RP 7523

ISA-RP75.
23-1995
Approved June 2, 1995
Recommended Practice
Considerations for Evaluating

Control Valve Cavitation
ISA-RP75.23, Considerations for Evaluating Control Valve Cavitation
ISBN: 1-55617-572-8
Copyright 1995 by the Instrument Society of America. All rights reserved. Printed in the United
States of America. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means (electronic, mechanical, photocopying, recording, or
otherwise), without the prior written permission of the publisher.
ISA
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, North Carolina 27709
Preface
This preface, as well as all footnotes and annexes, is included for informational purposes only
and is not part of ISA-RP75.23.
This recommended practice has been prepared as part of the service of ISA, the international
society for measurement and control, toward a goal of uniformity in the field of instrumentation.
To be of real value, this document should not be static but should be subject to periodic review.
Toward this end, the Society welcomes all comments and criticisms and asks that they be
addressed to the Secretary, Standards and Practices Board; ISA; 67 Alexander Drive; P. O. Box
12277; Research Triangle Park, NC 27709; Telephone: (919) 990-9227; Fax: (919) 549-8288;
e-mail: standards@isa.org.
The ISA Standards and Practices Department is aware of the growing need for attention to the
metric system of units in general, and the International System of Units (SI) in particular, in the
preparation of instrumentation standards, recommended practices, and technical reports. The
Department is further aware of the benefits to USA users of ISA standards of incorporating
suitable references to the SI (and the metric system) in their business and professional dealings
with other countries. Toward this end, this Department will endeavor to introduce SI-acceptable
metric units in all new and revised standards to the greatest extent possible. The Metric Practice
Guide, which has been published by the Institute of Electrical and Electronics Engineers as
ANSI/IEEE Std. 268-1992, and future revisions, will be the reference guide for definitions,
symbols, abbreviations, and conversion factors.
CAUTION: The information presented within this ISA Recommended Practice is believed to be
accurate and reflects the current state of knowledge within the field. The information is an
interpretation and condensation of a large volume of literature and experience, some of which is
contradictory and speculative. Therefore, application of the information to particular situations
requires the exercise of the independent professional judgement of the user. ISA is not
responsible for any results from such use of the information and shall not be liable for any
damages caused by such use.
It is the policy of ISA to encourage and welcome the participation of all concerned individuals and
interests in the development of ISA standards, recommended practices, and technical reports.
Participation in the ISA standards-making process by an individual in no way constitutes
endorsement by the employer of that individual, of ISA, or of any of the standards that ISA
develops.
The following people served as members of ISA Subcommittee SP75.16:

NAME
*F. Cain, Chairman
W. Weidman, Managing Director
G. Barb
*R. Barnes
S. Boyle
COMPANY
Valtek International
Consultant
Consultant
Neles-Jamesbury, Inc.
*One vote per company

ISA-RP75.23-1995
NAME
COMPANY
D. Buchanan
L. Driskell
L. Griffith
J. Harkins
H. Illing
C. Koloboff
G. Kovecses
J. Leist
H. Maxwell
H. Miller
T. Molloy
W. Rahmeyer
M. Riveland
A. Shea
E. Skovgaard
J. Stares
Union Carbide Corporation

Consultant
Retired/Consultant
Consultant
DeZurik Valve Company
Chevron Research & Technology Company
Yarway Corporation
Dow Chemical USA
Yarway Corporation
Control Components, Inc.
Pacific Gas & Electric Company
Utah State University
Fisher Controls International, Inc.
Copes-Vulcan, Inc.
Leslie Controls, Inc.
Masoneilan-Dresser
The following people served as members of ISA Committee SP75:

NAME
*D. Buchanan, Chairman
W. Weidman, Managing Director
*T. Abromaitis
H. Backinger
G. Baenteli
G. Barb
H. Baumann
K. Black
H. Boger
G. Borden, Jr.
S. Boyle
R. Brodin
F. Cain
C. Corson
*C. Crawford
L. Driskell
*J. Duhamel
H. Fuller
*J. George
L. Griffith
B. Hart
F. Harthun
B. Hatton
R. Jeanes
C. Koloboff
COMPANY
Consultant
Red Valve Company, Inc.
J. F. Kraus & Company
Bechtel
Consultant
H. D. Baumann & Associates, Ltd.
Cashco, Inc.
Masoneilan-Dresser
Consultant
Neles-Jamesbury, Inc.
Fisher Controls International, Inc.
Fluor Daniel, Inc.
Consultant
Red Valve Company, Inc.
Consultant
Richards Industries, Inc.
Consultant
M. W. Kellogg Company
Consultant
Honeywell, Inc.
TU Electric
Chevron Research & Technology Company

4
ISA-RP75.23-1995
NAME
G. Kovecses
C. Langford
J. Leist
A. Libke
R. Louviere
O. Lovett, Jr.
A. McCauley, Jr.
H. Miller
T. Molloy
L. Ormanoski
J. Ozol
W. Rahmeyer
J. Reed
*G. Richards
T. Rutter
K. Schoonover
A. Shea
E. Skovgaard
H. Sonderegger
R. Terhune
R. Tubbs
L. Zinck
COMPANY
Yarway Corporation
Consultant
Dow Chemical USA
DeZurik Valve Company
Creole Engineering Sales Company
Retired/Consultant
Chagrin Valley Controls, Inc.
Control Components, Inc.
Frick Company
Commonwealth Edison
Utah State University
Norriseal Controls
Richards Industries, Inc.
Fluid Controls Institute, Inc.
Con-Tek
Copes-Vulcan, Inc.
Leslie Controls
Grinnell Corporation
Cranmoor
Industrial Valve & Gauge Company
Consultant
This recommended practice was approved by the ISA Standards & Practices Board on
June 2, 1995.
NAME
COMPANY
M. Widmeyer, Vice President

H. Baumann
D. Bishop
P. Brett
W. Calder III
H. Dammeyer
R. Dieck
H. Hopkins
A. Iverson
K. Lindner
T. McAvinew
A. McCauley, Jr.
G. McFarland
J. Mock
E. Montgomery
D. Rapley
R. Reimer
R. Webb
Washington Public Power Supply System

H. D. Baumann & Associates, Inc.
Chevron USA Production Company
Honeywell, Inc.
Foxboro Company
Phoenix Industries, Inc.
Pratt & Whitney
Utility Products of Arizona
Lyondell Petrochemical Company
Endress + Hauser GmbH + Company
Metro Wastewater Reclamation District
Chagrin Valley Controls, Inc.
Honeywell Industrial Automation and Controls
Consultant
Fluor Daniel, Inc.
Rapley Engineering Services
Allen-Bradley Company

ISA-RP75.23-1995
NAME
COMPANY
W. Weidman
J. Weiss
J. Whetstone
C. Williams
G. Wood
M. Zielinski
Consultant
Electric Power Research Institute
National Institute of Standards & Technology
Eastman Kodak Company
Graeme Wood Consulting
Fisher Rosemount
ISA-RP75.23-1995
Contents
1 Scope ...................................................................................................................................... 9
2 Purpose ................................................................................................................................... 9
3 Definition of terms ............................................................................................................... 10
4 Nomenclature ....................................................................................................................... 12
5 Overview ............................................................................................................................... 15
6 Cavitation index and valve scale effects ........................................................................... 18
6.1 Application dependencies ........................................................................................... 18
6.2 Equations for scaling the cavitation coefficients ......................................................... 18
7 Applications ......................................................................................................................... 21
7.1
7.2
7.3
7.4
7.5
7.6
Method ........................................................................................................................ 21
Valve information ........................................................................................................ 21
Operating levels .......................................................................................................... 21
Considerations for selecting cavitation limits .............................................................. 22
Cavitation-resistant valve designs .............................................................................. 23
Examples .................................................................................................................... 23
8 Testing .................................................................................................................................. 28
8.1
8.2
8.3
8.4
8.5
8.6
Scope ......................................................................................................................... 28
Test system ................................................................................................................ 28
Test fluid ..................................................................................................................... 32
Test procedure ........................................................................................................... 32
Data evaluation ........................................................................................................... 33
Laboratory qualification .............................................................................................. 34
Annexes
A References.......................................................................................................................... 35
B Cavitation fundamentals ..................................................................................................... 37
C Cavitation damage intensity and service life....................................................................... 51
D Net pressure drop corrections ........................................................................................... 57
Figures
1
2
3
4
B.1
B.2
B.3
B.4
Cavitation parameter plot ............................................................................................... 11

Flow test system ............................................................................................................ 29
Cavitation detection equipment ...................................................................................... 29
Cavitation calibration test manifold ................................................................................ 30
Cavitation scale effects .................................................................................................. 47
Cavitation influences ...................................................................................................... 48
Typical flow curve appearance ....................................................................................... 49
Flow curve appearance: two flowpath butterfly valve .................................................... 50
ISA-RP75.23-1995
Tables
1 Numerical constants for cavitation equations ................................................................. 14
2 Pressure scale effect exponent ...................................................................................... 19
C.1 Range of estimated values of duty cycle factor, FDC ..................................................... 53
ISA-RP75.23-1995
1 Scope
This recommended practice is intended for control valves used in the control of process fluids
and is not intended to apply to fluid power components. The reader and user should be familiar
with fluid mechanics fundamentals and ISA standards ANSI/ISA S75.01 and ANSI/ISA S75.02
on valve sizing and testing. Definitions of terms in this document are intended for general
understanding; more rigorous definitions are found in the references.
Noise measurement and prediction methods are beyond the current scope of this document.
Methods of liquid flow noise measurement and prediction may be found in standards of the
International Electrotechnical Commission, CEI/IEC documents 534-8-2 and 534-8-4. The
relationship between cavitation parameters used in this recommended practice and those of the
IEC documents is discussed in Annex B.
2 Purpose
Cavitation as an applied science has not evolved to the highly refined level of that supporting the
more traditional control valve sizing calculations. However, there is a great need by users and
manufacturers alike for practical information in this area. The purpose of this document is to
supply that information, and to that end it is necessarily broad in scope. It embodies several
objectives:
a) to provide educational material in a background section that condenses the literature
and educates the reader in state-of-the-art valve cavitation knowledge and practice;
b) to establish a basis for communication by defining cavitation parameters and
nomenclature;
c) to propose methods for evaluating the cavitation characteristics of individual control
valves through testing procedures and application experience; and
d) to offer guidelines for selecting control valves for given applications.
ISA Subcommittee SP75.16 recognizes that the science of cavitation is in its infancy in terms of
defining the behavior of cavitation in complex valve geometry. The final objective of this
recommended practice is to promote additional research and testing. Subsequently, this practice
can serve as a starting point for those seeking to advance the state of the art.
ISA-RP75.23-1995
3 Definition of terms
Terms used are per ISA-S75.05 and additional terms as follows:

3.1
cavitation: A two-stage process associated with the flow of liquids. The first stage involves
the formation of vapor cavities or bubbles in the flow stream as a result of the local static pressure
in the flow stream dropping below the liquid vapor pressure. The second stage of the process is
the subsequent collapse or implosion of the vapor cavities back to the liquid state when the local
static pressure again becomes greater than the fluid vapor pressure. (The evaluation of "gaseous"
cavitation, i.e. the sudden dissolution of dissolved gases in a liquid, is not currently within the scope
of this document.)
3.2
cavitation coefficient: A characteristic number for (e.g., i, c, mv, id, ch), determined
for a given valve, valve opening, and pressure conditions, which corresponds to the numerical
value of the cavitation index at which the levels of incipient cavitation, constant cavitation, maximum
vibration cavitation, incipient damage, and choking cavitation occur.
3.3
cavitation index: The value for the operating service conditions of a valve, expressed as
and numerically equal to (P1 - Pv)/(P1 - P2).
3.4
cavitation level: The degree to which cavitation is occurring, i.e., incipient, constant, incipient damage, choking, or maximum vibration. Levels can be determined by testing for vibration,
pitting or metal loss, and changes in valve capacity (Cv).
3.5
choking cavitation: A limiting flow condition in which vapor formation is enough to limit
the rate of flow through the valve to some maximum value. Further increases in flow rate through
the valve are only possible by increasing the valve inlet pressure, because reducing downstream
pressure will no longer increase flow rate.
3.6
constant cavitation: An early level of cavitation characterized by mild, steady popping or
crackling sounds that may be audible or detected by vibration measurements. It is the next higher
inflection point on the cavitation profile above the point of incipient cavitation. (See Figure 1.) This
level is represented by the constant cavitation coefficient c.
3.7
duty cycle: The ratio of the amount of time a valve spends performing one particular function
to the valve's total installed time period. It may be expressed as a percentage of total time (service
time vs. installed time).
3.8
flashing: A flow condition in which vapor pockets formed inside a valve persist downstream
of the valve because the valve outlet pressure is at or below the fluid vapor pressure.
3.9
flow separation: A flow condition in which the fluid boundary layer flows away from the
boundary wall instead of flowing along the wall. A turbulent wake exists downstream of the point
of flow separation that is characterized by the presence of vortices. These vortices contain regions
of high local fluid velocities and hence low, local pressures. The areas of low pressure are potential
sites for vapor formation.
10
ISA-RP75.23-1995
Figure 1 Cavitation parameter plot

