01011

Issued November 2001

with Amendment A
April 2002
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Compendium of curves and

approximating algebraic
functions

Endorsed by
The Royal Aeronautical Society
 ESDU
TM

Engineering Sciences Data Unit 01011

ESDU DATA ITEMS

Data Items provide validated information in engineering design and analysis for use by, or under the supervision of,
professionally qualified engineers. The data are founded on an evaluation of all the relevant information, both published and
unpublished, and are invariably supported by original work of ESDU staff engineers or consultants. The whole process is
subject to independent review for which crucial support is provided by industrial companies, government research
laboratories, universities and others from around the world through the participation of some of their leading experts on ESDU
Technical Committees. This process ensures that the results of much valuable work (theoretical, experimental and
operational), which may not be widely available or in a readily usable form, can be communicated concisely and accurately
to the engineering community.
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

We are constantly striving to develop new work and review data already issued. Any comments arising out of your use of our
data, or any suggestions for new topics or information that might lead to improvements, will help us to provide a better service.

THE PREPARATION OF THIS DATA ITEM

The work on this particular Data Item was monitored and guided by the Performance Committee. This Committee first met in
1942 and now has the following membership:

Chairman
Mr K.J. Balkwill — Independent

Members
Mr P.V. Aidala* — Northrop Grumman ESID, Bethpage, New York, USA
Dr J.D. Avery — Qinetiq, Boscombe Down
Mr T. Bartup — BAE SYSTEMS, Woodford
Mr M. Broad — Independent
Mr N.J. Herniman — Airbus Industries, Toulouse, France
Mr R.G. Humpston — Rolls-Royce plc, Aero Division, Derby
Mr R.J. Orlowski* — Boeing Commercial Airplane Company, Seattle, Washington,
USA
Dr P. Render — University of Loughborough
Mr P.P. Renze* — Raytheon Aircraft Company, Wichita, Kansas, USA
Mr D.N. Sinton — Independent
Mr G.J.R. Skillen — Civil Aviation Authority, Safety Regulation Group, Gatwick
Mr G.E. Smith — Independent
Mr A. Stanbrook — Independent
Mr R. Storey — BAE SYSTEMS, Brough
Prof. E. Torenbeek* — Delf University of Technology, The Netherlands
Mr C.J. Tsach* — Israel Aircraft Industries, Jerusalem, Israel
Mr C.J. Turner* — Boeing Commercial Airplane Company, Seattle, Washington,
USA
Mr M. Wilson — Britten-Norman Ltd, Isle of Wight.
*
Corresponding Members.

The construction and subsequent development of the Data Item was undertaken by

Mr G.E. Smith — under contract to ESDU.

The person with overall responsibility for the work in this subject area is Mr D.J. Mitchell, Director of Engineering.
 ESDU
TM

Engineering Sciences Data Unit 01011

COMPENDIUM OF CURVES AND APPROXIMATING ALGEBRAIC FUNCTIONS

CONTENTS
Page

1. GLOSSARY 1
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

2. INTRODUCTION 2
2.1 Applications 2
2.2 Implementation 2

3. GENERAL MULTI-TERM FUNCTIONS SUITABLE FOR CURVE APPROXIMATION 3

3.1 Polynomials in Positive Integer Powers of the Independent Variable
(Table 7.1, Sketches 7.1 to 7.4) 3
3.2 Legendre Polynomials (Table 7.1, Sketches 7.1 to 7.4) 3
3.3 Rational Function (Table 7.1, Sketches 7.1 to 7.12) 4
3.4 Peak Function (Table 7.1, Sketches 7.13 to 7.15) 4
3.5 Trigonometric Functions (Table 7.1, Sketches 7.16 to 7.23) 5
3.5.1 General form, including odd and even harmonics in both
cosines and sines (Table 7.1, Sketch 7.16) 5
3.5.2 Cosine terms only (Table 7.1, Sketch 7.17) 5
3.5.3 Sine terms only (Table 7.1, Sketches 7.18 and 7.23) 5
3.5.4 Odd cosine terms only (Table 7.1, Sketch 7.19) 5
3.5.5 Odd sine terms only (Table 7.1, Sketches 7.20 to 7.22) 6

