CHAPTER 4: GEOMETRIC DESIGN OF HIGHWAYS

4.1 Introduction
Geometric design is an aspect of the highway design dealing with the visible dimensions of a
roadway. It is dictated, within economic limitations, by the requirements of traffic and includes
the design elements of horizontal and vertical alignment, sight distance, cross-section components,
lateral and vertical clearances, intersection treatment, control of access etc [Kadiyali, 2006].
The purpose of geometric design is to reduce the number and severity of road accidents while
ensuring high traffic flow with minimum delay to vehicles [Thagesen, 1996].
The safe, efficient and economic operation of a highway is governed to a large extent by the care
with which the geometric design has been worked out. Safety or the lack of it is an immediate
corollary of the various design features of the highway. Efficient and comfortable operation of
traffic is possible only if the design elements have been meticulously considered. A well designed
highway has to be consistent with economy. Too liberal standards may not fit in with the available
resources, whereas if the standards are too low, the cost of operation may mount up [Kadiyali,
2006].
The basic inputs are the Design speed and the Design hourly volume. The design speed governs
the design of vertical and horizontal curvatures while design hour volume governs capacity
required.
4.2 Considerations taken when selecting the design standards for a highway.
 Adequate geometric design in planning a highway facility ensures that the facility will not
become obsolete in the foreseeable future. Hence the volume and composition of traffic in
the design year should be the basis of design.
 Faulty geometrics are costly, and in some cases impossible to rectify at a later date and so,
due to consideration should be given to geometric design at the initial stage itself.
 The design should be consistent with and the standards proposed for different elements
should be compatible with one another. Abrupt changes in design should be avoided.
 The design should embrace all aspects of geometrics of the road, including signs, markings,
proper lighting, intersections, etc.
 The highway should be considered as an element of the total environment and its location
and design should enhance rather than degrade the environment. The highway should be
aesthetically satisfying. The design elements should strive to control pollution.
 The design should be so selected that not only the initial cost of construction of the facility,
but also the total transportation cost, including maintenance cost and road user cost should
be minimized.
 Safety should be inbuilt into the design elements.
 The design should enable all the road users (motor vehicles, cyclists, pedestrians and animal
drawn vehicles) to use the facility. The performance of the vehicles using the facility should
be given due consideration.
4.3 Highway Design Standards in Uganda
Some geometric standards in Uganda have been formulated by the Ministry of Works, Housing
and Communications e.g. The Uganda Road Design Manual Vol.1-Geometric Design Manual
2005. The AASHTO Standards represent the American practice, whereas the Department of
Environment (UK) standards give the current British practice.
It is important for engineers to exercise judgement in the use of a given design standard to ensure
that they come up with an economical solution for a geometric design. Sometimes, more than one
design standard is used for the purposes of comparing one pavement design with another so that
the comparison guides the engineer in selecting the most economical option.
4.4 Division of Roads into Functional Class
The rural roads in Uganda are divided into the following 5 classes according to their major function
in the road networks.
Class A: International Trunk Roads
Roads that link International Important Centres; Connection between the national road system and
those of neighbouring countries; Major function is to provide mobility;
Class B: National Trunk Roads
Roads that link provincial capitals, main centres of population and nationally important centres.
Major function is to provide mobility;
Class C: Primary Roads
Roads linking provincially important centres to each other or to a higher class roads (urban/rural
centres). Linkage between districts local centres of population and development areas with higher
class road. Major function is to provide both mobility and access;
Class D: Secondary Roads
Roads linking locally important centres to each other, to a more important centre, or to a higher
class road (rural/market centres) and linkage between locally important traffic generators and their
rural hinterland. Major function is to provide both mobility and access;
Class E: Minor Roads
Any road link to minor centre (market/local centre) and all other motorable roads; Major function
is to provide access to land adjacent to the secondary road system;
Roads of the highest classes, A and B, have as their major function to provide mobility and have
longer trip lengths. They are required to provide a high level of service with a high design speed.
The roads of Classes C and D serve a dual function in accommodating shorter trips and feeding
the higher classes or road. For these roads an intermediate design speed and level of service is
required. Road Class E has short trip length and their primary function is to provide access.
Design speeds and level of service for these roads may be low [MoWH&C, 2005].
4.5 Design Controls and Criteria
There are certain basic design controls and criteria which govern the geometric features of a
highway include the following:
 Topography,
  Traffic (its volume, directional distribution, and composition, including the future
estimates),
 Speed,
 Capacity design vehicle and control of access.
(i) Topography
Topography and physical features play an important role in the location and design of a highway.
The various design elements should be related to topographical features if an economical and
sound judgement is to emerge. The classification of terrain is normally done by means of the cross
(transverse) slope of the country, i.e. the slope approximately perpendicular to the centre-line of
the highway location.
(ii) Traffic
a) Importance of traffic data in Geometric Design
Traffic data crucial importance in highway design– both current and future estimates.
Traffic volume indicates the level of service for which the highway is being planned and directly
affects the geometric features such as width, alignment, grades etc. Without traffic data, it is futile
to design any highway.
b) Design Hour Volume (DHV)
The general unit for measuring traffic on a highway is the annual average daily traffic volume,
abbreviated as AADT. It is equal to the total annual volume of traffic divided by the number of
days in the year. Knowledge of traffic in terms of AADT is not of much use in geometric design,
since it does not represent the variations in traffic during various months of the year, days of the
week and hours of the day. A commonly used unit for geometric design is the 30th highest hourly
volume abbreviated as 30 HV. It is defined as the 30th highest hourly volume during the year.
Hence the design hourly volume (DHV) should be the
30 HV of the design (future) year chosen for design. Exceptions may be made on roads with high
seasonal fluctuation, where a different volume may need to be used [MoWH&C, 2005].
DHV is then expressed as DHV = AADT x K or ADT x K where K is estimated from the ratio of
the 30th HV to the AADT from a similar site. The 30th HV is the 30th highest hourly volume
during the year.
Table 4.1: The 30th HV is expressed as a fraction of ADT
Traffic Condition 30th HV as a fraction of ADT
Rural Arterial (average value) 0.15
Rural Arterial (maximum value) 0.25
Heavily trafficked road under Congested urban 0.08 – 0.12
conditions
Normal urban conditions 0.10 – 0.15
Road catering for recreational or Other traffic 0.20 – 0.30
of seasonal nature
Source: Uganda Road Design Manual, 2005
c) Directional Distribution of Traffic
For 2-lane highways, the design hour volume is the total traffic in both directions of travel. For
highways with more than 2-lanes, it is desirable to know the directional distribution of traffic.
Though this distribution has to be found from traffic surveys, a rough approximation can be to
assume 67% of total traffic to travel in one direction under the design conditions. The design has
to take into account both the morning and evening situations [Kadiyali, 2006].
d) Traffic Composition
Traffic composition has a vital effect on capacity and other design considerations. In Uganda, the
traffic is heterogeneous in character, consisting of fast driven cars, trucks, and buses. It is
customary in this country to express the traffic volume in terms of passenger car units (PCUs).
Table 4.2: Conversion Factor of Vehicle into Passenger Car

