Professional Documents
Culture Documents
PROGRESS REPORT
by
Victor Jinglue Hu
(251058373)
1|Page
Table of Contents
THE UNIVERSITY OF WESTERN ONTARIO FACULTY OF ENGINEERING SCIENCE ..................... 1
Chapter 1 – Introduction ............................................................................................................................... 3
1.1 Objectives ........................................................................................................................................... 3
1.2 Literature Review................................................................................................................................ 3
1.3 Scope of Work .................................................................................................................................... 6
Chapter 2 - Coefficient Table Method .......................................................................................................... 7
2.1 Methodology of Solution .................................................................................................................... 7
2.2 Numerical Modelling ........................................................................................................................ 10
2.3 Result ................................................................................................................................................ 13
2.4 Example ............................................................................................................................................ 13
2.5 Validation.......................................................................................................................................... 15
Chapter 3 - Equivalent Load Method.......................................................................................................... 16
3.1 Methodology of Solution .................................................................................................................. 16
3.2 Numerical Modelling ........................................................................................................................ 19
3.3 Result ................................................................................................................................................ 21
3.4 Validation.......................................................................................................................................... 22
Reference .................................................................................................................................................... 23
Appendix A - Coefficient Table for Obtaining 𝝀𝟏, 𝝀𝟐, 𝝀𝟑 ......................................................................... 24
Appendix B - 𝑹𝒙𝒏 Modelling Results ........................................................................................................ 27
Appendix D - Variation of equivalent ratio with each variable .................................................................. 29
Appendix E - Transmission Tower Model using SAP2000 ........................................................................ 32
2|Page
Chapter 1 – Introduction
1.1 Objectives
The objective of this study is to find a simplified method to account for “Load Case 3-Oblique” in
ASCE-74 --- the unbalanced longitudinal force on transmission tower due to non-symmetric
downburst conductor loads. The background theory of my research will be discussed in section 1.2,
including the downburst wind field and three critical load cases.
Downburst is a non-synoptic High Intensity Wind (HIW) whose mean velocity varies with time and can
cause significant damage to existing structures of transmission lines. Intensive studies have been conducted
at Western University in this area, triggered by reported incidents worldwide of transmission lines failing
due to downbursts throughout the years. This study will be focusing on the critical downburst load case that
includes non-symmetric load distribution on the conductors. Since it’s considered not feasible to obtain the
wind field data when real downburst events happen, numerical modeling of the wind field using
Computational Fluid Dynamics (CFD) (Hangan et al, 2003) was developed at Western University, from
which the wind field data was obtained from an impinging jet model and then scaled up to match the full-
scale events by a procedure created by Shehata et al (Shehata et al, 2005). Characteristics of the downburst
wind field depend on its location, size, and intensity; the parameters that can determine the structural
response of a downburst include jet diameter 𝐷𝑗, velocity 𝑉𝑗, and polar coordinates, R and 𝜃, of the
downburst’s center as shown in Figure 1-1 (a) (Elawady & El Damatty, 2016).
Figure 1-1
(Elawady & El Damatty, 2016)
Downward jet velocity 𝑉𝑗 creates a vortex when reaching the ground; the variation between elevation and
subsequent velocity components of the vortex: 𝑉𝑅𝐷 (radial component), 𝑉𝑉𝑅 (vertical 4 component) are
shown in Figure 1-1 (b) (Elawady & El Damatty, 2016). It is found that the vertical velocity is less
significant compared to radial velocity at the height of about 60 meters, which is a typical maximum height
of a transmission tower (Elawady & El Damatty, 2016), as a result, the vertical velocity can be considered
negligible in structural analysis. Furthermore, the radial velocity 𝑉𝑅𝐷 has two components relative to the
tower of interest—in the longitudinal direction of the transmission line as well as the transverse direction.
