1st Issue of Civil Engineering Journal

Civil Engineering

Journal

Editor in Chief:

Dr. M. R. Kavianpour

K.N.Toosi University of Technology (Iran)

Executive Manager:

Dr. O. Aminoroayaie Yamini


Senior Editor:

Msc. S. Hooman Mousavi


Editorial Board Members:

Prof. Dintie S. Mahamah

St. Martin's University (USA)

Dr. Kartik Venkataraman

Tarleton State University (USA)

Dr. Tanya Igneva

University of ACEG (Bulgaria)

Dr. Daniele Bocchiol

Polytechnic University of Milan (Italy)

Dr. Michele Iervolino

Second University of Naples (Italy)

Dr. Rouzbeh Nazari

Rowan University (USA)

Prof. Marta Bottero

Polytechnic University of Turin (Italy)

Dr. Xinqun Zhu

University of Western Sydney (Australia)

Dr. Srinivas Allena

Washington State University (USA)

Chris A. ORiordan-Adjah (PHD Candidate)

University of Central Florida (USA)

Dr. Yasser Khodair

Bradley University (USA)

Dr. Weidong Wu

University of Tennessee - Chattanooga (USA)

Dr. Viviana Letelier Gonzlez

University of the Frontera (Chile)

Dr. Paola Antonaci

Polytechnic University of Turin (Italy)

Dr. Davorin Penava

University of Osijek (Croatia)

Dr. Ricardo Monteiro

IUSS-Pavia (Italy)

Contents

Page 1-9

Application of Spatial Structures in Bridges Deck

Mohammad Hossein Taghizadeh, Alaeddin Behravesh

Page 9-18

Site Locating For Inspection Posts of Freight Cars in Railway

Network Using Analytical Hierarchy Process (AHP) and Geo-

graphic Information System (GIS) (CaseStudy: Iranian Rail-

way Network)

Saeed Monajjem, Mohammad Mahanpoor, Mohammad Sadathoseini

Page 19-30

Designing Manhole in Water Transmission Lines Using

Flow3D Numerical Model

Azin Movahedi, Ali Delavari, Massoud Farahi

Page 31-36

An Examination of Crash Severity Differences Between Male

and Female Drivers, Using Logistic Regression Model

Alireza Pakgohara, Mojtaba Kazemi

Page 37-49

Simulation of Flow Suspended Load in Weirs by Using

Flow3D Model

Mehdi Taghavia, Hesam Ghodousi

Mailing Address: Dr. Kavianpour office, 3rd Floor of Civil Engineering Faculty, K.N.Toosi University of Technology, No. 1346, ValiAsr St, Mirdamad Intersection, Tehran, Iran

Phone: +98-21-88779623 Fax: . +98-21-88779674 E-mail: [email protected] Website: www.CivileJournal.org

Vol. 1, No. 1, November, 2015

Dr. Jiliang Li

Purdue University North Central (USA)

Dr. Yaqi Wanyan

Texas Southern University (USA)

Dr. Jalil Kianfar

St. Louis University (USA)

Dr. Jorge Leandro

University of Bochum (Germany)

Dr. Saeed Khorram

Eastern Mediterranean University (Cyprus)

Prof. Nikolaos Eliou

University of Thessaly (Greece)

1st Issue www.CivileJournal.org

Focus and Scope

Civil Engineering Journal (C.E.J) is a multidisciplinary, an open-access, internationally double-blind peer-reviewed journal concerned with all aspects of civil engineering, which include but

are not necessarily restricted to:

Special Issues

Special Issues deal with more focused topics with high current interest falling within the scope of the

journal in which they are published. Special Issue proposals are welcome at any time during the year.

For most of the civil engineering conferences it is possible to submit papers presented at the conference

for subsequent publication in special issues of the C.E.J.

Civil Engineering Journal (C.E.J) is published monthly.

Civil Engineering Journal (C.E.J) has fast peer review process (3-4 weeks).

Civil Engineering Journal (C.E.J) International Editorial Board

Civ

il E

ngi

nee

rin

g Jo

urn

al

I

Building Materials and Structures Coastal and Harbor Engineering

Constructions Technology Constructions Economy and Management

Earthquake Engineering Environmental Engineering

Renovation of Buildings Geotechnical Engineering

Highway Engineering Hydraulic and Hydraulic Structures

Road and Bridge Engineering Structural Engineering

Surveying and Geo-Spatial Engineering Transportation Engineering

Tunnel Engineering Urban Engineering and Economy

Water Resources Engineering Urban Drainage

Available online at www.CivileJournal.org

Civil Engineering Journal


1

Application of Spatial Structures in Bridge Deck

Mohammad Hossein Taghizadeha*

, Alaeddin Behraveshb

a Ph.D. Student, Department of Civil Engineering, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran.

bProfessor, Department of Civil Engineering, University of Tabriz, Tabriz, Iran.

Received 23 October 2015; Accepted 22 November 2015

Abstract

Spatial structure is a truss-like, lightweight and rigid structure with a regular geometric form. Usually from these

structures is used in covering of long-span roofs. But these structures due to the lightness, ease and expedite of

implementation are a suitable replacement for bridge deck. However steel and concrete is commonly used to build bridge

deck, but heavy weight of steel and concrete decks and impossibility of making them as long-span bridge deck is caused

engineers to thinks about new material that besides lightness and ease of implementation, provide an acceptable

resistance against applied loads including both dead load and dynamic load caused by the passage of motor vehicles.

Therefore, the purpose of this paper is design and analysis bridge deck thats made of double-layer spatial frames compared with steel and concrete deck. Then allowable deflections due to dead and live loads, weight of bridge in any

model and also economic and environmental aspects of this idea is checked. As a result, it can be said that the use of

spatial structures in bridge deck is lead to build bridge with long spans, reducing the material and consequently reducing

the structural weight and economic savings. For geometric shape of the spatial structure bridge is used of Formian 2.0

software and for analysis of bridges is used of SAP2000 with finite element method (FEM).

Keywords: Spatial Structures, Bridge Deck, Steel and Concrete Deck, Finite Element Method, Deflection.

1. Introduction

In architecture and structural engineering, a spatial frame or spatial structure is a truss like, lightweight rigid

structure constructed from interlocking struts in a geometric pattern. Spatial frames can be used to span large areas

with few interior supports. Like the truss, a spatial frame is strong because of the inherent rigidity of the triangle,

flexing loads are transmitted as tension and compression loads along the length of each strut. A spatial frame truss is a

three-dimensional framework of members pinned at their ends. A tetrahedron shape is the simplest spatial truss,

consisting of six members which meet at four joints. Large planar structures may be composed from tetrahedrons with

common edges and they are also employed in the base structures of large free-standing power line pylons [1]. The

simplest form of spatial frame is a horizontal slab of interlocking square pyramids and tetrahedron built from

aluminum or tubular steel struts. Architects and engineers are always seeking new ways of solving the problem of

spatial enclosure. With the industrialization and development of the modern world there is a demand for efficient and

adaptable long-span structures. Spatial grid structures are a valuable tool for the architect or engineer in search for new

forms, owing to their wide diversity and flexibility. Before entering into a discussion of the design and use of spatial

grids in the late twentieth century, it is useful to look back at the early use of three-dimensional structures [2]. Until the

middle of the eighteenth century the main construction materials available to architects and engineers were stone,

wood and brick. Metals, being in relatively short supply, were used mainly for jointing of the other materials. Of the

widely available materials, stone and brick are strong in compression but weak in tension. Thus they are suitable for

three-dimensional structural forms such as domes and vaults. Impressive feats of vaulting were achieved by medieval

masons but the largest span masonry domes, St Peters Basilica in Rome (158893) and Santa Maria del Fiore in

Florence (142034) are both approximately 42 m diameter at the base. Good quality timber has strength in tension and

compression but is naturally available only in limited lengths and with limited cross-section [3]. For large-scale three-

* Corresponding author: [email protected]

Civil Engineering Journal Vol. 1, No. 1, November, 2015

2

dimensional structures jointing of timber becomes a major problem. Nevertheless, the Todai-ji temple at Nara in

