236
Macroscopic Analysis of Side Friction Impact on Speed of
Urban Roads-Case Study of “Jawahar Marg, Indore”
Ravi Parmar1, Anu Maru2, Rishi Singh Chhabra3
1
Design Engineer (Highway), GR Infra Pvt Ltd, New Delhi
2
Assistant Professor, Shri Vaishnav Institute of Technology and Science, Indore
3
Assistant Professor, Lovely Professional University, Punjab
[email protected], [email protected], [email protected]
Key words: Side friction, macroscopic analysis, friction factor, urban road, speed
1
Abstract
This study involved a macroscopic approach in which we calculate the combined friction factor
and also used SPSS software for finding the friction (FRIC). ‘FRIC’ is a short version of
‘friction’. ‘FRIC’ was determined by adding the weighted standardized coefficients of the
negatively correlated individual factors to the criterion variable. SPSS computer program was
used for the computation of the above equation and the results included the beta coefficients,
which are standardized coefficients. The results showed that on two-lane two-way roads, all
studied factors exhibited statistically significant impact on speed. On this basis it was possible
to measure the characteristic impact of the individual factors on speed. The macro analysis
approach was based on developing relationship models between speed and flow and
investigating the effect of aggregated side friction factors on such relationships.
2
Introduction
Road side friction mainly include road side activities like parking manoeuvres, frequent bus stops,
entries and exit from major roads etc. As a consequence of these factors, the traffic flow characteristics
of vehicles get affected. Side friction factors are defined as all those actions related to the activities
taking place by the sides of the road and sometimes within the road, which interfere with the traffic flow
on the travelled way. They include but not limited to pedestrians, bicycles, non-motorised vehicles,
parked and stopping vehicles. The objective of this research is to analyses the effect of these factors on
traffic performance measures on urban roads. An empirical case study methodology was adopted where
Jawahar Marg Road in Indore city was chosen as a representative case, three road sections on Jawahar
marg were selected. The scope was limited to include only road-link facilities. Data collection was
mainly performed by application of video method, which proved to be effective for simultaneous
collection of traffic and side friction data. Data reduction was conducted chiefly by computer, using
standard spreadsheet. This review involved a macroscopic approach where traffic and friction data were
collected and analyzed at an aggregated level. The analysis part was based on statistical methods, chiefly
regression analysis. The individual friction factors through regression analysis were weighted and
combined into one unit of measure of friction called ‘FRIC’. Free-flow vehicles were observed in the
recorded videocassettes interacting or not interacting with different friction factors. They were assessed
to be ‘free flowing’ on the screen and confirmed by computation of time headways between successive
vehicles. These were sorted according to types of friction factors they interacted with and also those,
which did not interact with any friction factor, were recorded in a different class. All free flow vehicles
that interacted with friction factors were labelled according to the type of friction factor they interacted
with, such as: those which interacted with pedestrians they were labeled (PED), Bicycles (BIC), NonMotorized Vehicles (NMV), and Parking/Stopping Vehicles (PSV). For those, which did not interact
with any friction factor, were labeled as ‘No Side Friction’ (NO SF).
3
Analysis of Side Friction
Analysis of side friction is the core of this research. The main objective was to identify if friction factors
explained any variation in mean vehicles speeds and capacity on the different studied roads. The studied
sites were firstly investigated individually and then aggregated according to road types. Four types of
side friction events were identified. These included pedestrians walking along or crossing the roadway,
bicycles, non-motorized vehicles, parked and stopped vehicles, all measured in numbers per hour per
length of study section. Figure below shows an example of the average 15-minute flow and friction
events data recorded over 2 hrs at Section 1:
Fig. 1 Flow rate and friction events (based on 10minutes-data)
Figure above indicates that in the study area there were several factors, which could be explanatory to
the dependent variables such as capacity and speed. From the figure, it is also noted that ‘flow’ was the
most prominent factor among all the above shown, and likely would have the highest explanatory power
than the others. Based on the above findings, it was proposed to study the impact of the above factors on
speed and capacity as described below.
3.1
Impact Analysis of Side Friction Factors
Effective way of performing impact analysis of side friction was to combine the individual frictional
components into one unit of measure. This had two advantages:
•
•
3.2
To reduce the no. of variables and consequently reduce the complexity of the analysis.
To make it easier to judge whether friction was significant or insignificant on a given site in
terms of one unit of measure instead of dealing with individual components.