4. LOGIC FUNCTIONS (TABLE 7.2) 6

4.1 Discontinuous Functions (Table 7.2, Sketch 7.28) 7
4.2 Ramp Functions (Table 7.2, Sketches 7.29 to 7.32) 7
4.2.1 Ramp shape defined by exponent n in algebraic expression 7
4.2.2 Ramp shape defined by cosine function 7

5. SINGLE-TERM FUNCTIONS (TABLE 7.3) 8

6. REFERENCES 8

7. TABLES 9 to 25

This page Amendment A

i
 ESDU
TM

Engineering Sciences Data Unit 01011

COMPENDIUM OF CURVES AND APPROXIMATING ALGEBRAIC FUNCTIONS

1. GLOSSARY

No “Notation and Units” Section is given in this Data Item – which makes use only of standard mathematical
symbols for variables and constants. However, some of the mathematical terms used may need explanation,
as follows.

Asymptote (to a curve): A straight or curved line that is closely approached by a plane curve so that the
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

perpendicular distance between them decreases steadily to zero as the independent variable governing the
curve approaches a critical value or tends to infinity.

Dependent Variable: The variable in a mathematical function whose value is entirely dependent on the
chosen values of the independent variable(s) and on the function itself. (It is customary, whenever practical,
for the dependent variable to appear on the left-hand-side of an equation and as the vertical axis (ordinate)
on a simple graph.)

Discontinuity: A point on a curve where the function is not continuous, i.e. its value takes a sudden jump.

Extrapolation: The process of estimating the value of a function beyond the range of values already known,
by the extension of its curve.

Even Function: A function, f ( x ) , is even if f ( x ) = f ( – x ) .

Independent Variable: The variable(s) in a mathematical function whose values are considered to be
selected independently, i.e. without reference to the function or any other variable(s). (It is customary,
whenever practicable, for independent variables to appear on the right-hand-side of an equation and as the
horizontal axis (abscissa) of a simple graph.)

Interpolation: The process of estimating the value of a function at a point within the range of values already
known.

Odd function: A function, f ( x ) , is odd if, f ( x ) = –f ( – x ) .

Overfitting: This occurs when a multiple-term mathematical function fitted to a set of data has too many
terms in comparison with the number of data points available. The resulting curve tends to follow the
irregularities represented by the individual points rather than the overall trends of the data. (See also
Reference 3.)

Singularity: A point on an algebraic function where it is not differentiable although it is differentiable in

the immediate neighbourhood of that point, e.g. an infinity in a curve.

Underfitting: This occurs when a multiple-term mathematical function fitted to a set of data has too few
terms to enable it to follow satisfactorily the overall trends of the data. (See also Reference 3.)

Issued November 2001 - 25 pages

With Amendment A – April 2002
1
 ESDU
TM

Engineering Sciences Data Unit 01011

2. INTRODUCTION

This Data Item presents families of curve shapes and algebraic functions as an aid to the effective
representation of experimental, numerical and graphical data by the selection of suitable such functions.
Three categories of functions are presented:

Section 3 and Table 7.1: General multi-term functions,

Section 4 and Table 7.2: Logic functions,

ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Section 5 and Table 7.2: Single-term functions.

In many cases more than one type of function can represent a curve or data-set satisfactorily. In such cases
it can be beneficial to compare the best-fitting curves produced by different functions before deciding which
gives the best overall representation.

2.1 Applications

No distinctions are made here between the various applications of such algebraic functions, i.e.

(i) representation of experimental results,

(ii) representation of output from calculation procedures,

(iii) representation of pre-existing curves from graphical representations of lines, “carpets” or “lattices”.

Reference 3 gives a range of examples of Applications (i) and (ii) using the least-squares method of fitting
and exploring the strengths and weaknesses of the functions presented here in Section 3 and Table 7.1.

Reference 1 (Data Item No. 81009) provides an example of Application (iii) to simple lines and carpets,
see Appendix B of Reference 1.

2.2 Implementation

When fitting an algebraic function to a set of points plotted on a graph it is helpful to start by drawing a
smooth “freehand” curve to follow the general trends of the points without being too influenced by any
irregularities or scatter*. In essence this should be a smoothing process rather than one of interpolation,
which produces curves constrained to pass through all the points individually see (Reference 3). Having
drawn a representative curve through the set of data points, a suitable algebraic function is chosen that is
capable of reproducing the curve’s salient features, see Sections 3, 4 and 5 and Tables 7.1, 7.2 and 7.3. The
use of the “least-squares” procedure in fitting the chosen function to the data is described in Reference 3.