e) Future Traffic Estimates

The design of the geometric elements has to be prepared for the traffic likely to use the road in the
design year. The design period used for a flexible pavement generally varies from 15 to 25 years.
A period of 20 years is widely used as a basis for design. The future traffic estimates should be
computed to include normal, diverted and generated traffic.

(iii) Design Vehicles

A design vehicle is a selected motor vehicle, the weight, dimensions and operating characteristics
of which are used to establish highway design controls to accommodate vehicles of a designated
type. The dimensions and operating characteristics of a vehicle profoundly influence geometric
design aspects such as radii, width of pavements, parking geometrics, etc. The weight of the axles
and the weight of the vehicles affect the structural design of the pavement and structures, as also
the operating characteristics of vehicles on grades. Because of its crucial importance the
standardisation of the dimensions and the weights of design vehicles is the first step in formulating
geometric design standards. In Uganda, the Ministry of Works, Housing and Communications’
Uganda Road Design Manual Vol.1 - Geometric Design Manual 2005, is being followed
[MoWH&C, 2005].
a) Dimensions of Design Vehicles
The present vehicle fleet in Uganda includes a high number of four-wheel drive passenger/utility
vehicles, buses and overloaded trucks. Accordingly the five design vehicles indicated in Table will
be used in the control of geometric design until a major change in the vehicle fleet is observed and
detailed information on the different vehicle types using the roads in Uganda becomes available.
Table 4.3: Dimensions of Design Vehicles

