Structural Behavior of A Guyed Mast
Structural Behavior of A Guyed Mast
Structural Behavior of A Guyed Mast
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/279199249
READS
69
1 author:
Alina de la Cruz
Universidade Federal do Rio Grande do Sul
1 PUBLICATION 0 CITATIONS
SEE PROFILE
All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Alina de la Cruz
letting you access and read them immediately. Retrieved on: 23 May 2016
1
ABSTRACT: The inherent nonlinearity in the structural behavior of guyed mast leads to difficulties in their structural analysis,
and prevents the formulation of a general-purpose design methodology [1]. The current design standard in Cuba [2] uses a static
approach to consider the fluctuating wind component, and a high occurrence rate of structural failures, causing the government to
lose millions of dollars on repairs, creates the need for the study, optimization and review of existing models of guyed mast. The
study of the structural behavior under extreme winds and the number of failures by type of elements was performed.
Two methodologies for calculating the fluctuating wind components were used: NC 285: 2003 (current Cuban code) and the
Patch Load proposed by Sparling and adopted by the British Standard [3]. The improvement of the tower design focused on the
aerodynamic factor in columns and changes of the width of the cross section of the mast. The results showed that the proposed
tubular column sections reduces the maximum axial force by up to 24% and significantly reduces the work of all elements of the
mast, cables and supports. The increased width of the transversal section of the shaft caused a detrimental increase to the internal
axial forces. Decreasing the cross section, however, produces lower axial forces in cables and columns. In horizontal members
there were no significant results and an increase in the compression axial forces occurs for the diagonal member.
From the comparison of the Patch Load method with the static method of the Cuban Standard, it is concluded that the internal
forces increase, in all types of elements, with the implementation of Patch Load method.
KEY WORDS: Guyed Mast; Structural Behavior; Wind Load; Spatial Arrangement.
1 INTRODUCTION
The analysis and design of masts and towers requires special knowledge and experience, especially when it concerns guyed
mast. The special problems related to these structures are underlined by many collapses during the years [4]. For example, in
Cuba, in the 1996-2006 period, 33 telecommunication towers failed due to strong winds associated with hurricanes and severe
storms. Likewise, in the hurricane season of 2008, 20 towers were damaged by high-speed winds [5]. The number of collapses of
guyed masts is relatively higher than other types of structures [6], due to the nonlinear behavior of guy ropes, which has led to
numerous investigations into its structural behavior. Furthermore, with the current knowledge on guyed towers and calculation
tools available, the study and revision of existing models established for decades is justified.
The predominant load on self-supporting towers and guyed mast is the wind load, although in some areas the atmospheric icing
and seismic loads may have a significant influence in the design. The wind is a dynamic load, and slender structures are sensitive
as they have low damping characteristics.
The resistance of a lattice tower dependent on two parameters: the meteorological parameter, which determines the wind speed
and the shape parameter, which determines the aerodynamic coefficients. In these structures, the member’s shape, dimensions and
spatial distribution, are aspects that significantly influence the wind load calculation. The flat-sided sections, commonly used in
Cuba, have a higher wind obstruction than the tubular sections, which leads to high aerodynamic coefficients and therefore high
wind pressure on the structure.
The towers fabricated with tubular sections have shown a higher efficiency compared to the flat-sided sections, the latter widely
used on existing models in the country. This is because, not only the beneficial aerodynamic properties of tubular cross-section,
but also the inertia properties compared to an angle bar member of the same cross-sectional area. Also in the design phase, tubular
towers are often less complicated due to the uniform inertia.
Moreover, the width variation of the cross section tower influences on the stability of the whole structure and individual
elements.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
2
Ø 13
90000
86000
68000
Ø9
50000
33000
17000
42000
82000
Figure 2. Geometry of MAR 300 tower. a) 3D view of the tower with the first atypical section. b) Typical face of the
tower. Units: mm.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
3
The elements that conform the mast are, in all cases, flat-sided sections. Specific dimensions are given in Table 1 and the cable
levels in Table 2 and Figure 3.
C-82
C-42
0°
12
12
C
A B
B-42
120
A-42
B-82
A-82
2.2 Support
Support conditions are given for ground anchorages of the cables and the shaft support. In all cases, articulated joints are
considered.
2.3 Material
A linear elastic material and constant properties in time was considered. Data of materials used for the shaft elements and cables
are given below:
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
4
3 LOADS
3.1 Wind load. Historical development.
The study and calculation of towers has changed over the investigations and the growing popularity of computational tools that
help resolve mathematical equations. The analysis, required multiple simplifications in the past due to its complexity, today is
done with greater precision, corresponding the model to reality and consequently leading to an economically better design.