The use of statistical techniques to represent the distribution of “scatter” about any function is introduced
in Reference 2 (Data Item No. 91017).

* This discussion is presented in terms of Applications (i) and (ii), rather than Application (iii) where the pre-existing curve(s) could
correspond to a “zero-scatter” case.

2
 ESDU
TM

Engineering Sciences Data Unit 01011

3. GENERAL MULTI-TERM FUNCTIONS SUITABLE FOR CURVE APPROXIMATION

This Section presents a few types of multi-term function that can be applied to a very wide range of curves.
These functions are general in nature and can be modified to suit specific characteristics of a curve, for
example by the omission of one or more terms.

Although the functions listed can be applied to an extensive range of engineering applications, the list is
not exhaustive and other functions might be contrived to suit a particular set of data. However, in general,
the use of over-complicated functions should be avoided as they can introduce extraneous features of their
own, such as waves or discontinuities, that are not inherent in the actual data.
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

3.1 Polynomials in Positive Integer Powers of the Independent Variable (Table 7.1, Sketches 7.1 to 7.4)

The general form, including odd and even powers of x, is

f ( x ) = C 1 + C 2 x + C 3 x 2 + ……… + C p x ( p – 1 ) (3.1)

where C 1 , C 2 , …… are coefficients of the independent variable x .

The symmetrical form, with only even powers of x, is

f ( x ) = C 1 + C 2 x 2 + C 3 x 4 + C 4 x 6 …… + C p x ( 2p – 2 ) . (3.2)

The antisymmetrical form, with only odd powers of x, is

f ( x ) = C 1 x + C 2 x 3 + C 3 x 5 + C 4 x 7 + …… + C p x ( 2p – 1 ) . (3.3)

(Other sequences, with larger differences in the exponents of x can be derived.)

If the curve to be fitted is known to be symmetrical, such that f ( – x ) = f ( x ) , then Equation (3.2) should be
used. If the curve is antisymmetrical, such that f ( –x ) = –f ( x ) then Equation (3.3) is appropriate. In all
other cases the general form of Equation (3.1) should be used.

These polynomials are easy to evaluate and are widely used for relatively simple curves. However, they
cannot adequately reproduce asymptotic behaviour, such as a levelling out to a constant value, or
singularities such as a rapid rise to infinity at some critical point in the data range. They also have a tendency
to diverge abruptly from the general trends of the data beyond the ends of the data range, and therefore
cannot safely be extrapolated. Another disadvantage is that if too many terms are employed they can be
prone to problems of overfitting and to numerical instabilities in the computation of the coefficients. These
features are discussed more fully in Reference 3.

3.2 Legendre Polynomials (Table 7.1, Sketches 7.1 to 7.4)

These take the form

f ( x ) = C1 P1 ( x ) + C 2 P 2 ( x ) + C3 P3 ( x ) + …… + C p P p ( x ) (3.4)
( 3x 2 – 1 ) ( 5x 2 – 3 )x
where P1 = 1 ; P 2 = x ; P 3 = --------------------- ; P 4 = ------------------------ etc.
2 2

3
 ESDU
TM

Engineering Sciences Data Unit 01011

They may be employed in fitting to simple curves in place of polynomials in powers of x . Legendre
polynomials are less likely than polynomials in powers of x to give rise to numerical instabilities such as
overflows and loss of significant digits. However, they also are unable to reproduce asymptotic behaviour
and cannot safely be extrapolated due to a tendency to diverge abruptly beyond the range of data. They are
also prone to the phenomenon of overfitting if too many terms are used. The features of Legendre
polynomials are discussed further in Reference 3.

3.3 Rational Function (Table 7.1, Sketches 7.1 to 7.12)

This is formed as the ratio of two polynomial functions. In its simplest form both polynomials could be in
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

powers of x , the numerator having, say, p terms and the denominator having q terms, thus

C 1 + C 2 x + C 3 x 2 + …… + C p x ( p – 1 )
f ( x ) = -----------------------------------------------------------------------------------------------------------------------------------------------
-. (3.5)
1 + C ( p + 1 ) x + C ( p + 2 ) x 2 + …… + C ( p + q – 1 ) x ( q – 1 )

The Rational function is extremely versatile and can produce a wide range of curves including those
obtained from power series or Legendre polynomials. Unlike the polynomial functions, the rational function
generates curves that show no tendency to diverge abruptly beyond the ends of the data range and which
therefore can be extrapolated to a limited extent. It is also less prone to overfitting than polynomials having
an equal number of terms.

