1. business and industrial
2. construction

Founded : December 1934
IRC Website: www.irc.org.in
The Indian Roads Congress
E-mail: [email protected]
Volume 76-1
JOURNAL OF THE INDIAN ROADS CONGRESS
january-march 2015
ContentS
Page
ISSN 0258-0500
Page
Advertisements
5
Inside Front Cover-JCB India Ltd.
Inside Back Cover-BASF India Ltd.
Outside Back Cover-Unisteel Engineering Works
4
Strata Geosystems (India) Pvt. Ltd.
17
Perma Construction Aids Pvt. Ltd.
24
Spectrum Chemicals
43
Mukesh & Associates
54
Tinna Rubber & Infrastructure Ltd.
Paper No. 628 “Quality of Water– Challenge for Highways”
by
13
by
19
Raja Mistry
Tapas Kumar Roy
Ankita Chowgule
M. Manjunath
M L Gupta
Dhananjay A Bhide
Prashant Dongre
Paper No. 632 “Effect of Utilization of Waste Marble on Indirect
Tensile Strength Properties of Bituminous Concrete Mixes”
by
44
Dr. Umesh Sharma
Paper No. 631 “Replacement of Damaged Suspended Span of Varsova
Bridge Across Vasai Creek on NH-8”
by
36
Dr.Tripta Kumari Goyal
Paper No. 630 “Analysis of T-Beam Skew Bridges under Live
Loads”
by
25
Mahesh Kumar
Paper No. 629 “Utilization of Rice Husk Ash in Hot Mix Asphalt
Concrete as Mineral Filler Replacement”
M.R. Archana
H.S. Sathish
G. Brijesh
Vinay Kumar
Paper No. 633 “Gap Acceptance Behavior of Right-Turning
Vehicles at T-Intersections - A Case Study”
Cover Page- Cable Stayed Bridge, kr Puram, Bangalore
by
Jamnagar House, Shahjahan Road,
New Delhi – 110 011
Tel: Secretary General: +91(11) 2338 6486
Sectt.: (11) 2338 5395, 2338 7140, 2338 4543, 2338 6274
Fax : +91 (11) 2338 1649
Gopal R. Patil
Jayant P. Sangole
Kama Koti Marg, Sector 6, R.K. Puram,
New Delhi – 110 022
Tel : Secretary General : +91 (11) 2618 5303
Sectt. : (11) 2618 5273, 2617 1548, 2671 6778,
2618 5315, 2618 5319, Fax : +91 (11) 2618 3669
No part of this publication may be reproduced by any means without prior written permission from the Secretary General, IRC.
Edited and Published by Shri S.S. Nahar on behalf of the Indian Roads Congress (IRC), New Delhi. Printed by Aravali Printers
& Publishers, Pvt. Ltd, W-30, Okhla Industrial Area, Phase-II, New Delhi on behalf of the Indian Roads Congress. The responsibility
of the contents and the opinions expressed in Journal of the IRC is exclusively of the author/s concerned. IRC and the Editor disclaim
responsibility and liability for any statement or opinion, originality of contents and of any copyright violations by the authors. The opinions
expressed in the papers and contents published in the Journal of the IRC do not necessarily represent the views of the Editor or IRC.
14000 Copies, January-March, 2015 (56 Pages)
Journal of the Indian Roads Congress, January-March 2015
Paper No. 628
QUALITY OF WATER– CHALLENGE FOR HIGHWAYS
Mahesh Kumar*, Dr.Tripta Kumari Goyal** And Dr. Umesh Sharma**
ABSTRACT
Water is one of the most important elements in construction but people still ignore quality aspect of this element. The water
is required for preparation of mortar, mixing of cement concrete and for curing work etc. during construction work. The
quality and quantity of water has much effect on the strength of mortar and cement concrete in construction work.
Water is a vital input both for commercial, industrial and Civil Engineering applications including structural engineering.
Depending upon the quality and characteristic of water, the life span may be influenced varying from 60 to 100 years.
India has 18 percent of the world’s population but it has only 4 percent of water resources of the world. Annual per capita
availability of water has decreased from 6042 cum in the year 1947 to 1545 cum in 2011. Annual per capita availability of
water was 1816 cum in 2001. Annual per capita availability of water will further reduce to 1340 cum by 2025 and to 1140
cum by 2050.
Depleting ground water level in India may be a real worry if one looks at the future demand of water in India. It is estimated
that the country would need 1180 Billion Cubic Meter (BCM) of water annually by 2050. India has, at present, annual
potential of 1123 BCM of utilizable water with 690 BCM coming from surface water resources and remaining 433 BCM
from ground water resources. In view of this projection, the country would not be able to meet its demand unless it recharges
its aquifers and uses water more efficiently and judiciously.
The characteristic of ground water in several parts of north India is changing fast resulting in a reduced life of Reinforced
Concrete structures which may vary from 15 to 30 years depending upon severity of character of ground water. Appropriate
care in use of water during construction and subsequently protecting the structure from the aggressive environment help
in giving a designed life to such like structures. It is a high time when appropriate investigations in respect of aggressive
environment which may be in form of harmful salts or water are conducted before taking the construction work in hand.
Based on these investigations, an appropriate modality need to be defined regarding execution of work.
1.
CONTROL OF ‘WATER’ AS quantities of alkalis, acid, oils, salt, sugar,
CONSTRUCTION INPUT
organic materials, vegetable growth and
other substances that may be deleterious
Water is turning to be a scarce commodity. to bricks, stone, concrete or steel.
It is because of its over exploitation, Potable water is generally considered
mismanagement and also contamination satisfactory for mixing. The pH value
by the inadvertent polluting activities of of water should be not less than 6.
industries and allied bodies.
As a result to avoid the damage to
The water used for mixing and curing existing bridges and culverts before the
should be clean and free from injurious design life of a structure and to maintain
the quality of work process, it is necessary
to investigate the reasons concerning
nature of water so used during the time
of construction. The main working area
at present has been confined to Haryana
in general and the National Capital
Region including Delhi. The remedial
measures in use of water or precautions
during the time of construction have
been suggested.
* (President, IRC) Engineer-in-Chief, Haryana PWD (B&R) Branch,Chandigarh Email: [email protected]
** Associate Professor, Civil Engineering Department, PEC University of Technology, Chandigarh
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
Mahesh kumar, Goyal & Sharma ON
6
In the present paper, it is entailed
to study the system in terms of
geological characteristics governing
salinity and sulphate attack, the
logistical constraints, standardization/
specifications requirements of the PWD/
MORTH and the construction attributes.
Various recommendations in terms
of methodology and improvement in
construction practices have been derived
from experiences of constructional
activities.
2.
SYSTEM ANALYSIS
The following figure (Fig.1) illustrates
System Diagram involving raw materials
and logistical constraints for desired
Quality of work, new experiences
and
learning,
revolving
around
process requirements, constraints and
technological imperatives.
3.
GEOGRAPHICAL STATUS OF available at 37.96 m deep, Similarly
WATER IN HARYANA
Siwan zone 37.65 m, Shahbad zone
36.88 m, Babain zone 34.50 m, Thanesar
The large scale urbanization has taken its zone 30.81 m, Ratia zone 30.65 m, Guhla
toll on the groundwater resources in the zone 29.91 m, Fatehabad zone 29.28 m,
districts falling in the National Capital Jakhal zone 28.28 m, Pehowa zone
Region of Haryana. While painting a 27.87 m, Ladwa zone 27.45 m, Panipat
gloomy picture on the exploitation of zone 24.79 m, Kaithal zone 24.28 m and
ground water, the Central Ground Water similar is the position in Rania, Alewa,
Board (CGWB) has termed 31 zones as Tohana, Samalkha, Pundri, Nilokheri,
Dark Zones (over-exploited areas). The Gharounda, Radaur, Assandh, Nishing,
Millennium City Gurgaon has four areas Ellenabad zone etc.
Gurgaon, Pataudi, Sohna and Farrukh
Nagar declared as the Dark Zones.
It is clear from the above situation that
According to CGWB report, Palwal,
Hodal and Hassanpur zones in the Palwal
district have been declared as Dark
Zones while Tauru, Ferozepur Zhirka
are included in the list in Mewat District,
Faridabad zone in Faridabad District is
also in the Dark Zone.
Fig. 1 : System Perspective of Water Deployment in Civil Infrastructural Development
In Panipat District, Bapoli, Israna,
Madlaunda, Panipat and Samalkha
Zone are in over-exploited category
while Ganaur, Rai, and Sonepat are
in the category in Sonepat District.
Nahar, Bawal, Rewari, and Khol in
Rewari District, Badra, Dadri, Kairu
and Loharu in Bhiwani District and
Ateli, Mahendergarh, Kanina, Nangal
Chaudhary and Narnaul in Mahendergarh
District are the other areas declared as
dark zones by the board.
concern. The level of water is falling
at fast speed in ten districts of paddy
growing areas continuously. In the
recent years, the groundwater level is
falling maximum. The falling of the
ground water level in many blocks in
Fatehabad, Kurukshetra, Kaithal, Karnal,
Panipat, Sonepat, Sirsa, Yamunanagar
and others have reached a very sensitive
stage. Earlier, the groundwater level was
available on an average at 8 m deep but
now the average level is 17 m.
Unauthorized exploitation of ground The most severely affected areas are:water has become a matter of great Sirsa zone where the groundwater is
the groundwater level is being over
exploited. If this over exploitation will
continue, the future of Construction
Industry in Haryana would be dark.
Haryana is a water deficit state with
respect to surface and ground water
resources. The ground water level
in the State particularly in the fresh
water zone is depleting fast due to
heavy exploitation of ground water
and is posing a serious problem.
Increasing demand and scarcity of
Ground Water Resource underlines the
importance of artificial recharge and
water conservation. The State Average
decline in water table from June 1974 to
June 2013 is -7.97 m and June, 1999 to
June, 2013 is -7.80 m. Based on average
decline in water table of Dynamic
Ground Water Resource estimation
as on 31.3.2011, the blocks have been
categorized as Over Exploited, Critical,
Semi Critical and Safe. A block is over
exploited where depletion of ground
water has taken place more than 100
percent, Critical where it is 90-100
percent, Semi Critical where it is 70-90
percent and Safe where it is less than 70
percent. Presently, the number of Over
Exploited, Critical, Semi Critical and
Safe blocks in the State is 68, 21, 9 and
18, respectively. One is quite concerned
about the fast depleting water table as
ground water is a precious resource
and sincere efforts are needed for the
conservation of this natural resource.
Journal of the Indian Roads Congress, January-March 2015
Quality of Water– Challenge for Highways
Table 1 :Districtwise Historical Fluctuation June, 1974 To June, 2013
Sr.
No.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
District
Ambala
Bhiwani
Faridabad
Fatehabad
Gurgaon
Hisar
Jind
Jhajjar
Kurukshetra
Kaithal
Karnal
Mahendergarh
Mewat
Palwal
Panipat
Panchkula
Rohtak
Rewari
Sonepat
Sirsa
Yamunanagar
State Average
Depth of Water Table (m)
June
1974
5.79
21.24
6.42
10.48
6.64
15.47
11.97
6.32
10.21
6.28
5.72
16.11
5.50
5.37
4.56
7.58
6.64
11.75
4.68
17.88
6.26
9.19
June
1999
5.45
16.19
8.71
6.42
15.22
5.87
5.92
4.49
16.72
7.78
7.59
25.01
7.14
5.72
8.53
11.17
3.80
13.07
5.33
9.45
7.13
9.36
Fluctuation (m)
June June 1974 to June 1999 to
2013 June 2013
June 2013
10.70
-4.91
-5.25
20.76
0.48
-4.57
16.25
-9.83
-7.54
23.05
-12.57
-16.63
26.03
-19.39
-10.81
7.35
8.12
-1.48
12.92
-0.95
-7.00
4.57
1.75
-0.08
33.24
-23.03
-16.52
23.07
-16.79
-15.29
17.62
-11.90
-10.03
45.68
-29.57
-20.67
11.21
-5.71
-4.07
9.38
-4.01
-3.66
16.97
-12.41
-8.44
16.42
-8.84
-5.25
3.71
3.92
0.09
22.96
-11.21
-9.89
8.52
-3.84
-3.19
17.34
0.54
-7.89
12.69
-.93
-5.56
17.16
-7.97
-7.80
*(+) indicates rise in water table and (-) indicates decline in water table
7
fluorides can lead to flourosis and dental
disorders.
While lab tests did not reveal the presence
of any pathogenic organisms in the water
so tested but mercury and arsenic beyond
the permissible limit is a matter of grave
concern.
The alarming presence of harmful
substances in ground water can be traced
to the continuous discharge of sewage
and industrial effluents into the Yamuna
and subsequently, into the groundwater
aquifer which, being sandy in nature,
allows pollution to spread at a rapid
rate.
The study has also been done on river
Yamuna in Delhi from Okhla to Mathura
where in 35 samples were taken.
Quantity of cadmium and lead found in
river water has been shown in Table 3. In
Faridabad, cadmium is 0.1 mg/l which is
10 times more than the permissible limit.
In Okhla, it is 4 times higher.
Table 3 : Quantity of Cadmium and
STUDY OF RIVER WATER
Lead in River Water
Palla and Okhla in Delhi. Chemical
ENTERING HARYANA
results of water samples have been
Location
River
Ground
The situation is getting alarming even shown in Table 2.
Water Water within
in respect of drinking water at present.
a Distance of
Shocking as it sounds, pollution has Table 2: Chemical Results of Water
100 m
Sample
trickled through the troubled waters of
Okhla
0.04 mg/L 0.03 mg/L
the Yamuna and percolated the rapidly
POLLUTANT RESULTS MPL*
Faridabad 0.1 mg/L 0.03 mg/L
depleting ground water reserves of
Nitrate
174 mg/L 100 mg/L
Varindavan 0.04 mg/L 0.03 mg/L
Delhi City. The study reveals that the Sulphate
680 mg/L 400 mg/L
concentration of arsenic, mercury, Flouride
Mathura
0.04 mg/L 0.01 mg/L
3.10 mg/L 1.5 mg/L
nitrates, sulphates and dissolved solids Mercury
4.60 mg/L 1 mg/L
in the Capital’s ground water exceeds Arsenic
Near Jasola, the contaminated water
69.5 mg/L 50 mg/L
permissible limit.
of Badarpur and Near Palla Bridge,
*Maximum permissible limit
the contaminated water of Sector 31
Keeping in mind the tremendous pressure These samples were subjected to a and Sector 3 of Faridabad falls in river
exerted by developmental activities on detailed study for the presences of Yamuna. New sewerage plants have
the Yamuna, a detailed impact analysis of chemicals, heavy metals and bacteria. The not been proved to be effective. Also
these activities to ascertain the changes quantum of pollutants detected, in turn chemical factories of city discharge their
in the chemical composition of ground makes for an unhealthy situation. While contaminated water into river Yamuna.
water, which is crucial for a sustainable the excessive presence of dissolved salts
supply of potable water, was carried out. in water affects the kidney and nitrates The above details make it clear
can trigger off the blue baby syndrome that guidelines for use of water for
The study entailed 50 samples of in infants. Besides, an overdose of construction purposes of culverts and
groundwater being lifted from random sulphates can cause gastric problems and bridges on highways as laid down needs
spots along a 22 kms stretch between
to be monitored very closely.
4.
Journal of the Indian Roads Congress, January-March 2015
Mahesh kumar, Goyal & Sharma ON
8
5.
GEOLOGICAL
CHARACTERISTICS
content either with depth at a particular location or with the location of the bore hole
did not show any regular pattern.
The geological survey reports on
the soils in India reveal that the soils on
the drier parts of Punjab, Haryana, North
Bihar, Uttar Pradesh, and Rajasthan tend
to be saline and alkaline efflorescence’s.
The soil contain many undecomposed
rocks and mineral fragments which on
weathering liberate sodium, magnesium
and calcium salts. Such soils are notably
impervious and therefore have impeded
drainage. Large areas, once fertile, have
become impregnated with these salts
(reh, kalar) destroying the value of the
ground. The salts are normally confined
to the top layers of the soil, being
transferred from below by capillary
action. Irrigation by canal water has
resulted soils in the canal irrigated areas
of Punjab, Haryana and elsewhere in the
country to change during the past three or
four decades. The alkali content in those
soils is high and there is a large excess
of free salts, combined with poverty in
nitrogen and organic plant food material.
Such lands pose problems not only for
cultivation but also for construction
of civil structures which are likely to
withstand the attack of salts in soils and
ground waters.
6.
CASE STUDIES OF HARYANA
A RCC structure is normally designed
for 60 to 100 years of life. A severe
damage to various culverts and bridges,
made us to think and comment on the
various aspects of water and eligibility
of water for construction purposes. Case
studies are given below:
6.1 Rohtak and Jhajjar Districts
The ground water table level varied
widely and is in the range of 1.2 to 4.7 m
deep.The pH values of the soil samples
of all depths are ranging from 7.6 to 9.6
and this is slightly in the alkaline range.
The chloride content in the samples
varied from traces to as high as 1.296
percent. Also, the variation of chloride
43 samples out of the 60 soil samples tested from the 22 bore holes and from different
depths contained chlorides in the range upto 0.10 percent, the remaining 17 samples
were in the range of 0.10 to 0.50 percent.
The sulphate content in the soil samples range from traces to as high as 0.556 percent.
Quite similar to the chlorides, no clear pattern of the variation of the sulphate content
either with depth at a particular location, or with the location of the bore hole was
observed. Out of the 60 soil samples tested from the 22 bore holes and at different
depths, 42 samples contained sulphates from traces to 0.1 percent, 12 samples
contained in the range from 0.10 to 0.20 percent and 6 samples contained more
than 0.2 percent. Results of soil samples regarding chloride contents and sulphate
contents are given in Table 4
Table 4 : Results of Soil Samples
No. of Samples
pH Value
Range
60
Samples
from 22 bore
holes tested
7.6 to 9.6
Chloride Contents
No. of
Content
Samples
%
43
0.10
17
0.10 to
0.50
*
Chloride content traces to as high as 1.296%
*
Sulphate content traces to as high as 0.556%
Sulphate Contents
No. of
Content
Samples
%
42
0.10
12
0.10 to
0.20
6
More than
0.20
The sub soil water samples collected at the water table level from 21 bore holes was
found to be neutral to slightly alkaline with pH ranging from 6.65 to 8.35. The water
samples also contained high concentrations of chlorides and sulphates in general. It
was seen that the water samples collected from 10 bore holes contained chlorides
from traces to 0.1 percent, 7 samples in the range 0.10 to 0.50 percent, and 4 samples
contained more than 0.50 percent. The sulphate content in 3 water samples was from
traces to 0.015 percent, in 5 samples in the range 0.015 to 0.10 percent, in 4 samples
in the range 0.10 to 0.20 percent and in 9 samples above 0.20 percent. These results
have been shown in Table 5.
Table 5 : Results of Water Samples
No. of
pH Value Chloride Contents
Sulphate Contents
Samples
Range
No. of
Content % No. of
Content %
Samples
Samples
21 Samples 6.65 to
10
0.10
3
0.015
from 21 bore 8.35
7
0.10 to
5
0.015 to
holes tested
0.50
0.10
4
More than
4
0.10 to
0.50
0.20
9
More than
0.20
Journal of the Indian Roads Congress, January-March 2015
Quality of Water– Challenge for Highways
Results to further analysis of selected
sub soil water samples revealed that
the sub soil water contains considerable
amounts of alkalis particularly as sodium
salts. They also contain small amounts
of calcium and magnesium salts. These
results indicate that the water contained
salts such as chlorides and sulphates of
sodium, calcium and magnesium. The
quantity of magnesium ions, though
much higher than in normal waters,
still seem to be in the safe range for
concrete.
