Cable Television Sag Calculations PDF
Cable Television Sag Calculations PDF
Cable Television Sag Calculations PDF
ABSTRACT Figure 1.
Understanding the mechanics of aerial cable installation
is essential to the cable engineer for proper plant design L
and for maximum cable plant reliability and longevity. A Sx
key part of that understanding is the calculation of sags S X
and tensions. Although basic sag and tension equations H
T WL
are available, little is available on calculating sag and
2
tension with changing temperature and load. H
This paper extends the basic sag and tension equations WL2
H = (1)
to address these effects, and gives the ability to solve 8S
integrally supported (Figure 8) cable, tightly lashed cable,
8S2
and unequal elevation problems. The equations are LC = L + (2)
applied to several basic tension and clearance require- 3L
ments of the 1993 National Electrical Safety Code
(NESC) which can only be addressed through calcula- Where:
tion. They are also used to analyze expansion loop life, H = Horizontal component of tension (pounds)
tight lashing, cable buckling, and center conductor
W = Linear weight of the wire (pounds/foot)
pullouts. Obscure cable failure modes caused by wind
gusting, solar heating and radiative cooling are discussed. L = Distance between supports (feet)
S = Sag (feet)
INTRODUCTION
LC = Actual length of wire (feet)
The calculation of sag and tension can be quite useful to
the cable television engineer. The calculations are used
to select the appropriate size support strand for a given These equations1 are a parabolic approximation of the
application or to determine if clearance requirements are actual form that a flexible wire assumes, which is a
met. With today’s clearance requirements the only way catenary. The equation for a catenary is:
to assure clearance requirements are met is by calcula-
tion. The calculations serve as an aid to the cable designers
who determine the mechanical stress that the cable must
be capable of withstanding or to evaluate such problems
S =
( )[ ( ) ]
H
W
cosh
WL
2H
-1 (3)
Lc = L +
( )[( ) ( )]
4
3
S12
L1
+
S22
L2
(10)
To determine the final sag and tension a mathematical
description of these events will be given.
Unstressed Length
Where:
The most important quantity in the process of determining
T = Tension in the direction of the cable (pounds)
the final sag and tension is the cable’s initial unstressed
With these equations 1 (with Eqs. 6 and 7 slightly modi- length at the initial temperature. The cable’s unstressed
fied from the referenced source) the initial sag or tension length is the length of the cable if it had no tension or stress
can be calculated. Of course, one or the other must be on it. It should not be confused with LC, the actual cable
known and span length and cable weight must also be length.
given. They can be used, for instance, to calculate clear-
The unstressed cable length is necessary to know be-
ance over roads and sidewalks, and to determine specific
cause, as additional tension is applied to the cable, its
size strand to be used under worst case loading. Note,
stressed length changes as a function of its unstressed The unstressed length, then is:
length (and also its elastic modulus). To determine the
appropriate amount of expansion and contraction due to LC
temperature changes, again the unstressed length must Lu0 = (16)
be used (along with its thermal coefficient of linear
expansion).
The stressed length of the cable (Lc ) can be determined
1+
( )H
AE
FL
= L+ (14) It should be noted that the thermal coefficient is not
AE
necessarily constant and again it is necessary to be
LC = L
( 1+
F
AE ) (15) familiar with the properties of the materials.
Most materials have positive thermal coefficients of lin-
ear expansion which cause them to get longer when the
It is assumed that the tension is the same along the entire temperature increases. When the temperature decreases,
length, but, because the wire has a finite weight per unit they get shorter. (In some cases, materials are specially
length, the ends of the wire near the support have more selected because they have negative expansion coeffi-
tension. The horizontal component of the tension is the cients so that when used in conjunction with other
same everywhere in the wire. The vertical component of materials the overall change in length is minimized or
the tension is zero at the lowest point since there is zero matched to some other component such as in fiber optic
cable weight. The vertical component of tension at the transmission lines.)
