122spec Design

Design
Handbook
Engineering
Guide
To SpringDesign
1987 Edition
AssogffilfiS
l0Mainstreet,
Brisro,,.*::$il3:
&ffi&ffiffi$ ffi
Table of Contents
section
Eor to Use the Handbook
Selccdng Spring Configurations

hrsh. Pull. Twist or Energy Stofage Applications: CommonAvailable Configurations.
Spring Materials
Common Specifications,Elastic Modulus, Magnetic Characteristics,Heat Treatment, Stress Relaxation,
Corrosion, Coatingsand Finishes.SpringWire: Tensile Properties,Cost and Availability.
Sprine Strip: Strength,Formability and Edge Condition.
Reidual Stress, Fatigue and Reliability
Load-CarryingAbility, FatigueTerminology, Modified GoodmanDiagram, Weibull Analysis, Load Loss.
Eelical Compression Springs
GeneralDefinitions, Squareness,Parallelism, Hysteresis,Design Equationsfor Spring Rate and Stress,Bucklinl
Choice of Operating Stress for Static and Cyclic Applications, Dynamic Loading Impict and Resonance,
RectangularWire, StrandedWire, Variable Diameter, Variable Pitch and Nested Springs,CommercialToleranc,
Hot-Wound Springs
DesignConsiderations,End Configurations, Materials, Choice of Operating Stress,Tolerances.
Eelirnl Extension Springs
Initial Tension, Types of Ends and Dimensions, Design Equations,Choice of OperatingStressfor Static
and Cyclic Applications, Commercial Tolerances.
D
D
-i
D
I
I
b
D
a
a
a
a
a
a
a
a
a
Garter Springs
Joint Design, Design Equationsand Tolerances.
Helical Torsion Springs

Mean Diameter, Length, Design Equations for Rate and Stress,End Configurations,Natural Frequency,Choice
of Operating Stress for Static and Cyclic Applications, Double Torsion and Rectangular Wire Springs, ioleranc,
l0
ll
a
a
a
I
f,staining Rings
External and Internal Types, Ends, Design Equations, Choice of StressLevel, Tolerances.
Belleville Spring Washers
Load-Deflection Characteristics,Mounting, Design Equations.Choiceof StressLevel for Static and
Cyclic Applications, Stackingand Tolerances.
12
Flat Springs . .
Design Considerationsand Equations for Cantilever and Simple Beams, Choice of StressLevel and
Tolerances.
13
Specid Spring Washers

Design Considerationsand Equations for Curved, Wave and Finger Washers, Choice of StressLevel and
Tolerances.
l{
Power Springs
GeneralDesign Considerationsand Equations, OperatingStressfor Power and hestressed Power
Springs.
lS
Constant Force Springs

Extension Type_,Design Equations, Mounting and Tolerances,Motor Type Design Equations for "A"
and "B" Type Motors, OperatingStress and Tolerances.
f6
Spird Springs
Design Equations for Hair Springsand Brush Springs.
17
Volute Springs
Design Equations and Choice of Operating Stress Level.
It
Wire Forms
GeneralInformation and How to Specify.
19
IndexandReferencelnformation
.......1
Glossary of Spring Terminology, Bibliography, Trademarks, Conversion Factors, Abbreviations and Symbols,
HardnessScaleConversions,Index and Lists of Thblesand Illustrations.
t
I
Page
il
il
il
il
il
n
il
n
M
H
H
M
M
ffi
ffi
ffi
ffi
ffi
a
I
Assg&tfi8&#*ffiffi$H
a
=)
SelectingSpring Configurations
Tabb 2-1. Spring Configurations.
T\'PT.
=)
-
Tv?u
COTNCUNATION
A.rTION
(Section5)
Helical Compression
Round and
Recrangular
\t-ire
Beam
cor'if"IstJ&ATIoli
(Section 12)
hrsh or pull-wide range
of loads, low deflection
range.
Cantilever,
RectangularSection
Push-wide load and deflection range-constant

rate'
Constant Pitch
ACTt0:ri
cantilever,
TrapezoidalSection
Pl
a
a
e
=*
:
,:
=
::=/
:/
^
Conrcal ::
I
Barrel
a
e
a
Push-wide load and deflection range. Conical

spring can be made with
minimum solid height and
with constantor increasing
rate. Barrel, hourglass,and
variable-pitchspringsused
to minimize resonant surging and vibration.
.>
Hourglass
ffi
SimpleBeam
Helical Torsion (Section9)
frS
RoundorRectang",." ffi
(2'il[
wire
Spiral (Section16)
#
r.-.1r.-fi
--.r--
Variable-PitchI
Spring Washer
Belleville
a
a
\lare
a
=a
5lotted
=f
Finger
(Sectionll and13)
/--\
\:r'
Vx
ql
/raR
rye
ft\1
V
=f
rl
a
a
a
a
/R
\ \ - / l
Curved
-:
Volute
,^
Hairspring
Twist.
rwistorhrsh'
Brush
Push-high loads, low deflections-choice of rates
(constant, increasing, or
decreasing).
rwist<onstantrate'
:ili',y:,::;'@
Twist---exerts torque over

many turns.
Push-light loads, low

deflection-uses limited
radial space.
Suppliedin retainer.
ad)
Removedfrom retainer.
Ptrsh-higher defl ection s

than bellevilles.
\/
hrsh-for axial loading of

bearings.
Prestressed
Power (Section
14)
Push-used to absorb axial

. end play.
Twist--+xerts torque over

many turns.
Suppliedin retainer.
(Section 17)
Removedfrom retainer.
Push-may have an inherently high friction damping.
Constant (Section15)
Force
Spring
Motor
Twist+xerts close-toconstant torque over many

turns.
Level Torq
a
a
h
AssogFifig/&sAnss
ffi
Spring Materials
Chemical and Physical Characteristics
While certain materials have come to be regarded as
ry.i.ng materials, they are not specially design-d alloys.
Spring materials are high strength alloys wtrictr ofien
exhibit the greatest strength in the alloy system. For
example: in steels, medium and high-carbon steels are
regarded as spring materials. Beryllium copper is frequently specified when a copper base alloy is required.
For titanium, cold-worked and aged Ti-l3v-llCr-3At is
used. The energy storage capacity of a spring is proportional to the square of the maximum operating stress
level divided by the modulus. An ideal spring material
has high strength, a high elasticlimit and a low modulus.
Becausesprings are resilient structures designedto undergo large deflections, spring materials must have an
extensive elastic range. Other factors such as fatigue
strength, cost, availability, formability, corrosion resistance, magnetic permeability and electrical conductivity
Tahle 3-1 Typical Properties of Common Spring Materials.
Com*on lihme
Y,o*qt*
Modulrrs S {l}
MPr | {ed}
r t ' l]
id
il{odnkm of
Rlddly G {l}
lWPr | {Dd}
rdl
id
&lcr .lesl
Corr&rc. Mtf
{t}
*iY,lty ttl
*1ffir5{tbl,tu'; % f*c$
$ias lioruelly
.tvrlhblc {2}
Min.
Mrr.
i
mm {hr.} i mm {h.}
7.86 (0.2E4)
7.85 (0.2E4)
7.E6 (0.2E4)
7.86 (0.284)
0.10(0.004)6.35(0.250) a
0.13(0.00s)16 (0.625) c
0.s0(0.020)16 (0.62s) c
1.3 (0.0s0) 6.35(0.250) a
r20
150
r50
150
250
250
300
300
r'rp|nl. ,:,.,ilil1fua
S{r&eG ,fi
..fcrycr.
i,,'i
, ({}
Qt&y, , ot
3F
t l
Carbon Steel Wires:

Music (5)
Hard Drawn (5)
Oil Tempered
Valve Spring
207
207
207
207
(30)
(30)
(30)
(30)
79.3
79.3
79.3
79.3
( l l .s)
( l l .s)
( l 1.5)
(l1.5)
Alloy Steel Wires:

Chrome Vanadium
Chrome Silicon
207
207
(30)
(30)
79.3
79.3
( l 1 . 5 ) 7.86 (0.284)
( l 1 . 5 ) 7.86 (0.2E4)
7
5
0.50(0.020)l r (0.435) a r b
0.50(0.020) 9.5 (0.375) arb
220
245
425
475
StainlessSteel Wires:
Austenitic Type 302
hecipitation
Hardening l7-7 PH
NiCr A2E6
193
203
(28)
(2e.s)
69.0
75.E
(10.) 7.92 (0.286)

(l l)
7 . E l (0.2E2)
2
2
0.13(0.00s) 9.5 (0.375) b

0.08(0.002)r2.5 (0.500) b
2@
315
500
600
200
(2e)
7r.7
(10.4) E.03 (0.290)
o.lm(0.016)5
510
950
Copper Base Alloy Wires:

PhosphorBronze (A)
Silicon Bronze (A)
Silicon Bronze (B)
Beryllium Copper
Spring Brass, CAz6/u.
103
103
t17
128
lr0
(15)
(15)
(17)
(1 E.5 )
(1 6 )
43.4
38.6
4.1
b
b
b
b
b
9s
95
95
205
95
200
200
2m
400
200
2r4
2t4
lE6
179
r79
(3l)
(3l)
(27)
(26)
(26)
75.E
79.3
62.9
6.2
6.2
(lt)
8.43 (0.304)
( l 1 . 5 ) 8.25 (0.298)
3.5
3
0.10(0.004)l2.s (0.500) b
0.10(0.004)12.5 (0.500) b
0.10(0.004)r2.5 (0.500) b
0.05(0.002) 9.s (0.375) b
0.05(0.002) 9.5 (0.375) b
320
595
9s
230
2ffi
700
ll00
200
450
500
207
2W
2W
207
207
(30)
(30)
(30)
(30)
(30)
79.3
79.3
79.3
79.3
79.3
( l l . s ) 7.E6 (0.2E4)
( l 1 . 5 ) 7.86 (0.284)
( l 1 . 5 ) 7.E6 (0.2E4)
( l 1 . 5 ) 7.86 (0.284)
7
7
7
7
7
0.25(0.010) 3
0.0B(0.003) 3
0.0E(0.003) 3
0.08(0.003) 3
0.10(0.004) l
(0.125) b
(0.125) b
(0.125) b
(0.125) b
(0.040) a
95
95
r20
r20
9s
200
2W
250
250
200
193
(28)
69.0
(10)
7.92 (0.2E6)
0.0E(0.003) 1.5 (0.063) b
315
600
203
(29.s)
75.8
(11)
7 . 8 1 (0.282)
0.08(0.003) 3
(0.125) b
370
700
103
r28
(1 5 )
(1 8 .5 )
43
48
0.08(0.003) 5 (0.18E)b
0.08(0.003) 9.5 (0.375) b
9s
205
200
400
Nickel Base Allovs:

Inconelo Alloy 600
Inconel Allov X750
Ni-Span-C@
Monilo Allov 400
Monel Alloy K500
Carbon Steel Strip:
AISI 1050
l06s
1074, 1075
r09s
Bartexo
StainlessSteel Strip:
Austenitic Types
301, 302
Precipitation
Hardening l7-7 PH
CopperBaseAlloy Strip:
PhosphorBronze(A)
BerylliumCopper
+8.t
42.0
(6.3)
(5.6)
(6.4)
(7.0)
(6.0)
8.86 (0.320) l 5
E.s3 (0.308) 7
E.75 (0.316) t 2
8.26 (0.298) 2r
8.53 (0.30E) t7
(e.7) 8.14 (0.2%)

(e.6) 8.83 (0.319)
(e.6) E.46 (0.306)
(l r.5) 7.E6 (0.2E4)
r.5
(6.3) E.16 (0.320) l 5

(7.0) 8.26 (0.298) 2 l
(l) Elastic moduli, density and electrical conductivity can vary with
cold work, heat treatment and operating stress. These variations are
usually minor but should be considered if one or more of these
properties is critical.
(2) Diametersfor wire; thicknessesfor strip.
(3) Typicd surface quality ratings. (For most materials, special processes can be specified to upgrade typical values.)
a. Maximum defect depth: 0 to 0.5Vo of d or t.
H Assog5ifiS
&ffienans
1.5
I
(0.200) b
0.10(0.004)12.5 (0.500)
0.10(0.004)12.5 (0.s00)
0.r0 (0.004)tz.5 (0.500)
0.0E(0.003)12.5 (0.500)
0.10(0.004)12.5 (0.500)
b. Maximum defect depth: l.$Vo of d or t.

c. Defect depth: less than 3.5Voof d or t.
(4) Maximum service temperatures are guidelines and may vary due
to operating stress and allowable relaxation.
(5) Music and hard drawn are commercial terms for patented and
cold-drawn carbon steel spring wire.
INCONEL, MONEL and NI-SPAN-C are registeredtrademarksof
International Nickel Company, Inc. BARTpi is a registered trademark of Theis of America. Inc.
G
C!
G
G
G
G
G
G
G
G
G
G
G
C
C
G
C
C
C
C
C
G
C
C
C
C
-
C
G
(F
,a
Spring Materials
il
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
It
il
tl
!t
a
a
a
a
a
a
a
a
t
--.
can also be important and must be consideredin light

of cost,benefit. Consequently,careful selectionsmust
te made to obtain the best compromise.
Table3- I lists some commonly used alloys along with
dau for material selectionpurposes.Data on mechanical
propenies are presentedin the Spring Wire and Spring
Strip subsections(Pages18 and 20 respectively).Specifications have been written by many national and intern a t i o n a l o r g a n i z a t i o n s .T h e s e s p e c i f i c a t i o n sa r e
cross-referenced
to AssociatedSpring specificationsin
Table 3-2r However, correlation between the specifications is only approximate.AssociatedSpring specificat i o n s w e r e d e v e l o p e de x c l u s i v e l y f o r h i g h q u a l i t y
material for spring applicationsand are generally more
detailedand stringentthan other specifications.
Surface quality has a major influence on fatigue
strength and is often not clearly delineated on national
specifications.It is important to use only those materials
uith the best surface integrity for fatigue applications,
particularly those in the high cycle region.
In steel alloys, for which processingcosts are a large
fraction of product cost, surface quality can vary over
an appreciable range. Depth of surface imperfections,
such as seams,pits and die marks, can be up to3.5%
of diameter for commercial spring wire grades (ASTM
A-227 and A-229). Various intermediate qualities can be
obtained. Highest levels are representedby music and
valve spring quality grades which are virtually free of
surface imperfections. Decarburization, which can also
adversely affect fatigue performance, follows a similar
pattern. Surface quality of spring materials is a function
of the care exercised in their production and processes
employed. Materials produced with a high level of surface integrity are more costly than commercial grades.
Elastic Modulus
The modulus of elasticity in tension and shear is vital
to spring design. Table 3-1 lists recommendedvaluesfor
commonly used spring alloys. For most steels and agehardenablealloys, the modulus varies as a function of
chemical composition, cold work and degree of agrng.
Usually variations ar0 small and can be compensatedfor
by adjustment of reference parameters of the spring design, (e.g. number of active coils, and coil diameter).
For most materials,moduli are temperature-dependent
and vary inversely with temperature by approximately
ZVcper 55'C (100"F).Sincenonambienttemperaturetesting is costly, design criteria should be specifiedat room
temperature after having made appropriate compensation for the application temperature. Certain nickelchromium-iron alloys are designed to have a constant
modulus over the temperaturerange from -5o to 65'C
(-50" to 150'F) and are exceptions to the above rule.
For true isotropic materials, the elastic moduli in tension
(E) and shear (G) are relatedthroughPoisson'sratio by the
expression:
E
P:6-r
sothat, for common spring materials,any one of the parametersmay be approximatedusingthe other two.
Magnetic Characteristics
For most applications, the question of "magnetic or
not" is adequatelyansweredwith the use of a permanent
magnet. For some applications,even very low levels of
magneticbehavior can be detrimental. Then, it is desirable to know the magnetic permeability of candidate
materials and reach agreement between parties on a
maximum allowable value. Table 3-3 lists approximate
values for a number of low permeability materials along
with other frequently used alloys.
Since permeability can be altered by cold work, some
variation can be expected. In general, low permeability
materials are more expensive so designersshould specify
low levels only when absolutely necessary.Often, nitrogen strpngthened manganese stainless steels are good
choices because they have good strength at moderate
cost.
Heat Treatment of Springs
Heat-treating temperatures for springs can be divided
into two ranges. Low temperature heat treatments in the
175'to 510'C (347'to 950'F) range are appliedto springs
after forming to reduce residual stresses and stabilize
parts dimensionally. For carbon steels, stainless steels
and some age-hardenablealloys, low temperature heat
treatments are used to increase or restore the set point.
Electroplated carbon steel parts are heat-treated at low
temperatures prior to plating, and baked afterward to reduce the susceptibilityto hydrogen embritflement. Most
low temperature stress relieving and age-hardening of
springs are done in air and a moderate amount of oxide
may be formed on the part. No detrimental effects of this
oxide have been noted.
High temperature heat treatments are used to strengthen annealed material after spring forming. High-carbon
steels are strengthenedby austenittzing in the temperature range 760'to 900"C(1480"to 1652"F),quenching to
form martensite and then tempering. Some nickel base
alloys are strengthenedby high temperature aglng treatments. Becausesubstantialoxidation occurs at these elevated temperatures,it is advisableto prevent excessive
oxidation by using an appropriate protective atmosphere.
Heat treatments suitable for many commonly used materials are listed in Table 3-4. Selection of a temperature
within a given range can only be made after considering
the material, size, strength level, application conditions
and desired characteristics.For additional guidance,Associated Spring engineersshould be consulted. Unless
otherwise noted, 20 to 30 minutes exposureat temperature is sufficient to obtain the bulk of the stress-relieving
effect.
Many spring-like parts involve forms which preclude
the use of prehardenedmaterial. In these cases, soft or
annealedmaterial must be used and heat-treatedto attain
spring propertiesafter forming. Thin high-carbon and
alloy steel parts become distorted when hardened by
quenching. Distortion may be reduced by fixture tempering; however,this process is costly and should be
avoided if at all possible by using pretempered materials.
AssgFifig
&ffi*ffiwsffi
il
Spring Materials
Table3-2. RelatedSpringMaterial Specifications.
.twoeint*d
$tri'n*
Corrrnon
Trldr F{ams
saa
ASTM
AM,S
Mfliarl'
EN
Ihderd
Britlih
8S
fieffirlr
':f,},[H,
,lltfsrye
.tI$
G3522,
SWP-A,B, V
G
G
Spring Wire
AS-5
MusicWire
1085
Jl78
1228
AS-10
Oil Tempered
CarbonSteel
1066
J316
L229
AS.2O
Cold Drawn
CarbonSteel
1066
J r1 3
A227
AS-25
Oil Tempered
Carbon Steel*
1070
42,30
5115
AS-32
Oil Tempered
Chrome
Vanadium*
6150
A232
&50
AS-33
Oil Tempered
9254
ChromeSilicon* n57
AS-35
Stainless Steel
30301 A 3 1 3 ; 5688
30302 Type301,
J230 Type302
AS.36
r7-7PH
J2T7
AS44
InconelX-750
5698.5699
AS-45
CopperBeryllium cA-t72 Bt97
4725,
Cond.A
AS-55
SpringBrass
AS-60A.
PhosphorBronze cA-510B159,
tr32
5tL2
s4@9
QQv/470
(obsolete)
1408or 5216
520r
17223,SheetI
1.1200
QQw428
2803,grade3
17223,Sheet2, G3560,
1.1230
swo-A,B
49B
w-22826
A40l
17222,SheetI
1408
5216NS or HS 1.0500
sw-A,B,c
2803,
GradesI & 2
swo-v
17225,
50CrV4
G3565,
QQw4l2 48A
17225,
67SiC15
G3566,
2056
A'313, 5678
I}pe 531
cA-2@ BTY,
n60
#5rc
AS-60C
PhosphorBrorue
cA-52r #521
AS-70
Chromium Steel
s160H A'304
4720
L7223.Sheet2 G3561,
QQw412 47
50
QQw423 58A
(obsolete)
C
G
G3521,
sus302
t7224,
1.4568
G4314,
2873,
cBl0l
1766,6,
2.t247.55
QQW-321,
n@
2786,
czrw
17660,
2.0265
QQw40l
2873,
PBlO2
17662,
2.1030.39
C
C
swosc-v
1.4300
1.4310
t7224
QQW-530,
Cond.A
swocvry
G43T4,
sus63IJl
C
C
C
-
C
C
C
C
C
C
C
C
C
C
C
C
C
C
970, Part5
A689
Spring Strip
A682
A684
5r2l
5t22
r074 A682
1075 A684
5r20
AS-102
1050
A682
A684
5085
AS-103
1065
A682
A684
5115
AS-100
1095
AS-l0l
AS-105
Bartex
1085
AS-135-A
AS-135-B
Stainless Steel
30301 At77
30302
AS-136
t7-7 PH
AS-144
InconelX-750
AS-145
Copper
SpringBrass
AS-160C
Phosphor
Brorze
G3311,
SK4M
42E. 14/19,Parrt38, t7222,
G3311,
s75CM
1449,Part38,
csl00
cs.cs80
l$g,
cA-260 836,
cA-5r0 8103,
#510
l.t2r0
G33ll,
s50cM
Paft3B,
cs50
42F
5517
5 5 1 85, 5 1 9 s-5059
5542
cA-r72 8194,
f260
AS-1@A
17232,
| .t274
44D
QQS-766 58A
cs60.cs70
17222,
l.1230
G3311,
s65CM
14,y'l9,Parrt4,
302-S-25
t7224,
1.4310
1.4300
G43r3,
sus-301-csP
sus-302-csP
t7224,
1.4568
G4313,
1440,Part38,
5528
s-25043
5529Spring Cond. A
Temper
#r72
AS-155
s-7947
Annealed
Cold-Rolled
sus-631-csP
N 7786
4s30(AT)
4s32(LtzHT)
QQC-533
2870,
cB101
r7666,
2.1247.55
4507.vzH
QQB-613,
Comp.2
TIzH
2870,
czt08
17660,
2.026s
4510
Spring
Temper
QQB-750,
Comp.A
2870,
PB1O2
17662,
2.1030.39
PhosphorBronze cA-521 8r03,

#521
*Valve spring quality.
f&ffi*ffffis
MAssogb?ifi8
C
_
a
a
Spring Materials
T&
3-3. Magnetic Characteristicsof Some Materials.