Japan, the largest historic timber building in the world, is 57 m by 50 m and 47 m high. Although these materials were

used to produce impressive large-scale structures, the spans were limited and the construction heavy. However, with

the Industrial Revolution came the wider production of iron and then steel, high-strength materials that permitted the

construction of more delicate structures of longer span or greater height [4]. At approximately the same time,

mathematical techniques were being developed to describe and predict structural behavior and understanding of the

strength of materials was advancing rapidly. Equally, with the advent of the Railway Age and the industrialization of

commodity production come an increasing demand for longer span structures for bridges, stations, storage buildings

and factories. With the wider availability of iron and steel and the demand for larger spans, there came a period of

development of new structural forms, initially a multiplicity of different truss configurations and eventually three-

dimensional spatial grids. Many structural forms including most spatial grid assemblies are modular. The concept and

efficiency of modular building construction was dramatically illustrated, almost 150 years ago, by the design,

fabrication and assembly of the metal framework of the Crystal Palace in Hyde Park, London, for the Great Exhibition

of 1851 [5]. Landmark structures such as the Eiffel Tower in Paris constructed from wrought iron between 1897 and

1899, bear witness to the stability and durability of modular three-dimensional metal construction. The tower, built as

a symbol for the centenary celebration of the French Revolution, and conceived as a temporary structure, has already

survived over 100 years. Sadly, the magnificent 114 m span Galerie des Machines by Contamin and Dutert, built at the

same time adjacent to the tower, has not. Such structures demonstrated the possibilities for the use of iron and steel in

high-rise and long-span buildings and challenged the ingenuity of architects and engineers to discover new and more

efficient ways for their construction. Probably the earliest examples of what we now commonly call spatial frames or

spatial grids (light, strong, three-dimensional, mass-produced, modular structures) were developed by the inventor of

the telephone, Alexander Graham Bell (18471922) [6]. In the first decade of the twentieth century he experimented

with spatial trusses composed of octahedral and tetrahedral units. Despite Bells development of lightweight three-

dimensional spatial trusses early in the century, they were not used in architecture until the introduction of the MERO

system, in 1943. This was the first spatial grid system widely available commercially and was developed in Germany

by Mengeringhausen (190388) [7]. Using what is still probably the most common method of spatial truss

construction the system consists of individual tubular members connected at ball-shaped node joints. The aesthetic

appeal and popularity of this system has endured to the present day, as confirmed by the many alternative tube and ball

systems now available. But the applications of spatial structures are not limited often used as roof for long-spans and

three-dimensional spatial structures can be used to build the bridge deck with long-spans [8]. So in this paper, a

double-layer spatial grid deck in which the node system is used compares with I-shaped steel bridges and concrete

deck bridges in terms of span deflection, structural weight and economic and environmental aspects.

2. Modelling of Decks

For geometric shape of the spatial structure bridge is used of Formian 2.0 software and for analysis of bridges is

used of SAP2000 with finite element method (FEM). For all models, including concrete, steel and spatial structure

deck are carried out the linear static analysis. These models are consists of single-span bridge with total length 33 m

and two-span bridge with length 33 m for each span. All decks are consists of two crossing line for passage of

vehicles. Design of this bridges are according to the AASHTO (5th Edition, 2010) design standards (LRFD).

2.1. Steel Deck

Figures 1 and 2 show the examples of single-span and two-span bridges that consist of the I-shaped steel beams in

combined with the concrete slab. According to figures 3, this deck consists of six beams and the total height of deck

that including the steel beam and upper concrete slab (the steel-concrete composite slab) is about 223 cm.

Figure 1. Single-Span Steel Deck Figure 2. Two-Span Steel Deck


3

Figure 3. Executive Details of Steel Deck

2.2. Concrete Deck

Figures 4 and 5 show the examples of single-span and two-span concrete bridge consist of precast concrete beams.

According to figure 6, the deck of this bridge is made of six concrete beams and total height of deck is equal to 220 cm

in both cases.

Figure 4. Single-Span Concrete Deck Figure 5. Two-Span Concrete Deck

Figure 6. Executive Details of Concrete Deck

2.3. Spatial Structure Deck

Figures 7 and 8 show the examples of double-layer spatial grid bridge in case of single-span and two-span.

Topology of these two-layer spatial grid deck is square on moved square and as seen in figure 9, the MERO system is

intended for the connection between members. Because of the connections between members in the spatial grid deck

is joint connections type, so there is no bending moment in the members and available forces in the members are axial

forces [9]. It should be noted, the upper concrete slab is located on top layer nodes of the spatial grid deck and to

prevent of bending moments development in the members, this concrete slab have no contact with the upper members

of the double-layer spatial structure [10].

Figure 7. Single-Span Spatial Structure Deck Figure 8. Two-Span Spatial Structure Deck


4

Figure 9. Executive Details of Spatial Structure Deck

2.4. Wearing Surface Dead Load

In the steel and concrete decks, dead load of wearing surface is applied to the concrete slab and is assumed to be

equally distributed to each girder. A wearing surface with a thickness of 20 cm is assumed. In fact, dead load is due to

concrete slab weight with thickness of 20 cm and weight of asphalt on it. According to equation (1), dead load of the

wearing surface for each girder is equal to multiplying unit weight of concrete slab by thickness of slab and dividing

by the number of girders gives the following:

Wearing Surface Dead Load (for each girder) = 0.25 x 2500

6= 104.17 kg/m2 (1)

But in spatial structures decks, dead load of the wearing surface that consists of concrete slab weight is applied to

the all nodes of structure and is assumed to be equally distributed to each node. In fact, no load is applied to the

members and as a result, the bending moments in the members is zero. Hence, there are only the axial forces in the

members. The studied spatial structure deck in this article have 48 nodes in top layer, so according to equation (2),

dead load of wearing surface for each node is equal to multiplying unit weight of concrete slab by thickness of slab

and dividing by the numbers of nodes gives the following:

Wearing Surface Dead Load (for each node) = 0.25 x 2500

48= 13.02 kg/m2 (2)

2.5. Vehicular Live Loads

The AASHTO LRFD (5th Edition, 2010) Specifications consider live loads to consist of gravity loads, wheel load

impact (dynamic load allowance), braking forces, centrifugal forces, and vehicular collision forces. Live loads are

applied to the composite section. In positive bending regions, the composite section is comprised of the steel girder

and the effective width of the concrete slab, which is converted into an equivalent area of steel by multiplying the

width by the modular ratio between steel and concrete. In other words, a modular ratio of n is used for short-term loads

where creep effects are not relevant. In negative bending regions the short-term composite section consists of the steel

girder and the longitudinal reinforcing steel, except for live-load deflection and fatigue requirements in which the

concrete deck may be considered in both negative and positive bending. The AASHTO LRFD (5th Edition, 2010)

vehicular live loading is designated as the HL-93 loading and is a combination of the design truck or tandem plus the

design lane load. The design truck is composed of an 35.58 KN lead axle spaced 4.27 m from the closer of two 142.34

KN rear axles, which have a variable axle spacing of 4.27 m to 9.14 m. The transverse spacing of the wheels is 1.83 m.

The design truck occupies a 3.05 m lane width and is positioned within the design lane to produce the maximum force

effects, but may be no closer than 61 cm from the edge of the design lane, except for in the design of the deck

overhang. In fact, the vehicular live loads are considered as follows [11]:

Load of truck (400 KN) three-axis with length 10m.

Load of tank truck (900 KN) six-axis.

Uniform loads about 15 KN/M on each crossing lane.

Concentrated load about 160 KN on each crossing lane (In unsuitable position).


5

3. Numerical Explain and Discussion

According to the Code of Practice for Skeletal Steel Space Structures of Iran (No.400), for spatial structures

bridge, the allowable deflection for dead loads is equal to

240 of the bridges span length. According to equation (3), the

dead load deflection must be under this amount [12].

(dead)

240 L(span) (3)

And the allowable deflection for live loads is equal to

000 of the bridge span length. According to equation (4), the

live load deflection must be under this amount [12].

(Live)

000 L(span) (4)

3.1. The Obtained Results for Single-Span Bridges

Figure 10 show the weight of double-layer spatial structures deck compared to the weight of steel and concrete

decks in case of single-span. According to the obtained results, the weight of double-layer spatial grid deck is

substantially less than weight of steel and concrete decks. In fact, the lightness of spatial structures is one of most

important advantages of decks built with these structures compared to other structural systems [13].

Figure 10. The Weight of Double-Layer Spatial Structure Deck Compared to

Steel and Concrete Decks in Case of Single-Span (in %)

Figures 11, 12 and 13 show deformation and allowable deflection of the steel, concrete and spatial structure deck in

case of single-span.