The most useful feature of the Rational function is its ability to reproduce asymptotic behaviour in a curve,
which the polynomial functions are unable to do. It may also be used in cases where the data suggest
singularities, such as when the curve tends to infinity at some point within the data range. This behaviour
can be reproduced by zero values in the denominator of a rational function but cannot be obtained from a
polynomial in powers of x . However, it is possible for a rational function to produce unwanted singularities
within the data range, in which case the function should be rejected. This emphasises the need to check the
plotted curve before a function is accepted as a fit to the data. The characteristics of the Rational function
are discussed further in Reference 3.

3.4 Peak Function (Table 7.1, Sketches 7.13 to 7.15)

The peak function can take the form

C
F p ( x ) = -------------------------------------- . (3.6)
x–K 2
1 + B  --------------
 ∆x 

This function produces a peak of height C (if the coefficient C is positive) or a trough of depth C (if the
coefficient is negative), at a specified point along the x-axis. The parameter K defines the value of x at
which the peak occurs. The parameter B controls the sharpness of the peak in terms of the interval ∆x .
The parameters K , B and ∆x must be prescribed so as to suit the position and shape of the peak.

Peak functions may be used in isolation but are best employed in conjunction with other functions which
can follow the background trends of the data, for example

C(p + 1 ) C(p + 2 )
f ( x ) = C1 + C2 x + C3 x 2 + ……C p x ( p – 1 ) + ------------------------------------- + ------------------------------------- + …… (3.7)
x – K1 2 x – K2 2
1 + B1  -------------- 1 + B 2  --------------
 ∆x   ∆x 

4
 ESDU
TM

Engineering Sciences Data Unit 01011

Properties of the peak function are discussed further in Reference 3.

3.5 Trigonometric Functions (Table 7.1, Sketches 7.16 to 7.23)

Trigonometric functions are useful in representing curves which are periodic in nature such that
f ( x ) = f ( x + 2L ) , where 2L is the period, or wavelength, over which the curve repeats itself. Such a curve
can usually be approximated by a series of cosine and sine functions, several types of which are introduced
in Sections 3.5.1 to 3.5.5 and illustrated in Sketches 7.16 to 7.24. For convenience they may be expressed
in terms of an angle θ = πx/L , so that θ varies between 0 and 2π as x varies from 0 to 2L.
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

3.5.1 General form, including odd and even harmonics in both cosines and sines (Table 7.1, Sketch 7.16)

f ( x ) = C1 + C2 cos θ + C 3 cos 2θ + C 4 cos 3θ + … + C p cos ( p – 1 )θ

+ C ( p + 1 ) sin θ + C ( p + 2 ) sin 2θ + …… + C ( p + q ) sin qθ. (3.8)

3.5.2 Cosine terms only (Table 7.1, Sketch 7.17)

If a curve is seen to possess certain degrees of symmetry, some of the terms in the general series (i.e.
Equation (3.8)) may be omitted. For example, if the curve is seen to be even, i.e. symmetrical about x = 0
such that f ( – x ) = f ( x ) , then only the cosine terms should be included, plus the initial constant term C 1 ,
to give

f ( x ) = C 1 + C 2 cos θ + C 3 cos 2θ + C 4 cos 3θ + … + C p cos ( p – 1 )θ . (3.9)

3.5.3 Sine terms only (Table 7.1, Sketches 7.18 and 7.23)

If a curve is seen to be odd, i.e. is antisymmetrical about x = 0 , such that f ( – x ) = –f ( x ) , then only the
sine terms should be included, as follows,

f ( x ) = C 1 sin θ + C 2 sin 2θ + C 3 sin 3θ + … + C q sin qθ . (3.10)

3.5.4 Odd cosine terms only (Table 7.1, Sketch 7.19)

Sometimes further degrees of symmetry may be observed in a curve, in which case certain other terms may
be omitted. For example, a curve which is even and also antisymmetrical about x = L/2 may be
approximated by a cosine series containing only the odd harmonics, as follows,

f ( x ) = C 1 cos θ + C 2 cos 3θ + C 3 cos 5θ + … + C p cos ( 2p – 1 )θ . (3.11)

