Spring Materials PDF
Spring Materials PDF
Spring Materials PDF
MECHANICAL SPRINGS
3.1 DEFINITION
A mechanical spring may be defined as an elastic body whose primary
function is to deflect or distort under load (or to absorb energy) and which
recovers its original shape when released after being distorted [4].
33
diameter Do and the inside diameter Di are of interest mainly to define the
minimum hole in which it will fit or the maximum pin over which it can be placed.
They are found by adding or subtracting the wire diameter d to or from the mean
coil diameter D. The minimum recommended diametric clearances between the
Do and a hole or between Di and a pin are 0.1 OD for D <13 mm or 0.05D for D>
13mm [5].
The helical springs have the following advantages:
(a) They are easy to manufacture
(b) They are available in wide range.
(c) They are reliable.
(d) They have constant spring rate.
(a) Their performance can be predicted more accurately.
(f) Their characteristics can be varied by changing dimensions.
34
Most light-duty springs are made of cold-drawn, round or rectangular wire or
thin, cold-rolled, flat-strip stock. Heavy-duty springs, such as vehicle-suspension
parts, are typically made from hot-rolled or forged forms. Spring materials are
typically hardened in order to obtain the required strength. Small cross sections
are work hardened in the cold-drawing process. Large sections are typically heat
treated. Low temperature heat treatment (175 to 510°C) is used after forming to
relieve residual stresses and stabilize dimensions, even in small-section parts.
High-temperature quenching and tempering is used to harden larger springs that
must be formed in the annealed condition. E.g. of spring materials are, steels for
cold-wound springs, Music Wire, Oil-tempered spring Wire, Hard-drawn spring
wire, Chrome-Vanadium steel wire, chrome-silicon wire, stainless-steel spring
wire, carbon- and alloy-steel bars for springs. Table 3.1 gives some common
spring materials and its description. Table 3.2 shows the mechanical properties of
some spring wires.
This gives an estimation of tensile strength when the intercept A, and the
slope m of the line are known. The values of these constants are given in
Table 3.3.
35
Table 3.1: Common spring wire materials [5]
36
Table 3.2 Mechanical properties of some spring wires [6]
Elastic Limit,
Diameter E G
Material percent of Sut
d, mm GPa GPa
Tension Torsion
Music wire A228 65—75 45—55 <0.8 203.4 82.7
0.8—1.6 200 81.7
1.6—3.0 196.5 81.0
>3.0 193 80.0
HD spring A227 60—70 45—55 <0.8 18.6 80.7
0.8—1.6 197.9 80.0
1.6—3.0 197.2 79.3
>3.0 196.5 78.6
Oil tempered A239 85—90 45—50 196.5 77.2
Valve spring A230 65—75 50—60 203.4 77.2
Chrome A 231 88—93 65—75 203.4 77.2
vanadium A 232 88—93 203.4 77.2
Chrome silicon A401 85—93 65—75 203.4 77.2
Stainless steel A31 3 65—75 45—55 193 69.0
17-7PH 75—80 55—60 208.4 75.8
414 65—70 42—55 200 77.2
420 65—75 45—55 200 77.2
431 72—76 50—55 206 79.3
Phosphor bronze 159 75—80 45—50 103.4 41.4
Beryllium copper 197 70 50 117.2 44.8
75 50—55 131 50.3
Inconel alloy X-750 65—70 40—45 213.7 77.2
37
Table 3.3 Constants A and m of Sut = A/d"" for Estimating Minimum
Tensile Strength of Common Spring Wires [6]
The torsional yield strengths of spring wire vary, depending on the material
and whether the spring has been set or not. Table 3.4 shows recommended
torsional yield-strength factors for several common spring wires as a percentage
of the wire's ultimate tensile strength. These factors should be used to determine
an estimated strength for a helical compression spring in static loading.
Torsional fatigue strength over the 10^ N 10^ cycles range varies with the
material and whether it is shot peened or not. Table 3.5 shows recommended
38
values for several wire materials in the peened and unpeened conditions at ttiree
points on their S-N diagrams at 10^, 10^ and 10^ cycles. Note that these are
torsional fatigue strengths and are determined from test springs loaded with
equal mean and alternating stress components (stress ratio R = T^i„/T^^=0).