Prfm**rbility rr W Onrr*tdn
Room Temperesrre
iffi
a
a
a
a
a
a
a
a
a
a
a
-{.!r
Brzsscs.Bronzes
Carbon Srcels
Flglol t
lnconel -{,lloys:
a
a
a
a
lr
a
a
a
I
a
a
a
a
I
I
Nonmagnetic
> 500
1.00003s
6m
l.0t
1.0006
1.0035
5r_<
x--50
Staintess
Steels:
Tlpe 301, spring temper
Tfpe 302, spring temper
631 07-7 PI{)
XV-28: Nitroniso 32*
\*itronic 50*
Titanium Alloys
>30
>12
>40
1.0il
r.004
Nonmagnetic
' \ itrogen-strengthenedmanganesestainless steels.

ELGILOY is a registered trademark of Katy Industries, Inc. NITRONIC
is a registeredtrademark of Armco., Inc.
Tabb 34. Typical Heat Treatments for Springs After

Forming.
1114
Hent,lftma-ffii:
tC, , ,:,
. ,,,,,,,
,,
.
Patentedand Cold-Drawn Steel Wire
TemperedSteel Wire:
Carbon
AllorAusteniticStainlessSteel Wire
l'.
,".,,; t'
190-230
2G400
3t5-425
23L510
PrecipitationHardening StainlessWire
rll-7 PH):
Condition C
480/l hour
ConditionA to TH 1050
760/l hour
cool to l5oC
followed by
565/ I hour
Vonel:
.{lloy'496
Allo.v K500, Spring Temper
Inconel:
.{lloy'600
Allol X-750:
* I Temper
SpringTemper
Copper Base, Cold Worked (Brass,
PhosphorBronze, etc.)
Beryllium Copper:
hetempered (Mill Hardened)
SolutionAnnealed,
Temper Rolled or Drawn
l:;i::T
l:zs-rso
lsruzso
600-800
I
45G950
900/l hour
1400/l hour.
cool to 60"F
followed by
1050/l hour
30L315
57ffio
525I 4 hours 98014hours
40G510
750-950
7301
16hours l350l16hours
4 hours n00l4 hours
6501
t75-205
3sG400
20s
400
3r5t2-3
hours
6W/2:3hours
AnnealedSteels:
Carbon(AISI 1050to 1095)
800$30*
1475-t525*
.{,llol'(AISI5150H6150,9254)
830{85*
1525-1625*
-Time dependson heating equipment
and sectionsize. Partsare austerutizedthen quenchedand temperedto the desired hardness.
Tempering is an effective stress-relieving treatment and

results in negligiblelevels of residual stress.Some spring
materials, such as beryllium copper and l7-7 PH, are
strengthened after forming by age hardening. This provides a good stressrelief, but may also cause distortion
unless special techniquesare used.
Environmental Considerations
Frequently, operating environment is the single most
important consideration for proper spring material selection. For successful application, material must be
compatible with the environment and withstand effects
of temperature and corrosion without an excessive loss
in spring performance. Corrosion and elevated temperatures decrease spring reliability. The effect of temperature on spring materials is predictable and discussed
below. Compatibility of spring materials and spring coating systemswith corrosiveenvironments is discussedin
general terms. For specific applications, the designer is
urged to rely upon previous experience or consult with
Associated Spring engineers.
Stress Relaxation
Primary concern for elevatedtemperature applications
of springs is stress rela:ration. Stress relaxation is the
loss of load or available deflection that occurs when a
spring is held or cycled under load. Temperature also
affects modulus, tensile and fatigue strength. For a given
spring, variables which affect stress reloration are:
stress, time and temperature, with increases in any parameter tending to increase the amount of rela;ration.
Stress and temperature are related exponentially to relaxation. Curves of relaxation versus these parameters
are concave upward as is shown in Figures 3-1 and 3-2.
Other controllable factors affecting relaxation include:
il
1. Alloy Type - more highly alloyed materials are generally more resistant at a given temperature or can be
used at higher temperatures.
2. Residual Stress - residual stresses remaining from
forming operations are detrimental to relaration resistance. Therefore, use of the highest practical stressrelief temperatures is beneficial. Shot peening is also
detrimental to stress relaxation resistance.
3. Heat Setting - various procedures can be employed
to exposespringsto stressand heat for varying times
to prepare for subsequentexposures. Depending on
the method used, the effect is to remove a usually
large first-stage relaxation and/or to establish a residual stress system which will lessen relaxation influences. In some cases,the latter approach can be so
effective that in application, compression springs may
"grow"
or exhibit negative relaxation. Increase in
free length does not usually exceed I to ZVo.
4. Grain Size - coarse grain size promotes relaxation
resistance. This phenomenon is used only in very
high temperature applications.
Becauseso many variablesare involved, it is impossible
to cite comprehensive data in a publication of this type,
but Table 3-l does show approximatemaximum service
a
I
---
o'"TFlfi8AH*ffilE$
H
|t
Spring Materials
Fig.3-1. RelaxationversusInitial Stressfor Spring Materials.
lnitiolstress( 103psi)
125
100
7'5
l8
Fig. 3-2. Relaxation versus Temperaturefor Spring Materials.
150
250
l8
Exposuretemperqture("F)
350
300
400
g
Corbon
steel
Ploinsprings Shot-peenedcrrr
Shot-peened ond Heot sei
l4
Corbon
steel
Ploinsprings --
Chrome
silicon
-
-.Shor-peened
-Shot-peened
'
$
fl
.l
Exposureof 100 hours ot l49oC (300"F)

Stressescolculoted ot room femPeroture
'
r2
E t
f I
o I
.gt
8 r o
o
x
o
Q
d,
8c r o
;rl
g.u I
.9
^
,f
fl
_f
3
a
g
ond Heot set
f
g
Chrome
silicon
m
-
ilr
.9
o
x
oc
lt
200
800
600
400
Initiqlslress(MPo)
120
1000
temperatures for many commonly used materials. It

should be remembered that, if a material is used at its
maximum temperature, a substantial reduction must be
made in applied stress from that used at room temperature.
Corrosion
The effect of a corrosive environment on spring performance is difficult to predict with certainty. General
corrosion, galvanic corrosion, stress corrosion and corrosion fatigue reduce life and load-carrying ability of
springs. The two most common methods employed to
combat effects of corrosion are to specify materials that
are inert to the environment and to use protective coatings. Use of inert materials affords the most reliable
protection against deleterious effects of all types of corrosion; however, this is often costly and sometimesimpractical. Protective coatings are often the most costeffective method to prolong spring life in corrosive environments. In special situations, shot peening can be
used to prevent stresscorrosion and cathodic protection
systems can be used to prevent general corrosion.
ffiessog51ffifu*ffiffi$
r40
160
r80
Exposuretemperoture(t)
Tahle 3 -5. Guidefor SeIecting Minimum Thicknessesfor
Zinc and Cadmium Coatings.
Cedmirrron LuardStscl
Prrrs per QQ'"I4I6 {Z}
'Sl*.i
Zixon tranerd Srnel

Part PerQ8rA325 t$
ilfisirnur$
3
Th*nm
m*n,till.l
S*li,$pqry
Tr*Sr''L
$rn$,,ill&turry!
ffi,:iH$4ry
l*;1ltry&
.*Ilri*h h:ftrr'..fion Thiekamrfiilf,$
Brr :White :n$d,rnar,{lnJ l}pr,' C
r92
0.025
(0.0010)
A
B
0.013
(0.000s0)
A
B
96
0.005
(0.00020)
A
B
r92
c
c
36
36
0.013
(0.00050)
96
0.008
(0.00030) B
96
0.005
(0.00020) B
96
(l) Requirements for zinc coating (electrodeposited).

(2) Requirements for cadmium plating (electrodeposited)'
Finish Type:
A. Without supplementary chromate or phosphatetreatment.
B. With supplementary chromate treatment.
C. With supplementary phosphate treatment.
C
C
C
G
G
G
G
G
G
G
C
G
C
C
C
C
C
C
C
C
C
C
|i
C
C
C
C
C
C
-1
Spring Materials
Coatings may be classified as galvanically sacrificial
or simple barrier coatings. Sacrificial coatings for high
carbon steel substratesinclude zinc, cadmium (and alloys thereof) and, to a lesser degree, aluntinum. Due to
its toxicity, cadmium coating should only be specified
when absolutely necessary.Because sacrificial coatings
are chemically less noble than steel, the substrateis protected in two ways. First, the coating acts as a barrier
between substrate and environment. Second, galvanic
action between coating and substrate cathodically protects the substrate. This characteristic allows sacrificial
coatingsto continue their protective role even after the
coating is scratched, nicked or cracked. The amount of
damagea sacrificial coating can sustain and still protect
the substrateis a function of the size of the damagedarea
and the effrciency of the electrolyte involved. The salt
spray life criteria for three thicknesses of sacrificial coatings are shown in Table 3-5. Use of conversion coatings, such as chromates,lengthensthe time of protection
by protecting sacrificial coatings. SaIt spray (fog) is an
acceleratedtest and results may, or may not, correlate
with corrosive activity in the actual environment. The
test is useful as a control to ensure the coating was
applied properly.
Metallic coatings are normally applied by electroplating. Since most high hardnesssteels are inherently very
susceptibleto hydrogen embrittlement, plating must be
carried out with great care to minimize embrittlement
and subsequentdelayed fracture. A baking operation after plating is also essential.The designershould observe
these points during design and specification:
l. Minimize sharp corners and similar stress-concentration points rn design
2. Keep hardnessas low as possible.
3. Keep operating stressdown, in accordancewith lowered hardness value.
4. Specify plating thickness, depending upon requirements.
Specify that parts be baked after plating.
6 . Consider use of HEPrM strips to monitor the plating
operation.
Residual stress from forming operations must be reduced by stress relief at the highest practical temperature. Otherwise the combined effect of residualtension and hydrogen absorbed during plating can induce
cracking even before plating is completed.
Similar cautions apply if acid cleaning procedures are
contemplated.
Mechanicalplating provides an effectivemeansof zinc

or cadmium protection with minimum hydrogen embrittlement. It is particularly recommended where parts
have high residual stress, have been hardened above
HRC48 and are used with high static loads. The process
can only be applied to parts that do not tangle and have
a clean, fully accessible surface. Hydrogen embrittlement, although unlikely, is still possible if parts are
cleanedby pickling. When appropriate,coatingsof zinc,
tin, cadmium, or an alloy of cadmium can be applied by
mechanical plating processes.
Cadmium, zinc or more commonly alloys of the two
can be applied to steel spring wire during its production,
and under some circumstances this alternative is highly
desirable. It is best suited to small diameter wire and,
in general, for the production of springs not requiring
grinding.
Springs are almost always in contact with other metal
parts. In a corrosiveenvironment, it is important that the
spring material be more noble than components in contact with it. Table 3-6 shows a partial list of alloys in
increasing order of nobility. When any two alloys are
placed in contact in the presence of an electrolyte, the
less noble alloy (higher on the list) will be attacked. The
anack will be significantly more vigorous than that of the
electrolyteactingby itself.
Table 34. Order of Nobility.
G*lvnnic Serics $itb,,'gl$
Magnesium
Zinc
Aluminum
Cadmium
Steel or Iron
Cast Iron
StainlessSteel, series
300 (active)
Hastelloy@C
Nickel (active)
Inconel (active)
Hastelloy B
Brasses,Bronzes
Monel
Nickel (passive)
Inconel (passive)
StainlessSteel, series
300 (passive)
Titanium
te ,gnehgs,,se*ry8t8r.
Least noble (+), Anodic
Most noble (-),
Cathodic
HASTELLOY is a registeredtrademark of Cabot Corporation.
ffi{*, Yffiffi
fFJ','r. iil;-
ffi
Spring Materials
The list of coatingswhich protect the base material by
acting as a barrier to the environment is extensive and
increasesas new finishes and techniquesare developed.
Table 3-7 shows protection available from some of the
common barrier finishes. This information is not for selection purposes;it simply shows the rangeof protection
afforded. In fact, the hours of salt spray protection may
only be valid for the specimenand test conditions employed in this series of tests. The tests were conducted
on springs which had undergone 4 million cycles in a
fatigue test prior to salt spray exposure.
While coatings frequenfly increase in effectiveness as
their thicknessesare incrbased, cautions are in order.
Tendencies to crack increase as coating thickness increases,and the coating increasesthe size of the spring.
For example, coatings increase the solid height and diametral clearancesrequired for compression springs.
Brittle coatings such as epoxy can chip under impact,
leaving unprotectedspots. Tough coatingssuch as vinyl
resist chipping, but bruises, tears or abrasions can expose the base material and trap corrosive agents. This
allows corrosionto continue after exposure,and in these
circumstancescoated springsoccasionallyexhibit shorter lives than uncoated springs.
Frequently oils, waxes or greasesprovide adequate
protection. Effectivenessof these coatings is often dependent on the nature of the surface to be protected. In
general,lustrousor smooth parts will not retain oils, and
waxes, paraffrn-basedoils or greasesare recommended.
Steelscan be phosphatecoatedby a conversionprocess.
Phosphate coatings have a high retention for oils,
greasesor paints. The combination of a phosphate and
oil coating becomesa colTosion inhibitor more effective
than either of the components. A similar effect is obtained by retaining or deliberately forming oxides on
metal surfacesto hold corrosion inhibitors or lubricants.
Oil tempered spring wire is a notable example of this
technique.
Spring Wire
Tensile properties of spring wire vary with size (Figure
3-3). Common spring wires with the highest strength
are ASTM 228 and ASTM 401 materials. ASTM A313
ffiA"o#fil,&F'*ffiK$
Type 302, A232 and A230 materials have slightly lower

tensile strength with surface qualities suitable for fatigue
applications. Hard-drawn (ASTM 227) and oil tempered
(ASTM 229) are also supplied at lower strength levels
and are most suitable for static applications.
Most spring wires can be wrapped on their own diameter (bent around a pin with a diameter equal to the
wire diameter). Exceptions include some copper-based
alloys and large diameter and/or high strength alloys.
Becausestress relieving increasesyield strengthof cold
drawn spring wire, all sharp bends of this grade material
should be made prior to stress relief.
G
G
G
G
G
SaIt Spray Resistance of Common Barrier

Finishes.
G
G
G
G
G
C
C
Tahles3J.
hotective
Materiat
Paints:
Japan
Stsnderd SaIt Spray

Test Resistancc,hours
15-20
Lacquer
31100
Enamel
Paint
5G400
2s-300
Oils, waxes
l-300
Phosphateswith
supplemental
oils, waxes,
etc.
2440
Cadmium, zinc
24-100
Description
Dark colored, usually
dipped, cured by baking.
Usually applied by spraying. Air dried.
Hard finish; applied by
spray, brush or dip; cured
by air or baking.
Lubricating, rustinhibiting, hard drying and
nondrying oils.
Chemical treatment converting steel surface to iron
phosphate crystalline
surface. Affords a bond
for oils and paints.
Electroplated or mechanically
plated.
This information is basedon laboratory-controlled applicationsand test.

The protective material selected,cleanliness of parts, method of application, subsequentoperations and part usage affect performance.The
choice of a spring finish must also consider shipping, assembly,end use
and total cost.
C
-
C
C
C
C
C
C
C
C
C
C
C
C
C
Spring Materials
Tabk 34. PreferuedDiameters for Spring Steel Wire.
ftrril
Hrrtnct
M*trk Sillri {ftm,

Sr*cnd
Third
PrEfrnmr
Prcfcrcns
0.10
0.r 2
EnglicbSirrs {la.)
First
$cmrd
hrcfcrrnc
hefttwmcc
0.r 4
0.004
0.005
0.006
0.008
0 .r 8
0.010
0.22
0.012
0.2E
0.014
0.35
0.016
0.45
0.01E
0.55
0.020
0 . 1I
0.16
0.009
0.20
0.01l
0.25
0.013
0.30
0.015
0.40
0.0r7
0.50
0.019
0.60
0.021
0.65
0.70
0.022
0.024
0.026
0.028
0.80
0.90
1.0
I.l
0.031
0.033
1.4
1.8
2.0
2.1
2 .4
2.5
2.6
2.8
3.0
3.2
3.5
3.8
4.2
4.5
5.0
Snt t
Wir!
,,,,,,il1,,,,,,:
,i,t
1 -Wrmll&wc
0.u7
0.04E
0.051
0.055
0.059
0.063
0.057
0.072
0.076
0.08r
0.085
0.092
0.098
0.r05
0.1r2
6.5
7.0
0.125
1.0
1.3
1.0
1.3
Music
Carbon Valve Spring
2.6
3.1
1.4
1.9
4.0
7.6
3.9
4. 7
8.0
6.7
8. 7
ASTM A,228
ASTM A23O
PhosphorBronze
ASTM
StainlessSteel (Type 631) ASTM A 313(631)
(17-7 PII)
ll
Beryllium Copper
Inconel Alloy X-750
27
44
0.120
7.5
0.130
Izrrn
Patentedand Cold Drawn ASTM A227

ASTM A229
Oil Tempered
ASTM A4OI
Chrome Silicon Valve
StainlessSteel (Type 302) ASTM A3l3 (302)
0.102
5.5
6.0
:::,,.,,::::
!|1}...'It*',...
0.u2
0.045
4.8
ffi,,#.2.mu
ffii;$
0.040
') ',
Table34. Ranking of Relative Costs of CommonSpring

Wires.
0.035
0.038
t.6
ASTM BI97
t7
31
0.135
0.140
E.5
0.148
9.5
0.t62
9.0
0.156
10.0
r1.0
12.0
0.170
0.t77
0.192
13.0
14.0
0.200
16.0
Table3-10. Standard Tolerancesfor Spring Wire.