Combining Friction Factors Process (determination of ‘FRIC’)
Based on the advantages mentioned above, it was decided to combine the individual friction factors into
one unit of measure. A name for this unit was called ‘FRIC’. The choice of the name was arbitrary to
suit the theme of the study. ‘FRIC’ is essentially a short version of ‘friction’. ‘FRIC’ was determined by
adding the weighted standardized coefficients of the negatively correlated individual factors to the
criterion variable. It was thus important, first to establish these coefficients by performing regression
analysis involving flow and the individual frictional items as independent variables and speed of the
light vehicles as the criterion variable. The following equation was applied:
Vo= A+B (FLOW)+ B (PED) + B (BIC)+ B (NMV) + B (PSV) ……….1
1
2
3
4
5
Where,
Vo = Observed average speed of light vehicle
A = Regression constant which is the free-flow speed
B , B , B , B , B = Regression coefficients for explanatory variables.
1
2
3
4
5
SPSS computer program was used for the computation of the above equation and the results included the
beta coefficients, which are standardized coefficients. These were weighted and summed up to determine
the unit measure of ‘FRIC’ by applying the following equation:
FRIC = A*PED + B*BIC +C*PSV + D*NMV …………………2
Where,
PED= Pedestrians (No. /hr)
BIC=Bicycles (No./ hr)
PSV= Parking and stopping Vehicles (No./ hr)
NMV=Non-motorized Vehicles (No./ hr)
A, B, C, D= Weighted standardized coefficients
In the table it is indicated that those factors, which showed positive correlation with the
dependent variable, were excluded from the derivation of ‘FRIC’. Negative correlation implied that the
factors reduced speed while positive correlation implied the opposite, such that the presence of friction
factors increased speed, which was considered as an anomaly and intuitively not expected, and hence
excluded. It is important to note that factors with negative correlation were included in the analysis
regardless of their significance levels. This was based on the assumption that despite showing low
significance levels as individual factors, yet they could anyhow have a contribution when combined
together.
Table.1 Determination of ‘FRIC’
Site
Section:1
Indepandent
Variable
Significance
Level
Regression
Coefficients
Beta
Coefficient
Flow
0.1856
0.015262
0.616485
Weighted
Coefficient
FRIC=A(PED)+
B(BIC)+C(NM
V)+D(PSV)
FRIC=.157(PE
D)+1(BIC)+0.3
(R2=0.571)
Section:2
(R2=0.092)
SIYAGANJ
PETROL
PUMP(R2=0.67
9)
46(NMV)
PED
0.8992
-0.00772
-0.15717
0.465
BIC
0.33861
-0.11129
-0.33728
1
NMV
0.79999
-0.05103
-0.1167
0.346
PSV
0.57222
0.083201
0.207819
Flow
0.879
0.003
0.075
PED
0.863
-0.015
-0.145
0.872
BIC
0.861
-0.022
-0.166
1.000
NMV
0.510
0.183
0.571
PSV
0.690
0.063
0.278
Flow
0.171
0.015
-0.135
PED
0.562
0.069
-1.064
BIC
0.273
-0.127
0.396
NMV
0.499
-0.166
-0.182
PSV
0.939
-0.016
0.281
Fig. 2 An example of the combined friction components into ‘FRIC’ are shown in figure for Section 1:
FRIC=0.872(PE
D)+1(BIC)
1.000
FRIC=1(PED)+
0.172(NMV)
0.172
3.3
Impact analysis with ‘FRIC’
‘FRIC’ was the practical unit to use in the analysis of the impact of friction on speed and capacity of the
studied roads. This was studied in three ways:
•
•
•
3.3.1
Preliminary investigation of speed-FRIC relationships. This is demonstrated for aggregated data
for the combined sites in the subsequent chapters.
Develop regression models for each site “with” and “without” ‘FRIC’ as an explanatory
variable
Develop speed flow models for different ‘FRIC’ intensity classes.