b) Selection of the Design Vehicle

The selection of the design vehicle for the design of a highway is governed by the type and volume
of traffic that is expected to use the highway. For instance the design of a superior facility such as
a motorway or an expressway should be based on the largest design vehicle. The design of streets
and junctions primarily in residential areas can be done by using the passenger car design vehicle.
(iv) Speed as a Design Factor
The value of a highway is largely indicated by the speed, safety and convenience afforded by the
facility for travel. Speed is important for economic operation and has a great bearing on safety of
the highway. It plays a vital role in determining the geometric design of any given highway.
a) Design Speed
Design speed is the speed determined for design and correlation of the physical features of a
highway that influence vehicle operation. It is the maximum safe speed that can be maintained
over a specified section of a highway when conditions are so favourable that the design features
of the highway govern.
The design speed obviously has to be correlated with terrain conditions and the classification of
the highway. There is considerable variation in the speed adopted by different drivers and by
different types of vehicles. This raises the question of what value of speed should be adopted for
design. The value selected should accommodate nearly all demands with reasonable adequacy, yet
the design should not fail completely under severe or extreme load.
The speed adopted should satisfy nearly all drivers with exception of those few who drive at
extremely high speed [Kadiyali, 2006].
The standard design speeds are 50km/h, 60km/h, 70km/h, 85km/h, 100km/h and 120km/h. These
speed bands are based on the premise that for a given highway, it is considered acceptable if 85%
of the drivers travel at or below the designated design speed, generally inducing a situation where
approximately 99% of the drivers travel at or below one design speed category above the design
speed. Thus if a chosen design speed is by definition the 85th percentile speed for the highway,
then the next speed band up will constitute the 99th percentile speed. Speed bands are related to
each other as follows:
99𝑡ℎ 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 85𝑡ℎ 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 4
= 50𝑡ℎ 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 = √2 …4.1
85𝑡ℎ 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒
The 85th percentile speed is selected as the design speed on the basis that it constitutes the most
appropriate choice. Use of the 99th percentile speed would be safer but extremely expensive while
use of the 50th percentile speed would be unduly unsafe for faster travelling vehicles [Rogers,
2003]. The curve depicting the cumulative distribution of speeds has a typical “S” shape. It is
important to note that designers use typical data previously obtained on similar roads.
4.6 Control of Access
Uncontrolled access to road side development along whose major function is to provide mobility
will result in an increased accident hazard, reduced capacity and early obsolescence of the roads.
In order to preserve major roads as high standard traffic facilities it is necessary to exercise access
control, whereby the right of owners or occupants of land to access is controlled by the Road
Authority. Although control of access is one of the most important means for preserving the
efficiency and road safety of major roads, roads without access control are equally essential as
land service facilities. The following three levels of access control are applicable:
1) Full access control: - means that the authority to control access is exercised to give
preference to through traffic by providing access connections with selected public roads
only and by prohibiting direct access connections.
2) Partial access control:- means that the authority to control access is exercised to give
preference to through traffic to a degree in that, in addition to access connections with
selected public roads, there may be (some) private access connections.
3) Unrestricted access: - means that preference is given to local traffic, with the road serving
the adjoining areas through direct access connection. However, the detailed location and
layout of the accesses should be subject to approval by the Road Authority in order to
ensure adequate standards of visibility, surfacing, drainage, etc.
Road function determines the level of access control needed. Roads of higher classes have their
major function to provide mobility, while the function of lower classes is to provide access.
Motorways should always have full control of access. For all purpose roads the following general
guidelines are given for the level of access control in relation to the functional road classification:
Table 4.4: Level of Access Control

4.7 Sight Distance

4.7.1 Definition
Sight distance is defined as the length of carriageway that a driver can see in both the horizontal
and vertical planes.
The design of a highway with adequate sight ahead of a travelling vehicle results in safe operation.
Knowledge of the sight distance requirements is needed in designing vertical curves. It also
governs the set-back of buildings, slopes, fence, and other obstructions adjacent to a carriageway
on a horizontal curve [Kadiyali, 2006].
Types of sight distance
There are two types of sight distance:
(i) Stopping sight distance and
(ii) Overtaking sight distance.
4.7.2 Stopping Sight distance, SSD
This is defined as the minimum sight distance required by the driver in order to be able to stop the
car before it hits an object on the highway. It is primary importance to the safe working of a
highway. Because of its importance to safety, all highways must be designed for the minimum
stopping sight distance.
It is made up of two components:
(i) The distance travelled during perception and brake-reaction time; and
(ii) The distance travelled during the time the brakes are under application till the vehicle
comes to a stop.
When sensations received through the eyes, ears or body are strong enough to be recognized and
interpreted, they become perceptions.
In the cases of a motorist, it is the time which elapses between the instant the driver perceives the
object on the carriageway and the instant that he realises that braking is needed. The time lag or
the brief interval between the perception of danger and the effective application of the brakes is
called the brake-reaction time.
The perception time and the brake reaction time depend upon a variety of factors, e.g. age, gender,
alertness and visual acuity of the driver, visibility, vehicle design, the size and type of the object,
among others.
According to Ugandan practice, a perception reaction time of 2.5s, eye height of 1.07m above the
road surface and an object height of 0.15m are used in computing stopping sight distance.
The distance travelled during this interval, d1 is given by:
𝑑1 = 𝑣𝑡 = 0.278𝑉𝑡 ….4.2