Dlrrmttr: rnar{ln,}
tohrw:
ulur,{lil;}
0.218
0.225
0.250
0.262
tl&#m:Ost
,d"Mnrrr:
,,il;il,,,{*n.}
0.207
15.0
C
C
C
C
C
C
C
C
C
C
0.030
t.2
Spring Strip
Most flat springsare made from AISI grades 1050,1065,
1074and 1095steel strip. These compositions are listed
in ASTM specifications A.682 and A684.
Tensile strength and formability characteristics are
shown in Figure 3-4. The vertical inclined bands delineate three strength levels as functions of stock thickness
and hardness. Horizontal curves indicate minimum bending radii required for the strength levels they intersect.
Interpolations can be made between any two bands or
lines for intermediate levels. Formability criteria are given for relatively smooth bends made at reasonablebending rates. Operations which apply forming forces other
than smooth bending, or have impact characteristics,
may require larger radii to prevent fracture. Four-slide
part manufacture, progressive die work and secondary
forming are examples of operations that often produce
less than ideal bending.
0.514.71 (0.020-0.028) 0.olo(o.ooo4)

(0.0006)
0.71-2.00(0.0284.078) t 0.015
0.010(0.000+)
0.015(0.0005)
2.00-3.00(0.0784.I tE)
3.00{.00 (0.118J.240)
0.020(0.0008)
0.030(0.0012)
6.0G9.00 (0.24G4.354)
9.50-r6.00(0.37s4.62s)
r 0.050(0.00197)
+ 0.070(0.00276)
0.2E1
0.306
0.312
0.343
0.362
0.375
0.406
0.437
0.459
0.500
C
C
C
C
C
C
C
E
C
C
C
C
C
C
0.050(0.002)
0.070(0.0028)
C
C
C
Most spring wires can be purchased to tighter tolerances. Music wire

and most nonferrous materials are regularly made to closer tolerances.
0.020(0.000E)
(0.0011E)
0.030
ffiAssEslf;g&ffi*ffws
C
I
SpringMaterials
Direction of bending with respect to rolling direction
is an important consideration. Formability of strip is
greater in transverse than in longitudinal directions (Figure 3-5). If a part is designedwith two identical bends
at 90" to each other, it is common practice to orient the
part so that both bends are made at 45oto rolling direction. Dmensionless parameter2rlt, often referred to as
bendfactor, is frequently used as a measureof formabilit1'. Materials with low values are more formable than
materialswith high values. This measureis only a guide
since it does not allow for tooling considerationsand
complex strains associated with forming operations.
Spring steels are nonnally produced to specifiedhardnesslevels which are related to tensile strength (Figure
3{). Composition is not shown in Figure 34 because
the lowest carbon level (AISI 1050)can be used at high
strengthlevels and the highest carbon level (AISI 1095)
canbe temperedto the lowest strengthlevels. In general,
higher carbon levels are used when strength is critical
and lower carbon levels when formabilitv is critical.
Fig. 34. Minimum TransverseBending Radii for Various Tempers and Thicknessesof Tempered
Spring Steel.
E
E
1.0
S o.7s
.c
Hardness levels above HRC 50 result in high strength

but are not generally recommended due to notch sensitivity. Surfaceand edge smoothnessbecomecritical and
plated parts become highly susceptible to static fracture
due to trapped hydrogen.
Parts which cannot be made within formability limits
of pretempered strip are made from annealedstrip and
hardened and tempered after forming. To maintain critical dimensions,it is often necessaryto fixture temper
these parts. Sharp bends are not only diffrcult to fabricate but are also undesirablein service becauseof stress
concentration. The formability limits of annealedspring
steels are presented in Table 3-1 1.
In flat spring designs where the edge of the strip becomes an edge of the part, the type of edgeis important,
particularly for cyclic applications. Common types of
edges available are shown in Figure 3-7. Slit edge (No.
3) and deburred (No. 5) are preferred for blanked parts
and static applications. No. I round edge is recommended for cyclic applicationsto reduce the stressconcentration and eliminate the edge flaws due to slitting.
Configurations shown in Figure 3:7 are approximate,
and it is advisableto use both the numerical designation
and a description when specifying edge condition.
Commercial thickness tolerances for spring steel strip
EIrepresentedin Table 3-12. Many flat springsand spring
washer designscan tolerate this variation. Sincethe load
varies as the cube of the thickness, critical designsmay
require closer tolerances.
Fig. 3-6. TensileStrength versus Hardness of Quenched
and TemperedSpring Steel.
l0 k9. DPH or Vickers (VHN)
350
44
46
48
50
450
52
RockwelI Hordncss(HRC)
Fig. 3-5. orientation of Bend Axis to Rollins Direction

for Transverseand Longitudinal B-ends.
-)
24o E_
al
2zo *,
lndicotesDirectionOf Rolling
g
6
N 1 : 2 r
Ntn
2oo+
Nrffi
,o
g
.o
r8o E
5
---t,.
Tronsverse Bend
Longitudinol Bend
Also known os ocross the groin,

p e r p e n d i c u l o rt o t h e r o l l i n g
direction. Eosy or good woy.
Also known os with the groin,

porollel to rolling direction.
Hord or bod woy.
160
t40
120
28 32 36
40
444648 50 52 54
Rockwell Hordness (HRC)
ffi
Spring Materials
Other Spring Materials
A variety of materials other than carbon steel strip is
used for flat springs(Table 3-13). When high conductivity is required, copper base alloys are usually specified.
Stainlesssteelsare used in applicationsrequiring heat or
corrosion resistance.Typical tensile strength levels'for
these alloys after heat treatment are shown in Table
3-13. Bend factors and tensile elongationsare for alloys
in "as received" condition prior to final heat treatment.
Specifyrng Hardness
Hardness tests are used extensively to inspect strip
and flat springsand it is necessaryto specify the correct
scale. Recommended hardness scales for steels are
presented in Table 3-14. To obtain accurate readings
free from the effect of the anvil, it is important to limit
the thickness of the material for each hardness scale as
shown in Figure 3-8 for hard materials and Figure 3-9
for soft materials.
Fig.3-7. EdgesAvailable on Steel Strip.
No. I Edge
_l
l
:l
SQUARE
Stondord
moximum
corner
rodius:0.08
mm
(0.003')
No.3 Edge
C
C
ROUND
Stondord
C
C
C
NOR'VIALAS SIIT
BLUNT
ROUND
Speciol
'No.5 Edge
OVAL
Speciol
_l
_l
BROKEN
CORNERS
Speciol
No.3DEBURRED
Table3-11. Formability of Annealed Spring Steels.
*Iffi r0s
Ntll
Ttulcknmr{(}
mrn (in,l
Di:recti*n
nf: Ssn6
AI$I T065
Hrlt
Ann*n|ed
rd
{*ta
tpr*iffit
rnnx.)
*nn
tr$st.
t.!l{x"}i
Annwkd
($tardflrd
lsw{st
mnx.)
lmregt ,,
'::il$#$
*IS,*qid tfrs
t:
. :,: ::,:il,{i
Annerlsd Aaar*bd
{sp l.; {ffiid
'lwnst ,
bnffit'.
filflX,l*
I:
.,.,
.i..'t,,.,.
...[-{itt,..'
,,.{W.,,.. t
,,,i{ffi:
,,.i#ffiffi'.
,.l{illffit::,,.
:..,,i,[Ufte$:
',.m&X;},,' : ,.,'1114.1g;)i,:,,,r,s**:l} ,, ,,,.ifm.i|,*
1 . 9m m
(0.076)-over
I
ll
2
4
0
3
2
4
2
4
0
3
3
)
I
4
0 .9 -1 . 89m m
(0.036{.075")
I
2
0
I
I
2
0
I
I
2
0
I
2
3
0
2
0.374.89mm
(0.0154.035)
il
0
I
0
0
0
IU2
0
I
I
IU2
0
I
1
2
0
I
il
0
0
0
0
0
0
0
0
1
I
0
0
I
I
0
U2
0.2-O.36
mm
(0.008{.014j
ll
Formability is determinedby slowly bendinga sampleover l80ountil its

ends are parallel. The measureddistance between the ends is Nt.
For example,if N1 = 4 and t:2, then Nr/1 = 2
*Availableas Barco-Form@
from WallaceBarnesSteelsubsidianvof
Theisof America, Inc.
C
C
Table3-12. TyoicalHigh-CarbonStrip ThicknessTolerances.

Thhkns*: mrn{in,}
'#*tqgry,,Pffi;1iffi. rl;,,
t#idth:
Srrip
t?.7*?6.1
(0.50*?.99)
t
|
rnm.{tor.}
76,3*]S4.fl
f3 jfffl*12.00)
0. l0-0.25 (0.004-0.010)
0.005(0.00020)
0.906(0.00025)
0.25-0.5
I (0.010-0.020)
0.006(0.00025)
0.009(0.0m35)
0.5l-0.76(0.020-0.030)
0.009(0.00035)
0.013(0.00050)
0.7Gt.02(0.030-0.erc)
0.010(0.00040)
0.0r3(0.000s0)
t .02-| .s2(0.040-0.
060)
0.013(0.00050)
0.019(0.00075)
| .52-2.03(0.060-0.080)
0.025(0.00100)
0.038(0.001s0)
2.03-2.s4(0.0E0-0.l 00)
0.038(0.001s0)
0.051(0.00200)
(0.r00-0.125)
2.54-3.r8
0.051(0.00200)
0.063(0.002s0)
Precisionrolled high-carbonsteelstrip is availablecommerciallyat

tolerancesconsiderablvlessthan the valuesstatedabove.
C
C
C
C
C
C
t
C
C
C
C
C
C
C
c
c
C
-
a
a
a
a
a
Spring Materials
DPH( I kg)
DPH( I ks)
500
600
2
P
0.0r0
55 60 65 70
Hordness N.umber
Tendlc@rySr
ntra {rtr rd}
a
a
a
a
a
Monel 400
Monel K500
Inconel 600
a
a
a
a
a
a
a
,til
;E
at
5
E
'7
i
=
o . E
=
.=
:
60
65 70 7s 80
Hordness
Number
85
Tabb 3-13. Typical Properties of Spring Temper Alloy Strip.
r70o(246)
(r89)
1300
!D
r40
35 40 45 50
Steel, spring temper

Stainless301
Stainless302
a
a
a
iS
ldfrtrr,*$l
r
0.020
r20
2
E
a
a
a
a
a
r00
900
800
.r_----t0.040
Fig. 3-9. Minimum Safe Thicknesses

for Hardness Testing Soft Materials.
FA. J-d. Ilinimum Safe Thicknesses

for Hardness Testing Hard Materials.
,,,,,
, :8fsd
,
:, .';F
l'{.1
,}
in
t:l:t ,;,',,,., .,rif.1;tl:':.,,rr,t;,,,;,,;,;:.

:.,fst.m|,.
,,, h,qns., b)
b
:l,:t;i1,,::,
c50
c40
c40
2
8
5
5
3
4
20.7(30)
19.3(28)
19.3(28)
0.30
0.31
0.31
6e0(100)
rz0o(r74)
895
(151)
1040
c34
c30
2
40
2
5
5
2
17.9(26)
17.9(26)
2r.4(3r)
0.32
0.29
0.29
Inconel X-750
Copper-Beryllium
Ni- Span-C
1050(152)
1300(189)
1400(203)
c35
c40
c42
20
2
6
3
5
2
2r.4(3r)
0.29
0.33
BrassCA 260
PhosphorBronze
I1J PH RH95O
l7J PH Condition C
620 (e0)
6e0(100)
1450(210)
1650(239)
B90
890
c4
6
1
1300(189)
c46
2.5
flat
2.5
12.8(18.5)
r8.6(27)
ll (16)
10.3(1s)
20.3(2e.s)
20.3(29.s)
0.33
0.20
0.34
0.34
(l ) Before heat treatment.
Tsblc3-14. Recommended Hardness Scales for Hard

and Soft Spring Alloys.
"tnncrlcd Stcd
and Finnfcrmnr Albys
Thickness:mm {in.}
0.89(0.035)
andover
0.64J.86(0.025-0.034)
0.35{.61(0.01
5-0.024)
(0.0084.014)
0.204.36
Under0.20(0.008)
C
A
30N
r5N
DPH
B
457
307
r5 T
DPH
Assogfrlfi8
&tr*ffiwsffi
Helical
Compression
Springs
G
C
C
C
G
G
C
C
C
I
C
'., C
C
..: C
,,115
,t,,
c
,:;,.;;',
',
::: 5
li:',',C
1
*<,
c
!n,-1
ff
k* Asso8Ftfigi%Fr*'Fs
C
} C
I
I- C
l c
I- C
3 c
C
C
C
C
Helical CompressionSprings
=
-r
-t
-,
-
-l
-,
4t
-,
-,
-,
-,
-,
-.
at
-
a
a
-
a
l|
a
a
a
a
Introduction
Helical compressionspringsare used to resist applied
compressionforces or to store energyin the push mode.
They have the most common spring configuration and are
found in many applications such as auiomotive, aerospace and consumer goods. While the most prevalent
form o-fcompression spring is a straight cylindri-d spring
made from round wire, many other forms are produced.
Conical, barrel, hourglass or cylindrical forms Ere available, with or without variable spacing between coils.
Suchconfigurationsare usedto reducesolid height, buckling and surging, or to produce nonlinear load deflection
characteristics. Energy storage capacity is greater for
round wire compression springs than for rectangular wire
compression springs and can be increased by nesting.
Rectangular wire is sometimesemployed to reduce solid
heightor increasethe spaceeffrciencyof the design.Most
die springs are made from rectangular wire for this reason.The SPEC line of springscontainshundredsof compressionspring designsusing wire sizesfrom 0.15 to 5.26
mm (0.006" 1o 0.207') diameter music or stainlesssteel
wire. Specifying SPEC springssavesdesigntime, reduces
cost for low volume applications and offers improved
delivery.
Helical Compression Spring Terminologr
Special terminology has evolved in the spring industry
to describefeaturesof helicalcompressionsprings.These
terms are defined and the relationship between terms is
reviewedin Figure 5-1. Communicationbetweendesigner and springmaker is improved if these common terms
are used.
Spring Diameter
Outside diameter, inside diameter and mean diameter
are all used to describe helical compressionspring dimensions. Mean diameter is equal to the sum of O.p.
and I.D. divided by two, and is employed in spring design calculations for stress and deflection. The O.D. is
specified for springs that operate in a cavity, while the
I.D. is specified for springsthat operateover a rod, seat
or shaft. Minimum diametral clearance between the
spring and cavity or rod is:
0.05D - when D. is greater than 13 mm (0.512')
0.10D- when D. is less than 13 mm (0.512,)
D. is the diameter of the rod or cavity.
Diameter increases when a spring is compressed.

Although the increase in diameter is usually small,
it must be considered when minimum clearances are
established. The increase in diameter is a function of
initial spring pitch and can be estimated from the
following equation where p : pitch.
O.D.4*1;6 =
P2-d:
*2
tl
+ d
(s-t)
If the spring ends are allowed to unwind, the O.D. at

solid may be greater than calculated by this equation.
Long springs buckle (see Figure 54, page 35) and may
require lateral support and larger diametral clearances.
Spring Index
Spring index is the ratio of mean diameter to wire
diameter or radial dimension of the cross section (Figure
5-15, page 40). The preferred index range is 4 to lZ.
Springs with high indexes tangle and may require individual packagrng, especially if the ends are noi squared.
Springs with indexes lower than 4 are diffrcult to form.
Free Length
Free length is overall spring length in the free or unloaded position (Figure5-1). If loads are not critical, free
length should be specified. When definite loads are required, free length should be a referencedimension that
can be varied to meet load requirements. Pitch is the
distance between centersof adjacent coils and is related
to free length and number of coils.
Type of Ends
Types of ends available are: plain ends, plain ends ground, squaredendsand squaredends-ground (Figure
5-2).To improve squarenessand reduce buckling during
operation, a bearing surface of at least 270" is required.
Sguared and ground springs are normally supplied with
a bearing surface of 270 to 330". Additional grinding results in thin sections."squared ends only" are preferred
on springs with small wire diameters (less than 0.5 mm
or 0.020), a large index (gtreaterthan 12) or low spring
rates. Squared ends cost less to manufacture than
squared and ground ends.
fi
Fig. 5-1. Dimensional Terminology for Helical Com-
a
a
a
a
-T
+r Il f",.
Porolleli s m ( e p )
an.h1
l
|
d L r
a
I
I
nssgpffi,&gmp$ffi
Number of Coils
Total number of coils should be specified as a reference number. For springs with squaredends, the total
number of coils minus two is the number of active coils.
There is some activity in end coils, but during deflection
someactive material comesin contact with the end coils
and becomes inactive. Experience indicates that this
equation is a good approximation. The number of active
coils in springs with plain ends is greater than those with
squaredends and dependsupon the seatingmethod employed. Some useful guidelinesfor estimatingthe number of active coils are presented in Table 5-1.
Solid Height
Solid height is the length of a spring with all coils
closed. For ground springs, solid height is the number
of coils multiplied by wire diameter. For unground
springs, solid height is the number of coils plus one,
Fig. 5-2. Typesof Endsfor Helical CompressionSprings.
ffi@ffiffi@
Squoredond Ground Ends
Coiled left-hond
PloinEnds
Coiled Right-hond
PloinEndsGround
Coiled Lefi-hond
Squoredor ClosedEnds
Not Ground, Coiled Righrhond
Table 5-1. Guidelinesfor Dimensional Characteristicsof

Compression Springs.
Ttxtqd':SS$
,Di
'
:,Oif,;1
Ollftior
,, . , f i , . ,
,,,fhfu,,:,
$gurd Ouly
{tr{t:,,$K[{l {G
+ l)d
r.sc*d
',',,,,G
Solid Height
(Lr)
(Nr + l)d
Ntd
(Nt
Pitch
(p)
Lr-d
N.
Lr
Lr-3d
Lr-2d
Nt
Na
Na
Active Coils
(NJ
Lr-d
p
L r r
p
Lr-3d
p
Lr-2d
p
Total Coils
(Nt)
Na
Na+l
N"*2
N"+2
Free Length
(Lr)
pNt+d
PNt
,"
Ntd*
pNa+3d
*For small index springslower solid heights are possible.
pN6+2d
multiplied by wire diameter (Table 5-1). If critical, solid

height should be specified as a maximum dimension.
After allowances are made for plating or other coatings,
it is good practice to add one-half of the wire diameter
to determine maximum solid height. With larger wire
sizes and fewer coils, this allowancecan be decreased.
Solid height is often measuredby applying a force equal
to 110 to l5Vo of the calculatedload at solid. If solid
heigtrt is not critical, this dimensionshould be omitted.
Direction of Coiling
A helical compression spring can be either left or righthand coiled. If the index finger of the right hand can be
bent to simulate direction of coil; so that the fingernail
and coil tip are approximately at the same angular position, the spring is right-hand wound (Figure 5-3). If the
index finger of the left hand simulatesthe coil direction,
the spring is left-hand wound. If direction of coiling is
not specified, springs may be coiled in either direction.
Nested springs with small diametral clearancesshould be
coiled in opposite directions.
Squarenessand Parallelism
Squarenessof helical compressionspringscan be measured by standing a sample spring on end on a horizontal
flat plate and bringing the spring against a straightedge
at right angles to the plate. The spring is rotated to produce a maximum out-of-square dimension e, (Figure
5-l). Normally squared and ground springs are square
within 3owhen measuredin the free position. Squareness
should be checked at both ends. Specifying squareness
or parallelism in the free position does not assuresquareness or parallelism under load.
Parallelism (Figure 5-1) refers to the relationship of
the ground ends, and is determined by placing a spring
on a flat plate and measuring the maximum difference
in free length around the spring circumference ep.
Hysteresis
Hysteresis is the loss of mechanicalenergy under cyclic loading and unloading of a spring. It results from
frictional losses in the spring support system due to a
tendency of the ends to rotate as the spring is compressed. Hysteresis for compressionsprings is low and
Fig.5-3. Direction of Coiling Helical Compression
Springs.
Coiled
Right-hond
e
e
e
e
e
e
a
a
I
a
a
a
a
e
e
a
I
a
I
I
I
I
I
(:-f
---
I
C
-*--)
( ---fL---
I
I
I
I
-l-
Coiled
Left-hond
(
(
ts-
WA"SS;$ &ffi&ffiffi$
F
P
ts
F
F
t
a
a
a
a
a
a
a
a
a
a
a
I
a
a
a
a
a
HelicalCompression
Springs
the contribution due to internal friction in the spring
material itself is generally negligible.
Spring Rate
Spring rate for helical compression springsis defined
as the changein load per unit deflection and is expressed
as shown:
,r
(s1)
gD3N.
This equation is valid when the pitch angle is less than

l5o or deflection per turn is less thanD/4. For large
deflections per turn, a deflection correction factor (Reference 3, page 102) should be employed.
The load deflection curve for helical compression
springsis essentially a straight line up to the elastic limit,
provided that the amount of active material is constant.
The initial spring rate and the rate as the spring approachessolid often deviate from the averagecalculated
rate. When it is necessaryto specify a rate, it should be
specified between two test heights which lie within 15
ro 85Voof the full deflection range (Figure 54).
When compressionsprings are used in parallel, the
composite rate is the sum of the rates for individual
springs. For compressionsprings in series, the rate is
calculatedfrom:
k-
kr
kz
- J - - I -
(s-3)
kr"'kn
This relationship is often used to calculate the rate for

springs with variable diameters. The technique involves
dividing the spring into many small increments and calculating the rate for each increment. The rate for the
whole spring is computed from the rate of the increments
according to the equation above.
Fig. 54.
Typical Load Deflection Curve for Helical