Regression models “with” and “without” ‘FRIC’
Linear regression models may include various explanatory variables that describe the variation of the
dependent variable. If all possible explanatory variables are included, the explanatory power of the
2
model (R -value) approaches 100 per cent or 1. If only few of them are included, the explanatory power
2
becomes much less than one. In essence the R -value measures the proportion of variation in the
dependent variable that is explained by the predictor variables. In this study one of the approach was to
2
study the impact due to ‘FRIC’ on each site by comparing the R -values of the models “with” and
“without” ‘FRIC’ as an independent variable. The interpretation is that speed is influenced by several
factors in the traffic environment, which cannot all possibly be studied together. In this study thus, two
were studied, which included ‘flow’ and ‘FRIC’. The logic was that both of these variables would likely
account for higher proportion of variation in the speed variable as compared to only one of them. Based
on the above explanations, firstly regression model that included only ‘flow’ as an explanatory variable
2
and speed as a response variable was analyzed, and the R -value thereon obtained was compared to the
results of the other model, which included both ‘flow’ and ‘FRIC’ as explanatory variables. The
2
different R -values obtained in these two situations, indicated the impact of ‘FRIC’ on speed. The SPSS
program was applied in the analysis of the regression models and the results are depicted in table below:
Table.2 Comparison of explanatory powers of models “with” and “without” friction factors
Road Type
SITE
R2
Model “with” ‘FRIC’
Model “without” ‘FRIC’
2lane-2way
Section 1
.51
.328
Section 2
.016
.10
Section 3
.517
.002
From above table the following facts were observed:
2
At S- 2 site, R -values for both models were nearly equal, hence ‘FRIC’ had negligible effect
•
•
3.4
2
At Section 3 site, the difference between the models’ R -values was substantial; hence ‘FRIC’
had the greatest impact.
Generally ‘FRIC’ exhibited impact on each of the different sites at varying degrees depending
on site-specific characteristics.
Speed flow models for different FRIC-classes
In this approach, speed-flow relationships were compared during different intensities of ‘FRIC’ on each
site. ‘FRIC’ was categorized into three classes of intensity representing low, medium and high levels of
intensity. This classification was arrived at after transformation of friction factors into ‘FRIC’ units,
which was observed to range between 70 ‘FRIC’-units/10 min. to 149 FRIC units/10 min. on different
sites. Based on this range, and a revisit on site observation in the recorded video clipping, ‘FRIC’ was
classified as shown in table below:
Table. 3 FRIC classes
FRICTION CLASS
CODE
FRIC units/h/200m
LOW
L
0-70
MEDIUM
M
70-90
HIGH
H
Above 90
Under the above classification, the observed data on some sites were short of full range, i.e. on many
sites only low to medium FRIC levels were observed hence making the comparison between low and
high FRIC levels less effective. However, for demonstration purposes the observed FRIC data along
with their corresponding speed and flow data were rearranged in ascending order and divided into three
equal percentiles where the upper, middle and bottom percentiles represented low, medium and high
FRIC-levels respective. This is exemplified in table below:
Table. 4 An example categorizing speed-flow data according to FRIC
Site
Speed
(Km/H)
Flow
(PCU/H)
22.05
11687
Descending Order of ‘FRIC’ Data
(‘FRIC’Average Units/Hr/30m)
Section 1
‘FRIC’ Class
132.14
HIGH
Section 2
18.67
18992
127.04
Section 3
16.63
11702
100
The above procedure was adopted for all sites, where the upper and the lower portions of data were
compared in order to see any visual impact between the two. There was generally a clear impact that was
visually observed on many sites.
T-test was performed to identify if this impact was statistically significant. The results showed that on all
sites except one the impact was significant as shown in table below:
Table.5 Significant value from t test
Average speed (km/hr)
T-test two tails
Low ‘FRIC’
High ‘FRIC’
Significance level: 0.05
Impact
Significance of
FRIC
impact;
Site
2LANE-2WAY ROADS
Section 1
17.54
14.23
0.1760
Significant
Section 2
16.65
14.39
0.0000
Significant
Section 3
4
20.32
16.54
0.0204
Significant
Conclusion
This research has been concerned with the concepts, theories, and methods related to side friction impact
on speed of urban road, and were performed in Jawahar marg, Indore as a case study. A major part of
this work has focused on issues related to identification of side friction factors and application of various
measurement methods and analysis techniques used for determining their impact on speed at an
aggregated level (macroscopic) and at an individual level (microscopic). On the basis of the macroscopic
study, modelling based on regression analysis was performed for this purpose. Analysis of individual
friction factors per se is quite rare in the prevailing research literature: On this basis the microscopic
study was performed where measuring and evaluation methods were devised with a view to obtaining
important knowledge concerning their potential impact on speed. Generally, the study was performed in
two parts based on these two methods of approach (macroscopic and microscopic). Both parts were
related in terms of objectivity and only differed in approach and detail. They both focused on attaining
the prescribed objectives, which were addressed by the two methods as follows:
5
Discussion on Attainment of Objectives
1. Greenshield model is fit for all three sections which give the free flow speed for all the sections. The
standard equation has been developed to determine the FRIC value with the help of SPSS.