The braking distance is the distance within which a moving vehicle comes to a stop after the
application of the brakes. On a level road, the distance is given by;
𝑉2
𝑑1 = 254𝑓 ….4.3

The coefficient of friction (for a wet pavement condition) is assumed to vary from 0.40 at 30 km/h
to 0.28 at 120 km/h. The above considerations yield the values in Table 4.5.
Table 4.5: Stopping Sight Distance on Level Ground for Wet Pavement Condition
Some slight adjustments are needed in the values of the braking sight distance to take into account
the effect of grades. The following amended formula may be used to calculate d1.
𝑉2
𝑑2 = 254(𝑓±𝐺) ….4.4

The positive sign is used when the gradient is upgrade and the negative sign may be used if the
gradient is downgrade. Correction for grade should not be applied on undivided roads with two-
way traffic but must invariably be considered for divided highways which have independently
designed profiles.
The safe stopping sight distance, SSD is given by d1 + d2.
𝑉2
𝑆𝑆𝐷 = 0.278𝑉𝑡 + 254(𝑓±𝐺) ….4.5

Example:
Calculate the safe stopping distance of a vehicle travelling at a speed of 80kph on an upward
gradient of 2%. Make suitable assumptions.
4.7.3 Full Overtaking Sight Distance, FOSD
Overtaking sight distance is that distance which should be available to enable the driver to overtake
another vehicle safely and comfortably without interfering with the speed of an oncoming vehicle
travelling at the design speed should it come into view after the overtaking maneuver is started
[Kadiyali, 2006].
Overtaking sight distance is of central importance to the efficient working of a given section of
highway. Overtaking sight distance only applies to single carriageways. There is no full overtaking
sight distance (FOSD) for a highway with a design speed of 120km/h since this speed is not
suitable for a single carriageway road.
Full overtaking sight distances are much larger in value than stopping sight distances. Therefore,
economic realities dictate that they can only be complied with in relatively flat terrain where
alignments, both vertical and horizontal, allow the design of a relatively straight and level highway
[Rogers, 2003].
Full overtaking sight distance is measured from vehicle to vehicle (the hazard or object in this case
another car) between points 1.05m and 2.00m above the centre of the carriageway. FOSD is made
up of three components: d1, d2 and d3 as described below:
 d1 = distance travelled by the vehicle in question while driver in the overtaking vehicle
completes the passing maneuver (Overtaking Time);
 d2 = distance between the overtaking and opposing vehicles at the point in time at which
the overtaking vehicle returns to its designated lane (Safety Time);
 d3 = distance travelled by the opposing vehicle within the above mentioned ‘Perception –
reaction’ and overtaking times (Closing Time).
In order to establish the values for full overtaking sight distance, it is assumed that the driver
making the overtaking maneuver commences it at two design speed steps below the designated
design speed of the section of highway in question. The overtaking vehicle then accelerates to the
designated design speed. During this time frame, the approaching vehicle is assumed to travel
towards the overtaking vehicle at the designated design speed.
The safety time, d2 is assumed to be 20% of d3. These assumptions yield the following equation:
𝐹𝑂𝑆𝐷 = 𝑑1 + 𝑑2 + 𝑑3 ….4.6
 𝑡 𝑣
𝐹𝑂𝑆𝐷 = 2 [ 4 4 + 𝑣] + 0.2𝑣𝑡 + 𝑣𝑡 = 2.05𝑣𝑡 = 0.5𝑉𝑡 ….4.7
√2 √2

The value of ‘t’ is generally taken as 10 seconds, as it has been established that it is less than this
figure in 85% of observed cases [Rogers, 2003].