Compression Springs.
t:H*-
(s4)
Bending stressesEuepresent but can be ignored except

when the pitch angle is greater than 15oand deflection
of each coil greater thanD/4 (Reference3, page 102).
Under elastic conditions, torsional stress is not uniform
around the wire cross section due to coil curvature and
a direct shear load. Murimum stress occurs at the inner
surfaces of the spring and is computed using a stress
correction factor. The most widely used stress correction factor Kwr is attributed to Wahl. It is shown below
and in Figure 5-5.
4C - I
0.615
r(wr:4c-4 c
(s-s)
In some circumstancesafter yielding occurs, resultant

stressesare distributed more uniformly around the cross
section. Then, a stress correction factor Ks,2 which accounts only for the direct shear component is used.
Kwz:1+
0.5
C
(s4)
In other circumstances,such as static loading at elevated

temperatures, stress distribution tends to become uniform around the cross section and can best be estimated
by using no correction factor. Use of different stress
correction factors can lead to confirsion. In published
data, it is essentialto know which stress correction factors were used. (The stress correction factor used by a
designer must be the same as that used to develop the
data.) Methods to calculate stress for different applications and the use of stress correction factors will be
Fig. 5-5. Wahl SrressCorrection Factorsfor Round Wtre
Helical Compressionand Extension Springs.
2.2
I
4C-1
.,
Awr=4c-4-
0.615
c
2.0
For 2e/o set point
or fotigue
n q
Kw2=l*t
J r.8
For springs with

set removed
a
a
: - : -
Stress
Wire in a helical compressionspring is stressedin
torsion.Torsionalstressis expressedas:
;
Y
o
't.6
o
I
{*,*'
t r.4
3
a
a
a
a
a
I
6
C=D/d
Assosl$lfi8&ffi*ffiffiffiw
discussedin the following paragraphson choice of operating stresses.
loads
When deflection is known, loads are determined by
multiplying deflection by the spring rate (Equation 5-Z).
When the stress is known or assumed,loads are determined from Equation 5-4. The procedure used to determine loads of variable rate springs is complex. In this
case, the load deflection curve is approximated by a
series of short chords. The spring rate is calculated for
each chord and multiplied by deflection to obtain the
load. The load is then added to that calculated for the
next chord. The processis repeateduntil load has been
calculated for the desired value of deflection (Figure
s4).
Loads should be specifiedat a test height. Becausethe

load deflection curve is often not linear at very low loads
or at loads near solid, loads should be specified at test
heights between 15 and 85% of the full deflection range
(Figure 54).
Loads are classified as static, cyclic or dynamic. In
static loading applications, the spring is expected to
operate between specified loads only a few times. Frequently, springs in static applications remain loaded for
long periods of time. In typical cyclic applications,
springs are required to cycle between specified loads
from 10,000to more than a billion cycles. During dynamic loading, the rate of load application is high and causes
a surge wave in the spring which will induce stressesthat
exceed the value calculated from Equation 5-4.
Buckling of Compression Springs
Compression springs that have lengths greater than
four times the spring diameter can buckle. If properly
guided, either in a tube or over a rod, buckling can be
minimized. However, friction between the spring and
tube or rod will affect the loads, especially when the
aspect ratio (I4lD) is high.
Critical buckling conditions are shown in Figure 5-7

for axially loaded springs with squared and ground ends.
Curve A is for springs with one end on a flat plate and
the other end free to tip (Figure 5-8). It indicates that
buckling will occur when the spring design is above and
to the right of the curve. A tendency for buckling is
considerablyless for springs compressedbetweenparallel plates as shown in curve B. For applicationsrequiring
springs with a high aspect ratio and large deflections,
several springs can be used in series in a tube or over
a rod, with guides between the springs to prevent binding.
Choice of Operating Stress - Static Conditions
For static applications, the yield strength or stress relaxation resistance of the material limits the loadcarrying ability of a spring. The spring is required to
operate for a limited number of cycles, and the velocity
of the end coils is low to preclude high stressesdue to
surgtng or impact conditions. Maximum allowable torsional stressfor helical compression springs usedin static applications is presented in Table 5-2 as a percentage
of the tensile strength for common spring materials. For
springs that do not contain beneficial residual stresses
induced by set removal, maximum allowable torsional
stress values are from 35 to SVo of the tensile strength.
To calculate the stressbefore set removal, it is necessary
to use the Ks,1 correction factor. If the calculated stress
at solid is greater than the indicated percentageof tensile
strength, the spring will take a permanent set when deflected to solid. Amount of set is a function of the
amount that calculated stress at solid exceeds the indicated percent of tensile strength.
Fig. 5-7. Critical Buckling Condition Curves.
Fig. 54. Load Deflection Curve for a Variable Rate

Spring.
CD
0,
o
.9
u
q,
o
o
ii
o
d,
f1
t2
f3
Defleaion --1.
f4
ps = krfr + kz(fr - fr)...ks(fs- fr)
C
C
C
C
C
C
C
C
C
C
C
5
C
C
5
C
C
C
C
C
C
C
C
C
C
C
5
C
/
?
E
'J
h
E\
-,
-
To increase the load-carrying ability of springs in static applications,it is common practice to make the spring
longer than its required free length and to compressthe
spring to solid. This causesthe spring to set to the desiredfinal length and inducesfavorable residualstresses.
This processis called removing set or presettingand can
be conducted at either room or elevated temperatures.
The loss of deflection from the free position to solid by
cold set removal shouldbe at least lVo.If the set is less,
it is diffrcult to control the spring's free length. Ratios
of stress greater than 1.3 lead to distortion and do not
appreciably increase the load-carrying ability. This is illustrated schematicallyin Figure 5-9.
Allowable torsion stressesin springswith set removed
(Table 5-2) are significantly hiefier than for springs that
have not had set removed. It is important to note that
because yielding has occurred during presetting, the
stressis relatively unifonn around the cross section and
it is calculatedusing the Kwz stresscorrection factor. Set
removal is an added springmaking operation which increasesthe manufacturing cost but gfeatly increasesthe
energy storage capacity of the spring. Set removal is
common for critical springs made from premium materials. In some instances, springs have the set removed
during an assembly operation.
Fig.54.
End Conditions Used to Determine Critical

Buckling.
B
End FixedAgoinstTipping
End Free to Tip
P++
ffi
m
ffi
ry
--'/A
rA_--
=
vokelaxation
x too
(sJ)
P" is load at test height before testing.

Pr is load at test height after testing.
Typical stress relaxation data (Figure 5-10) indicate that
at high stresses, some spring materials such as music
wire exhibit appreciable stress relaxation after only 100
hours at temperatures as low as 100"C(zn"q. These
data are only representative of the conditions indicated.
Stress relaxation is affected by material, spring processingvariables, time, temperatureand stress. Associated Spring engineers should be contacted for critical
applications involving stress rela,ration resistance.
When set is removed at an elevated temperature, the
process is called heat setting. It significantly improves
the stress relaxation resistance of springs (Figure 3-2,
page 16) at moderate temperatures and is frequently a
more cost-effective method for achieving low levels of
stress relaxation than specifying a more costly spring
material.
- V A
H
\\ \\\\\\ \\\
Fig. 5-9. Spring l-oad-Carrying Ability versusAmount of
Fixed End
Table 5-2. Maximum Allowable Torsional Stressesfor

HeIicaI CompressionSprings in Static Applications.
If the calculatedstressusing the Ks,2stress correction

factor exceedsthe percentageof tensile stength indicated in Table 5-2, the spring cannot be made. In this case,
it is necessary to either lower the stress by altering
spring design or selecting a higher strength material.
In some applications, maximum operating stress is
limited by material stress relaxation resistance and
amount of load loss that the design can tolerate. When
load is constant, these designs are limited by material
creep resistance. When the spring is compressed at a
fixed test height, stress relaxation resistance of the material is limiting. Designs limited by stress relaxation resistanceaxemore common than designslimited by creep
resistance.It is suggestedthat creep-limited designsbe
reviewed by Associated Spring engineers.
Stress relaxation is defined as percent loadlossaccording to the following relationship:
Bendingor buckling stressesnot included.
E
o
E s
o .D
c
Maximun EaoI Tcmile

Mitsrirk
Patented and cold
drawn carbon steel
Hardenedand tempered
carbon and low allov
steel
Austenitic stainless
steels
\onferrous allovs
::..,B rc.r5tl
,{Ktvi}
r $ct
Erurgrcd fi[#r]
60-70
65-75
o
3n tn
o o
0.6
StressesAre
Colculoted
At Solid.
tt
g
o
o
0.4
o
.t1
ol
0.2
d,
55-65
1. 0
55-65
r.l
r.2
1.3 1.4
Stress Before Set Removol
"t:ffi:i,
1.5
c
Assog&ifi8
f&ffi#ffiffis
ffi
il
Choice of Operating Stress - Cyclic Applications
In cyclic applications, the load-carrying ability of a
spring is limited by material fatigue strength. Velocity of
end coils is low compared to the natural frequency. To
selectthe optimum stresslevel, it is necessaryto balance
spring cost versus reliability. Reducingoperating stresses increasesspring reliability as well as cost. A complete
knowledge of operating environment, expected life,
stressrange, frequency of operation, speedof operation
and permissible levels of stress relaxation are required
in order to make the best choice between cost and reliability.
Becausemaximum stress is at the wire surface, any
surface defects such as pits or seams severely reduce
fatigue life. Shot peening improves fatigue life and minimizes the harmful effect of surface defects, but it does
not totally remove them.
Maximum allowable design stressesfor fatigue applications should be calculated using the Kwr correction
factor and are shown for common spring materials in
Table 5-3. These values are for a stress ratio of 0 in an
ambient environment with no surging. Note that shot
peeningincreasesthe fatigue strength by as mudr asZVo
at lives of 10 million cycles.
Values in Table 5-3 are guidelines for designers and
shouldonly be used in the absenceof specific data. Most
springsdesignedto recommendedstress levels will exceedthe indicated lives; however, in the absenceof detailed information on material, manufacturing method
and operating conditions, it is not possible to quantify
the reliability level.
Fatigue Life Estimation Example
Fatigue life at other stress ratios can be determined
from Table 5-3 according to the procedures outlined in
Section 4. A short example illustrates the procedure:
Estimate the fatigue life of a not-shot-peened helical
compression spring loaded sinusoidally at a rate of one
cycle per second.The spring is flooded with oil and operates at a maximum temperature of 40'C (104"F). The material is ASTM A228 wire and ends are squared and
ground. The design is given here:
d = 1.00mm (0.039')
C = 8
I+ = 20.5 mm (ref) (0.807')
Lr : 17.5mm (0.689?
Lz: l0 mm (0.394')
L,:8mm(0.315')
Nt:8
Stressesare calculated using Equation 5-4 and are:
s=#r*,
Tensile strength of the wire is 2180MPa (Figure 3-3, page

19). The stress at solid is 44Voof the tensile strength.
Referring to Table 5-2, the maximum stress allowable
before sit removal for ASTM A228 is 45% of tensile
strength. Therefore, the spring can be made and does not
require set removal.
To estimate the fatigue life, it is necessary to:
l. Plot an S-N curve on a modified Goodman diagram
(Figure 5-11) using the data from Table 5-3 for notshol-peened springs and a tensile strength of 2180
MPa.
2. Plot point A on the 45" line at 67Voof the tensile
strength.
d
o=3'2Nimm
.
t= GEtr
Loads are calculated from the deflections and found to

be:
P , : ( 2 0 . 5- 1 7 . 5 )x 3 . 2 : 9 . 6 N
P 2: ( 2 0 . 5- 1 0 . 0 )x 3 . 2 : 3 3 . 6N
P , = ( 2 0 . 5- 8 ) x 3 . 2 : 4 0 N
C
C
C
C
C
3. Plot the stressrange coordinates, point B.
4. Estimate the life by drawing a line through AB. At the

intersection of this line with the vertical axis, point C,
' draw a horizontal line to intersect a S-N curve. The
point of, interse'Jon, D, is the estimated life of
2,500,000cycles.
C
C
Dynamic
-When Loading - Impact
a spring is loaded or unloaded, a surge wave is
established which transmits torsional stress from the
point of loading along the spring length to the point of
iestraint. The surg wave travels at a velocity approxi-mately I / l0 of a normal torsional stress wave. Velocity of
the torsional stresswave (V1) is given by:
i:
I=
C
C
s
C
in./sec 64)
Vr = 10.1./9
m/sec (or) Vr : ./9
- - Y
p
Y p
Velocity of the surge wave V. varies with material and

spring design, but is usually in the range of 50 to 500
m/sec. The surge wave limits the rate at which a spring
can absorb or release energy by limiting impact velocity
V. Impact velocity is the spring velocity parallel to the
Tabte 5-3. Maximum Allowable Tbrsional Stesses for

Round Wire Helical CompressionSprings in
Cyclic Applications.
Pcrccat af Ttuih
Spring rate is determined from equation:

,
Sr : 232 MPa
Sz : 810MPa
S, : 955 MPa
C
C
C
C
C
f*tiglrr
Llfr {*yrhr}
*$rll{ A?p8',*s
$teinllrs $e?clrtrd
l{mfrrrous
tlot Shot
F*ilrrd
I$5
t06
r$?
36
''J
3{1
'fu.',
f*md
4?
39
3S
S&eagth
A$TM AEN d
Al3Z
shd.
llld;$hot-,
f,,e,nd
PoGnGd
at
*0
t?,
49
4?
4S
This information is basedon the following conditions:no surging,room

temperatureand noncorrosiveenvironment-
s
s
=g
ratioin fatigue= Pi,lil+
Stress
s maxlmum
wnsso5ff1ffi
,'ffi,m*ffirys
7
-
Helical Compression Springs

Fig. 5-10. Spring RelaxationDataforVarious
Materials. Springswerepresetat roomtemperature
andtestedlM hourso
indicatedtemperatures.
Theinitial stressis Kwtcorrected.
StoinlessSreel302 Wire, ASTMA313, 1.57 mm (.062")
MusicWire,ASTM4228, 1.57 mm (0.062")
140
rl40
r30
r30
e
a
r20
? 8oo
3
=
CL
o-
! 7oo
ar,
l-roo3
o
v,
600
o
a.n
o
tn
l-80 s
s00
400
Reloxolion, Lmd Loss (o/o)
Reloxolion,
Wire,ASTMA401, 1.57 ro 3.76 mm (0.062,'toO.l4S").

Chrome-Silicon
r000
Lood Loss (o/o)
ChromiumVonodiumWire, ASTMA232, 1.57 io 3.76 mm (0.062"to 0. 140")
140
l-140
I
r30
*9
i-r20
llo
i.
a
a
roo3
o
rtl
9 0 ;
E
a
a
a
a
a
a
i-rooI
rt
60
10
Reloxotion,LoodLoss("/o)
CorbonVolve SpringWire, ASTM A230, l. 57 to 3.76 mm (0.062"ro 0.148,,)
InconelAlloy X 750 Wire, 1.93mm (0.075')
r20
oi-----<,-
t,
e/
ilo?
'100
=
tt
v,
70
60
50
r5
g*" 1,.j
70
rD
8 0 =
tt
a
a
a
ll0
0
4
al
Reloxolion, Lood Loss (o/o)
C
Helical ComPressionSPrings
spring axis and is a function of stress and material constants shown as:
IT
(or)V = tV#
v = 10.tt{ft m/sec
in./sec(s4)
This is a surprising result becauseimpact velocity and

stress are independentof the spring configuration. For
steels, impact velocitY reduces to:
phehigh velocities are frequently subject to resonance

nomena
-'Wh;; (page 39).
itreiatio'of the weight to be accelerated to the
weijtrt of the spring is lesslhan 1, surge-wave theory
(Figure 5-12).
;";;;"iy
frea-icls-design performance
an energy
velocities,
Ai frigtt iveight ratios and lower
projected
weight
a
of
velocity
is ujed to predict
t"d;
when
spring
the
of
deflection
o,
"ia
th;- rprng
il;are redeflection
and
velocity
."rr.
u
uv
i.pu.t"o
lated as:
C
C
-
C
6-tot
(or) v:*in./sec
( s- 11) C
V:*m/sec
For horizontal loading:
C
If a spring is compressedto a given stress level and
v,ffii"'
(or)
r:
velocity
3l.ovfY.spRqg
r:
maximum
releasediistantanebusly, the
C
as tti stressdivided by 35.5. Similarly, if
il;6;;rseo
instantaneous
velocity,
known
a
at
loaded
is
u tpting
(s-12)
C
For vertical loading:
stresscin be calculated.At very high loading velocities,
instantaneous stress will exceed the stress calculated
. C
- m + w m'
- / w+ w mm (or) f : v{*[
from the conventional static formula (Equation 5-4) and
31.6v#i
imr:
for
equatigns
Thesg
1performance.
design
*ifi fitnit
u
pu.t u.focity are bttt' concerned with the_primary surge
C
*uu". Frequently, this wave will reflect from the other
w/g is the mass that is being accelerated or decelerated
at
end of the-spring, in.t."ring stress. Springs loaded
"nO-V is the axial velocity of the spring'
5
C
Fatigue Ltfe'
Fig. 5-11. Modified Goodman Diagram for Estimating
C
Minimum Stress(103Psi)
c
c
:=
4
o
A
o
ttl
E
t
o
tn
E
.E
x
E
x
o
c
c
C
C
(
(
StressCycles
(
I
These equations assumethat the spring is massless
and should only be used when the spring mass is less
then 114 of the mass to be acceierated.
When the ratio of spring load to weight is less than
4. the energy required to acceleratethe spring itself becomes appreciable. By assuming that all mass of the
springis concentratedat the moving end, Equations5-10
and l-l I can be corrected by substituting (W + W,/3)
for W where W, is the spring weight.
Qnamic Loading - Resonance
Resonanceoccurs in a spring when the frequency of
the cyclic loading is near natural spring frequency or a
multiple of it. Resonancecan increase individual coil
deflectionand stresslevels well above amountspredicted by static or equilibrium analysis. Resonancecan also
causespring bounce, which results in loads considerably
lower than calculated at the minimum spring deflection.
To avoid resonance,natural spring frequency should be
at least i3 times the operating frequency.
The natural frequency of a compressionspring ls lnversely proportional to the time required for a surge
wave to traverse the spring. For a compression spring
without damping and with both ends fixed:
(s-r3)
" : %#,ry'
rorsteel
n: $fi!}
n = gD2N" ,./qg: for steelrr:

\ p
metric
l4oood F
English
ffiF
n is in hertz.
If a spring cannot be designedso the natural frequency
is more than 13 times operating frequency, or if the
springis to serve as a vibration damping device, it must
utilize one of several methods of energy absorption.
Generally,these are friction devices in which the spring
rubs against another element such as an internal damper coil, arbor, housing or another portion of the spring.
Variablepitch springs and springs in combination aie
also occasionally used to avoid or minimize resonant
frequencyeffects.
Fig. 5-12. Velocity of an Object Propelled by a Compression Spring.
For a vibration isolation system, the essential characteristic is that the natural frequency of the springmass system be as far as possible from the disturbing
frequency. Filtering of disturbing forces may be calculated as:
% offorce transmitted:
x 100 6-14)
+,
(na/n)'- I
where no is the frequency of the disturbingforce and n

the natural frequency of the spring-masssystem (Figure
5-13).
If rq/n is lessthan l, the denominatorin Equation 5-14
should be changed to I - (no/n)2. Note that the frequency n in this equation is the frequencyof the springmass system and not the natural spring frequency. In
fact, the most commonly used equation neglecti the
splng weight and is only deflection dependenr.The general equation is:
n:
15'8
.E metric(or) n: j,p
7r YP
ensrirn6-ts)
Special Springs
Previously in this section, design considerationsfor
round wire helical compression springs of uniform diameter were discussed.These design techniquesare modified below and applied to many special spring configurations. Special springs are chosen to fulfill a unique set
of design criteria. Springs from rectangular wire and
stranded wire as well as variable diameter springs with
conical, hourglassand barrel shapes,zlrdiscussedbelow. Helpful guidelines for nested springs are also reviewed.
RectangUlar Wire
In applications where space is limited and particularly
where solid height is restricted, springs designedfrom
rectangularor keystonedwire are often selected.Associated Spring manufactures hundreds of rectangular wire
spring designs.These springs are commonly referred to
as die springs and are available for immediate delivery.
Fig. 5-13. Transmissibilityof Spring Mounting.
1.4
sts
1.0
For moss rotios of :

Over 4 - use cose (i )
t
l -4 - use cose (2)
'l
|
| Under - use cose (3)
t
l
t
d Moss Theory (2)
|
I
I
\\
Concenirole
9 1
5 r
'Er
ot
J--
\o
F
0.6
Spring Theory(l )
\Mortt"ss
fl!\'\ n /
x l0O = 7o of Force
tronsmitted
|
<
Vo = Velocity
0.4
Vm = Moximum Velocity
3.0
2
3.5
Moss Roiio PiW,
Asso(;iated,r.*
y!'n,T3
Sprinq r'l.##*trile#$
Springs made from rectangular wire, with the width of
the rectangle perpendicular to the spring axis, store more
energy in a smaller space than equivalent round wire
springs. Even though stressdistribution around the rectangular cross section is not as uniform as the round wire
section, the energy storage capacity is higher because
more material can be incorporated into the allocated
space.Rectangularwire is more costly than round wire,
but less costly than keystoned wire. Keystoned wire is
processed specially so that deformation during spring
winding or coiling causes the cross section to become
approximately rectangular. Distortion of the cross section can be estimatedfrom the equationpresentedin Figure 5-14. However, distortion dependsupon the manufacturing technique employed and this equation is at best
an approximation. Axial dimensional change of the wire
must always be consideredwhen calculating solid heights
of rectangularsprings.
The rate for a compression spring made from rectangular wire is expressedas follows:
k: p/r=ffi*,
(s-r7)
Since the wire is loaded in torsion, the rate is the same

whether the wire is wound on flat or on edge (Figure
5-15). Values for the constant K2 are shown in Figure
5-16. Stressis expressedas:
s--*ftr,or#*.
Stranded Wire Springs

Long springs with many coils, when subjected to higb
rates of load application Etsin automatic weapons, encounter shock wave motion and can literally be torn
apart. Stranded-wire springs are often the most successful solution to such problems because of the frictional
resistancebetween the strands.
To function properly, the helix of a spring must be
opposite in direction to the helix of the strands, so that
the strandsbind together when the spring is compressed.
The stranded-wire spring may be wound with 2, 3 or
more strands. Springs with four or more strands are
made with a center wire core to assure necessary stability. Ends should be soldered,brazed or welded to prevent unraveling.
Recognizingthat a stranded-wire spring can be considered as single-wire springs aranged in parallel, spring
rate is derived on the basis given by:
KnGda
*:
8DN
where K : correction factor and n = number of strands.