FRIC = A*PED + B*BIC +C*PSV + D*NMV
Where,
PED= Pedestrian, BIC= Bicycle, PSV= Parking and stopping vehicle, NMV= Non
motorize vehicle.
A, B, C, D is weightage coefficients (from SPSS Analysis)
‘FRIC’ value (‘FRIC’AVERAGE units/hr/30m)
Section 1
= 132.14
Section 2
= 127.04
Section 3
= 100
2. A visual inspection method is also developed with the help of Speed-Flow curve with low
high FRIC value. (Macroscopic analysis)
5.1
and
Practical Implications
The characteristic impacts of the individual side friction factors identified in this study can equally be
useful to transportation planners in developing countries where the problem of friction is apparent by
guiding their priorities in planning and traffic operations. For instance if existing or new arterial road is
planned for high speed, measures should be taken to reduce as much as possible the presence of factors
with high characteristic impact values. Undoubtedly knowledge of individual factors can be useful for
development of ‘friction management’ programs seeking to limit levels of friction on high mobility
urban arterial streets so that would improve traffic safety and operation efficiency. Such programs may
include solutions for mitigation like:
•
•
•
5.2
Design standards for the arterial streets intended for high mobility should appropriately address
prevailing conditions such as to adequately provide for the needs of the road users, i.e. wide
sidewalks for pedestrians, adjacent ways for bicycles and non-motorized vehicles, and wide
shoulders for stopping and parking manoeuvres for motor traffic. If the available road width
does not permit separate facilitation for such needs, traffic management measures such as
parking regulations, restriction of non-motorized vehicles, bicycles and pedestrians along the
main road must be considered.
Implementation of sound urban-planning policies where land use activities are limited or rather
restricted along high mobility streets, including various social and economic activities. This
would in effect limit accessibility to the roadway.
Implementation of strategic education-based campaigns aimed for awareness and law
enforcement programs on road use regulations to be abided by all road-users.
Recommendations and Conclusion
This study was arguably conducted on a small scale. This study included only limited number of friction
factors. This was reflected by the explanatory power of the regression models relating speed to flow and
friction, which varied from poor to good depending on site. There were clearly site-specific factors in
operation that had not been identified or measured. It is thus recommended to conduct this study on a
much larger scale including a wider range of all frictional components in order to account for much of
the variation in the criterion variable. Similarly, larger scale-study would imply to include wider
spectrum of facilities such as intersections, roundabouts, ramps and different terrains. It is likely that the
effect of different friction factors would vary for different facilities and different terrains.
It has been shown that side friction can have effects on in Jawahar Marg, Indore which
indicated considerable effects like other commonly used factors in capacity analysis. This leads to the
recommendation that highway capacity studies, particularly in the developing world, should include this
variable, though in a form suited to their own particular circumstances.
References
1.
2
3
4
5
6
Chiguma M L. Sweden “Analysis of side friction impacts on urban road links. Case study, Dares-salaam” , Swedish National Road and Transport Research Institute (VTI), 2007, pp. 176.
Patel C R, Joshi G.J. “Mixed Traffic Speed–Flow Behavior under Influence of Road Side
Friction and Non-Motorized Vehicles: A Comparative Study of Arterial Roads in India” World
Academy of Science, Engineering and Technology International Journal of Civil,
Environmental, Structural, Construction and Architectural Engineering, 2014, 8(11),pp. .
Chandra S, Chand S. “Impact Of Bus Stop On Urban Traffic Characteristics – A Review Of
Recent Findings” , Journal of Society for Transportation and Traffic Studies (JSTS), 2014,5(2),
pp. 1-16.
Sharma HK, Swami M, Swami B L. “Speed flow analysis for interrupted oversaturated traffic
flow with heterogeneous structure for urban roads”, International Journal for Traffic and
Transport Engineering, 2012, 2(2), pp. 142 –52.
Greenshields B D, “A Study of Traffic Capacity.” , Proceedings of Highway Research Board,
Washington D.C. 1935, 14, pp. 458-67.
IRC: 106-1990, “Guidelines for Capacity of Urban Roads in Plain Areas”, Indian Roads
Congress: New Delhi, 1990.