Sulphate (as SO4) contents of the order of
1,000 to 2,000 ppm in the ground water
is considered to be detrimental even to
brickworks of bridges and culverts. As
far as the pH is concerned, the range of
values beyond permissible have effect
on concrete or brickwork of culverts and
bridges.
9
and results showed that vol. of 0.2 N
H2SO4 required to neutralize 100 ml.
of H2O sample using mixed indicator
required 22.1 to 33.57 ml. against limit of
maximum 25 ml. which is incorporated
in IS:456.
These results indicate that the ground
The following conclusions can be drawn
water of Sonepat District is alkaline
from the chemical analysis results on the
in nature and is not meeting the
soil and ground water samples:
requirements of MORTH Specifications.
●● The soil in the ground water is
This situation is more prevalent with
slightly alkaline with pH ranging
an increase in distance from Yamuna
The maximum chloride content in Rohtak
from 7.6 to 9.6.
and Jhajjar districts was 0.5 percent
River. But with minor addition of acids
(5,000 ppm) in the soil and 1.296 percent ●● Considerable amount of chlorides of required normality, the water can be
and sulphates are present in the
(12,960 ppm) in the ground water.
brought to meet with the requirements of
soil samples. The chloride content
ranged from traces to as high as 0.5 above specifications of IS:456.
Simultaneous presence of deleterious
percent and the sulphate content
salts such as sulphates and chlorides
ranged from traces to as high as A sample calculations for bringing
in the soil and ground water in large
water compatible to use for construction
0.556 percent.
quantities are harmful for the concrete
foundations of bridges and culverts, ●● The sub soil water is neutral to purposes is given below:
their possible interaction and the
slightly alkaline in nature with pH
HCl Acid of Specific gravity N
periodic fluctuation of the ground
ranging from 6.65 to 8.35. It also
N1 V1 = N2 V2
water table can further aggravate the
contained considerable amounts
situation arising out of such type of
Acid
of chlorides and sulphates in some Water
salinity attack. For example, presence
samples. The chloride content in the
N x 34 (0.02N=N)
of chlorides will reduce the sulphate
sub soil water samples ranged from N ×100=
1
resistance of a concrete corresponding
50
50 traces to as high as 1.296 percent.
to a certain amount of sulphates present,
The sulphate content ranged from
and periodic fluctuations of water table
traces to as high as 0.49 percent.
N × 34
will enhance the dangers of sulphate
N
(Normality
of
water)
=
1
50 × 100
attack compared to another situation ●● The sub soil samples also contained small amounts of sodium, calcium
where such movements of water table
and magnesium salts.
are absent. Even the requirements of
100
Mol. Wt.
the chemical composition of cement to The chloride and sulphate contents found Equivalent =
=50
=
Valency
2
combat sulphate attack and corrosion of to be present at different locations do not Weight
steel in the presence of chlorides are to indicate any regular pattern of occurrence
be considered on a different footing vis- inter-alia site conditions except that the Strength of Alkalinity in term of CaCo3
à-vis presence of only one of the two. later may be considered as mild to severe Equivalent
The same survey, as referred to above, for concrete foundations of bridges and
had shown that when both sulphates culverts thereby ruling out the approach
34 × 50 gm/L (Eq. Wt. of
and chlorides were present, the more of any rearrangement of the locations
Alkalinity =
imminent cause of reinforcement steel proposed for the different structures on 50 × 100 CaCo3) = 340 mg/L
corrosion leading to distress to concrete considerations of their susceptibility to
Strength (Alkalinity)
foundations was the presence of chemical attack.
Normality =
Equivalent
chlorides and the magnitude of distress
was more dependent upon the chloride 6.2 Sonepat District
340 mg/L – 250 = 90 mg/L
concentration in the soil and ground The ground water samples were also
water rather than sulphates.
studied in respect of Sonepat District
Journal of the Indian Roads Congress, January-March 2015
Mahesh kumar, Goyal & Sharma ON
10
580 mg. of Alkalinity is neutralized with to severe cracking of concrete structures
HCl = 1 ml,
exposed to sulphatic environments. The
cracking also create conditions for the
90 mg. of Alkalinity is neutralized with reinforcement corrosion which in turn
HCl = 1 × 90
may lead to spalling of concrete.
580
Chlorides such as MgCl2 and AICl3 reacts
For 90 mg. of Alkalinity, the requirement with lime and forms thereby unstable
of HCl is 0.155 ml.
and water soluble compounds which are
detrimental to concrete. The chlorides of
1 L of water needs 0.155 ml. to neutralize alkali metals (NaCl, KCl) which do not
the Alkalinity
react with lime or with other components
of the hardened concrete are harmless
1000 L of water need 0.155 ml. to
but in concentrated solutions, they tend
neutralize the
to leach lime from concrete. On the other
hand, the soluble chlorides may induce
Alkalinity 0.155 × 1000 =155 ml.
corrosion of the steel reinforcement
So 155 ml of HCl of Normality N-12 is and as such reinforced concretes
required to be added to 1000 L of water exposed to chloride environments
so as to make it safe for use on bridges need to be protected against attack and
and culverts on highways. Same can vary deterioration.
from area to area.
7.
SALINITY PROBLEMS - EFFECT
ON THE DURABILITY OF
CONCRETE OF BRIDGES AND
CULVERTS
Portland cement on hydration gives rise
mainly to calcium silicate hydrates and
calcium hydroxide. The calcium silicate
hydrates are responsible for the strength
of concrete and the calcium hydroxide
is uniformly distributed in the calcium
silicate hydrate matrix. Acidic waters
with pH below 6 or so dissolve the
calcium hydroxide and also effect the
stability of the calcium silicate hydrate
matrix both leading to deterioration of
concrete. On the other hand alkaline
waters with pH in the range of 7.5 to 10
or so are harmless to concrete.
The sulphates react with the calcium
hydroxide giving rise to gypsum with a
molar volume nearly 2.20 times that of
calcium hydroxide. Also the sulphates
react with the tricalcium aluminate in
portland cement gives rise to a calcium
sulphoaluminate solid solution with a
molar volume 2.5 times the original
volume. Thus, both these reactions lead
Magnesium ions are introduced into
ground water mainly in the form of
MgSO4, MgCl2 and MgHCO3. The
majority of magnesium ions originate
from dolomitic rocks. Surface waters
rarely contain more than 25 mg/L of
Mg++ions whereas the ground water may
contain as much as 300 mg/L As such
in water, the magnesium ion contents
areusually 1.3 g/L
All salts of magnesium with the exception
of hydrocarbonate are destructive to
concrete of bridges and culverts and are
even more aggressive than CaSO4 and
Na2SO4, because in this type of attack,
the entire calcium content of the binding
agent of the concrete may be replaced
gradually by magnesium which may lead
to the deterioration of concrete.
precautions and treatments. The possible
solutions to make durable constructions
will involve proper choice of materials,
adoption of proper construction practices,
quality control and use of protective
coatings or barriers, wherever necessary.
The choice of construction material is as
below:
8.1 Cement
The type, chemical composition and
physical characteristics of the cement
greatly influence the resistance of
plain and reinforced concrete in the
presence of sulphates and chlolrides.
So far as the sulphates are concerned,
calcium aluminate or part of cement are
susceptible to sulphate attack of concrete
and so in such cases, use of sulphate
resisting cements is the obvious solution.
From this point of view, cements in
decreasing order of preference are:
●●
High alumina cement
●●
Super sulphated cement
●●
Sulphate
cement
●●
Portland Pozzolana cement or
Portland slag cement
●●
Ordinary Portland cement
resisting
Portland
Among the various cements listed above,
high alumina cement is not suitable for
use under tropical conditions. Super
sulphated cements are not to be used
above 400C and as such both high
alumina cement and super sulphated
cement should not be considered.
8.MEASURES TO COUNTER Sulphate resisting Portland cement has
been recommended for countering the
THE SALINITY ATTACK
attack of sea waters which contain both
Investigations of the soil and ground sulphates and chlorides in amounts
water of Rohtak and Jhajjar districts, ranging up to 3.65 g/l and 19.00 g/l
as discussed above, shows that the respectively. As such this type of cement
foundation conditions will be aggressive conforming to the specifications for
to the usual concrete and brickwork ASTM type V (ASTM C-150) will be
constructions without any special preferable. Accordingly, the Portland
Journal of the Indian Roads Congress, January-March 2015
Quality of Water– Challenge for Highways
cement should contain C3A not more
than 5 percent and 2C3A+C4AF or solid
solution (4CaO.Al2O3 + 2 CaO.Fe2O3)
not more than 20 percent. A Portland
cement as per IS: 269 with the additional
stipulation of C3A not more than 5 percent
and 2C3A + C4AF (or its solid solution)
up to 20 percent could also be considered
for use. However, in the absence of
this type of cement, the choice may
fall respectively on such Portland slag
cement (IS: 455) and Portland pozzolana
cement (IS: 1489) which are known
to possess sulphate resistance greater
than that of ordinary Portland cement.
However, except when sulphate resistant
portland cement is being used, in case
of use of all other cements including
ordinary Portland cement as per IS: 269,
suitable precautions and protections can
be adopted.
●●
Reinforced
foundations
11
concrete
pile plaster (1:4). On top of the plaster, a coat
of bitumen (blown type grade 85/25,
IS: 702) at the rate of 1.5 kg/m2 on a coat of
The relevant construction practices primer (of asphalt and kerosene in equal
recommended for these three types of proportions) should be applied upto the
foundations are described below:
HFL and also around the base concrete
footing. This type of construction is likely
9.1 Brickwork on Plain Concrete to cost approximately 58 percent more
Footing
than the usual brickwork foundations.
If this type of foundation is adopted, the
concrete footing should be of minimum
M 20 grade (corresponding to 1:1.5:3)
and sulphate resistant Portland cement
or other alternative cements as discussed
above should be used. The bricks should
be of first class having not more than 10
percent water absorption. The mortar
should be of 1:4 proportion preferably
with sulphate resistant cement. The
concrete footing should have a formed
finishso as to result in a more impermeable
construction. Such a construction is likely
8.2 Canal Water
to cost approximately 48 percent more
Water used for mixing and curing than the usual brickwork foundation.
concrete and mortars should not contain
harmful amounts of dissolved salts. The If for some reason, the above type of
ground water samples containing large foundation is not feasible, eg. for nonamounts of deleterious salts should not be availability of bricks having less than
used. Water conforming to requirements 10 percent water absorption or sulphate
of IS: 456 and meeting specifically the resistant cement, then an alternative
limitations of solids content shall only foundation can be adopted. This type of
foundation envisages use of first class
be used.
bricks as per PWD Specifications and
Water from the irrigation canal in Rohtak ordinary Portland cement or Portland
and Jhajjar districts was also tested. The pozzolana cement or Portland slag
sample tested had pH 7.2 and contained cement. In this case, the foundation
chlorides and sulphates in traces only should be excavated, 75 mm thick limebrickbat concrete in proportions 1:2:5
and was found fit for construction use.
(one part lime, two parts of sand and five
9. CONSTRUCTION PRACTICES parts of overburnt brick bats by volume)
FOR
BRIDGES
AND should be laid and consolidated. The top
CULVERTS
surface should be finished smooth with
Cement-lime-sand plaster. 1:1:6 (one part
The foundations for the various structures cement, one part lime and six parts sand).
in Rohtak and Jhajjar districts can be of When this surface has dried, it should
one of the following types:
be coated with two coats of bitumen.
Then the usual base concrete (M 20 or
●● Brickwork on plain concrete 1:1.5:3) for the footing shall be laid. This
footing
concrete should be form finished. The
brickwork in cement mortar 1:4 should
●● Reinforced concrete foundations
be provided with 12 mm thick cement
9.2 Reinforced Concrete Foundations
for Bridges and Culverts
Such concrete foundations should be
made with proper type of cement of
richer proportions and should have
adequate cover and protective coatings.
Cement used should preferably be
sulphate resistant type. If the same is
not available, any other type of cement
as discussed above may be used. The
minimum cement content in the mix
should be at least 370 kg/m3 when
sulphate resistant cement is used or
400 Kg/m3 when any other cement is
used for a nominal maximum size of
aggregate of 20 mm. If for some reason
the nominal size of aggregate has to be
restricted to 10 mm, the cement content
in the mix shall be increased by 50 Kg/
m3. The water-cement ratio shall not
exceed 0.45. The concrete cover upto the
plinth level shall be increased by 25 mm
over and above the minimum specified
in IS: 456. The concrete foundation shall
rest on 75 mm thick lime sand brickbat
concrete (1:2:5) finished smooth with
12 mm thick cement-lime-sand plaster
(1:1:6), as in the case of brickwork
foundations. Concrete of bridges and
culverts shall be protected all around
including on the top of the plaster, upto the
plinth level, with two coats of bitumen.
This type of construction is likely to cost
approximately 37 percent more than the
usual (unprotected) reinforced concrete
foundation as per PWD Specifications.
In this case, however, only sulphate
resistant Portland cement should be
used.
Journal of the Indian Roads Congress, January-March 2015
12
Mahesh kumar, Goyal & Sharma ON Quality of Water– Challenge for Highways
If more durable foundations are needed,
then concrete shall be protected with
a five course bituminous protective
treatment all around (instead of bitumen
coatings). Such a protective treatment
should be laid on top of the above plaster
and taken right up to the plinth level. At
the base, a 50 mm thick layer of cement
concrete (of identical proportions as in
the rest of the foundations) should be
laid over the protective layer to prevent
puncturing of the bitumen layers while
placing the reinforcements. The cover of
concrete shall be in addition to this. Also
vertical and top surfaces 120 mm thick
brick lining should be provided to ensure
that the protective treatment is secured to
the concrete surface. Such a construction
is likely to cost approximately 100
percent more than the usual (unprotected)
reinforced concrete foundation as per
PWD Specifications.
Another alternative solution will be to use
two coats of epoxy paints on the concrete
surface instead of two coats of bitumen
as described in the first alternative above,
the other details remaining the same.
This type of construction is likely to cost
approximately 175 percent more than the
usual (unprotected) reinforced concrete
foundation as per PWD Specifications.
9.3 Reinforced
Concrete
Foundations for Bridges
Pile
If RCC pile foundations are to
be adopted, special considerations and
precautions become necessary. Generally,
precast RCC piles with suitable protective
treatments lowered in prebored holes
will be preferable. For this, the type of
cement, cement content, water- cement
ratio and the cover thickness shall be
the same as above. The precast concrete
piles should be protected with two coats
of bitumen coatings after proper curing
and then lowered into the prebored holes,
the intermediate space being filled with
bentonite slurry. Epoxy paints will be
more durable but the cost will be nearly purify the increasingly polluted water in
three to four time more.
our cities.
9.4 Other Precautions
Fresh concrete surfaces should not be
allowed to come in contact with the
aggressive soil and ground water from
the foundation trenches at least for the
first 14 days.
It is the most efficient and effective
method of water purification known to
man. It uses a special, semi-permeable
membrane which removes impurities as
small as 0.0001 micron (i.e. 0.00000004
inches) in size, cleansing water of
all biological impurities, suspended
particles, Total Dissolved Solids (TDS),
salts, metals and chemicals. Most nonRO systems can filter particles only up
to 0.5-10 microns in size, leaving out
almost all dissolved impurities (like badtasting salts) and some finer physical
impurities.
Concrete surfaces exposed to such
aggressive
environments
should
preferably be free from construction
joints. For this concrete should be cast
in single uninterrupted operation. If
construction joints become inevitable,
then, they should be located beyond the
range of levels of the fluctuating water 11. CONCLUSIONS
table. Steps should be taken so that a
perfect bond between the hardened and The results of the investigation show that
fresh concrete is obtained in such a case. the water and soil in and around Rohtak,
Jhajjar and National Capital Region
9.5 Quality Assurance
contains high percentages of sulphates
It is well known that properly placed, and chlorides. The water in Sonepat
dense, impermeable and well cured district is alkaline and does not meet the
concrete is very necessary to ensure its requirements of IS: 456.
durability. Therefore, proper quality Suitable measures are required to protect
control must be exercised at all stages foundations of bridges and culverts from
of construction and proper workmanship surrounding aggressive environment so
ensured. Materials conforming to the that structures may live upto the normal
required specifications should only be expected life. Appropriate treatment of
used, after proper testing. All aspects of water is also needed to be done so as to
concrete making right from the choice of bring water as per standards.
materials through the stages of batching,
mixing, placing, compaction and curing
should receive adequate attention. A References
proper quality control plan should be
1. British Standards Institution –
drawn before the commencement of
Structural Use of Concrete.
construction and strictly adhered to.
2. Specifications for Road & Bridges
Works by Ministry of Road,
Wherever we do not have appropriate
Transport & Highways.
ground water or if there are no canals or
water courses then, we have no alternative 3. Study Report of Guru Govind Singh
but to use principle of Reverse Osmosis
Inderaprastha University; Delhi.
to purify water for construction.
10. WHY REVERSE OSMOSIS?
Reverse Osmosis (RO) is an advanced 4. Haryana Vidhan Sabha Report, 2014
on Water Conservation Measures in
water purification technology being
Haryana.
adopted for use in homes and offices to
Journal of the Indian Roads Congress, January-March 2015
Paper No. 629
UTILIZATION OF RICE HUSK ASH IN HOT MIX ASPHALT
CONCRETE AS MINERAL FILLER REPLACEMENT
Raja mistry *and tapas kumar roy**
ABSTRACT
An experimental study was conducted to investigate the use of agro-industrial by product namely Rice Husk Ash (RHA)
as filler instead of conventional material filler in dense bituminous macadam (DBM) mix. For this purpose, samples with
five different asphalt content were prepared by using 2% cement as conventional filler and different proportions of RHA
ranging from 2% to 4% as alternative filler and the amount of optimum bitumen content and other Marshall properties were
determined. Comparing the test result of Marshall mix design for different filler mixes, it has come in view that RHA can be
used effectively as mineral filler by reducing optimum bitumen content in asphalt concrete mix.
1. INTRODUCTION
The use of good quality conventional
materials in road construction is
becoming increasingly expensive in
India due to the increasing demand as
well as its scarcity in nature. Further the
development and use of new modified
paving materials in road construction
results in high performance pavement
to meet the communities. So, attempts
should be made to utilize industrial
and agricultural wastes effectively in
construction to address environmental and
economic concerns.In usual practice the
mineral filler used in asphalt-aggregate
mixture as tail end product conforms to
aggregate specification (Serkan Tapkin,
2007). The use of Portland cement, lime
stone powder, fly ash (Serkan Tapkin,
2007), marble dust (Karasahin and Terzi,
2007), Glass powder (Jony et al.2011) in
place of conventional mineral filler like
stone dust is universally accepted.
Rice is a primary source of food
especially in Asian region. In 2002, the
annual global production of paddy rice
was 579.5 million tones and of this
21.2% was produced by India alone.
Rice husk, basically an agricultural
residue is obtained from rice paddy
milling industries and each ton of dried
rice paddy produces about 20% husks.
After burning rice husk at rice mill and
electricity generating power plant as
fuel, rice husk ash (RHA) is produced
as by product.As the ash to husk ratio is
18%, therefore the total ash production
in India could be 4.43 million tones per
year (Bronzeoak Ltd, 2002). RHA is a
highly pozzolanic material; it contains
non-crystalline silica and high specific
surface area that are accountable for
his high pozzolanic reactivity (Della et
al.2002). RHA has been used in limepozzolana mixes and could be a suitable
partly replacement for Portland cement
(Smith et al., 1986; Zhang et al., 1996;
Nicole et al., 2000; Sakr 2006; Sata et
al., 2007; etc). So, an effort was made to
evaluate the usefulness of using RHA as
filler instead of conventional filler in hot
mix asphalt concrete that may mitigate
the problem of waste management as
well as the cost of land required for
disposal of wastes.