For a three-strand spring, K = 1.05.
Fig. 5-15. RectangularWire CompressionSpring Wound
on Flat or Edge.
till.
-l
- T r nI l.-
(s-r8)
| rt
t tF
l*-u-
Dt2-*, S p r i n g W o u n d " = 9
-
On Edge
and d is equal the wire diameter for the equivalent round

wire spring. A typical value for a width to thicknessratio of 2 may be assumedin the initial design calculations.
Before
Coiling
l-l
il
Spring
Axis
Rectongulor
K2
I-]
l
/
/
b1
After
Coiling
/r,
l-l
l
l-l
/
/
/
l
I
0
0.r0
t,=t(C+'5)
ffiAssogFifig
/#i, R**Fs
0.20
0.40
K' ondK,
0.50
G
G
G
G
l0
C
G
Fig. 5-14. Wire Cross Section Before and After Coiling

Keyslone
C
G
G
G
G
C
C
Fig.5-16. Constantsfor Rectangular Wire in Torsion.
,- _ L
l+b/t
_lSpringWound
c- = 9
?t t l ; <
t 1l - D h a cI i . i i l , ' - - *
l l l
Values for Kr are shown in Figure 5-16, while values

for the stress correction factor for springs wound on the
flat (Kr) and springs wound on edge (K.) are shown in
Figure 5-17 and 5-18. When rectangular wire is produced
by rolling round wire or if the cross section of the wire
deviates significantly from a rectangle, additional correction factors are required. Whenever a round wire cannot
be usedbecausethe solid height exceedsspecifications,it
is possibleto try a rectangularwire coiled on edgewhere:
0.60
0.70
C
C
C
C
C
C
C
C
C
C
C
C
C
I
7
=,
An approximation for torsional stress in each wire of
the strand is given by Equation 5-4:
8PD--
S =;;l(qr1
=l
=r
?
?
a
a
a
-
Maximum allowable stress after set removal should

not exceed 55 to 60% of the material tensile strength.
Wire diameter (d,) for a single strand in a stranded wire
spring is less than the wire diameter for a monolithic
spring with the same mean diameter and rate.
lr,lumber of Strands
Wire Size
2
3
d, greaterthan 0.79 d
d. greaterthan 0.69 d
d, greaterthan 0.63 d
4
Stress in the stranded-wire spring is also less than the
stress in an equivalent monolithic spring.
Fig, 5-17. Stress Correction Factors for Rectangular
Wire Compression SpringsWound on FIat.
r.5
Lr:N".,ffi
\
:
v
a
a
a
3
I
1.2
C=8
C- l0
C- l2
't.5
2.0
a
a
2.5
3.0
3.5
4.0
Rotio b/t
Fig. 5-18. ,StressCorrection Factors for Rectangular

Wire Compression Springs Wound on Edge.
Fig. 5-20. Typical Load Deflection Curve for Variable

Diameter Springs (Solid Line).
1.5
and Hourglass
.9
r.0
a
a
Fig.5-19. Typical Conical, Barrel

SpringsRespectively.
-c=,1
1.3
ID
(5-re)
where u : the O.D. large end minus the O.D. small end
divided by 2N".
1.4
lD
E)
+2d
=)
a
a
Yariable Diameter Springs

Conical, hourglass and barrel-shaped springs (Figure
5-19) are used in applications requiring a low solid
height, increasedlateral stability or resistance to surgxng.
Conical springs can be designed so that each coil nests
wholly or partly into an adjacent coil. Solid height can
be as low as one wire diameter. Rate for conical springs
usually increases with deflection (Figure 5-20) because
the numberof active coils decreasesprogressivelyas the
spring approaches solid. By varying the pitch, conical
springs can be designedto have a uniform rate. Rate for
conical springsis calculated,as indicated previously, by
consideringthe spring as many springs in series. Rate for
each turn or fraction of a turn is calculated using Equation 5-2. Rate for a complete spring is then determined,
remembering that the spring rate follows the series relationship given previously in Equation 5-3.
To calculate the highest stress at a given load, the
mean diameter of the largest active coil at load is used.
Solid height of a uniformly tapered, but not telescoping,
spring with squared and ground ends made from round
wire can be estimated from:
\
:
g
Y
1.4
\
\
1 . 3\
\
.9
g
I
I
C=3
r.2
C=4
\
\
1.0
-14
C=6
C=10-
-:y
'r.5
C=12
2.0
2.5
3.0
Rotio b/t
Assogslfig
lhff*ffiffisffi
il
Designing a variable diameter spring so that adjacent
coils rub against one another during deflection increases
resistanceto resonancephenomenabut may also shorten
spring life due to wear.
Barrel and hourglass springs are calculated as two
conical springs in series.
Variable Pitch
Variable pitch springs (Figure 5-21) are used to
achieve a variable rate similar to that shown in Figure
5-20 or in dynamic applications where the cyclic rate of
load application is near the natural spring frequency. As
turns of lesser pitch becomeinactive during deflection,
the natural frequency of a spring changes. Throughout
the cycle, the spring has a spectrum of frequency responseand not a single resonantfrequency. Thus, surging and spring resonance Ere minimized.
Nested Compression Springs
Helical compressionspringsare often used in combination becauseof spacelimitations and resonanceconsiderations.A nest of compressionsprings can store
more energy but will have lower natural frequencies
than a single equivalent spring. Nested springs are not
recommendedwhen the diametral space is so restricted
that a single spring would have an index of 5 or less.
The following design practicesapply to nestedsprings:
1. To prevent internesting,the springs should be wound
alternately left and right-hand.
2. Clearancebetween springsmust be at least twice the
diameter tolerance.
3. The most efficient distribution of load between individual springs varies with their indexes and the clearances between them. For a first approximation in designing a nest with two springs,one-third of the load
should be on the inner spring and two-thirds on the
outer spring.
4. Solid heights and free heights should be about the
same for all springs.
These practices result in springswith approximately the
same index.
Commercial Tolerances
Standard commercial tolerancesfor free length, diameter and load are presentedin Tables 54, 5-S and 5{.
Tolerance on squarenessis 3o. These tolerances represent a good trade-off between manufacturing costs and
performance in most applications. Certain premium
spring materials and processingmethods can be used to
Fig. 5-21. Typical Variable Pitch Helical Compression
Spring.
achieve tighter tolerances. If the application requires

tighter tolerances,the required tolerancelevels should be
discussedwith an Associated Spring engineer.
For fatigue applications, spring life is often specified.
Unless otherwise stated, life is interpretedas the Srolife.
This is the life at which 9Vo of the springs are expected
to survive with a 50Voconfidence level basedon Weibull
analysis.
Acceptable Quality l-cvel (AQL)
Quaitv levels are often expressed by an AQr, (for
example MIL-STD-l05, Sampling hocedures and Tables for Inspection by Attributes). Only critical attributes
should be subject to an AQL on the drawing. Unnecessarily tight AQL's will increase manufacturing and inspection costs. If tolerances must be close for proper
functioning and if, for instance, nonconforming parts can
be discarded at assembly, a standard AQL will minimize the parts cost. Springs assembled automatically
often require tight AQL on dimensions, while springs
used in instruments and critical machines often require
tieht AQL on loads and life. A close liaison between
Aisociated Spring engineers and the designerduring design and prototype phases is the best way to ensure
optimum quality.
Packaging
Normally, compression springs and other custom
parts are packaged in bulk. Compression springs with
high pitch angles and large indexes are subject to tangling. Tangling not only makes it diffrcult to separate
springs upon arrival but can also cause distortion. SpeciA pictcigrne systemssuch as the Spring FlowrM system
where springs are packaged in rows (Figure 5-22) is one
method to prevent tangling. Another method is to place
springs on adhesive-coatedcomrgated panels. There are
Toble54. Free Length Tolerances of Squared and
Ground Helical CompressionSprings.
I!*ur*lbar of
Aeli?e Coi.l'*
par rnrfi{in.}
ffinrmt*
4
t:rntu/nm {h.ltn,} of fr&cril

E
ro
t2
T{
,5
0.02
(0.5)
0.010 0 . 0 1 l 0.012 0.013 0.015 0.0r6 0.016
0.04
(l)
0 . 0 1 l 0.013 0.015 0.016 0.017 0.018 0.019
0.08
(2)
0.013 0.015 0.017 0.019 0.020 0.022 0.023
0.2
(4)
0.016 0.018 0.021 0.023 0.024 0.026 0.027
0.3
(E)
0.019 0.022 0.024 0.026 0.028 0.030 0.032
0.5
(12)
0.02r 0.024 0.027 0.030 0.032 0.034 0.036
0.6
(16)
0.022 0.026 0.029 0.032 0.034 0.036 0.038
0.8
(20)
0.023 0.027 0.031 0.034 0.036 0.038 0.040
for 12-7
For springsless than 12.7mm (0.500) long, use-the-tolerances
mm (0.500). For closedends ndt glound, multiply abovevaluesby 1.7'
G
G
-
C
C
G
C
C
C
C
C
C
C
-
C
C
C
C
C
C
C
C
C
C
C
C
C
5
s
I
l.
b
I'b
a
a
I
a
a
a
a
a
a
HelicalCompression
Springs
!r'n:nl other packagingmethodsused to prevent tangling
and reduce shipment bulk.
Em to Specitv
Tbere are many ways to specify compressionsprings.
Becauserhe number of variablesis large, it is useful for
the designerto usethe specificationchecklist on the next
page to be sure that all critical aspects are specified.
Compression Spring Desrgn Example
Given: Squaredand ground compressionspring to work
in a hole DH : 40 mm (1.575')and exert Pr:275 N (61.8
lbfl at a height of Lt : 60 mm (2.362') and Pz : 500 N
t lll lbf) at a height of Ia : 50 mm (l .96y). Application:
sadc at room temperature. Material: oil tempered wire
ASTI1 N29. Spring must not set when compressedto
solid height.
A. First estimatethe wire diameter by solving equation
(Equation A, page33) using approximatevalues for
unknown factors and Kq,1= 1.
Then, calculateO.D. and D.
Substitute this wire size in the load deflection equation
(Equation 5-2, page 33) and solve for N". Repeat this
process until a satisfactory solution is obtained.
l. Rearrangng Equation 5-4 for uncorrected stress:
A _: vlz.ss
o
T PD
!D
I
l. Assum" ,.nril. strength of ASTM Al2}gis 1500MPa

and S: = 700 MPa uncorrected:
a
a
3. For clearance,assumeO.D. :0.95 Ds:

O.D. : 0.95x 40 : 38.0 mm
a
I
I
t
a
a
'/ffi,
d -
r r - - T . ! r r l l l l
700
Tablc 5-5. Coil Diameter Tolerances of Helical Compressionand Extension Springs.

Tolemmffr: t:mm:,(lui)
Wirc
IXa.,
$pring lrdn {D/d}

t
IB
L2
l{
t6
0.38
0.05
0.05 0.08 0 . 1 0 0 . 1 3 0 . 1 5 0 . l E
( 0 . 0 1 t (0.002) (0.002)(0.003)(0.004) (0.00s) (0.006) (0.007)
E(iD.)
0.58
0.05 0.08 0 .1 0 0 . 1 5 0 . 1 8 0.20 0.25
r0.023) (0.002) (0.003)(0.004)(0.006) (0.007)(0.008)(0.010)
0.E9 0.05 0 . 1 0 0 . 1 5 0 . l E 0.23 0.28 0.33

r0.035)(0.m2) (0.004)(0.006) (0.007) (0.00e) (0 .0 1 l )(0.013)
1.30
0.08 0 . 1 3 0 . 1 8 0.25 0.30 0.38 0.43
(0.05r) (0.003) (0.005) (0.007) (0.010)(0.012)(0.0r5) (0.017)
1.93
0.10
0.lE
0.25 0.33 0 . 4 1 0.48
0.53
(0.076) (0.004) (0.007) (0.010)(0.013)(0.016) (0.019) (0.021)
:.90
0 . 1 5 0.23 0.33 0.46
0.53 0.9
0.74
(0.00e)(0.013)(0.01E) (0.021)(0.025)(0.029)
r 0 . 1 1 4 (0.006)
)
1.y
0.20 0.30 0.43 0.58 0 .7 1 0.84 0.97
( 0 . 1 7 1 )(0.008) (0.012)(0.017)(0.023) (0.028)(0.033)(0.038)
D = 38.0- 4.2- 33.8mm
c =? = # : 8 . 0
59- ?-T:22.5
Rate= k =
N/mm
-50
60-
Gd4
N":
8m
|rJ"
" =-
(7.93 x 104)(4.2)o
=
8G3*8n22i)
B. Find amount of space left between L2 and L5:

1. Compareto f2.
2. Find the corrected stress at solid height.
3. Compare to tensile strength of material. See Figure
3-3, page 19.
L , : 5 . 5 5x 4 . 2 : 2 3 . 3
)1\
P,
,
I4:
* Lt :
+ 60 :72'2 mm
kt
zL:t
fz: 72.2- 50 : 22.2mm
Lz - L, : 50 - 23.3:26.7 mm
f, = 72.2 - 23.3 : 48.9 mm
15%of 48.9 : 7.3 mm
4C - l
.615
K w r :f f i +
1.18
a:=
P , : f . x k : 4 8 . 9 x 2 2 . 5 : 1 1 0 0N
cr :_ 2.55P,D__-_ (2.55)(il00)(33.8)(1.18)
b'
-5-K*t
: 1510MPa
4 . Tensile strength of 4.2 mm diameter wire : 1400
MPa. Before set is removed,ma:cimumallowable torsionai stress is 5Vo of TS or 700 MPa (Table 5-2,
page 35). S, = 1510is greater than 700 MPa, and the
spring will set.
C . Because (L, - L,) : 26.7> 0.15f, :7.3, there is
more spaceavailable. Try a larger preferred wire size
(Table 3{, page 20) of 4.8 mm.
TS = 1400MPa, D : 38.0- 4.8 : 33.2mm, C = 6.9
N":
(7.93x l0a)(4.8y
:
8(33.2)3
Qz.s)
6.4
L, : 8.4 x 4.8 : 40.3mm

Lz-L.:50-40.3:9.7mm
f, = 72.2- 40.3 - 31.9mm
(Lz - L,) : 9.7 > 0.15f, = 4.8 mm
= 718N
P, : (31.9)(22.5)
(+X6.e)_l*#:r.zz
a
a
6.3,(
0.28 0.38 0.53 0 . 7 1 0.90
1.07
1.24
(0.250) ( 0 . 0 1 l )(0.01s) (0.021) (0.028) (0.035) (0.042) (0.049)
9.53
0.41
0 . 5 1 0.66
0.94
t.l7
1.37
1.63
( 0 . 3 7 t (0.016) (0.020) (0.026) (0.037) (0.046) (0.054) (0.064)
s , - @ : 6 7 t (4.8)'
Mpa
11.70
0.53 0.76
r.02 r.57 2.03 2.54 3 . 1 8
r0.500) ( 0. 021)(0.030) (0.040)(0.062)(0.0E0)(0.100) (0.12s)
S. : 671MPaot
a
a
3.55
rawt=(4x6.9)-4-et=
ffi
x loo - 48voof rS
Assos$tf;g&
wB
ffi*ffirys
C
C
CHECKLIST
SPRINGSPECIFICATION
COMPR,ESSION
(Fill in requireddoto only)
Moteriol:
Frequency of looding
Required life
Required reliobility (see Section 4)
Wo*ing Conditions:
Speciol Informotion:
m(in.) diometer hole
m(in. dismeter shoft
To work in
To work over
(1b0,*
Lood
tl (lb0
(lbf/in.)
(lbf/in.), t -N/mm
Rote -N/mm
qnd -mm(in.)
between -rnm(in.)
mm(in.)
Moximum solid height
Direction of coil (right-hond, left-hond or opfionol) -
:c("F)
Wire diometer
Outside diom
lnside diometer
Free length
Totol number of coils
mm(in.)
Tablc 54. Load Tolerancesof Helical CompressionSprings.

X.*ugth
Tskrarr*
::mm {in.}
r"*,l""*'*qb."r1yrd..y{,*t
Tl.e.rypry|,]:i
, l.Y'H.bl,f:
,,to',.[ , mm {in.)
fien f,romfre*,,I
"28*, t*.1 fO;il,.,'l?6 ,,.1,.i,ltrEl ,,:1 ,:
'19*,1
'l;3;7' ?.54 3,*r. 5"S0 ,,f,$$ "V,67, 10.2
1:,7
,'
(s.s3o){0,ltr} {0,1$6} {0"m,}{0,2sJ t0.300)(0.400,1{ 0 } {s-.?50}{r;tr} (r (2'ffi1 {,s,;} {*$1 {6iffi}
0.13(0.00s) ,:
0.25(0.010)
0.51(0.020)
0.76(0.030)
1.0 (0.040)
r.3 (0.050)
1 . 5 (0.060)
1 . 8 (0.070)
2 .0 (0.080)
7.
t2.
22.
6.
8.5
15.5
5.
7.
t2.
17.
22.
.:I
6.5
10.
l.s
5.
7.
8.s
6.
r8
))
9.5
t2.
1 5 . 5 12.
14.5
19.
8
l0
12
6.
7.5
9.
5
6
7
s
5.5
22.
25.
l4
l6
l8
10.
11.
12.5
8
9
10
6.
6.5
7.5
20.
14.
15.5
1l
t2
22
T4
2.3 (0.0e0)
2.s (0.100)
5.1 (0.200)
7.6 (0.300)
10.2 (0.400)
12.7 (0.500)
17.
19.5
22.
Fis.5-22. Tansled Helical CompressionSprings (Left)

and-Spring Flow P ackaging.
5.
).)
6.
6
8.
8.5 7
1 5 . 5 t2
Firstloadtestat not lessthanl5Voof availabledeflectron.