Wind studies focused on three ways to determine this load: by wind tunnel simulation, by using static approach and the dynamic
methods, based on the time-dependent equation of the wind. It is noteworthy that the latter alternative is relatively recent compared
to the others.
Theoretical studies applied finite element models for the analysis of guyed towers [7], [8] and insulated cables [9] under the
effect of wind and earthquake.
Several researchers have studied the gust factor to simulate the fluctuating nature of wind. In 1967, Davenport [10] took the
first steps in determining a gust factor to represent the fluctuating component of the wind. At first, all researchers agreed with the
use of a gust factor to take into account the increased of wind pressure, but more recent studies argue for limiting its use for towers
up to 150m. In the case of a taller structure, the use of a specific dynamic method or the process developed by Davenport and
Sparling called Patch Load is recommended. This method is a simplification to determine the dynamic response of the tower using
successive overlapping patches to obtain the final effect of wind loads.
Based on this, the research is focused on comparing the results obtained by the gust factor, the Patch Load method and existing
theoretical models so far [11] [12].
The Cuban Wind Standard [2] allows an acceptable design of short self-supporting towers, however tall guyed mast have
characteristics that involve the use of a special criteria. To do this, many countries have developed specific standards for
calculating towers or exclusively developed sections to deal with these cases.
In Cuba, a deficiency appears in the absence of a specific standard or chapter for towers that allows a structural analysis based
on the theory of second-order (geometric nonlinearity) and the effects on the shaft and cables due to the lateral loads.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
5
The dynamic analysis provided in the Cuban Standard (NC 285:2003) is used when the oscillation period is greater than 1
second (𝑓 ≤ 1 𝐻𝑧). In the case of the tower studied, the period 𝑇 = 0.54 𝑠𝑒𝑐, therefore this case does not require dynamic
analysis.
The Cuban Standard establishes the equation that static wind component is calculated:
Where:
𝑞: Wind static pressure component.
𝑞10 : Basic wind pressure depending on the region of the country where the tower is located, 𝑞10 = 1.3 𝐾𝑛⁄𝑚2 (zona 1).
𝐶𝑡 : Recurrence period, adopted 𝐶𝑡 = 1 for 50 years.
𝐶𝑠 : Site coefficient, adopted 𝐶𝑠 = 1.10 corresponding to exposed site. In our country, it is very common that the towers are
placed in exposed sites.
𝐶ℎ : Height coefficient.
For open terrain (type A), the expression for a given height 𝑧 is:
𝑍 0.32
𝐶ℎ = ( ) (2)
10
𝐶𝑟 : Gust factor. The fluctuating nature of wind and its interaction with the structures is taken into account through the gust factor
depending on the height of the construction and the terrain type. For a height of 90 m and A terrain type, the 𝐶𝑟 = 1.06.
𝐶𝑟𝑎 : Coefficient of area reduction. In the case of lattice towers, the exposed areas are relatively small, so it takes 𝐶𝑟𝑎 = 1.
𝐶 𝑓 : Aerodynamic coefficient either drag coefficient. This value takes into account the spatial characteristics of the bodies exposed
to the wind, i.e. the coefficient involving the aerodynamic nature of the structure. For spatial lattice towers, this coefficient
becomes 𝐶 𝑓𝑒 (spatial), which is calculated by the following expression:
𝐶 𝑓𝑒 = 𝐶 𝑓 (1 + 𝑁) (3)
Where 𝐶 𝑓 = 1.9, for the case of flat-sided section. The value of 𝑁 is set to the dependence 𝐴𝑝𝑟𝑜𝑗𝑒𝑐𝑡𝑒𝑑 ⁄𝐴𝑡𝑜𝑡𝑎𝑙 and 𝑏/ℎ
relationship i.e. the dimensions. A quadruple interpolation was used to pinpoint the value of 𝑁.
The cables load is considered uniformly distributed over the entire length. Equation 4 describes the wind load on the cables:
𝑞𝑐𝑎𝑏𝑙𝑒𝑠 = 𝑞10 ∙ 𝐶𝑡 ∙ 𝐶𝑠 ∙ 𝐶ℎ ∙ 𝐶𝑟 ∙ 𝐶𝑟𝑎 ∙ 𝐶 𝑓𝑑 ∙ 𝐷 [𝐾𝑛⁄𝑚] (4)
In which 𝐷 is the corresponding cable diameter. The coefficients remain the same values described above except that the shape
coefficient, defined by:
𝐶 𝑓𝑑 = 𝐶 𝑓 ∙ 𝑠𝑒𝑛3 𝛼 (5)
The 𝐶 𝑓 coefficient is taken as equal to 1.2. The value of 𝛼 corresponds to the angle between the wind vector and the cable.