So, they are not directly comparable to any of the fully reversed fatigue strengths
generated from rotating-bending specimens because of both, the torsional
loading and the presence of a mean stress component. We use the designation
Sfw, for these wire fatigue strengths to differentiate them from the fully reversed
fatigue strengths. These fatigue strengths Sfw are nonetheless very useful in that
they represent an actual (and typical) spring-fatigue-loading situation and are
generated from spring samples, not test specimens, so the geometry and size
are correct. Note that the fatigue strengths in Table 3.5 are declining with
increasing number of cycles, even above 10® cycles, where steels usually display
an endurance limit.
Table 3.5 Maximum Torsional Fatigue Strength Sfw for Round-Wire Helical -
Compression Springs in Cyclic Applications (Stress Ratio, R=0) [5].
39
contact. Once shut, spring can support much larger "indefinite" loads up to the
compressive strength of wire. The clash allowance Ydash is the difference
between the minimum working length and the shut height, expressed as a
percentage of the working deflection. A minimum clash allowance of 10-15% is
recommended to avoid reaching the shut height in service with out-of-tolerance
springs, or with excessive deflections.
naload
preload
maximum
working indefinite
load load
rcbsH
^
free length auembled length min wmrking length shut length
Four types of end details are available on helical compression springs and
they are, plain, plain-ground, squared and squared-ground as shown in fig 3.3.
Plain ends result from simply cutting the coils and leaving the ends with the same
pitch as the rest of the spring. This is the least expensive end detail, but provides
poor alignment to the surface against which the spring is pressed. The end coils
40
can be ground flat and perpendicular to the spring axis to provide normal
surfaces for load application. Squaring the ends involves yielding the end coils to
flatten and remove their itch. This improves the alignment. A flat surface on the
end coil of at least 270° is recommended for proper operation. Squaring and
grinding combined provides a 270-330° flat surface for load application. It is the
most expensive end treatment but is nevertheless recommended for machinery
springs unless the wire diameter is very small (<0.02 in or 0.5 mm), in which case
they should be squared but not ground.
C=^ 3.1
d
The preferred range of C is from 4 to 12. At C<4, the spring is difficult to
manufacture, and at C>12 it is prone to buckling and also tangles easily when
handled in bulk.
41
)
Fig 3.4 Forces and Torques on the Coils of a Helical Compression Spring
Figure 3.4 shows a portion of a helical coil spring with compressive axial
loads applied. The load on the spring is in compression, the spring wire in torsion,
as the load on any coil tends to twist the wire about its axis. A helical
compression spring is, in fact, a torsion bar wrapped into a helical form, which
packages better. The deflection of a round-wire helical compression spring is
easily obtained using Castigliano's theorem. The total strain energy for a helical
spring is composed of a torsional component and a shear component. The strain
energy is
T2| F2|
U= 3.2
2GJ 2AG
fF^D^^F^DNa
3.3
d^G d2G
42
SFD^Na^ SFD^Na
y= 1+- 3.5
d^G 2C^
Where F is the applied axial load on the spring, D is mean coil diameter, d is wire
diameter, Na is number of active coils, and G is the shear modulus of the
material.
The spring rate or stiffness or spring constant is defined as the load required
per unit deflection of the spring. Spring rate, k =F/y Where F is the Load, and
y = Deflection of the spring. The standard helical compression spring has a spring
rate k that is essentially linear over most of its operating range, as shown in
Figure 3.5. The first and last few percent of its deflection have a nonlinear rate.
When it reaches its shut height Lg, all the coils are in contact and the spring rate
becomes the stiffness of the solid coils in compression. The spring rate should
be defined between about 15% and 85% of its total deflection and its working
deflection range La - Lm kept in that region.
The equation for spring rate is found by rearranging the deflection equation.
d^G
k= 3.6
SD'^Na
0 15 85 100
% Deflection
43
Fig 3.6 Stress distributions across wire in a helical compression spring
_Tr^F_F(D/2)(d/2)^ F
t max 3.7
J A pdi^/32 pd^M
8FD 4F 3.8
pd'^ pd^
We can substitute the expression for spring index C from equation 3.1 in equation
3.8.
8FC^_4F_ 8FCH-4F
t 3.9
max pd^ pd' Pd2
8FD
*max '^s 3.10
pd-^
Where Kg = f l + ^ 3.11
To consider the effect of curvature and direct shear A IV! Wahl [4] determined the
stress-concentration factor for this application.