1.1 Objective of the Present Work
In order to solve the environmental
pollution caused by RHA usually dumped
near by the plant, it is tried to use as filler
material in the construction of roadway
pavement by improving the properties
of asphalt mix. In view of the same, the
present investigation is targeted:
•
To use of RHA as replacement
of conventional filler in a certain
percentage to minimize the cost of
utilization of cement, lime or stone
dust which are used conventionally.
•
To evaluate the modified Marshall
properties for utilization of RHA.
•
To check the validity of utilization
of RHA for construction of HMA
as alternative filler that may bring
the economic and environmental
benefit.
* Ph.D., Student, E-mail: [email protected]
Indian Institute of Engineering Science and Technology Shibpur, (WB)
** Assistant Professor, E-mail: [email protected]
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
14
2.
mistry
LITERATURE REVIEW
In construction industry considerable
research work was made on the use
of RHA after detection of the said
materials as waste, particularly in
cement industry, RHA was mixed with
cement in concrete work.Chatveera
and Lertwattanaruk (2011) ground and
used the black rice husk ash (BRHA)
from a rice mill as a partial cement
replacement and investigated the
durability of conventional concretes
with high water–binder ratios including
drying shrinkage, autogenous shrinkage,
depth of carbonation, and weight loss
of concretes exposed to hydrochloric
(HCl) and sulphuric(H2SO4) acid
attacks. Memon et al. (2010) evaluated
the utilization of RHA as viscosity
modifying agent in self compacting
concrete (SCC) by satisfying the slump
flow, L-Box, V-funnel, V-funnel at
T5min, compressive strength, density
of hardened SCC and water absorption
characteristics and indicated as the cost
effective ingredients for specific SCC
mix upto 42.47%. The various properties
of concrete like compressive strength,
splitting tensile strength, modulus of
elasticity, water permeability and rapid
chloride permeability were determined
by Ramezanianpour et al. (2009) by
& roy
Table 2: Evaluated Properties of Aggregates
Property Tested
Test Methods Results
Aggregate Impact Value
IS:2386(IV)
16.0 %
MoRT&H
Specifications
24% max
Los Angeles Abrasion Value IS:2386(IV)
19.0%
30% max
Water Absorption Value
Specific Gravity1. Coarse aggregate
2. Fine aggregate.
Combined Flakiness and
Elongation Index
1.2%
2% max
IS:2386(III)
IS:2386(III) 2.83
IS:1202-1978 2.68
2.5-3.0
IS:2386(I)
30%
adding RHA with cement and the result
of investigation showed the enhanced
durability criteria of concrete by
reducing the chloride diffusion. Further
Roy (2013) utilized RHA to improve
the strength of subgrade constructed
by alluvial soil without increasing
any significant demend of water for
achieving desired compaction.
27.45%
3.1 Aggregate
Basalt and granite type of crushed aggregate
was used in this investigation. The coarse
and fine aggregates were sieved and mixed
in proportion to achive the desired grading
as per IS: 2386 part-I. The combined
grading of coarse and fine aggregates and
added filler for the particular mixture fall
within the limit shown in Table 1. The
percentage content of aggregates reduced
3.MATERIALS USED IN THIS
accordingly as the percentage of RHA
INVESTIGATION
increased in the gradation. The evaluated
properties of aggregates used in the present
In the present investigation coarse study are tabulated in Table 2.
aggregate, fine aggregate, cement, Rice
Husk Ash (RHA) and bitumen were 3.2Mineral Filler
used to check the potentiality of using
RHA as mineral filler in the preparation The mineral filler used in this study was
of Dense Bituminous Macadam Ordinary Portland Cement (OPC) of 43
grade and rice husk ash as alternative
(DBM) mix.
of cement after satisfying the MoRT&H
Specification mentioned in table 500-9.
Properties of OPC were tested as per
Table 1: Aggregate Gradation of Dense Bituminous Macadam Grading 2 from IS 4031 (Part 1) – 1988 and results are
MORT&H Table 500-10
furnished in Table 3.
Normal aggregate size
Layer Thickness
IS sieve (mm)
37.5
26.5
19
13
4.75
2.36
0.300
0.075
Bitumen
25 mm
50-75 mm
Specified grading 2
Cumulative % by weight of total aggregates passing
Specified limit
Adopted Gradation
100
100
90-100
95
71-95
83
56-80
68
38-54
46
28-42
35
7-21
14
2-8
5
minimum 4.5%
minimum 4.1%
Table 3: Evaluated Properties
of Cement
Property Tested
Results
Specific Gravity
3.14
Setting time
1. Initial
2. Final
135
180
Fineness, m2/kg
225
Compressive
Strength, MPa
Journal of the Indian Roads Congress, January-March 2015
3 day
7 day 28 day
20
44
54
Utilization of rice Husk ash in hot mix Asphalt Concrete As mineral Filler Replacement
Rice husk ash collected from the rice
mill of Bardhaman, a district of West
Bengal. Now for using as filler materials
such ashes were sieved through 75 µ
sieve and then different physical and
chemical properties were evaluated.
The characteristics of RHA are given in
Tables 4 and 5.
Table 4: Evaluated Physical
properties of RHA Used
Sl. Properties
No.
1. Grain size distribution
(a) Sand size particles
(b) Silt size particles
(c) Clay size particles
2. Light compaction
(a) Proctor’s maximum
dry density
(b) OMC
3. Specific gravity
4. CBR
(a) Unsoaked
(b) Soaked
Test results
63.0%
29.0%
8.0%
11.1 kN m-3
27.50%
2.00
16.50%
11.50%
were evaluated by using RHA in three
different proportions (2%, 3% and 4%)
as alternative of conventional filler for
the said five different bitumen contents.
Three samples were prepared for each
proportion of bitumen and total 45
number of samples were prepared for
determining Marshall Stability, flow
value, percentage air void (Vv), voids in
mineral aggregate (VMA), voids filled
with bitumen (VFB) values.
15
the said value has been increased to
3500 kg for mixing of 4% RHA as
alternative filler. But all such
experimental values as shown in Fig. 1
(a) have satisfied the minimum Marshal
Stability value of DBM mentioned
in the specification of Ministry of
Road Transport and Highways, 2001
(MoRT&H) as 900 kg at 60ºC.
Flow value corresponding to OBC
as illustrates in Fig 1(b) decreases
Table 6: Physical Properties of 60/70 gradually for mixing RHA in increasing
proportion. For addition of RHA as
Grade Bitumen
4%, the said value becomes 2.5 mm,
which is nearly 35% lesser than that
Property Tested
Test Method Results
value obtained by using 2% cement as
Penetration (100 g, 5 IS 1203-1978
67
filler as shown in Table 8. According
sec at 25°C) (1/10th
to MORT&H, the flow value range
of mm)
between 2 mm to 4 mm and satisfies all
Softening Point ,
IS 1205-1978
49
the experimental values.
°C (Ring & Ball
Apparatus)
Ductility at 27°C (5
cm /min pull), cm
IS 1208-1978
>100
Specific Gravity
IS 1202-1978
1.024
3.3 Bitumen
5.
RESULTS AND DISCUSSIONS
Bitumen used in this investigation
was collected from the PWD, Howrah
Sub Division, West Bengal. Different
evaluated properties of the same as per
standard code of practices have been
furnished in Table 6.
Further, Fig. 1(c) shows the changes
in air voids (Vv) against the varying
bitumen content ranging from 4.1%
to 5.7% for all kinds of fillers. The
said value of against OBC for mixes
with 2% cement is 3.4% but for RHA
of increasing proportion from 2% to
4% shows the increasing trend having
range from 4.6% to 5.12% satisfies the
standard specifications of MORT&H
by increasing the stability value.
The Marshall Test result conducted on
mixes by using 2% cement and varying
proportions of RHA as filler are
shown in Fig. 1 and different Marshall
properties of Dense Bituminous
Macadam corresponding to OBC are VMA for all the specimens against
varying proportion of bitumen for
4.METHODOLOGY
also shown in Table 7.
different filler materials shown in Fig
For determining the optimum bitumen The OBC values are remain same as 1(d).The VMA value corresponding
content (OBC) of DBM mix, samples 5.433% on addition of 2% cement as OBC for conventional mix is 15.8%.
well as 2% RHA as filler, however
were prepared with varying asphalt further addition of RHA the said value But for addition of RHA as alternative
content (4.1%, 4.5%, 4.9%, 5.3% decreases gradually and for mixing of filler from 2% to 4%, it decreases
and 5.7%) by using 2% cement as 4% RHA, the OBC value decreases to from 16.7% to 15.64%. According to
conventional filler. Then Marshall Test 5.167%.
MoRT&H, minimum VMA value is
was conducted as per outline of ASTM:
12.5% against nominal Dmax value and
D-1559 on the prepared samples In conventional mix, stability value designed air voids in this study.
in view of determining Marshall corresponding to OBC is 2585 kg for
Properties. Further the said properties addition of cement, however after that All the calculated values of VFB as
shown in Fig 1(e) indicates a decreasing
tread with mixing RHA in increasing
Table 5: Evaluated Chemical Properties of RHA
ratio and become 15.64% for addition of
SiO2 CAO Al2O3 Fe2O3 MgO K2O SO3Loss on insoluble 4% RHA, while for mixing 2% cement,
aggravation residue
the VFB is 15.8%. All such values
86.17 1.98 1.52 1.26
0.87 0.35 0.11 6.28 satisfy the standard specification.
Journal of the Indian Roads Congress, January-March 2015
16
mistry
& roy
VFB as specified in the Ministry
of Road Transport and Highway
(2001).
6.4 Hence, rice husk ash can be usedas
an alternative filler material instead
of conventional mineral filler in
asphalt concrete mixture as a cost
effective solution by reducing the
environmental hazard created by
such wastes.
Fig. 1a: Bitumen Content vs Marshall Stability
Fig. 1e: Bitumen Content vs VFB valus
7.
ACKNOWLEDGEMENT
Fig 1: Marshall Properties of DBM
mixture with Cement and RHA as
The Authors are thankful to the technical
filler Materials
Fig. 1b: Bitumen Content vs Flow value
persons of Transportation Engineering
Laboratary of Indian Institute of
6. CONCLUSIONS
Engineering Science and Technology,
From this experimental study, the Shibpur, Howrah, West Bengal.
following conclusions can be made:
REFERENCES
6.1 Addition of RHA in DBM mix,by
replacing conventional mineral filler 1. ASTM: D-1559, “Test for Resistance to
like cement effectively reduces the
Plastic Flow of Bituminous Mixture Using
optimum bitumen content.
Marshall Apparatus”
6.2 Experimental results also indicate
that mixing of different proportions
of RHA as alternative filler
have better strength with lesser
deformation compare to that of the
conventional mix.
Fig. 1c: Bitumen Content vs Vv value
6.3 Further, addition of RHA as a filler
up to 4% by weight of the bitumen
may be used satisfactorily in asphalt
concrete by satisfying the Marshall
parameters i.e. stability value, flow
value, percent air void, VMA and
2.
Bronzeoak Ltd., Rice husk ash market study.
<www.berr.gov.uk/files/file15138.pdf>.
3.
Chatveera, B. and Lertwattanaruk, P. (2011)
“Durability
of
Conventional
Concrete
Containing Black Rice Husk Ash” Journal
of Environment Management, Vol. 92,
pp. 59-66.
4.
Della, V.P., Kuhn, I. and Hotza, D. (2002)
“Rice Husk Ash as an Alternate Source for
Active Silica Production.” Materials Letter,
Vol.57 (4), pp. 818-821.
Table 7: Marshall Properties of DBM with Cement and RHA Corresponding
to Optimum Bitumen Content
Fig. 1d: Bitumen Content vs VMA value
Optimum
Filler Added
Bitumen
Content (%)
5.433
2% Cement
Marshall
Stability
(kg)
2585
Flow
Value
(mm)
3.825
Vv
(%)
VMA
(%)
VFB
(%)
3.4
15.8
74.7
5.433
5.33
5.167
2675
3228
3500
3.53
2.9
2.5
4.6
4.427
5.12
16.7
16.721
15.64
72.5
72.65
69
2% RHA
3% RHA
4% RHA
Journal of the Indian Roads Congress, January-March 2015
Utilization of rice Husk ash in hot mix Asphalt Concrete As mineral Filler Replacement
5.
Jony, H. H., Al-Rubaie, M. F. and Jahad,
I. Y. (2011) “The Effect of Using Glass
Powder Filler on Hot Asphalt Concrete
Mixtures Properties” Eng. & Tech. Journal,
Vol.`29, pp. 44-57.
6.
Karasahin,
M.
and
Terzi,
S.
(2007)
“Evaluation of Marble Waste Dust in the
Mixture of Asphaltic Concrete” Journal of
9.
Memon, S. A., Shaikh, M. A. and Akbar,H.
(2010) “Utilization of Rice Husk Ash
as Viscosity Modifying Agent in Self
Compacting Concrete” Journal of Materials
in Civil Engineering. Vol.25, pp. 1044-1048.
10. Nicole, P.H., Monteiro, P.J.M. and Carasek,
H. (2000) “Effect of Silica Fume and Rice
Husk Ash on Alkali-Silica Reaction” Journal
of Materials. Vol. 97 (4), pp. 486-492.
7.
Kartini, K., Mahamum B.H. and Hamidah,
M.S. (2008) “Improvement on Mechanical
Properties of Rice Husk Ash Concrete with
Superplasticizer” International Conference on
13. Sakr,K. (2006) “Effects of Silica Fume and
Rice Husk Ash on the Properties of Heavy
Weight Concrete.” Journal of Materials in
Civil Engineering. Vol.18 (3), pp. 367-376.
14. Sata, V., Jaturapitakkul, C. and Kiattikomol,
K. (2007) “Influence of Pozzolan from
Various by-Product Materials on Mechanical
Properties
of
High-Strength
Concrete.”
Journals of Construction and Building
Construction and Building Materials, Vol. 21,
pp. 616-620.
17
11. Ramezanianpour,
A.A.M.,
Khani,
M.
and
Ahmadibeni,
Gh.
(2009)
“The Effect of Rice Husk Ash on
Mechanical Properties and Durability
of Sustainable Concretes” International
Journal of Civil Engineering. Vol.7 (2),
pp. 83-91.
Materials. Vol. 21 (7), pp. 1589–1598.
15. Tapkin, S. (2008) “Mechanical Evaluation of
Asphalt-Aggregate Mixture Prepared with Fly
Ash as a Filler Replacement” Journal of Civil
Engineering, Canada, Vol.35, pp. 27-40.
Construction and Building Technology, A (20),
pp. 221-230.
8.
Ministry of Road Transport and Highways
(MoRT&H) (2001), “Specification for Roads
and Bridge Works”, New Delhi.
12. Roy, T.K. (2013) “Evaluation of Properties
of Alluvial Soil with addition of Wastes from
Thermal Power Plantand Rice Mill” International
Journal of Geotechnical Engineering, Vol.7
(3), pp. 323-329.
16. Zhang, M.H. and Mohan, M.V. (1996) “HighPerformance Concrete Incorporating Rice
Husk Ash as a Supplementary Cementing
Material” ACI Materials Journal, Vol. 93 (6),
pp. 629-636.
Journal of the Indian Roads Congress, January-March 2015
18
Statement about ownership and other particulars about Newspaper
(JOURNAL OF THE INDIAN ROADS CONGRESS)
to be published in the first issue of every year
after the last day of February
Form IV
(See Rule 8)
1.
Place of Publication
…
Delhi
2.
Periodicity of its Publication
…
Quarterly
3.
Printer’s Name
…
S.S. Nahar
Nationality – whether citizen of India
(if foreigner, state the country or origin)
Address
…
Indian
…
Secretary General, Indian Roads Congress,
Jamnagar House, Shahjahan Road,
New Delhi-110011
Publisher’s Name
…
S.S. Nahar
Nationality-whether citizen of India
(if foreigner, state the country or origin)
Address
…
Indian
…
Secretary General, Indian Roads Congress,
Jamnagar House, Shahjahan Road,
New Delhi-110011
Editor’s Name
…
S.S. Nahar
Nationality-whether citizen of India
(if foreigner, state the country or origin)
Address
…
…
Secretary General, Indian Roads Congress,
Jamnagar House, Shahjahan Road,
New Delhi-110011
Names and address of individuals
who own the newspaper and partners of
shareholders holding more than one
percent of the total capital
…
Indian Roads Congress,
Jamnagar House, Shahjahan Road,
New Delhi-110011.
4.
5.
6.
Indian
I, S.S. Nahar, Secretary General, Indian Roads Congress, hereby declare that particulars given above
are true to the best of my knowledge and belief.
S.S. Nahar
Publisher
Dated: 1 March 2015
Journal of the Indian Roads Congress, January-March 2015
Journal of the Indian Roads Congress, January-March 2015
Paper No. 630
Analysis of T-Beam Skew Bridges under Live Loads
Ankita Chowgule* And M. Manjunath**
abstract
T-beams are most commonly used by the designers for small and medium span bridges. For the safety of fast moving traffic,
the roadway must be maintained as straight as possible. In order to cater to this requirement of the bridge, provision of skew
bridge becomes necessary. With increase in skew angle, the stresses in the bridge deck and reactions on the abutment vary
significantly from those in straight slab [1, 2]. The analysis is carried out on reinforced concrete T-beam skew bridges using FE
software SAP2000. The live load considered is Class 70R wheeled vehicle as per IRC. The results in terms of longitudinal
bending moment, transverse bending moment, shear force and torsional moment for varying skew angles (0, 10, 20, 30, 40,
50 and 60 degrees) and different spans (16m, 20m and 24m) are compared with the respective straight bridges.
1.
Introduction
in straight slab. Special characteristics of The superstructure consists of three
skew deck slab are:
longitudinal girders at 2.5m centre
to centre spacing and cross girders at
1. Variation in the direction of every 4m interval. Three single-span
maximum bending moment across lengths of 16m, 20m and 24m, simply
width, from near parallel to span at supported bridge for various skew angles
edge and orthogonal to abutments in (as specified in Table 1) are modelled.
central region.
Details of the bridge and constituent
material are given below:
2. Hogging bending moments near
obtuse corners.
Deck slab Thickness
200 mm
Skewed bridges are most commonly
encountered at highways, river crossings
and at other extreme grade changes
when the provision of skew bridge
becomes necessary due to limitation of
space. T-beams are most commonly used
by the designers for small and medium
span bridges. The skew angle can be
defined as the angle between the normal
to the centreline of the bridge and the 3. Considerable torsion in decks.
centreline of the abutment or pier cap.
With the increase in skew angle, the 4. High reactions and shear forces near
stresses in the bridge deck and reactions
obtuse corners.
on the abutment vary significantly from
those in straight slab. [1, 2]
5. Low reaction and possibly uplift
near acute corners, especially in case
In normal bridges, the deck slab is
of slab with high skew angles.
perpendicular to the supports and as
such the load placed on the deck slab 6. The points of maximum deflection
is transferred to the supports which are
nearer obtuse angled corners.[3]
placed normal to slab. Load transfer from
a skew slab bridge is a complex problem 2. Details
of
the
with uncertainties regarding spanning of
Structure
the slab and load transfer to the supports.
With increase in skew angle, the stresses The bridge selected for the study is
in the bridge deck and reactions on the two-lane T-beam concrete bridge with
abutment vary significantly from those kerb and parapet (without footpath).
Grade of Concrete
M 25
Grade of Steel
Fe 415
The dimensions of main girders and
diaphragms are as shown in Table 1.
The analysis is carried out using FE
software SAP2000[4]. The live load
considered is IRC Class 70R wheeled
vehicle [5, 6].
3.