5
 ESDU
TM

Engineering Sciences Data Unit 01011

3.5.5 Odd sine terms only (Table 7.1, Sketches 7.20 to 7.22)

In a similar manner, if a curve is odd and also symmetrical about x = L/2 then only the odd sine terms are
necessary, as follows,

f ( x ) = C 1 sin θ + C 2 sin 3θ + C 3 sin 5θ + … + C q sin ( 2q – 1 )θ . (3.12)

4. LOGIC FUNCTIONS (TABLE 7.2)

ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Instead of a mathematical variable, the argument of a logic function consists of a conditional expression,
the outcome of which determines the value taken by the function. A logic function takes the value unity if
the condition is true, and zero if it is false.

Four general types of logic function are represented in Equations (4.1) to (4.4) (by convention enclosed in
square brackets). These are illustrated in Table 7.2, Sketches 7.24 to 7.27 respectively to show how their
values change according to the truth of the conditional expression.

y = [ x > x1 ]
= 0 for x ≤ x 1 (4.1)

= 1 for x > x 1 .

y = [ x < x1 ]
= 1 for x < x 1 (4.2)

= 0 for x ≥ x 1 .

y = [ x > x 1 AND x < x 2 ]

= 1 for x 1 < x < x 2 (4.3)

= 0 for all other x .

y = [ x < x 1 OR x > x 2 ]
= 0 for x 1 < x < x 2 (4.4)

= 1 for all other x .

Any logic function can be employed as a multiplier of other mathematical functions, when it serves the
purpose of a switch to turn them on or off at specific values of the independent variable x . Examples of
this are given in Sections 4.1 and 4.2.

6
 ESDU
TM

Engineering Sciences Data Unit 01011

4.1 Discontinuous Functions (Table 7.2, Sketch 7.28)

A curve may show discontinuities, such as an abrupt change of slope, at one or more values of x , and it is
then not possible to fit a single continuous function over its whole range. In such cases the curve may be
divided into a number of segments between the discontinuities at x1 , x2 etc. and separate continuous
functions f 1 ( x ) , f 2 ( x ) etc. fitted to each segment. Each function is then multiplied by a suitable logic
function to switch it on and off over its segment, for example,

f ( x ) = [ x ≤ x 1 ]f 1 ( x ) + [ x > x 1 AND x < x 2 ]f 2 ( x ) + [ x ≥ x 2 ]f 3 ( x ) . (4.5)

ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

The result of such an application is shown in Table 7.2, Sketch 7.28. The fitting of functions to discontinuous
curves is discussed further in Reference 3.

4.2 Ramp Functions (Table 7.2, Sketches 7.29 to 7.32)

Ramp functions may be used to replicate a ramp, or jump, from one value of y to another over a prescribed
range of x , e.g. a jump from point ( x1, y1 ) to ( x2, y2 ) . Various types of function can be chosen to suit the
shape of ramp required; examples are given in Sections 4.2.1 and 4.2.2.

4.2.1 Ramp shape defined by exponent n in algebraic expression

A general expression takes the form

x – x1 n
f ( x ) = y1 + [ x > x1 AND x < x2 ]h  ------------- + [ x ≥ x 2 ]h (4.6)
 l 
where l = ( x 2 – x 1 ) and h = ( y 2 – y 1 ) . (4.7)

The exponent n in the second term of Equation (4.6) controls the shape of the ramp, for example,

if n = 1 the ramp is a straight line of constant slope (see Sketch 7.29),

if n > 1 the ramp is curved, with zero slope at x = x1 (see Sketch 7.30),

if n < 1 the ramp is curved, with an infinite slope at x = x1 (see Sketch 7.31).

Note the use of the logic functions, contained in square brackets, which act as controlling switches on the
terms they multiply.

4.2.2 Ramp shape defined by cosine function

In this case,

h π 
f ( x ) = y 1 + [ x > x 1 AND x < x 2 ] ---  1 – cos --- ( x – x 1 )  + [ x ≥ x 2 ] h . (4.8)
2 l 

Here the ramp between points ( x1, y 1 ) and ( x 2, y2 ) is a smooth, symmetrical cosine curve with zero slopes
at x = x1 and x = x2 . See Table 7.2, Sketch 7.32.