Figure 4.1: Stopping and Passing Sight Distances on a crest curve

Source: Uganda Road Design manual Vol.1, 2005
4.7.4 Sight Distance for Multi-Lane Roads
Divided highways with 4 or more lanes need only to be designed for safe stopping sight distance.
Undivided highways with 4 lanes have enough opportunities for overtaking within one half of the
carriageway. Such roads therefore need only be designed for safe stopping sight distance.
4.7.5 Set-back Distance at Obstructions of Horizontal Curves
On horizontal curves with obstructions on the inside, an important consideration is the lateral
clearance so as to obtain the sight distance. It should be noted that:
 Sight distance is measured along the arc of the curve;
 If the pavement has two or more lanes, sight distance is measured along the arc at the centre
of the inner lane.
The presence of obstructions adjacent to the highway such as boundary walls, buildings, slopes of
cuttings may constrain the limiting radius of the horizontal alignment. To provide the necessary
horizontal sight distances, it may be necessary to set back obstructions.
In cases where the obstructions are immovable, it may be necessary to redesign the road alignment
in order to meet the safety requirements. It is therefore necessary to estimate the offset clearance
necessary to secure the required horizontal distance by considering two cases as in the following
sections.
a) Required Sight distance, S lies wholly within the length of the curve, L (S ≤ L)
Figure 4.2: Sight Distance Requirements on a horizontal curve with S ≤ L
Source: Rogers, 2003
The offset M can be approximated by considering the vehicle truck to be along the chords AC and
CB.
When the radius of horizontal curvature is large, then it can be assumed that the required sight
distance, S, approximates to a straight line. When S lies within the curve length, the minimum
offset M from the centreline to the obstruction can be estimated by considering the triangle
OAM and ACD. Thus:

b) Required Sight distance, S lies outside the length of the curve, L (S > L)
S is greater than the available length of the curve L and overlaps on the tangents for a distance l on
each side.
Figure 4.3: Sight Distance Requirements on a horizontal curve with S > L
Source: Rogers, 2003
Assuming a large horizontal radius of curvature and considering triangles ACP and OAP,

Example
A 2-lane 7.3 m single carriageway road has a horizontal curve of radius of 600 m. If the minimum
sight stopping distance required is 160 m, calculate in metres the required distance to be kept clear
of obstructions if the length of the curve is:
(a) 200 m;
(b) 100 m.
Solution
From the question, S = 160 m, R = 600 m.
(a) The length of the curve 200 m > 160 m. So the required sight distance S lies wholly within the
length of the curve. Applying equation (4.8), the required offset
(b) The length of the curve 100 m < 160 m. So the required sight distance S lies outside the length
of the curve. Applying equation (4.9), the required offset

4.8 Horizontal Alignment

Horizontal alignment deals with the design of the directional transition of the highway in a
horizontal plane. A horizontal alignment consists, a horizontal arc and two transition curves
forming a curve which joins two straights. In some cases the transition curve may have zero length.
The design procedure itself must commence with fixing the position of the two straight lines which
the curve will join together. The basic parameter relating these two straight lines is the intersecting
angles.
Minimum permitted horizontal radii depend on the design speed and the super-elevation of the
carriageway, which has a maximum allowable value of 7% in the UK, with designs in most cases
using a value of 5%.
4.8.1 Basic Formula for Movement of Vehicles on Curves
When a vehicle is moving on a curved path, it is subjected to an outward force, commonly known
as the centrifugal force. In order to resist this force, it is the usual practice to superior-elevate the
roadway cross-section. Figure 4.4 shows the forces acting on the vehicle at a super-elevated
section.

Figure 4.4: Forces acting on a vehicle on a horizontal curve

Source: Kadiyali, 2006
Let; M = mass of the vehicle;
v = speed of the vehicle in m/s;
V = speed in km/h;
R = radius of the curve in metres;
g = acceleration due to gravity (=9.81m/s2);
 N = normal force;
μ = coefficient of lateral friction;
α = angle of super-elevation;
e = rate of super-elevation, normally given as a percentage (= tanα)
C = centrifugal force.
The centrifugal force acting on the vehicle,