44
W 4C-4 C
t -K ^^^ '^^'^
pd
Since Wahl's factor Kw includes both effects, we can separate them into a
curvature factor Kc and the direct shear factor Ks using
Kw=KsKc 3.14
K^=^ 3.15
When the free length of the spring (LF.) is more than four times the mean or
pitch diameter (D), then the spring behaves like a column and may fail by
buckling at a comparatively low load. The critical axial load (Wcr) that causes
buckling may be calculated by using the following relation, i.e.,
Wcr=kxKBXLF 3.16
45
Table 3.6 Values of buckling factor (KB)
LF/D Hinged end Built-in end LF/0 Hinged end Built-in end
spring spring spring spring
1 0.72 0.72 5 0.11 0.53
2 0.63 0.71 6 0.07 0.38
3 0.38 0.68 7 0.05 0.26
4 0.20 0.63 8 0.04 0.19
Sometimes, the load on the springs does not coincide with the axis of the
spring, i.e., the spring is subjected to an eccentric load. In such cases, not only
the safe load for the spring reduces, the stiffness of the spring is also affected.
The eccentric load on the spring increases the stress on one side of the spring
and decreases on the other side. When the load is offset by a distance e from the
spring axis, then the safe load on the spring may be obtained by multiplying the
D
axial load by the factor where D is the mean diameter of the spring.
2e + D'
U = -F.y 3.17
2
t=Kx-8F.D or F = 3.18
pd"^ 8K.D
46
We know that deflection of the spring,
v=
SF.D-^.Na Sxpd^.t
= —!- V
D'^.Na = 1-pt.D^.Na 3 19
G.d^ 8K.D G.d'^ K.d.G
4K2.G V4 J 4K2.G
Any device with both mass and elasticity will have one or more natural
frequencies; springs are no exception to this rule and can vibrate both laterally
and longitudinally when dynamically excited near their natural frequencies. If
allowed to go into resonance, the waves of longitudinal vibrations, called surging,
cause the coils to impact one another. The large forces from both the excessive
coil deflections and impacts will fail the spring. To avoid this condition, the spring
should not be cycled at a frequency close to its natural frequency. Ideally, the
natural frequency of the spring should be greater than about 13 times that of any
applied forcing frequency.
Wn =
- ' •—
^ ^^ Rad/sec f^,f =
n -= -—^
- ^ Hz 3.23
" VW., " 2\W,
47
Where k is the spring rate, Wg is the weight of the spring's active coils, and g Is
the gravitational constant. It can be expressed either as angular frequency (o„ or
linear frequency/„. The weight of the active coils can be found from
\N^= ^ 3.24
Where y is the material's weight density, G is shear modulus. For total spring
weight substitute Nt for Ng.
fn=-^4J—Hz 3.25
" TTNaD2V32y
When a wire is coiled into a helix, tensile residual stresses are developed at
the inner surface and compressive residual stresses occur at the outer surface.
Neither of these residual stresses is beneficial. They can be removed by stress
relieving (annealing) the spring.
48
at shut height about 10 to 30% greater than the yield strength of the material.
Less than that amount of overload will not create sufficient residual Stress. More
than 30% overstress adds little benefit and increases distortion.
Helical springs of square or rectangular bar section are sometimes used for
cases where a large amount of energy must be stored within a given space.
Particularly if the spring is coiled flat wise, it is clear that a larger amount of
material may be provided within a given outside diameter and compressed length
than if a circular section were used. More energy may be stored within a given
space for such a design than would be the case if a circular bar section were
used. Although the rectangular bar section theoretically does not have as
favorable an elastic stress distribution as does the round bar section, for static
loading or loads repeated only a few times this disadvantage is of no particular
importance, since local yielding of the highest stressed portions can occur without
appreciably affecting the performance of the spring or the capacity for storing
energy. However, where fatigue or repeated loading of the spring is present, this
non uniformity of stress distribution will be a disadvantage. A further
disadvantage is the fact that the quality of material used is generally not as good
as would be the case where round wire is used; also, the rectangular-bar material
may be difficult to procure.
In general, when bar stock of rectangular section is coiled to a helical form, a
keystone or trapezoidal shape of cross section finally results, and this tends to
reduce the space efficiency and energy storage capacity. Springs with
rectangular cross sections having the long side of the section parallel to the axis
are sometimes used in the design of precision scales in order to obtain a more
nearly linear load-deflection characteristic.