7
 ESDU
TM

Engineering Sciences Data Unit 01011

5. SINGLE-TERM FUNCTIONS (TABLE 7.3)

Many single-term functions of an independent variable, x , can serve to represent data or curves. Their
primary advantage is ease of calculation of the function (usually pre-programmed). However, since they
contain only one term they can be rather inflexible and may not give close fits to the data or curves of
interest. Greater flexibility may be obtained by combining several of these functions, each with its own
multiplying constant, as illustrated in Section 3 with simple polynomials in integral powers of x . The
following is a list of single-term functions whose graphs are illustrated in Table 7.3.

Sketch 7.33: y= xr (5.1)

ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Sketch 7.34: y= sin x (5.2)

Sketch 7.35: y= cos x (5.3)
Sketch 7.36: y= tan x (5.4)
sin x
Sketch 7.37: y = ---------- (5.5)
x
cos x
Sketch 7.38: y = ----------- (5.6)
x
Sketch 7.39: y = sinh x (5.7)
Sketch 7.39: y = cosh x (5.8)
Sketch 7.40: y = tanh x (5.9)
tanh x
Sketch 7.41: y = ------------- (5.10)
x
Sketch 7.42: y = log ex (5.11)
log ex
Sketch 7.43: y = ------------ (5.12)
x
Sketch 7.44: y = x log ex (5.13)
Sketch 7.45: y = ax (a ≥ 1) (5.14)
Sketch 7.46: y= ax (0 < a < 1) (5.15)
(a ≥ 1)
1/x
Sketch 7.47: y=a (5.16)
Sketch 7.48: y=a 1/x
(0 < a < 1) (5.17)
Sketch 7.49: y= x bx (b ≥ 0) (5.18)
Sketch 7.50: y = x bx (b ≤ 0) (5.19)

6. REFERENCES

1. ESDU Estimation of windmilling drag and airflow of turbo-jet and turbo-fan

engines. ESDU Data Item No. 81009, June 1981 (with Amendment A, April
1984).
2. ESDU Statistical methods applicable to analysis of aircraft performance data.
ESDU Data Item No. 91017, October 1991 (with Amendments A to D,
September 1997).
3. ESDU ESDU Data Item to be issued.

8
 ESDU
TM

Engineering Sciences Data Unit 01011

7. TABLES

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS

CAPABLE OF FITTING THEM
Suitable types of Equation Section
Typical Curves
function number number

y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Polynomial (3.1) 3.1

Legendre polynomial (3.4) 3.2
Rational function (3.5) 3.3

Sketch 7.1

Polynomial (3.1) 3.1

Legendre polynomial (3.4) 3.2
Rational function (3.5) 3.3

Sketch 7.2

Polynomial (3.2) 3.1

(symmetric)
Legendre polynomial (3.4) 3.2
Rational function (3.5) 3.3
x

Sketch 7.3

9
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Polynomial (3.3) 3.1

(antisymmetric)
Legendre polynomial (3.4) 3.2
x
Rational function (3.5) 3.3

Sketch 7.4

Asymptote

Rational function (3.5) 3.3

y with p = q

Sketch 7.5

y Asymptote

Rational function (3.5) 3.3

with p = q

Sketch 7.6

10
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Rational function (3.5) 3.3

Asymptote with p = q

Sketch 7.7

Rational function (3.5) 3.3

with p > q
As
ym
pt
ot
e

x
Sketch 7.8

Rational function (3.5) 3.3

with p > q
e
ot
pt
ym
As

Sketch 7.9

11
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Rational function (3.5) 3.3

with p < q
As
ym
p tot
e

0 x

Sketch 7.10
As
ym

y
pt
ot
e

Rational function (3.5) 3.3

with p < q

Asymptote

Sketch 7.11

Rational function (3.5) 3.3

with p = 1 , q = 2
Asymptote

Sketch 7.12

12
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

The peak function (3.7) 3.4

(2 peak functions)

k1 k2 x
Sketch 7.13

The peak function (3.7) 3.4

(2 peaks, 1 trough
function)

k1 k2 k3
x

Sketch 7.14

The peak function (3.7) 3.4

(1 peak function)

k1 x

Sketch 7.15

13
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

y
1
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

General form of (3.8) 3.5, 3.5.1

trigonometric function
0

-1
L 2L 3L
x
Sketch 7.16

1
y

Trigonometric (3.9) 3.5.2

function; cosine terms
0 only

-1
0 L 2L 3L
x
Sketch 7.17

1
y

Trigonometric (3.10) 3.5.3

function; sine terms
0 only
(cf. Sketch 7.23)