support is the weight of the cable between the support
and the lowest point. There are several factors that affect temperature. The
temperature of the cable or wire is of course affected by
the air temperature. The air temperature changes on a
daily basis (diurnal) as well as yearly basis (annual). If
electrical current is carried through the wire, a tempera-
ture rise will also occur, but, generally speaking, this rise As an example, the vertical weight components may
is small for CATV cable. Another factor is radiation. include: the weight of the support strand, the weight of
During the day the cable is heated from the sun. At night all the cables, the weight of the lashing and possibly a
the cable radiates its heat toward clear skies. The wind cylinder of ice around the group. The horizontal force is
tends to minimize this effect. only attributed to wind loading. In some cases a weight
Measurements made in Connecticut on the longest day constant (e.g., in heavy loading districts, 0.3 pound/foot)
of the year when the sun’s rays are most direct indicate is added to the resultant final weight for safety.
that a temperature rise of about 45°F above ambient can Ice Loading
be expected on black jacketed cable and about 24°F rise
above ambient for unjacketed aluminum sheathed co- The additional weight caused by ice build-up on the cable
axial cable. At night, in the same area, the temperature is usually calculated based on a hollow cylinder having
of black jacketed cable was about 8°F below ambient an inside diameter equal to the outside diameter of the
and 4°F below ambient for unjacketed aluminum cable. bundle of the cables and the support strand and with a
given thickness which is usually 0.25 inches or 0.5 inches
LOAD CALCULATIONS depending on which loading district is used. Using 57
pounds per cubic foot as the weight of ice3, the linear ice
Before the final sag and tension can be calculated, it is
weight around the cable can be calculated from:
necessary to evaluate the span of wire in terms of load.
Initially we assumed that the wire’s weight was the only Wice = 1.244 ( Dit + t 2 ) (19)
load applied. Under worst case conditions, the wire may Where:
have ice formed around its circumference and at the
same time a strong wind blowing on it. The total load on Wice = Linear ice weight (pounds/foot)
the wire is the resultant of all vertical and horizontal Di = Diameter over cable bundle (inches)
loads. (The horizontal load here is perpendicular to both
t = Thickness of the ice (inches)
the vertical and the wire itself.)
Wind Loading
Figure 3.
The wind loading on a circular surface can be calculated
WEIGHT from3:
P = 0.002 56V2 (20)
Where:
P = Horizontal wind pressure (pounds/square foot)
V = Wind velocity (miles/hour)
In order to use this equation, the pressure must be
converted to a load or force on the projected surface of
the bundle that faces the wind. Assuming that the wind
The total resultant load of the wire is the vector sum of
is perpendicular to the cable and we wish the results to
all the horizontal and vertical components:
be in terms of pounds/foot:
It should also be pointed out that horizontal loading From Eqs. 5, 6, and 7, S 1 and S2 can be set in terms of
results in sags which are not vertically directed. The S as:
horizontal and vertical components of the final sag can
be resolved since they are proportional to the horizontal SL 12
and vertical loads described in Eq. 18, i.e., S1 = (27)
L2
Sf = ( SV2 + S H2 ) 1/2 (22) SL 22
So, for example the actual vertical component of the sag S2 = (28)
L2
would be:
So Eq. 26 can be transformed to:
Sf ( ΣWV )2
SV = (23)
( ΣWV )2 + ( ΣWH )2
Lcf = Luf
( 1+
Hf
) (24)
( )
AE h
L2 = L 1+ (31)
4S
Where:
Lcf = Final stressed length of the wire (feet) After a bit of work we find:
Luf = Final unstressed length of the wire (feet)
Hf = Final horizontal tension (pounds)
A = Cross sectional area of the wire (square inches)
L13 + L23= L3
( 2+
3h2
8S2 ) (32)
E = Elastic modulus, tensile (pounds/square inches) which can be plugged into Eq. 29 to yield:
If the supports are at equal elevations, the final length
of the wire is also (from Eq. 2):
Lcf = L+
8Sf2
(25)
Lcf =
L+ ( )( )
8S f 2
3L
h2
2L
3L
Where:
Back to the problem at hand, we were in search of
Sf = Final sag (feet) another equation to set equal to Eq. 24; our search is over
This is actually a special case of supports at unequal with Eq. 33. Actually Eq. 24 would be better in terms of
elevations. The following progression puts Eq. 10 into S f , like Eq. 33. This is easily done via Eq. 1.