Figure 11. Deformation of Single-Span Steel Deck Figure 12. Deformation of Single-Span Concrete Deck

Figure 13. Deformation of Single-Span Spatial Structures Deck


6

Figure 14 show the structural and allowable deflection for the mentioned bridges. As seen, the structural deflection

in double-layer spatial grid deck is more than steel and concrete decks. Because of the allowable deflection in the

spatial structure deck is equal to

240 of span length for dead loads and

000 of span length for live loads, hence,

according to equations (5) and (6), the allowable deflection for the double-layer spatial grid deck with length of span

30 m is equal to 12.5 cm for dead loads and 3 cm for live loads. Since the structural deflection of this double-layer

spatial grid deck is below of allowable deflection range, so this bridge is stable in terms of deflection due to dead and

live loads. Increasing the structural deflection in the double-layer spatial grid deck compared of steel and concrete

decks is due to use of joint connections in spatial grid deck and also the more rigidity of steel and concrete decks.

Certainly in the double-layer spatial grid deck, the freely movement of elements is more than steel and concrete decks.

Therefore the structural deflection is increased [14], [15].

(Dead)

240 L(span) =

240 x 30 = 12.5 cm (5)

(Live)

000 L(span) =

000 x 30 = 3 cm (6)

Figure 14. Structural and Allowable Deflections for the Steel, Concrete and

Spatial Structures Decks in Case of Single-Span

3.2. The Obtained Results for Two-Span Bridges

Figure 15 show an estimate of deck weight percent thats made of double-layer spatial structure compared to steel

and concrete decks in case of single-span. According to this obtained results, the weight of double-layer spatial grid

deck is substantially less than steel and concrete decks. Because of the spatial structures is lighter than another

prevalent structural systems.

Figure 15. The Weight of Double-Layer Spatial Structure Deck Compared to

Steel and Concrete Decks in Case of Two-Span (in %)

Figures 16, 17 and 18 show deformation and allowable deflection of the steel, concrete and spatial structure deck in

case of two-span.


7

Figure 16. Deformation of Two-Span Steel Deck Figure 17. Deformation of Two-Span Concrete Deck

Figure 18. Deformation of Two-Span Spatial Structures Deck

As seen in figure 19, the structural deflection in double-layer spatial grid deck is more than steel and concrete

decks. Because of the connections between members in the spatial grid deck is joint connections type, the structural

deflection in double-layer spatial grid deck is more than steel and concrete decks. Because in double-layer spatial grid

deck, the freely movement of elements is more than steel and concrete decks. Therefore the structural deflection is

substantially increased. As seen, the structural deflection in the double-layer spatial grid deck is below of allowable

deflection range, so this bridge is stable in terms of deflection due to live loads.

Figure 19. Structural and Allowable Deflections for the Steel, Concrete and Spatial Structures Decks in Case of Two-Span

4. Conclusions

Use of spatial structures in bridge deck is a better approach to build bridge with long spans, reducing material and

consequently reducing structural weight and economic savings. Also use of these structures can lead to ease and

expedite construction operations where need to build bridge (e.g. for the military purposes) in the shortest possible

time. In other side, because the construction operations are done on the ground, so use of these structures in bridge

deck is lead to decreasing dangers of work in height.

5. References

[1] Reis, A. J. "Bridge decks: composite systems for improved aesthetics and environmental impact." In Proc. 3rd Int. Meeting on

Composite Bridges, pp. 645-59. 2001.

[2] Fu. Bridge design and evaluation: John Wiley & Sons, 2013.

[3] Braz, J. "Composite truss bridge decks." PhD diss., Msc. Thesis, ISTTU Lisbon, 2009.

[4] Dauner, Hans G., A. Oribasi, and D. Wery. "The Lully Viaduct, a composite bridge with steel tube truss." Journal of

constructional steel research 46 (1998): 67-68.

[5] Reis, Antnio, Oliveira Pedro, and J. Jos. "Composite truss bridges: new trends, design and research." Steel Construction 4, no.

3 (2011): 176-182.


8

[6] Makowski, Z. S. "Analysis, design and construction of double-layer grids, 1981." Applied Science, London.

[7] Koushky, Dehdashti, and Fiouz. "Nonlinear analysis of double-layer grids with compositive nodes under symmetric and

unsymmetrical gravity loads." International Journal of Space Structures 22, no. 2 (2007): 133-140.

[8] Sheidaii, M. R., K. Abedi, and A. Behravesh. "An investigation into the Collapse behavior of double layer grid space structure."

In Annual LUSAS User Conference for the Construction Industry, pp. 2-3.

[9] Sheidaii, M. R., K. Abedi, and A. Behravesh. "Collapse Behaviour of Double Layer Space Trusses." In IASS Symposium 2001:

International Symposium on Theory, Design and Realization of Shell and Spatial Structures, Nagoya, Japan, 9-13 Oct. 2001, pp.

220-221. 2001.

[10] Sheidaii, M. R., K. Abedi, A. Behravesh, and G. A. R. Parke. "An investigation into the collapse behaviour of double-layer

space trusses." Iranian Journal of Science and Technology 27, no. B 1 (2003): 7-20.

[11] American Association of State and Highway Transportation Officials (AASHTO): LRFD Bridge Design Specifications. 6th

ed. AASHTO. 2012.

[12] Vice presidency for Strategic Planning and Supervision: Code of Practice for Skeletal Steel Space Structures of Iran, No 400,

2010.

[13] Taniguchi, and Saka. "Effect of Covering Plates on Buckling Behaviour of Double Layer Grids." Proceeding of Asia-Pacific

Conference on Shell and Spatial Structures, (1996).

[14] Supple, W. J., and I. Collins. "Limit state analysis of double layer grids." The Analysis, Design and Construction of Double

Layer Girds (1981): 93-117.

[15] Levy, R., A. Hanaor, and N. Rizzuto. "Experimental investigation of prestressing in double-layer grids." International Journal

of Space Structures 9, no. 1 (1994): 21-26.




9

Site Locating for Inspection Posts of Freight Cars in Railway

Network Using Analytical Hierarchy Process (AHP) and Geographic

Information System (GIS)

(Case Study: Iranian Railway Network)

Saeed Monajjema, Mohammad Mahanpoor

b*, Mohammad Sadathoseini

c

a Associate professor of K.N.Toosi University of Technology, Faculty of Civil Engineering, Tehran, Iran.

bPhD. student of Road and Transportation, K.N.Toosi University of Technology, Faculty of Civil Engineering, Tehran, Iran.

cAssistant professor, Head of Iranian Railway Research Center, Tehran, Iran.


Abstract

Freight car inspection and maintenance system have an undeniable role in total costs imposed on system for repair and

rehabilitation process of different components of it. This issue shows its importance when railway transportation comes

to competition with other modes of transportation. In this competition, lower total cost means more demands and more

benefits for optimum systems. Using preventive maintenance methods for rolling stock are among appropriate solutions

in order to lower the costs. These methods require to have an exact monitoring system to achieve a reliable scope of

system. Inspection posts play an essential role as wise eyes on inspection system.

Using Analytical Hierarchy Process (AHP) in this article has developed decision tree including goals, criteria, sub

criteria and alternatives. The main criteria are 1-traffic, 2-geographical position, 3-loction of station on railway network,

and 4-repair and maneuver equipment of station. These criteria and sub criteria have been weighted and quantified using

experts opinion. The use of Geographical information system, 403 stations had been evaluated with 26 criteria and sub

criteria and prioritized. By considering coverage of network in next step, 43 stations are recommended as required station

numbers in railway network to provide 70.53% Coverage of railway network traffic.

Keywords: Analytical Hierarchy Process (AHP), Geographical Information System (GIS), Freight-car, Inspection posts.

1. Introduction and research background

The aim of a site-selection is to find the optimum location that satisfies a number of predetermined selection factors.

The process of problem solving typically involves two main stages: screening and evaluating. The first stage is to

identify limited numbers of candidate sites, from a broad geographical area, taking into account the selection criteria.

The second stage includes careful examination of alternatives to find the most appropriate site (Chang, Parvathinathan

and Breeden 2008). The second stage is an important issue; for example, in the waste management the selection of an

appropriate solid waste landfill requires to be considered by multiple alternatives and evaluation criteria (Guiqin,

Li,Guoxue and Lijun 2009). In recent years, several decision-making methods have been proposed for different site

selection applications. For example, Ballis (2003) used the analytical hierarchy process (AHP) for an airport-site

selection on the Island of Samothraki, Greece. Also, Guiqin et al. (2009) applied geographical information systems

(GIS) and AHP to solve the problem of selecting a landfill site for solid waste in Beijing, China. Similarly, Vahidnia,



10

Alesheikh (2009), and Alimohammadi (2009) suggested a fuzzy AHP method for determining the optimum site for a

hospital.