-1
0 L 2L x 3L

Sketch 7.18

14
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

1
y
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Trigonometric (3.11) 3.5.4

function; odd cosine
0 terms only

-1
0 L 2L 3L
x

Sketch 7.19

1
y 1
1
0
Trigonometric (3.12) 3.5.5
0
function; odd sine
0 terms only
-0 (cf. Sketches 7.21,
-0 7.22)
-1
-1
-1
0 L1 22L x 33L

Sketch 7.20

1
1
y
1
0
0 Trigonometric (3.12) 3.5.5
0
function; odd sine
terms only
-0
(cf. Sketches 7.20,
-0
7.22)
-1
-1
-1
0 L1 2 2L 33L
x
Sketch 7.21

15
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.1 TYPICAL CURVES AND GENERAL MULTI-TERM FUNCTIONS
CAPABLE OF FITTING THEM (Continued)
Suitable types of Equation Section
Typical Curves
function number number

1
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y Trigonometric (3.12) 3.5.5

function; odd sine
0 terms only
(cf. Sketches 7.20,
7.21)
-1

0 L 2L
x
Sketch 7.22

1
1
y 1
0
0
0 Trigonometric (3.10) 3.5.3
-0 function; sine terms
-0 only
-1 (cf. Sketch 7.18)
-1
-1
0 0 0 1 1 1L 1 1 2 2 22L
x
Sketch 7.23

16
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.2 LOGIC FUNCTIONS AND EXAMPLES
Equation Section
Typical Curves Suitable types of function
number number

Logic Function:
y
1
y = [ x > x1 ] (4.1) 4
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

= 0 for x ≤ x 1
x1 x
= 1 for x > x 1

Sketch 7.24

Logic Function:
y

1 y = [ x < x1 ] (4.2) 4
= 1 for x < x 1
x1 x
= 0 for x ≥ x 1

Sketch 7.25

Logic Function:
y
1
y = [ x > x 1 AND x < x 2 ] (4.3) 4
= 1 for x 1 < x < x 2
x1 x2 x = 0 for all other x

Sketch 7.26

Logic Function:
y

1
y = [ x < x 1 OR x > x 2 ] (4.4) 4
= 0 for x 1 < x < x 2
x1 x2 x
= 1 for all other x

Sketch 7.27

17
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.2 LOGIC FUNCTIONS AND EXAMPLES (Continued)
Equation Section
Typical Curves Suitable types of function
number number

y
Discontinuous functions (4.5) 4.1
f3(x)
f2(x)
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

f1(x)

x1 x2 x

Sketch 7.28

Ramp with shape defined (4.6) 4.2.1

by n = 1 in algebraic
y2
expression
y

y1

x1 x2
x

Sketch 7.29

Ramp with shape defined by (4.6) 4.2.1

y2 n = 2 in algebraic
expression
y

y1

x1 x2
x
Sketch 7.30

18
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.2 LOGIC FUNCTIONS AND EXAMPLES (Continued)
Equation Section
Typical Curves Suitable types of function
number number

Ramp with shape defined (4.6) 4.2.1

y1
by n = ½ in algebraic
expression
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y2

x1 x2
x

Sketch 7.31

Ramp with shape defined (4.8) 4.2.2

y2 by cosine function

y1

x1 x2
x
Sketch 7.32

19
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5
Typical Curves Function

1
0.1
0.2
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

0.5 y = xr
y r=1

2
3
5
10

0 x 1
Sketch 7.33

1
1
1
0
0 y = sin x
y
0
-0
-0
-1
-1
-1
-2π -π 0 π 2π
x
Sketch 7.34

y = cos x
y
0

-1
-2π -π 0 π 2π
x
Sketch 7.35

20
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5 (Continued)
Typical Curves Function