the same form as Eq. 25. Consider:
LC = L +
( )[( ) ( )]
4
3
S12
L1
+
S22
L2
(26) [ ( )]
Lcf = Luf 1 +
Wf L 2
8S f AE
(34)
If
()() a
+
-b
<0 (42)
( )( ) ( )
8S f 2 h2 LufWf L2 3 2
L+ + - Luf - =0
3L 2L 8S f AE
Then
( )
{ [ ]}
-b
(35) 1/2
() ()
-a 1 2
Sf = 2 cos cos-1
3 3 3/2
()
Which can be put into the form of: -a
3
Sf3 + Sf
( )( 3L
8
L+
h2
2L
- Luf -
) ( )( )
3L
8
LufWf L2
8AE
=0 (43)
[ () ]
a = 3 L2 +
h2
2
- LL uf (38)
CLEARANCE CALCULATIONS
Sometimes when checking to assure that the proper
clearances are met (e.g., during make ready), it is nec-
8 essary to know the elevation of the wire at points other
than at the lowest point. For equal elevations Eq. 4 can
- 3W f L 3L uf be used. The sag at any point is given with respect to the
b= (39)
64AE lowest point which, for problems of equal elevation, is
always exactly half-way between the supports.
The solution of this particular form of cubic equation is For problems of unequal elevation, it is not particularly
as follows4: helpful to know what the sag is with respect to the
3 2 position of the lowest point sag because not only is the
If
()() a
3
+
-b
2
≥0 (40) lowest point sag not exactly half-way between the two
supports, it may not be (mathematically) between the
two supports at all!
Then Equation 4, however, can be adapted.
Figure 4.
3 2 1/2 1/3
Sf =
{( ) [( ) ( ) ] }
-b
2
+
a
3
+
-b
2
L2
Sy
3 2 1/2 1/3
{( ) [( ) ( ) ] }
-b a -b S2
+ - + (41)
2 3 2 X Y
From the above figure (which is simply Fig. 2 without during the course of time the strand is exposed to such
some of the details, but includes some new information stress that would cause 0.09 percent creep. The implica-
about the sag at any point) it can be seen that x can be tion can be approximated from Eq. 2. The resulting sag
stated in terms of y, which is the horizontal distance would be 2.09 feet instead of returning to 1 foot as might
from the highest support as: be expected. Although the strand should not be exposed
to such stresses, the purpose of this illustration is to show
x=
( ) L2
2
-y (45)
the limitations of the calculations and highlight the need
for caution and good engineering judgement.
Where:
SPECIAL CONFIGURATIONS
x = Distance from the lowest point (feet) Multiple Elements
y = Distance from the highest point (feet) In some cases the wire may be comprised of several
different materials with different cross sections. Cable
L2 = Span length (feet)
with integral messengers (“Figure 8”) and drop cable are
S2 = Sag at lowest point (feet) examples. Each material has its own expansion coeffi-
Substitution Eq. 45 into Eq. 4 we obtain: cient and elastic modulus. When suspended, each element
takes a different portion of the total load. The amount of
{ }
2
[( ) ]
L2 the load that any particular cable element takes is a
4 -y function of its elastic modulus and area as compared to
2
Sy= S2 1- (46) the total resultant modulus of all the materials and the
total area of all the materials.