(i)
For equilibrium, resolving forces parallel to the incline plane gives,

(ii)
Resolving forces perpendicular to the incline plane gives,

(iii)
Substituting equation (iii) into equation (ii) gives,

Dividing through the above equation by Mg cos α gives,

Equation 4.10 above is the basic equation relating the speed of vehicles, the radius of the curve,
the super-elevation and coefficient of friction. This equation forms the basis of design of horizontal
curves.
Equation 4.10 can be rewritten as seen in Equation 4.11, and is known as the minimum radius
equation:
If the entire centrifugal force is counteracted by super-elevation, then frictional force will not come
into play. In this case, μ = 0 in equation 4.10. The super-elevation then provided is said to be
‘equilibrium super-elevation’. In such a case, the pressures on the inner and outer wheels would
be equal.
Design practice is based on the assumption that at absolute minimum radius the 99th percentile
speed vehicle should not experience more than the maximum level of centrifugal acceleration
acceptable for comfort. Its value is 0.22g.
𝑣2
Thus if g, = 0.22𝑔, then the total centrifugal acceleration at the design speed (85th percentile
𝑅
speed) should not exceed:

4.8.2 Value of the Coefficient of Lateral Friction, μ

The value of the coefficient of lateral friction depends upon a number of factors, chief among them
being the vehicle speed, type and condition of roadway surface, and type of and condition of the
tyres. AASHTO recommends the values given in Table 4.6. However, constant value of 0.15 is
generally recommended.
Table 4.6: Coefficient of Lateral Friction as Recommended by AASHTO

4.8.3 Maximum super-elevation Value, emax

If equation 4.10 is to be used for design, it is desirable to know the maximum super-elevation that
can be permitted. Practice in this regard varies from country to country. The AASHTO practice
limits it to 0.12 (12%), whereas the UK practice limits it to 0.07 (7%). In Uganda the value is
limited to 0.08 (8%) [MoWH&C, 2005].
4.8.4 Super-elevation Rates
Super-elevation on curves is intended to counteract a part of the centrifugal force, the remaining
part being resisted by the lateral friction. Also, super-elevation results in economies in
maintenance. This is because skidding and unequal pressures on the wheels of vehicles, which
result from high value of sideway force between the tyres and the roadway surface, necessitate
frequent attention to the surface.
Super-elevation can be provided either to fully counteract the centrifugal force or to counteract a
fixed proportion of the centrifugal force. In the former case, the super-elevation needed would be
more than 1 in 15 (6.67%) on sharp curves causing inconvenience to slow moving vehicles.
Since super-elevation has to be limited to 7% or 8% as per Ugandan practice, maximum friction
would have to be relied upon when the sharpest possible curve is traversed. When a vehicle
negotiates a flat curve, friction would not be developed to the maximum and this would not be a
balanced design. It is desirable that the super-elevation should be such that a moderate amount of
friction is developed while negotiating flat curves and friction not exceeding the maximum
allowable value should be developed at sharp curves. Therefore designing the super-elevations to
fully counteract the centrifugal force developed at a fraction of the design speed will provide the
necessary balance.
The above is achieved as per UK practice by providing full super-elevation for a speed of 67.082%
of the design speed such that 45% of the centrifugal force is balanced by super-elevation while
55% of the centrifugal force is balanced by friction. Therefore Equation 4.10 becomes:

The super-elevation computed from Equation 4.13 is restricted to a value of 7% (0.07) or 8%

(0.08) as per Ugandan practice.
4.6.5 Radii of curves for which no super-elevation is required
The normal cambered section of a highway can itself be continued on a curve where the super-
elevation calculated is less than the camber. From Equation 4.13

Substituting the values of camber for e in Equation 4.14 above, the minimum radius beyond which
no super-elevation is required is obtained.
In such cases where the radius is greater than those given by the above formula it is desirable to
remove the adverse crown in the outer half of the carriageway and super-elevate at the normal
crown slope.
5.8.6 Method of Attainment of Super-elevation
The normal cambered surface on a straight reach of road is changed into a super-elevation surface
in two stages.
First stage: the outer half of the camber is gradually raised until it is level as seen in Figure 4.5.
It is desirable to accomplish the raising of the outer-half till it is level before the starting point of
the transition curve. The raising of the outer edge should be done in a slope not exceeding 1 in 150
for plain and rolling terrain and 1 in 60 for hilly terrain.
Figure 4.5: Stages involved in attainment of super-elevation
Second stage: any of the three methods given below may be adopted to attain the full super-
elevation:
a) The surface of the road is rotated about the centre-line of the carriageway, gradually lowering
the inner edge and raising the outer edge while keeping the level of the centerline constant (Figure
4.5 d);
b) The surface of the road is rotated about the inner edge, raising the centre and the outer edge
(Figure 4.5 e);
c) The surface of the road is rotated about the outer edge depressing the centre and the inner edge
(Figure 4.5 f);
In most circumstances method (a) is generally used results in the least distortion of the pavement.
Figure 4.6 below shows the method of attaining super-elevation using method (a).
Figure 4.6: Attaining Super-elevation by revolving about the centre line
Example
Calculate the super-elevation to be provided for a horizontal curve with a radius of 400m for a
design speed of 100kph in plain terrain. Comment on the results. What is the coefficient of lateral
friction mobilised if super-elevation is to be restricted to 7%.
Solution
1.0 Data Summary
a) Curve Radius, R = 400m
b) Design Speed, V = 100kph
c) Maximum super-elevation, e = 7%
2.0 Maximum Elevation, emax
According to the UK practice, the super-elevation is calculated on the assumption that it should
45% of the centrifugal force developed at 67.082% of the design speed.
Therefore from equation 4.13