Finalload test at not morethan85%of availabledeflection.
c
C
Design Dofo (Reference):
Hours/doys mm/sec (in./sec)
C
C
Porollelism
Squoreness
Finish
Moximum operoting fem Peroture
Operoting environment
Electricol/ mognetic
x (tbO
Type of ends
Alfowobfe reloxotion -o/o
lmpoct looding
cYcles'
17
2l
25
5.
5.5
8.5
7.
9.5 7 .
t2.
8.5
1 5 . t2.
18.5 14.5 10.5
Again referring to Table 5-2, page35, it is clear that the

spring can be made without presetting. Tolerances are
obtaineOfrom Tables 54, 5-5 and 5-5. The final design
becomes:
Final Design Specifications:
Material: ASTM A229
Wire Diameter d: 4.8 mm (0.189) Reference
O.D.: 38.0 -r 0.4 mm (1.500+ 0.050)
Free Length I-r: 72.2 mm (2.843') Reference
Test Height Lr: 60 mm (2.362')
Test Height Lz: 50 mm (1.96Y)
Pr Load at Lr: 275 N (61.8 lbO = ll.0%
Pz Load at L2: 500 N (112 lbO =7Vo
Final Design StressS,: 671 MPa (97,300psi) or 48%TS
N,: 8.4
C
5
C
C
C
C
C
5
5
5
E
s
s
C
C
5
5
HelicalExtension
7
Springs I
D
a
a
a
a
a
I
I
I
I
I
o"tF'!fi349*nrrs
!t
HelicalExtensionSprings
Introduction
Helical extension springs store energy and exert a pulling force. Usually, they are made from round wire and
are close-wound with initial tension. Typical applications
includetape cassetteplayers,balancescales,toys, garage
doors, auiomatic waJhing machines and various types of
tensioning devices
spring
Helical extension springs are stressed in torsion in the
body. Design proceduresfor the body are similar to those
discussedpleviously for compressionsprings(Section5)
with the following major exceptions. Most helical extension springs are coiled with initial tension, equal to the
minimum force required to separateadjacent coils. Helical extension springs do not normally have set removed.
Furthermore, untit<e compression springs, extension
springs do not have a solid stop to prevent overloading.
For thesereasons,designstresslevels are generally lower
for extension than for compression springs. A special
type of extension spring, known as a drawbar spring (figurl 7-1), has a solid stop. It is essentiallya compression
spring with special hooks.
Fig.7-1. Drawbar Spring Provides a Solid Stop.
Fie. 74. Load Deflection Curvefor a Helical Extension

Spring with Initial Tension.
The pulling force exerted by an extension spring body

is tranimitted to mating parts througfi hooks or loops.
When stressesin the hooks are higher than in the spring
body, the hooks limit spring performance.
Alsociated Spring includes hundreds of different extension spring designs with full twist loops in its SPEC
line of stbck iprings. These extension springs are made
from.either music wire or stainless steel and are preengineeredto meet a wide range of applications.
Initiat Tension
Initial tension in an extension spring is rneasured according to the procedure illustrated in Figure 7-2. T\e
linear portion of the load deflection curve is extrapolatgf
to zero deflection. The point of intersection on the ordinate is initial tension Pr. The amount of initial tension
that can be put into a spring depends upon its index,
material, method of manufacture and postcoiling stressrelief treatment. Occasionally, in critical applications
when stress is high, a high stress-relief'temperatureis
required to minimize unfavorable residual stresses due
to Loiling or forming the hooks. High temperature stress
relief reduces the amount of initial tension. Typical values of initial tension are shown in Figure 7-3. High
strength materials such as small diameter music wire are
able io support higher levels of initial tension than low
strength materials such as large diameter hard-drawn
wire.
of Ends
Types
-Extension
springs require a method of attachment to
parts
in
an assembly. A wide variety of ends has
other
Fig.7-3. Torsional StressDue to Initial Tension as a
Function of Index in Helical Extension
Springs.
275
Lood (P)
P1
250
2s E
.g
9 c
5 P
20;E
g s
j f rso
o . 9
125
C
C
C
.
c
.
c
a
c
e
c
c
c
c
C
I
.n
r s . E TC
;
;tg
.E
T roo
I t
o
o -
to " c
gg t7s
U|.=
:
8.
30
t v
3s
tn
22s
U =
Deflection(f)
=3 2oo
9
300
p 6
7s
l0
50
c
I
25
l0
lndex--+
12
I4
l6
I
I
(
I
HelicalExtensionSprings
been developed and used successfullyfor many years for example,threadedinserts, swivel hooks, twisi loops,
side loops, cross-center loops and extended hooks.
l-oops are attachment ends that have small gaps (Figure
-{).
while hooks.are loops with a large gap. In fact, the
'ariet-v-of ends is almost unlimited. The most common
configurations are those that can be formed during the
springmaking operation. Typical types include twist,
crosscenter, side loops and extendedhooks (Figure 74).
Man), of these configurations are made by bending the
last coils of an extension spring to form loops. Most
special hooks are formed from straight sectioni of wire
on the so-called "tangent ends" of an extension spring
body.
Guidelines for the lengths of common loops are
presented in Figure 74. Alrhough other configuiations
and lengths are available, common loops of preferred
lengths are generally the most eccnomical. If possible,
a spring should be designed with one or both loops at
the prefered length. For example, if a design requires
a-total loop length equal to five times the I.D., a popular
choice is one twist loop with a length equal to the I.D.
and one extended loop with length equal to four times
the I.D. Wheneverpossiblefor extendedloops, the designer should allow for a straight section approximately
three wire diameters long at the end of the wire (A,
Figure 74). Loops at each end can be made with a controlled angular relationship. Specifying an angular relationship may add to the cost; therefore, whenever an
application permits, a random angular relationship
should be allowed. Production of special end configurirtions may involve tool charges and generally resufts in
increasedcosts.
Stressesin loops are often higher than in the spring

body. This limits spring performance, particularly in cyclic applications. Generous bend radii in loops and reduced end coil diameters are two methods frequently
employed to reduce stresses.In a full twist loop, stress
reachesa maximum at point A in bending and a maximum in torsion at point B (Figure 7-5). Stress at these
locations is complex, but can be estimatedwith reasonable accuracy by:
se: S*,
t-rt
#bending
lRr
whereK
? t -:4+Ctt 2
f f- Ci at - tn d c r : ?
2R,
A : 8DP/4C,-l\
Ss
andCz:
torsion e4t
A \4ffi)
?
Recommendedpractice is to make C2geater than four.
Fig.7-5. r,ocation of Maximum Bending and Torsion
Stressesin Twist Loops.
P
rl
\ ltl
,-r;lF(--=r:f--
Torsion Slress
otB
Fig. 74. common End configurations for Helical Extension Springs.
?ypr
hceCImm*ndd,,Lrn#h*
Min"-t\{sx,
::'--::i"T_
--W
mJ -
Twist
Loop or
Hook
CrossCenter
Loop or
Hook
A\UZ
/A
-Tr-rr-7r-
- 7r
ullv/
I.D.
\Jlz
-uNzrz
Side
Loop or
Hook
I.D.
0.5-1.7
rmA
-TrnF
#@
v p-
I.D.
0.9-1.0
Extended
Hook
Special
Ends
l . l I . D .a n du p
as requiredby design
vvvvvvv -ffi
'
@-F
t
A A
av )i Ll
<
I
'
6\ i lZl Y
l
As requiredby design
' length
is distancefrom last body coil to inside of end. I.D. is inside diameter of adjacentcoil in spring body.
A'"oEFIf;g
&wwyvW
Helical ExtensionSprings
Extension Spring Dimensions
Free length of an extension spring is the distance between the inner surfaces of the ends (Figure 7-6). It is
equal to the spring body length plus ends, where spring
body length is given by Lur, = d(N + 1). The gap, which
is sometimes referred to as hook or loop opening, can
be varied by the springmaker. Certain manufacturing
processes require a minimum gap and the designer
should consult Associated Spring engineersif a gap must
be less than one-half wire diameter. The number of active coils in a spring is approximately equal to the number of coils in its body. For springs with threaded inserts
or swivel hooks, the number of active coils is less than
the total coils in the body. Hooks and loops add to the
number of active coils. Allowances of 0.1 N" are occasionally made for one-half twist loops. Allowances as
large as 0.5 N" can be made for some cross center, full
twist and extended loops.
Desigp Equations
Design equations for extension spring are similar to
compressionsprings. The rate is given by:
16:
P-Pr
:m
Gd4
(7-3)
wherePr is initial tension.Stressis given by:

8PD-5 = --liTKw
(74)
Dynamic considerations discussed previously in Section

5 are generally applicable to extension springs. Natural
frequency when one end is fixed is given by:
5.6 x ldd
n:5;;V7
EE
t76xlffd
forsteeln=j::+W-
metnc
metric
metn(
n: ofu.',tr English
Tooog
fcr r s t e e l n : N F English
Fig. 74. Typical Extension Spring Dimensions.
(7-s)
Choice of Operating Shess - Static

Recommended maximum stresses for extension
springs used in static applications (Table 7-1) are similar
to levels recommendedfor compression springs without
set removal. For springs that cannot be adequately
stress-relieved due to high initial tension requirements,
the ma,rimum recommendedstress in the body should be
reduced to that recommended for their ends. Maximum
recommended stress in the ends is lower than in the
body because the wire is often stretched, marked or
distorted during loop-making.
Choice of Operating Stress - Cyclic
Maximum recommended stresses for extension
springs used in cyclic applications are presentedin Table
7-2. These data are for stress-relieved springs with low
levels of initial tension.
Tabte7-1. Maximum Allowable Stresses('Kn,Corrected)
for Helical Extension Springs in Static Applications.
Pcrst
fi4rnniak
Patented, cold-drawn or
hardened and tempered
carbon and low alloy steels
Austenitic stainlesssteel
and nonferrous alloys
of Twih
Sftng{h
In
InT
Ddf
affd
End
45-50
40
75
35
30
55
Tobb 7-2. Maximum Allowable Stresses

for ASTM A228
and Tvpe 302 StainlessSteel Helical Extension Sirings in Cyclic Applications.
Scrwt-of
..
Tadk
36
33
30
@$
In Ucntrry
In tordffi
Epdl,
ld
106
107
s
-
This information is based on the following conditions: set not removed

and low temperanrre heat treatment applied.
For springs that require high initial tension, use the same percent of
tensile strength as for end.
,.,ill
ff,'ffd6
C
C
C
C
C
C
C
C
C
C
C
ffiil
34
30
28
trffi
51
47
45
This inforrration is based on the following conditions: not shotpeened, no surging and ambient environment with a low temperature heat treatment applied. Stress ratio : 0.
C
C
C
C
C
C
5
C
C
C
C
5
5
C
EA""Th!8$M
s
I
=
r-
a-
-,
.-,
Helical Extension Springs

Clearances
Extension springs,when deflected,do not require central arbors or holes to prevent buckling. When a spring
is dynamically loaded or unloaded suddenly (as a cam
drop-off), it may vibrate laterally, inducing additional
stresses.If clearanceis not allowed, this lateral vibration
may be noisy and result in premature failure from abrasion of the ppring or adjacent parts.
r-,
Tolerances
Since requestingclose tolerancescan increasemanufacturing costs, only those characteristicswhich are crit-
r-
Table 7-3. Commercial Free Lensth Tblerancesfor Helical Extension Springs Wtttt Initial Tdnsion.
rl
r-,
<,
-
$pring Sreelxngttu tirsidn koolrri

rnm fin-)
Up to 12.7(0.500)
over 12.7to 25.4(0.500to 1.00)
Over25.4to 50.8(1.00to 2.00)
Over50.8to 102(2.00to 4.00)
Over 102to 203(4.00to 8.00)
Over 203to 406(8.00to 16.0)
Over406to 610(16.0to 24.0)
Tohrsscs
* nrm {in.}
0.51(0.020)
0.76(0.030)
r.0 (0.040)
1.5 (0.060)
2.4 (0.093)
4.0 (0.155)
5.5 (0.2r8)
ical to spring performance should have tolerances

specified. Commercial free length, angular relationship
of ends,and load tolerancesare presentedin Thbles7-3,
7-4, and 7-5 respectively.O.D. tolerancesare the same
as for compressionsprings (Table5-5, page43). These
tables should be used only as a guide since some manufacturing operations have different process capabilities
which can cause variations in tolerancevalues. For special applicationsrequiring closer tolerances,consult AssociatedSpringengineers.
How to Specify
For minimum cost, it is important to specify springs
properly. The following checklist is presentedas a guide.
Table74. Toleranceson Angular Relationshipof Extension Spring Ends.
4
0.75
Arryular Tohranm ptr Coih I Deperx

Index
6
7
5
E
l0
t2
9
0.9
l.l
1.3
1.5
t.7
1.9
2.3
t4
l6
2. 6
;ffi
For example, tolerance for a lGcoil spring with an index of 8

l0x+1.5=tl5o.
If angulartolerance is greater than * 45o,or if closer tolerancesthan
indicatedmust be held, consult with AssociatedSpring.
a
a
a
H E L I C AE
L XTENSION
SPRINGS
SPECIFICATIO
CN
HECKLIST
(Fill in requireddoto only.)
SuggestedDesign Doto:
Moteriol
Wire diometer
Working Conditions:
=l
Moximum outside diometer
In i ti o lt ens ion
Lood
!r
rD
N(lbf), +
of
Lood
mm(in.)
N(l b0
length mm(in.)
N(lbf), +
ot
lmpoct Looding
a
a
Rote
Moximum extended length in service
r}
optionol
Typ" of ends
Posifion of ends ond toleronce
Gop opening ond toleronce
mm(in.)
Outsidedio
Direction of coil: right
length mm(in.)
sec (in. /sec)
N/mm(lbf/in)
mm(in.)
Totol number of coils

Free length inside ends
mm(in.)
Fi ni sh
Moximum operoting temperoture
Frequency of Looding
Required life
Required Reliobility (see Section 4)
hertz
) during instollotion
left
mm(in.)
I
I
I
L,'
F
L
I
-
iwffii
..*,,, :':ri' :r:;,,r:,.,xff:
':':
';;:ffi
"'"'
"M:'6'" "':;";: '
SPring
Associatgd
Helical Extension SPrings
Table7-5. Load Tolerancesfor Helical Extension Springs.
Sodg
I"e.qgth
J
1 Uirldcd
,bf
Ildlecth*r
Tokrrm*r: * Pstcnt of:,$hgroilld ld

qpcl
{Cb*rr tohnnct* nry rt@ll dlt*ud
i
k&x
a i
d
I
i
*p
r
*JS
t0..CIls)
r2
20.0
18.5
r6.8
15.0
13.1
8
6
4.5
2.5
1.5
0.5
L2
l4
l6
18.5
17.5
16.1
t4.7
LZ.4
9.9
5.4
,,,,.,f;{
*s1
{0;03?} t0;0*tl
17.6
16.7
15.5
14.l
r2.r
9 .3
4 .t
16.9
15.8
t4.7
13.5
l1.8
8.9
4.6
tr"f
(0.ffiat
16.2
15.0
13.8
12.6
10.6
8.0
4.3
14.3
13.2
11.8
r0.3
8.5
6.5
3.8
ta;ffSt
',:,1'8
{off,f,}
t1.l
@8n,
13.8
t2.5
tl.2
9.7
8.0
6.1
3.6
r3.0
12.3
I1.5
l0.l
8.7
7.4
6.2
3.7
12.0
10.6
9.3
8.0
6.5
5.0
3.4
11.5
9.9
8.4
6.E
5.3
3.3
t2.6
11.0
9.4
7. 9
6.2
4. 8
3.2
13.5
12.6
I1.6
10.5
9.4
7.3
4.5
l3.l
12.2
10.9
10.0
9.0
7.0
4.3
12.7
tL.7
10.7
9.6
E.5
6.7
4.1
12.0
11.0
10.0
9.0
7.9
6.4
4.0
11.5
10.5
9.4
E .3
7.2
6.0
3.7
tt.2
10.0
8.8
7.6
6.2
4.9
3.5
10.7
9.5
8. 3
7.1
6.0
4.7
3. 4
1 3.0
12.l
rt.2
l0.l
8.8
7.1
4 .8
t2.l
12.0
11.0
r0.0
9.0
7.9
6.7
4.4
I1.5
10.6
9.7
E.7
7.7
6.5
4.2
r0.8
r0.2
l 0.l
9.3
8.3
7.4
6.2
4.1
9.4
E .6
7.E
5.9
5.E
3.9
10.0
9.0
8.1
7.2
6.1
4.9
3.6
9. 5
E. 5
7. 6
5.6
5. 6
4. 5
3.5
1 2 .0
1r.6
1 0 .E
r0.0
9.0
7 .5
5.1
11.9
1l.l
10.5
9.8
9.0
8.0
6.7
4.7
10.9
10.5
9.7
8.9
8.1
6.3
4.3
9.9
9.2
E.6
7.E
7.0
6.0
4.2
9.3
8.6
8.0
7.3
6.5
5.6
4:0
9.2
8.3
7.6
6.8
5.9
5.0
3.E
E. 8
8. 0
7. 2
6.4
5.5
4.6
3. 7
1l.l
10.7
10.2
9.4
8.5
7.2
5.3
10.8
9.7
9.0
E .0
6.E
5.1
10.1
9.6
9.0
E.4
7,E
6.5
4.9
9.E
9.3
8.5
E .0
7.4
6.3
4.7
9.5
8.9
8.2
7.6
7.0
6.1
4.5
9.0
E.4
7.9
7.2
6.6
5.7
4.3
E .5
7.9
7.4
6.E
5.1
5.4
4.2
8.2
7.5
6.9
6.3
5.6
4.E
4.0
7. 9
7. 2
6.4
5. E
<)
4. 5
3.3
l1.3
10.9
10.4
9.7
8.8
7.7
5.9
r0.2
9.1
8.7
8.3
7.8
7.1
6.3
5.0
8.8
8.3
7.7
8.4
8.0
7.5
7.0
6.5
5.8
4.6
8.1
7.6
7.2
6.7
6.2
5.5
4.4
7.6
1' '
9.8
9.3
8.7
8.1
7.0
5.4
9.7
9.2
8.8
8.2
7.6
6.7
\)
1' '
4.3
5.8
6.3
5.8
<)
4.7
4.2
7. 0
6.4
5.9
5.4
5. 0
4.5
4.0
10.3
10.0
9.6
9.1
8.4
7.4
5.9
9.2
8.9
8.5
8.i
7.6
6.6
5.5
8.6
8.3
8.0
7.5
7.0
6.2
5.3
8.1
7.8
7.5
7.2
6.7
6.0
5.1
7.4
7.2
6.9
6.5
6.1
5.6
4.8
7.2
6.E
6.5
6.2
>.t
5.3
4.6
6.8
6.5
6.2
5.8
5.4
5.1
4.5
6.3
6.0
5.7
5.3
4.9
4.6
4.3
6.1
5.7
5.4
5.1
4.7
4.4
4.1
12
8
6
4. 5
2. 5
1.5
0. 5
17.0
16.2
t5.2
13.7
11.9
9.9
6.3
15.5
t4.7
14.0
12.4
10.8
9.0
5.5
14.6
13.9
12.9
11.5
10.2
8.3
4.9
t2
8
6
4.5
15.8
13.1
12.5
1.5
0. 5
12.8
tt.2
9.5
6.3
t4.3
13.7
1 3 .0
rt.7
10.2
8 .6
5 .6
t2
8
6
4. 5
2. 5
1.5
0. 5
14.8
14.2
t3.4
t2.3
1 0 .8
9.2
6.4
13.3
12.E
12.l
10.8
9.6
8 .3
5 .7
t2
8
6
4.5
) <
1.5
0.5
14.0
13.2
12.6
12.3
11.8
10.5
8.9
6.5
1 0 .2
9.2
8 .0
5.8
T2
E
6
4. 5
2. 5
1.5
0.5
13.1
t2.4
11.8
ll.l
10.1
8.6
6.6
t2
8
6
4.5
2.5
1.5
0.5
1 2 .3
ssociated
15.0
14.0
t2.7
11.5
9.1
7.0
4.0
{0;lft}
, ::g;l['::.,
t4.I
13.4
t2.3
11.0
9.8
7.7
4.7
? s
1.5
0.5
11.0
10.5
9.7
8.3
6.7
15.5
14.5
13.2
12.0
10.0
7.5
4.1
. :,
:,'1,!l;
1l.l
9.6
8.0
4.4
14.8
13.5
12.3
rr.7
3.t
{0"r2ft
l2.E
15.5
t4.3
13.2
tl.7
10.0
E.4
4 .6
rr.7
,,3
f0,0e!)
13.5
12.5
tl.4
10.2
8.7
7.2
4.1
t5.7
t4.9
13.8
t7.3
1 0 .8
E.9
4 .9
r5.0
t4.2
t4.l
13.0
11.8
10.5
9.0
7.7
4.2
16.5
15.7
14.8
l3.l
1l.3
9.3
5.4
? s
r0
r0.2
6.2
0;56
(0.f22)
l7.E
l6.E
15.8
14.2
12.3
10.0
6.2
t2
E
6
4. 5
Whc Dfiril#cr:
"}
rr.2
rr.7
1.0.7
9 .5
7 .8
5 .0
snrindf&S&'ffiflq$
rr.2
10.5
9.5
E.4
6.9
4.9
r0.2
rr.4
10.6
9.7
8.3
6.9
4.5
r0.2
9.3
8.5
7.7
6.5
4.5
1 J
6.7
6.0
4.8
1 1
7.4
7.1
6.8
6.3
5.8
5.0
t.5
tr.7
10.7
9.6
E.l
6.5
3.9
rt.z
r0.0
8.8
1 <
6.8
6.3
5.7
< t
6.r
4.5
3.3
G
C
C
IE
C
C
C
C
-
C
-
3
C
C
.
C
C
C
C
C
C
C
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
I
I
a
a
a
a
Helical ExtensionSPrings
r-rtrrdon Spring Desi$ Example
-{ spring is to be incorporated into an overload circuit
rreaker. It is to be preloadedat length Lr :25.00 mm
0.84-l and must exerta load of 17.5N(3.93lbf), + IsVo,
sten the circuit breaker is closed. If overload occurs,
Cre circuit breaker is tripped and the spring is extended
u a lengrh Lz : 29.00mm (1.142').The load must be 30
\ - ll%. to operatea lock, preventing accidentalresetnng before the malfunction is corrected. Either twist or
ertended loops with generousradii are satisfactory.Because of surroundingcomponentsthe maximum O.D. is
nrm (0.276'). Probability of overloads is small and
t'reaker operation is expected only three or four times
rn a )'ear. The spring will not be extended beyond Li
during service or installation.
For static application, in an ambient environment, the
material selected is ASTM A227.
7 . Calculate free length I+ and deflections fr and fz;
l. Assume a clearanceon O.D. of lWo:

O . D . - ( 0 . 9 )( 7 ) : 6 . 3 m m
9. Check stressesin the hooks:

Bending Stress:
l. Assume Sz : 700 MPa uncorrected; let D = O.D.