L22
The resultant modulus can be found by multiplying the
modulus of each material by the ratio of the material’s
Where: area to the total area:
Sy = Sag at y from the highest support (feet)
Equation 46 can also be used for equal elevation prob-
lems. Notice from Eq. 9 that L 2 is not fixed. L2 will be
different from its original value depending upon the
Er =
() ()
A1
AT
E1 +
A2
AT
E2 +
The resultant coefficient of linear expansion (ar ) is cal- And the tension on element n is:
culated as follows:
AnEn ( Lcf - Lufn )
Hn = (52)
(A1E1a1 + A 2E2a2 + .... + AnE nan) Lufn
ar = (48)
(A1E1 + A2E2 + .... + AnEn)
Where:
By using these values for Er, Ar, and ar the final sag and Hn = Tension on element n (pounds)
final tension can be determined at the new temperature
with the new load. An = Cross sectional area of element n (square
inches)
Because the above equations which include elastic
modulus only work when the materials are stressed En = Elastic modulus of element n (pounds/square
within their elastic region, it is important to make sure inch)
that each element’s elastic limit is not exceeded. Lcf = Final cable length (feet)
The unstressed length of all the elements is the same Lufn = Unstressed final length of element n (feet)
only at the original temperature. It is assumed that any
stress built into the cable is small compared to the
installed stress. Once the temperature changes, each The stress in Eq. 51 should be tested to assure that the
element expands and contracts at its own rate. Also, each elastic limit not be exceeded. Aside from exceeding the
element has its own elastic modulus and cross sectional elastic limit of any particular element, one additional
area. Unless the elements are bonded together in some caution should be noted. Generally, “Figure 8” cable
fashion they may move independently of one another. used in CATV applications is for distribution purposes
and so taps are installed. To install a tap the cable must
In order to determine the stress on any particular ele- be cut and separated from the messenger wire. Cutting
ment, the unstressed final length must first be deter- the cable relieves the stress in the cable at the point
mined. It can be calculated, based on the change in where it is cut. This tension, however, does not
temperature and the expansion coefficient of the element necessarily disappear. A significant portion, if not all
in question, as follows: of it, is diverted to the steel messenger wire. So,
Lufn = Luon [ 1 + an ( T f - To ) ] (49) wherever taps are installed in messenger cable, you
must as sure that the
Where: total tension does not exceed the safe limits of the
Lufn = Final unstressed length of any particular messenger as if it were alone.
element n (feet)
Cable Tightly Lashed
Luon = Original unstressed length of any particular
In cases where cable is lashed to the strand so tight as
element n (feet)
to restrict cable from moving independently of the steel
a n = thermal coefficient of linear expansion of any strand support, a similar analysis can be performed. This
particular element n ( inch/inch / °F ) condition can occur even if the lashing is not restrictive:
Note: Luon is the same for all elements and exactly the by having no expansion loops.
same as that given in Eq. 16. We will call this third case “cable restrained”. In this
The final cable length (under stress) is the same for all case, the steel strand is installed and loaded with the
elements. It can be determined from either Eq. 2 or Eq. cable. The cable is then lashed so tight that cable move-
10. The strain that the cable element is under is then: ment is restricted. The resultant condition is that the
Lcf - Lufn steel strand is under stress and the cable has zero stress.
en = (50) After the temperature, load or both change, the load is
Lufn
not distributed like it was in the case of “Figure 8” cable.
The stress is the modulus times the strain: They differ in that there is no original cable tension; so
En ( Lcf - Lufn ) the original unstressed cable length is the same as the
σn = (51) actual cable length ( Lco ). The original unstressed steel
Lufn length is slightly shorter.
( )
H
1+ Lufn = Lco [1 + an ( Tf - To ) ] (59)
ATEr
Where the subscript “n” refers to one of the cable ele-
Er is found in the same way as for Figure 8 cable, Eq. 47. ments.
The tension, H, in Eq. 53 is the tension of the steel The stress on that particular element is then:
support alone. The original unstressed length can be
used to find the final unstressed length at the final En ( L cf - L ufn)
temperature from Eq. 17 except that the resultant expan- σn = (60)
Lufn
sion coefficient (ar ) should be used in place of a. Finally,
the final sag can be calculated from Eq. 41 or 43. The total And the tension on that element is:
tension on the assembly can then be found from Eq. 44.
The original tension on the components was easily de- H f n = A nσn (61)
termined; the steel had all the tension, the cable had
none. At final state, that is, after the load, temperature Unfortunately, it is difficult to first assume full load at the
or both change, the tension on each component may not lowest temperature for restrained cable. For example,
be quite so easy to determine. We will divide the tensions you may want the tension on the strand to be 60 percent
into two groups: namely, the tension on the steel and the of its breaking strength at full load at the lowest tempera-
tension on the other components. ture. The tension on the other components can be
anywhere between zero and the material’s elastic limit,
To analyze the final tension on the steel, we must first depending on the original temperature.
determine its initial unstressed length. It is:
In the other two cases, the original temperature and load The unstressed length of the steel support strand is
have no restrictions. But in this case, cable restrained, found from Eq. 16 and the elastic modulus of the strand
the original conditions must be at the temperature at from Table 2 and 3.
which the cable was lashed to the strand. Of course, if
enough information is known about the final conditions, 125.075 0
the initial conditions can be determined.