NOTE This is a classical curve illustrating acceleration versus sigma. Tested valves may
not result in this specific configuration or exhibit all the inflection points or coefficients shown
above. Test data are subject to expert interpretation.
ISA-RP75.23-1995
11
3.10 incipient cavitation: The onset of cavitation, where only small vapor bubbles are formed
in the flow stream. (See Figure 1.) This level is represented by the incipient cavitation coefficient
i or 1/xFz.
3.11 incipient damage: A cavitation level sufficient to begin minor, observable indications of
pitting damage. (This is not to be confused with incipient cavitation. See 3.10.)
3.12 influences: Factors or effects that change the damage rate or extent of damage but do not
change the numerical value of cavitation coefficients. See Figure B.2.
3.13 manufacturer's recommended cavitation limit: An operational limit expressed as a cavitation coefficient mr supplied by the valve manufacturer for a given valve type, size, opening, and
reference upstream pressure. Application of the limit may require scale effect and influence factors
if the service conditions and valve size are different than for the reference pressure and size.
3.14 maximum vibration cavitation: The level of cavitation associated with peak vibration
measurements and determined from a cavitation level plot at the peak separating Regime III and
Regime IV. The test conditions at this point define the conditions for calculating the valve cavitation
coefficient mv. See Figure 1.
3.15 pressure recovery: The increase in fluid static pressure that occurs as fluid moves through
a valve from the vena contracta to the valve's outlet and downstream piping. The recovery, which
may be expressed as the difference P2 - Pvc, is caused by the velocity-reducing, diffusing action
of the downstream geometry.
3.16 scale effects: Differences in cavitation coefficients occurring between the flow test conditions and actual valve operating conditions. These scale effects result from differences in valve
size and operating pressures. Scaling equations are used to modify the reference values of
cavitation coefficients supplied by valve manufacturers in order to evaluate equipment at other than
reference conditions. See Figure B.1.
3.17 vapor pressure: The pressure, for a specified fluid temperature, at which both the liquid
and vapor phases of a fluid exist in equilibrium. The vapor pressure is more commonly thought of
as the thermodynamic saturation pressure.
3.18 vena contracta: The minimum area of a flow stream. It is smaller than the area causing
the flow constriction, because the streamlines continue to converge for a short distance beyond
the constriction. Average flow velocity is highest and mean static pressure is lowest in the vena
contracta. However, local vortex pressures in separation regions and turbulent boundary layers
can be lower than the vena contracta pressure.
4 Nomenclature
Nomenclature used is per ANSI/ISA S75.01 and additionally as follows:
Empirical characteristic exponent for calculating PSE
A characteristic exponent for calculating SSE; determined from reference valve data for
geometrically similar valves.
12
ISA-RP75.23-1995
Cv
Valve flow coefficient*, C v = q(Gf/P)1/2
CvR
Valve flow coefficient of a reference valve
Valve inlet inside diameter, inches (mm)
dR
Valve inlet inside diameter of tested reference valve, inches (mm)
D1
Internal diameter of upstream pipe, inches (mm)
D2
Internal diameter of downstream pipe, inches (mm)
Napierian base, e = 2.71828... (for natural logarithms)
FDC
Duty cycle factor for modifying the intensity index
FF
Liquid critical pressure ratio factor*
FL
Liquid pressure recovery factor*
Fp
Piping factor for ISA valve sizing*
FT
Temperature factor for modifying the intensity index
FU
Velocity factor for modifying the intensity index
Gf
Specific gravity of the liquid at inlet flowing conditions
Intensity index
KB1
Bernoulli coefficient for upstream pipe reducer*
KB2
Bernoulli coefficient for downstream pipe expansion*
K1
Head loss coefficient for upstream pipe reducer*
K2
Head loss coefficient for downstream pipe increaser*
N1-4
Numerical constants for units of measure used in equations. See Table 1.
Pa
Atmospheric pressure, psia (kPa)
P1
Valve inlet static pressure, psia (kPa)
P2
Valve outlet static pressure, psia (kPa)
PSE Pressure scale effect

Pv
Absolute fluid vapor pressure of liquid at inlet temperature, psia (kPa)
Pvc
Fluid static pressure in valve vena contracta*, psia (kPa)
Volumetric flow rate, gpm (m 3/h)
SSE Size scale effect

SPL
Sound Pressure Level referenced to 20 x 10-6 Pascal (2.0 x 10-5 N/m 2)
Fluid Temperature, F (C)
Tave
Average temperature between a liquid's freezing and boiling temperatures for a specified
pressure, Tave is equal to (TF + T B)/2, F (C)
TB
Boiling temperature of a liquid for specified pressure, F (C)
TF
Freezing temperature of a liquid, F (C)
Average velocity at the valve inlet, ft/s (m/s)
*More completely defined in ANSI/ISA S75.01 and ANSI/ISA S75.02

ISA-RP75.23-1995
13
Uo
Pitting threshold velocity determined at the valve inlet, ft/s (m/s)
xFz
Coefficient of incipient cavitation per IEC-534-8-2; xFz 1/i
Measured valve differential pressure, psi (kPa)
Pch Pressure drop at choking, psi (kPa)
Cavitation index equal to (P1 - Pv)/(P1 - P2) at service conditions, i.e., (service)
Alternate cavitation index equal to (P2 - Pv)/(P1 - P2) at service conditions. See B.5.6.
Coefficient for constant cavitation; c is equal to (P1-Pv)/P at the conditions causing mild,
steady cavitation
ch
Coefficient for choking cavitation; ch is equal to (P1-Pv)/[FL2(P1-FFPv)] at the point

associated with choking in the valve
Coefficient of incipient cavitation; i is equal to (P1-Pv)/P at the point where incipient

cavitation begins to occur
id
Coefficient of incipient damage for cavitation; id is equal to (P1-Pv)/P at the conditions

causing onset of damage by cavitation
mr
Coefficient of manufacturer's recommended minimum limit of the cavitation index for a

specified valve; mr is equal to minimum recommended value of (P1-Pv)/(P1-P2)
mv
Coefficient of cavitation causing maximum vibration as measured on a cavitation

parameter plot (See Figure 1.)
Cavitation coefficient v that has been adjusted for the effects of installing a smaller-thanline-size valve with reducers in the pipeline.
A reference value of a cavitation coefficient.
ss
Cavitation index scaled for service pressure and size effects for use in intensity index
calculations; ss is equal to [(/SSE-1)/PSE] +1
Cavitation coefficient for a valve, scaled for a valve size and pressure other than the
originally tested size and pressure, that has geometric similarity to the tested valve. It
does not include the effect of reducers.
Table 1 Numerical constants for cavitation equations

Constant
N
Nc1
Nc2
Nc3
Nc4
14
Units Used in Equation

d, D
U, U0
1.00
0.00155
890
0.00214
1.00
25.4
0.078
0.256
in
mm
in
mm
in
mm
-
ft/s
m/s
ISA-RP75.23-1995
5 Overview
Cavitation is a phenomenon that can accompany the flow of liquids through control valves.
5.1
Failure to account for cavitation can result in potentially costly performance problems. To prevent
this situation, it is important that personnel responsible for control valve specifications understand
the nature of cavitation and fundamental abatement technology. The purpose of this section is to
provide the reader with a brief introduction to the subject. For a more comprehensive treatment
of the same subject, the reader is directed to Annex B. Familiarity with this material is encouraged.
Successful solutions to cavitation problems still rely heavily on engineering judgments stemming
from insight into cavitation basics.
Simply viewed, cavitation consists of the formation, growth and rapid collapse of cavities in
5.2
a liquid. These vapor cavities (bubbles) are formed whenever the prevailing fluid pressure falls
below the vapor pressure of the liquid. They subsequently collapse if the pressure again rises
above the vapor pressure.
Different specific sources of pressure changes cause cavitation, but they all arise from the
5.3
flow of the liquid through the control valve. Cavitation usually begins in the low pressure regions
associated with boundary layer separation. This may occur even though the mean pressure is
greater than the vapor pressure. Mean pressure (the average static pressure in the plane perpendicular to the flow path) will decrease as the liquid passes through the various restrictions in the
valve trim. The degree and extent of cavitation escalates when the mean pressure falls below the
vapor pressure in these regions.
Unacceptable noise levels, excessive vibrations, and physical damage to the valve and
5.4
adjacent hardware are the foremost problems associated with cavitation. These problems all arise
from the collapse of the vapor cavities. Material damage results from shock waves and micro-jets,
established during cavity collapse, impinging on the boundary surfaces. Corrosion further aggravates these mechanical attack mechanisms. The physical appearance of cavitation damage varies
from a "frosted glass" appearance to a rough, cinder-like surface texture.
Another "side effect" of cavitation is an apparent decrease in the efficiency of the valve. The
5.5
compressibility introduced to the fluid when a portion of the liquid vaporizes can ultimately lead to
a choked flow condition similar to a flashing fluid.
While treated simply in this section, cavitation is a very complex sequence of events. Not
5.6
all cavitation necessarily results in the problems mentioned above. However, attempts to model
the behavior of the cavitating liquid have not met with universal success. Distinguishing "problem
causing" cavitation from acceptable behavior presents some very real challenges.
Historically, the control valve industry has adopted the practice of describing cavitation
5.7
applications in terms of a single, unadjusted parameter. In this approach, the suitability of a given
control valve is determined by comparing the value of this parameter evaluated at operating conditions to an "operating limit" for that control valve.
While appealing from a user standpoint, the approach described above suffers from some
5.8
major drawbacks. First, the definition of the parameter and the manner in which it is used have
varied significantly from manufacturer to manufacturer. While the principle underlying the method
is basically the same, the differences in appearance lead to much confusion. Furthermore, the
complexity of cavitation renders it difficult to predict the exact behavior in any given service on the
basis of a single, unadjusted parameter. Many service factors can affect the apparent level of
ISA-RP75.23-1995
15
cavitation. Unfortunately, no currently known model fully describes the intensity or extent of cavitation under universally varying conditions regardless of the number of parameters employed.
The operating limit used as the basis for comparison has, in many instances, been equal to
5.9
the value of the pressure recovery factor, FL. If a valve is operated at the limit defined by the
pressure recovery factor, the valve is at or near choked flow conditions. Substantial vapor has
been formed in the flow stream, and significant levels of cavitation can exist. As discussed elsewhere, using FL in this manner is not a universally correct solution and is, in general, valid only for
specially designed valve trims. The vast majority of valves cannot operate problem free under this
condition.
5.10 A modified single parameter will be adopted for use in this document. A specific parameter,
as defined in Equation 1, is recommended. Adjustments to this parameter are supplied wherever
they are known to account for the variations associated with different application conditions. While
it is recognized that even this technique may have limitations, it is believed that it will provide a
justifiable blend of ease of use and meaningful predictions.
5.11
The parameter chosen for use in this document is the cavitation index
= (P1 - Pv)/(P1 - P2)
(1)
where P1 is the absolute pressure upstream of the valve, P2 is the absolute pressure
downstream of the valve, and Pv is the absolute vapor pressure of the fluid at the inlet
temperature.
5.12 As noted above, other parameters have been used and probably will continue to be used in
the future for essentially the same purpose. In many instances, well defined mathematical relationships exist between these parameters and the index defined in Equation 1. Several other
parameters, relationships to , and the associated advantages or disadvantages are discussed in
Annex B.
5.13 The index, in effect, quantifies only the service conditions. By itself it does not convey
any information about the performance of a particular valve in that particular application. Different
valves can tolerate different levels of cavitation, and different applications are concerned about
different aspects of cavitation (for instance, noise versus damage). Therefore, must be evaluated
at the service conditions and then compared to some benchmark.
5.14 The benchmark value for any specific application obviously will depend on both the problem
of concern (e.g., noise) as well as the valve style selected. Various limits have been suggested
and used in the control valve industry in the past. The following benchmarks, hereafter referred
to as levels, are used throughout this document:
a) incipient cavitation;
b) constant cavitation;
c) incipient damage;
d) choking cavitation; and
e) maximum vibration cavitation.
Definitions of these various levels are given in Section 3. More complete descriptions of their
significance are provided in Annex B. A discussion of the methods of determination is presented
in Section 8.
5.15 The different levels of cavitation listed in 5.14 merely define different significant cavitation
conditions that exist. No specific levels can be universally recommended. The appropriate level
16
ISA-RP75.23-1995
to use for a given application is not always self-evident and will usually embrace a degree of
subjectivity. In addition to the service conditions, factors such as valve style and opening, duty
cycle, location, desired life, and past experience should be considered. The valve manufacturer
always should be consulted. A manufacturer may recommend an application dependent valve
operating limit called "manufacturer's recommended limit" or mr. Additional discussion may be
found in Section 7 and Annex B of this document.
5.16 These various levels are a strong function of the internal geometry of the control valve. It
can be expected that different values of any given cavitation coefficient will be associated with
different valve styles or even different openings of the same valve.
5.17 Furthermore, the numerical values of these cavitation coefficients must be adjusted if a
reduced scale model was used to determine them. Any factor that changes the numerical value
of a cavitation coefficient as that factor is varied is known as a "scale effect." The numerical values
of coefficients can be corrected for "size scale effect" (SSE) through the use of scaling equations
presented in Section 6.
5.18 Research has shown that many factors contribute to the total nature of cavitation and to the
resulting problems. Some of these factors are associated with the valve geometry as noted in
5.16. Others are associated with the service environment. The foremost service condition scale
effect is the "pressure scale effect" (PSE). The numerical values of the various cavitation coefficients for a particular valve will change as a function of the pressures at which they are evaluated.
Consequently, an adjustment in the values is necessary if the service pressures are different from
the test pressures. Equations to calculate these adjustments are presented in Section 6.
5.19 At this point it is helpful to introduce another category of effects other than "scale effects"
as defined in 5.17. Application "influences" include factors that do not change the numerical value
of cavitation coefficients as in scale effects, but that do affect the intensity of the cavitation. The
list of influences is long, but for engineering purposes can be pared down to the following primary
effects (Knapp, ref. 1; Barnes and Cain, ref. 13):
a) viscosity;
b) velocity;
c) dissolved and undissolved gases in the liquid;
d) thermal properties of the liquid; and
e) duty cycle.
More detailed discussion regarding the nature of these effects and methods of accounting for
them are presented in Annexes B and C.
5.20 The pressure drop, P, measured in testing valves for cavitation is the pressure difference
between upstream and downstream pressure taps of the test manifold. For higher recovery valves
(Cv/N1d2>20) and critical applications, it may be necessary to adjust this pressure drop to account
for actual piping configuration in evaluating Cv, P, valve opening, and . Annex D discusses this
in more detail.
5.21 Cavitation will continue to be a major problem in industrial process control. An understanding
of the nature of the subject and utilization of current quantitative information will aid in formulating
effective problem abatement. However, the ultimate benefit of analyzing valves and service conditions for cavitation depends upon the quality and completeness of service and valve information
available. Valve users and manufacturers should make every reasonable effort to share clear and
accurate data (see ISA-RP75.21, Process Data Presentation). The data will make possible the
comparisons between the service conditions and valve capabilities.
ISA-RP75.23-1995
17
5.22 The procedures contained in this document are intended to provide the best practical knowledge currently available on the subject. However, practitioners always should avail themselves of
proven new technology as it becomes available.
6 Cavitation index and valve scale effects
6.1 Application dependencies

6.1.1 The cavitation index (Equation 1) is based on the assumption that the size of the valve
and the process fluid properties (other than vapor pressure) would have little effect on the value
of the index. (Knapp et. al., ref. 1). In actuality, the cavitation behavior and cavitation coefficients
are not independent or constant with either different upstream pressures or valve size. The change
in the value of a cavitation coefficient associated with change in pressure is known as the "pressure
scale effect" (PSE) (Tullis, refs. 2, 4; Rahmeyer, ref. 3; Stripling, ref. 6). Likewise, the change in
the value of a cavitation coefficient associated with valve size is known as the "size scale effect"
(SSE) (Tullis, refs. 2, 4; Rahmeyer, ref. 3; Stripling, ref. 6). In addition to direct valve size changes,
the presence of pipe reducers or increasers can also affect the value of a coefficient.
6.1.2 Methods are presented in this section to account for these effects. Equations are available
that adjust the value of the coefficients for the difference between the actual size or pressure
(operating conditions) and the test size or pressure (reference conditions) (Tullis, ref. 2). Modifications to the coefficients for pipe fittings are also provided. It is important to note that the equations
for scale effect adjustments were developed using water as the test fluid. They are assumed valid
for fluids other than water.
6.1.3 Other factors besides pressure and valve size also can influence the nature of the cavitation.
Fluid properties such as viscosity, density, and surface tension are recognized effects (Knapp, et
al., ref. 1; Barnes and Cain, ref. 13). Effects such as these do not tend to change the value of the
coefficients, but can change the degree of cavitation associated with any particular cavitation level.
These are referred to as "influences", which are further discussed in Annex B.
6.2 Equations for scaling the cavitation coefficients

