygen
¹⁄₄"– 2" 80–100 (Dry Air)
3"– 8" 30–40 (Dry Air)
Bubbletight (0 bpm)
60
Chlorine
80 psig (Dry Air or Nitrogen)
0 bpm
60
Cryogenic
80 psig (Helium)
0 bpm
60
Metal Seats
Alpha (Teflon
Impregnated)
0 bpm
30
Gamma (Graphite
Impregnated)
See Table A
TP-7D-1
Flow Control Division
Worcester Control Valves
PRESSURE RATINGS
TRAPPED CAVITY PRESSURE TEST
Here is a clarification of the “pressure terminology”:
INTRODUCTION
Water Oil Gas: WOG
“WOG”, derived from “water, oil and gas” denotes the highest
pressure the valve is capable of at a temperature between -20°F to
100°F. It is the maximum rating for the valve because most materials
have a reduction in allowable working stress at elevated temperatures.
It was speculated that trapped cavity pressure was caused by a
change of state (i.e. from liquid to gas) of the media in the valve
cavity. Other sources of information implied it was the expansion of
the liquid itself that caused the damage attributed to trapped cavity
pressure. So, a simple experiment was conducted in the lab to
determine what actually does transpire.
Cold Working Pressure: CWP
“Cold Working Pressure” is another designation for “WOG”.
Working Steam Pressure: WSP
This is the on/off pressure rating for a steam valve. It is the
maximum saturated pressure which the valve will shut off. In this
particular case, it is the pressure generated by saturated steam.
WOG, CWP and WSP are for on/off valves.
Maximum Allowed Throttling Drop
A control valve functions as a variable resistance in a pipeline. It
provides a pressure drop by changing the turbulence in the process
fluid or in the case of laminar flow, the pressure drop is caused by
the changed valve resistance or “drag”. This pressure drop process
is often called throttling. The allowable throttling pressure drop is
determined by the mechanical strength of the seat. The seat rating
is an engineering evaluation based on the material, strength at
ambient and elevated temperatures for applications in saturated
steam. Allowable pressure drop ratings are for clean service, dirty
service and steam.
VALVE BODY CAVITY PRESSURE RELIEF
Valve body cavity pressure relief is necessary in liquid service to
prevent the thermal expansion of liquid in a closed valve from causing
damage or torque that is so high the valve cannot be opened until the
pressure rise is bled off. The seat areas acted upon by upstream
pressure are designed to relieve the pressure automatically to the side
of lowest line pressure. Lockup of valves in liquid service has not
been uncommon in the past before this type of design was produced.
Cavity Pressure Relief (CPR) Seats are standard on 3" and larger
valves. V-3 vent hole can be used on all valves.
DESCRIPTION OF EXPERIMENT
The handle was removed from a two inch valve and inverted. A hole
was tapped on the bottom of the valve and a long piece of stainless
steel tubing was threaded into the hole. The valve was partially
opened and submerged into a tank of water with the tubing attached
in the inverted position. The valve and tubing was then filled with
water and subsequently closed to trap the water. It was assumed that
the valve was completely filled with water even though there had to be
some air trapped in the tubing. For a small increase in temperature,
this trapped air would alter the corresponding pressure increase
significantly. However, in this experiment, since there was a large
increase in temperature, it was of less concern. The valve was placed
in the oven by snaking the tube through the top hole of the oven. A
pressure gauge was then attached to the end of the tubing.
RESULTS
When the temperature of the valve was elevated to 200°F, the
pressure INCREASED to about 100 psig. The temperature of the valve
then elevated to approximately 300°F. The corresponding pressure
reading was 1100 psig. A rather significant increase. It would have
been more had there not been air in the system and had the Teflon
seats not deformed which increased the volume inside the cavity. The
Teflon seats actually started to “blow” through the port hole; having
reduced in diameter about .040. The seats completely flattened out on
the back surface of the pipe end (recall that there is chamfered back
surface on the seat). This simple test demonstrated the physical
phenomena of “trapped cavity pressure”.