2 The relationship between the shaft bending stiffness and lateral stiffness of the cables, defined as 𝛽𝑠 , must be less than unity.
The 𝛽𝑠 parameter is given by:
4∙𝐸𝑚 ∙𝐼𝑚 /ℎ𝑠2
𝛽𝑠 = 𝑁 𝑖 𝐾 ∙𝐻
(6)
(1/𝑁𝑙 )∙∑𝑖=1 𝐺𝑖 𝐺𝑖
The equivalent inertia and the equivalent elastic modulus was calculated according to Kalha [14] (Figure 5), because the shaft of
the tower is fabricated with elements of different cross sections.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
6
Figure 5. Tower spatial representation for the calculation of the equivalent inertia and equivalent elasticity modulus, according
to Kalha [14].
The obtained value 𝛽𝑠 = 0.1711 validates the use of the Patch load method, since it is less than 1.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
7
3 The inertial resistance parameter 𝑄 must be less than 1. This value takes into account the relationship between inertial forces
and damping properties on the tower shaft:
1 3 𝐻∙𝑉𝐻 𝑚𝑎𝑉
𝑄= √ 𝑏 √𝐻∙∑ 𝑅 (9)
30 𝑎𝑉 𝑤
Where:
𝐻: Shaft height including the cantilever (m), 𝐻 = 90𝑚.
𝑉𝐻 : Average wind speed for one hour interval, at the top of the shaft (m/s), 𝑉𝐻 = 55 m/s
𝑏𝑎𝑉 : Average width of the shaft (m), 𝑏𝑎𝑉 = 1.02𝑚
𝑚𝑎𝑉 : Average mass per length unit of the shaft including the accessory elements (kg/m), 𝑚𝑎𝑉 = 5 kg/m.
∑ 𝑅𝑤 : Average value of net areas and aerodynamic coefficients for the shaft and accessories (m2/m), ∑ 𝑅𝑤 = 0.82.
𝑟 = 𝑟̅ ± 𝑟̂𝑃𝐿 (10)
Where:
𝑟̅: Mean response component.
𝑟̂𝑃𝐿 : Peak fluctuating response component.
The mean component was estimated by the coefficients established by the Cuban Standard, as discussed in section 3.3. For the
Patch Load method, the basic pressure corresponding to the mean speed interval of an hour is established, so that, the value taken
from Cuban Standard was 0.95 kN/m2, and the gust coefficient was not considered.
For the mean component a nonlinear analysis is performed while the patches of the fluctuating component were linearly applied
to the structure.
The fluctuating response 𝑟̂𝑃𝐿 was obtained as a series of static analysis for different stages of loading, taking in consideration
the stiffness properties of the system calculated at the equilibrium position. These static load patterns were applied to the mast in
succession, with the position to the scheme on figure 6:
Where:
rPL : Resultant patch load response.
rPLi : Response due to the patch load 𝑖.
𝑛: Total number of patch loads.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
8
Where:
rPL : Resultant patch load response.
λB : The background-scaling factor.
λR : The resonant magnification factor.
λTL : Turbulent length scale factor.
g: Statistical peak factor.
Using conservative values (λB = 0.75 , λR = 1.2 , λTL = 1.05 and g = 4), equation 12 reduces to:
5 CASE STUDIES
The first modification proposed, was the replacement of the columns flat-sided section in the typical tower by tubular elements
(Figure 7). To this, an equivalent angle-section to circular section is required, not involving modifications in the weight of the
tower. The tubular section chosen for the legs, after a trial and error process, corresponds to HSS3X216 (AISC catalog), which
has a diameter D=0.0762m and a thickness t=0.005486m.
Figure 8 describes the second change performed in the width shaft of the tower, from the typical dimension of 1.02m. Proposed
widths were 1.42m and 0.8m to increase and decrease the original design.
Elevation
1420 800
1020
Plan-view
1020 800
1420
Figure 8. Change proposed in width of the cross section mast, units: mm.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
9
6 RESULTS
For comparison purposes, the most stressed members were analyzed. The results in the legs, diagonals, horizontal members,
cables, anchors reactions and base tower reactions were studied. Finite element model (FEM) analysis was performed using SAP
2000 v.12 in order to obtain the internal forces for each case. For the processing of the results, the shaft is divided into columns
A, B and C. The horizontals and the diagonals were divided according to the faces of the tower in AB, BC and CA. The naming
of the cables considers the corresponding column and horizontal distance between column and anchorage point.