Since the force is non-linearly proportionate to the velocity, each radial velocity component will introduce
different loading conditions to both the tower of interest and the conductor on either side of the tower,
depending on the downburst wind field. It is necessary to obtain such reactions for practitioners to analyze
the structural response of the transmission tower. The parametric study carried out by EL Damatty and
Elawady (El Damatty & Elawady, 2018) has established three critical load cases to account for such loading
conditions, and the procedure was incorporated into ASCE-74:
3|Page
• Load Case 1: 𝜃 =0
• Load Case 2: 𝜃 = 90
4|Page
Figure 1-5 (b)- transverse radial velocity distribution over six spans
In Load Case 1 and 2, the transverse reactions from the conductor to the tower of interest can be calculated
using equivalent velocity obtained from the graph, as shown in Figure 1-3 (b) and 1-5 (b). However, the
transverse radial velocity distribution is non-symmetrical on each conductor adjacent to the tower of interest
as shown in Figure 1-5 (b); as a result, there will be an unbalanced force, 𝑅𝑥, being introduced to the
longitudinal direction of the tower of interest. The parametric study (Elawady & El Damatty, 2016) shows
that 𝑅𝑥 is related to the following parameters: Insulator length (h), wind pressure (𝛼), sag ratio (s),
conductor’s weight per unit length (w); based on different range of those parameters shown in Table 1-1,
eight groups of charts are provided by the study.
𝑅𝑥 can be obtained using three-dimensional linear interpolation using the equations below:
Where 𝑅𝑥1, 𝑅𝑥2, …, 𝑅𝑥8 can be respectively obtained from the charts based on their defined group using
known span length, sag ratio, and wind pressure; refer to the upper and lower limits of each group for
5|Page
𝛼𝑚𝑖𝑛/𝑚𝑎𝑥, ℎ𝑚𝑖𝑛/𝑚𝑎𝑥 (Elawady & El Damatty, 2016).
The scope of this research would be a continued study of Elawady & El Damatty (2016), and is composed
of two parts that will be discussed further in Chapter 2 and 3 of this report:
The Coefficient Table Method discussed in Chapter 2 is an alternative to the graphical method
incorporated in ASCE-74, using a 16 x 60 table shown in Appendix A and a procedure described in
Section 2.3 of this report. The scope of this method includes all eight groups shown in Table 1-1 with
validation attached in Appendix C.
The Equivalent Load Method discussed in Chapter 3 will propose an approach discussed in section 3.3 to
account for the numerical solution of the unbalanced force Rx using FEM analysis by SAP2000. Due to
the iteration process and time constraint, this chapter will focus on one Group listed in Table 1-1 (Group
III).
6|Page
Chapter 2 - Coefficient Table Method
As mentioned in section 1.2 of this report, the unbalanced force, Rx, in the longitudinal direction, is related
to the following parameters: Insulator length (h), wind pressure (𝛼), sag ratio (s), conductor’s weight per
unit length (w); the value of Rx can be subsequently obtained using three-dimensional linear interpolation,
shown below, where 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8 can be respectively obtained from the charts based on their defined
group using known span length, sag ratio, and wind pressure;
Firstly, define variables 𝑘1 , 𝑘2 , 𝑘3 , related to wind pressure, conductor’s self-weight and insulator height,
respectively as shown below,
𝛼 − 𝛼𝑚𝑖𝑛
𝑘1 =
𝛼𝑚𝑎𝑥 − 𝛼𝑚𝑖𝑛
𝑤𝑚𝑎𝑥 − 𝑤
𝑘2 =
𝑤𝑚𝑎𝑥 − 𝑤𝑚𝑖𝑛
ℎ𝑚𝑎𝑥 − ℎ
𝑘3 =
ℎ𝑚𝑎𝑥 − ℎ𝑚𝑖𝑛
7|Page
Substitute into the three-dimensional linear equations to get the unbalanced force Rx in terms of 𝑅𝑥𝑛 , 𝑛 =
1,2, … ,8:
𝑅𝑥 =
+(−𝑅𝑥7 + 𝑅𝑥8 ) × 𝑘1
+(𝑅𝑥5 − 𝑅𝑥7 ) × 𝑘2
+(𝑅𝑥3 − 𝑅𝑥7 ) × 𝑘3
+𝑅𝑥7 × 1
8
∗
𝑤ℎ𝑒𝑟𝑒 𝑅𝑚 = ∑ 𝜇𝑚𝑛 𝑅𝑥𝑛 ; 𝑚 = 1, 2, … , 8
𝑛=1
𝜇𝑚𝑛 is a parameter accounting for the contribution of additivity in terms of 𝑅𝑥𝑛 , and the values are shown
in Table 2-1 below:
Table 2-1. Values of 𝜇𝑚𝑛
For each sag ratio, 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8, can be obtained from the graph in ASCE-74 – Appendix K. By
numerical modelling discussed in Section 2.2 of this report, it is found that 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8, can be
approximated by a quadratic equation in following terms:
Where 𝜑1𝑛 , 𝜑2𝑛 , 𝜑3𝑛 , 𝑛 = 1,2, … , 8 , are coefficients in numerical values that can be obtained from
function modelling of 𝑅𝑥𝑛 .