( )
Lu = 457.3
In review, three basic configurations have been dis- 1+
( 0.035 185 ) ( 28 x 106 )
cussed with regard to sag and tension. The first was a
simple case of a single wire. The second considered
multiple elements which can be applied to “Figure 8” type = 125.016 97 feet
cable or drop wire. The last, “cable restrained”, ad-
dressed what happens when the cable has zero tension
initially but supports some of the load as conditions Now the unstressed length at any other temperature
change. Each configuration was considered after the can be determined. From Table 4, the temperature for
temperature, load or both changed. a heavy loading district is 0°F. From Eq. 17 and the
expansion coefficient of steel in Table 2, the unstressed
APPLICATIONS length of the steel at 0°F is:
Strand Tension Under Worst Case Load Luf = ( 125.016 97 ) [ 1 + 7.2 x 10 -6 ( 0 - 60 ) ]
To illustrate the application of these equations a few = 124.962 96 feet
examples will be given. Consider, first, a 125 foot span
having a 500 and 750 size jacketed cables supported by Assuming that the cables are stacked one on top of the
a 0.25 inch EHS strand. Assume that the cable has other and from the dimensions in Tables 1 and 3, the
expansion loops and is free to move independently of the width of the cable plus the strand is 1.620 inches. From
steel support strand so that there is no tension on the Eq. 19, 0.5 inch thick ice over the cable weighs 1.319
cable under any condition. The cable was installed at pound/foot. So the total vertical weight (i.e., ice plus the
60°F and has a 1.5 percent sag (1.875 feet). The cable cable) is 1.757 pound per foot. See Figure 5.
is installed in a heavy loading district with extreme wind
loading of 21 pounds/square foot. The question is, what Figure 5.
is the tension on the strand under worst case conditions? 1/4 IN. EHS STRAND 500 J
0.240 IN. OD 0.560 IN. OD
From Table 1 and 3, the total weight of the strand, plus
the 500J, plus the 750J, plus the double lashing is 0.439 LASHING WIRE 750 J
pound/foot. The maximum tension on the strand must 0.045 IN. 0.820 IN. OD
not exceed 60 percent of its rated break strength under
worse case expected loading. The initial strand tension
and cable length at 60°F can be calculated from Equa-
tions 1 and 2.
= 457.3 pounds
8 ( 1.875 feet)2
Lc = 125 feet +
3 (125 feet)
The horizontal load from the wind is calculated from This is the tension (horizontally directed) in the strand
Eq. 22. The projected diameter over the ice is 2.620 under heavy loading conditions. This is well within the
inches (i.e., 2 x 0.5 inch + 1.620 inches). The wind 3990 pound maximum limit.
pressure for a heavy loading district, from Table 4, is 4 One more check is necessary and that is under extreme
pounds/square foot. The loading is then: wind loading at 60°F. From Figure A2 the wind loading
is 21 pounds/square foot. The resulting wind load from
(4) (2.620) Eq. 22 is:
Fwind =
12
(21) (1.620)
= 0.873 pound/foot Fwind =
12
The vector sum of the horizontal and vertical loads
calculated from Eq. 18 is: = 2.835 pounds/foot
W = [ (1.757)2 + (0.873)2 ]1/2 By adding the horizontal and vertical load components,
= 1.963 pounds/foot the final load is 2.869 pounds/foot. Using the unstressed
length at 60°F:
The weight adder for a heavy loading district is 0.30
3
()
pounds/foot and results in a final weight of: a
Wf = 1.963 + 0.30 = - 0.018 645 2
3
= 2.263 pounds/foot
2
()
To solve for the final sag in Eq. 37, a and b are first found -b
from Eq. 38 and 39. The difference in support elevations, = 277.746
2
h, is zero so:
3 [ (125)2 - (125) (124.962 96) ] The final sag and tension is found in the same way as
a = before.
8
Sf = 3.30 feet
= 1.736 088 75 H = 1698 pounds
So the strand tension under extreme wind loading is well
-3 (2.263) (125)3 (124.962 96) under the 3990 pound limit.
b=
64 (0.035 185) (28 x 106) The above example was carried out in several steps:
3
2. Calculate the initial cable length from Eq. 2.
() a
3
= 0.193 80 3. Calculate the initial unstressed cable length from Eq. 16.
4. Calculate the final unstressed cable length from Eq. 17.
2
5. Calculate the final load from the loading tables and Eqs.
() -b
2
= 172.604
18 and 19.