6.2.1 Scaling equations and exponents have been derived to adjust or extrapolate cavitation
coefficients from one system pressure and size to another. Equation 2 gives the relationship by
which a coefficient for the valve, v, can be calculated from a reference coefficient, R. The value
of R may be chosen as the value of a cavitation coefficient or a manufacturer's limit mr. Equation
2 reflects the correlation of published data (Rahmeyer, ref. 3 and Tullis, ref. 4).
v = (RSSE - 1)PSE + 1
(2)
After v has been calculated, it can be compared to (service) calculated by Equation 1. If

(service) is greater than v, the valve will operate at a level of cavitation less severe than that for
which the valve's R was determined by the manufacturer. If the valve will be installed in a larger
diameter pipe, p must be calculated from v by Equation 7. The piping adjusted p is then
compared to (service).
18
ISA-RP75.23-1995
6.2.2 Intensity and level of cavitation increase with increasing (P1-Pv). The pressure scale effect
or scaling correction PSE can be calculated from the power relationship of Equation 3.
PSE = [(P1-Pv)/(P1-Pv)R]a
(3)
The subscript R refers to reference pressures. In most cases, the actual test pressure difference
(P1-Pv) will be less than 100 psi (690 kPa). For high recovery valves with Cv/N 1d2 > 20, refer to
Annex D for suggested piping loss corrections. For the purpose of uniformity in data presentation, it is recommended that the test data be scaled up to and presented at a reference value of
(P1-Pv)R is equal to or less than 100 psi (690 kPa). For example, for reasons of system capabilities and concerns for safety, a rotary valve might be tested at (P1-Pv) equal to 40 psi (276 kPa).
The exponent a of Equation 3 is found by measuring the slope of a log-log plot of n (n=i, c, mv,
id...) versus P1-Pv. Table 2 shows a list of typical values of the coefficient for different valve types
and the different levels of cavitation. The value of zero for the coefficient a indicates there is no
pressure scale effect for choking (Tullis, ref. 4; Rahmeyer & Odeh, ref. 5).
Table 2 Pressure scale effect exponent

Valve type
Cavitation
level
Exponent
a
Quarter-turn valves
(e.g., ball, butterfly)
Incipient
Constant
Incipient Damage
Choking
0.22 - 0.30
0.22 - 0.30
0.10 - 0.18
0
Segmented ball and

eccentric plug
Incipient
Constant
Incipient Damage
Choking
0.30 - 0.40
0.30 - 0.40
N/A
0
Single-stage globe
Incipient
Constant
Incipient Damage
Choking
0.10 - 0.14
0.10 - 0.14
0.08 - 0.11
0
Multi-stage globe
Incipient
Constant
Incipient Damage
Choking
0.00 - 0.10
0.00 - 0.10
N/A
0
Orifice
Incipient
Constant
Incipient Damage
Choking
0
0
0.20
0
N/A = Not Available
ISA-RP75.23-1995
19
6.2.3 Intensity and level of cavitation also increase with valve size. The size scale effect correction
SSE can be calculated from the power relationship of Equation 4. The exponent b of Equation 5
was derived from limited testing for the size scale effect for the cavitation levels of incipient, constant,
and incipient damage. Note that there is no size scale effect for the cavitation level of choking.
Therefore, the coefficient b has a value of zero for the level of choking cavitation (Rahmeyer &
Odeh, ref. 5; Stripling, ref. 6).
SSE = (d/dR)b
(4)
C v 1/4
b = 0.068 ------------2-
N d
(5)
where CvR and dR refer to the reference valve.

It also must be noted that this equation for scaling cavitation coefficients is rigorously valid only
when the valves are geometrically similar. Most valves usually are not geometrically similar for
different sizes. However, limited testing has suggested that these equations may be used
whenever two valves are of the same style (e.g., globe, butterfly, or ball), flow is in the same
direction, and
Cv/d2 = CvR/dR2
(6)
NOTE It is difficult to apply size scale effects to specialty, multi-stage, anti-cavitation

valve designs. For this reason, the valve manufacturer should be consulted for applications
of this type of product.
6.2.4 Another important effect on valve cavitation is the installation of a valve in a larger sized
pipeline for example, a 6-inch (150 mm) valve in an 8-inch (200 mm) pipeline. The upstream
pipe reducer and downstream expansion cause a variation in cavitation levels and coefficients.
Equation 7 can be used to calculate the corrected cavitation coefficient P for the installation of a
valve in a larger sized pipeline. The coefficient P (for the proposed valve and reducers) is then
compared to the (Equation 1) of the service system. The coefficient v (for the valve without
reducers) first should be calculated from Equation 2, before it is used in Equation 7. Equation 7
can be used with all of the cavitation levels (Rahmeyer, ref. 3, (&Odeh), 5; Cain, ref. 19).
P = FP2 [v + (K1+KB1)Cv2/(N2d4)]
(7)
FP = [1+(K) Cv2/(N2d4)]-1/2
(8)
KB1 = 1 - d4/D14
(9)
KB2 = 1 - d4/D24
(10)
K1 = 0.5 (1 - d2/D12)2
(11)
K2 = 1.0 (1 - d2/D22)2
(12)
K = KB1 - KB2 + K1 + K2
(13)
Equations 7 through 13 are theoretical expressions that account for the combined head losses
associated with the upstream and downstream reducers. Other potential effects of close coupled
reducers are not included. Although these equations have been supported by testing, the
accuracy of the method becomes less certain as the relative capacity (Cv/d2) increases. Actual
P should be used for maximum accuracy.
20
ISA-RP75.23-1995
7 Applications
7.1 Method
The method of using for determining the level of cavitation in a valve is a straightforward
procedure. A value of is calculated for the service conditions; is then compared to a selected
cavitation coefficient for the valve. If the (service) is greater than the of the selected limit, the
valve will experience an intensity of cavitation less than that associated with the selected limit.
This assumes that the service pressures and valve size are the same as those used in testing for
the different levels or regimes of cavitation. (See Annex B.7 and B.8) Section 6 explains how
data from cavitation tests may be "scaled" for comparison with other service pressures and valve
sizes.
7.2 Valve information

Evaluation or selection of a valve for cavitation applications can be reasonably objective after
certain decisions have been made and the necessary information is available. This section will
describe how to use cavitation coefficients discussed in Sections 5 and 6 to evaluate a valve in
specific service conditions. The following steps are prerequisites to the evaluation:
a) Gather system and service data for the valve. Pressures P1 and P2 should be
determined at the valve inlet and outlet, respectively. For valves with Cv/N 1d2 > 20
(e.g., ball, butterfly, and plug valves), see Annex D.
b) Make a selection of the general type valve (i.e., select globe, ball, butterfly, plug, etc.)
unless the valve type is already known.
c) Gather data from cavitation tests for the type and size of valve being considered. See
Section 8 on cavitation testing. Consult the manufacturer for available coefficients (i.e.,
c, id, etc.) or the recommended operating limit (i.e., mr).
7.3 Operating levels

The various levels of cavitation are introduced in 5.14 and described in Annex B. The
recommended cavitation coefficient for a particular valve depends on the application
requirements, valve performance capabilities, and cost considerations. A valve's reference
cavitation index R usually is set at a value of c, id, or mr depending on the tolerable level of
cavitation. Certain designs may not exhibit cavitation parameter plots with clear, meaningful i,
c, or other characteristics. Some valves may not have been tested due to test facility limitations.
In these circumstances, a manufacturer may provide only a mr based on field experience,
damage testing, or other criteria agreeable to the user.
ISA-RP75.23-1995
21
7.4 Considerations for selecting cavitation limits

7.4.1 Manufacturer's recommended limit, mr
The manufacturer's recommended limit for cavitation, mr, is the limit suggested by the
manufacturer for a given valve. It may or may not coincide with other cavitation coefficients such
as incipient damage or constant cavitation. Published values of this limit are based on
experience and on the normal type of application for the valve. Published values may not be
suitable for all applications. The manufacturer also should publish the criteria for the selection of
mr. The manufacturer always should be contacted to verify the recommended limit for each type
of valve application.
7.4.2 Incipient cavitation, i
Selecting incipient cavitation, i, as a limit restricts all operation to a cavitation-free regime. This
regime exists only with relatively low pressure drops. For instance, typical values of i for a high
recovery valve and a low recovery valve are approximately 15 and 8, respectively. Moderate to
high pressure drop applications require larger, more costly, multi-stage trim to maintain
cavitation-free operation. This regime of operation may be necessary in sensitive laboratory test
applications or in certain biochemical processes where all vaporization must be prevented.
7.4.3 Constant cavitation, c
Constant cavitation, c, is a mild level of cavitation that produces low levels of vibration and cavity
formation and poses no danger to valve equipment. This is considered a conservative
application limit; generally, no objectionable noise, vibration, or damage occurs at this condition.
7.4.4 Incipient damage cavitation, id
Operating at the incipient damage level of cavitation, id, may produce objectionable noise and
vibration in a control valve; minor indications of pitting begin on softer materials. Testing for id is
a complex process beyond the scope of this document. Additional guidance for damage testing
is available in the references (refs. 1,4,6,7,10,15). Many control valves can operate at this level in
moderate pressure drops if minor trim erosion is tolerable. For smaller valve sizes, stainless
steels or hardened materials often provide enough resistance for economical operation without
additional design enhancements. Special materials, multi-stage, or tortuous-path designs often
are required to operate at this limit in very high pressure drop applications. When operation
beyond this level cannot be avoided, evaluation of the intensity index per Annex C is suggested.
7.4.5 Choking cavitation, ch
The choking cavitation coefficient, ch, is not determined from a cavitation parameter plot (see
Figure 1), but it is calculated (Equation B.4) to correspond to the liquid pressure recovery factor,
FL. It is important to note that scale effects for pressure and size differences do not apply to FL or
ch. Most control valve applications should not use ch as a reference limit for operation. When
liquid flow becomes choked, the cavitation has become severe enough to develop very large
volumes of vapor in the valve and piping until pressure recovery causes the vapor cavities to
collapse. Flashing, on the other hand, does not produce cavity collapse because the
downstream pressure remains below the vapor pressure. The increase in volume develops
extremely high velocities. Most materials are subject to severe damage when exposed to choked
conditions for any significant length of time. Valves designed for this service are intended for
infrequent operation (e.g., pressure relief valves); some use sacrificial components (e.g., chokes
and liners); and they usually exhaust into tanks, headers, or vents to protect downstream piping
and equipment from erosion. Infrequent upset or start-up conditions may justify operating briefly
in the choked condition, but this should be done with great caution. Annex C on intensity and
service life presents an evaluation procedure for quantifying these situations.
22
ISA-RP75.23-1995
7.4.6 Maximum intensity of vibration, mv

As with choking, the conditions at the point of maximum intensity of vibration, mv, can subject
valve and piping to severe damage when exposed for a significant duration. Valves in this
service are designed for infrequent operation, sacrificial components, and for exhausting into
vessels without impingement on vessel walls. Annex C presents an evaluation procedure for
estimating damage intensity.
7.5 Cavitation-resistant valve designs

Frequently, it is necessary to use special designs to reduce the effects of cavitation or to
eliminate cavitation altogether. Manufacturers produce valves or special trim for cavitation
applications. Designs exist for rotary valves and for linear motion valves. Cavitation resistant
control valves vary widely in cost and performance capabilities. Understanding the application
requirements and accurate process data from valve users is essential in applying cavitationresistant valve designs.
7.6 Examples
The following examples illustrate the application of this cavitation evaluation method.
7.6.1 Rotary valve application (US Customary units)
Service data
Fluid: Water
T = 74 F
Line Size: 10-inch Sch. 40
Pv = 0.41 psia
P1 = 82 psia
Gf = 0.998
P2 = 70 psia
q = 3500 gpm
Results of valve sizing calculations
Cv = 1009
Preliminary valve selection
8-inch, ANSI Class 150 throttling rotary disk valve

Approximate opening = 75%
Calculate (Equation 1)
= (82-0.41)/12 = 6.80
Compare with manufacturer's

cavitation data. (mr and a are
based on incipient damage for
this application.)
i = 12.5
c = 7.0
id = 4.0
mr = 4.1
(P1-Pv)R = 100 psi; a = 0.12; dR = 6-inch
CvR/dR2 = 16 at 75% opening
Calculate PSE, b, SSE

(Equations 3, 4, 5)
PSE = [(82 -.41)/100]0.12 = 0.976

b = 0.068[1009/(1)(82)]1/4 = 0.14
SSE = (8/6)0.14 = 1.04
Calculate v (Equation 2)
Let R = mr = 4.1
v = [(4.1)(1.04)-1](0.976)+1 = 4.186
ISA-RP75.23-1995
23
Calculate effects of pipe reducers:

(Equation 9)
(Equation 10)
(Equation 11)
(Equation 12)
(Equation 13)
(Equation 8)
(Equation 7)
Evaluation
Valve inlet d = 8.00 in.

Pipe inside diameter D = 10.0 in.
KB1 = 1 - (8.0)4/(10.0)4 = 0.59
KB2 = 1 - (8.0)4/(10.0)4 = 0.59
K1 = 0.5[1-(8.0)2/(10.0)2]2 = 0.065
K2 = [1 - (8.0)2/(10.0)2]2 = 0.13
K = 0.59 - 0.59 + 0.065 + 0.13 = 0.195
Fp = {1+(0.195)(1009)2/[890(8)4]}-1/2 = 0.974
P = (0.974)2{4.186+(0.065+0.59)(1009)2/[890(8)4]}
P = 4.14
Valve has an acceptable P (i.e., P). Minor
cavitation may be present, because is slightly
less than c.
7.6.2 Globe valve in ammonia service (US Customary units)

T = 20 F
Pv = 48.2 psia
Gf = 0.65
q = 850 gpm
Service data
Fluid: Ammonia
Line Size: 3 inch Sch 40
P1 = 149.7 psia
P2 = 64.7 psia
P = 85 psi
Cv = 74.3
3-inch, ANSI Class 300 globe valve

Full Area trim, Linear Characteristic
= (149.7-48.2)/85 = 1.19

recommended based on incipient
damage for this application.
Trim Style mr
Standard 2.0
Trim A
1.15
Trim B
1.002
Trim A is checked for size and

pressure scale effects
Manufacturer's data for Trim A

(P1 - Pv)R = 90 psi; a = .20; dR=3.0 inch
PSE = (101.5/90)0.20 = 1.02
SSE = (d/dR)b = (3/3)b = 1.00
Calculate v for Trim A

(Equation 2)
Let R = mr = 1.15
v = [(1.15)(1.00)-1](1.02)+1 = 1.153
Evaluation
Trim A in a 3-inch globe has an acceptable v

(i.e., v).
7.6.3 Boiler feedwater application for globe valves (US Customary Units)
Service data
24
Fluid: Water
Line Size: 8-inch Sch. 80
Pv = 135 psia
P1 = 1600 psia
P2 = 1500 psia
T = 350F
Gf = 0.89
q = 1800 gpm
ISA-RP75.23-1995
Cv = 170
6-inch, ANSI Class 900, globe valve, d = 5.75 in.