8|Page
∗
As a result, 𝑅𝑚 , 𝑚 = 1, 2, … , 8, can be expressed in quadratic term as well:
𝑦𝑖𝑒𝑙𝑑𝑠 (2-3)
∗ ∗ ∗ ∗
→ 𝑅𝑚 = 𝜑1𝑚 𝐿2 + 𝜑2𝑚 𝐿 + 𝜑3𝑚 , 𝑚 = 1,2, . . ,8
𝑤ℎ𝑒𝑟𝑒
8
∗
𝜑1𝑚 = ∑ 𝜇𝑚𝑛 𝜑1𝑛 𝑚 = 1,2, … ,8
𝑛=1
8
∗
𝜑2𝑚 = ∑ 𝜇𝑚𝑛 𝜑2𝑛 𝑚 = 1,2, … ,8
𝑛=1
8
∗
𝜑3𝑚 = ∑ 𝜇𝑚𝑛 𝜑3𝑛 𝑚 = 1,2, … ,8
𝑛=1
𝐿𝑒𝑡
∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗
𝜆𝑖 = 𝜑𝑖1 𝑘1 𝑘2 𝑘3 + 𝜑𝑖2 𝑘2 𝑘3 + 𝜑𝑖3 𝑘1 𝑘3 + 𝜑𝑖4 𝑘1 𝑘2 + 𝜑𝑖5 𝑘1 + 𝜑𝑖6 𝑘2 + 𝜑𝑖7 𝑘3 + 𝜑𝑖8 (2-4)
𝑤ℎ𝑒𝑟𝑒 𝑖 = 1, 2, 3
∗
𝜑𝑖2
𝑎𝑖1 = ∗
𝜑𝑖1
∗
𝜑𝑖3
𝑎𝑖2 = ∗
𝜑𝑖1
∗
𝜑𝑖4
𝑎𝑖3 = ∗
𝜑𝑖1
∗
𝜑𝑖1
𝛽𝑖 =
𝑎𝑖1 𝑎𝑖2 𝑎𝑖3
∗
𝑏𝑖1 = 𝜑𝑖5 − 𝛽𝑖 𝑎𝑖1
∗
𝑏𝑖2 = 𝜑𝑖6 − 𝛽𝑖 𝑎𝑖2
∗
𝑏𝑖3 = 𝜑𝑖7 − 𝛽𝑖 𝑎𝑖3
∗
𝑐𝑖1 = 𝜑𝑖8 − 𝛽𝑖
𝑦𝑖𝑒𝑙𝑑𝑠
→ 𝑅𝑥 = 𝜆1 𝐿2 + 𝜆2 𝐿 + 𝜆3
9|Page
From equation (2-1) (2-2) (2-3) (2-4), the unbalanced force 𝑅𝑥 , can be expressed in quadratic term shown
above, where 𝜆1 , 𝜆2 , 𝜆3 is the coefficient for the second order, first order and constant term, respectively.
Additionally, 𝜆𝑖 , 𝑖 = 1,2,3, is a known function of variables 𝑘1 , 𝑘2 , 𝑘3 :
The coefficients in the above function can further be determined numerically in a systematic approach for
each sag ratio and organised into a table for all eight groups. Practitioners can use such table to obtain the
value of 𝑅𝑥 subsequently. The approach to obtain the coefficients numerically will be discussed in section
2.2 of this report.