In Contrast to round-wire helical compression or tension springs where
curvature effects can be neglected in calculating deflections, such effect are
particularly important in rectangular-Wire springs coiled flat wise. In such cases
neglecting curvature may result in errors 15 per cent or more.
For springs made of rectangular wire, as shown in Fig 3.7, the maximum shear
stress is given by
49
F.D(1.5t + 0.9b)
t =Kw 3.26
b2.t2
This expression is applicable when the longer side (i.e., t > b) is parallel to the
axis of the spring. But when the shorter side (i.e., t < b) is parallel to the axis of
the spring, then maximum shear stress,
,F.D(1.5b + 0.9t)
t =Kw 3.27
b2.t2
And deflection of the spring.
2.45F.D^.Na
y=. 3.28
G.b^(t-0.56b)
For springs made of square wire, the dimensions b and t are equal. Therefore,
the maximum shear stress is given by
2.4F.D
t=Kx- 3.29
5.568F.D^.n 5.568F.C^.n
y= 3.30
G.b^ G.b
^ 4C-1 0.615 ^ ^ D
K= + , and C = —
4C-4 C b
50
3.14 DESIGN OF SPRINGS
The primary objective of the spring designer is to design a spring which will
do the job required and at the same time be the most economical, all factors
considered. This means that the spring must fit into the space available and have
a satisfactory life in service not only from the standpoint of fatigue breakage but
also from that of excessive relaxation or set in service.
In other cases where failure may endanger life or property, the designer may
require maximum reliability even at increased cost. This is the case, for example,
in aircraft-engine valve springs. In such cases, more expensive materials, such
as valve-spring-quality wire, and additional processing costs, such as those due
to shot peening, may be justified. For certain applications, the designer may wish
to obtain a spring of minimum weight, volume, or length.
One of the most important decisions to be made by the spring designer is the
choice of the proper spring material. Since the primary purpose of most springs is
to store energy, and since the energy stored for a given volume of material is
proportional to the square of the stress, it is of advantage to use a high-strength
material which will permit operation at relatively high stresses. For example, a 10
per cent increase in working stress generally means a 20 per cent reduction in
amount of spring material required. This explains the wide use of spring steel,
which has relatively high tensile strength, and also explains why springs are
generally stressed higher than is the case for other applications.
51
Change in torsional modulus with temperature, Vibration and Impact effects,
Effects due to load eccentricity, Tolerances, etc.
The type of loading to which the spring is subjected is the important factor to
consider in designing various types of mechanical springs.
Static loading-normal temperature. This category refers to springs subject to a
steady load or one repeated, say, less than 100 to 1,000 times. (An example is a
spring used to apply gasket pressure.) In such cases the chief problem is usually
to avoid excessive set or load loss. Thus, if a helical spring is compressed by a
certain amount, the load may drop or relax with time; if the spring is loaded by a
constant load, the spring may take a set or creep. In practical design, this
relaxation must usually be limited to the initial load.
In general, at normal temperatures if the peak stress in the spring is kept
below the elastic limit of the material, trouble from set or relaxation will seldom
occur. In the case of springs which are preset, the nominal working stress may, in
some cases, be higher than the elastic limit. This is due to favorable residual or
trapped stresses which are set up during the presetting operation and which
subtract from the load stress.
In the case of statically loaded springs it is common practice to neglect
stress-concentration effects in calculating stresses. (Such effects occur because
of sharp bends, holes, notches, etc.) This is justified since the usual spring
material has sufficient ductility that stress relief may occur at localized points.
Static loading-elevated temperatures. At elevated temperatures, it is found that
the effects of creep or relaxation become much more pronounced than at normal
temperatures, i.e., the springs tend to set or lose load more rapidly. This load loss
or relaxation is also a function of time. For example, for temperatures around 250
to 350° F, music or oil-tempered wire helical springs may be used, but for higher
temperatures other materials such as stainless steel, Monel, Inconel and Inconel
"X" are generally required. In such cases the design must be based on the creep
and relaxation properties of the material at the operating temperature.
52
Time
times in certain spring applications (valve springs) while in others, it may occur
perhaps only a few times. The allowable stresses in the two cases are
considerably different. In general, the stress rangecr^ is of particular importance
where fatigue is involved, since for many materials the limiting endurance range
is approximately constant provided the yield stress is not exceeded.