Reduced Trim, Equal Percentage Characteristic
= (1600 - 135)/100 = 14.6
Manufacturer's recommended
operating and scaling data for
incipient damage
mr = 2.5
(P1 - Pv)R = 100 psi; a = 0.11;
dR = 3.0 inch; CvR/dR2 = 5.5
Calculate PSE (Equation 3)
PSE = [(1600-135)/100]0.11 = 1.34
Calculate b and SSE

(Equation 4, 5)
b = 0.068 (5.5)1/4 = 0.102

SSE = (5.75/3.0) 0.102 = 1.07
Let R = mr = 2.5
v = [(2.5)(1.07)-1](1.34)+1 = 3.24
Calculate effects of pipe reducers.
Valve inlet d = 5.75 in.

Pipe inside diameter D = 7.62 in.
KB1 = 1 - (5.75)4/(7.62)4 = 0.68
KB2 = 1 - (5.75)4/(7.62)4 = 0.68
K1 = 0.5[1 - (5.75)2/(7.62)2]2 = 0.093
K2 = [1 - (5.75)2/(7.62)2]2 = 0.185
K = 0.68 - 0.68 + 0.093 + 0.185 = 0.278
Fp = {1+(0.278)(170)2/[890(5.74)4]}-1/2
Fp = 0.996
p = (0.996)2{3.24+(0.093+0.68)(170)2/[890(5.75)4]}
p = 3.24
(Equation 9)
(Equation 10)
(Equation 11)
(Equation 12)
(Equation 13)
(Equation 8)
(Equation 7)
Evaluation
Preliminary selection is acceptable, because p

(3.24) is less than the operating (14.6). Also note
that piping effects are negligible for small values of
Cv/N1d2.
7.6.4 Rotary Valve Application (SI units)

Service data
Fluid: Water
T = 23.3 C
Line Size: NPS 10, Sch. 40, D = 254 mm
Pv = 2.83 kPa
P1 = 565.39 kPa
P2 = 482.65 kPa
Gf = 0.998
q = 795.6 m3/h
Cv = 1009
NPS 8, ANSI Class 150 throttling rotary disk valve

Approximate opening = 75%, d = 203 mm
= (565.39-2.83)/82.74 = 6.80
ISA-RP75.23-1995
25

cavitation data. (mr and a are
based on incipient damage for
this application.)
i = 12.5
c = 7.0
id = 4.0
mr = 4.1
(P1-Pv)R = 690 kPa; a = 0.12; dR = 152 mm
CvR/N 1dR2 = 16 at 75% opening
Calculate PSE, b, SSE

(Equation 3, 4, 5)
PSE = [(565.39-2.83)/690]0.12 = 0.976

b = 0.068[1009/(0.00155)(2032)]1/4 = 0.14
SSE = (203/152)0.14 = 1.04
Let R = mr = 4.1
v = [(4.1)(1.04)-1](0.976)+1 = 4.186
Valve inlet d = 203 mm

Pipe inside diameter D = 254 mm
KB1 = 1 - (203)4/(254)4 = 0.59
KB2 = 1 - (203)4/(254)4 = 0.59
K1 = 0.5[1-(203)2/(254)2]2 = 0.065
K2 = [1 - (203)2/(254)2]2 = 0.13
K = 0.59 - 0.59 + 0.065 + 0.13 = 0.195
Fp = {1+(0.195)(1009)2/[0.00214(203)4]}-1/2 = 0.974
P =
(0.974)2{4.186+(0.065+0.59)(1009)2/[0.00214(203)4]}
P = 4.14
(Equation 9)
(Equation 10)
(Equation 11)
(Equation 12)
(Equation 13)
(Equation 8)
(Equation 7)
Evaluation
Valve has an acceptable P (i.e., P). Minor

cavitation may be present, because is slightly less
than c.
7.6.5 Globe valve in ammonia service (SI units)

T = -6.67 C
D = 76 mm
Service data
Fluid: Ammonia
Line Size: NPS 3, Sch 40,
Pv = 332.4 kPa
P1 = 1032.4 kPa
P2 = 446.2 kPa
P = 586.2 kPa
Gf = 0.65
q = 193.2 m3/h
Results of valve sizing

calculations
Cv = 74.3
NPS 3, ANSI Class 300 globe valve, d = 76 mm

Full Area trim, Linear Characteristic
= (1032.4-332.4)/586.2 = 1.19

recommended based on
incipient damage for this
application.
Trim Style mr
Standard
2.0
Trim A
1.15
Trim B
1.002
26
ISA-RP75.23-1995
Trim A is checked for size

and pressure scale effects
Manufacturer's data for Trim A

(P1 - Pv)R = 620 psi; a = .20; dR=76 mm
PSE = (700.3/620)0.20 = 1.02
SSE = (d/dR)b = (76/76)b = 1.00
Calculate v for Trim A

(Equation 2)
Let R = mr = 1.15
v = [(1.15)(1.00)-1](1.02)+1 = 1.153
Evaluation
Trim A in a 3-inch globe has an acceptable v

(i.e., v).
7.6.6 Boiler feedwater application for globe valves (SI units)

T = 176.7 C
D = 193.5 mm
Gf = 0.89
q = 409.1 m3/h
Service data
Fluid: Water
Line Size: NPS 8, Sch. 80,
Pv = 931.0 kPa
P1 = 11 034 kPa
P2 = 10 344 kPa
Cv = 170
NPS 6, ANSI Class 900, globe valve, d = 146 mm

Reduced Trim, Equal Percentage Characteristic
= (11034 - 931)/690 = 14.6
Manufacturer's recommended
operating and scaling data
for incipient damage
mr = 2.5
(P1 - Pv)R = 690 kPa; a = 0.11;
dR = 76 mm; CvR/N1dR2 = 5.1
Calculate PSE (Equation 3)
PSE = [(11034-931)/690]0.11 = 1.34
Calculate b and SSE

(Equation 4, 5)
b = 0.068[170/(0.00155)(146)2]1/4 = 0.102
SSE = (146/76)0.102 = 1.07
Let R = mr = 2.5
v = [(2.5)(1.07)-1](1.34)+1 = 3.24
Valve inlet d = 146 mm

Pipe inside diameter D = 194 mm
KB1 = 1 - (146)4/(193.5)4 = 0.68
KB2 = 1 - (146)4/(193.5)4 = 0.68
K1 = 0.5[1 - (146)2/(193.5)2]2 = 0.093
K2 = [1 - (146)2/(193.5)2]2 = 0.185
K = 0.68 - 0.68 + 0.093 + 0.185 = 0.278
Fp = {1+(0.278)(170)2/[0.00214(146)4]}-1/2
Fp = 0.996
p =
(0.996)2{3.24+(0.093+0.68)(170)2/[0.00214(146)4]}
p = 3.24
(Equation 9)
(Equation 10)
(Equation 11)
(Equation 12)
(Equation 13)
(Equation 8)
(Equation 7)
ISA-RP75.23-1995
27
Evaluation
Preliminary selection is acceptable, because p (3.24)

is less than the operating (14.6). Also note that piping
effects are negligible for small values of Cv/N1d2.
8 Testing
8.1 Scope
This section provides a method of testing to determine the following control valve performance
characteristics in a cavitating fluid service. Section 6.2 describes how the test results of one
valve may be scaled to larger or smaller valves and other pressure conditions.
a) i (end of Regime I and beginning of Regime II), incipient cavitation coefficient;
b) c (end of Regime II and beginning of Regime III), constant cavitation coefficient; and
c) mv (end of Regime III and beginning of Regime IV), point of maximum vibration
cavitation coefficient .
NOTE The above cavitation coefficients are not intended to identify a point of unacceptable or damaging cavitation. Sections 5, 6, 7, and Annex B of this recommended practice
explain in detail their use and description.
8.2 Test system

8.2.1 General description
The flow test system shall be as shown in Figure 2. (Except for the cavitation detection
equipment, the flow test system is the same as that required for testing to ISA-S75.02, "Control
Valve Capacity Test Procedures.") It includes:
a) a test specimen;
b) a test section;
c) upstream and downstream throttling valves;
d) a flow measurement device;
e) pressure taps and measuring devices (upstream and downstream);
f) a temperature sensor; and
g) cavitation detection instrumentation shown in Figure 3.
28
ISA-RP75.23-1995
Test Section
Test Specimen
Fluid
Source
Temp.
Sensor
Accelerometer
Downstream
Throttling
Valve
Flow
Flow
Meter
Upstream
Throttling
Pressure
Taps
Valve
Cavitation
Detection
Equipment
Figure 2 Flow test system
Accelerometer
Measuring
Device
Voltmeter,
Oscilloscope,
or Spectrum Analyzer
Accelerometer
Output
Amplifier
Accelerometer
High Pass
Filter
Acceleration
Recording Device
Figure 3 Cavitation detection equipment
ISA-RP75.23-1995
29
8.2.2 Test specimen

The test specimen can be any valve or test apparatus for which test data are required. The initial
test specimen for any laboratory should be the calibration test section shown in Figure 4. The
test results from the test orifice plate shall be recorded and shall be within the calibration limits in
Section 8.6 to qualify the laboratory for testing to this recommended practice.
4X 90 o
.06
(1.5)
TYP
O1.534 + .001
(38.96 + .05)
O7.50
(190.5)
O3.068
(77.9)
O5.0
(127.0)
Sharp edge
free of nicks
and burrs
O6.0
(152.4)
.50
(12.7)
.125
(3.2)
4X O.75 (19)
60 o
NPS 3. Class 150 flanges
.06
(1.5)
6.0
(152)
Pressure tap
30.0
(762)
Pressure tap
Flow
NPS 3. SCH. 40
3.068 I.D.. (77.9) Pipe
Flat gaskets
Orifice plate
60.0
(1524)
36.0
(914)
NOTE: Dimensions are in inches (mm)
Figure 4 Cavitation calibration test manifold
30
ISA-RP75.23-1995
8.2.3 Test section

The test specimen upstream and downstream piping shall conform to the nominal size of the test
specimen connection and to the following length requirements.
The upstream pressure tap shall be two nominal pipe diameters from the test specimen
connection, while the downstream pressure tap shall be six nominal pipe diameters from the test
specimen connection. There shall be at least 18 nominal pipe diameters of straight pipe (eight if
straightening vanes are used) upstream of the upstream pressure tap, and at least one pipe
diameter of straight pipe downstream of the downstream pressure tap.
An effort should be made to match the inside diameter at the inlet and outlet of the test specimen
with the inside diameter of the adjacent piping.
NOTE For valves with Cv/N1d2 > 20, see Annex D.
8.2.4 Throttling valves
The upstream and downstream throttling valves are used to control the pressure differential
across the test section pressure taps and to maintain a specific downstream pressure. The
downstream valve should be of sufficient capacity to ensure that choked flow (in the case of the
standard calibration test manifold) and the other desired cavitation coefficients (i, c, mv) can
be achieved. Care should be taken to assure that noise or cavitation from these valves during
the testing does not influence the test specimen results.
8.2.5 Flow measurement
The flow measuring instrument may be any device that meets the required accuracy. This
instrument shall measure the true time average flow rate within an error not exceeding 2% of
the actual value. The resolution and repeatability of the instrument shall be within 0.5%.
8.2.6 Pressure taps
Pressure taps shall be provided on the test section piping in accordance with the requirements of
8.2.3 and shall conform to the construction described in ISA-S75.02.
8.2.7 Pressure measurement
All pressure and pressure differential measurements shall be made to an error not exceeding
2%.
8.2.8 Temperature measurement
All temperature measurements shall be made to an error not exceeding 2 F (1.1 C).
8.2.9 Accelerometer installation
An accelerometer shall be rigidly mounted on the test specimen downstream pipe wall. The main
sensitivity axis of the accelerometer shall be perpendicular to the pipe axis. The exact location
on the downstream pipe should be determined by test (to obtain maximum vibration sensing);
however, one nominal pipe diameter downstream of the test specimen connection is a good
starting point. As the higher frequency vibrations (5-50 kHz) are of interest in this testing,
mounting on the lower mass pipe wall (as compared to the flange mounting) will provide
improved sensitivity from a high frequency accelerometer.
8.2.10 Vibration measurement
Although this testing does not require accurate quantitative acceleration results (5% required), a
piezoelectric or other accelerometer that has a high enough resonant frequency (> 100 kHz)
ISA-RP75.23-1995
31
should be used. This will assure consistent results for the full frequency range being analyzed
(5-50 kHz). A vibration preamplifier should be used as recommended by the accelerometer
manufacturer. For ease of data analysis, an optional high pass filter can be used to differentiate
the low frequency noise (< 5 kHz) resulting from background and turbulent flow and the high
frequency noise resulting from cavitation. Figure 3 depicts suggested instrumentation.
The exact frequency range of interest for each valve type may vary, and it is the responsibility of
the valve tester to determine the optimum range of frequencies for evaluation.
8.2.11 Installation of test specimen
The alignment between the centerline of the test section piping and the centerline of the inlet and
outlet of the test specimen shall be within 1/16-inch (1.6 mm) for pipe sizes up to NPS 6 (150
mm), and within 1% of the diameter for pipe sizes NPS 8 (200 mm) and larger.
When rotary valves are being tested, the valve shaft shall be aligned with the test section
pressure taps. All gaskets should be positioned so that they do not protrude into the flow stream.
8.3 Test fluid

Water at relatively constant temperature shall be the basic fluid used in this test procedure. The
water shall be sufficiently free of suspended particles, air, or other gases so as not to affect the
test results. Successful completion of the laboratory qualification testing per Section 8.6 of this
procedure will verify the suitability of the test fluid. Inhibitors may be used to prevent or retard
corrosion or to prevent the growth of organic matter. The effect of additives on density, vapor
pressure, and viscosity shall be verified.
8.4 Test procedure