This section will focus on the numerical process of obtaining the Coefficient Table using Microsoft Excel,
based on the theoretical development discussed in previous section. Figures and Tables in this section use
Group I, S= 2% as an example to demonstrates the numerical process and can be broken down into three
steps in a systematic approach, such that it can be repeated for other sag ratios and all eight groups.
Eight groups of charts are given in ASCE-74-Appendix K; with each group has eight sets of graphs,
representing 𝑅𝑥1 , 𝑅𝑥2 , … , 𝑅𝑥8 , respectively. Five curves in each graph corresponds to five different sag
ratios (S=2.0%, 2.5%, 3.0%, 3.5%, 4.0%), respectively, as shown in Figure 2-1.
10 | P a g e
Table 2-2. Group I, S = 2 %
After digitizing the given graphs into Excel, as shown in Table 2-1, the first step would be fitting each curve
of a specific sag ratio with a polynomial function. It is found that the best fit would be using a polynomial
function with order of 2; the regression analysis yields 𝑅 2 = 0.99, applicable for all eight groups; Figure
2-1 illustrates such fitting process for instance: the horizontal axis represents wind span L, while the vertical
axis is the value of 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8 in 𝑘𝑁.
The coefficients 𝜑1𝑛 , 𝜑2𝑛 , 𝜑3𝑛 of 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8 can be organised into a table format as shown in Table
2-2, in order to further compute the coefficients of λi , i = 1,2,3 in Step 3 using the built-in functions of
Excel. The complete fitting results of 𝑅𝑥𝑛 , can be found in Appendix B of this report.
11 | P a g e
Step 3 - Solving coefficients of 𝜆𝑖 , 𝑖 = 1,2,3
Figure 2-3 illustrates the process of obtaining the numerical values needed for the Coefficient Table; with
𝜑1𝑛 , 𝜑2𝑛 , 𝜑3𝑛 , 𝑛 = 1,2, … ,8 as inputs in Step 2, the coefficients value 𝑎𝑖1 , 𝑎𝑖2 , 𝑎𝑖3 , 𝛽𝑖 , 𝑏𝑖1 , 𝑏𝑖2 , 𝑏𝑖3 , 𝑐𝑖1 , can
be subsequently obtained as outputs in a table format as shown in Table 2-3; the complete Coefficient Table
can be found in Appendix A of this report.
12 | P a g e
2.3 Result
The procedure to obtain the unbalanced longitudinal force 𝑅𝑥 are shown below:
1. Obtain 𝑘1 , 𝑘2 , 𝑘3
𝛼 − 𝛼𝑚𝑖𝑛
𝑘1 =
𝛼𝑚𝑎𝑥 − 𝛼𝑚𝑖𝑛
𝑤𝑚𝑎𝑥 − 𝑤
𝑘2 =
𝑤𝑚𝑎𝑥 − 𝑤𝑚𝑖𝑛
ℎ𝑚𝑎𝑥 − ℎ
𝑘3 =
ℎ𝑚𝑎𝑥 − ℎ𝑚𝑖𝑛
3. Calculate 𝑅𝑥
𝑅𝑥 = 𝜆1 𝐿2 + 𝜆2 𝐿 + 𝜆3
2.4 Example
Design data:
• Wind span L = 400 m
• Jet velocity 𝑉𝑗 = 65 𝑚⁄𝑠 2
• Conductor weight per unit length w = 30 𝑁⁄𝑚
• Conductor projected diameter d = 0.0345 m
• Insulator height h = 2 m
• Line sag ratio S = 3%
120 ≤ 𝛼 ≤ 160
25 ≤ 𝑤 ≤ 40
13 | P a g e
𝑦𝑖𝑒𝑙𝑑𝑠
→ 𝐺𝑟𝑜𝑢𝑝 𝑉𝐼
𝑤𝑚𝑎𝑥 − 𝑤 40 − 30
𝑘2 = = = 1.62
𝑤𝑚𝑎𝑥 − 𝑤𝑚𝑖𝑛 40 − 25
ℎ𝑚𝑎𝑥 − ℎ 5−2
𝑘3 = = = 0.75
ℎ𝑚𝑎𝑥 − ℎ𝑚𝑖𝑛 5 − 1
𝑖=1
⇒ 𝜆1 = 𝛽1 [(𝑎11 𝑘1 + 1)(𝑎12 𝑘2 + 1)(𝑎13 𝑘3 + 1) + 𝑏11 𝑘1 + 𝑏12 𝑘2 + 𝑏13 𝑘3 + 𝑐1 ]
= (1.65 × 10−7 )[(−7.63 × 0.64 + 1)(−7.07 × 1.62 + 1)(−3.26 × 0.75 + 1) + 122.26 × 0.64
+ 99.87 × 1.62 + 102.12 × 0.75 + 159.12)] = 5.933 × 10−5
𝑅𝑥 = 𝜆1 𝐿2 + 𝜆2 𝐿 + 𝜆3
14 | P a g e
2.5 Validation
This section will discuss the validation of Coefficient Table Method, including the method used for
validation, possible source and errors, and potential approach to get better accuracy. The complete
validation can be found in Appendix C of this report.