This paper explores the problem of finding the optimum location of a railway station to provide service as

inspection posts in Iranian railway network, using a hierarchy structure. We introduced the inspection posts site-

selection problem as a hierarchy model consisting of four levels, each with its own main criteria. The main criteria are:

(1) traffic related, (2) repair equipment (3) position on railway networks and (4) geographical position. Each of these

main criteria is then divided into several sub criteria, giving a total number of 26 sub criteria. In addition, the hierarchy

model has 403 railway stations, as candidates or alternatives. We use expert judgment to perform them in individual

pair wise comparisons in the AHP. Furthermore, in next steps the number of inspection posts, are determined with

traffic coverage of these stations.

Iranian railway network with about 7192 kilometers length, connects four corners of Iran to each other. Located in

heart of Middle East, its east-west corridor acts as a link that connects Europe to Asian countries and in north-south

corridor, connects Caspian Sea to Persian Gulf. Its strategic location and more than 20*106 ton-kilometers of freight

transportation, the maintenance issue plays a vital role in total costs of transportation. With good inspection program it

would be possible to reduce the costs of damages, equipment failures and derailments. Inspection posts, as wise eyes

of maintenance system, have an important role in this matter. The main issue to watch the system is the location of

these posts to ensure best performance of system.

2. Site selecting by Analytical Hierarchy process

The AHP, initiated by Saaty (1980), is a flexible multi-criteria decision-making methodology that transforms a

complex problem into a hierarchy with respect to one criterion or more.

The AHP method has been used for a wide variety of decision making problems in the fields such as government,

business, industry, healthcare, and education (Boroushaki and Malczewski 2008; Forman and Gass 2001; Jyrki et al.,

2008; Linkov, Satterstrom, Steevens, Ferguson and Pleus 2007; Raharjo, Xie, and Brombacher 2009; Saaty 2008), and

also for site selection problems. For example, Ballis (2003) used the AHP method for an airport-site selection on the

Island of Samothraki, Greece, Korpela, Lehmusvaara, and Nisonen (2007) selected a warehouse operator network

using a combination of the AHP and DEA methods. Also, Onut and Soner (2008) used the method for trans-shipment

site selection. Rosenberg and Esnard (2008) used a hybrid version for a transit site selection. Furthermore, Hsu, Tsai,

and Wu (2009) used the method to analyze tourist choice of destination, Dagdeviren, Yavuz, and Kilinc (2009) to

analyze the problem of weapon selection, Garcia-Cascales and Lamata (2009) to choose a cleaning system for engine

maintenance. The AHP method requires the following pair wise comparison matrix, A, which contains the relative

weights of the criteria:

(

)

(1)

Where wi is the importance weight of the i-th criteria with respect to goal, or the importance weight of the i-th sub-

criteria (i = 1,. . ., n) with respect to criteria and so on. Furthermore, the importance weights can be obtained using the

following equation (Saaty 1980, 2008):

(2)

Where is the maximum Eigen value of the matrix and w= (w1, w2 . . . , wn) is the corresponding eigenvector of A.

3. Application of ArcGIS on site locating

Numerous researches have been done based on Geographical information system on site locating for different sites

and facilities. Delevar and Naghibi (2003) have been investigated pipeline routing using geospatial information. Ocalir

and et al. presented an integrated model of GIS and Fuzzy Logic (FMOTS) for location decisions of taxicab stands.

Beheshtifar (2006) has investigated Site selection of thermal power plants using GIS system. Alesheikh and et al.

(2008) did a Land assessment for flood spreading site selection using geospatial information system.


11

In this paper, for geographical analysis on railway network, a model of Iranian Railway Network on ARCGIS

(version 9.3) has been developed. In GIS software the spatial analysis, data process and inquiry extraction has been

used in space analysis and for slope analysis of railway network a DEM (Digital Elevation Map) with 40*40 meter

pixels has been used.

Using GIS also helped to manage huge comprehensive data for 403 stations (about 52 data for each) that without

using this software was quite impossible. It is also used for data generating of slopes and distance calculations in C-1-

1, C-1-2, C-3-3, C-3-1, and C-4-1criteia, furthermore for data management in C-1-3, C-2-3 data.

4. Identifying the criteria and sub criteria

First step is obtaining an insight to frame work of inspection posts. Tasks that have been accomplished in these sites

are divided into 3 main categories:

1- Reaction-base activities; such as orders to change failed components of rolling stock,

2- Detection-base activities; such as checking the components for defects and watching the repair schedules of wagons.

3- Prediction-base activities; such as brake control and pre-failure change of components.

To be congruous with their frame work of posts the decision making tree is proposed in Figure 1 and would be

illustrated in next section.

Figure 1. Proposed decision making tree

The criteria are considered to be relevant with Framework of inspection posts. These criteria are also considered to

take into account these three factors: reactions, detection and prevention activities of inspection posts.

5. Description of multi objective decision making hierarchy

The components of decision making process have been divided into four main categories named C-1to C-4 from

main branches into more sub-categories. each part would be more discussed in this section.

C-1: Traffic related includes:

C-1-1: Freight car traffic (except dangerous materials): This item is one of the most important tasks in

inspection post selection because it is one of the indices that inspection post performance on network can be measured.

This item includes three main sub criteria that are:

C-1-1-1: Loading in station: according to Iranian railway code, it is essential to inspect the train after loading. If

there is an inspection post in loading center this task will be done with inspection post technicians. Otherwise it should


12

be done with train managers. It is obvious that inspection with train manager is more time consuming and not as exact

as inspection post assess of rolling stock.Figure1 shows the loading tonnage of each station.

C-1-1-2: Unloading in station: for unloading, train should be detached and discharge its loads. Then it should be

connected together and form the train again. This train should be inspected before re-entrance to railway network. And

like loading stations, it is better to be there and do the inspection posts there as well.

C1-1-3: Passing traffic of station: This item plays a reaction and detection role on passing trains. When a train

enters the inspection post, the inspector should go to entrance zone of station and check the cars while train enters the

station.Figure5 shows the passing traffic in railway network.

C1-2: Dangerous material freight car traffic: This item will be investigated and weighted separately with C-1-3

because loading, transporting and unloading of dangerous materials need more care and investigation; The sub-criteria

are same as C-1-3:

C-1-2-1: Loading in station: The same as C-1-1

C-1-2-2: Unloading in station: The same as C-1-2

C1-2-3: Passing traffic of station: The same as C-1-3

C-1-3: Characteristics of passing wagons; every type of trains needs special care and investigation. it is different

in type and time of inspection. More vulnerable types need more attention and inspection post is more necessary for

these types.

C-2: Geographical position

C-1-1: Dominant slope: This item is an index for vulnerability of covered length (distance between stations and its

adjacent stations). Sharper slopes need more intact brake system, coupling, axle and wheel system to prevent disaster.

C-1-2: Slope change: This sub-criteria is assumed to consider the effect of slope variation where the extra tension

force is needed (extra locomotive). It results in train to dissection and re-formation. According to rules, the brake test

in every train re-formation is mandatory, so existence of inspection posts have logical justification.

C-1-3: Environmental vulnerability of covered route: More vulnerable routes cause more loss if an accident

occurs. So the necessity of inspection post in this region will be increased by preventive performance of inspection

posts.

C-3: Station position on railway network: Includes:

C-3-1: Coverage length: As stated before, this length is defined as the distance between station and its adjacent

stations. More coverage length means more need of inspection in long distance between stations.

C-3-2: History of failures of rolling stock in covered route: Needless to say, failure repetition means there is a

need for more inspection of system in vulnerable routes. Figure 8 shows accident and failure repetition in railway

network.

C-3-3: Intersecting lanes position: If a station locates in a junction, it would provide service for two or three

routes instead of single route.

C-3-4: Curves on covered length: Existence of curves with less than 1000 meter diameter means that there are

some vulnerable points in block. The frequency of these curves on a block increases the risk of accidents or derailment

if these curves are followed with defects on rolling stock. Figure 6 shows the frequency of curves on blocks with

under 1000 meter radius in railway network.

C-4: Maneuver or repair equipment of station: Which includes:

C-4-1: Distance with repair shops: According to detection duty of station post, if a wreck has been detected or

deadline for repair has been arrived, it should be sent to a repair shop that are categorized into:

C-4-1-1: Four-year repair shops: In Iranian repair schedule the fundamental repairs should be done in every 4

year period. Nine repair shops and zone that every station provide service shown in Figure 3.

C-4-1-2: Annual repair shops: Some minor check and repair actions have been done annually in freight-cars to

prevent accumulation of wagons in repair shops. The undercover region of each repair shop (for 1 and 4 year repair

service) proposed considering capacity of repair, loading tonnage of undercover stations and neighborhood of every

repair shop.