Geometric Dimensions
T-beam construction consists of vertical
rectangular stem with a wide top flange
as shown in Fig. 1; the wide top flange is
usually the transversely reinforced deck
slab and the riding surface for the traffic.
* Post Graduate Student, E-mail: [email protected]
** Assistant Professor, E-mail: [email protected]
Department of Civil Engineering, KLEMSSCET, Belgaum, Karnataka, (India)
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
Chowgule & Manjunath on
20
Table 1: Dimensions of T-beams and Diaphragms
Model Span
No.
(m)
Skew
Angle
T – beams (m)
hw
hf
ht
Diaphragms (m)
hw
h’f
ht
1
16
0°
0.3
0.2
1.6
0.3
0.2
1.6
2
16
10°
0.3
0.2
1.6
0.3
0.2
1.6
3
16
20°
0.3
0.2
1.6
0.3
0.2
1.6
4
16
30°
0.3
0.2
1.6
0.3
0.2
1.6
5
16
40°
0.3
0.2
1.6
0.3
0.2
1.6
6
16
50°
0.3
0.2
1.6
0.3
0.2
1.6
7
16
60°
0.3
0.2
1.6
0.3
0.2
1.6
8
20
0°
0.3
0.2
2.0
0.3
0.2
2.0
9
20
10°
0.3
0.2
2.0
0.3
0.2
2.0
10
20
20°
0.3
0.2
2.0
0.3
0.2
2.0
11
20
30°
0.3
0.2
2.0
0.3
0.2
2.0
12
20
40°
0.3
0.2
2.0
0.3
0.2
2.0
13
20
50°
0.3
0.2
2.0
0.3
0.2
2.0
14
20
60°
0.3
0.2
2.0
0.3
0.2
2.0
15
24
0°
0.3
0.2
2.4
0.3
0.2
2.4
16
24
10°
0.3
0.2
2.4
0.3
0.2
2.4
17
24
20°
0.3
0.2
2.4
0.3
0.2
2.4
18
24
30°
0.3
0.2
2.4
0.3
0.2
2.4
19
24
40°
0.3
0.2
2.4
0.3
0.2
2.4
20
24
50°
0.3
0.2
2.4
0.3
0.2
2.4
21
24
60°
0.3
0.2
2.4
0.3
0.2
2.4
The stem widths (hw) vary from (0.25 to
0.40m). The depth of wide top flanges (hf)
and stem depth (hs = ht -hf) must satisfy
the requirement of moments, shear and
deflection under critical combination of
loads.
4.Modelling using SAP2000
In the analysis, dead load and live load
without impact factor are considered.
Dead load includes the self weight of the
deck slab, longitudinal girder, transverse
girder and wearing coat load. The
live load considered is IRC Class 70R
wheeled vehicle
Fig. 2: Cross Section of T-beam
Fig. 3: Plan of T-beam
4.
Results and Discussions
4.1 Comparison of Results
The longitudinal bending moment,
its percentage variation and ratio
of transverse bending moment to
longitudinal bending moment for
16 m, 20 m and 24 m spans and 0, 10,
20, 30, 40, 50 and 60 degrees skew
angles is studied and is presented in the
form of various graphs for each span
respectively.
SAP2000 supports predefined T-Beam
deck element which includes various
geometric data such as material
4.2 Discussions
properties and sectional details of bridge
For 16 m span:
components. The sectional details of
Fig. 1: Section in T-beam
T-Beams used in this study are arrived as
per the codal provisions and the same is support data which is partially hinged 1) The maximum longitudinal bending
moment decreases from 0 degree
i.e. the translational displacement across
incorporated in the model.
skew angle to 30 degree skew angle
the span of the bridge is not permitted
and further it increases to 50 degree
The boundary condition used in and thereby allowing for vertical
skew angle.
analysing the model includes the displacement.
Journal of the Indian Roads Congress, January-March 2015
Analysis of T-Beam Skew Bridges under Live Loads
21
Table 2: Dead Load Results for 16m Span
Skew Angle
(degree)
Maximum
Longitudinal
Bending
Moment
‘Tmax‘ (kNm)
0
(Ref.)
10
20
30
40
50
60
Skew Angle
(degree)
0
(Ref.)
10
20
30
40
50
60
Ratio of
Tmax
Lmax
1120.741
Maximum
Transverse
Bending
Moment
‘Lmax’
(kNm)
291.388
1123.306
1121.154
1127.390
1145.921
1156.123
1165.274
297.156
310.209
315.333
329.799
371.397
361.3753
0.265
0.277
0.280
0.288
0.321
0.310
0.260
Percentage Shear Force
Variation Of at 0m end
Longitudinal
(kN)
BM as
Compared to
0 deg. skew.
0
-329.639
0.229
0.037
0.593
2.247
3.157
3.974
-332.847
-343.856
-362.201
-382.418
-430.724
-329.639
Table 3: Live Load Results for 16m Span
Maximum
Maximum
Ratio of
Percentage Shear Force
Longitudinal Transverse
Tmax
Variation Of at 0m end
Bending
Bending
Lmax
Longitudinal
(kN)
Moment
Moment
BM as
‘Tmax‘ (kNm) ‘Lmax’ (kNm)
Compared to
0 deg. skew.
2371.324
520.170
0.219
0
-905.711
2342.165
2333.744
2333.254
2335.877
2363.726
2341.571
532.895
557.681
569.385
606.961
765.691
763.322
Skew Angle
(degree)
Maximum
Longitudinal
Bending
Moment
‘Tmax‘ (kNm)
0
(Ref.)
1937.387
10
20
30
40
50
60
1940.902
1950.207
1966.217
1989.880
2027.456
2083.131
0.227
0.239
0.244
0.260
0.332
0.326
1.230
1.585
1.605
1.495
0.320
1.255
-897.948
-936.193
-989.770
-955.941
-943.114
-977.079
Table 4: Dead Load Results for 20m Span
Ratio of
Maximum
Percentage Shear Force
Transverse
Variation Of at 0m end
Tmax
Bending
(kN)
Longitudinal
Moment
BM as
Lmax
‘Lmax’
Compared to
(kNm)
0 deg. skew.
386.233
0.199
0
-439.877
387.973
390.716
395.407
399.219
415.051
460.977
0.200
0.200
0.201
0.201
0.205
0.221
0.181
0.662
1.488
2.709
4.649
7.523
-448.69
-459.565
-473.797
-498.042
-531.078
-601.717
Journal of the Indian Roads Congress, January-March 2015
Shear Force
at 16m end
(kN)
Torsional
Moment
(kNm)
317.572
-35.915
314.594
362.382
306.739
307.672
299.316
296.981
-41.799
-46.581
-48.917
-62.182
-65.881
-118.659
Shear Force
at 16m end
(kN)
Torsional
Moment
(kNm)
908.300
-222.025
867.024
845.461
815.184
765.200
726.801
694.479
-230.638
-246.360
-258.294
-286.443
-325.950
-415.393
Shear Force
at 20m end
(kN)
Torsional
Moment
(kNm)
439.877
-42.771
432.653
426.937
422.85
423.39
422.016
425.668
-47.747
-49.530
-49.788
-67.304
-90.273
-114.404
Chowgule & Manjunath on
22
Skew Angle
(degree)
Maximum
Longitudinal
Bending
Moment
‘Tmax‘ (kNm)
0
(Ref.)
10
20
30
40
50
60
3174.762
3167.933
3165.668
3181.247
3170.391
3160.747
3152.346
508.118
514.202
529.561
545.936
584.572
699.648
Skew Angle
(degree)
Maximum
Longitudinal
Bending
Moment
‘Tmax‘ (kNm)
0
(Ref.)
10
20
30
40
50
60
3084.247
Skew Angle
(degree)
0
(Ref.)
10
20
30
40
50
60
3085.155
3113.287
3145.556
3190.026
3283.072
3416.473
Table 5: Live Load Results for 20m Span
Maximum
Ratio of
Percentage Shear Force
Transverse
Tmax
Variation Of at 0m end
Bending
Lmax
Longitudinal
(kN)
Moment
BM as
‘Lmax’
Compared to
(kNm)
0 deg. skew.
505.145
0.159
0
-978.365
0.160
0.162
0.167
0.172
0.185
0.222
0.215
0.286
-0.204
0.138
0.441
0.706
-987.112
-1021.577
-1062.192
-1044.261
-1038.109
-1068.370
Table 6: Dead Load Results for 24m Span
Maximum
Ratio of
Percentage Shear Force
Transverse
Tmax
Variation Of at 0m end
Bending
Lmax
Longitudinal
(kN)
Moment
BM as
‘Lmax’
Compared to
(kNm)
0 deg. skew.
488.815
0.158
0
-614.928
500.830
521.605
548.088
576.203
634.050
731.5033
0.162
0.168
0.174
0.181
0.193
0.214
0.029
0.942
1.988
3.430
6.446
10.772
-621.615
-632.542
-645.698
-669.034
-706.871
-776.440
Table 7: Live Load Results for 24m Span
Maximum
Maximum
Ratio of
Percentage Shear Force
Longitudinal Transverse
Tmax
Variation Of at 0m end
Bending
Bending
Lmax
Longitudinal
(kN)
Moment
Moment
BM as
‘Tmax‘ (kNm) ‘Lmax’ (kNm)
Compared to
0 deg. skew.
4040.601
550.576
0.136
0
-1095.074
4024.205
4021.197
4026.506
4008.677
4016.809
4002.743
564.205
586.759
621.467
639.588
707.120
837.299
0.140
0.146
0.154
0.159
0.176
0.209
0.406
0.480
0.349
0.790
0.589
0.937
-1094.715
-1113.330
-1139.739
-1105.907
-1089.943
-1108.757
Journal of the Indian Roads Congress, January-March 2015
Shear Force
at 20m end
(kN)
Torsional
Moment
(kNm)
985.013
-247.779
947.579
923.379
902.309
865.991
815.462
781.250
-254.330
-261.458
-270.705
-289.327
-336.697
-409.011
Shear Force
at 24m end
(kN)
Torsional
Moment
(kNm)
614.928
-48.912
609.983
611.183
602.768
600.300
595.976
592.368
-55.591
-58.120
-57.181
-74.849
-101.390
-141.595
Shear Force
at 24m end
(kN)
Torsional
Moment
(kNm)
1098.425
-247.532
1072.184
1056.856
1041.447
1001.936
947.824
891.845
-254.315
-261.026
-267.312
-282.652
-326.748
-396.282
Analysis of T-Beam Skew Bridges under Live Loads
2) The maximum transverse bending For 24m span:
moment increases from 0 degree
skew angle to 50 degree skew angle 1) The maximum longitudinal bending
and further it decreases to 60 degree
moment decreases from 0 degree
skew angle. This is because the
skew angle to 30 degree skew angle
stress concentration at the supports
and further it increases to 50 degree
varies significantly.
skew angle.
3) The maximum shear force at the
obtuse angle of the deck decreases 2) The maximum transverse bending
moment increases from 0 degree
from 0 degree skew angle to 60
skew angle to 50 degree skew angle
degree skew angle.
and further it decreases to 60 degree
4) The maximum torsional moment
skew angle. This is because the
increases from 0 degree skew angle
stress concentration at the supports
to 60 degree skew angle.
varies significantly.
5) The ratio of maximum transverse
bending moment to maximum 3) The maximum shear force at the
longitudinal
bending
moment
obtuse angle of the deck decreases
increases from 0 degree skew angle
from 0 degree skew angle to 60
to 50 degree skew angle and further
degree skew angle.
it decreases to 60 degree skew
angle.
4) The maximum torsional moment
increases from 0 degree skew angle
For 20m span:
to 60 degree skew angle.
1) The
maximum
longitudinal
bending moment decreases from
0 degree skew angle to 30 degree
skew angle, then it increases to 40
degree skew angle and further it
again decreases to 60 degree skew
angle. This is because the stress
concentration at the supports varies
significantly.
2) The maximum transverse bending
moment increases from 0 degree
skew angle to 60 degree skew
angle.
23
4) The maximum shear force at the
obtuse angle of the deck decreases
as the skew angle increases.
5) The maximum torsional moment
increases as the skew angle increases.
This is due to the variation in load
dispersion area on the bridge deck.
6) The ratio of maximum transverse
bending moment to maximum
longitudinal
bending
moment
increases as the skew angle
increases.
7.
References
1.Kar Ansuman, Vikash Khatri,
Maiti P.R., Singh P.K., “Study on
Effect of Skew Angle in Skew
Bridges” International Journal
of Engineering Research and
Development, Vol. 2, Issue 12,
August 2012, pp. 13-18.
2.Menassa C., Mabsout M., Tarhini
K., Frederick G., “Influence
of Skew Angle on Reinforced
5) The ratio of maximum transverse
Concrete Slab Bridges” Journal of
bending moment to maximum
Bridge Engineering, Vol. 12, No. 2,
2007, pp. 205-214.
longitudinal
bending
moment
increases from 0 degree skew angle
3.Khatri Vikash, Maiti P. R., Singh
to 60 degree skew angle.
P. K., Kar Ansuman, “Analysis Of
Skew Bridges Using Computational
5. Conclusions
Methods” International Journal
of Computational Engineering
Based upon the results presented in
Research, Vol. 2, No. 3, 2012, pp.
this study, the following conclusions
628-636.
emerge:
4.SAP2000
(2009),
User’s
manual SAP2000, Computers
1) SAP2000 Software is useful to
and Structures, Inc., Berkeley,
develop the finite element (FE)
California, U.S.A.
models of T-beam skew bridges.
3) The maximum shear force at the
obtuse angle of the deck decreases
5.IRC:
6-2010,
“Standard
from 0 degree skew angle to 60
2) The maximum longitudinal bending
Specifications and Code of Practice
degree skew angle.
moment decreases upto 30 degree
for Road Bridges”, Section II
4) The maximum torsional moment
skew angle and further it increases
Loads and Stresses, Indian Road
increases from 0 degree skew angle
to 60 degree skew angle. This is
Congress, New Delhi.
to 60 degree skew angle.
because the plane of maximum stress
21-2000,
“Standard
in skew deck bridges is not parallel 6.IRC:
5) The ratio of maximum transverse
Specifications
and
Code
of
to the centerline of the roadway.
bending moment to maximum
Practice for Road Bridges”,
longitudinal
bending
moment 3) The maximum transverse bending
Section III Cement Concrete (Plain
increases from 0 degree skew angle
and Reinforced), Indian Road
moment increases as the skew angle
Congress, New Delhi.
to 60 degree skew angle.
increases.
Journal of the Indian Roads Congress, January-March 2015
Paper No. 631
Replacement of damaged suspended span of Varsova
Bridge across Vasai Creek on NH-8
M. L. Gupta* Dhananjay A. Bhide** And Prashant Dongre***
SYNOPSIS
The existing Varsova Bridge across Vasai Creek, called as Bassein Creek Bridge when constructed, was opened to traffic
in 1968. It is in Mumbai Ahmedabad section of NH-8, about 35 Km from Mumbai. Major cracks were noticed on 12th
December 2013 in the west side girder of the penultimate span from Mumbai end. As an immediate measure the traffic was
restricted to single lane of light vehicles only. The site was inspected for assessment of the damage. It was decided to replace
the said span with composite steel girder. The road being one of the busiest National Highway in India connecting MumbaiDelhi carrying heavy traffic,, an immediate replacement in very short duration was warranted. The real task was to dismantle
the existing PSC span without any debris falling in the creek and working in very restricted location.
This paper discusses the constraints imposed by the design of the existing structure, formulating replacement scheme, design
of new structure, dismantling the damaged PSC span and its replacement with composite steel girders and concrete deck
slab, in detail.
INTRODUCTION
The Varsova Bridge across Vasai creek
is 555.32m long with 8 spans. The span
arrangement is 48.46 + 2 x 57.3 + 2 x
114.6 + 2 x 57.3 + 48.46m spans. The
main spans were built with balanced
cantilever construction from central three
piers for spans of 2*57.3 + 2 * 114.6 +
2*57.3 m. The cantilever construction
was continued in adjacent penultimate
spans with overhangs of 8.84m each.
These overhangs supported suspended
spans of 48.46m. Thus the bridge has
6 spans that constitute a continuous
module for a length of 458.4m. This is
quite a long continuous module even at
prevailing standards.
Bridge is with 7.32m carriageway, 1.525m
wide raised footpaths on both side and
0.180m wide railing kerbs. Overall deck
width is 10.77m. The suspend spans and
end spans are with two numbers of precast, post tensioned girders of 3.05m
total depth. The deck slab in these spans
is 275mm thick in between girders and
depth of cantilevers vary from 275mm
to 150mm at the tip. Part of the deck
slab thickness constitutes 2.285m wide
flanges of pre-cast girders. 100mm thick
slab is cast in-situ over precast flanges.
Rocker and rocker cum roller bearings
are used at all the locations. Expansion
joints at articulations, penultimate
foundations and abutments are of
single strip seal type. Thin plate piers
are provided at all locations. Except at
penultimate foundation locations, only a
single row of bearings is provided over
all piers. All foundations are with well
foundations.
dia strands, provided externally for each
of the girders. This strengthening was
carried out on two occasions, first in
year 1989 by external prestress with 10
nos. of strands and subsequently with
additional 10 nos. of strands in year 2001.
(Photos 1 & 2).
THE DAMAGE
Existing girder on West side of suspended
span penultimate span on Mumbai end
developed cracks at center (Photo 3).
Cracks were predominant on one side of
central diaphragm. The major crack was
widest at the bottom flange, measuring
about 150mm and progressed towards
top and away from center. Crack had
branched out while progressing. A large
chunk of concrete was missing at the
The suspended spans as well as end crack location. The reinforcement and
spans i.e. all 4 simply supported spans internal prestressing cables were exposed
were strengthened with 20 nos. 12.7mm at this point (Photo 4). Some of the wires
* Director ( Technical), IRB Infrastructure Developers Ltd., Mumbai
** Vice President (Design)
*** Manager (Technical)
Modern Road Makers Pvt. Ltd., Mumbai
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
Gupta, Bhide & Dongre on
26
provided at the support of
overhang was sufficient to carry
the design loads only. Live
load part of this was same for
existing as well as new structure.
As such the weight of the new
structure could not exceed than
that of damaged structure so as
not to overstress the cantilever
overhang.
Photo 1
Photo 2
External Pre-stressing to Simply Supported Spans
of internal cables were snapped. At top 6. The articulated end was about
of deck, a perceptible dip was observed
8.84m away from nearest pier
at the location of the crack. Some cracks
and therefore no firm support
were observed in intermediate diaphragm
for any construction activity was
at one of the suspenders of the deviator
available at that end.
block as well. The deviator block at this
location was twisted also. The girder was 2.Constraints dictated by the
held in position due to external prestress
arrangement
of
existing
provided and this prevented a total
structure:
collapse, averting a catastrophe.
1. The span rested on articulated
support at one end. At this
CONSTRAINTS FOR PROPOSED
location a rocker bearing was
STRUCTURE
provided. The bearings did not
1. Physical Constraints:
have any pedestal. Therefore gap 1. The suspended span was with 2
between the two articulations was
pre-cast PSC beams with in-situ
very small only 175mm.
deck slab.
2. The
dimensions
of
the
2. The dimensions of the existing
articulations were just sufficient structure were very slender
to accommodate the bearings
compared with present day
only.
requirement of various codes /
practices.
3. The suspended span was
3. The damaged span was between
a strengthened span on Mumbai
end and cantilever span on Surat
end, imposing restrictions on
type and weight of equipment for
replacement due to their limited
load carrying capacities.
4. The span was in tidal zone.
5. Head room between HTL and
soffit of the structure was very
small to allow any equipment to
work underneath.
supported on overhang from
adjacent span. The prestress
Photo 3: Cracks in Damaged Girder
4. The adjacent span on Surat end
was with an overhang. Therefore
the sagging moments in the span
were controlled by the load from
the overhang and span supported
by it. Any reduction in the weight
of suspended span would have
resulted in increase in the sagging
moments in this span.