A random point of each sag ratio within design parameters’ (jet velocity 𝑉𝑗 , conductor’s weight per unit
length w, and insulator height h, and wind span L) defined range, based on their respective groups, were
chosen to validate the Coefficient Table Method. The obtained values of unbalanced longitudinal force 𝑅𝑥
using Coefficient Table Method were then compared to the Graphical Method given in ASCE-74. The
errors were subsequently obtained by assuming the results of Graphical Method to be the true value. The
average error is found to be 1.2 %, with possible source of error coming from:
• The curve fitting of 𝑅𝑥𝑛 , 𝑛 = 1,2, … ,8, in the numerical modelling process discussed in Step 1,
section 2.2 of this report. Specifically, Figure 2-4 shows the two curves of 𝑅𝑥1 , 𝑅𝑥2 with sag ratio
S=2% in Group V; although the regression analysis, using polynomial function with order of 2,
yields 𝑅 2 = 0.99, some points like L=200 will have minor deviation and can not be perfectly fitted.
The solution to get better accuracy is to use a polynomial function with order of 3; as a result, the
Coefficient Table will expand from 𝑖 = 1,2,3 to 𝑖 = 1,2,3,4; the procedure will remain the same.
15 | P a g e
Chapter 3 - Equivalent Load Method
This study will focus on finding an equivalent load 𝑄𝑒𝑞 in 𝑁⁄𝑚 using FEM analysis, that can be
applied uniformly to the cable in order to obtain the unbalanced longitudinal force 𝑅𝑥 on the tower
of interest.
As discussed in Chapter 1, the unbalanced force Rx is related to the following variables: jet velocity
𝑉𝑗 , conductor’s weight per unit length w, and insulator height h, and wind span L, line sag ratio S.
With the above variables defined, the value of Rx can be computed by the numerical solution
(Elawady & El Damatty, 2016) using three-dimensional linear interpolation as outlined in section
1.2.
By assuming the numerical solution to be the true value of the unbalanced force 𝑅𝑥 (𝑡𝑟𝑢𝑒), a
transmission tower model can be constructed by SAP2000 using six spans in responding to the
parametric study conducted by Elawady & El Damatty (2016). A uniformly distributed load 𝑄 is
then applied to the cable on one side to the tower of interest to obtain the value of 𝑅𝑥 (𝐹𝐸𝑀). Then,
equating 𝑅𝑥 (𝐹𝐸𝑀) = 𝑅𝑥 (𝑡𝑟𝑢𝑒), the output 𝑄𝑒𝑞 , can be subsequently determined. The detail of
the iteration process to obtain 𝑄𝑒𝑞 will be demonstrated in section 3.2 of this report.