13

C-4-1-2: Unscheduled repair shops: These repair shops are more prevalent in network than scheduled repair

shops (26 stations in total) but are less equipped. every station provides support for nearest stations because of its

unpredictable nature of failure and necessity of repairing (in contrast with scheduled repair shops).

C-4-2: Locomotive existence: If a freight car has been recognized as a failed car it should be separated from

wagons and sent to repair shops. Existence of locomotive accelerates this action.

C-4-3: Total length of repair and maneuver lanes: Maneuver line length of every station provide space for

changing line, stopping and repair of trains. Furthermore it is implies the size and equipment existence of a station.

6. Determining AHP weights

The proposed hierarchy model for the railway station site-selection problem is shown in figure1, where the overall

objective is in the first level. Also, the figure shows the main criteria in the second level, 26 sub criteria in the third,

and five potential stations in the last level. According to the AHP method, the elements of each level are pair wise

compared with the element in the next higher level. This results in a number of pair wise comparison matrices (Saaty

1980, 2008). We assess the importance of the i-th criteria against the j-th criteria using the following five-point

assessment values:

(i) If criteria i and j are equally important then the corresponding element of the comparison matrix will

be

.

If criteria i be moderately more important than criteria j then

&

.

If criteria i be extremely more important than criteria j then

&

.

Also, values 2 and 4 are used to show an intermediate importance between the criteria. Using the hierarchy model

and the criteria we developed standard questionnaires that were filled by railway transportation experts, inspection

posts directors and college professors in railway fields. Also 2 economists and 2 environmental experts had been

interviewed about consequences, advantages and disadvantages of site selection for inspection posts. With

accomplishing a pair wise comparison and with help of Expert Choice (EC) software, valid data (with inconsistency

rate less than 0.1) has been taken into account and final weights has been calculated. Final local and overall weights

have been presented in Table 1 to 9.

Table 1. Main criteria comparison results

criteria weight

C-1 0.685

C-2 0.104

C-3 0.085

C-4 0.126

Total inconsistency rate = 0.026

Table 2-5. Sub criteria (level 1) priority comparison results

Table 2.

sub criteria Weight

C-1-1 0.55

C-1-2 0.24

C-1-3 0.79


Table 3.

sub criteria weight

C-2-1 0.41

C-2-2 0.38

C-2-3 0.21


Table 4.

sub criteria weight

C-4-1 0.77

C-4-2 0.11

C-4-3 0.12


Table 5.

criteria weight

C-3-1 0.19

C-3-2 0.33

C-3-3 0.43

C-3-4 0.05



14

Table 6-8: Sub criteria (level 2) priority comparison results

Table 6.

criteria weight

C-1-1-1 0.71

C-1-1-2 0.11

C-1-1-3 0.18


Table 7.

criteria weight

C-1-2-1 0.75

C-1-2-2 0.15

C-1-2-3 0.10


Table 8.

criteria weight

C-4-1-1 0.13

C-4-1-2 0.15

C-4-1-3 0.71


Table 9. Total weights of end-branches of decision making tree

Figure 2. Annual and four-year repair shops and

covered route of each shop

Figure 3. Existing repair shops

Sub-criteria Criteria weight Sub criteria

weight(1)

Sub criteria

weight(2) Total weight

C-1-1-1 0.685 0.55 0.71 0.267493

C-1-1-2 0.685 0.55 0.11 0.041443

C-1-1-3 0.685 0.55 0.18 0.067815

C-1-2-1 0.685 0.24 0.75 0.1233

C-1-2-2 0.685 0.24 0.15 0.02466

C-1-2-3 0.685 0.24 0.10 0.01644

C-1-3 0.685 0.79 - 0.54115

C-2-1 0.104 0.41 - 0.04264

C-2-2 0.104 0.38 - 0.03952

C-2-3 0.104 0.21 - 0.02184

C-3-1 0.085 0.19 - 0.01615

C-3-2 0.085 0.33 - 0.02805

C-3-3 0.085 0.43 - 0.03655

C-3-4 0.085 0.05 - 0.00425

C-4-1-1 0.126 0.77 0.13 0.012613

C-4-1-1 0.126 0.77 0.15 0.014553

C-4-1-1 0.126 0.77 0.71 0.068884

C-4-2 0.126 0.11 - 0.01386

C-4-3 0.126 0.12 - 0.01512


15

Figure 4. Loading (tonnage) in stations Figure 5. Passing traffic

Figure 6. Curvatures rate in railway corridors Figure 7. Slopes rate in railway corridors

Figure 8. Failure and accident occurrence in different corridors of Iran railway network


16

7. Determining number of stations in railway network

Main idea for determining quantity of inspection posts is highest coverage of system. Traffic coverage is the most

important factor that its relative importance has been proved by experts pair-wise comparison results (Table 9).

Therefore for each station the network coverage of freight cars total transport has been calculated (different types of

traffics that described in C-1-1 and C-1-2). This coverage has been sorted by priority of each station that has been

determined in previous chapters, and the accumulative traffic coverage versus station numbers has been calculated.

This diagram presents Figure 9.

Figure 9. Increase trend of network coverage with increasing inspection posts.

As shown in Figure 9, we can cover 70.53% of loading for 43 recommended stations in railway network. After this

station, because other factors have more proportion in prioritization of stations comparing with traffic tasks in 13

stations we have very mild increase in coverage; so with respect to economical evaluation, this increase in station

numbers is out of financial justification. Therefore with respect to the total number of 43 stations and prioritization

order of stations, the final arrangement of stations is suggested in Figure 10.

Figure 10. Position of recommended inspection posts within Iranian railway network

0

10

20

30

40

50

60

70

80

90

100

0 50 100 150 200 250 300 350 400

Acc

um

ula

ted

Ne

two

rk C

ove

rage

(%

)

Numbers of Posts


17

8. Summary and conclusion

In this research, main factions of inspection posts were taken into accounts, a decision making tree was suggested

and with Analytical hierarchy process and pair-wise comparison the weight of each factor was gained. Then 403

stations were prioritized using GIS software. By considering coverage length of network, as most important factor and

one of the leading important tasks that can figure out with site locating of inspection posts, the total number of 43

station posts with provided priority were recommended.

9. References

[1] Saaty, Thomas L. "The analytic hierarchy process: planning, priority setting, resources allocation." New York:

McGraw (1980).

[2] Beheshtifar M. Site selection the thermal power plants. Master thesis of GIS, K.N. Toosi University of

Technology;2006; p.36-50.

[3] Poormohamadi M. Schematization of urban land use. Tehran, Iran: SAMT publication; 2003, p.70-71

[4] Ocalir, Ebru Vesile, Ozge Yalciner Ercoskun, and Rifat Tur. "An integrated model of GIS and fuzzy logic

(FMOTS) for location decisions of taxicab stands." Expert Systems with Applications 37, no. 7 (2010): 4892-4901.

[5] Alesheikh, Ali Asghar, Mohammad Jafar Soltani, Nahal Nouri, and M. Khalilzadeh. "Land assessment for flood

spreading site selection using geospatial information system." International Journal of Environmental Science &

Technology 5, no. 4 (2008): 455-462.

[6] Raharjo, Hendry, Min Xie, and Aarnout C. Brombacher. "On modeling dynamic priorities in the analytic

hierarchy process using compositional data analysis." European Journal of Operational Research 194, no. 3 (2009):

834-846.

[7] Saaty, Thomas L. "Relative measurement and its generalization in decision making why pairwise comparisons

are central in mathematics for the measurement of intangible factors the analytic hierarchy/network process."

RACSAM-Revista de la Real Academia de Ciencias Exactas, Fisicas y Naturales. Serie A. Matematicas 102, no. 2

(2008): 251-318.

[8] Vahidnia, Mohammad H., Ali A. Alesheikh, and Abbas Alimohammadi. "Hospital site selection using fuzzy

AHP and its derivatives." Journal of environmental management 90, no. 10 (2009): 3048-3056.

[9] Forman, Ernest H., and Saul I. Gass. "The analytic hierarchy process-an exposition." Operations research 49,

no. 4 (2001): 469-486.

[10] Garca-Cascales, Mara Socorro, and Mara Teresa Lamata. "Selection of a cleaning system for engine

maintenance based on the analytic hierarchy process." Computers & Industrial Engineering 56, no. 4 (2009): 1442-

1451.

[11] Wang, Guiqin, Li Qin, Guoxue Li, and Lijun Chen. "Landfill site selection using spatial information

technologies and AHP: a case study in Beijing, China." Journal of environmental management 90, no. 8 (2009): 2414-

2421.