6. Calculate the final sag from either Eq. 41 or 43.
7. Calculate the final tension from Eq. 44.
Clearly, Eq. 40 is true and the final sag (which is not
vertically directed) can be found from Eq. 41 and the final For unequal elevations problems, LC is calculated from
tension from Eq. 1. Eq. 10 and h is used in Eq. 38. These equations can be
loaded into a computer or programmable calculator for
Sf = 2.779 feet easy calculations.
(2.263) (125)2
H =
8 (2.779)
= 1590 pounds
Clearance Over a Traffic Light Assuming that the worst case is at 32°F with 0.5 inch ice,
no wind and a 0.3 pound/foot weight constant, the fol-
Consider, next, a situation where you wish to keep a 30
lowing results are obtained:
inch clearance, under all expected conditions, over a
traffic light. If the steel support is installed so that it is Wf = 2.057 6 pounds/foot
loaded to about 5 percent of its rated strength will the Luf = 150.168 274 feet
proper clearances be met? Figure 6 shows a sketch of
Hf = 1280 pounds
the span.
Sy = 4.02 feet
Figure 6.
Sf = 4.52 feet
50 FT. Assuming Heavy Load Conditions:
150 FT.
Wf = 2.263 pounds/foot
Hf = 1431 pounds
Sf = 4.45 feet
TRAFFIC 30 FT. Assuming 120°F with no load:
LIGHT
21 FT.
Wf = 0.439 pound/foot
Hf = 296 pounds
Sf = 4.17 feet
The following information is available. For heavy loading, 1431 pounds is about 22 percent of
Support Strand: the strand’s rated strength of 6650 pounds. The expected
0.25 inch EHS creep is approximately 0.015 percent. Doubling it for
margin, 0.03 percent,
E = 28 x 106 pounds per square inch
Lcf = 150.363 381 feet
A = 0.035 185 square inches
This is a first order approximation of the increased sag
a = 7.2 x 10-6 inch/inch °F over time due to creep. The sag in the middle of the span
Cable: would be about 4.783 feet instead of 4.52 feet. As long
1 - 750J as the attachments are: 21 + 2.5 + 5 = 28.5 feet high, there
should not be any clearance problem.
1 - 500J
Double Lashed Tension on Composite Materials
Diameter = 1.620 inches Consider a 150 foot span at 60°F with 0.5 percent sag
(0.75 foot). In this example let us assume that a single
Weight = 0.439 pounds/foot 750 non-jacketed cable is either tightly lashed to the
The cable is installed in a heavy loading district with strand or has no expansion loops so that it can not move
16 pounds/square foot extreme wind. independently of the 0.25 inch EHS support strand.
The initial (loaded) conditions from Eq. 1, 2, 16 and L = 150 feet
46 are: S = 0.75 feet
H = 330 pounds W = 0.299 pounds/foot
Lu = 150.198 554 feet
S = 3.74 feet
Sy = 3.33 feet at 50 feet from the support.
A E a
Component Material (square inch) (pounds/square inch) (inches/inches °F)
The resultant is calculated from Eqs. 47 & 48. The Notice that the cable is under compressive forces.
conditions at 60°F are: The force (F) required to cause the aluminum to buckle
S = 0.75 feet under compressive forces is:
Hsteel = 1,121 pounds
π 2 EI
Htotal = 1,121 pounds F=
X2
Halum = 0 pounds
Assume the temperature drops to -40°F. Where:
Sf = 0.27 feet I = Moment of inertia
Htotal = 3,129 pounds X = Unsupported distance
The moment of inertia, I, for 750 cable is approximately
Stress 0.005 inch4 so the maximum unsupported distance that
E Tension (pounds/ the cable can be exposed to without buckling is about 34
(pounds) square inch)
inches. Although the lashing wire spacing is much less
HSteel = 1,775 Eq. 57 50,448 than 34 inches, it is quite conceivable that, at the pole,
the cable will be unsupported for over 34 inches and
HShield = 980 Eq. 61 12,136
cable buckling is imminent.