The procedure for evaluation of the data collected in these tests is given in 8.5.
8.4.1 The test section shown in Figure 2 shall be used with the test specimen (valve) set at a
specified travel. Travel positions of 20%, 40%, 60%, 80%, and 100% rated travel, as a minimum,
shall be used for a valve, and the following test should be run for each travel position. Since the
parameters being tested are a direct function of valve geometry, additional travel positions for some
valves may be needed. (Figure 1 only applies to a single open position.)
8.4.2 The downstream throttling valve shall be in the fully open position, or in a travel position that
provides for cavitating conditions in the test valve. Then, with a preselected upstream pressure,
the flow rate, upstream pressure, downstream or differential pressure across the test valve, and
accelerometer levels shall be recorded. The use of differential pressure measuring instruments is
preferred whenever possible to ensure accuracy. During a test run, P1 shall be held constant to
within 5%. This establishes a maximum pressure differential for the test valve in a test system
at a particular travel setting.
CAUTION DO NOT EXCEED THE MAXIMUM PRESSURE DROP RATING OF
THE VALVE BEING TESTED. CARE SHOULD BE TAKEN TO ASSURE THAT
GAS IS NOT TRAPPED IN THE DOWNSTREAM PRESSURE TAPS.
8.4.3 A series of additional tests shall be made at subsequently decreasing pressure drops by
throttling the downstream throttling valve. Each test should decrease the test specimen pressure
32
ISA-RP75.23-1995
drop by suitable increments to detect inflection points in the vibration curve; and the same recordings of flow, pressure, and acceleration shall be made.
8.4.4 The following data shall be recorded:
a) Valve travel measurement error shall not exceed 0.5% of rated travel;
b) Upstream pressure (P1) instrument measurement error shall not exceed 2% of
the actual value;
c) Pressure drop (P) (preferred) or downstream pressure (P2) instrument
measurement error shall not exceed 2% of the actual value;
d) Measured flow rate (q) measurement error shall not exceed 2% of the actual value;
e) Fluid temperature (T) measurement error shall not exceed 2 F (1.1 C). Care
should be taken to monitor and record changes in fluid temperature so that vapor
pressure can be properly determined for all test points. This is especially important for
recirculating test systems in which fluid temperature may increase during a test.
f)
Downstream pipe wall vibration measurement error shall not exceed 5% of full
scale. Instrument full scale should be selected for optimum measurement of the range
of vibration amplitudes.
g) Barometric pressure measurement error shall not exceed 2% of the actual value.
Convert to absolute pressure (Pa), psia (kPa).
8.5 Data evaluation

For each test:
a) Evaluate Cv and FL in accordance with ISA-S75.02 or Annex D, as applicable.
b) Calculate sigma cavitation index for each test point, preferably using differential pressure
measurement data.
= (P1 - Pv) / (P)
(14)
c) Calculate the average acceleration in G, ft/s2 (m/s2), or equivalent to as high a frequency

as permitted by the instrumentation within the frequency range 5-50 kHz. These data
can be obtained from a frequency analyzer, oscilloscope, or voltmeter (with a high-pass
filter).
d) Obtain cavitation coefficients by making a log/log (or semi-log) plot of acceleration
versus sigma () for each test made at a particular travel. A separate plot should be
made for each travel position tested.
From these plots, determine cavitation curve inflection points i, c, and mv where
possible by intersection of the straight line segments, Regimes I, II, III, and IV, as
shown in Figure 1. Note that some valves will not exhibit all inflection points in the
test data which are always subject to expert interpretation; and some multistage
valves may not provide meaningful data by this test method alone.
ISA-RP75.23-1995
33
8.6 Laboratory qualification

8.6.1 To assure that a particular laboratory can successfully test a piece of equipment to this
procedure, a qualification test is recommended. The test procedure and data evaluation of 8.4
and 8.5 shall be performed on the calibration test manifold shown in Figure 4, using 3-inch test
section piping. A P1 of 100 psig (690 kPa) is recommended as a test pressure. The resultant
sigma values shall be within the following acceptable ranges to qualify the laboratory and equipment.
i
= 2.7 5%
= 2.3 5%
mv = 1.4 25%
Cv
= 52 5%
FL
= 0.86 5%
8.6.2 When use of the 3-inch piping manifold is impractical, alternate pipe sizes (having inside
diameter D1) may be used for calibration test manifolds if the following manifold design conditions
are met:
a) the ratio of orifice diameter to pipe inside diameter is 0.50 1%;
b) a minimum of twenty (20) nominal pipe diameters of straight pipe is provided upstream
of the orifice;
c) a minimum of twelve (12) nominal pipe diameters of straight pipe is provided downstream
of the orifice;
d) upstream and downstream pressure taps are located two (2) nominal pipe diameters
upstream and ten (10) nominal pipe diameters downstream of the orifice; and
e) the orifice is a conventional sharp-edged design, (similar in construction to the orifice
in Figure 4, to match with manifold piping), thick enough to resist vibration and pressure
forces from cavitation testing, and approximately centered in the pipe.
When other calibration manifold sizes are used for testing, size scale effects must be considered
(Equations 2 through 5), assuming PSE = 1.00. The alternate piping test values shall be within
the following acceptable ranges (which include adjustment for size scale effect) to qualify the
laboratory and equipment.
i
= 2.7[(D 1/3.068N3)0.104] 5%
(15)
= 2.3[(D 1/3.068N3)0.104] 5%
(16)
mv
= 1.4 25%
Cv/N1D 12 = 5.52 5%
FL
34
= 0.86 5%
ISA-RP75.23-1995
Annex A References
1.
Knapp, R.T., Daily, J.W., and Hammitt, F.G., Cavitation, McGraw-Hill, New York, 1970.
2.
Tullis, J.P., "Cavitation Scale Effects for Valves", Journal of the Hydraulics Division,
ASCE, Vol. 99, No. HY7, p. 1109, 1973.
3.
Rahmeyer, W., "Cavitation Sizing and the Prediction of Cavitation in Control Valves",
Valve Manufacturers Association of America, Valve Industry Technical Seminar,
Washington, D.C., November 1987.
4.
Tullis, J. Paul, Hydraulics of Pipelines: Pumps, Valves, Cavitation, Transients, John

Wiley and Sons, Inc., 1989, ISBN 0-471-83285-5.
5.
Rahmeyer, W. and Odeh, M., "Prediction Hydrodynamic Noise from Cavitating Valves",
American Society of Mechanical Engineers, paper No. 88-PVP-23, June 1988.
6.
Stripling, T., "Cavitation Damage Scale Effects: Sudden Enlargements", Ph.D.

Dissertation, Colorado State University, Fort Collins, Colorado, 1975.
7.
Cain, F.M. and Barnes, R.W., "Testing for Cavitation in Low Pressure Recovery Control
Valves", ISA Transactions, Vol. 25, No. 2, Instrument Society of America, 1986.
8.
Knapp, R.T. and Hollander, A., "Laboratory Investigations of the Mechanism of

Cavitation", Transactions of the ASME, Vol. 70, 1948.
9.
Hutchison, J.W. (ed.), ISA Handbook of Control Valves, 2nd ed., Instrument Society of
America, Pittsburgh, PA, 1976.
10.
Riveland, M.L., "The Industrial Detection and Evaluation of Control Valve Cavitation",
ISA Transactions, Vol. 22, No. 3, 1983.
11.
Ball, J.W. and Tullis, J.P., "Cavitation in Butterfly Valves", Journal of the Hydraulics
Division, Vol. 99, No. HY9, p. 1303, 1973.
12.
Rahmeyer, W.J., "Cavitation Testing of Control Valves", ISA 0-87664-781-6/83/1253-9,

Instrument Society of America, 1983.
13.
Barnes, R.W. and Cain, F.M., "Proposed Universal Method for Quantifying the Severity
of A Cavitation Control Valve Service", Second International Conference on
Developments in Valves and Actuators for Fluid Control, BHRA, Manchester, England,
March 28-30, 1988 and Addendum Feb. 15, 1989.
14.
Ivany, R.D., and Hammitt, F.G., "Cavitation Bubble Collapse in Viscous, Compressible
Liquids - Numerical Analysis", Journal of Basic Engineering, ASME, pp. 977-985, 1937.
15.
Mousson, J.M., "Pitting Resistance of Metals Under Cavitation Conditions", ASME

Trans, Vol. 59, pp. 399-408, 1937.
16.
Hammitt, F.G., Cavitation and Multiphase Flow Phenomena, McGraw-Hill, New York,
NY, 1980.
17.
Stepanoff, A.J., "Cavitation in Centrifugal Pumps with Liquids Other Than Water",
Journal of Engineering for Power, ASME, Vol. 83, Series A, pp. 79-90, 1961.
ISA-RP75.23-1995
35
18.
Rahmeyer, W. and Driskell, L., "Control Valve Flow Coefficients", American Society of
Civil Engineers, Pipeline Division, Journal of Transportation Engineering, Vol. 111, No.
4, July 1984, pp. 358-64.
19.
Cain, F.M., Correspondence to ISA-SP75.16, Subject: dRP75.23, Derivation of p; July

19, 1991.
20.
Kirik, M.J. and Driskell, L., Flow Manual for Quarter-Turn Valves , Rockwell International
Corp., Flow Control Division, Pittsburgh, PA, 1986.
21.
IEC Publication 534-8-2, Laboratory Measurement of Noise Generated by Liquid Flow

Through Control Valves, International Electrotechnical Commission, Geneva,
Switzerland, 1991.
22.
NRC Publication NUREG/CR-6031, Cavitation Guide for Control Valves, U.S. Nuclear
Regulatory Commission, Washington, D.C., 1993.
36
ISA-RP75.23-1995
Annex B Cavitation fundamentals
B.1 Introduction
B.1.1 Cavitation is a subject that has been of interest to both the theorist and the industrial
practitioner for nearly a hundred years. Unfortunately, many aspects of the cavitation process
remain a frustrating mystery in spite of intense study during this time. Theory has supplied much
understanding of the behavior of single cavities as well as the nature of the influence of many
variables. Likewise, contending with cavitation in pumps, valves, propellers and hydrofoils has
spawned many "rules-of-thumb" governing prediction and control practices. However, the
state-of-the-art does not currently offer a satisfactory technology that bridges the gap between
theory and practice.
B.1.2 The purpose of this section is to provide more insight into the cavitation events from both
a theoretical and a practical perspective. This level of familiarity provides a broad foundation for
dealing with cavitation in the process control industry. Not only does it supply the generally agreed
upon quantitative methods, it also establishes a background on which to base the inevitable engineering judgments that characterize this type of application.
B.2 Cavity behavior

B.2.1 Cavity dynamics are governed by changes in fluid pressure and can be very complex. An
analysis of even a single cavity in a well defined environment requires methods of thermodynamics,
heat transfer, mass diffusion, fluid mechanics, statics, and dynamics. A detailed discussion is
obviously not in keeping with the overview nature of this document. However, some of the salient
points on the subject of cavity inception, growth, and collapse are presented below.
B.2.2 The pressure level conventionally associated with the onset of vaporization of a liquid is
the thermodynamic vapor pressure. This model holds that a portion of the liquid will vaporize when
the prevailing fluid pressure decreases below the vapor pressure. This vaporization occurs at a
"weak" spot within the liquid continuum. This weak spot is usually associated with a free surface
within the liquid that was created by an entrained foreign particle, microscopic gas bubble or other
such "nucleus." In the absence of such nuclei, vaporization would not occur at the thermodynamic
vapor pressure of the fluid. Very large fluid forces would be required to overcome the effects of
surface tension at the radius of an infinitely small bubble within a pure liquid.
B.2.3 Since nuclei are required to promote vaporization at the vapor pressure of the liquid, the
subject of nuclei content naturally arises. In most industrial applications an ample supply of such
nuclei are present in the form of various contaminants. Even in the case of highly purified liquids,
small amounts of gas can be stably entrained in the liquid and the surrounding solid surfaces. Thus,
for practical intents and purposes the fluid can be assumed to vaporize at the vapor pressure of
the liquid.
B.2.4 This model assumes thermodynamic equilibrium and, therefore, is accurate only when the
pressure change rate is slow enough that sufficient time exists for exchange of the various thermal
energy quantities. When the pressure changes are very large and occur very rapidly, it is possible
for non-equilibrium or "meta-stable" thermodynamic states to exist. For the specific case at hand,
ISA-RP75.23-1995
37
this would result in the fluid existing in the liquid state under pressure and temperature conditions
that ordinarily correspond to a saturated state.
B.2.5 When non-equilibrium conditions occurs, the fluid pressure may actually drop below the
vapor pressure of the fluid with little or no vaporization occurring. As the thermal dynamics "catch
up" to the inertial dynamics, vaporization occurs very rapidly (physically displaced slightly downstream from the location where the local pressure actually drops below the vapor pressure).
B.2.6 A forming cavity experiences a period of stable initial growth. During this growth phase
the force balance on the cavity is such that energy is required for sustained growth. However, when
the cavity attains a "critical" size, this growth is no longer stable. Now the force balance is such
that energy is liberated on additional growth. Under these circumstances the growth is "explosive"
(Knapp, et al., ref. 1).
B.2.7 If the fluid pressure should rise above the vapor pressure, the reverse process will occur;
i.e., cavity growth will cease and collapse of the cavity and condensation of the vapor will occur.
This latter event occurs very rapidly and is the primary source of the noise, vibration and material
damage. In some cases additional growth and collapse cycles, known as rebound, occur.
B.2.8 At this point it is appropriate to mention another phenomenon, which is outwardly similar
to cavitation but requires distinction. This is the phenomenon of "out-gassing." Pockets of gas may
form when dissolved gases (and possibly the fluid's own vapors) come out of solution upon depressurization of the fluid. Sometimes this is referred to as "gaseous cavitation." It may occur
separately from the aforementioned type (i.e., vaporous cavitation) or in combination with it. The
behavior of this type of cavity formation is markedly different. It generally is considered less
detrimental, because the growth, and particularly the collapse rates, are significantly reduced.
B.3 Damage mechanisms