Fatigue tests on helical or leaf springs show that the surface condition is
particularly important where repeated loading is involved. In hot-wound helical
springs especially, decarburization of a thin surface layer usually occurs, and this
reduces fatigue life. Surface defects such as quench cracks, pits, coiling-tool
marks, and seams are also detrimental to spring endurance. Hence, where
repeated loading is present, such defects should be avoided, while the amount of
decarburization should be kept to a minimum.
53
In general, it is also advantageous to use the shot-peening process for
springs subject to fatigue loading. This process consists in propelling hardened
steel shot at high velocity against the spring surface with an air blast or
centrifugal-type machine. This peening action cold-works the surface and sets up
beneficial compressive stresses, both of which tend to increase the endurance
strength. This process, however, cannot compensate for an excessive thickness
of decarburized layer or for defects present.
Another factor which the spring designer should keep in mind is that there is
always an unavoidable variation in the size of wire or plate used in making
springs. The effect of these variations may often be large, especially when it
comes to obtaining proper load-deflection characteristics. For example, in the
case of helical springs, a cumulative variation in both coil and wire diameter of
only 1 per cent will result in a 7 per cent change in the load-deflection
characteristic of the spring. Thus for 0.1-in, wire, a 1 per cent variation would
correspond to a change in diameter of only 0.001 in. Such variations are easily
possible in commercial practice. Hence, it may be necessary to allow the spring
manufacturer some leeway in choosing the other spring dimensions to
compensate for unavoidable variations in sizes of commercial wire stock.
The functional requirements for a spring design can be quite varied. There
may be a requirement for a particular force at some deflection, or the spring rate
may be defined for a range of deflection. In some cases there are limitations on
the outside diameter, inside diameter or working length. The approach to design
will vary depending on these requirements. In any case, spring design is
inherently an iterative problem. Some assumptions must be made to establish the
values of enough variables to calculate the stresses, deflections, and spring rate.
Because wire size appears to the third or fourth power in the stress and
deflection equations, and because material strength is dependent on wire size,
the safety of the design is very sensitive to this parameter.
54
Many approaches may be taken to spring design, and more than one
combination of spring parameters can satisfy any set of functional requirements.
It is possible to optimize parameters such as spring weight for a given set of
performance specifications. To minimize weight and cost, the stress levels should
be made as high as possible without causing static yielding in service.
The stress state is compared to the yield strength for static loading. The
safety factor for static loading is
N-=-^ 3.31
^ t
If the calculated stress is too high compared to the material strength, the wire
diameter, spring index, or material can be changed to improve the result. When
the calculated stress at the required operating force seems reasonable compared
to the material strength, a trial number of coils and a clash allowance can be
assumed and further calculations for spring rate, deflection, and free length done
using equations 3.5 and 3.6. Unreasonable values of any of these parameters will
require further iteration with changed assumptions.
After several iterations, a reasonable combination of parameters can usually
be found. Some of the things that need to be checked before the design is
complete will be the stress at shut height, and the Di, Do and free length of the coil
with respect to packaging considerations. In addition, the possibility of buckling
needs to be checked.
55
3.16 DESIGNING COMPRESSION SPRINGS FOR FATIGUE LOADING
mm
Let us now consider a spring subjected to a force varying between Fmax and
Fmin in a load cycle the mean and amplitude forces are given as:
Fmax + Fmin
Mean force, Fm = 3.32
56
Fmax-Fmin
Amplitude force, Fa = 3.33
8FmD
The mean torsional shear stress tm =Ks 3.34
Where Ks = shear stress concentration factor = 1 + 0.5 / C It is used only for mean
shear stress calculation
The torsional shear stress amplitude
8FaD
t.=KX 3.35
nd"^
Where K is the Wahl's stress corrective factor
The design of a spring under fatigue load conditions is based upon the modified
soderberg line approach as shown in fig 3.11. Accordingly the point A can be
found by drawing a line OA inclined at 45° with the X-axis, which intersects the
original soderberg lir^e at poir^t A. The point A on this diagram indicates the
limiting value of stress due to pulsating load condition. The point B on the X-axis
indicates the limiting value of stress due to static load condition. Thus the line
joining the points A and B is called soderberg line of failure. To take into account
the effect of factor of safety (FoS), a line CD is drawn parallel to the line AB, from
point D on the X-axis where OD = ty/FoS, The line CD is called the modified
soderberg line of failure.
According to the soderberg hypothesis, any point lying either on line CD or within
triangle O'CD is considered safe. The modified design equation is
1
=tm-ta/ty+2ta/t« 3.36
Fos
57