HDiel = 111 Eq. 61 327
HCenter = 264 Eq. 61 12,348 Long Spans - Cable Movement and
Expansion Loops
Consider a long span, approximately 200 feet of 750
cable, with 1.5 percent sag at 60°F. Suppose that the
The stresses on the shield and center conductor have cable has expansion loops and the cable can move freely
exceeded their elastic limits. This translates to excessive as the temperature changes. Consider then that the
stress on the conductors. This span will be prone to temperature changes from 50°F at night to 110°F during
center conductor pullout problems and stress fatigue. the day. What is the differential cable movement? This
Now assume instead that the cable was installed on a cold much temperature change is not unreasonable to as-
day, say at 40°F, and during the summer the cable sume since solar heating and radiative cooling can account
temperature rose to 120°F. The initial sag and tension for a 28F temperature change on unjacketed cable even
are the same as before; the final conditions are: if the ambient temperature does not change.
Sf = 2.14 feet
HSteel = 1023 pounds
HTotal = 393 pounds
HShield = -435 pounds
HDiel = -82 pounds
HCen = -113 pounds
Expansion Loop Life For the specific case given above where the loop is
The life of an expansion loop can vary significantly de- exposed to 1.4 inch excursions, the life of the 750 cable
pending on a number of factors: the surface finish of the loop is about 8 years. Therefore, even with modest tem-
aluminum, the type of loop, the depth of the loop, the size perature changes, a single expansion loop may not be
of the cable, the wall thickness of the aluminum outer adequate for long spans.
conductor, whether it’s jacketed or not, and the excur-
sion distance per cycle. Without going into expansion Wind Gusting
loops to any great extent, the following expansion loop Aside from length changes due to temperature changes,
lifes are typical of semiflexible cable with measured 6 changes can also occur due to load changes. Although
inch depth and total 1 in excursion (i.e., +0.5 inch and load changes due to ice build-up are infrequent, perhaps
-0.5 inch from the neutral point). The expansion loops in a few times a year, load changes due to wind can occur
0.500 inch were formed with a LEMCO G120 and the quite frequently especially if gusting is considered.
1.000 inch with a G240. The frequency and amplitude distribution of wind gust-
Cycles to Outer Conductor Fracture ing is complicated and can vary significantly from one
(1 Inch Excursion) region to the next. In general, gusting is far more severe
in urban areas with buildings than in flat, level country.
Although there is some good data available on wind
Type Unjacketed Jacketed gusting, the next example simplifies the frequency and
amplitude distribution of gusting to a single data set.
0.500 29,900 51,800 Consider a 150 feet span of cable, one 500J and one 750J
0.750 18,000 22,500 cable, suspended between supports with 1.5 percent sag
1.000 8,000 17,900 at 60°F in a town with buildings. Consider next that the
distribution of wind gusts for a one year period can be
represented by 1000 gusts from 20 miles/hour to 60
The life of other cable sizes can be roughly approximated miles/hour. The following table can be generated:
by interpolation. If each cycle is equivalent to one day,
then the life of 0.750 inch jacketed cable would be about Wind Speed Wind Force Cable
49 years. Measurements show that as you double the (miles per (pounds per Length
excursion distance, the life of the loop drops by a factor hour) square foot) (feet)
for jacketed cables and by a factor of about 20 for
unjacketed cables. The depth of the loop is extremely 0 0 150.090 0
important. For a 0.500 inch unjacketed cable with a 20 1.024 150.092 7
3-1/2 inch depth instead of 6 inches, the aluminum 60 9.216 150.182 7
fractures at 1,500 cycles instead of 29,900 cycles.
So, the change in length that the expansion loop must REFERENCES
accommodate is 1.08 inches. Assuming that there are 365
1 Edward C. Jordan, Ed. in Chief., Reference Data for
one inch temperature cycles per year and 1000 inch gust
Engineers: Radio, Electronics, Computer, and Commu-
cycles per year, then the life of the loop will decrease by
nications, 7th ed. (Indianapolis: Howard Sams, 1985),
a factor of 0.27 as compared to its life due to temperature
pp. 48-19-20.
changes. For the 0.750 inch cable mentioned above, the
life of the loop should be derated from about 49 years to 2 Donald G. Fink, Ed. in Chief, H. Wayne Beaty Assoc.
about 13 years. Ed., Standard Handbook for Electrical Engineers ,
11th ed., (New York: McGraw Hill Book Co., 1978),
CONCLUSION pp.14-64.