ORIGIN OF CAVITY PRESSURE
Printed here is an experiment done in the lab pertaining to this very
subject:
TABLE A: Gamma Seats
51/51
Series
¹⁄₂"
¹⁄₂"
³⁄₄"
³⁄₄"
¹⁄₂"
1
1"
1"
³⁄₄"
1
1"
2
1¹⁄₂"
3
1-¹⁄₄"
TP-7D-1
1¹⁄₂"
1¹⁄₂"
2"
2"
3"
3"
4"
4"
59
Series
45
Series
Maximum
Acceptable Leakage
In Bubble/Minute
44
Series
1
2
2"
2¹⁄₂"
3"
4
6
3"
4"
11
7
Flow Control Division
Worcester Control Valves
CONCLUSION
Steam tables show that saturated steam can only exist at 300°F when
the pressure is approximately 53 psig. The media inside the valve
cavity was most definitely in the liquid state because the pressure
indication was 1100 psig. Consequently, it was this LIQUID that
caused the damage to the seat. Liquid is emphasized here because a
change in the state of the media did not occur which was earlier
speculated. This is due to the incompressible nature of water and that
there was practically no room for expansion in the valve cavity
considering the rather large increase in temperature.
As an aside, keep in mind that trapped cavity pressure not only
occurs from an external heat source (similar to this experiment), but
more often than not from process heat such as hot water, steam, etc.
used to purge lines. This can expose one seat and the filled cavity to
the elevated temperatures resulting in “trapped cavity pressure”.
standpoint, because most of the effects and evidences of cavitation are
related directly to the collapse, rather than the formation of cavities.
As the fluid stream approaches the restriction in the line, its crosssectional area must decrease in order to pass through the orifice. The
velocity is inversely proportional to the stream area and, therefore,
must increase. Since the sum of the pressure and velocity heads will
remain approximately equal, an energy interchange must take place
with the pressure head losing what the velocity head gains.
Immediately downstream of the orifice, the stream will reach its
minimum cross section and thus its maximum velocity and minimum
pressure. This point is called the vena contracta (Pvc). If the velocity
is increased sufficiently, the pressure will fall to the vapor pressure,
thus permitting the formation of voids in the stream, which is the first
stage of cavitation.
PRESSURE RECOVERY
CAVITATION AND FLASHING
CAVITATION
Cavitation is a two stage phenomenon, the first stage of which is the
formation of voids or cavities within the fluid system. The second stage
is the collapse or implosion of these cavities back into an all liquid
state. Some have defined cavitation as merely the formation of cavities,
but this seems to be an impractical definition from a control valve
Downstream from the vena contracta, fluid friction causes the
stream to decelerate with resultant increases in both stream cross
section and pressure. This reversal of the energy interchange
between the velocity and pressure heads is called “pressure
recovery”. Vapor bubbles, formed as a result of reducing the
pressure at the vena contracta to the vapor pressure, cannot exist at
the increased pressure downstream and are forced to collapse or
implode back into the liquid state. When this has occurred, the
cavitation process is complete.
Analogous to a
control valve
(Vena Contracta)
P1
Pvc
P2
Velocity (fps)
Pressure (psi)
Pressure
Velocity
Liquid Vapor Pressure
Pressure and velocity changes caused by a restriction in a line.
8
TP-7D-1
Flow Control Division
Worcester Control Valves
FLASHING
If the pressure in the downstream piping system were being
maintained at a level equal to or less than the inlet vapor pressure, the
fluid proceeding downstream would have an increasing percentage of
vapor, the velocity of the stream would continue to increase and the
end result would be flashing rather than cavitation.
As cavitation and flashing have been defined, the following restrictive
statements may be made regarding their occurrence in control valves.
CAVITATION
of metal is removed. Bubbles that implode at a sufficient distance
from a solid surface are believed to be incapable of physical damage,
because their energy is absorbed by the flowing liquid.
FLASHING EVIDENCES
When a fluid is flashing in a control valve, there are two major
problems which must be considered: physical damage and decreased
efficiency. It will be recalled that these were also two of the problems
associated with cavitation.
1. The fluid at both the inlet and outlet must be in an all-liquid
condition. That is, there can be no vapor present in the piping
upstream or immediately downstream of the valve.
Physical damage associated with the flashing process has a very
smooth appearance in contrast to the cinder-like resemblance of
cavitation damage. The appearance is often compared to a fine
sandblasted surface, although in some material the surface may be
even smoother.
2. The liquid must be in a sub-cooled state at the inlet. Obviously if
the fluid were all liquid at the inlet, but in a saturated state, any
pressure drop across the valve would result in residual vapor
present downstream.
FLASHING MECHANISM
3. The valve outlet pressure must be either at or above the vapor
pressure of the liquid. Conceivably, it would be possible for
cavitation to exist if a saturated, but all-liquid, condition were
present at the downstream location.