Columns, diagonals, horizontals and cables are subjected to axial forces, compression or tension. Moment values, shear and
torsion are very small compared to the resulting axial forces, so they were negligible in the analysis.
The base of the tower and anchor of the cables were processed to take into account differences in models in the maximum design
values of the corresponding foundation. The resulting force is divided into three axes "X", "Y" and "Z", corresponding the "Z" to
the vertical component and is the most important to the overall analysis.
Figure 9. Maximum axial force in columns according to the Cuban Standard NC 285:2003 and the Patch Load method.
In short, for all elements of the tower and the support reactions, increases occur in the axial force resulting from the application
of Patch Load method compared with the methodology provided by the Cuban Standard, in the 0 o direction.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
10
members, the most unfavorable working condition occurs for the 900 wind direction. For horizontals AB and BC, the peak axial
force is found at 68m, while for CA, this happens at 3m from the ground.
The diagonals are subjected to compression and tensions forces. However, compressive values are dominating the design of
these elements and for which the diagonals show a more vulnerable state.
The position of the cable levels in the tower determines an increase in the axial force of the diagonals. Similarly, to what occurs
in the horizontals, the diagonal members of the structure are affected by the vibratory movements of the cable due to their
stabilizing function.
The horizontals are only affected on a well-defined point at the height of each cable level. By contrast, the axial forces on the
diagonals increase over a larger height, which comprises of the elements found in the upper and lower sections of this level.
If the diagonals of the three faces of the tower are compared (Figure 11 a), it is observed that no group has a noticeable
predominance over another, as was the case with the columns.
The cables that hold the shaft of the tower are subjected to axial tension force. This value increases with height to the point
where it is critical at 68m and then, begins to decline again (Figure 11 b).
The internal forces in the B cable dominate over all the others, and making this element the most critical of the entire structure
(Figure 11 b).
Figure 10. a) Max axial force in columns b) Max axial force in horizontal members
Figure 11. a) Max axial force in diagonal member b) Max axial force in cables
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
11
The most critical condition for the tower proved to be the 900 direction, where failures occurred in 495 elements of the array.
Figure 12 shows the number of columns, diagonals and horizontals belonging to the tower, and the proportion of these elements
that failed the critical condition. In the columns, the number of resistant elements outnumber those that failed. On the diagonals,
unlike the columns, the number of elements in failure is much greater than the resistant ones under extreme wind loading. In the
horizontals, no losses occurred.
Due to the loads imposed on the structure, three failures occurred in the cables that hold the shaft. These cables are located at
the second and fourth levels of Column B, which are at heights of 33m and 68m. The other failure is located at the Column A and
a height of 33m. Despite this result, the cable is considered to be working at 100% capacity, however, some regulations such as
TIA/EIA [16] establish that only 60% of the ultimate load should be considered for greater safety of these elements. Considering
this, the amount of cables in failure will increase significantly.
Table 7 shows the maximum axial force registered and the resistance capacity of each element. The demand/capacity ratio is a
measure of each member’s performance. A closer relationship to 0 indicates that the element works conveniently and sufficient
to withstand an increase in loads. For values close to 1, the element is at the limit capacity and a value exceeded, the structural
failure will occur.
From the results shown in Table 7, it is concluded that axial forces exceeding the ultimate strength limit were recorded in all
elements leading to failure, except for the horizontal members. In addition, it is observed that in the columns, demand/capacity
ratio is greater than the other elements, indicating that that loading has been imposed over its capacity.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
12
Figure 13 shows the results of the wind load on the model with tubular sections in comparison to the original flat-sided model,
over the entire height. The behavior shown in the figure is similar for all other wind directions.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
13
in the AB horizontals. In the model of tubular sections, contrary to the above, the maximum value is located in the BC horizontals.
If these two peaks of forces are compared, the resulting difference is 10%.
Figure 17. Wind load in models with change in width of the cross section mast
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
14
The maximum axial force recorded for each element varies after change in the width of the shaft. The 1.42m model aggravates
the compression (Figure 18), in each column of the tower. The 0.8m model, unlike the first, registers a decrease in the resultant
axial force in columns. It is therefore an advantage to decrease the width of the cross section shaft.
The maximum axial forces registered in the typical model and 0.8m model correspond to C columns, while for the 1.42m model,
the maximum compression is located in B column. If we compare these three peaks, it is appreciated that the 0.8m model decreases
the axial force 9% and the 1.42m model increases 7% the axial force, respect to typical model.
In the horizontals, higher values of axial force produced with an increased width of the cross section of shaft (Figure 19).
The 0.8m model registered a decrease in the values of the maximum axial force in the members of the BC and CA sides. In AB
horizontals is observed a slight increase in this value, but is negligible because it’s minor than 1%.