The basic sag and tension equations have been expanded 3 National Electrical Safety Code, 1990 Edition. (New
to account for changes in temperature and load. Special York: The Institute of Electrical and Electronic Engi-
configurations, such as tight lashing and “Figure 8” type neers, Inc. 1989), pp. 276-81.
cable with composite materials, were also addressed. 4 Murray R. Spiegel, Mathematical Handbook of Formu-
The application of these equations can help the cable las and Tables, Schaum’s Outline Series (New York:
engineer respond to the local utility about maximum McGraw Hill, 1968), p.32.
tension and proper clearance of aerial plant. The equa-
tions can also be applied to help understand various 5 Eric Winston, “Stress Analysis of CATV Aluminum-
failure modes such as center conductor pullouts and Sheathed Cables Lashed to Hanging Steel Strand”,
premature expansion loop cracking. IEEE Transactions on Cable Television, (Oct. 1976),
pp. 18-24.
ACKNOWLEDGEMENTS 6 Jack Park, The Wind Power Book, (Palo Alto, CA:
The author would like to express his thanks to Times Cheshire Books, 1981), p. 58, p. 164.
Fiber Communications for their support and to those in 7 Robert R. Lukems ed. et al, 1977 Annual Book of ASTM
the field who, over the years, have provided him with Standards, (Philadelphia: American Society for Testing
numerous firsthand experiences. and Materials, 1979), A475-78.
APPENDIX
Table 1.
CABLE PROPERTIES:
Unjacketed Trunk And Feeder
Type Weight Diameter Er ar Ar
(pound/foot) (inches) (pounds/square (inches/inch (square inches)
inch x 10 6) °F x 10 -6)
412 0.058 0.412 2.79 13.3 0.133 3
500 0.078 0.500 2.44 13.5 0.196 4
625 0.121 0.625 2.42 13.5 0.306 8
750 0.171 0.750 2.37 13.5 0.441 8
875 0.225 0.875 2.26 13.5 0.601 3
1000 0.325 1.000 2.62 13.4 0.785 4
TABLE 1,
Continued
59 Quad
Single 0.027 0.262 1.17 10.8 0.042 5
Messengered 0.039 0.417 2.03 8.89 0.052 6 0.051
6 Quad
Single 0.034 0.297 1.22 10.6 0.055 6
Messengered 0.046 0.452 1.90 9.03 0.065 8 0.051
7 Quad
Single 0.045 0.350 1.21 10.4 0.079 8
Messengered 0.066 0.527 2.24 8.66 0.094 1 0.072
11 Quad
Single 0.071 0.434 1.06 10.2 0.126 7
Messengered 1 0.100 0.657 1.91 8.59 0.149 4 0.083
Messengered 2 0.096 0.658 2.60 8.44 0.153 2 0.109
Note: The messenger used on the 412 and 500 cable is a solid 0.109 inch EHS steel wire,
a stranded 0.1875 inch EHS steel wire is used on the 625 and TX565.
Er, ar and Ar are the resultant elastic moduli, expansion coefficients and areas respectively.
TABLE 2.
Mechanical Properties Of
Selected Materials
TABLE 3.
Properties of Selected
Messenger Wires And
EHS Steel Strands7
Type Stranding Diameter Weight Area Break
(inches) (pounds/foot) (square inches) Strength x
1000 pounds
TABLE 4. TABLE 5.
NESC Loading Table3 NESC Clearance Requirements3
32°F, No Wind, With Ice,
Heavy Medium Light Extreme
Or At 120°F, No Wind
Wind
Radial Ice
Roads, alleys and other land subject to
(inches) 0.5 0.25 0.0 0.0
truck traffic or farm vehicles 15.5
Wind See
Residential driveways, no truck traffic 15.5
(pounds/square foot) 4 4 9 Fig. A2
Pedestrians only 14.0
Temperature (°F) 0 + 15 + 30 + 60
Railroad crossings 23.5
Weight Adder
(pounds/foot) 0.30 0.20 0.05 0.0 Water areas sailboats prohibited
(unless otherwise specified) 14.0
See the NESC for restrictions that apply.
Figure A1.
NESC Loading District Map
(Printed With Permission)
Figure A2.
NESC Extreme Wind Loading Map
(Printed With Permission)