• Step 1- Choose a Group for the five variables listed above (Group III for example) and find the
range of each variable:
48 ≦ 𝑉𝑗 ≦ 59 𝑚⁄𝑠
1≦ℎ≦5𝑚
10 ≦ 𝑤 ≦ 25 𝑁⁄𝑚
100 ≦ 𝐿 ≦ 500 𝑚
2≦𝑆 ≦4%
• Step 2- Choose a value of each variable within their own range, respectively:
𝑉0 = 50 𝑚/𝑠
ℎ0 = 2 𝑚
𝑤0 = 15.7 𝑁/𝑚
𝐿0 = 400 𝑚
𝑆0 = 2.5 %
𝑉𝑒𝑞
• Step 3- Assume the ratio 𝑟 = 𝑉𝑗
is function of 𝑉𝑗 , ℎ, 𝑤, 𝐿, 𝑆, where:
2𝑄𝑒𝑞
𝑉𝑒𝑞 = √
𝜌𝑑
16 | P a g e
We have:
𝑟 = 𝐹(𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 )
Vj h w L S
x1 = x2 = x3 = x4 = x5 =
V0 h0 w0 L0 S0
, , , ,
• Step 4- Varying the value of one variable within their respective range and control the values of
other variables unchanged; five tables can then be constructed showing the variation of ratio 𝑟 with
each variable. The variation of r with each variable can then be approximated use a polynomial
function with the maximum order of 3, with all the coefficients as numerical values as shown
below:
3
r = g i ( xi ) = ain xi3− n (i = 1,2,3,4,5)
n =0
The detail of the numerical process to find 𝑔𝑖∗ (𝑥𝑖 ) 𝑖 = 1,2, … ,5, will be discussed in section 3.2.
• Step 5- Find the final format function 𝑟 = 𝐹(𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 ); the general process can be described
by the mathematical form blow:
5
Assume r = F ( x1 , x2 , x3 , x4 , x5 ) = g i ( xi ) ------(1)
i =1
Where:
Vj h w L S
x1 = x2 = x3 = x4 = x5 =
V0 h0 w0 L0 S0
, , , ,
3
g i ( xi ) = ain xi3− n (i = 1,2,3,4,5)
n =0 ------(2)
a10 = 0, a30 = 0, a40 =0
where:
from(1)and(3):
17 | P a g e
5
g m (1)
g ( xi ) = g i ( xi )
i
m =1
(i = 1,2,3,4,5)
g i (1)
Let:
5
g m (1)
Ai = m =1
(i = 1,2,3,4,5)
g i (1)
g i ( xi ) = Ai g i ( xi ) (i = 1,2,3,4,5)
We have:
3 3
= g i (1) (i = 1,2,3,4,5)
Ai
5.
Since
5
g m (1)
Ai = m =1
(i = 1,2,3,4,5)
g i (1)
We have:
5
A
m =1
m =4
------(7)
5
1
r= g i ( xi ) ------(8)
4 i =1
18 | P a g e
Since:
a10 = 0, a30 = 0, a40 =0
a11a20 a31a41a50
Let: = ------ (9)
4
Where:
g1 ( x1 ) g 2 ( x2 ) g 3 ( x3 )
f1 ( x1 ) =
f 2 ( x2 ) =
f 3 ( x3 ) =
a11 a20 a31
, , ,
g ( x4 ) g ( x5 )
f 4 ( x4 ) = 4
f 5 ( x5 ) = 4
a41 a50
,
We have:
5
r = f i ( xi ) ------ (11)
i =1
∗ ∗ ∗ ∗ ∗
𝑎11 𝑎20 𝑎31 𝑎41 𝑎50
𝜅=
𝜉4
k ij
(i = 0,1,2,3; j = 1,2,3,4,5) can be presented numerically in a table format shown in
section 3.3
• The numerical process of Step 3 discussed in section 3.1 can be illustrated as the flowchart
presented by Figure 3-1. The results of the variation of equivalent ratio r with each variable can be
found in Appendix D of this report. The result of 𝑔𝑖∗ (𝑥𝑖 ) 𝑖 = 1,2, . . ,5 is shown below:
• The detailed procedure for SAP2000 model can be found in Appendix E of this report.
• The numerical result of Step 5 discussed in section 3.1 is resented in section 3.3 of this report.