FLASHING
1. The fluid at the inlet must be in an all-liquid condition while some
vapor must be present at the valve outlet. Obviously, if vapor were
present at the inlet, any pressure differential taken across the valve
would result in the formation of additional vapor. This restriction,
requiring absence of vapor at the inlet, is imposed primarily
because the valve sizing procedures are considerably more
complicated when the inlet stream contains vapor.
2. The fluid at the inlet may be in either a saturated or a subcooled
condition.
Actually, the damage mechanism probably closely resembles
sandblasting. In sandblasting the grit is borne along by some
compressible media, usually steam or air, and achieves a velocity
approaching that of the carrier. These high velocity particles impact
on the surface causing deformation and ultimately removal of the
surface material.
In a flashing liquid, the volume of vapor is frequently greater than the
volume of liquid so that the liquid droplets tend to achieve the high
velocity of the vapor. Their impact upon a surface will have the same
tendency to deform and remove material as will solid particles,
although perhaps not to the same extent.
Physical damage to control valves passing flashing liquids is usually
confined to the downstream section and it is not uncommon for the
piping immediately downstream to be eroded. If the liquid is near
saturation at the valve inlet, so that flashing begins in the upstream
section of the body, the valve plug’s seating surface and flow
contours may also be affected.
3. The valve outlet pressure must be either at or below the vapor
pressure of the liquid.
CAVITATION EVIDENCES
The occurrences of cavitation in a control valve is always
accompanied by one or more of several characteristic evidences
which include noise, vibration, physical damage and a decrease in
efficiency. The numbers manifested, and their magnitudes, are
dependent upon the proximity of the instantaneous operating point to
the point of incipient cavitation. All of the effects of cavitation can be
traced to the implosion of the vapor bubbles, with the exception of the
decrease in efficiency which is traceable to their formation.
CAVITATION DAMAGE MECHANISM
Several theories have been advanced relative to the mechanism by which
cavitation damage occurs. One fact that is known, however, is that
damage always takes place in close proximity to the imploding bubbles.
Earlier investigators have reported pressures as high as 100,000 psi
in collapsing cavities. One theory suggests that a high pressure shock
wave emanates from each and when these shock waves originate near
a solid boundary, a highly concentrated, but minutely small blow is
struck. Repeated strikes on any given increment of surface tend to
fatigue it until eventually the endurance limit is reached and a tiny bit
TP-7D-1
9
Flow Control Division
Worcester Control Valves
10
TP-7D-1
Flow Control Division
Worcester Control Valves
TP-7D-1
11
Flow Control Division
Worcester Control Valves
Flowserve Corporation has established industry leadership in the design and manufacture of its products. When properly selected, this Flowserve product is designed to perform its intended function
safely during its useful life. However, the purchaser or user of Flowserve products should be aware that Flowserve products might be used in numerous applications under a wide variety of industrial
service conditions. Although Flowserve can (and often does) provide general guidelines, it cannot provide specific data and warnings for all possible applications. The purchaser/user must therefore
assume the ultimate responsibility for the proper sizing and selection, installation, operation, and maintenance of Flowserve products. The purchaser/user should read and understand the Installation
Operation Maintenance (IOM) instructions included with the product, and train its employees and contractors in the safe use of Flowserve products in connection with the specific application.
While the information and specifications contained in this literature are believed to be accurate, they are supplied for informative purposes only and should not be considered certified or as a guarantee of
satisfactory results by reliance thereon. Nothing contained herein is to be construed as a warranty or guarantee, express or implied, regarding any matter with respect to this product. Because Flowserve
is continually improving and upgrading its product design, the specifications, dimensions and information contained herein are subject to change without notice. Should any question arise concerning
these provisions, the purchaser/user should contact Flowserve Corporation at any one of its worldwide operations or offices.
For more information about Flowserve Corporation, contact www.flowserve.com or call USA 1-800-225-6989.
FLOWSERVE CORPORATION
FLOW CONTROL DIVISION
1978 Foreman Drive
Cookeville, Tennessee 38501 USA
Phone: 931 432 4021
Facsimile: 931 432 3105
www.flowserve.com
© 2003 Flowserve Corporation, Irving, Texas, USA. Flowserve and Worcester Controls are registered trademarks of Flowserve Corporation.
TP-7D-1 6/03 Printed in USA