[12] Korpela, Jukka, Antti Lehmusvaara, and Jukka Nisonen. "Warehouse operator selection by combining AHP

and DEA methodologies." International Journal of Production Economics 108, no. 1 (2007): 135-142.

[13] Linkov, Igor, F. Kyle Satterstrom, Jeffery Steevens, Elizabeth Ferguson, and Richard C. Pleus. "Multi-criteria

decision analysis and environmental risk assessment for nanomaterials." Journal of Nanoparticle Research 9, no. 4

(2007): 543-554.

[14] Nijkamp, P. (2004). Transport systems and policy. USA: Edward Elgar Publishing, Inc.Onut, S., &Soner, S.

(2008). Transshipment site selection using the AHP andTOPSIS approaches under fuzzy environment. Waste

Management, 28(9),15521559.

[15] American Planning Association. Planning and urban design standards. John Wiley & Sons, 2006.

[16] Boroushaki, Soheil, and Jacek Malczewski. "Implementing an extension of the analytical hierarchy process

using ordered weighted averaging operators with fuzzy quantifiers in ArcGIS." Computers & Geosciences 34, no. 4

(2008): 399-410.


18

[17] Chang, Ni-Bin, G. Parvathinathan, and Jeff B. Breeden. "Combining GIS with fuzzy multicriteria decision-

making for landfill siting in a fast-growing urban region." Journal of environmental management 87, no. 1 (2008):

139-153.

[18] Charnes, Abraham, William W. Cooper, and Edwardo Rhodes. "Measuring the efficiency of decision making

units." European journal of operational research 2, no. 6 (1978): 429-444.

[19] Cooper, William W., Lawrence M. Seiford, and Kaoru Tone. Data envelopment analysis: a comprehensive text

with models, applications, references and DEA-solver software. Springer Science & Business Media, 2007.

[20] Emrouznejad, Ali, Barnett R. Parker, and Gabriel Tavares. "Evaluation of research in efficiency and

productivity: A survey and analysis of the first 30 years of scholarly literature in DEA." Socio-economic planning

sciences 42, no. 3 (2008): 151-157.

[21] Ballis, Athanasios. "Airport site selection based on multicriteria analysis: the case study of the island of

Samothraki." Operational Research 3, no. 3 (2003): 261-279.

[22] Dadeviren, Metin, Serkan Yavuz, and Nevzat Kln. "Weapon selection using the AHP and TOPSIS

methods under fuzzy environment." Expert Systems with Applications 36, no. 4 (2009): 8143-8151.




19

Designing Manhole in Water Transmission Lines Using

Flow3D Numerical Model

Azin Movahedia*

, Ali Delavarib and Massoud Farahi

c

a M.Sc. Faculty of Civil Engineering, K.N. Toosi University of Technology, Tehran, Iran

bB.Sc. Irrigation Engineering, Uremia University, The Manager of Water and Water Waste Installations Affairs of Moshaver Yekam

Engineers Company, Tehran, Iran

c B.Sc. Faculty of Mechanical Engineering, K.N. Toosi University of Technology, The Designer Expert in Moshaver Yekam Company,

Tehran, Iran


Abstract

Using cascades and drops existing in flow path has a history of 3000 years. Particularly, Roman engineers employed

stepped spillways with the same idea in several countries; however, there are few information about the hydraulic

performance of aqueducts. Most of these channels have flat long cross sections with low torsions (variable slope) such

that they can encompass cascade and steep spillways or dopshaft. Given that there are few studies conducted on

dropshafts, the present paper attempted to discuss about such structures in flow path and water transmission lines as well

as introducing the existing principles and relations and present, the obtained results of designing though Flow3D. The

obtained error percentage was about 20% which is acceptable for numerical studies.

Keywords: Drop manhole, Vertical shaft, Projectile, Finite volume, Flow3D.

1. Introduction

In studies related to Roman structures, cost and time of implementing projects depended on various issues such as

tunnels, piers, arcades, raised foundations, and siphon. Roman projects have been completed during 3 to 15 years with

an average cost of 23 to 69 million dollars each kilometer. Their structures have been designed for low discharge

flows (0.2-2 m3/s) and low longitudinal slopes (about 1-3 m each kilometer, on average) [1-4]. Their studies include

the three following areas:

Smooth sharp shots

Stepped channels

Cascades and dropshafts

Using the third alternative (dropshaft) as the main branch of their channel requires a certain science of engineering

and is considered as new designs. Hydraulically, dropshafts includes the followings:

The possibility of implementing vertical drop in balance of trade

Kinetic energy dissipation of fluid flow

Flow aeration

In the first case, dropshaft will allow the relation between two flat channels which are placed in various trades in a

very short distance from each other. The second case of these structures uses is kinetic energy dissipation of fluid

flow which is used to optimize the performance of structures and prevent scouring and erosion of downstream

*Corresponding author: [email protected]


20

hydraulic structures [5, 6 and 7]. The third case is flow aeration which is used to prevent cavitation and corrosion of

water duct. Figure 1 shows a schematic of dropshaft. Table 1 also presents a summary of Roman studies on water

channels [8].

Table 1. A summary of Roman studies on water channels [2]


21

Figure 1. Schematic of drop manholes [1]

2. Introducing Flow Conditions in the Model

The purpose of the present study is to design a dropshaft which can be circular, rectangular or square. Since

concreting square dropshaft is easier than concreting circular dropshaft, square dropshaft is the base of designing in

the present paper. The height difference of upstream and downstream pipes (drop height) is 3 m and the diameter of

upstream and downstream pipes is 1600 mm with the slope of 0.01 (1%).

Discharge of the design is 6 3/. It is fed by its upstream spillway and falls into an open channel with the length

of 18.5 m. The purpose of designing is to reach a state of flow regime which firstly, transmission lines pipes contain

fluid as much as 75% of cross section (80% of the pipe diameter), i.e. flow is not under the pressure in the pipes;

secondly, the length of dropshaft should be in such a way that the projected fluid jet from the upper pipe to the

downstream wall have slight or no collision. The height of dropshafts, depending on the depth of burial in soil, is

different and its other dimensions are implemented as tip across transmission line in 5 points of a 500-m-path. Finally,

after passing this path by the fluid, transmission line and dropshafts are discharged into a river in downstream. Flow in

the upstream of dropshaft and in conversions channel is supercritical and enters into the dropshaft with a normal depth

of 20 cm, critical height of 51 cm (according to Eq. 1), velocity of about 4.5 m/s, and the slope of 0.02 (2%).

Therefore, it can be stated that flow in pipe lines have supercritical flow which requires appropriate design to supply

the objectives of the design. Figure 2 shows a schematic of hydraulic parameters used in the governing relations.

Critical depth in rectangular channel: =

(1)

Where indicates critical depth of flow in the open channel (m); indicates flow discharge (3/) and b

indicates the width of channel (m) which is 8 m in this research.

Figure 2. The employed parameters [1]


22

Selecting appropriate flow regime in designing is perform to prevent the collision of the jet projected to lower

wall of dropshaft since the collision of this jet to the wall causes corrosion and destruction of the wall and disturbs the

dropshaft use by the time pass. In the following, a schematic of various flow regimes has been presented:

Figure 3. Various flow regimes and the place of projectiles collision with the wall [1]

3. Relations Governing Flow

The relations which should be investigated in the present study are presented. To start computations, using the

available relations, the primary dimensions for the dropshaft are assumed. After investigating the governing relations,

the available model is simulated through Flow3D to compare the results. Selecting flow regime of , the relations are

as following [2]:

(2)

Where, L indicates length and width of the dropshaft (m).

Substituting the critical depth value, the length of the dropshaft will be (3.4 m and 14.5 m) considering

affordability of the design; the length of the design is regarded 4.5 m. Since the best performance of the dropshaft for

supercritical regimes is considered , for the accuracy of the primary assumption of the dropshaft length selection,

the following control relation can be used [1]:


23

(3)

(

)

Where denotes the velocity of inlet flow in the dropshaft and H indicates the height of fluid drop (m). Other

parameters have been previously introduced. Substituting the study dimensions in Eq. 3, heterogeneity assumption is

established, confirming the accuracy of the primary assumption in the dropshafts dimensions selection. Another

important dimension is p value which can be extracted from the next graph:

Figure 4. The non-dimensionalized height of the dropshaft relative to dc/h and Dab/h

Since the 80% of the lower pipes diameter should be filled with fluid, thus / = . Given to Figure 2, it is

clear that = . Accordingly, p is considered 1.7 m. Therefore, the dropshaft has been designed with the length and width of 4.5 m and the height (p) of 1.7 m. These dimensions should be simulated by Flow3D to confirm flow model in the pipe and manhole.