20

y = tan x
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y 0

-20
-π -π / 2 0 π/2 π
x

Sketch 7.36
1

sin x-
y = ---------
y x
0

-1
0 10 20
x

Sketch 7.37
1

cos x-
y = ----------
y x
0

-1
0 10 20
x

Sketch 7.38

21
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5 (Continued)
Typical Curves Function

cosh

y = sinh x
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y 0 y = cosh x

sinh

-5
-2.5 0 2.5
x
Sketch 7.39

y = tanh x

y 0

-1
-2.5 0 2.5
x

Sketch 7.40
1

tanh x
y = -------------
y 0.5
x

0
-10 -8 -6 -4 -2 0 2 4 6 8 10
x

Sketch 7.41

This page Amendment A

22
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5 (Continued)
Typical Curves Function

1
y = log ex
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y
0

-1

-2
0 1 2 3 4 5 6 7 8 9 10
x
Sketch 7.42
0.5

log ex
y 0 y = ------------
x

-0.5
0 5 10
x

Sketch 7.43
2

1
y = x log ex
y

-1
0 1 2
x

Sketch 7.44

This page Amendment A

23
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5 (Continued)
Typical Curves Function

20
a=5
4

y = ax
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

y
10 3 with a ≥ 1
e

0 1 2
x
Sketch 7.45
1

a = 0.9
y y = ax
with 0 < a < 1
0.8

0.7
0.6
0.5
0
0 5 10
x

Sketch 7.46
a=2 3 4 5
10

1
y = a /x
y
5 with a ≥ 1

a=1

0 1 2
x

Sketch 7.47

24
 ESDU
TM

Engineering Sciences Data Unit 01011

TABLE 7.3 CURVE SHAPES OF PARTICULAR SINGLE-TERM
FUNCTIONS IN SECTION 5 (Continued)
Typical Curves Function

1 0.9
1 0.5
0.2
1 0.1
1 0.05
0.01 y = a 1/x
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

1
y
1 a with 0 < a < 1
0
0
0
0
0
x 10

Sketch 7.48

5 b=3 2

y = x bx
y
with b ≥ 0
0.5

1 0

0 1 2
x
Sketch 7.49

5
b = -4

-3 y = x bx
y with b ≤ 0
-2

-1
1
0

0 1 2
x
Sketch 7.50

25
 ESDU
TM

Engineering Sciences Data Unit 01011

KEEPING UP TO DATE

Whenever Items are revised, subscribers to the service automatically receive the material required to update the appropriate
Volumes. If you are in any doubt as to whether or not your ESDU holding is up to date, please contact us.

Please address all technical engineering enquiries and suggestions to:

ESDU International plc Tel: 020 7490 5151 (from the UK)
+44 20 7490 5151 (from outside the UK)
Fax: 020 7490 2701 (from the UK)
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

+44 20 7490 2701 (from outside the UK)

E-Mail: [email protected]
Website: www.esdu.com

For users in the USA, please address all Customer Service and Support enquiries and suggestions to:

IHS Engineering Products Tel: 1 800 525 7052 (toll free number)
and Fax: 1 303 397 2599
Global Engineering Documents Website: www.ihs.com
www.global.ihs.com
 01011
Compendium of curves and approximating algebraic functions
ESDU 01011

ISBN186246 168 6, ISSN 0141-4054

Available as part of the ESDU Series on Performance. For information on all

ESDU validated engineering data contact ESDU International plc, 27 Corsham
Street, London N1 6UA.

Data Item No. 01011 presents families of curve shapes and algebraic functions
as an aid to the effective representation of experimental, numerical and graphical
data by the selection of suitable such functions. The functions presented are
divided into three categories: General multi-term functions; Logic functions;
ESDU product release: 2005-03. For current status, contact ESDU. Observe Copyright.

Single-term functions.

No distinctions are made here between the various applications of such algebraic
functions, i.e.
representation of experimental results;
representation of output from calculation procedures;
representation of pre-existing curves from graphical
representations of lines, “carpets” and “lattices”.

A separate Data Item will give examples and guidance on the use of different
classes of function for particular purposes.

© ESDU International plc, 2001

All rights are reserved. No part of any Data Item may be reprinted, reproduced, or transmitted
in any form or by any means, optical, electronic or mechanical including photocopying,
recording or by any information storage and retrieval system without permission from ESDU
International plc in writing. Save for such permission all copyright and other intellectual
property rights belong to ESDU International plc.

ESDU
TM

Engineering Sciences Data Unit

An IHS GROUP Company