5. Any change in weight of new span
would have induced differential
moments in penultimate pier and
foundation.
6. From the conditions as in 3 to5
it was imperative to maintain
weight of new span same as that
was replaced.
7. The existing structure complied
with code provisions that were
prevailing in Sixties. Structure
with 2 girders was then allowed,
instead of minimum three girders
now required. The thickness of
web of the girders was only
Photo 4: Failed Internal Cables
Journal of the Indian Roads Congress, January-March 2015
Replacement of damaged suspended span of Varsova Bridge across Vasai Creek on NH-8
190 mm. Prestressing cables
important, absence of proper jetty
now available cannot be
facilities.
accommodated in this thickness.
The pre-stressing was with 3. The pre-stressed girders had to be
removed from span either as a whole 28 nos. of cables. Only 14 cables
or with cut segments that were to be
were anchored at end and balance
temporarily held from equipment
cables were anchored in the
spanning across the damaged span.
deck at top. The resulting web
thickness at end of 600mm would
be insufficient to accommodate 4. A standard launching girder for
all the cables at end. It was
50m span and capacity for girder
therefore a foregone conclusion
weighing 180T was not readily
that concrete structure complying
available at such a short notice.
with present codes and still
maintaining weight restrictions PROPOSED STRUCTURE:
was impossible.
1. Structural arrangement:
3. Construction constraints:
i. The replacement structure had to
be launched/moved over existing
end span.
ii. The launching/movement within damaged span had to be over
intact girder as well as within its
capacity.
iii. The number of girders had to be
two to accommodate on existing pre-stressed
articulations,
thereby making individual
girders heavier.
i.With due consideration to the
foregoing constraints, it was
decided to provide two simply
supported steel girders with cast
in-situ deck slab.
27
bottom plates, without breaking
pre-stressed articulation was
impossible.
iv.The visual inspection of the
bearings reveled that these were
maintained in good condition
without any corrosion and
damage. Therefore it was
decided to use the same for new
structure.
v.The strip seal type expansion
joints at both ends of the
damaged span were found in
good condition. It was decided
to use same type of joint.
CONSTRUCTION SCHEME AND
SEQUENCE:
The major construction steps were as
below (Fig. 1A & 1B show schematic
sequence).
ii.The weight and overall depth of
the deck was exactly matched 1. Remove the service/utility cables
with that of original structure,
laid over footpaths.
with slab weight adjusted for the
same.
2. Cut the footpaths and part of the
cantilever slab, keeping the flanges
iii.Arrangement of raised footpaths
of the precast girders intact along
was retained as it helped to
with deck slab over them.
maintain the weight of the
deck.
3. Remove
existing
bituminous
wearing coat.
iv.At support over cantilever
overhang, the articulations 4. Assemble steel girders in the
similar to existing girders were
approach, in line with intact girder of
1. The movement of any equipment
provided. The depth at other
the damaged span. Launch the steel
over damaged span had to be within
end was maintained, same as
girder over intact girder across the
the zone and capacity of intact
existing structure.
damaged span. Support this girder
girder.
beyond diaphragm at articulations
of cantilever overhang on Surat end,
2. The bed comprised of marine 2. Bearings and expansion joints:
in line with girder/webs and over
clay/soft soil overlaying rocky i.The cast steel rocker bearings
beams on Mumbai end. Side shift 1st
stratum. Thus any support from bed
were used for existing structure,
necessarily had to be from bed rock
girder to place it over the damaged
rocker on articulation end and
for its stability.
girder. Launch 2nd girder across the
rocker cum sliding on pier end.
intact span. Fix cross diaphragms for
lateral stability.
Support from bed for holding cut ii.The bottom plates of the
bearings were embedded in presegments was therefore difficult
stressed articulation and fixed to 5. Clear headroom of 1.5m was
and cost prohibitive due shallow
it with rag bots, as per practice
draft, limited head room, depth of
required at Surat end (higher end)
in those days. Removal of the
founding stratum and last but equally
for placing the concrete, cutting
CONSTRAINTS FOR DISMANTLING EXISTING STRUCTURE:
Journal of the Indian Roads Congress, January-March 2015
28
Gupta, Bhide & Dongre on
Fig. 1A: Schematic Arrangements for Construction Methodology
machinery etc. As such lift both the
temporarily supported from already 8. Remove the cut segments with
girders and support them on trestles.
a moving trolley over main steel
launched steel girders.
The resulting head room at Mumbai
girders with strand jacks mounted
7. The pre-cast girder was proposed to
end was about 3m.
over the trolley.
be cut in segments, about 3m long and
weighing 10t each, on an average.
9. Lower the cut segments in the barge
6. Remove the deck in between girder
Cross members along flanges were
stationed below.
flanges and cross diaphragms, with
proposed at 1.5m centers to hold the
segments of about 3m in length.
cut segments through bolts drilled
Since the slab and diaphragms for
and anchored in the flanges. For 10. Retrieve the embedded parts of
the bearings from cut segments.
some length were essential for lateral
connecting cut segments suspenders
Clean all the components for reuse.
stability of the damaged girder these
from cross members across main
Reassemble and fix the bearings in
in central portion were to be cut only
girders at top were proposed and
position.
after girder segments were cut and
were to be erected at this stage.
Journal of the Indian Roads Congress, January-March 2015
Replacement of damaged suspended span of Varsova Bridge across Vasai Creek on NH-8
Fig. 1B: Schematic Arrangements for Construction Methodology
Journal of the Indian Roads Congress, January-March 2015
29
Gupta, Bhide & Dongre on
30
11. Lower the main girders
assembled bearings.
over
arrangements were designed as per
relevant IRC/BIS Codes.
12. Erect shuttering for the deck slab 4. Arrangement for lifting and
and cast the same. Follow this with
lowering Main Steel Girders:
casting of raised footpaths and
Girders were to be lowered within
parapets.
the gap created by removing
13. Fix expansion joints, lay bituminous
the span and since no space to
support the girders was available
wearing coat and open to traffic.
at pier cap/articulations, additional
DESIGN
AND
DETAILING
length of the girder were required
OF THE
STRUCTURE AND
to enable to support them from
ARRANGEMENTS:
spans on either side. To facilitate
this, brackets from top flange was
1.Main Steel Girders:
provided at either ends. These
In order to limit the weight of the
brackets were bolted to top flanges
steel girder to a minimum, medium
for easy removal before casting of
tensile steel of grade E350 BR was
deck slab. (Fig. 2)
used for main girders and of grade
E250 BR for other components The temporary supports for the
girders were in form of trestles at
such as cross diaphragms, bracings,
each end (Fig. 3). Trestles were
web stiffeners etc. HTS Bolts of
made in pieces that could be easily
grade 8.8 were used for spliced
handled for raising and lowering.
connections. All other joints were
Two trestle arrangement was
welded joints.
necessary.
To expedite the fabrication activity
5.Procedure for dismantling of
the depth of webs were limited to
damaged span:
2.45m, to suit available plate width,
2.5m. The splices in web were also The cantilever footpaths along with
some part of deck slab, till edge of
decided on similar lines to suit
flange of original precast girder
available lengths, 12m. Splice at
were cut in segments of 2.5 to 3m
center of the span was avoided.
length.The slab in between girders
2. Composite Deck Slab:
was cut in two stages. Firstly
transverse cuts were completed for
RCC slab for full deck width in
segments prior to launching of the
M40 grade concrete was adopted.
steel girders. After launching of
The required slab thickness for
the steel girders and lifting them
matching the weight of the structure
in position the individual segment
to that of old one was 340mm thick
of slab was cut. Diaphragms were
between girders and for cantilevers
cut after damaged main PSC girder
with thickness reducing to 200mm at
was cut into segments.
tip and provided.
Existing precast girders weighed
3. Analysis and design:
about 180 T each with in-situ slab
Analysis of the structure i.e. new
over their flanges. Individual PSC
composite deck was done with
girder was to be cut with damaged
STAAD analysis program. Design
girder first followed by intact one.
was as per latest IRC codes. Similarly
Initial cuts were planned at L/4
all
the
components/members
location. These were in two stages,
for various supporting/lowering
first cutting bottom bulb and web
followed by top flange. The 20 nos.
of strands for external pre-stress
were to be cut sequentially.
After completing first two cuts
and cutting strands of external prestress, the cutting of girder in 3m
long segments was allowed as per
site convenience.
Diaphragms were cut in parallel
to the activity of cutting smaller
segment of first girder.
6.Suspending arrangement
holding cut segments:
for
Basic arrangement for suspending
cut segments of the girders was
with beams across the two main
girders. These were spaced at 1.5m
C/C with respect to corresponding
segment that was to be held. Thus
four suspenders were planned for
each of the segments. Each of the
suspenders was designed to carry
half the load of the segment at
working stress level.(Fig. 4)
Since the girders were with external
prestress as well some residual
prestress in damaged girder and
full in intact girders, each of the
segmentsof L/4 length resulting
from first cuts was expected to act
one unit. Therefore the suspenders
on either side of first cuts were
designed to carry load of the said
length of the girder.(Fig. 5)
The suspenders held the girder
segments through a beam connected
to segment at top of slab.
The slab and diaphragm panels
were comparatively light, weighing
about 2.5t. These were suspended
directly from jacks over the trolley
during cutting operation itself.
7.Arrangement for securing cut
segments to suspenders:
The cut segments of the girders were
connected to suspenders through
Journal of the Indian Roads Congress, January-March 2015
Replacement of damaged suspended span of Varsova Bridge across Vasai Creek on NH-8
Fig. 2: Bracket Arrangement for Lifting and Lowering of Main Girders
Fig. 3: Trestles for Supporting Main Girders During Dismantling of Damaged Girder
Journal of the Indian Roads Congress, January-March 2015
31
Gupta, Bhide & Dongre on
32
a beam that in turn is connected
to anchor bolts in segments. This
indirect system was necessary as
the suspenders at top had to clear
the 1150mm wide bottom flange
of the steel girders. Suspenders on
either side of the first cuts at L/4
locations were connected to beams
across bottom of the girder instead
of anchors as the loads to be catered
were substantially higher.
Anchor bolts were placed in drilled
holes for a depth of 200mm in
275mm thick slab. These holes were
kept at haunch of cantilever slab.
The reinforcement in cantilevers
of existing slab was very nominal,
10mm dia (MS) at 300mm C/C.
As such slab did not have any
significant capacity in flexure
and load had to be predominantly
transferred through shear.
Slabs and diaphragms panels were
planned to be held directly during
cutting
operations.
However
anchor bolts at 4 locations for
slab panel and 2 locations for
diaphragm panels were provided
to connect these through a beam to
holding equipment. No suspenders
were provided for these.
All anchors were expected to carry
about 6T load.
8.Arrangement for Removal of
Cut Segments:
The cantilever footpath and part
slab panels were planned to be held
in position with slings through
drilled holes, connected to holding
and removing equipment.
As already explained the cut
segments of the girder were
suspended through 2 beams spaced
at 1.5m. Two lifting beams, one at
each end of the beams connected
to suspenders were planned.
Strand from jacks on trolley over
Fig. 4 Fig. 5
girders was to be attached to lifting
beams.
REMOVAL
OF
DAMAGED SPAN:
Five anchors were tested for 13.6T
load each while one was tested till
18T load. (Photos 5 & 6)
EXISTING
3.Fabrication, Erection, Launching
and Positioning of Main Steel
1.Cantilever Footpaths and Part of
Girders:
Deck Slab:
Fabrication of the girder segments
Cutting
and
removal
was
as well as other steel components
commenced from Mumbai end,
started in different workshops to
starting at damaged girder side.
facilitate simultaneous activity.
Initially the segments were held in
The joints in main girders were
position by hydra of 10T capacity.
planned to enable transporting of
However it was observed that
each with commercially available
the front axle of hydra had to be
trailers.
positioned on damaged girder
itself. Therefore the system A rail track with standard BG rails
was discontinued as a safety
was laid in line with intact girder
precaution.
of the span over approach of the
bridge. It was laid over intact
A 50T capacity crane was mobilized.
girder to transport the steel girders
in damaged span. Trolleys were
A careful study and check was made
erected over track to launch the
to decide the position of the crane
girders.
and its outriggers to maintain the
loads transferred within carrying
Girders were launched over the
capacity of the deck.
damaged span one after the other.
First girder after launching across
2.Installing and Testing of Anchor
the span was side shifted. The
Bolts:
support to girder was provided
As a precaution and ensuring the
over the diaphragms of the span to
workability and capacity of the
avoid any load transfer to damaged
anchor bolts and existing slab
girder. The second girder was then
concrete, it was decided to install
launched in similar way. Cross
six test anchor bolts. As per the
diaphragms and bracings were
design of the anchor bolt, anchor
installed. The girder assembly
depth of 150 mm, for 20 mm dia.
was then lifted to its desired
HTS anchor bolt, was necessary.
Journal of the Indian Roads Congress, January-March 2015
Replacement of damaged suspended span of Varsova Bridge across Vasai Creek on NH-8
Photo 5 : Pull out Test in Progress
position over trestles to support
cut segments of the damaged span.
The assembly was raised and
maintained in horizontal alignment.
(Photos 7 & 8)
4. Intermediate Slab and Diaphragm
Panels:
Photo 6 : Pull out Test Assembly
33
Prior to start of the initial cuts
suspenders of the proposed
segments
were
tightened.
Tightening continued during the
cutting operations, prior to cutting
of each of the external strand
set as well as after cutting them.
Suspenders on either side of the
cut were invariably found to be
tight indicating the load transfer to
them and three segments spanning
across them, as predicted.
cut at the haunch point leaving After completion of first main cuts
further cutting was continued at
part of the pre-cast portion of
two locations in the girder till all
diaphragms with girders only. This
was a parallel activity to cutting of
segmentation was complete. A
damaged girder.
separate set of equipment cut the
diaphragms along with.
5. Segments of the Girders:
The cutting was done with diamond For removal of the segments
the lifting beams were attached
studded wire saw in conjunction
to segments. These beams were
with concrete saw whenever
connected to overhead jacks on
required.The cut was from bottom
trolleys. The jacks then lowered
upwards.As explained earlier the
the segments in the barge
first two cuts were at L/4 positions.
positioned below the segment for
The external strands were cut as
transporting them to dump yard.
planned during this cut.
(Photos 9 & 10)
The intermediate slab panels were
removed, starting from one end.
The panels on either side of the
central diaphragm were retained
till first two main cuts in damaged
girders were complete. This was to
ensure the lateral stability of the
girder, especially at the main crack Unexpectedly wire saw was getting
location.
frequently stuck while cutting the The overall estimated time was
an underestimate on account of
webs. This appeared to be due to
frequently stuck wire saw as well as
The diaphragms were cut only
pre-stress from external as well as
only cutting or lowering operation
after first two cuts in damaged
internal cables. This slowed down
possible at any given time.
girder were complete. These were
the progress quite significantly.
Photo 7
Photo 8
Main Girders Launched Over the Damaged Span
The end segments of both the
girders and connecting diaphragms
could not be tackled along with
respective girders. Trolley holding
the segments directly below the
jacks could not be stationed at
required position as brackets
holding the girder and supporting it
on trestles were at higher position
than rails. Length of the trolley
beyond jacks also added to the
eccentric position. The segments
then had to be further cut in five
pieces at each end for both girders
put together.
Journal of the Indian Roads Congress, January-March 2015
Gupta, Bhide & Dongre on
34
Steel girders were then lowered in
position. The rocker bearings at Surat
end were aligned and matched with
existing bottom plates, embedded in
articulations. The rollers of the bearings
on Mumbai end were set at required
position for ambient temperature for
resting the rocker assemblies over them.
COMPLETING BALANCE
ACTIVITIES:
Photo 9
Photo 10
Lowering of Girder Segments by Using Strand Jack
While lifting for lowering the last,
(fifth piece) at Mumbai end anchor
bolts in the concrete failed one after
the other. Fortunately the piece was
on pier cap and was laterally held
hence it did not fall in the creek. The
saddle plate of rocker cum roller
bearing at this end slipped into the
creek. Rollers were fortunately
held by bottom plate and remained
in dangling position. The segment
and rollers were secured over cap
with wooden wedges immediately.
Slings were passed around by
drilling holes in the segments
and then removed. Except this,
all dismantling operations were
without any untoward incident.
(Photos 11 & 12)
RE-FIXING OF BEARINGS AND
LOWERING OF STEEL GIRDERS:
Loss of one saddle plate was a setback.
The bearings were manufactured with
cast steel of yield stress of 425 Mpa and
UTS of 550 Mpa. Cast steel plate of this
grade was very difficult to get of-the-
The unfamiliar and challenging activities
for replacement of the span were almost
over. Casting of deck slab and providing
other miscellaneous items however
became a new challenge on account of
fast approaching monsoon i.e. almost
end of working season.
shelf. After a concentrated search a plate
of this grade was located at ongoing
works in Jamnagar, about 800 Km away
from Mumbai. The plate of required size,
shape and finish was fabricated from the
1.Removal of Trolley Rails etc.
said plate but it took a full week to get
from Top of Girder and Erection
it at site.
of Shuttering:
The top plates of the rocker bearings
This was done with the help of
were removed from cut segments
50T crane having 25m reach that was
simultaneously. All the plates were
deployed for erecting these items.
cleaned and inspected. Only surface
This had to be done after lowering
rusting and some dried grease stains
the main steel girders in position
existed. These were thoroughly cleaned
as otherwise the crane boom would
and reassembled. The bolts connecting
need to be with steep angle to reach
bearings to girders were removed. Existing
near mid span and capacity of crane
bolt threads were inspected and new bolts
would be quite large, beyond the
manufactured to suit the same as well as
allowable load on the existing deck.
for the required length to fix with steel
girders. The shear pins in rocker bearings Trolley and jacks were moved
were also removed and replaced.
to Mumbai end and removed.
Rails and cross beams along
The main steel girders were provided with
with suspenders were removed.
slotted holes in bottom flanges to pass
Cantilevering brackets as well as
the bolts connecting bearing plates and to
beams between steel girders were
fasten them. This enabled fixing of bearings
fixed to them to receive longitudinal
quite smoothly to the new girders.
members of the shuttering system.
Half the span was done from each
end with the help of the crane.
2.
Casting of Deck Slab:
Expansion joints were cast along
with deck slab to minimize the
overall time. The edge beams of
adjacent spans were used along
with new edge beams in the
damaged spans.
Mix design was specially made to
allow pumped concrete as well as
Photo 11
Photo 12
Lowering of Existing Damaged Girders Completed
Journal of the Indian Roads Congress, January-March 2015
Replacement of damaged suspended span of Varsova Bridge across Vasai Creek on NH-8
achieve strength of 40 Mpa at age
of 4 days.
tanker; 3000 Lit storage water
tank; 3 nos. 125 KVA DG sets; 10
T hydra; 50 T capacity hydraulic
crane.
Road kerbs were cast a day after
concreting of deck slab and parapet
walls after three days.
2. Removal of Segments:
The solid footpaths were cast a Movable trolley on girders; 4 no.
week after casting deck slab.
strand jacks 185 T capacity; 12
nos. 15.2mm dia strands 15m long;
3.Laying Bituminous Wearing
1 no. 125 KVA DG set; 200 T
Coat:
capacity barge with tug.
After deck slab concrete had
3. Assembling and Erection of main
achieved strength of 35 Mpa, i.e.