4. Flow3D Numerical Model

In Flow3D model, the equations governing fluid flow includes continuity and momentum equations. Flow

continuity equation is obtained from the law of conservation of mass as well as by writing mass balance equation for

compressible and viscous fluid simple element. In general, this equation is written as following:

(4)

( )

( )

( ) =

Where VF indicates the ratio of the volume of the fluid passing through an element to total volume of the element

and denotes the density of the element. Velocity components (u,v,w) are in (x,y,z) directions. Ax indicates the ratio of

the area of the fluid passing through an element to total area of the element at the direction of x. Ay and Az, similarly,

are flow levels at the directions of y and z. Navier-Stokes equations of fluid with velocity components of (u,v,w) have

been shown in 3-dimensional coordinates:

V {

} =

P

G f

(5)

V {

} =

P

G f

V {

} =

P

G f


24

In these equations, (Gx, Gy, Gz) are the terms of mass acceleration and (fx, fy, fz) are the terms of viscous

acceleration [10 and 11]. To numerically simulate the equations governing these flows, Flow3D is employed. Flow3D

is powerful software in CFD area. Flow3D has been designed for 1-2-3 dimensional problems. One of the main

advantages of this software for hydraulic analyses is the ability of modeling flows with free surface. Free surface

refers to the interval between gas and liquid. Free surface is simulated using Volume of fluid (VOF) [12]. Flow

environment is divided into networks with fixed rectangular cells such that for each cell, there are average values of

dependent quantities. On other words, all variables are computed at the center of the cell except that velocity which is

computed at the center of cell faces. In this software, two numerical techniques have been used for geometrical

simulation [10].

1. Volume of fluid (VOF) method: it is used to show the behavior of fluid at free surface and includes the following three components [13]:

Displaying the position of surface

Meshing

Boundary conditions of surface

2. Fractional Area-Volume Obstacle Representation (FAVOR): it is employed to simulate surfaces and rigid volumes such as geometrical boundaries. For numerical modeling, turbulence numerical model is required. In

Flow3D, for this purpose, five turbulence models have been introduced including: Prandtl's Mixing Length, k- one

and tow-equations, RNG models, and Large Eddy Simulation Model. RNG-based models less rely on constant

empirical figures. RNG model uses equations which are similar to k- turbulence model equations. Constant values of

the equation which have been practically received in the standard mode of k- have been taken from RNG. The

presence of an additional term in the equation causes the increase of computations accuracy in strain flow in RNG

model. RNG, compared to standard k- model, has a higher efficiency in strain flow; and unlike the standard model,

analytical relation is used to determine Prandtl turbulence figures. Therefore, this model has an appropriate accuracy

in low Reynolds numbers and it is more considered to determine turbulence values of flow in curved fields or

geometrical complexity [14]. Accordingly, RNG has been used in the present study.

5. Simulation Results

In this section, the rigid model constructed in Solidworks 2011 software is referred. Then, its geometrical meshing

and Boundary conditions are discussed. Finally, the obtained results are presented. In this model, three meshing blocks

have been selected. The first block pertains to conversion channel at upstream with the length of 18.5 m. in inlet

section of channel, volume flow rate with the discharge of 6 3/ and outlet section has symmetry Boundary

conditions. Other faces also have wall Boundary conditions. The second block which encompasses dropshaft has

symmetry condition in inlet and outlet sections and walls have been also defined. The third block which encompasses

about 14 m of transmission line has symmetry inlet and outlet Boundary conditions as well as other conditions of wall.

The following figure shows Boundary conditions:

Figure 5. The Boundary conditions applied to the numerical model


25

After selecting appropriate dimensions of meshing and obtaining steady flow, the extracted numerical results are

investigated, the following diagrams show steady flow state in which flow reaches a balance and does not change after

40 sec. These figures indicate that simulation time for this study has been appropriately selected (80 sec) and it can

also be decreased to 45 sec.

Figure 6. The diagram of fluid volume and flow stability by the time pass

Figure 7. Diagram of inlet and outlet flow passing through Boundaries by the time pass; inlet Boundary (right) and outlet

Boundary (left)

Selecting appropriate dimensions of meshing cells causes that the curvatures of the rigid model are well modeled

and the rigid model is closer to the real states (the simulation accuracy is increased). Figure 8 shows a schematic of

FAVOR method to observe the rigid model after meshing which indicates its closeness to real geometry model.

Figure 8. FAVOR method in modeling the rigid model geometry


26

In the following figures, flow pattern and hydraulic parameters (depth, velocity, and Froude number) in

transmission line are presented:

Figure 9. Flow depth at final moment of numerical simulation

Figure 10. Flow depth in transmission channel (upstream)

Figure 11. Flow depth in downstream pipe


27

Figure 12. Flow velocity at the final moment of numerical simulation

Figure 13. Flow velocity in transmission channel (upstream)

Figure 14. Flow velocity in downstream pipe


28

Figure 15. Froude number at the final moment of numerical simulation

Figure 16. Froude number in transmission channel (upstream)

Figure 17. Froude number in downstream pipe


29

Figure 18. Schematic of size and direction of velocity vectors in drop manhole

Considering Figures 9 to 17, it is revealed flow pattern has a supercritical trend which flows with a relatively

constant and stable depth in upstream channel. After entering into the dropshaft, this flow pattern with its relatively

high velocity firstly causes flow obstruction in the beginning of the pipe and then, exists from downstream pipe with

free surface flow states. The depth and velocity of the flow in the model is very close to the manually computed value

and it has an error percentage of below 20% (this value is acceptable in numerical studies up to 30% and numerical

results can be relied on to reach these values). Additionally, in the model, there is no trace of projectile jet collision

from the upstream channel to downstream wall. It can be stated that flow regime of R1 is established in the model and

the structure damage is in its minimum state. Finally, it can be said that the dimensions which have been selected from

manual computations method for this design are acceptable values and the design can be confirmed.

6. Conclusion

In the present paper, using relations available in other researches, a drop manhole was designed in downstream

water transmission line of a spillway by formulating the considered research hypotheses and objectives. Then, using

Flow3D, the dropshaft with the height of 3 m and discharge of 6 m3/s was simulated to extract its flow and hydraulic

parameters. The extracted parameters have been well consistent with manual computations values. The obtained error

percentage was about 20% which is acceptable for numerical studies. Figures, flow pattern and hydraulic parameters

were also presented.

7. References

[1] Chanson, Hubert. "Hydraulics of Roman aqueducts: steep chutes, cascades, and dropshafts." American Journal of Archaeology

(2000): 47-72.

[2] Chanson, Hubert. "Hydraulic design of stepped cascades, channels, weirs and spillways." (1995): 1-292.

[3] Fevrier, P.A. "The Roman Army and the Construction of Aqueducts." (1979).

[4] Hodge, A. T., Roman Aqueducts, and Water Supply. "Duckworth." London, United Kingdom (1992).

[5] Jain, Subhash Chandra, and John Fisher Kennedy. Vortex-flow drop structures for the Milwaukee metropolitan sewerage district

inline storage system. Iowa Institute of Hydraulic Research, the University of Iowa, 1983.

[6] Apelt, C. J. "Goonyella railway duplication drop structures and energy dissipators at culvert outlets. Model studies." Report


30

CH27 84 (1984).

[7] Rajaratnam, N., A. Mainali, and C. Y. Hsung. "Observations on flow in vertical dropshafts in urban drainage systems." Journal

of Environmental Engineering 123, no. 5 (1997): 486-491.

[8] Ervine, D. A., and A. A. Ahmed. "A Scaling relationship for a two-dimensional vertical dropshaft." In Proc. Intl. Conf. on

Hydraulic Modelling of Civil Engineering Structures, pp. 195-214. 1982.

[9] Chanson, Hubert. "Energy dissipation and drop structures in ancient times: the Roman dropshafts." In Water 99 Joint Congress,

25th Hydrology & Water Resources Symposium and 2nd International Conference on Water Resources & Environmental Research,

vol. 2, pp. 987-992. 1999.

[10] Hoseini, A. and Abdipour, A. Numerical modeling of velocity profile in continuous muddy flows and investigating the effect

of slope, concentration and discharge on the flow. Civil engineering magazine of Azad Islamic University, the 3rd year, no. 3.

(2010).

[11] Flow Science. "FLOW-3D Users Manuals Version 9.2.1." Flow Science, Inc., Los Alamos, New Mexico, USA. (2008).

[12] Baghdadi, H.; Ershadi, S. and Rostami, M. Numerical investigation of topical scouring due to sought horizontal directions

using Flow3D. The 10th Iranian hydraulic conference, Gilan University. (2011).