B.3.1 Material damage is a side effect of cavitation that receives considerable attention for obvious reasons. Several theories have been offered to explain the observed levels of damage;
however, no single mechanism is currently considered universal. It appears that all of the identified
theories apply to varying degrees.
B.3.2 The microjet theory asserts that vapor cavities collapse asymmetrically and, in doing so,
set up a very small, high velocity liquid jet (Knapp & Hollander, ref. 8). This jet, if it impinges on a
material surface, imparts damage by eroding away a portion of the material. This theory has been
substantiated with both high speed photographic observations of collapsing cavities, and also
numerical studies of individual cavity collapses. Interestingly, the preferred orientation of this jet
is influenced by the nature of the various boundaries in close proximity to the jet. Rigid surfaces
tend to attract the microjet, while compliant surfaces appear to repel the microjet (Knapp, et al.,
ref. 1). This appears to be a dominant damage mechanism on the initial collapse of travelling
cavities.
B.3.3 Another theory, the shock wave theory, advances the notion that shock waves are established by the rapid movement of the vapor-liquid interface during collapse. The compressive loading
from shock wave impingement on the material surface can result in failure by either fatigue (multiple
loadings) or plastic deformation (single loading). This appears to be characteristic of the symmetrical collapse mode, typical of rebound collapses.
B.3.4 These two forms of cavitation attack constitute the mechanical component of attack. The
damage resulting from either of these types of mechanical attack possesses some similar attributes.
38
ISA-RP75.23-1995
The physical appearance of cavitation damage varies from a "frosted glass" appearance to a rough,
cinder-like surface texture. Also, the collapsing cavity must be near the material surface to cause
any damage, regardless of the exact mechanism of damage. Both the shock wave intensity and
the microjet dissipate to a harmless level within a fairly short propagation distance.
B.3.5 Finally, chemical attack or corrosion can occur simultaneously with mechanical attack.
This form of attack usually is considered secondary. It is important in that it tends to accelerate or
reinforce the mechanical attack process. The extent to which this form of attack is a factor depends
on the chemical compatibility of the process fluid and the materials of valve construction.
B.3.6 The combined effects of mechanical attack, chemical attack, and the particular material
of construction give rise to an "incubation period" in some circumstances. This is a period of time
(that varies with changing conditions) during which no material loss is apparent. Plastic deformation
or mass loss then commences, even though the cavitation intensity remains constant.
B.4 Pressure dynamics that cause cavitation

B.4.1 The pressure changes that cause cavitation arise from the flow of the liquid through the
control valve. However, pressures sufficiently low to cause cavitation may result from different
sources. The conventional industrial treatise usually accounts for only the mean fluid pressure.
Pressure changes associated with various boundary layer phenomena also play an important role
in industrial cavitation.
B.4.2 Cavitation in control valves begins from highly localized pressure reductions typically associated with boundary layer separation. Flow streams in control valves are characterized by
irregularly shaped passages, abrupt cross-sectional area changes and various protrusions. These
features subject the fluid to adverse pressure gradients that in turn cause the flow stream to
separate from the bounding surface. The resulting high turbulence and re-entrant vortex formation
produce very low pressures that initiate cavitation. This can occur even though the mean fluid
pressure at the vena contracta is greater than the fluid vapor pressure.
B.4.3 Mean fluid pressure changes are associated with changes in the kinetic energy of the fluid
as it passes through the valve. The fluid velocity increases as the liquid moves into the reduced
area of the throttling restriction. This causes a corresponding decrease in the mean fluid pressure.
Conversely, when the fluid exits the throat region into a larger cross-sectional flow area, the velocity
decreases and the mean fluid pressure is partially restored. Thus, a minimum pressure exists in
the valve in the general vicinity of the minimum flow area. If this mean pressure falls below the
vapor pressure of the fluid, the degree and extent of cavitation increases.
B.4.4 Both of these sources of cavitation (i.e., mean pressure changes and localized pressure
changes) are important. It cannot be categorically stated that one dominates the other throughout
the cavitating flow regime. The total behavior of the cavitating liquid is heavily dependent on valve
style and other factors. This document will not distinguish between the origins of cavitation in
further discussion of the subject matter.
B.5 Cavitation parameters

B.5.1 As noted in Section 5, the complexity of cavitation and its sensitivity to a variety of factors
(discussed later) render it difficult to quantify cavitation in terms of a single or few parameters.
Nonetheless, it is necessary to construct a mathematical representation of the application in order
to meaningfully deal with it during process design. In the case of control valves it is conventional
ISA-RP75.23-1995
39
to describe the application in terms of a single parameter and then to compare this parameter to
different limits of operation for a given control element.
B.5.2
The parameter selected for use in this document was given in Section 5 as
= (P1 - Pv)/(P1 - P2)
(B.1)
where, P1 is the upstream absolute pressure of the valve, P2 is the downstream absolute
pressure of the valve, and Pv is the absolute vapor pressure of the fluid at the inlet temperature.
For high recovery valves with Cv/N1d2 > 20, see Annex D.
B.5.3 The cavitation index sigma () is a form of another dimensionless parameter, the Euler
Number. Sigma is constant for either SI or US Customary units as long as the same units of
pressure are used throughout the equations for the parameter. This index () is the ratio of fluid
forces trying to prevent cavitation (the system or service pressure) to the forces trying to cause
cavitation (the pressure drop). The smaller the value of the cavitation index for a flow system, the
more likely or the more severe cavitation will be.
B.5.4 Different cavitation parameters have been used for defining cavitation in control valves
(Hutchison, ref. 9; IEC, ref. 21). Parameters such as K or xF may be defined as
K = xF = 1/ = (P1 - P2)/(P1 - Pv)
(B.2)
The operational limiting value of K is often referred to as Kc (Hutchison, ref. 9). This limit was
originally defined as the set of conditions where cavitation will begin to affect the flow pattern in
the valve. This manifests itself as apparent decrease in the flow coefficient of the valve. Some
publications have referred to this as the incipient cavitation point. However, for many valves,
cavitation damage, heavy vibration, and noise could occur for flows and pressure drops less than
that for this definition of Kc (NRC, ref. 22). Another parameter used to indicate incipient cavitation
for the purpose of predicting noise levels is xFz (IEC, ref. 21). While xFz is approximately
equivalent to 1/i evaluated for the same valve, size, opening, and test pressure, equations
contained in this document are rigorously developed only for i. However, for purposes of
evaluating control valves for cavitation, the value of 1/xFz may be used for i. Users are advised
to check with the valve manufacturer for the definition and proper use of any cavitation parameter.
B.5.5 Sigma () is recommended because of numerous performance correlations published
using (refs. 1-7, 11-13, 16, 20, 22) and because of the confusion that exists both nationally and
internationally over the use and definitions of the K-type parameters. Sigma will typically range
from 1 to 15. As a rule of thumb, a valve with service conditions corresponding to a sigma value
greater than 15 for low pressure loss valves or greater than 8 for high pressure loss valves usually
will not experience cavitation.
B.5.6 Alternate forms of the sigma index also exist. One particular expression that has been
used in the past is based on the relationship of the vapor pressure to the downstream pressure
(as opposed to the upstream pressure). The following equation can be used to convert between
these different forms of :
= 1 + 2
(B.3)
where
P2 Pv
2 = -----------------P1 P2
40
ISA-RP75.23-1995
B.5.7 The pressure recovery factor, FL, is a coefficient that is useful for predicting the choked
flow rate through a control valve. However, because of its relationship to the pressure recovery
characteristics of the valve, it also has been used as a cavitation limit. This does not constitute a
correct use of this factor. In fact, when the pressure drop meets or exceeds the pressure drop
calculated from FL, substantial levels of cavitation can already exist. The significance of this
parameter to the analysis of control valve cavitation is discussed further in B.6.3 and B.8.4.
B.6 Levels of control valve cavitation

B.6.1 The cavitation index, , by itself does not convey any information about the performance
of a particular valve in a particular application. Different valves can tolerate different levels of
cavitation, and different applications are concerned about different aspects of cavitation (for instance noise versus vibration versus damage). Therefore, must be evaluated at the service
conditions (service ) and then compared to some benchmark for the valve that reflects the permissible degree of cavitation (limiting ) for the application.
B.6.2 The degree and extent of cavitation in a control valve will be a function of many things.
Clearly, it is a function of the valve style and valve opening since these factors directly affect the
character of the flow through the valve. It follows that the desired benchmarks (limiting coefficients) will be unique to a given valve style and opening, if not a particular valve. A "good" benchmark will delineate a meaningful cavitation threshold as well as be measurable. However, in the
quest for practical benchmarks one or the other of these two characteristics should not be overemphasized. It is possible to define a level of cavitation that is convenient to measure but has little
meaning to valve selection. On the other hand, it may be desirable to have a benchmark that
predicts a particular event, such as damage, but is difficult to accurately measure or predict.
B.6.3 The following levels of cavitation, which often serve as benchmarks for the valve selection
process, are believed to represent a good blend of measurability and meaning. An additional
parameter, which can be based on valve manufacturers experience with a particular valve, is also
included. The corresponding coefficients are shown in parenthesis.
B.6.3.1 Incipient cavitation (i)
The first level of cavitation is incipient cavitation, i. This level is associated with the flow
conditions for which cavitation can first be detected. A typical means of establishing this
condition is by measuring increased SPL (IEC, ref. 21) or vibration (Tullis, ref. 2; Riveland, ref.
10). Incipient cavitation is intermittent and usually is the result of a few isolated collapses of vapor
pockets very near the source of the flow separation. Incipient cavitation is extremely mild, and
often cannot be heard over the flow noise and vibration produced by other components in a
piping system.
B.6.3.2 Constant cavitation (c)
Constant cavitation, c, is one of the earliest levels of cavitation in which the intensity has
noticeably increased above the incipient level (Ball and Tullis, ref. 11). At this point the cavitation
involves a sufficiently large volume of vapor to produce a uniform and constant level of cavitation
that is readily detected. The collapse region is beginning to move downstream from the point of
cavitation inception within the valve. The observable effects of constant cavitation are light
vibration and mild noise levels. Usually there is no undesirable damage associated with this level
of cavitation. However, as the flow rate is increased, cavitation increases above this level at a
steady and moderate rate. Constant cavitation can be determined in a flow laboratory by
comparing the relationship of the acceleration or vibration produced by cavitation with different
flow rates.
ISA-RP75.23-1995
41
B.6.3.3 Incipient damage (id)

Incipient damage, id, is the level of cavitation in which an increase in cavitation intensity first
produces any detectable damage to either the valve or the downstream piping. This level is
much more difficult to detect. It cannot be determined from an acceleration or vibration curve as
other points might be. A suggested method is to measure the pitting rate of damage on samples
of soft materials such as aluminum (Rahmeyer, ref. 12). The pitting rate of one pit per minute per
square inch in aluminum is a good limit since it represents the most mild and conservative stage
of cavitation damage. The rate and magnitude of damage have been found to rapidly increase
with velocity after the level of incipient damage in aluminum.
B.6.3.4 Choking cavitation (ch)
"Choking" cavitation, ch, is a most severe level of cavitation. Under "fully choked" flow
conditions an additional decrease in the downstream pressure will not increase the flow rate
through the valve at a given inlet pressure. The value of associated with this condition may be
estimated from the following equation:
( P 1 Pv )
ch = ---------------------------------------2
FL ( P 1 F F P v )
(B.4)
This relationship is only an approximation of the fully choked condition. In control valve sizing
this condition usually is associated with the following pressure differential:
Pchoked = (FL)2 (P1 FFPv)
However, the choking "process" actually occurs over a range of pressure drops as a
consequence of the compressibility changes accompanying vapor formation. The fully choked
condition may actually occur at a pressure differential slightly larger than that calculated by this
equation.
The maximum levels of noise, vibration, and material damage have been observed to occur at or
just prior to this condition. This level of cavitation usually is avoided in most applications.
However, consideration of other offsetting factors such as trim design or reduced flow rate, size
or pressure may justify operation at this level without significant problems. These factors are
discussed later.
B.6.3.5 Maximum vibration cavitation (mv)
Maximum vibration cavitation, mv, is a level of cavitation corresponding to the "peak" of the
characterizing vibration curve. As noted in the preceding paragraph, maximum vibration, noise,
material damage, and flow rate all have been observed to occur very close to the same value of
. However, this is not to imply that they are necessarily coincident points. Use of a noise-based
laboratory cavitation detection method may yield a slightly different maximum noise coefficient.
Categorically, these points may be utilized in somewhat similar fashion; they all represent a
severe level of cavitation that is ordinarily avoided. On occasion, however, certain mitigating
factors may justify operation in this region. As noted above, these factors are discussed later.
B.6.3.6 Manufacturer's recommended limit (mr)
Manufacturer's recommended limit, mr, is an operational limit supplied by the valve
manufacturer. The evaluation of this point may be based on factors other than a single laboratory
test, such as accumulated experience with a particular design in a particular application or
understanding of specific design features. It may or may not coincide with one of the levels of
cavitation already discussed. This mr limit concept has been introduced to allow the most
42
ISA-RP75.23-1995
effective use of available control valve hardware and knowledge. The valve user should seek
clarifications from the valve manufacturer of the information used in this evaluation.
B.6.4 The general evaluation of these different levels of cavitation is the subject of other sections
of this document. (See Sections 6, 7, and 8 for more information.)
B.7 Control valve design and application factors

B.7.1 Ostensibly, the sizing procedure would appear complete with the introduction of the various
cavitation levels and their coefficients. However, in actuality, several issues still confront the potential user. Examples of these issues follow:
a) What factors affect the choice of mr limits?
b) What factors might affect the actual value of cavitation coefficients?
c) What factors might affect the value of the apparent operating index?
B.7.2 The different levels of cavitation described in the preceding paragraphs merely define
different significant cavitation conditions that exist. The appropriate coefficient to use for a given
application is not always self-evident and usually embraces a degree of subjectivity. In addition to
the service conditions, factors such as valve style and opening, duty cycle, location, desired life
and past experience should be considered (Barnes and Cain, ref. 13). There are no firm guidelines
for this decision, and the valve manufacturer should be consulted in this matter. Additional discussion may be found in Section 7 and Annex C of this document.
B.7.3 Factors that affect the numerical values of the coefficients established by test, and subsequently the level of cavitation, may be broadly grouped into valve factors (associated with the style
of the control valve) and service factors (e.g., pressure, temperature, fluid, and flow rate).
B.7.4 Accounting for all of these various effects can be difficult and controversial. The extent of
cavitation-related problems is a function of the total amount and intensity of the cavitation formed
in the flow stream, and of the response of the surroundings (material surfaces, acoustic field, etc.)
to that cavitation. Research has shown that many factors contribute to the total nature of cavitation
and to the resulting problems.
B.7.5 To aid in understanding the nature of some of these effects, it is helpful to divide them into
two categories that will be labeled "scale effects" and "influences." Scale effects are factors that
change the numerical value of any of the defined cavitation coefficients as that factor is varied.
"Influences," on the other hand, are those factors that do not change the numerical value of the
defined cavitation coefficients, but may affect the intensity of cavitation or cavitation damage. The
differences in these two types of effects are depicted in Figures B.1 and B.2. Knowledge of potential
factors in either category is necessary in both the testing and application phases. Thus, it is
necessary to sufficiently standardize the testing procedure and supply correcting equations for
application purposes where possible. In spite of the efforts and advancements of research in this
area, a degree of subjectivity still is present in the application process.
B.7.6 Two very important scale effects are associated with the cavitation parameter : the
pressure scale effect, PSE, and the size scale effect, SSE. The various cavitation coefficients are
not independent of or constant with either different upstream pressures or the valve size. Failure
to account for the difference between the actual service pressure and size and the test values
(reference values) can introduce significant errors. Thus, it becomes important to conduct all
evaluation tests at a prescribed or known reference condition and scale the values to the service
conditions with a known scaling equation.
ISA-RP75.23-1995
43
B.8 Effects on by size and other control valve geometry