19 | P a g e
Figure 3-1. Flowchart for Step 3 discussed in section 3.1
20 | P a g e
3.3 Result
𝑟 = 𝜅 ∏ 𝑓𝑖 (𝑥𝑖 )
𝑛=1
𝜅 = 2.44 × 10−5
ℎ
2 1 0.806 -9.563 -22.080
ℎ0
𝑤
3 0 1.000 -1.802 -6.884
𝑤0
𝐿
𝐿0
∈ [0.25,0.45] 0 1.000 -3.835 18.340
4 𝐿
∈ [0.45,1.25] 0 -161.772 132.910 -10.341
𝐿0
𝑆
5 1 -3.621 3.827 0.875
𝑆0
𝑉𝑒𝑞 = 𝑟𝑉𝑗
1
𝑄𝑒𝑞 = 𝜌𝑉𝑗2 𝑑 (𝑢𝑛𝑖𝑡 𝑜𝑓 𝑁⁄𝑚)
2
The unbalanced longitudinal force 𝑅𝑥 for the tower of interest can be subsequently obtained by applying a
uniformly distributed load with the value of 𝑄𝑒𝑞 along the cable.
21 | P a g e
3.4 Validation
Five random points were chosen based on the range of respective variables (Group III) to validate the
Equivalent Load Method described in section 3.3. The errors were then obtained by assuming the
numerical values of Rx to be the true value as discussed in section 3.1 of this report and the results are
shown in Table 3-2 below.
• Not enough points when constructing the variation points of each variable
• The iteration process assumed 𝑅𝑥 (𝐹𝐸𝑀) = 𝑅𝑥 (𝑡𝑟𝑢𝑒) to the accuracy of one decimal point,
as illustrated in the Flowchart in section 3.2. However, the ratio r is sensitive to two
decimal points. As a result, all the points in the variation table listed in Appendix E should
be further adjusted to increase the overall accuracy of this method.
• Step 5 discussed in section 3.1 assumed the obtained function 𝑔𝑖∗ (𝑥𝑖 ) 𝑖 = 1,2, . . ,5 as listed in
Table 3-1 to be the exact curve that fits all the data points, in other words, the regression
analysis would yield 𝑅 2 = 1 in order to satisfy the mathematical equation (6) listed in Step 5.
The actual 𝑅 2 = 0.99, and this would ultimately affect the accuracy of the coefficient 𝜅 listed
in section 3.3. For instance, the value of obtained in section 3.3 has an accuracy to two decimal
points; however, the accuracy of 𝜅 would be better to increase one more decimal point in order
to make the ratio r sensitive to two decimals.
22 | P a g e
Reference
Hangan, H., Roberts, D., Xu, Z., & Kim, J. (2003). Downburst simulation. Experimental and numerical
challenges. In Proceedings of the 11th International Conference on Wind Engineering, Lubbock, Texas,
Electronic Version.
Shehata AY, El Damatty AA, Savory E. Finite element modeling of transmission line under downburst
wind loading. Finite Elem Anal Des 2005;42(1):71–89.
Elawady, A., & El Damatty, A. (2016). Longitudinal force on transmission towers due to
non symmetric downburst conductor loads. Engineering Structures, 127, 206–226.
https://doi.org/10.1016/j.engstruct.2016.08.030
El Damatty, A., & Elawady, A. (2018). Critical load cases for lattice transmission line structures
subjected to downbursts: Economic implications for design of Transmission Lines. Engineering
Structures, 159, 213–226. https://doi.org/10.1016/j.engstruct.2017.12.043
23 | P a g e
Appendix A - Coefficient Table for
Obtaining 𝝀 , 𝝀 , 𝝀
24 | P a g e
25 | P a g e
Appendix A – Coefficient Table for Obtaining 𝝀𝟏 , 𝝀𝟐 , 𝝀𝟑
26 | P a g e
Appendix B - 𝑹𝒙𝒏 Modelling Results
27 | P a g e
Appendix C – Validation of Coefficient
Method
28 | P a g e
Appendix D - Variation of equivalent ratio with each variable
(Group III)
29 | P a g e
Table D-5. Variation of r with S
30 | P a g e
Figure D-3. Variation of r with w/wo
31 | P a g e
Appendix E - Transmission Tower Model using SAP2000
Insulator section:
32 | P a g e
Cable section
33 | P a g e
Step 4-Define Load Case
34 | P a g e
Step 5-Define Cable Geometry
35 | P a g e
Step 7- Applying uniformly distributed load Q
36 | P a g e
Step 7-Set Load Case to Run
37 | P a g e