[13] habibi, M. and Khanjani, M. J. Investigating scouring phenomenon in long vertical shoap drop and comparing laboratory

model with the results obtained by Flow3D, the 8th international congress on civil engineering, Shiraz University. (2009).

[14] Movahedi, A.; Kavianpour, M. R. and Aminoroayayi Yamini, A. Nurmerical analysis of downstream scouring hole of cup

launchers using Flow3D, international conference on civil engineering, architecture and sustainable development, Tabriz. (2013).




31

An Examination of Crash Severity Differences Between Male

and Female Drivers, Using Logistic Regression Model

Alireza Pakgohara, Mojtaba Kazemi

b*

a PhD Student, Department of Statistics, International Campus of Ferdowsi University of Mashhad, Iran.

bMSc. Highway and Transportation engineering, Department of civil engineering, Roudsar and Amlash branch, Islamic Azad

University, Roudsar, Iran


Abstract

One person in every 2539 people gets killed and one in every 253 suffers injuries due to driving crashes each year in Iran. Such that driving incidents are second rank factor of death and the first rank reason for lost lifetimes in this country. 60% of total incidents which lead to deaths or injuries are actually driving incidents in Iran. That is while the same ratio is only 25% worldwide average. In this article, we report a probabilistic relationship between vehicle drivers gender and severity of the accidents. The model accuracy rate is more than 91%. Coefficient values show that if an crash happens and all other variables are under control, the probability of suffering injuries for a man is 1.597 times more than for a woman (1.40 1.79, 99% CI) in comparison with the case that the person does not get injured at all. Similarly, the probability of death for a man is 1.462 times higher than for a woman (1.13-1.79, 90% CI) again in comparison with case of no injury at all.

Keywords: Gender, Road Crashes, Crash Severity, Logistic Regression.

1. Introduction

Taking into account the 6,342,000 population of the world in 2004, one person in every 5,285 dies and one in every

127 people suffers injuries due to driving incidents each year. These figures are 2,539 and 253 respectively (Pakgohar,

2012). Total yearly direct and indirect costs imposed by driving crashes amount to 180,000 billion Rials. This estimate

cost amount is equal to 6.23% of GDP of the country in 2007. Since the GDP growth rate was 6.7% that year it can be

concluded that driving crashes cost swallows almost all the growth of GDP in Iran. A statistical report from Health

Ministry shows that driving crash is the second rank cause of death and first rank cause of lost lifetime in Iran (average

world statistics show this factor in rank nine). 60% of total incidents which cause death or injury in this country are

actually driving incidents, while worldwide average is only 25% (Pourmoalem and Ghorbani, 2011). Humans are

different in terms of physical, psychological, social, and recognition abilities. This is true in driving as well; Such that

people with higher sensory skills, lower reaction time, and higher precision are more successful in driving. Sensation

seeking is a personal characteristic which influences peoples driving behavior. The person in this case tends to

experience new things and risks for them. Males and females are different in this sense (Soori, 2005). Significant

differences have been observed in other traits like intelligent cognition and sensation (Esmaily, 2010).

This research addresses the effect of persons gender on crash severity. Binary logistic regression method is used in this research to tackle this job. The level of probability for occurrence of some situation can be determined using this specific method.



32

2. Research background

The primary goal of the research was to examine how human factors influence crash severity prediction and

categorization in Iran. Data regarding crashes happened in 2007 were used for this task. Results obtained using tree

regression and logistic regression methods suggested that having driving license, using seat belt, age, and gender all

influence the severity of road crashess as human factor indices (Pakgohar et al., 2011).

Waylen and McKenna (2000) have shown that crash involvement patterns are different for two genders. Men are more

probable to get involved in crashes on the bends, low light or overturn situations, while women on the other hand are

more prone to get involved in crashes at the intersections and junctions than men (Waylen 2000)

Men perform better on assessing time intervals. Although women recognize the movement faster than men; but they

estimate the distance shorter. These differences are reasons why female drivers keep longer distances and break faster

and bore severely at emergencies. As the studies have shown, women react faster than men to dangerous situations, but

sometimes act to control the vehicle, for example turn the steering wheel and apply breaks with more delay in

comparison to men, because of physical conditions (Leen, 2004). Landaur and colleagues (1980) believe that on

average, women react faster than men on a time task. Reaction time average was 0.485 sec. for women and 0.534 for

men. Parker and Lajunen, (2001) study has shown that aggressive driving is far more seen in men than women

(Esmaily, 2012).

3. Methodology

Usually it takes to use logistic regression when multi-value data are to be processed as dependent variables. Especially,

this method is more prominent in organic assay and audit analysis. On the other hand, since prediction by logistic

regression is in fact some type of classification, we can set audit analysis in this framework as well. Logistic

regression is one of the most applicable generalized linear models used to analyze relations between one or more

descriptive variable/s and a scalar response variable. Therefore, Logistic regression opens wider fields of statistical

analysis before peoples eyes (Mojtaba Kazemi, 2011).

Our methodology is categorized as Descriptive Research in the realm of social studies and descriptive-analytic in

terms of viewpoint and problem addressing. The statistical analyses used in this paper include statistical descriptive

measures such as average, percents, etc and Logistic regression (LR) model.

We use LR as the primary model to recognize patterns of crash severity applied to the driver based on driver gender.

Many papers have used Logit model (for example, Pakgohar and Khalili (2010), Pakgohar et al., (2011), and Esmaeili

et al., 2012). There are a number of reasons to use this method. First, the Logit model has been widely used and well

developed. Second, it is relatively easy to understand and is integrated readily in most software packages. Our last

reason is that the Logit model is well known as an accurate and reliable tool for predictions.

Logistic Regression mode is a nonlinear transformation of linear regression model (LN transformation). The logistic

distribution is an S shaped distribution function similar to standard normal distribution. Like multivariate regression,

researchers are interested to find a suitable arrangement for predictor variables which helps with interpreting binary

results.

With logistic regression, the probability of occurrence for a certain event is directly estimated. In case only one

predictor exists, logistic regression can be formulated as:

Occurrence probability = ( )

(1)

In which 0 and 1 are coefficients that would be estimated using data (original samples), and x is the predictor. The formulation with more than one predictor variable is as follows:

Occurrence probability =

(2)

In which

Clearly the probability for event not to occur is

. These relations are called multi-variable logistic functions. A

linear pattern at Logit scale would be fitted using Logit transformation introduced above:

( ) (3)


33

So Logit modeling in respect to

is regarded as a linear function of predictors. Equation (4) can be obtained

from this relation (Mojtaba Kazemi, 2011).

( ) [

( )

( )

] (4)

The accident severity applied to the driver is represented by a random binary variable Y in this study which would be

given as:

{

}

In which:

( | ) ( ) (5)

Although Logit model is called non-parametric, but function h postulates are totally parametric in statistical deduction.

Especially h is a logical accumulated distribution function and is formulated as:

( ) ( )

( ) (6)

3.1. Model criteria

Model test

Likelihood ratio test for the main model - which is called chi square test as well - is used for comparing

researchers model against a trivial model as a basis with a constant value. Chi square likelihood ratio test will be given by deducting deviation (-21l) of final (complete) model from deviation of sheer intersection model. Number of degrees of freedom would be equal to number of terms minus 1 for this test. (Munizaga and Alvarez-

Daziano, 2005)

Gauges of data fit information model

Biesian information criterion (BIC) and Akaike coefficient or AIC are general information theory statistics and are

used when we want to compare alternative models. Lower value for them indicates better fit for the models

(Munizaga and Alvarez-Daziano, 2005).

3.2. Model efficiency measures

1. Classification Accuracy: This measure shows the ratio of correct predictions over positive and negative input. This measure is largely dependent on dataset distribution, and therefore can easily lead to wrong results

regarding system efficiency.

2. Classification Sensitivity: This measure evaluates the ratio of true positives, i.e. gives the extent of system ability to predict correct values out of total input items.

3. False positive ratio: In binary regression, the number of wrong predictions in which the dependent variable is predicted to have value 1, but it really has the value 0. This ratio is stated as a percent of total observations.

In multivariate regression, the number of wrong predictions for which the predicted value of the variable is

higher than actual observed value. This is stated as a percent of total items on or above diagonal.

4. False negative ratio: In binary regression, the number of wrong predictions in which the value of 0 is predicted for dependent variable, but the actual observed value is 1. This is stated as a percent of all

observations with value 1. In multivariate logistic regression, the number of wrong predictions in which the

predicted value for dep

1st Issue of Civil Engineering Journal

Documents

Transcript of 1st Issue of Civil Engineering Journal