B.8.1 The various cavitation coefficients, as well as the suitability of the valve to operate at a
given level, are strong functions of the geometry of the control valve. It can be expected that
different values of any given coefficient will be associated with different valve styles, or even different
openings of the same valve. Further, it can be expected that different levels of cavitation are
appropriate limits for different valve styles. Some of the major factors that have an effect on
values are discussed below.
B.8.2 Perhaps one of the most self-evident geometric features is the physical size of the valve.
There are strong size scale effects associated with most of the various cavitation levels discussed
earlier. In general, as the size of the valve increases, the numerical value of the cavitation coefficient
increases for a true scale model. An understanding of the general behavior of broad classes of
control valves has lead to the development of scaling equations. Thus, if cavitation coefficient
information is available from test data for a given valve, application limits can be conditionally
estimated for different size valves of the same style (Tullis, ref. 2).
B.8.3 Size scale effects need to be accounted for only when a reduced scale prototype has been
used to establish the various cavitation coefficients for the full size valve. It is important that scaling
is done from an actual scale model of the valve in question. Use of a smaller size valve model of
a particular valve style line will introduce geometric dissimilarities. These may have different effects
in the installed valve performance, which are not reflected in the scaled results. Additional information on these considerations is supplied in Section 6.
B.8.4 Another characteristic that is different from valve to valve is the pressure recovery characteristic. As noted previously, this often is associated with cavitation. This is not altogether
inappropriate, but too much emphasis often has been placed on this one feature. Qualitatively,
pressure recovery characteristics help distinguish valves inherently better suited for cavitating
applications. Low recovery valves maintain higher overall pressures within the valve body under
otherwise equal conditions. Therefore, the tendency for the valve to cavitate, and the intensity of
any cavitation that may occur are both reduced.
B.8.5 The degree of pressure recovery is heavily dependent on the valve geometry and is characterized by the pressure recovery factor, FL. Note, however, that this is not a cavitation index.
The singular quantitative use of this parameter is to determine the maximum flow (i.e., choked flow)
for the valve under a given set of conditions.
B.8.6 Trends in the pressure recovery characteristic usually are reflected in the other values
accordingly. High recovery valves have the effect of relatively increasing the numerical value of
corresponding levels of cavitation compared to a low recovery valve. However, no verified functional
relationship between the cavitation coefficients and FL exists, with the exception of that between
ch and FL. Actual coefficients should be obtained for each valve and should not be predicted
from the pressure recovery factor.
B.8.7 Another valve geometry consideration, which has a pronounced effect on the magnitude
of the various values, is the shape and number of primary restrictions in the control valve. Highly
irregular cross-sectional flow areas (e.g., such as the "lens shape" associated with certain rotary
valves) may result in substantial amounts of cavitation at lower operating sigmas. Cavitation may
begin in one localized region (for the reasons cited earlier) and gradually develop. Ultimately the
cavitation will extend throughout most of the entire flow area. The compressibility effects of the
forming vapor on the overall flow characteristics may be seen in the associated flow curve
44
ISA-RP75.23-1995
(Figure B.3). Typically such valves begin departing from the "straight line" portion of the curve at
much lower pressure differentials relative to the choked flow point.
B.8.8 The typical butterfly valve geometry creates two flow paths of differing geometries. Each
flow path possesses individual flow parameters and passes flow accordingly. One path may choke
prior to another and give rise to a "segmented" flow curve as shown in Figure B.4. The slope of
the flow curve is proportional to the magnitude of the Cv value associated with the valve. The steep
slope of the first segment represents the full flowing capability of both flow paths. The second
segment has a slightly lower slope as a result of one flow path choking, and subsequently flowing
a fixed amount regardless of the pressure drop. The third segment has zero slope, since both flow
paths are choked and flowing fixed volumes of liquid. "Safe" operating limits may be tied to the
flow path that cavitates first and may consequently be lower than anticipated when considering the
choked flow limit. The degree of this effect will be a function of the valve opening.
B.8.9 Certain control valves capitalize on particular geometric effects and are designed especially for cavitation service. These designs may include special or proprietary features that enhance
the numerical values of the various cavitation levels and the acceptable operating limit. Other
times they control where and to what extent the cavitation occurs. This allows a less conservative
limit to be used in the valve selection process. For instance, consider the case when c is chosen
as an appropriate operating limit when considering a butterfly valve for a particular application.
However, the use of a valve specially designed to control cavitation may allow the choice of ch as
the limit for that same application.
B.8.10 A note of caution is appropriate at this point. The values of the various cavitation levels
are clearly a function of the valve style and valve opening. To a certain extent this can be exploited
when selecting valves for specific applications. However, the user must resist the temptation to
arbitrarily change styles and sizes to effect favorable numbers. For example, the use of a large
butterfly valve at small openings in lieu of smaller valves near wide open could result in additional
damage to seating and shutoff surfaces from throttling too close to the seat. More factors to
consider are discussed in the sections that follow.
B.9 Pressure scale effects and other application influences on

B.9.1 The foremost scale effect associated with specific application service conditions is the
pressure scale effect (PSE). Similar to the size scale effect, the actual value of the various
cavitation levels will change as the operating pressure changes. In general, as the service pressures increase, the numerical value of the cavitation level also increases for a given valve.
Empirical scaling equations based on general behavior of broad classes of control valves are
available. Thus, if values are available from test data at a specific pressure, application limits
can be estimated for other service conditions.
B.9.2 Other factors, particularly fluid properties, can have an effect on the degree and extent of
cavitation, even though the numerical value of a cavitation level is unchanged. Several prominent
considerations follow:
a)
b)
c)
d)
e)
viscosity
dissolved gas volume
surface tension
"thermal" properties
fluid composition
ISA-RP75.23-1995
45
B.9.3 The influence of viscosity usually is considered to be one of damping, although for most
practical applications the effect is negligible (Ivany & Hammitt, ref. 14). Again, the numerical value
of a particular will not change. No analytical or empirical method for accounting for viscosity has
been developed. However, high viscosity fluids are not likely to present significant cavitation
problems.
B.9.4 Undissolved gases can have a pronounced influence on the behavior of cavitation. Small
additional amounts of undissolved gases provide additional nuclei for the inception process and,
in theory, thus result in "additional" cavitation (Knapp et al., ref. 1). However, larger amounts (i.e.,
a few percent by volume) tend to suppress the effects of cavitation by disrupting the collapse
process (Mousson, ref. 15). As emphasized in the section on testing, it is necessary to assure
minimal air content so that the test results are not obscured by this influence.
B.9.5 The surface tension term appears in the theoretical equations that describe the cavity
behavior. It acts to restrain nucleation and growth, yet promotes cavity collapse. The former effect
tends to diminish the total of cavitation that occurs and is a mitigating effect. The latter has the
opposite effect by intensifying the collapse. It is likely that the magnitude of the restraining effect
would dominate (Hammitt, ref. 16); however, for most fluids of practical interest, the total effect
probably is negligible.
B.9.6 Most thermal property influences are accounted for by utilizing the absolute vapor pressure
(at service condition temperature) in the sizing and scaling equations. There are some additional
effects that are less quantifiable. When the thermal energy terms are negligible compared to the
inertial terms, the pressure dynamics govern the rate at which vaporization and collapse occur.
When the thermal terms become significant, the time associated with the rate of heat transfer can
become the controlling factor, and the process becomes more like boiling (Stepanoff, ref. 17).
However, most fluids of industrial interest usually do not fall into this category.
B.9.7 The discussion to this point has concerned fluids that are comprised of a single chemical
species. Many process fluids are actually mixtures of different fluids. The fluid composition of
such mixtures can have a direct bearing on the degree to which cavitation-related problems occur.
When the process fluid consists of a mixture of fluids with widely varying vapor pressures, the
classical "single fluid" model breaks down. Vaporization occurs over a range of pressuresas
opposed to the constant pressure, constant temperature vaporization of a "pure" liquid. Furthermore, the compositions of both the vapor and liquid phases change as a result of this fractionation.
The combined effect is, in general, to reduce the severity of cavitation-related problems. A good
example of a mixture of fluids is a broad mixture of hydrocarbons.
B.10 Closure
B.10.1 Cavitation will continue to be a major problem in industrial process control. An understanding of the nature of the subject and utilization of current quantitative information will aid in formulating
effective problem abatement. However, the ultimate benefit of analyzing valves and service conditions for cavitation depends upon the quality and completeness of service and valve information
available. Valve users and manufacturers should make every reasonable effort to share clear and
accurate data (see ISA-RP75.21, Process Data Presentation). These data will make possible the
comparisons between the service conditions and valve capabilities.
B.10.2 The procedures contained in this document are intended to provide the best, practical
knowledge currently available on the subject. However, practitioners should always avail themselves of proven new technology as it becomes available.
46
ISA-RP75.23-1995
Operating
Condition
Reference
Condition
Observed change
in value
for a specific
cavitation level
Log
Figure B.1 Cavitation scale effects
ISA-RP75.23-1995
47
Observed change
in magnitude
of a specific
cavitation effect
Reference
Condition
Operating
Condition
Log
Figure B.2 Cavitation influences
48
ISA-RP75.23-1995
Figure B.3 Typical flow curve appearance
ISA-RP75.23-1995
49
Both flowpaths choked
One
flowpath
choked
Both flowpaths "unchoked"
P/G f
Figure B.4 Flow curve appearance: two flowpath butterfly valve
50
ISA-RP75.23-1995
Annex C Cavitation damage intensity and service life*
C.1 Purpose
The safety of personnel, environment, and equipment must be ensured by proper monitoring,
inspection, and maintenance programs in addition to proper valve selection. Within safety
constraints, the desire to eliminate all cavitation damage sometimes meets with other limitations
such as physical space, process variables, infrequent use, and cost. The following method is
proposed in order to quantify cavitation damage intensity or relative service life reduction of
control valves. In the past, this has been a judgment based on "guess work" and personal
experience. Although based on theory and experience, the method and its results have not been
substantiated by as broad a database as, for example, the scaling equations for . The
presentation of the method acknowledges the need to better quantify the severity of cavitating
service and the need to promote research, testing, and exchange of experience data.
C.2 Intensity index

The intensity index "I" (Equation C.1) can be considered as a damage intensity or life reduction
factor. The magnitude represents an approximation of how many times faster erosion will occur
over the threshold damage rate for the given operating conditions and service environment.
NOTE The intensity index can be used only when a valve must operate in cavitation
levels more severe than the incipient damage level, i.e., at levels that are not recommended
for the valve because of potential damage to the valve. It cannot be used where id for the
valve is unknown. This method must be approached with caution with some types of valves,
e.g., rotary butterfly valves, for which id is difficult to determine.
Once a numerical measure of severity of the service is calculated using ss (Equation C.2), valve
styles and trim can be selected. If the recommended application guidelines for any style of valve
are exceeded, damage and vibration can be expected to be approximately proportional to the
intensity index, defined as Equation C.1.
id 1
I = F U F T F DC -------------------
SS 1
(C.1)
------------ 1
SSE
SS = --------------------- + 1
PSE
(C.2)
*The procedure outlined in Annex C was adapted from a work by Barnes and Cain (ref. 13) wherein the
authors theorized a method for evaluating valve damage potential in the cases where damaging cavitation
for short durations cannot be totally eliminated. The method proposes some relationships from experience
and the general literature, but it has not been rigorously validated by testing. It is presented here for
information and to encourage further investigation. The method is not presented to justify the operation of
valves under damaging cavitation conditions, but rather to illustrate the consequences of doing so.
ISA-RP75.23-1995
51
where FU, FT, and FDC are intensity modifiers defined below. Calculate the application from
Equation 2. Adjust by using Equation C.2 to calculate ss. This "normalizes" to the reference
data (i.e., test data or manufacturer's reference values for c, id, etc.) Compare the selected R
to ss. If R is less than ss, then the valve will operate at an intensity level less than the level
corresponding to the selected R. The basis for interpreting values of I is that a value of 1 (one)
indicates "normal" wear, noise, and vibration for a valve operating at conditions for incipient
damage. Values between zero and one mean proportionally less wear, noise, and vibration.
Values greater than one indicate greater than normal wear, damage, vibration and noise. The
intensity index is not defined for values less than or equal to zero.
C.3 Intensity modifiers

The Intensity is a function of influences in addition to . These factors should be determined by
appropriate damage testing, but the following approximations may be used for estimating
purposes. The incipient damage reference sigma (id) can be determined from model tests or
tests on valves in service with means provided to inspect damage coupons. Reference values of
id for different valve styles are provided by valve manufacturers.
C.3.1 Velocity factor, F U
The velocity factor, FU, may be the most important factor, because it is a function of the difference
between the pitting threshold velocity (Uo) and the actual velocity (U). It can be shown that once
damage or pitting has commenced at the incipient damage sigma or velocity, the rate of pitting is
a strong function of the difference U-Uo. The following velocity correction factor for intensity
appears to be in general agreement with published data (Knapp, et al., ref. 1) on pitting rates
versus velocity.
F U = 1.0
F U = 0.18 + 0.82 e
for
N4 ( U U 0 )
U U0 < 0
(C.3)
for
U U0 0
where:
U is the average fluid velocity through valve outlet area or other characteristic area,
Uo is the damage threshold velocity (average) through the valve's outlet or characteristic area,
and
N4 is a constant based on units used for U and Uo (see Table 1).
From Equation C.3, it can be observed that values of FU approach the value 1.0 as fluid velocity
approaches Uo. As U exceeds U o, values of FU increase exponentially, which reflects the
generally observed trends of pitting rate and material removal rate in the cavitation zone with
sigma values less than id. Due to the sensitivity of FU to the velocity difference (U-Uo), the value
of Uo should be verified by testing. Care should be taken in such tests to allow sufficient
incubation time for pitting to be observed in the test materials. The importance of testing each
valve style for its cavitation coefficients (i.e., c and id) and Uo cannot be over-emphasized
where intensity evaluations are concerned. Minor differences in design geometry can result in
significant differences in valve cavitation parameters.
52
ISA-RP75.23-1995
C.3.2 Fluid temperature factor, FT