3.0 Comment on the Result

Since, as per UK practice, the maximum super-elevation allowable is 7%, then the computed
super-elevation is too high and should be restricted to 7%. The balance of the centrifugal force will
be taken care of by the friction which is mobilised. If μ is the coefficient of friction, then from
equation 4.10:

This is less than the recommended value of 0.15

4.8.7 Transition Curves
A transition curve is a curve in which the radius changes continuously along its length and is used
for the purpose of connecting a straight with a circular curve, or two circular curves of different
radii.
a) Need for Transition
When a vehicle travelling on a straight course (i.e. R=∞) enters a curve of infinite radius, it
suddenly subjected to the centrifugal force which causes shock and sway. In order to avoid this, it
is customary to provide a transition curve at the beginning of the circular curve, having a radius
equal to infinity at the end of the straight and gradually reducing the radius of the circular curve
where the curve begins. Incidentally, the transition portion is also used for gradual application of
the super-elevation, curve widening and improvement of the general appearance.
The transition curve is also used to achieve the following:
  They reduce the tendency of vehicular skidding;
 They minimize passenger discomfort;
 They provide convenient sections over which super-elevation or pavement widening may
be applied;
 They improve the appearance of the road by avoiding sharp discontinuities in alignment at
the end and beginning of circular curves.
b) The Spiral
Various forms of curves are suitable for highway transitions, but the most popular and
recommended for use in this country is the spiral. It is easy to set out in the field and the rate of
acceleration is uniform through the length of transition. In Figure 4.7, the main elements of a
circular curve provided with spirals for transition at its two ends.

Figure 4.7: Main Elements of a Circular Curve Provided with Transitions

The following nomenclature applies,
φmax = Spiral angle
θ = Deflection angle
T = Tangent length
R = Radius of circular curve
S = Shift
L = Length of Spiral (or transition curve)
I = Point of intersection
T = Beginning of spiral
T1 = Beginning of circular curve
T2 = End of circular curve
U = End of spiral
Some of the important properties of a spiral are given below:

c) Length of Transition
The length of the transition should be determined from the following two considerations:
(i) The rate of change of centrifugal acceleration adopted in the design should not cause
discomfort to the drivers. If C is the rate of change of acceleration then:

Where;
aT1 = radial acceleration at T1 (= v2/R)
aT = radial acceleration at T (= 0)
t = time taken (= L/v)
Substituting the above in equation 4.15 gives

Where v is speed in m/s and V is speed in Km/hr. The value of C is usually taken as 0.3m/s3.
(ii) The rate of change of change of super-elevation should also be such as not to cause higher
gradients and unsightly appearances.
This could be kept in 1 in 150 for roads in plain and rolling terrain and 1 in 60 for roads in hilly
terrain. Since the super-elevation can be given by rotating about the centreline, inner or outer edge,
the length of the transition will be governed accordingly. In calculating the length of transition,
the pavement width should include any widening that may have been provided at the curve.
The higher of the values given by the above two methods should be adopted.
4.8.8 Curve Widening
Widening of pavements is needed on curves for the following reasons:
a) On curves, the vehicles occupy a greater width because the rear wheels track inside the front
wheels (See Figure 4.8)
b) On curves, drivers have difficulty in steering their vehicles to keep to centre line of the lane.
c) Drivers have psychological shyness to drive close to the edges of the pavement on curves.
From Figure 4.8, considering the triangle OCB, right angled at B,