Steel Girders:
th
on 4 day after concreting deck
slab, bituminous wearing coat was 300m long 52 kg rails; 2 nos.
laid. In order to limit loads within
trolleys; 1 no. 50 T capacity crane;
the capacity of newly laid deck
1 no. 10 T capacity winch with
slab the mix was brought to site in
225m wire rope; 5 nos. 100 T
light loads i.e. limited to 10t only.
capacity hydraulic jacks with 200
As a precaution only one truck was
mm stroke.
allowed on the new span besides
4. Casting of Deck Slab and Laying
the paver and roller.
Wearing Coat:
EQUIPMENT USED FOR THE
6 nos. transit mixers; concrete
OPERATIONS:
pump; vibrators; sensor paver
1. Cutting
of
the
Concrete
etc.
Structure:
5.Miscellaneous:
3 nos. wire saw sets; 2 nos. wheel
Motor boat for movement in creek;
saw sets, 10000 Lit capacity water
safety nets; wood items (planks,
CHRONOLOGICAL EVENTS:
(From noticing the damage till completion of the replacement)
Sr. No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
35
ply wood, sleepers etc.); grinders;
concrete and steel drills, welding
sets, flame cutters; leveling
instruments; chain pulley blocks;
flood lights; platforms below
decking etc.
CREDITS:
1. Owner:
National Highway
Authority of India
2. Independent Intercontinental
Engineer:
Consultants &
Technocrats Pvt.
Ltd.
3. Executed by: IRB Surat – Dahisar
Tollway Pvt. Ltd.
4. Design
STUP Consultants
Consultant: Pvt. Ltd.
ACKNOWLEDGEMENTS:
The
Authors
express
sincere
thanks to NH Division, Public
Works Department, Government of
Maharashtra, Gammon India Ltd.
and Mr. D K Kanhere, Retired CE,
PWD, Government of Maharashtra
for making available the original
drawings and valuable information
on old structure and repairs.
Event
Cracks noticed and communicated by Police petrol
Joint inspection by NHAI, IE and concessionaire
Traffic restricted to single lane
Inspection by experts and recommendation for replacing span and closure of traffic
Closure of traffic
Concept proposal
Review of detailed designs and drawings
In-principle approval for replacement of span
Acceptance to proposal with composite construction
Dismantling of footpaths and part slab
Erection of steel girders at site
Launching of girders across damaged span
Fixing of cross diaphragm, end brackets and lifting of girders to desired level
Commencement of segment cutting
Complete removal of span
Fixing of reassembled bearings
Placing steel girders in position
Casting of deck slab
Laying wearing coat using light loads
Opening of the rehabilitated Bridge to traffic
Journal of the Indian Roads Congress, January-March 2015
Date
12th December 2013
19th December 2013
19th December 2013
24th December 2013
25th December 2013
27th December 2013
10th January 2014
17th January 2014
31st January 2014
11th March 2014
31st March 2014
05th April 2014
16th April 2014
26th April 2014
24th May 2014
31st May 2014
02nd June 2014
05th June 2014
10th June 2014
12th June 2014
Paper No. 632
EFFECT OF UTILIZATION OF WASTE MARBLE ON INDIRECT
TENSILE STRENGTH PROPERTIES OF BITUMINOUS
CONCRETE MIXES
M.R. Archana*, H.S. Sathish**, G. Brijesh*** And Vinay Kumar***
ABSTRACT
Generally, the aggregate material passing the No: 200 sieve has been called filler. The amount of filler material is specified as
percentage of the weight of the mix, and becomes part of the bituminous mix design. The mechanical properties of properly
compacted pavements are dependent on the interlocking of the aggregate and the consistency of the binder. For the present
study bituminous concrete grade II as per MoRT&H (IV revision), binders 60/70 and PMB SBS 70 were used. Aggregates,
binders and marble dust were tested for their basic properties to ascertain their suitability for further tests.
Marshall properties, viz., stability, bulk density, air voids, flow, VMA and VFB were determined both for neat and modified
bituminous concrete mixes to find optimum binder content and optimum filler content. It was found that mixes with 10%
marble filler replacement exhibited optimum results both neat and modified bituminous concrete mixes. However mixes
with marble fine aggregate and filler replacement showed better Marshall results as compared to conventional mixes. As
the next part of investigations degradation tests were conducted on bituminous concrete mixes with 10% marble filler and
marble (fine aggregate and filler) replacement. All of which indicated breakage within standard limits. Modified bituminous
concrete mixes exhibited lower breakage as compared to neat bituminous concrete mixes. To determine the temperature
susceptibility, bituminous concrete mixes were subjected to indirect tensile strength test at 25°C and 40°C temperature.
1.
INTRODUCTION
A flexible pavement is built up of several
layers consisting of the wearing course,
surface course, base course, sub base
course, and compacted sub grade. The
pavement is built to a depth where stress
on any given layer will not cause undue
rutting, shoving and other differential
movements resulting in an uneven
wearing surface. The chief function of a
surfacing course is to provide a smooth
wearing surface with high resistance
to deformation. The thickness of the
pavement largely depends on the load to
be carried and the strength characteristics
of the sub grade. The effect of
magnitude of loads, tyre pressure, wheel
configuration influences largely the
stress, strain and deflection induced in the
flexible pavement. These factors must be
thoroughly analyzed and understood for
the pavement thickness design, damage
analysis and sensitivity analysis in both
ideal masses and layered systems.
2.
in solid waste management resulted
in alternative construction materials
as a substitute to traditional materials
like bricks, blocks, tiles, aggregates,
cement, lime, soil, timber and paint.
To safeguard the environment, efforts
are being made for recycling different
SOLID WASTE GENERATION wastes and utilize them in value added
IN INDIA
applications.
Presently in India, about 960 million
tonnes of solid waste are being generated
annually as by-products during industrial,
mining, municipal, agricultural and
other processes [2]. Out of 960 million
tonnes, 350 million tonnes are organic
wastes from agricultural sources and 290
million tonnes are inorganic waste of
industrial and mining sectors. Advances
2.1Necessity
Materials
of
Using
Waste
Growth of population, increasing
urbanization, rising standards of living
due to technological innovations have
contributed to increase both in the
quantity and variety of solid wastes
generated. Globally the estimated
*Assistant Professor, Department of Civil Engineering, RVCE, Bangalore
** Associate Professor, Department of Civil Engineering, BMSCE, Bangalore
*** Post Graduate Students, Department of Civil Engineering, RVCE, Banglaore
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
Effect of Utilization of Waste Marble on Indirect Tensile Strength Properties of Bituminous Concrete Mixes
quantity of wastes generation was 12
billion tonnes in the year 2002 of which
11 billion tonnes were industrial wastes
and 1.6 billion tonnes were municipal
solid wastes (MSW). About 19 billion
tonnes of solid wastes are expected
to be generated annually by the year
2025. Annually, Asia alone generates
4.4 billion tonnes of solid wastes and
MSW comprise 790 million tonnes
(MT) of which about 48 (6%) MT is
generated in India. By the year 2047,
MSW generation in India, is expected
to reach 300 MT and land requirement
for disposal of this waste would be 169.6
km2 as against which only 20.2 km2 were
occupied in 1997 for management of
48 MT. Fig. 1 shows the details of solid
waste (non-hazardous and hazardous
waste) generation from different sources
in India.
37
Table 1: Industrial Waste Product Usage in Road Construction [3]
Waste product
Fly ash
Blast furnace slag
Construction and
demolition waste
Colliery spoil
Spent oil shale
Foundry sands
Mill tailings
Cement kiln dust
Used engine oil
Marble dust
Waste tyres
Glass waste
Nonferrous slags
China clay
Source
Thermal power station
Steel industry
Construction industry
Possible usage
Bulk fill, filler in bituminous mix, artificial aggregates
Base Sub-base material, Binder in soil stabilization (wound slag)
Base Sub-base material, bulk-fill, recycling
Coal mining
Petrochemical industry
Foundry industry
Mineral processing
industry
Cement industry
Automobile industry
Marble industry
Automobile industry
Glass industry
Mineral processing
industry
Bricks and tile industry
Bulk-fill
Bulk-fill
Bulk-fill, filler for concrete, crack-relief layer
Granular base/sub-base, aggregates in bituminous mix, bulk fill
Stabilization of base, binder in bituminous mix
Air entraining of concrete
Filler in bituminous mix
Rubber modified bitumen, aggregate
Glass-fibre reinforcement, bulk fill
Bulk-fill, aggregates in bituminous mix
Bulk-fill, aggregates in bituminous mix
are a major source of pollution. Due
to environmental degradation, energy
consumption and financial constraints,
various organizations in India and abroad,
apart from the regulatory frame work of
United States Environmental Protection
Agency (USEPA), have recommended
various qualitative guidelines for
generation,
treatment,
transport,
handling, disposal and recycling of
non-hazardous and hazardous wastes.
Safe management of hazardous wastes
is of paramount importance. It is now a
global concern, to find a socio-technoeconomic,
environmental
friendly
solution to sustain a cleaner and greener
environment.
The
heterogeneous
characteristics of the huge quantity of
wastes generated lead to complexity in
recycling and utilization.
Generation of all these inorganic
industrial wastes in India is estimated to
be 290 MT. In India, 4.5 MT of hazardous
wastes are being generated annually
during different industrial process like
electroplating, various metal extraction
processes,
galvanizing,
refinery,
petrochemical industries, pharmaceutical
and pesticide industries. However, it is
envisaged that the total solid wastes from
municipal, agricultural, non-hazardous
and hazardous wastes generated from
different industrial processes in India
seem to be even higher than the reported 2.2Solid Waste Generation from
data. Already accumulated solid wastes
Mining Operations
and their increasing annual production
India has considerable economically useful
minerals and they constitute one-quarter
of the world’s known mineral resources.
In India, Rajasthan, Chhattisgarh, Bihar,
Madhya Pradesh, Orissa and Andhra
Pradesh are rich in minerals. Mining
operations are the primary activity in
any industrial process and the major
sources of pollutants include overburden
waste disposals, tailings, dump leaches,
mine water seepage and other process
wastes disposed by near-by industries.
Management of mining wastes is likely
to be of some significance in many
Fig. 1: Current Status of Solid Waste
Generation in India [2]
developing countries where recycling/
extraction and processing of minerals
have important economic values. In coal
was hery operations about 50% of the
material is separated as colliery shale or
hard rock. Most of this spoil is used as
filler in road embankments. Some spoils
can be considered for use in producing
lightweight aggregate. Presently most
of these wastes are being recycled and
used for manufacture of various building
materials and details are shown in Table 1.
2.3Construction Debris, Marble
Processing Waste and their
Recycling Potentials
In India, about 14.5 MT of solid wastes
are generated annually from construction
industries, which include wasted sand,
gravel, bitumen, bricks, and masonry
and concrete. However, some quantity
of such waste is being recycled and
utilized in building materials. The share
of recycled materials varies from 25%
in old buildings to as high as 75% in
new buildings. In India, about 6 MT of
waste from marble industries is being
released from marble cutting, polishing,
processing, and grinding.
Rajasthan alone accounts for almost
95% of the total marble produced in the
country and can be considered as the
world largest marble deposits. There are
about 4000 marble mines in Rajasthan and
about 70% of the processing wastes are
being disposed locally. The marble dust
is usually dumped on the riverbeds and
posses a major environmental concern.
Journal of the Indian Roads Congress, January-March 2015
Archana, Sathish, Brijesh & Kumar on
38
In dry season, the marble powder/dust
is lifted and carried in the air, flies and
gets deposited on vegetation and crop.
All these acts significantly affect the
environment and local ecosystems. The
marble dust disposed in the riverbed and
around the production facilities causes’
reduction in porosity and permeability of
the topsoil and results in water logging.
Further, fine particles result in poor
fertility of the soil due to increase in
alkalinity. Marble industry has few human
impacts with minor environmental risks,
however, each factory needs an intensive
evaluation to determine the certain norms
to regulate their action and to control the
possible impact produced. However,
new factories must be established
within industrial zones to prevent
environmental-community damages and
to allow better and safe competition. On
the other hand, existing factories have to
introduce mitigation actions to minimize
gradually the environmental impacts
through providing proper managements
relevant to environmental performance
test. Attempts are being made to utilize
marble dusts in different applications
like road construction, concrete and
bituminous concrete aggregates, cement,
and other building materials. It is evident
from such studies that there is a great
potential for recycling of wastes released
from different industrial processes.
Work carried out by earlier researchers
has shown that marble process residues
could be used in road construction since
they help to reduce permeability and
improve settlement and consolidation
properties.
3
Laboratory Investigations
3.1Main Constituents of BC Mix
binder. Material passing 2.36 mm IS MoRT&H (IV revision) specifications.
sieve and retained on 0.075 mm or Table 3 shows the specifications of
75 micron IS sieve is taken as fine bituminous concrete mix grade-I.
aggregate.
c) Filler: Fills the voids between the
fine aggregates, stiffens the binder
and reduces permeability. Ordinary
portland cement, stone dust and
Marble dust are used as filler.
d) Binder: Fills the voids and also
causes particle adhesion. Neat
bitumen 60/70 grade or Polymer
modified bitumen SBS 70 grade is
used.
3.2 Aggregates
Aggregates to be used should be
sufficiently strong, hard, tough, and
durable and of desirable right shape
to with stand the traffic effects. In the
present investigation crushed granite
aggregates were used.
Table 2: Tests Results of Aggregates
Sl. Aggregate tests
No.
Test
result
Requirements
as per Table
500-14 of
MoRT&H
Specifications
(IV revision)
1
Crushing value
(%)
12.80
Max 30%
2
Impact value (%) 18.50
Max 27%
3
Los Angeles
abrasion value
(%)
21.50
Max 35%
4
29.85
Flakiness and
Elongation Index
(Combined) (%)
Max 30%
5
Water absorption 0.162
(%)
Max 2%
6
Aggregate
specific gravity
(i) Coarse
aggregate
(ii) Fine
aggregate
Table 3: Aggregate Gradation as per
MoRT&H (IV revision) for Bituminous
Concrete Grade-I Mix
Sieve Size, in mm
26.5 – 19
19 - 13.2
13.2 - 9.5
9.5 - 4.75
4.75 - 2.36
2.36 - 1.18
1.18 - 0.6
0.6 - 0.3
0.3 - 0.15
0.15 - 0.75
0.75 – pan
Adopted Mid
Gradation
100
89.5
69
62
45
36
27
21
15
9
5
Table 3.1: Physical Properties of Marble
Dust
Physical Properties
Bulk Density (gm/cc)
1.38
Specific gravity
2.72
Particle size
Less than 350
micron
3.4 Filler
In the present study stone dust, marble,
and cement has been used as filler
material. Their specific gravity test
results are presented in Table 3.3
Table 3.2: Chemical Composition of
Marble Dust in Percentage
Chemical Composition
2.65
a) Coarse
Aggregates:
Offer
compressive and shear strength and
2.69
show good interlocking properties.
Material retained on 2.36 mm IS
sieve is taken as coarse aggregate.
3.3 Aggregate Gradation
2.5-3.0
b) Fine Aggregates: Fills the voids in Bituminous concrete mix of grading-I,
the coarse aggregate and stiffens the was chosen as per Table-500-18, of
CaCO3
MgCO3
SiO2
TiO2
Al2O3
MnO
MgO
CaO
K2O
P2O5
BaO
LOI
Journal of the Indian Roads Congress, January-March 2015
0.548
0.322
0.099
25
0.7
0.022
<0.05
0.054
0.58
0.02
0.34
37.06
Effect of Utilization of Waste Marble on Indirect Tensile Strength Properties of Bituminous Concrete Mixes
39
Table 3.3: Specific Gravity of Filler
Material
Mineral filler
Specific gravity
Stone dust
Marble
Cement
2.78
2.72
3.01
3.5 Bitumen
Properties of bitumen to be used in the
mix design were tested for Ductility,
viscosity, flash and fire points etc. and
the results tabulated in Tables 3.4 and
3.5
Table 3.4 Basic Test Results of Neat
Bitumen 60/70 Grade
SL Test conducted Test
Requirements
No.
results
as per
IS: 73-2002
1
2
Penetration at
25°C
(1/10th of mm)
66
Softening
47.25
point(R&B) (°C)
Fig. 2: Variation of Bitumen Content with Varying Marble Filler for Both Neat and
Modified Bituminous Concrete Mixes
60-70
45-55
3
Ductility
@27°C, cm
80+
75 minimum
4
Specific gravity
1.01
0.99 - 1.02
5
Flash point, °C
230
175 minimum
Table 3.5: Basic Test Results of PMB SBS
70 Grade Bitumen
Requirements
SL Test conducted Test
No.
results
as per
IRC: SP: 532002
1
Penetration at
25°C, 0.1 mm
100 gm, 5 sec.
62
50-90
2
Softening point
(R&B) (°C), min.
59
55 Min.
3
Specific gravity
1.05
-
4
Flash point, °C,
min.
226
220 min.
5
Elastic Recovery
of half thread in
ductilometer at
15 °C, %, min.
85
50 min.
6
Separation
difference in
softening point
R&D, °C, max.
2.45
3 max.
7
Ductility @
27°C, cm
100+
40+
Fig. 3: Variation of Bulk Density with Varying Marble Filler for Both Neat and
Modified Bituminous Concrete Mixes
Fig. 4: Variation of Stability with Varying Marble Filler for Both Neat and Modified
Bituminous Concrete Mixes
Journal of the Indian Roads Congress, January-March 2015
40
Archana, Sathish, Brijesh & Kumar on
3.6 Degradation
test
results:
Degradation test were conducted on the
neat and modified bituminous concrete
mixes. The results are shown in Fig. 8
and 9.
3.7 Indirect Tensile Strength Test (ITS)
Results: Indirect Tensile Strength test
were conducted on the neat and modified
bituminous concrete mixes. The results
are shown in Fig. 10 and 11, The variation
in ITS ratio for both the mixes is shown
in Fig.13.
Fig. 5: Variation of Air Voids with Varying Marble Filler for Both Neat and Modified
Bituminous Concrete Mixes
4.CONCLUSIONS
i.A portion of large size
aggregates in the bituminous
concrete mixes break down
during compaction and get into
the voids between the smaller
aggregates. This results in
creating a more dense mixture
and further degradation in
Marshall testing will be
minimum.
ii.The tests shows that degradation
of course aggregates in neat
bituminous concrete mixes
prepared with conventional
mixes is 10.02% and 2.87%
lower degradation compared
to 10% marble dust and partial
marble
replacement
(fine
aggregate and filler).
iii.The tests shows that degradation
of course aggregates in modified
bituminous concrete mixes
prepared with conventional
mixes is 6.21% and 3.19%
lower degradation compared
to 10% marble dust and partial
marble
replacement
(fine
aggregate and filler) showing
better performance for modified
bituminous mixes.
iv.The bituminous concrete mixes
prepared with PMB SBS 70
performed better than neat
bituminous concrete mixes.
Fig. 6: Variation of Voids Filled with Bitumen with Varying Marble Filler for Both
Neat and Modified BC Mixes
Fig. 7: Variation of Flow with Bitumen with Varying Marble Filler for Both Neat and
Modified Bituminous Concrete Mixes
Journal of the Indian Roads Congress, January-March 2015
Effect of Utilization of Waste Marble on Indirect Tensile Strength Properties of Bituminous Concrete Mixes
41
4.1Effect of marble dust on indirect
tensile strength properties
i.The indirect tensile strength
values for neat bituminous
concrete mixes tested at 25°C
for conventional mixes showed
15.24% and 9.35% lesser values.
Similarly for conventional
mixes prepared with modified
bituminous concrete mixes
showed 20% and 12.92% lesser
values compared to bituminous
concrete mixes prepared with
10% marble dust and partial
marble
replacement
(fine
aggregate and filler).
ii.The indirect tensile strength
values for neat bituminous
concrete mixes tested at 40°
C for conventional mixes
showed 29.04% and 18.46%
lesser values. Similarly for
conventional mixes prepared
with modified bituminous
concrete
mixes
showed
25.66% and 17.19% lesser
values compared to bituminous
concrete mixes prepared with
10% marble dust and partial
marble
replacement
(fine
aggregate and filler).