FT is the fluid temperature factor. Research on water at low pressures shows that, at fluid
temperatures roughly half-way between the freezing point and boiling point (based on upstream
pressure), cavitation damage is approximately three times more severe; that is, the rate of
material removal is three times faster than the removal rate near the freezing or boiling
temperatures (Knapp, et al., ref. 1). Equation C.4 can be used as an approximate factor to take
this into account. This temperature effect may vary at higher pressures and with liquids other
than water (Hammitt, ref. 16). Effects of temperature on damage rate are not well known as
pressure nears the critical pressure; FT should be ignored (i.e., set to 1) for pressures
approaching or exceeding critical pressure. This factor applies only to liquids of a single
chemical species.
T T ave
F T = 3 2 -----------------------
T B T ave
(C.4)
where:
TB = boiling temperature at upstream pressure
TF = freezing temperature at upstream pressure
T = fluid service temperature
Tave = (TB + TF)/2
T - Tave = absolute value of (T-Tave)
C.3.3 Duty cycle factor, FDC
The duty cycle of the control valve within the cavitating condition (i.e., continuous, intermittent,
rare) can also be taken into account with the duty cycle factor FDC. If the valve will experience
the cavitating condition only during a rare upset, short operation in this condition may not
jeopardize the overall performance of the valve, since the damage effects are time dependent. If
the valve cavitates only during certain service conditions, such as start-up, a more intense
cavitating condition might be withstood than in a continuously throttling valve. If the valve is
continuously throttling, or is in critical service, or both, a more conservative factor must be taken.
Table C.1 represents some possible values of FDC that might be used for estimating purposes if
service history data are lacking. Values for FDC are highly application dependent, and estimated
values may introduce large uncertainties in the resulting calculation of intensity.
Table C.1 Range of estimated values of duty cycle factor, FDC

Frequency of occurrence
Rare upset
Start-up
Throttling
Continuous or critical duty
ISA-RP75.23-1995
FDC
0.1 - 0.3
0.5 - 0.8
1.0 - 1.5
2.0 - 3.0
53
C.4 Example
Calculations of the intensity index were originally proposed (Barnes and Cain, Ref. 13) for a
cavitation index based on P2 (see Equation B.3). The conversion of equations from the P2-based
index to a P1-based index involves adding one (1.0). For very small values of , this changes the
apparent precision of the calculations and magnifies the effects of cumulative round-off errors.
Therefore, determination of significant figures and round-off should be postponed until the final
calculation of intensity index (I).
C.4.1 Boiler feedwater start-up application (US Customary units)
Service data (same valve and
piping as example 7.6.3)
Fluid: Water
P1 = 1600 psia
P2 = 150 psia
P = 1450 psi
Cv = 10.5
Same valve as in example 7.6.3

Body outlet velocity = 4.9 ft/s.
= (1600 - 0.70)/1450 = 1.103
Compare manufacturer's
recommended operating for
possible alternative trim styles
Trim Style
Standard
Trim A
Trim B
Evaluate Trim A, a multi-hole

trim. Use exponent a for an
orifice, because multi-hole
geometry determines the
pressure scaling behavior.
Manufacturer's cavitation reference data

(P1-Pv)R = 100 psi; a = 0.20
Calculate PSE, (Equation 3)

Assume SSE = 1 for small,
multi-hole type trim.
PSE = (1599/100)0.20 = 1.741
v = [(1.20)(1.00)-1](1.741)+1 = 1.348
Calculate effect of reducers.

(Equations 8, 7)
Fp = {1+(0.278)(10.5)2/[890(5.75)4]} = 1.00
p = (1)2{1.348+(0.093+0.68)(10.5)2/[890(5.75)4]}
p = 1.348
Evaluation
is not P for Trim A. Evaluate damage potential

per Annex C or select another trim type.
Use intensity index evaluation.

Calculate ss (Equation C.2)
ss = [(1.103)/(1)-1]/1.741+1 = 1.059
Determine velocity factor, Fu

(Equation C.3)
Since U-U o = 4.9 - 33 < 0, then Fu = 1.
54
mr
2.5
1.2
1.02
T = 90F
Pv = 0.70 psia
Gf = 0.995
q = 400 gpm
Uo
30 ft/s
33 ft/s
--
id
2.3
1.2
--
ISA-RP75.23-1995
Calculate temperature factor, FT

(Equation C.4)
FT = 3-2[90-318/(605-318)]
FT = 1.411
Estimate duty cycle factor

FDC (Table C.1)
For start-up, assume FDC = 0.5.
Determine intensity index

(Equation C.1)
Let id = mr = 1.2
I = (1)(1.411)(0.5)[(1.2-1)/(1.059-1)]
I = 2.4
Evaluation
Trim A will wear (erode) approximately 2 times faster

than at the incipient damage level. Other trim options
(Trim B) should be checked, e.g., multi-stage trim.
CAUTION RAPID TRIM EROSION MAY
BE ACCOMPANIED BY BODY WALL
EROSION.
NOTE The results of intensity calculations are
sensitive to the accuracy and precision of the terms.
Comparisons of intensity indexes should consider the
relative precision of the terms and their accuracy.
C.4.2 Boiler feedwater start-up application (SI units)

Service data (same valve and
piping as example 7.6.6)
Fluid: Water
T = 32.2 C
Pv = 4.83 kPa
P1 = 11 034 kPa
P2 = 1034 kPa
P = 10 000 kPa
Gf = 0.995
q = 90.85 m3/h
Cv = 10.5
Same valve as in example 7.6.6

Body outlet velocity = 1.49 m/s.
= (11034 - 4.83)/10000 = 1.103
Compare manufacturer's
recommended operating for
possible alternative trim styles
Trim Style
Standard
Trim A
Trim B
Evaluate Trim A, a multi-hole trim.

Use exponent a for an orifice,
because multi-hole geometry
determines the pressure scaling
behavior.
Manufacturer's cavitation reference data

(P1-Pv)R = 690 kPa; a = 0.20
ISA-RP75.23-1995
mr
2.5
1.2
1.02
Uo
9.14 m/s
10.06 m/s
--
id
2.3
1.2
--
55
Calculate PSE, (Equation 3)

Assume SSE = 1 for small,
multi-hole type trim.
PSE = (11029/690)0.20 = 1.741
Calculate effect of reducers.
(Equations 8, 7)
v = [(1.20)(1.00)-1](1.741)+1 = 1.348
Fp = {1+(0.278)(10.5)2/[0.00214(146)4]} = 1.00
p = (1)2{1.348+(0.093+0.68)(10.5)2/[0.00214(146)4]}
p = 1.348
Evaluation
is not P for Trim A. Evaluate damage potential per

Annex C or select another trim type.
Use intensity index evaluation.

Calculate ss (Equation C.2)
ss = [(1.103)/(1)-1]/1.741+1 = 1.059
Determine velocity factor, Fu

(Equation C.3)
Since U-U o = 1.49 - 10.06 < 0, then Fu = 1.
Calculate temperature factor, FT

(Equation C.4)
FT = 3-2[32.2-158.9/(318.3-158.9)]
FT = 1.41
Estimate duty cycle factor

FDC (Table C.1)
For start-up, assume FDC = 0.5.
Determine intensity index

(Equation C.1)
Let id = mr = 1.2
I = (1)(1.41)(0.5)[(1.2-1)/(1.059-1)]
I = 2.4
Evaluation
Trim A will wear (erode) approximately 2 times faster

than at the incipient damage level. Other trim options
(Trim B) should be checked, e.g., multi-stage trim.
CAUTION RAPID TRIM EROSION MAY
BE ACCOMPANIED BY BODY WALL
EROSION.
NOTE The results of intensity calculations are
sensitive to the accuracy and precision of the terms.
Comparisons of intensity indexes should consider the
relative precision of the terms and their accuracy.
56
ISA-RP75.23-1995
Annex D Net pressure drop corrections
D.1 Nomenclature
Cv meas Valve flow coefficient calculated from the measured pressure drop P during testing per
SP75.02.
Cv net
Valve flow coefficient calculated from the net pressure loss through the valve only.
Valve inlet inside diameter, inches (mm).
The Darcy-Weisbach pipe friction factor.
Gf
Specific gravity of the liquid at inlet flowing conditions.
Pmeas Measured valve pressure drop which included pipe friction loss, psi (kPa).
Pnet
Net valve differential/pressure drop, psi (kPa).
meas
Cavitation index calculated with the measured pressure drop.
net
Cavitation index calculated with the net pressure drop.
D.2 Measured and net pressure drop

D.2.1 The specific pressure drop utilized in testing valves versus that specified when applying
them in a cavitating service is very important. Ordinarily, the pressure drop used in valve flow tests
is the pressure difference between upstream and downstream pressure taps within the test section.
However, when sizing and selecting a control valve, the upstream and downstream pressures are
often specified at the upstream face and downstream face of the valve, respectively. Failure to
account for this can lead to unexpected results.
D.2.2 The differences as described above give rise to the concept of net pressure difference
versus total (or measured) pressure difference. The net pressure drop is the net effect on system
pressure of installing a control valve into a piping system. The measured pressure drop is the total
of the net drop and the additional loss associated with the length of pipe between the pressure
taps. The flow coefficient, Cv, and the cavitation index, , calculated from the measured pressure
drop can be significantly different from the values of Cv and calculated from the net pressure
drop (C v net and net).
D.2.3 The difference in calculated values of Cv, valve opening, and , is generally within the
accuracy of the method itself for low recovery valves, and is therefore negligible. Low recovery
valves typically are considered to be valves where Cv/N1d2 < 20, as in the case of many globe valves.
D.2.4 For higher recovery valves (i.e., when Cv/N1d2 > 20), the differences may be significant,
and it becomes necessary to ensure consistency between test methods and application practices.
Differences between measured and net pressure drop can be as large as 50% for low loss or high
pressure recovery valves (Rahmeyer and Driskell, ref. 18). The difference between measured and
net values can be even more important for cavitation calculations, because the cavitation index
can vary considerably with small changes in valve opening. Small differences in calculating the
Cv and pressure drop can produce significant differences in valve opening, actuator requirements,
and valve cavitation coefficients.
ISA-RP75.23-1995
57
D.2.5 If the net measured pressure drop is to be determined directly from testing, the downstream
pressure taps can be located 10 diameters downstream (instead of 6 diameters) to ensure more
complete pressure recovery. The downstream pressure of a control valve with a Cv/N 1d2 greater
than 20 may not be fully recovered at the distance of the 6 diameters as specified by ISA-S75.02.
D.2.6 As implied in the above discussion, the effect of adjacent piping should be considered
when applying high recovery valves in cavitating service. The following equations describe methods to convert the various coefficients between total (measured) and net values. This technology
is still evolving and is subject to certain restrictions. The methods described herein are limited to
straight pipe installations. As of this writing, no inference is made as to the nature of the net effect
of a valve on a system other than when straight pipe is immediately adjacent to the valve both
upstream and downstream of the valve.
D.3 Equations
For control valves with a Cv/N1d2 greater than 20, the following equations can be used to convert
pressure drops, Cv, and from measured values to net values. Equations D.1 and D.2 were
originally published by Rahmeyer and Driskell (ref. 18) to convert values of Cv. Equations D.3
through D.6 were then derived from the original equations.
1
---
2
C v net
C v meas 2
-------------- = ------------------ 0.008986 fG f
2
N d2
N1 d
1
(D.1)
1
---
2
C v meas
C v net 2
------------------ = -------------- + 0.008986 fG f
2
N d2
N1 d
1
(D.2)
C v meas2
P net = P meas 1 0.008986 fGf ------------------
N d2
(D.3)
C v net 2
P meas = P net 1 + 0.008986 fGf --------------
N d2
1
(D.4)
meas
net = ----------------------------------------------------------------------- C v meas 2
1 0.008986 f Gf ------------------
N d2
(D.5)
net
meas = ------------------------------------------------------------------- C v net 2
1 + 0.008986 fGf --------------
N d2
(D.6)
58
ISA-RP75.23-1995
Developing and promulgating technically sound consensus standards,

recommended practices, and technical reports is one of ISA's primary
goals. To achieve this goal the Standards and Practices Department
relies on the technical expertise and efforts of volunteer committee
members, chairmen, and reviewers.
ISA is an American National Standards Institute (ANSI) accredited
organization. ISA administers United States Technical Advisory
Groups (USTAGs) and provides secretariat support for International
Electrotechnical Commission (IEC) and International Organization for
Standardization (ISO) committees that develop process measurement
and control standards. To obtain additional information on the
Society's standards program, please write:
ISA
Attn: Standards Department
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, NC 27709
ISBN: 1-55617-572-8

RP 7523

Uploaded by

Copyright:

Available Formats

RP 7523

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RP 7523

Uploaded by

Copyright:

Available Formats

What is the purpose of this document?

What is the purpose of this document?

What organization developed this document?

What organization developed this document?

ISA-RP75.

Approved June 2, 1995

Considerations for Evaluating

ISA-RP75.23, Considerations for Evaluating Control Valve Cavitation

The following people served as members of ISA Subcommittee SP75.16:

*One vote per company

Union Carbide Corporation

The following people served as members of ISA Committee SP75:

*One vote per company

M. Widmeyer, Vice President

Washington Public Power Supply System

*One vote per company

Cavitation parameter plot ............................................................................................... 11

Terms used are per ISA-S75.05 and additional terms as follows:

Figure 1 Cavitation parameter plot

Nomenclature used is per ANSI/ISA S75.01 and additionally as follows:

Empirical characteristic exponent for calculating PSE

Valve flow coefficient*, C v = q(Gf/P)1/2

Valve flow coefficient of a reference valve

Valve inlet inside diameter, inches (mm)

Valve inlet inside diameter of tested reference valve, inches (mm)

Internal diameter of upstream pipe, inches (mm)

Internal diameter of downstream pipe, inches (mm)

Napierian base, e = 2.71828... (for natural logarithms)

Duty cycle factor for modifying the intensity index

Liquid critical pressure ratio factor*

Liquid pressure recovery factor*

Piping factor for ISA valve sizing*

Temperature factor for modifying the intensity index

Velocity factor for modifying the intensity index

Specific gravity of the liquid at inlet flowing conditions

Bernoulli coefficient for upstream pipe reducer*

Bernoulli coefficient for downstream pipe expansion*

Head loss coefficient for upstream pipe reducer*

Head loss coefficient for downstream pipe increaser*

Numerical constants for units of measure used in equations. See Table 1.

Atmospheric pressure, psia (kPa)

Valve inlet static pressure, psia (kPa)

Valve outlet static pressure, psia (kPa)

PSE Pressure scale effect

Absolute fluid vapor pressure of liquid at inlet temperature, psia (kPa)

Fluid static pressure in valve vena contracta*, psia (kPa)

Volumetric flow rate, gpm (m 3/h)

SSE Size scale effect

Sound Pressure Level referenced to 20 x 10-6 Pascal (2.0 x 10-5 N/m 2)

Fluid Temperature, F (C)

Boiling temperature of a liquid for specified pressure, F (C)

Freezing temperature of a liquid, F (C)

Average velocity at the valve inlet, ft/s (m/s)

*More completely defined in ANSI/ISA S75.01 and ANSI/ISA S75.02

Pitting threshold velocity determined at the valve inlet, ft/s (m/s)

Coefficient of incipient cavitation per IEC-534-8-2; xFz 1/i

Measured valve differential pressure, psi (kPa)

Pch Pressure drop at choking, psi (kPa)

Coefficient for choking cavitation; ch is equal to (P1-Pv)/[FL2(P1-FFPv)] at the point