:6.3 mm
Calculate wire diameter d:
o:
' i(2.ssx30)(6.3)
/zffi
\r::
Y
700
: 0.88 mm
Tensile strength taken from Figure 3-3, page 19, is

1790MPa.
Calculate mean diameter.D and coil index C:
D = O . D . - d = 6 . 3- 0 . 9 : 5 . 4 m m .
7 4
4 . Calculate mean stressat the extended length:
Lz = 29.00DD, Pz = 30 N
2.55PzD.,
)=: -a3-n*t
1 - 0 . 6 1 5 _ 4 ( 6 ) - I _ -0 . 6 1 _5 1
46)----:1.25
c
(2.55X30X5.4)(1.2s)
(0.eOf
2.55 PrD
sd r = - - F
(2.ss)(7
.42)(s.4)
: 140MPa
(0.e0)'
Referring to Figure 7-3,it can be seenthat this is in the

preferred range for initial stress for an index of 6.
- C- I
l6PD..
A :fff'
f, = 4Cr2
- 4P tKr
Se
#
ffi
: 708 MPa or 4Vo TS
Calculate rate k:
g =,Pt P' : 39^-Y;t :3.13

r ' L e N/mm
Lz-L,
29-25
,
; letCr:C
4 ( . 6 ) 2 - 6 - r:
1.142
4(6X6- l)
(16)(30X5.i1)9.142)
* 1^3%,: LJ
So:
vA
'v 'Mpa or
1340
Kr:
z(0'90)z
74.9%o
TS
Torsional stress, where Rz : 2.70 mm:
8PD14C2-1\ ^
c r =;Ar
DB
t,:
\rc= )
2Rz
6 _ (8)(30)(s.4)t4(6) - l\
ss:
=1ffi
C:Drdi;:6
Su:
8 . Calculate initial tension P1 and uncorrected stress

due to initial tension 51:
Pr : Pr - kf' : 17.5- (3.13)(3.22):7.42N
z(0'90)3
letd:0.9mm
tK w r_: 4
NC -
assumefull twist loops:

I-r : 2(ID) + (N" + l)d : (2)(4.5)+ (13.2+ 1X0.9)
: 2 1 . 7 8m m
f, = L, - I4 : 25.00- 21.78= 3.22 mm
fz: Lz - Lr :29.00 - 21.78: 7.22mm
rs
(Affi ):6sr MPaor 36vo
Final Desig Specifications:

Free Length l-t 21.78 mm (0.854') Reference
Outside Diameter: 6.3 + 0.10 mm (0.248f 0.004'J
Wire Diameter d: 0.9 mm (0.035t Reference
Initial Tension Load Pi: 7.45 N (1.68lbf) Reference
Extended Length L1: 25.00 mm (0.9&4')
Extended Length Lz: 29.00mm (1.142')
P1Load at L1: 17.5+2.0 N (3.93*0.45 lbO
PzLoad atLz:30 +2.5 N (6.74+0.55lbO
Final Design Stress Sz: 708 MPa (103,000psi) 40"76TS
N": 13.2Coils
Refer to the load tolerances for helical extension springs
(Table 7-5). Tolerance on load for P1is -r llVo, which is
less than the required + lsVo, and Pz is = 8Vo,which is
less than the required + lTVo.
6 . Calculate number of coils N":

Gd4 Q.g3 x 104)(0.90)4
N":d
m:ffi:13.2
rY
a
I
AssoS&!fi$,'eP*sqres
ffi
C
C
C
C
C
C
C
C
HelicalTorsion
Springs
C
C
.
g----''-';
'' "
ff
&
Lff
3
5
c
I
a
a
!D
!D
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
a
t
t
D
b
L
a
a
Helical TorsionSprings
lntroduction
Helical springs used to apply a torque or store rotadonal energy' are commonly referred to as torsion
springs. The two most common types are single and
double-bodiedsprings (Figure 9-l). Torsion springs are
found in clothes pins, window shades, counterbalance
mechanisms,ratchetsand various types of machine components. They are also used as couplings between concentric shafts such as in a motor and pump assembly.
Torsion springsare generally mounted around a shaft or
arbor. and must be supported at three or more points.
Yarious kinds of ends are available to facilitate mounting.
Torsion springs are stressedin bending. Rectangular
uire is more efficient in bending than round wire, but
due to the premium cost of rectangular wire, round wire
is preferred. If possible, a torsion spring should always
be loaded in a direction that causesits body diameter to
decrease.The residual forming stressesare favorable in
this direction, but unfavorable when the spring is loaded
in a direction which increases body diameter. Unless
there are unfavorableresidual stressesin the end bends,
springmakers normally heat-treat these springs at a low
temperatureto stabilize the end positions rather than to
fulll' stressrelieve them. If the direction of loading tends
to increase body diameter, the springmaker should be
advised to stress relieve the springs.
The Associated Spring SPEC line contains many torsion spring designsusing stainlesssteel and music wire,
either left or right-hand wound. These springs have tangent ends and are available for immediate delivery.
\umber of Turns
The number of active turns in a helical torsion spring
is equal to the number of body turns, plus a contribution
from the ends. For straight torsion ends, this contribution is equal to one-third of the moment arms and is
usually expressed as an equivalent number of turns:
N":t#
(e-t)
Lr : length of the moment arm of the first end.

Lz: length of the moment arm of the second end.
(e-2)
N":Ns*N.
shaft or tube and spring at all times to prevent binding.

The ideal shaft size is equal to, or slightly less than,90Vo
of the I.D. when the spring is fully deflected (minimum
diameter). Shafts significantly smaller than 907a should
be avoidedto prevent buckling during large deflections.
Length
Most torsion springs Ere close-wound, with body
length equal to the wire diameter multiplied by the number of turns plus one. When a spring is deflected in the
direction that will reduce the coil diameter, body length
increasesaccording to:
(9-4)
L = d ( N o + 1 + 0 )
For applications that require minimum hysteresis,

springsshould be designedwith space between adjacent
coils to reduce frictional losses.
Spring Rate
Spring rate for helical round wire torsion springs is
given by:
,
t : - _
(e-s)
l0.8DN"
The 10.8factor is greater than the theoretical factor of

10.2 to allow for friction between adjacent spring coils
and between the spring body and the arbor. This factor
is basedon experienceand has been found to be satisfactory. Loads for torsion springs should be specified at
a fixed angular position and not at a fixed deflection from
the free position (Figure 9-1). Presenfly, there is no standard way to test loadsfor torsion springs. Consequently,
in critical applications,it is advisableto contact Associated Spring engrneersto establish a test method during
prototype work.
Fig. 9-1. Specifiing Load and Deflection Requirements

for Torsion Springs.
l-L
Nu : number of body turns.

Mean Dinmeter
Mean diameter is equal to I.D. plus O.D. divided by
tu'o. When the direction of loading tends to reduce the
bodl' diameter,the mean diameter changeswith deflection accordineto:
D - -
D,N,
(9-3)
N o + 0
*here D, is initial mean diameter and d is deflection in

revolutions. Clearancemust be maintained between the
E n d si n
Free Position
E n d si n
Free Position
Specify:
c =
P=
L=
0 =
Angle between ends

Lood on ends ol a
Moment orm
Angulor Deflection from Frce Posiiion
AssogFifiS&ff*nfn$
@
Helical Torsion Springs

Stress
Stress in torsion springs is due to bending, and for
round wire is given by:
n : 32M,
(e-6t
d
f"
n: ,ffiV7;
During elastic deflection of a curved beam, the neutral

axis shifts toward the center of curvature, causinghigher
stress at the inner surface than the outer. Wahl (Reference 3, page 102) has calculated the stress correction
factor at the I.D. of a round wire torsion spring as:
n:
f7
ABrD:
C -
(e-7)
rc16 _ t)
A convenient approximation for engineeringcalculations
is:
4C-l
f,
llBrD:4C_4
4C+1
l(gop:4C+4
(e-8)
(e-e)
At low indexes, stress is significantly higtrer on the inner

surface than the outer. These factors are useful to determine the stress range for cyclic applications and the
set point for fully stress-relievedsprings in static applications. A stress correction factor of 1 is recommended
to determine the set point of springs that have favorable residual stresses induced by yielding during forming. Yielding results in a more uniform stress distribution
over the round cross section. Therefore, the actual
stress correction factor approaches the recommended
value of l.
End Configurations
Some of the more common end configurations available are shown in Figure 9-2. Special configurations are
available on request. In designing ends, it is important
to recall that bends, loaded to decrease their radius of
curvature, have favorable residual stresses. They can
operate at higher applied stress levels than bends that
increasethe radius by loading. Frequently, spring performance is limited because the sharply bent ends have
greater stress than the body. Equation 9-6 is generally
employed to determine maximum bending stress in the
ends. Torsion springs are subject to surging and resonance phenomena. The natural frequency n for a torsion
spring with one end fixed is:
n:
t.26xldd r
ffi
and with both ends fixed:
ff*,
4C2 -
: 2 x t d d meric9-Ioi
VT ; forsteet ffi
C
C
8040dF-^,:-,English
forsteel:
ffi
i-
2.5 x ld,
4 x tfd _--^--:
metricQ-II.
VT ; forsteel:ffi
ffid
d
E
n:
A",tFi.I" !-
lE8od
rfor steel : tr-I\a English
To avoid or minimize resonancephenomena,the natural

frequency must be much greater than the operating frequency and/or the spring should contain initial tension.
Choig of Operating Stress - Static
Recommendedmaximum operating stressesfor static
applications are given as a percentages of tensile
strength in Table 9-1. For spring bodies or ends loaded
in a direction that increases the radius of curvature,
"stress-relieved" column are most
stress levels in the
appropriate. These stresses should be calculated using
tfie appropriateKB stresscorrection factor (Equation 9-8
or 9-9). When the outer surface is in tension, springs
with a low index usually yield at the inner surface, while
those with a higfr index may yield at the outer surface.
For springs not stress-relievedand loaded in a direction
that decreasesthe radius of curvature, the stress levels
recommended for springs with favorable residual stress
are appropriate. No stress correction factor is used since
the spring has yielded.
Fig.9-2. Com.monHelical Torsion Spring End Config'
uratrcns.
A
O-O
/
,',,}t
*s
I
ftfif l
lil ltl
t(
)) ililll
I'UU
\r/
SHORTHOOK ENDS
/6\)) [m
((
]iltl
\r/
ENDS
SPECIAT
tulj
HINGEENDS
-^.
/-\\
illt
(v/ n5liltr
tuu
I
\ l
I I
bm N
STRAIGHT OFFSET
DOU
C
C
C
C
c
.
a
.
-
c
3
e
a
c
C
C
T ORSION
TORSION
STR.AIGHT
c
c
Associatec'
sprindl&**firys
t
I
b
b
t,
F
p
F
F
P
p
F
F
F
t
P
p
F
T
T
P
p
I
a
I
Helical TorsionSprings
CLoi- of Operating Stress - Cyclic
\tarimum allowed operating stressesfor cyclic appli;arions are presentedin Table 9-2 as percentagesof tensile suength. All stressesare assumedto be calculated
rrith the appropriate Ks correction factor. This infortnaion can be used to estimatefatigue lives at other stress
rangesb1'methodsdiscussedpreviously (Section4). Frequentll'. bending stressesare higher in the ends than in
the bodl'. In this situation, bear in mind that during forrr.ng of sharp bends, the wire may be stretched or
rrarked. resulting in stress concentrations that reduce
design stresslevels below those recommended.Because
of friction, the point of contact between torsion end and
arbor is often the highest stressed area.
DouHe Torsion Springs
Double-bodied torsion springs are designedusing the
sarnemethods as for single-bodiedtorsion springs. The
rate for a double-bodied torsion spring is equal to the
sum of the rates for eachcomponent. For the samewire
diameter, coil diameter and wire length, double-bodied
torsion springs have rates four times those of singlebodied types. Double-bodiedtorsion springs should be
designed so they are coiled out from the center rather
than in from the ends (Figure 9-3).
Rectangular Wire
Rectangular wire torsion springs have higher energy
storage capacities than similar round wire springs. The
general comments on round wire torsion springs apply
to springs with rectangular wire. In producing springs
Tablc 9-1. Maximum RecommendedBending .Stresses
for Helical Torsion Springs in Static Applica'
tions.
Fr,*mi nf :Tcrdle,Slrqfh
M.lrtrid
Patented and
Cold Drawn
from rectangular wire, the wire cross-section distorts

and becomes "keystoned" (Figure 94).The wire axial
dimension br can be estimated from:
br:bF#)
When axial length is critical, keystone-shapedwire can

be purchased. This wire will have a near rectangular
shape after coiling. The rate equation is:
k:M/0=6*K
Hardened and Tempered

Carbon and Low
Alloy Steels
85
100
Austenitic Stainless
Steelsand NonFerrous Alloys
60
80
t l) Also for springs without residual stresses.

i2) Springs that have not beenstress-relievedand which have bodies and
ends loaded in a direction that decreasesthe radius of curvature.
Fig. 9-3. Preferred Windingfor Double-Bodied Torsion

Springs.
t:#*"
a
a
I
(9-r4)
These equations are for springs wound either on edge or

on flat (Figure 94).Stress correction factor Ks is slie[tly
lower than for round wire and can approximated by:
Kgro :
4C
4Cl
(e-r5)
Ksoo =
4C
4C+3
(9-r6)
Sharp corners on rectangular wire cause stress concentrations and should be avoided, while generous corner
radii of rolled wire reduce the wire cross sections sufficienfly to lower the rate.
Table9-2. Maximum Recommended Bending Stresses
(Kp Corrected) for Helical Torsion Springs in
Cyclic Applications.
fucml of Tnnr*lelitrugth
ASTeilA23t
rild TIF t&l $trtudwscel
lJf"
fi$kl
105
106
Noil Sbo{M
A"5TlilA?*rO
sd he3,2,
L$oe:. 8*
I
l snot-n t*O.
62
60
53
50
'
Sbqt-ften*d*
55
53
&
62
This information is based on the following conditions: no surging,

springs are in the "as-stress-relieved" condition
*Not always possible.
Fig. 94. KeystonedCross Sectionsof SpringsWoundon

Edge and Flat
Wound On Edge
Wound On Flot
-l-
l-br
a
a
(e-I3)
and the stressequation is:
lf,frl:f*v,otl&,,
$trr*r'&trcv ,{,tr} *eeied SrcrF {2}

l , {lio Corrac'thnFrctori
{XrCo
r00
EO
(9-r2)
br
_i_ _l_
F'-l
Spring
Axis
-l ' l--
A"TFtfi8ArynffisH
BellevilteSPringWashers
Fig. I1-2. Load Deflection Curvesfor BeIIeviIIeWashers
with Various h lt Ratios.
hh = 0.4
Fig. 11-5. Comparison of S7 and Sn for Various Deflections, h lt Ratios and Diameter Ratios (R
Values) of Belleville Washers.
L r
hh = o.7
Tl-T.c
h/r = L4l
l
h/t = 2'83
Srz Hi g h e r
t l I
#W{ffi
bd*
9-':
/,
zt
t1\
4eij
- 0.6
a
-cil
Srr Higher
.
Fig. 11-3. Mounting a Belleville Washerfor Deflection

Past the FIat Position.
Fig. 11-6. Compressive Stress Constantsfor Belleville

Washers.
2.5
6
" t = , ' l n R fl R - r l
2 )
=,
1.5
o
d
(J
Fig.114. Highest Stessed Regions in

Washers.
Belleville
o
;
1.0
o
(J
6 l - ( R- l ) ' l
=;"e-l
*r-1
777777V,
Spring
Axis
Assogb?ifi8,r,.fuffi*
ffi

Fig. 11J. Tensile Jtress
Washers.
Constants fo,
Belleville
3.0
)' 2.O
RlnR-(R-l)
"=--r.F;t1n-1y
_l .
0.5R
I
|
+ ' R - r
E
o
(J
l.)
Determining an optimum solution to a belleville washer design problem is a trial and error process which may
have to be repeated many times. A simple approach,
designed to minimize the number of repetitions, is
presentedbelow. All of the graphs are basedon belleville
washer designswith a ratio of O.D. to I.D. of 2 (R : Z).
Designs that have R approximately equal to 2 have muu(imum energy storage capacity.
The first step is to select an appropriate h/t ratio based
on the load, outside diameter and stress constraints given. For example, (referring to Figure l lJ), assumethe
desired load at flat is 1125N and outside diameter is 76
mm. A washer with an h/t equal to 1.41 would have a
maximum stress S. of 1500MPa. Loads at intermediate
deflection can readily be computed with the aid of Figure
ll-9. Material thickness is then determined from:
t=*Vmmetric
)
Before finalizing a design based on these graphs, it is
best to check results using the equations, making final
adjustmentsas required. For cyclic applications, stress
levels Srr and Sp mustbe determinedin order to estimate
belleville washer life.
T z :
I l r r r
3
R
Fig. 11-8. Inads and CompressiveStresses

Srfor SteelBelleville Washerswith VariousOutsideDiameters and h/t Ratios.
A
o.
o
o
ra)
o
!o
q,
c
v,
o
o
Lood At Flot - (lbf)
O
(n
F\
U)
e e'E
o
o
r
60 80 r00
200
400
When R = 2
6008001000
o
o
2000
4000 60008000
tt)
.tt
g E
Fi E.
8 f r
I
I
r 0I l
81
I
6l
sJ
100
80
41
60
3J
''rl
21
2
160
I
I .5.{
l"
l 2'
1 2 0 60 l-30
o
80 40 120
r.l
751rI
Il
).s.1
40 2 0 F r o
8
6'
o
5. E
4.o
J.
2.
30 r 5 F7.s
37J0.
). 1 . 5
I(
2s1
Metric Units
20
r ,4b- 1F,.'-
,4ffi
to-\/ t 32A (h/r)
19.2xlo? (h4)
Metric
E n g l l s h1 0 0
r0.
60 3 0 F 1 5
200
4
6
t
l
400 600800I 000
ffiAssogb?lfi
F$
3;"fu$.*ff
6000 r
L o o d A t F l o t-
II
I
( H ) W h e n R= 2
e
e
r : l /\@tg.z
E nx gto?(h/t
rish
0.5
e
c
c
c
c
1.0
e
e
e
e
e
e
e
a
e
e
a
e
e
a
e
a
a
a
a
a
a
.a
It
Bellevitle SPring Washers
F
F
!p
b
a
a
!|
-
a
a
Ch)ic! of Sress Level - Static