Fig. 8: Degradation Test Results for Specimens Prepared with Neat Bituminous
Concrete Mixes
Fig. 9: Degradation Test Results for Specimens Prepared with Modified Bituminous
Concrete Mixes
4.2Effect of Marble Dust on Indirect
Tensile Strength Ratio
The indirect tensile strength ratios for neat
bituminous concrete mixes tested at 25°C
for conventional mixes showed 17.10%
and 10.49% lesser values. Similarly
for conventional mixes prepared with
modified bituminous concrete mixes
showed 23.53% and 13.37% lesser
values compared to bituminous concrete
mixes prepared with 10% marble dust
and partial marble replacement (fine
aggregate and filler).
REFERENCES
Fig. 10: Shown the Variation of Indirect Tensile Strength for Specimens Tested at 250°C
1.Ministry of Road Transport and Highways
(MORTH), Indian Road Congress, 4th
Edition, 2001 New Delhi.
Journal of the Indian Roads Congress, January-March 2015
42
Archana, Sathish, Brijesh & Kumar on Effect of Utilization of Waste Marble on Indirect Tensile Strength Properties of Bituminous Concrete Mixes
2.Asokan Pappua(a), Mohini Saxenaa(a) and
Shyam R(b). “Solid Wastes Generation
in India and their Recycling Potential in
Building Materials”, A Regional Research
Laboratory (CSIR), Bhopal–462026, India,
b CESE, Indian Institute of Technology,
2006, Bombay, India.
3.Amit Goel and Animesh Das “Emerging
Road
Materials
and
Innovative
Applications”, National Conference on
Materials and their Application in Civil
Engineering, Aug. 2004, Hamirpur, India.
Fig. 11: Shows the Variation of Indirect Tensile Strength for Specimens
Tested at 400°C
4.Fakher J. Aukour and Mohammed I.
Al-Qinna
“Marble
Production
and
Environmental Constrains”, Jordan Journal
of Earth and Environmental Sciences, Vol 1,
Mar. 2008, pp 11 -21.
5.Ibrahim Ugur1 and Erhan Sener2 “Site
Selection for Marble Wastes Using Multi
criteria Decision Analysis and Geographic
Information Systems”, Ecology and
Environmental Protection, Jan. 2007,
pp 765 – 771.
6.Guidance for Collecting, Disposing and
Reusing Solid and Liquid Residuals from
Marble and Stone Extraction and Cutting
“Marble and Stone Waste Management”,
Palestinian Marble Hebron Municipality,
Hebron, Oct. 8th, 2007.
Fig. 12: Shows the Variation of Indirect Tensile Strength Ratio for Specimens
Prepared with Neat Bituminous Concrete Mixes
7.Paryavaran Bhawan, “Clean Technology
& Waste Minimization”, Ministry of
Environment & Forests Government of
India, 2006.
8.Santhosh Kumar M.M. “Comparison of
Marshall Stability and Indirect Tensile
Strength Using 60/70 Grade Bitumen,
Crumb Rubber Modified and Polymer
Modified Bitumen on Bituminous Concrete
Mix”, Seminar work, Bangalore University,
University Visveswaraya College of
Engineering Jnana Bharathi, Bangalore560056, 2007-08.
9.Peter E. Sebaaly, “The Benefits of Hydrated
Lime in Hot Mix Asphalt”, Prepared for the
National Lime Association, Apr. 2006.
Fi.g 13: Shows the Variation of Indirect Tensile Strength Ratio for Specimens
Prepared with Modified Bituminous Concrete Mixes
10.Randy C. West and Robert S. James,
“Evaluation of a Lime Kiln Dust as A
Mineral Filler for Stone Matrix Asphalt”,
Submitted at the 85th Annual Meeting
of the Transportation Research Board,
Washington, D.C., Jan. 2006.
Journal of the Indian Roads Congress, January-March 2015
Paper No. 633
GAP ACCEPTANCE BEHAVIOR OF RIGHT-TURNING VEHICLES
AT T-INTERSECTIONS - A CASE STUDY
Gopal R. Patil∗ And Jayant P. Sangole**
SYNOPSIS
This paper deals with the study of the gap acceptance of major to minor road and minor to major road right turning vehicles
at limited priority T-intersections in India (vehicles are driven on the left side in India). The unsignalized intersections in
India are uncontrolled or partially controlled; analyzing such intersections is complex. Limited priorities are observed at
partially controlled intersections, where major and minor roads are perceived by drivers based on intersection geometry and
traffic volume and speed on the approaches.Field data were collected at four T-intersections with limited priority using video
camera. The data extracted include gap/lag, subject vehicle type, conflicting vehicle type, and driver’s decision (accepted/
rejected).Different distributions are fitted to the available and accepted gaps. Based on Kolmogorov-Smirnov (K-S) test, it
is found that gamma distribution fits the available gaps well, whereas accepted gaps are better represented with lognormal
distribution. Binary logit models are developed for gap acceptance for the both turning movements. For model development,
80% of the extracted data (total data observations are 722 for major road right turning vehicles and 1066 for minor road
right turning vehicles) are used and remaining are used for model validation. The percentage of correct prediction by binary
logit models are 74.48% (for major road right turning) and 81.51% (for minor road right turning).Critical gaps estimation
revealed that maximum likelihood method gives the most consistent results. The critical gaps are smaller than the values
reported for developing countries.
1.
Introduction
In developed countries, unsignalized
intersections are usually controlled
by stop and yield (give way) signs,
which decide the priorities of various
movements. Efficient enforcement of
priority rules has made it possible to cross
the intersections with minimum conflicts.
However, the situation is very different
in India where most of the unsignalized
intersections do not have stop or yield
sign, and even if they exist, drivers do
not follow the priorities indicated by
the signs. At a few intersections, where
drivers are well aware of the major and
minor roads, limited priority is observed;
drivers usually interpret the major
and minor approaches based on the
intersection geometry, traffic volumes,
and vehicle speeds on those approaches.
At a significant number of unsignalized
T-intersections, especially in uncongested
areas,limited priority is observed. At
such intersections, the drivers on the
minor approach and vehicles taking
right turn from the major approach will
behave as if there is a give way or yield
sign, even if a stop sign is present.The
driver will not perform the mandatory
stop if adequate gap is available to
maneuver. Overall, the drivers’ behavior
at unsignalized intersections in India is
significantly different than that in the
developed countries. Thus, it is important
to investigate and model the traffic
characteristics at such intersections.
Many analytical and simulation models
for unsignalized intersections are
based on the gap acceptance behavior
of drivers (Tanner 1962, 1967. The
Highway Capacity Manual (HCM
2010) uses gap acceptance approach for
determining the capacity of unsignalized
(sign controlled) intersections. Thus, it is
necessary to verify if the gap acceptance
approach is suitable for at least some
type of unsignalized intersections in
India. Data is collected and extracted at
four T-intersections with limited priority
and analyzed gap acceptance for the
major road and minor road right turning
movements. Vehicles are driven on the
left side in India, thus the right turns are
critical. Binary logit models are developed
that predict the gap acceptance of right
turning vehicles.Various methods in the
literature are used to estimate critical gap
values.
This paper is organized in seven sections
including this section. The second section
gives a brief background and literature
related to this study. Section three
discusses the data collection procedure
* Assistant Professor, Email: [email protected]
** Ph. D. Student, Email: [email protected] Department of Civil Engineering, Indian Institute of Technology Bombay,
Written comments on this Paper are invited and will be received by the 10th June, 2015
Journal of the Indian Roads Congress, January-March 2015
Gap Acceptance Behavior of Right-Turning Vehicles At T-Intersections - A Case Study
and geometry of intersections selected
for this study. Preliminary data analysis
and distribution fitting for gaps is given
in section four. Section five gives the
structure and procedure for development
of binary-logit models. Section six
discusses the validation of models
with field observations of binary-logit
models and its prediction. Section seven
discusses the estimation of critical gap
and section eight concludes the paper.
2.
(a) Turning from Major to Minor Approach
45
(b) Turning from Minor to Major Approach
Fig.1: Measurement of Lag
BACKGROUND
AND roads at all the intersections considered
LITERATURE REVIEW
in this study are four-lane divided and
the lane discipline is not observed at the
Gap is defined as the time interval intersections. Therefore, gap between
between passing the rear bumpers of the two consecutive vehicles irrespective of
leading vehicle and the front bumper of their lateral position along the width of
the following vehicle on the same point the road have been measured.
moving in the same direction. Lag is
defined as the time elapsed after a right A good amount of work is available
turn intended vehicle reaches the stop in the literature on the sign controlled
line until a major approach conflicting unsignalized intersections. Some of the
vehicle reaches the conflict point (refer earlier studies are by Tanner (1962),
Fig. 1). As shown in Fig. 1, conflict Tanner (1967), and Mahmassani and
point of right turning vehicles with the Sheffi (1981). A model to estimate
through vehicles of the opposing road average delay for minor road is
is used for measuring gap/lag. The developed by Tanner (1962), and Tanner
measurement of lag for right turning
(1967) focused on capacity estimation
vehicles from major approach and minor
of minor road. Cowan (1987) extended
approach are explained with the help of
Tanner’s work by developing delay
Fig. 1(a) and 1(b) respectively. Let t­0
model considering wider class of arrival
be the time at which the subject vehicle
A, intending to take right turn, reaches stream. The effect of waiting time on gap
section Y1-Y1; and Y2-Y2 is the section acceptance is studied by Mahmassani
where the first conflicting vehicle B is and Sheffi (1981) using probit model.
located at time t­0 (Fig. 1(a)). Assume t1 is The authors observed that the probability
the time at which the vehicle B reaches of accepting a gap increases with the
sectionY3-Y3, that is, the conflict point. increased waiting time.
In Fig. 1(b), let section Y’-Y’ be the
location of conflicting vehicle D at time A probit model to estimate the probability
t0­ when right turning vehicle C reaches of a driver to accept a given gap for
section X-X and t1 be the time at which minor road left turning at T-intersection
vehicle D reaches the conflict point at is developed by Hamed et al. (1997).
section Y’-Y’. The lag in both the cases Ruskin and Wang (2002) through their
is calculated as the difference in times t1 cellular automata (CA) model observed
that CA model captures many feature
and t0­.
of traffic behavior at unsignalized
The definitions of gap and lag as discussed intersection than gap acceptance models.
above are suitable if the major road is two- The gap acceptance decision is affected
lane (one lane in each direction) and all by driver characteristics, characteristics
vehicles follow lane disciple. The major of the gap, and the choice situation. The
driver characteristics include gender,
age, driving skill, etc (Sheikh, 1997.)
Gap and choice situation factors may
include gap size, speed of vehicles, type
of subject and conflicting vehicle types,
waiting time, type of sign control, etc.
Limited work has been done for
unsignalized intersections with no
priority or limited priority, which are
prevalent in India. Ashalatha and Chandra
(2011) proposed a method of critical gap
estimation based on clearing behavior
of vehicles. The authors estimate the
critical gaps using six different methods
available in literature and conclude
that the results by these methods are
not acceptable because the critical gap
values obtained are smaller than the
values recommended in HCM 2000 and
the values by different methods vary.
Venkatesan (2011) developed probit
based gap acceptance models for normal
merging, forced merging, group merging
and vehicle cover merging for left turning
vehicles at uncontrolled T-intersection.
The gap acceptance behavior of
right turning vehicles which involve
crossing major road through vehicles
is not considered. Recently, Sangole
et al. (2011) used ANFIS for modeling
gap accepting behavior focusing twowheelers at uncontrolled intersections.
The studies that focus on some aspects
of uncontrolled intersections include
the work by Chandra et al. (2009),
Rengaraju and Rao (1995), carried out
a study to identify suitable probability
distribution models for vehicle arrivals
at uncontrolled intersections under
Journal of the Indian Roads Congress, January-March 2015
Patil & Sangole on
46
mixed traffic conditions. It was observed
that Poisson distribution gives a close
fit to vehicle arrivals, if traffic volume
is less than 500 vehicles/hour/lane. For
higher traffic volumes, multivariate
distribution is suggested. The authors
in another study developed a model
to estimate possible conflicts at urban
uncontrolled intersection. In yet another
study on uncontrolled intersection they
used simulation to model the conflicts at
uncontrolled intersections. Chandra et al.
(2009)developed models for estimating
service delay for various vehicle types.
Based on the data collected at five
uncontrolled intersections, eighteen
exponential models for different
movements and vehicle combinations
are developed. No other variable except
conflicting traffic is used for all models.
Based on the literature review, it is clear
that more studies are required to develop
proper understanding of the traffic flow
at uncontrolled intersections. In this
process, it is also important to evaluate
the applicability of existing approaches
of sign controlled intersections. The gap
acceptance approach may not be suitable
for fully uncontrolled intersections
found in congested urban parts. Thus,
it becomes necessity to explore the
applicability of gap acceptance approach
to partially controlled intersections.
3.
DATA COLLECTION
EXTRACTION
AND
The data used in this study were
collected at four T-intersections: three
in Aurangabad city and one in Thane
city, both are in Maharashtra state.
Aurangabad is a city of about 1.2 million
people with two-wheelers as the primary
personal mode of transportation. Thane
city has about 1.8 million people and
is within Mumbai Metropolitan Region
(MMR), which is heavily dependent on
transit and para-transit (auto-rickshaws
and taxis) modes. Consequently, Thane
has lower penetration of two-wheelers
than Aurangabad. Data at Aurangabad
intersections were collected in July 2010
and at Thane intersection in November
2010. Data were collected during
morning hours (approximately between
10:00 to 11:00 am) of typical weekdays.
All the intersections are in plain terrain
and there is adequate sight distance
available for all approaches. The
intersections are sufficiently away from
upstream or downstream intersection, so
the flow was not affected by the nearby
intersections. Additionally, at the time of
data collections the intersections were
free from any encroachments.
The geometric features of two
intersections in Aurangabad (Intersection
1 and Intersection 4) and an intersection
in Thane (Intersection 3) are similar
(refer Fig.2 (a)). Both major as well
as minor roads are four-lane divided.
The major road Intersection 2 (from
Aurangabad city) is also four-lane
divided, but the minor road two-lane
undivided. Video camera was placed
on the terrace of a nearby building in
such a way that a good view of all the
three approaches is obtained for getting
attributes of the traffic stream (refer Fig.
2 (b). Recording was done for about
60 minutes at each intersection. The
recordings were played at slow speed on
a screen in laboratory to extract various
parameters.
Lags/gaps were measured in 1/100th
of second. The data extraction resulted
in 722 lags/gaps (both accepted and
rejected) for major road right turning
movement (WS movement in Fig. 2(a)
and 1066 lags/gaps for minor road
right turning (SE) movement. These
observations are from 384 major road
and 530 minor road vehicles.
A summary of traffic composition at
all intersections, as well as the mode
wise share (in %) is given in Table 1.
For the traffic compositions of all the
movements shown in Fig. 2(a) refer
Sangole et al. (2011). It can be observed
that the proportion of two-wheelers at all
intersections is much higher than other
modes. This proportion is significantly
high at the three intersections in
Aurangabad city, (74.47%, 69.93%, and
78.76% respectively at intersections 1,
2, and 4). Public transit is heavily used
in Mumbai and neighboring cities, thus
Thane has lesser share of two-wheelers
than Aurangabad.
The gap acceptance of movements WSthe right turning movement from major
to minor approach and SE–the right
turning movement from minor to major
approach are the focus of this study
(see Fig. 2). The conflicting movement
for both WS and SE is EW-the through
traffic from east to west on the major
road. The movements WS and SE are
conflicting to each other; however, it is
observed that the WS has priority over
SE at all intersections.
4.
DISTRIBUTION OF GAPS
Fitting distribution to available gaps is
useful in traffic flow simulation and may
help in analyzing and evaluating the
traffic on transportation facility.
Fig. 3 shows the histogram of available
gaps (accepted and rejected gaps) in
the major stream (EW movement) for
the both right turning movements. As
seen in the Fig. 3, different distributions
fitting are tried to the available gaps data.
The probability distribution functions
(pdf) of normal distribution, lognormal
distribution, exponential distribution,
and gamma distribution fit for all the
four intersections as shown in the Fig.
3. Kolmogorov-Smirnov (K-S) test is
used to measure the goodness of fit for
fitted distribution. K-S test value is the
maximum distance between the empirical
distribution function of the sample and
the cumulative distribution function of
the reference distribution.
Journal of the Indian Roads Congress, January-March 2015
Gap Acceptance Behavior of Right-Turning Vehicles At T-Intersections - A Case Study
47
weight coefficients. For a binary logit
model, the probability of choice i by
driver k is given as:
a) Geometry of intersection
b) Location of video camera
Fig.2: Geometry of Intersections and Locations of Video Camera
Table 2 gives the mean, standard deviation
of gap data, and K-S test values for the
fitted distribution. K-S test critical values
at 95% are also given in the last column.
As seen in the table, the K-S test values
for Gamma distribution are the lowest
and less than the critical values for three
intersections (intersection 1, 2, and 4);
intersection 3 has the lowest K-S value
for lognormal distribution, but the value
for gamma distribution is also within the
threshold values.
lowest K-S test values, which are also
within the threshold values.
5.
DEVELOPMENT OF BINARYLOGIT MODELS
NLOGIT 4.0, popular econometric
software, is used for developing logit
models. Two separate models are for
major road right turning and minor
road tight turning. About 20% of the
data for each movement are selected
randomly and kept for validation and
the remaining data are used for models
development. The major road and minor
road right turning models are developed
with 576 and 852 gaps/lags, respectively.
Different combinations of different
variables are considered but models with
one variable have been chosen, GAP
for both the major and minor road right
turning. The models along with various
model parameters are presented in
Table 4 and Table 5 respectively.
Some earlier studies have used discrete
choice models for the driver’s decision
of accepting or rejecting gaps. In this
section, the implementation of binary
logit model to capture the gap acceptance
behavior of right turning drivers is
discussed. The formulation of logit
Also different distributions are fitted for model is based on random utility theory. 6. VALIDATION
AND
accepted gaps only at each intersection. The deterministic component of utility
PREDICTION OF MODELS
Fig. 4 shows the normal distribution, of an alternative i is expressed as:
lognormal distribution, exponential
As mentioned above about 20% of the
distribution, and gamma distribution
data (146 gap/lag for major and 213
fit of accepted gaps. Various statistical
for minor road) are randomly selected
parameters including K-S test values where, Vi = deterministic component and kept for the validation of the
for accepted gaps are given in Table 3. of utility of choosing a particular various models developed Simply using
It is clear that for all the intersections, alternative; α = constant; X1,X2­,.., ­­Xn = prediction by model can sometimes
lognormal distribution results in the independent variables; and β1, β2,… βn= result in mislead interpretations and
hence Receiver Operator Characteristic
(ROC) curve are used for binary decision
problem. ROC curves shows how correct
Table 1: Traffic Composition and Mode Wise (%) Share at All Intersections
predictions of model vary with incorrect
predictions of model. Precision-Recall
Intersection no. Two-wheeler Auto-rickshaw Car
Bus/Truck Total
(PR) curve is an alternative to ROC
Intersection 1
2643
328
401
177
3549
curves and can be used if data is highly
(Aurangabad)
(74.47)*
(9.24)
(11.30) (4.99)
(100) skewed. In ROC curves data should be
Intersection 2
1495
271
262
110
2138
present in the upper left-hand corner, and
(Aurangabad)
(69.93)
(12.68)
(12.25) (5.14)
(100)
in PR curves, it should be in the upper
Intersection 3
1197
818
1038
205
3258
right-hand corner. Model results are
(Thane)
(36.74)
(25.11)
(31.86) (6.29)
(100)
validated by calculating the Type I and
Intersection 4
2422
378
258
17
3075
Type II error, and comparing ROC and
(Aurangabad)
(78.76)
(12.29)
(8.39) (0.55)
(100)
PR curves. The output of binary logit
* Percentage share
Journal of the Indian Roads Congress, January-March 2015
Patil & Sangole on
48
model is the probability of accepting
the lag/gap and it varies from 0 to 1. If
the probability of accepting lag/gap is
greater than 0.5, then that particular lag/
gap is accepted.