The maximum traction was recorded in the AB elements for widths of 1,02m and 0,8m, while for the 1.42m model, it is located
at the BC face. If these maximum axial forces in each model are compared, the behavior is unfavorable in both cases relative to
the typical model. Increases in percentage terms are 7% for 1.42 m model.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
15
7 CONCLUSIONS
1. The structural revision of the MAR 300 tower indicated that, regardless of the method used for calculating wind load,
either the Cuban Proceedings (NC285:2003) or the Patch Load, the tower does not resist the regulation set speed for the Cuba
western region.
2. Changing legs to tubular section improves the performance of the MAR 300 tower. The values decrease in axial forces in all
the elements of the shaft, the cables and the support reactions. The percentages for each element forces are:
24% in legs
10% in horizontal elements
11% in diagonals
15-18% in cables
22% in the support reactions at the base
11-22% in the anchors of cables
3. Changes in the width of the cross section shaft produce significant changes in the structural behavior of the MAR 300
tower, reflected as follows:
The increased in the width of the cross section of the shaft not improve the structural behavior of the tower. Increases occurred
both in the interior forces of the elements of the tower and the support reactions.
Decreasing the cross section of the shaft produces lower axial forces in the cables in both diameters and legs. In horizontal
members there were no significant changes and for the diagonals occurs an increase in the compression axial forces.
4. From the comparison of the Patch Load method with the static method of the Cuban Standard, it is concluded that the
internal forces increase in all elements of the implementation of the first method.
REFERENCES
1. Gantes, C., et al., Modeling, loading, and preliminary design considerations for tall guyed towers. Computers &
structures, 1993. 49(5): p. 797-805.
2. Oficina Nacional de Normalización, Carga de viento. Método de cálculo NC 285:2003. 2003 Ciudad de La Habana,
Cuba.
3. British Standards Institution, Lattice towers and masts Part 4: Code of practice for loading of guyed mast, BS 8100-
4:1995. 1995: London, UK.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015
16
4. STØTTRUP-ANDERSEN, U. Masts and Towers. in Proceedings of the International Association for Shell and Spatial
Structures (IASS) Symposium 2009,Evolution and Trends in Design, Analysis and Construction of Shell and Spatial
Structures. 2009. Valencia, España.
5. Elena, V., Influencia de la asimetría y la no linealidad de los cables en la vulnerabilidad estructural de torres
atirantadas., in Departamento de Ingeniería Civil. 2008, Instituto Superior Politécnico José Antonio Echeverría: La
Habana, Cuba.
6. Nielsen, M. Guyed Masts Exposed to Guy Failure. in Structures Congress 2006: Structural Engineering and Public
Safety. 2006 Caracas, Venezuela: CD ROM, ISBN: 0-7844-0791-6.
7. Gerstoft, P. and A.G. Davenport, A simplified method for dynamic analysis of a guyed mast. Journal of Wind
Engineering and Industrial Aerodynamics, 1986. 23(1-3): p. 487-499.
8. Sparling, B.F., The dynamic behavior of guys and guyed masts in turbulent winds, in Department of Civil Engineering.
1995, The University of Western Ontario: London, Ontario.
9. Smith, B.W., Communication structures. First ed, ed. A. Thomas Telford. 2007, London, Great Britain.
10. Davenport, A.G., Gust loading factors. Journal of Structural Division. ASCE, 1967. Vol. 93 p. 11-34.
11. Davenport, A.G. and B.F. Sparling, Dynamic gust response factors for guyed towers. Journal of Wind Engineering and
Industrial Aerodynamics, 1992. Vol 41: p. 2237-2248.
12. Loredo-Souza, A.M. and A.G. Davenport, The influence of the design methodology in the response of transmission
towers to wind loading. Journal of Wind Engineering and Industrial Aerodynamics, 2003. 91: p. 995–1005.
13. International Association of Shell and Spatial Structures, Recommendations for the Design and Analysis of Guyed
Masts. 1981, International Association of Shell and Spatial Structures: Madrid, España.
14. Kahla, N.B., Equivalent Beam-Column Analysis of Guyed Towers. Computers & Structures, 1995. 55(4): p. 631-645.
15. American Institute of Steel Construction, Load and Resistance Factor Design Specification for Structural Steel
Buildings. 1999: Chicago, Illinois.
16. TIA-EIA/222-F, Normas estructurales para torres y estructuras de acero para antenas. 1996, Telecommunications
Industry Association, USA.
14th International Conference on Wind Engineering – Porto Alegre, Brazil – June 21-26, 2015