For static applications,stressat the convex inner cor:er S. us,uellvcontrols the spring set point. Carbon steel
3g{lrrlle u-asherswill start to set when stress (S") reachEs l,l[Fr of tensile strength (Table 1l-1). Set is removed
:r nilostbellevillewashers,and in this case stress(S.)can
:each T56c of tensile strength before additional set ocJ.rrs. These calculated stressesEueconsiderably higher
:hen actual stressesdue to yielding. If washers are to be
piarcd or operated at elevated temperatures,these values
must be reduced.
Cbob of Stress Level - Cyclic
For cy'clic applications, it is necessary to consider
both the stress level and stress range at the concave
conrcrs Srr and Sr2.Minimum and maximum stress must
be evaluatedat both Srr and Srzusing a modified Goodman diagam. The location with the more severe condiions u'ill control washer life. The modified Goodman
diagram (Figure 11-10) illustrates fatigue strength for

various thicknesses of carbon and alloy steel washersat
HRC 47 to 49. (Use of this diagram is discussedon page
27.) Shot peening increasesfatigue strength while burrs,
edge cracks and surface imperfections reduce it.
Stacks of Belleville Washers
To increasedeflection or loads, belleville washerscan
be used in series, parallel or a combination of seriesand
parallel (Figure 1l-11). Deflection for a series stack of
identical belleville washers is equal to the number of
washers multiplied by the deflection of one, while the
load is equal to the load of one washer. When belleville
washers with an h/t ratio greater than 1.3 are used in
a stack, the load deflection curve will be erratic as some
washerswill snap through the flat position. To avoid this
problem, the h/t ratio for each washer in a seriesstack
should not exceed 1.3.
Fig. 114. Load Deflection Characteristicsfor BellevilleWashers.

tf
ootl .{lo o/ouruolpolreo
!)
a
a
t
t
t
a
a
a
+ ,
0
o
CL
=
s
o
:,
. E r
o
o
o-
o
oE
o
!
o
o
t2.
o
3
s
:
o
a
plv
o
o
t
t
I
I
90
too
4
I
Deflecionino/ool f' lf OO
f
.:
o
o
o3
B
a
t
lf a *'asher is supported and loaded at its edges so that it is deflected

:c1'ond the flat position, then the greatest possible deflection can be
:ru-lizcd.Since the load/deflection curve beyond the horizontal position
:s slmmetrical with the first part of the curve, this chart has been
labeled at the right and top to be read upside down for deflections
beyond horizontal. Dotted lines extending beyond the chart indicate
continuation of curves beyond flat.
AssogFlf;S&H*ffirns
ffi

Load on a parallel stack of identical belleville washers
is equal to the load of one multiplied by the number of
washers, while deflection is equal to the deflection of
one washer.
The load deflection curve for both series and parallel
stacks has a hysteresis due to friction. Hysteresis (Figure 1l-I2) is greater for parallel than series stacks and
can be minimized by lubrication. The energy absorbed
by this hysteresishelpsto dampen vibrations. By careful
selection, stacks can be designed with increasing, approximately linear or decreasingrates. Stacked belleville
washers must be guided either over a pin or in a tube.
Hardness of the guides should be at least HRC 50 to
minimize wear. Clearancebetween washer holes and pin
of the relevant diameter.
or tube should be about 1,.5Vo
Fig. 11-10. Modtfted Goodman Diagram for Fatigue
and
Strength of Belleville Washers. Carbon
Alloy Steel at HRC 47-49 with Set Removed but Not
Shot-Peened.
Tolerances
To ensure proper clearance, it is good practice to
specify outside diameter with a minus-only tolerance and
inside diameter with a plus-only tolerance. Recortmended tolerancesare shown in Table ll-2. Load tolerances
should be specified at a test height. For belleville washers with h/t < 0.25, reqommendedload tolerances zue
=, l1Vo. For washerswith h/t > 0.25, use + lVo. The
recommendedtolerance for washers made of nonferrous
materials generally is + lsVo. Closer diameter and load
tolerances are available.
How to Specify
A checklistto aid the springdesignerin specifyingbelleville spring washersis shownon the next page. For washers with critical load requirements,it is recommendedthat
a test fixture be developed.
Fig. 11-11. Stacksof Belleville Washers.
ffiRR:
Lower Tensile Stress (103 psi)
r00
50
I 50
200
250
Combinolion of
Series ond Porollel
Fig. 11-12. Hysteresisin StackedBelleville Washers.

Deflection (in.)
o.
o
=
o
aA
o
v,
o
o
F
o
F
o
o
o
I
Nonferrous and
Austenitic Stainless Steel
Ten in Series
Test Five in Porollcl
C
Dcflcction
(mm)
C
Table11-2. Belleville WasherDiameter Tolerances.
Lll, m,(h-l
D*rrn*tcr, mnn{in.}
Carbon or Alloy Steel
up to 5 (0.197)
s-10(0.1924.3%)
tL25 (0.3,94-0.984)
2s-s0(0.98f 1.%9)
5L100(r.%9-3.937)
*0.ffi
-0.20(-0.00E)
-0.2s(-0.0r0)
-0.30(-0.012)
-0.40(-0.016)
-0.50(-0.020)
+0.20(+0.00E)
+0.25(+0.0r0)
+0.30(+0.012)
+0.40(+0.016)
+0.50(+0.020)
Based on R = 2, increased tolerances are required for lower R ratios.
ffiAssoget#Affieffiffis
C
C
C
C
C
C
C
Figurc| l - 10 maybereadas follols:

A bclleville washer0.8mm (0.030") thick maybe expectedto havea life of approximately106cycles
whenstrcssedbetweeneither
0-820MPa (0-I 17,000psi)
or 350-990MPa (50,000-141.000
psi)
or 700-l 170MPa (100.000-167,000
psi)
andmaybe expcctedto harrea lifc of approximarely
l0: cycleswhens!rcssed
bctwceneither
psi)
0-740MPa (0-105.000
psi)
or 3 15-890MPa (45.000-127,000
or 630-1050MPa (90,000-150.000psi)
Table 11-1. Maximum RecommendedStressLevels

S, for Belleville Washers in Static
Applications.
C o l c u l o t c dC u w e
--Tcst
Lower Tensile Stress (MPo)
C
C
C
3
xxx
C
C
C
C
C
C
C
C
5
C
C
I

B.{leritle Spring VYasberDesign Example
ln a clutch. a minimum pressureof 900 N (202 lbfl is
rcquired. This pressuremust be held as nearly constant
as possible*'hile the clutch facing wears down 0.80 mm
r0.031-r.The washer O.D. is 76 mm (2.99). Material
sclected for this application is carbon steel HRC 47-50.
l. Base the load on a value l0Vo above minimum load,
or 900+ lVo :990 N. Assume O.D./I.D. = 2.
From Figure 1l-9, select a load deflection curve
*'hich gives approximately constant load between 50
and l$Vo of deflection to flat. Choose the hlt : l.4t
curve.
2. From Figure 1l-9, the percent load at 50Vodeflection
to flat is 88%.
3. Flat load is Pr:990/0.88 : 1125N
(follow line A-B from 1125N to
4. Using Figure ll{
h/t = 1.41, and line B-C to approximately 76 mm
O.D.), estimatedstress is 1500MPa.
5. From Table11-1, maximum staticstresswithout setremovedis 120% of tensile strength. From Table l9-1,
page 103, tensile strengthat HRC 48 will be approximately 1650 MPa. Yield point without residual stress
will be 1650 x L.20 = 1980 MPa. Therefore,stress
1500MPa is lessthan maximumstressof 1980MPa.
'
I /-qDF
6. Stock thickne; s S
rS':
mv 132-4hh
8 . Referring to Figure 11-9, the load of 990 N will be

reached at fr = 5Wo of maximum available deflection.
fr : 0.50 x 1.93:0.97 mm deflection, or the load of
990 N will be reached at Hr : fl - fr = 3.30 - 0.97
: 2.33 mm height at load. To allow for wear, the
spring should be preloaded at Hz : Hr - Af (wear)
:2.33 - 0.80: 1.53mm height. This preloadcorrespondsto a deflection fz : H - Hz : 3.30 - 1.53:
1.77 mm. Then fzlh : 1.7711.93: 0.92or 92%.
9 . Because 92Voof h exceeds the recommended85Vo
(the load-deflection cunre is not reliable beyond 85Vo
deflection when a washer is compressedbetweenflat
surfaces), increase the deflection range to 4Vo to
85%. From Figure ll-9, the percent load at 40Vodeflection is78.5Voand Pr : 990 + 0.785 = l26L N. Repeat previous procedures4, 5, 6, 7 and 8, and find
that fz + h x 100: SlVo of h.
Final Desrgn Specifications:

Material: AISI 1074, 1075
O.D.: 76 +0.00,-0.5 mm (2.99 +0.00,-0.020)
I . D . : 3 8 + 0 . 0 4 0 , - 0 . 0 0m m ( 1 . 5 0 + 0 . 0 1 6 , - 0 . 0 0 )
Thickness t: 1.40 mm (0.055') Reference
Height h: 1.97 mm (0.078') Reference
Load: 990 N (223 lbf) + lVo at h1 : 1.18 mm (0.046')
Compressive S.: 1216MPa (185,000psi) at fz ESVoof h
Tensile Stress Srr: -203 MPa (-29,500 psi) at f2 85Vo
ofh
Tensile Stress Srz: +710 MPa (103,000psi) at fz 85%
ofh
7 . h = 1 . 4 1t = l . 4 l x 1 . 3 7: 1 . 9 3m m
H : h * t = 1 . 9 3+ L . 3 7= 3 . 3 0 m m
BELLEVILLE
WASHERSPECIFICATION
SPR,ING
CHECKLIST
(Fillin requireddoto only.)
Moteriol
Moximum operoting temperoture

*
To be used in o stock (type)
Working Conditions:
To work in
To work over
Lood
Testheight
"c("F)
mm(in.) diometer hole

diometer pin
N (1b0,I -N(lb0
Reloxotion
Requiredlife
mm(in.)
o/o
Reference Doto:
Thickness
mm(in.)
Outside Dio
Inside Diomefer
Free height
m(in.)
h/r
Assog&r#&H*ffitr$
ffi
SpecialSpring
Washers
ruffi
C
C
.
C
C
C
C
c
.
"@
e
e
c
t
a
SpecialSpring Washers
Introductiron
Special spring washersexert a thrust load and absorb
rrbration. reduce end play or apply pressure. The state
of stress is primarily bending, and most of the general
desienconsiderationsfor flat springs(Section 12) apply.
Spnng u'ashers are used in seals,bearings, motors and
o$er rotating mechanisms, and because of the trend
tosard miniaturization and compactness, demand for
tbem is increasing.
The Associated Spring SPEC product line contains
rnanv precision engineered wave, curved and finger
spring washers. These washersare made to close tolerancesand are available for immediate delivery. Selecting
SPEC washer designs savesdesigntime, avoids tooling
costs and is generally more cost-effective than specifying
custom-designedparts.
Cured Washers
Curved washers exert a relatively light thrust load and
are often used to absorb axial end play. Designersmust
provide space for diametral expansionin a direction perpendicular to the A dimension (Figure l3-1). Bearing
surfaces should be hard to prevent washer corners from
scraping or digging in. The spring rate is approximately
linear up to 80Voof available deflection. Beyond 80Vothe
rate increasesand is considerablyhigher than calculated.
Design equations for spring waJhers are similar to
those for simple beams, discussedin Section 12, except
for an empirical correction factor K. The equation for
load is:
o
@
^ = (oD)'K
(r3-r)
O.D. is outside diameter in the flat position and the

equation for stress is:
r:T*
Wave Washers
Wave spring washers,Figure 13-3,are especiallyuseful to apply moderate thrust loads when radial space is
limited. The rate is linear between 20 and 8Vo of available deflection. During forming, the washer is often
stretched at the crest and trough of the waves. Washers
that are round in the free position go out-of-round when
deflected. Generally, a ratio of D/b = 8 is a good balance betweenflexibility and load-carrying ability. When
the ratio of D/b is substantiallylower than 8, a belleville washeris preferred.
The number of wavesN" can be equal to 3 or more
and is usually selectedon the basis of desired spring
rate, since spring rate is proportional to the number of
wavesraised to the fourth power, as:
Ebt3N.oD"
ft=Plf=r/ot-,
(13-s)
This formula is basedon the equations for a simple beam

with correction factors based on experience to improve
accuracy. Stress is given by:
3zrPD
(t 3-4)
"s 4bttN.'
The outside diameterof the washer changesupon deflection and at flat is given by:
(I 3-s)
Do is outside diameter in the free position. The above
equations for load, stress and diametral change are not
exact solutions, but do provide useful engineeringestimates for design purposes.
(13-2)
Correctionfactor K is shown in Figure l3-2. Theseequations are approximate and yield satisfactory solutions
only for deflections up to 80Voof h where f is less than
li 3 of O.D. Associated Spring engineersshould be consultedwhen clearancesare critical or more exact designs
required.
Fig. 13-2. Empirical Stess Correction Factor K for

Curved Spring Washers.
4.0
3.0
Fig. 13-1. Typical Curved Spring Washer.
Y
o
o
e
.9
o
o
e,
2.5
2.0
\
\
\
2.O
2.5
3.0
3.5
4.0
RofioO.D./1.D.
At Flor
*Long
oxis of the wosher
in free position
Assog&'lfi3&ffi*ffiF$
@
Finger Washers
Finger washers,Figure 134, combine the flexibility of
curved washers and the distributed loading points of
wave washers. Load, deflection and stressare approximated by assuming that the fingers are cantilever
springs; then samplesare made and tested to prove the
design. Finger washers are used in static applications
such as applying an axial load to ball bearing races to
reduce vibration and noise.
Choice of Operating Stress - Static
Operating stressesrecommended for special spring
washersare similar to stresslevels recommendedfor flat
springs and are shown in Table l3-l as a percent of
tensile strength. Finger washers are generally produced
in the stress-relievedcondition. If favorable residual
stressesare required, consult AssociatedSpring.
Choice of Operating Stress - Cyclic
Maximum recommended operating stressesfor cyclic
conditions are shown in Table l3-2 for curved and wave
washers. Finger washersare not recommendedfor cyclic
applications.
Tolerances
Dimensional tolerances are similar to those on flat
springs. Load tolerancesdepend primarily on strip thickness tolerances and are listed in Table l3-3. All load
tolerances should be specified at a test heigttt and only
those dimensions critical to spring function should have
tolerances. Special tolerances are available for demanding applications.
Fig. 134. Typical Wave Spring Washer.
How to Specify
The specification checklist on the next pageis provided as a guide to all critical aspects of special spring
washers.
Special Spring lVasher Design Example
A wave washer is needed to go into a 80 mm (3.15')
bore and over a 60 mm (2.362') shaft, to support a load
of approximately 500 to 550 N (112 to 124 lb) with 1.8
mm (0.071')deflection. The applicationrequires a steady
load and is therefore a static application. The washer will
operate in an ambient environment. AISI 1075 is the
preferred material.
Since deflection is comparatively large for a spring of
this type, select the most flexible design - a three wave
configuration.
Assumea 75 mm (2.953") outside diameter and a 64
mm (2.520") inside diameter to fit the given conditions.
This would make the mean diameter (D) 69.5 mm
(2.736").
1. Substituting these values in the load-deflection equation, solve for thickness:
,: \
Y(207,000)
(1.s)(5.t (3r0t
= l . 3 0m m
&
P,=
D,
= 530N
2.4 D3
Fig. 134. Typical Finger Spring Washer.
'-ll'-
c
C
C
C
C
C
C
C
C
C
C
c
C
C
h\-
+:+.:+,*il.b
C
C
C
h = H - t
Table13-1. Maximum Recommended Operating S/ress

Levels for Special Spring Washersin Static
Applications.
C
C
C
Table13-2. Maximum RecommendedOperating Stress

Levels for Steel Curved and Wave Washers
in Cyclic Applications.
C
C
C
Pcreent of T,mih Strerylh
c
Finger washers are not generally supplied with favorable residual stresses.
H#Assos&tfi8
/fu H#,ffirp$
This information is based on the following conditions: ambient environment, free from sharp bends, burrs, and other stress concentrations.
AISI 1075
C
I
l. Set the maximum stress at solid at 8Vo of tensile
suenglh. Steel with a hardnessof HRC 49 has a tensile strength of 1725MPa (250,000psi) (Table l3-1 and
Figure 3-6, page 2l). \Vo of 1725MPa is 1380MPa.
Solve for deflection at that stress. Using the equation:
pPsD _
fr r_= fe.6
fi-
(9.6x69.t'(1380x64)
Deflection to load of 1.8 mm is 75% of deflectionto

solid, which is satisfactory.Diameter in the deflected
position:
Do'=ffi:
Do': 75.1 mm
There is adequate clearance.
f, = 2.39mm
finat Dexf,gnSpecifications:
Tablc13-3. Load Tolerancesfor Special Spring
Washers.
0.r-1.0 (0.0044.03e)
0.1-0.25(0.004{.010)
0.2il.30 (0.0104.012)
0.H.5 (0.0124.020)
0.5-1.0 (0.020{.039).
1.G2.0 (0.039-0.079)
33
25
20
15
t2
Material: AISI 1075

O.D.: 75 *,0.2 mm (2.953+ 0.008)
I.D.: S,!,+ 0.2 mm (2.520<- 0.008)
Thickness t: 1.30 mm (0.055'r 0.002)
H: 3.69 mm (0.145') Reference
Load P1: 530 N + lTVo(119 lbf + l7Vo)
H,: 1.89 mm (0.074')
SPECIAL
SPRINGWASHERSPECIFICATION
CHECKLIST
(Fill in required doto only.)
Speciol
Type of Wosher:
Mqximum operoting temperoture

Operoting environmeni
Curved
wove
finger
designer
recommendotion
Reference
Moteriol:
Working Conditions:
Thickness
To work in
To work over rm(in.)
I Lood
,"c('F)
mm(in.) diometer hole

diometer pin
Outside diometer
Inside diom
mm(in.)
N(lb0+ -N(lbo
i Test height
mm(in.)
Describeone cycle
Assog&'ifi8/&,H*ffis
ffi

122spec Design

Uploaded by

Document Information

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

122spec Design

Uploaded by

Copyright:

Available Formats

Design

Eor to Use the Handbook

Selccdng Spring Configurations

Helical Torsion Springs

Specid Spring Washers

Constant Force Springs

Push-wide load and deflection range-constant

Push-wide load and deflection range. Conical

Helical Torsion (Section9)

Twist---exerts torque over

Push-light loads, low

Ptrsh-higher defl ection s

hrsh-for axial loading of

Push-used to absorb axial

Twist--+xerts torque over

Push-may have an inherently high friction damping.

Twist+xerts close-toconstant torque over many

Tahle 3-1 Typical Properties of Common Spring Materials.

Carbon Steel Wires:

Alloy Steel Wires:

(10.) 7.92 (0.286)

0.13(0.00s) 9.5 (0.375) b

(10.4) E.03 (0.290)

Copper Base Alloy Wires:

0.0E(0.003) 1.5 (0.063) b

Nickel Base Allovs:

(e.7) 8.14 (0.2%)

(l r.5) 7.E6 (0.2E4)

(6.3) E.16 (0.320) l 5

b. Maximum defect depth: l.$Vo of d or t.

can also be important and must be consideredin light

Table3-2. RelatedSpringMaterial Specifications.

CopperBeryllium cA-t72 Bt97

42E. 14/19,Parrt38, t7222,

PhosphorBronze cA-521 8r03,

*Valve spring quality.

3-3. Magnetic Characteristicsof Some Materials.

' \ itrogen-strengthenedmanganesestainless steels.

Tabb 34. Typical Heat Treatments for Springs After

Tempering is an effective stress-relieving treatment and

Fig. 3-2. Relaxation versus Temperaturefor Spring Materials.

Exposureof 100 hours ot l49oC (300"F)

ond Heot set

temperatures for many commonly used materials. It

Tahle 3 -5. Guidefor SeIecting Minimum Thicknessesfor

Zinc and Cadmium Coatings.

Zixon tranerd Srnel

(l) Requirements for zinc coating (electrodeposited).

Mechanicalplating provides an effectivemeansof zinc

Most noble (-),

HASTELLOY is a registeredtrademark of Cabot Corporation.

Type 302, A232 and A230 materials have slightly lower

SaIt Spray Resistance of Common Barrier

Stsnderd SaIt Spray

This information is basedon laboratory-controlled applicationsand test.

M*trk Sillri {ftm,

Patentedand Cold Drawn ASTM A227

Table34. Ranking of Relative Costs of CommonSpring

Table3-10. Standard Tolerancesfor Spring Wire.

0.514.71 (0.020-0.028) 0.olo(o.ooo4)

Most spring wires can be purchased to tighter tolerances. Music wire

Hardness levels above HRC 50 result in high strength

Fig. 3-5. orientation of Bend Axis to Rollins Direction

Also known os ocross the groin,