(a) Intersection 1
(b) Intersection 2
(c) Intersection 3
(d) Intersection 4
Fig. 3: Distributions Fitted for Accepted Gaps
In a binary decision problem, the
validation model outputs can be
represented as a confusion matrix or
contingency table (Provost et al., 1998).
Table 6 gives the format of confusion
matrix along with some definitions. From
the validation results, it is necessary to
group the responses into the following
categories to develop a confusion
matrix: i) True Positives (TP) (response
correctly labeled as positives), ii) False
Positives (FP) (response incorrectly
labeled as positive), iii) True Negatives
(TN) (negatives correctly labeled as
negative), and iv) False negatives
(FN) (response incorrectly labeled as
negative). It is considered that gap
acceptance as condition positive and
gap rejection as condition negative.
The False Positive Rate (FPR) is the
proportion of negative responses that
are misclassified as positive, whereas
the True Positive Rate (TPR) also
known as Sensitivity is the proportion
of positive responses that are correctly
labeled.
Calculations of confusion matrices
are done by grouping the gap values
into three ranges: less than or equal to
5 sec, greater than 5 and less than or
equal to 10, and greater than 10 and
less than or equal to 21 (see Table 7).
Calculations can also be carried out
for whole validation data as a single
group. But in that case the variation of
true positive rate with respect to false
positive rate, over particular ranges
of gap values, cannot be observed.
Cutoff points for different groups can
be decided by trial and error method to
improve sensitivity and specificity. The
cutoff point of 5 sec was found to be
suitable for the first group. Since the
number of observations for gaps larger
Journal of the Indian Roads Congress, January-March 2015
Gap Acceptance Behavior of Right-Turning Vehicles At T-Intersections - A Case Study
49
Table 2: Statistical Parameters for Available Gaps Distribution Fitting
Intersection
Inter. 1
Corr.
volume
(Veh/hr)
Mean
3549
2.68
Std. dev.
2.43
K-S test values
Normal Dist. Lognormal Dist. Exponential Dist.
0.1388
0.0788
0.1413
Gamma
Dist.
0.0459
Critical
values
0.0738
Inter. 2
2138
4.44
3.79
0.1413
0.0629
0.0866
0.0363
0.0840
Inter. 3
3258
3.66
2.01
0.1230
0.0495
0.2835
0.0658
0.1025
Inter. 4
3075
4.74
4.20
0.1350
0.0460
0.0965
0.0369
0.0426
Gamma
Dist.
Critical
values
Table 3: Statistical Parameters for Accepted Gaps Distribution Fitting
Intersection
Corr. volume (Veh/
hr)
Mean
Std.
dev.
Inter. 1
3549
3.78
2.83
0.1807
0.0452
0.2720
0.0794
0.1469
Inter. 2
2138
6.85
3.70
0.1372
0.0461
0.2715
0.0668
0.1201
Inter. 3
3258
4.37
2.10
0.1618
0.0713
0.3307
0.1010
0.2243
Inter. 4
3075
7.87
4.42
0.1615
0.0534
0.2580
0.0903
0.1097
Normal
Dist.
K-S test values
Lognormal Dist. Exponential Dist.
Table 4: Logit Model Estimation Results for Major Road Right Turn
Estimated Parameters
Name
Coefficient Standard error
Constant
-2.779
0.2530
GAP
0.9145
0.0816
McFadden Pseudo R-squared: 0.3841
Log likelihood function:
hi squared:
306.4122 Restricted log likelihood:
t-stat
-10.987
11.211
-245.6777
-398.8837
Table 5: Logit Model Estimation Results for Minor Road Right Turn
Estimated Parameters
Name
Constant
GAP
McFadden Pseudo R-squared:
Chi squared:
Coefficient
-2.3761
0.6810
0.238
277.132
Standard error
0.1687
0.0533
Log likelihood function:
Restricted log likelihood:
t-stat
-14.08
12.75
-443.516
-582.082 than 10 sec were less, all such data is
grouped into one range.
The confusion matrices calculations for
the major right turning and minor right
turning are presented in Table 8 and
Table 9 respectively as per the format
given in Table 6. The part (a) of Table
8 indicates that there are 20 gaps of
less than 5 sec are accepted in the data
and the model predicts 17 observations
correctly, resulting in the positive
prediction value of 0.85. The lower
value of positive prediction rate in
Table 9 (a) is due to the small number of
accepted data for gaps less than 5 sec in
the validation data set. In general, it can
be noted that prediction rates of positive
values are higher than of the negative
value. In other words, the models are
predicting gap acceptance decisions
better than the gap rejection.
(a) Intersection 1
Journal of the Indian Roads Congress, January-March 2015
Patil & Sangole on
50
Table 7: Data for Model Validation
Major Road
Minor Road
Gap
range
Accepted Rejected Accepted Rejected
(in sec)
0 to 5
20
73
16
119
5 to 10
30
4
49
3
10 to 21
13
0
24
0
Table 8: (a), (b), (c) Calculations of
Confusion Matrices for Major Road Right
Turning Model
(a) At cut-off point less than or equal to 5
(b) Intersection 2
Logit
Model
Prediction
Actual Output
17
3
21
57
0.85
0.7307
0.4473
0.95
0.5526
0.05
(b) At cut-off point less than or equal to 10
Logit
Model
Prediction
Actual Output
43
7
(c) Intersection 3
23
59
0.86
0.7195
0.6515
0.8939
0.3485
0.1060
Logit
Model
Prediction
(c) At cut-off point less than or equal to 21
Actual Output
56
7
0.8889
23
0.7195
59
0.7088
0.8939
0.2911
0.1060
(d) Intersection 4
Fig. 4: Distributions Fitted for Available Gaps.
Table 6: Format of Confusion Matrix (Provost et al., 1998)
Test
Test
Outcome Outcome
Negative Positive
Test
Outcome
Actual Condition
Condition Positive
Condition Negative
True Positive
False Positive
(Type I error)
Positive predictive value =
Σ True Positive
Σ Test Outcome Positive
False Negative
(Type II error)
True Negative
Negative predictive value =
Σ True Negative
Σ Test Outcome Negative
Sensitivity =
Specificity =
Σ True Positive
Σ True Negative
Σ Condition Positive
Σ Condition Negative
False negative rate (β) = type II error = 1 − sensitivity
False positive rate (α) = type I error = 1 − specificity
Journal of the Indian Roads Congress, January-March 2015
Gap Acceptance Behavior of Right-Turning Vehicles At T-Intersections - A Case Study
51
Table 9 (a), (b), (c) Calculations of
Confusion Matrices for Minor Road Right
Turning Model
(a) At cut-off point less than or equal to 5
Logit
Model
Prediction
Actual Output
1
3
0.25
15
116
0.8854
0.0625
0.9747
0.9375
0.0252
(a) ROC curve (b) PR Curve
(b) At cut-off point less than or equal to 10
Fig. 5: ROC Curve and PR Curve for Major Road Right Turning Vehicles
Logit Model
Prediction
Actual Output
44
4
0.9166
21
118
0.8489
0.6769
0.9672
0.3230
0.0327
(c) At cut-off point less than or equal to 21
(a) ROC curve (b) PR Curve
Fig. 6: ROC Curve and PR Curve for Minor Road Right Turning Vehicles
Logit Model
Prediction
Actual Output
68
4
0.9444
21
118
0.8489
0.7640
0.9672
0.2359
0.0327
Fig. 7: Predicted Probabilities
The Receiver Operating Characteristic
(ROC) and Precision Recall (PR)
curves presented in Fig. 5 and Fig. 6
are constructed using these confusion
matrices of Tables 8 and 9. Recall used
in the PR curves is the same as TPR and
it is the fraction of responses classified
as positive that are truly positive i.e.
positive predictive value. For obtained
a ROC curve the true positive rate
(Sensitivity) is plotted against the false
positive rate (1 - Specificity) for different
cut-off points. Each point on the ROC
curves represents a sensitivity/specificity
Journal of the Indian Roads Congress, January-March 2015
52
Patil & Sangole on
Table 11: Critical Gap by Raff’s Method for Different Subject Vehicle Type and
pair corresponding to a particular group
Conflicting Vehicle Type
of gaps decision. ROC curve lying in
the upper left corner indicates that the
Subject Vehicle
Conflicting Vehicle Type
model predictions are well. It can also
Type
be observed that specificity increases
Major road right turn
Minor road right turn
slightly with increase in TPR and gap
TW
AR
CAR
TW
AR
CAR
values.Ideally, a PR curves should
TW
2.8
3.25
3.35
3.10
3.25
3.60
overlap with the diagonal lines, which
AR
2.65
2.70
2.80
3.55
3.30
2.90
will happen if the models’ prediction is
3.40
2.30
3.00
3.25
2.60
3.50
100% accurate. The PR curves presented CAR
are not deviating much from the diagonal
lines, indicating good prediction by
critical gap is the minimum gap required well. For these distributions Ashworth’s
models. Presence of PR curve on upper
for the road user or driver to make the method gives close approximation. The
part of diagonal line indicates more
maneuver safely. Different methods maximum likelihood method is based on
positive responses in the data.
are available for determining critical the assumption that a driver’s critical gap
Fig. 7 shows the predicted probabilities of gap; some of them are Raff’s method, is within the range of his/her accepted
developed logit models. The percentage Lag method, Ashworth’s method, and gap and the largest rejected gap during a
of correct prediction by binary logit Maximum likelihood estimation method. single decision making. In logit method,
models are 74.48% for major road right Raff defines the critical gap as the gap the gap for which the probability of
turning and 81.51% for minor road right for which the number of accepted gaps acceptance is 0.5 is taken as the critical
shorter than it is equal to the number gap.
turning.
of rejected gaps longer than it. In lag
method, only lags are considered for Table 10 gives the critical gap values
7. ESTIMATION OF CRITICAL critical gap estimation. Ashworth’s for the four intersections, calculated by
GAP
method estimates critical gap based on various methods. Almost all critical gaps
accepted gaps/lags only. It is based on the values vary between 2 to 5 seconds.
Critical gap is a primary parameter for the assumption that major stream gaps are The mean values of critical gaps and
deterministic gap acceptance models. It exponentially distributed and accepted standard deviations are also given in
is generally used to estimate the capacity gaps are normally distributed. It should the Table. Among all the methods, the
and delay at unsignalized intersections. be noted that the Ashworth method maximum likelihood method has the
The accuracy of the capacity and delay assumes normal distribution for accepted lowest standard deviation of critical
estimation is mainly dependent on the gaps, but normal distribution is not a gaps at different intersections showing
accuracy of the critical gap estimation. As good fit to the data in this study; gamma its consistency in estimating critical
per Highway Capacity Manual (HCM), and lognormal distributions fit the data gaps. Raff’s method gives critical gaps
comparable to maximum likelihood
method and it also has lower standard
Table 10: Critical Gap Values (In Seconds) by Different Methods (for Individual
deviation than Lag and Ashworth’s
Intersections)
methods. Ashworth’s method has the
highest variability among critical gaps at
Method
Intersections
Mean
Std. Dev.
different intersections.
1
2
3
4
The values of critical gap and critical
Mjr. Mnr. Mjr. Mnr. Mjr. Mnr. Mjr. Mnr. Mjr. Mnr. Mjr. Mnr.
lag obtained from Raff’s method for
Raff’s Method 2.40 2.50 3.35 3.00 2.25 3.70 3.75 3.20 2.94 3.07 0.73 0.60 left turning vehicles by other studies
Lag Method 2.35 2.00 3.60 3.32 2.25 3.25 4.05 4.75 3.06 3.33 0.90 1.12 in developed countries on a twolane roadway are 4.2 sec and 5.6 sec
Ashworth’s
1.76 2.26 5.25 4.45 2.85 4.86 2.32 5.26 3.05 4.21 1.54 1.34 respectively. The base critical gap on
Method
a four-lane major road given in HCM
Maximum
2.55 2.83 3.26 3.06 2.32 3.30 3.05 2.97 2.80 3.04 0.43 0.20 2000 for left turning from minor road
Likelihood
is 7.5 sec. HCM 2010 also recommends
Method
the same value but uses term critical
headway instead of critical gap. These
Note: Mjr: Major Approach; Mnr: Minor Approach
Journal of the Indian Roads Congress, January-March 2015
Gap Acceptance Behavior of Right-Turning Vehicles At T-Intersections - A Case Study
values are much higher than the values For major road right turn:
reported in this study. These insights
clearly prove that the drivers in Thane
and Aurangabad are aggressive and
choose smaller gaps than the drivers in
developed countries. Although, the data
in this study are not enough to make
For minor road right turn:
the countrywide generalized statement,
the traffic situation in other cities is not
totally different. Thus, it is expected that
similar aggressive gap acceptance will
Critical gap = 3.48 sec.
be observed in other parts of the country.
More similar studies with the data from Table 12: Critical Gap Values (In Seconds)
the different parts are needed to confirm
by Different Methods (Combined Data)
this observation.
camera. The following are some of the
conclusions based on our analysis of the
study intersections:
●●
The results in this paper indicate
that the gap acceptance concept can
be used in India at limited priority
intersections
in
uncongested
conditions.
●●
The available gaps follow gamma
distribution and accepted gaps
follow lognormal distribution.
●●
The prediction success of the
developed binary logit models for
major road right turning and minor
road right turning is about 75% and
82% respectively.
●●
Critical gap obtained by developed
logit models are comparable with
maximum likelihood method
and Raff’s method. Critical gaps
obtained in this study are smaller
than that obtained in developed
countries and suggested in HCM
2010.
Method
The different vehicle types in Indian
traffic have significantly different
characteristics from one another. In order
to underst and the gap acceptance of
different vehicles, critical gaps values are
estimated considering different subject
vehicle types and conflicting vehicle
types. Two-wheelers, three-wheelers and
cars are considered and the values are
given in Table 11. The values for heavy
vehicles are not given because the less
number of heavy vehicles making right
turns. It was generally observed that
when the subject vehicle and conflicting
vehicle are the same, the critical gap
values are smaller. Additionally, since
the right turning vehicles on minor road
perceive lower priority than the right
turning vehicles on the major road, the
critical gap values of the former are
larger.
Table 12 gives the critical gap values for
minor road and major road right turning
movements considering all intersections
together. Also the critical gap values from
the logit models (developed in section
5 (Table 4 and 5)) are calculated. The
binary models are developed combining
data for all the intersections; thus the
method is used only for the combined
critical gap values given in Table 12.
Using the utility function developed
logit model takes the following forms for
critical gap estimation:
Major Minor
Major and
right right
Minor Right
turn turn Turns Combined
Raff’s Method
3
3.3
3.18
Lag Method
3.05
3.3
3.2
Ashworth’s
3.6
4.01
3.84
Method
Maximum
2.79 3.04
2.91
Likelihood
Method
Logit method
3.03 3.48
3.26
As seen in Table 12, although the
critical gaps of right turning from major
road and minor road are comparable,
the values of the latter are found to be ●●
larger by about 10%. At intersection 2,
the critical gap by all the methods for
major right turn is smaller than that for
minor right turn.
8.
CONCLUSIONS
Unlike in developed countries, the
unsignalized intersections in India
are not properly controlled, that is,
the priorities of different movements
are not fully respected by drivers. At
a few intersections, limited priorities
are observed where the right turning
vehicles look for suitable gaps in
through vehicles. The primary focus
of this paper is to understand the gap
acceptance behavior of vehicles at
such limited priority intersections. Gap
acceptance data were collected at four
T-intersections with the help of video
53
The critical gaps obtained by
various methods indicate the
drivers at studied intersections are
aggressive and accept lower gaps.
Among different method maximum
likelihood method found to give
more consistent results.
Although the findings of this study can
become a building blocks for developing
detailed methodology for unsignalized
intersections in India, the study has some
limitations. The data for this study have
been collected at only four intersections
in two cities of Maharashtra. The
applicability of the models and
conclusions need to be verified at more
places in India. Many more such studies
are needed focusing traffic conditions in
India and other developing countries. In
this study, only right turning movements
at T-intersection are considered; the
Journal of the Indian Roads Congress, January-March 2015
54
Patil & Sangole on Gap Acceptance Behavior of Right-Turning
study can be extended to four legged
intersections and more movements can
also be studied. Effect of some other
parameters such as approach speed,
geometric
characteristics,
driver’s
characteristics, etc. will further increase
the understanding of the traffic behavior
at uncontrolled intersections.
ACKNOWLEDGMENT
This study is partially funded by
Department of Science and Technology
(DST), Govt. of India, through project
SR/FTP/ETA-61/2010. The authors
would like to thank Mr. Prasad Patare
for helping during the data collection
and the preliminary analysis.
REFERENCES
1.Tanner, J.C. (1962). A Theoretical Analysis
of Delays at an Uncontrolled Intersection.
Biometrika 49(1/2):163-170.
2.Tanner, J.C. (1967). The Capacity of an
Uncontrolled Intersection. Biometrika
54(3/4):657-658.
Vehicles At T-Intersections - A Case Study
3.Mahmassani, H., Sheffi, Y. (1981).
Using Gap Sequences to Estimate Gap
Acceptance Functions. Transportation
Research Part B: Methodological
15(3):143-148.
8.Venkatesan K. (2011). Study of Merging
at Urban Uncontrolled Major-Minor Road
Intersections Under Heterogeneous Traffic
Condition. PhD dissertation. Indian Institute
of Technology Madras, Chennai.
4.Cowan, R. (1987). An extension of Tanner’s
Results on Uncontrolled Intersections.
Queueing Systems 1(3):249-263.
9.Ashalatha R., Chandra, S. (2011). Critical
Gap through Clearing Behavior of Drivers
at Unsignalised Intersections. KSCE Journal
of Civil Engineering, 15(8), 1427-1434.
5.Hamed, M. M., Easa, S. M., Batayneh, R.
R. (1997). Disaggregate Gap-Acceptance
Model for Unsignalized T-intersections.
Journal of Transportation Engineering,
123(1), 36-42.
6.Ruskin, H.J., Wang, R. (2002). Modeling
Traffic Flow at an Urban Unsignalized
Intersection. In Computational Science
— ICCS 200 (Lecture Notes in Computer
Science) Springer Berlin Heidelberg, vol.
2329, pp. 381-390.
7.Abu Sheikh, A.M. (1997). Developing
Behavioral Models for Driver Gap
Acceptance at Priority Intersection. Ph.D.
Thesis, King Fahd University of Petroleum
and Minerals Dhahran, Saudi Arabia.
10.Sangole J.P., Patil G.R., Patare P. S. (2011).
Modelling Gap Acceptance Behavior of
Two-Wheelers at Uncontrolled Intersection
Using Neuro-Fuzzy. Procedia-Social and
Behavioral Science. 20, 927-941.
11.Rengaraju, V., Rao, V. (1995). VehicleArrival Characteristics at Urban Uncontrolled
Intersections. Journal of Transportation
Engineering 121(4):317-323.
12.Chandra, S., Agrawal, A., Rajamma, A.
(2009). MicroscopicAnalysis of Service Delay
at Uncontrolled Intersections in Mixed Traffic
Conditions.JournalofTransportationEngineering
135(6):323-329.
Journal of the Indian Roads Congress, January-March 2015