Figure 4.8: Widening on Curves

Assuming a wheel base of 6m for a vehicle corresponding to AASHTO single unit, widening in
metres, m is given by

Where; R = radius in metres

The widening due to psychological reasons is a function of speed and can be assumed to be given
by the empirical formula, Wp;

Where; V = speed in kph; R = radius in metres

The total widening for pavements, We is given by the following formula:

Where; n = number of lanes

Example
A two-lane (7.0m wide) pavement on a National highway in hilly terrain has a curve radius of
250m. The design speed is 80kph, maximum super-elevation is 7%, camber is 2.5%, the relative
longitudinal gradient is 0.5% (1 in 200), the angle of deviation is 60o and the rate of change of
radial acceleration is 0.3m/s3. Determine the following assuming that; a) the curve will need to
bet widened if the curve radius is less than 300m, b) the super-elevation is obtained by rotation
about the centre line, and c) the design vehicle is a DV-2 single unit truck with a wheel base of
6.1m.
i) The length of transition curve;
ii) The tangent length
iii) The total length of the curve.
Solution
1.0 Data Summary
o Pavement width, W = 7.0m
o Curve Radius, R = 250m
o Design Speed, V = 80kph
o Maximum super-elevation, emax = 7%
o Camber, eo = 2.5%
o Relative longitudinal gradient, S = 0.5%
o Angle of deviation, θ = 60o
o Rate if change of radial accn., C = 0.3m/s3
2.0 Sketch drawing
Refer to Figure 4.6
3.0 Transition Length, L
3.1 Based on the rate of change of centrifugal acceleration, La
From equation 4.21, the transition length, L, required for safety and comfort is given by;

3.2 Based on the rate of change of super-elevation, Ls

Since super-elevation is obtained by rotation about the centre line then the transition length,
L, will be given by;

Where; W=pavement width, e = super-elevation, S=longitudinal gradient, and eo =camber

Since radius, R < 300m, extra widening, We, of the carriageway is required.
Therefore the pavement width will be
4.8.9 General Controls for Horizontal Alignment
The following general controls for horizontal alignment should be kept in view in a sound design
practice:
(i) The alignment should be as directional as possible;
(ii) The alignment should be consistent with topography and should generally conform to the
natural contours. A line cutting across the contours involves high fills and deep cuts, mars
the landscape and is difficult for maintenance;
(iii)The number of curves should, in general, be kept to a minimum;
(iv) The alignment should avoid abrupt turns. Winding alignment consisting of short curves
should be avoided, since it is the cause of erratic vehicle operation;
(v) A sharp curve at the end of along tangent is extremely hazardous and should be avoided.
If sharp curvature is unavoidable over a portion of the route selected, it is preferable that
this portion of the road be preceded by successive sharper curves. Proper signage, well in
advance of a sharp horizontal curve is essential;
(vi) Short curves giving the appearance of kinks should be avoided, especially for small
deflection angles. The curves should be sufficiently long to provide a pleasing appearance and
smooth driving on important highways. They should be at least 150m long for a deflection
angle of 5 degrees, and the minimum length should be increased by 30m for each 1 degree
decrease in the deflection angle;
(vii) For a particular design speed, as large a radius as possible should be adopted. The
minimum radii should be reserved only for the critical locations;
(viii) The use of sharp curves should be avoided on high fills. In the absence of cut slopes,
shrubs, trees, etc., above the roadway, the drivers may have difficulty in estimating the extent
of curvature and fail to adjust to the conditions;
(ix) While abrupt reversals in curvature are to be avoided, the use of reverse curves becomes
unavoidable in hilly terrain. When they are provided, adequately long transitional curves
should be inserted for super-elevation run-off;
 (x) Curves in the same direction separated by short tangents, say 300m -500m long, are called
broken-back curves. They should be avoided as they are not pleasing in appearance and are
hazardous;
(xi) Compound curves may be used in difficult topography in preference to a broken-back
arrangement, but they should be used only if it is impossible to fit in a single circular curve.
To ensure safe and smooth transition from curve to curve, the radius of the flatter curve should
not be disproportional to the radius of the sharper curve. A ration of 2:1 or preferably 1.5:1
should be adopted;
(xii) The horizontal alignment should blend with the vertical harmoniously. General
controls for the combination of horizontal and vertical alignments should be followed
[Kadiyali, 2006].

4.9 Vertical Alignment

Vertical alignment design refers to the arrangement of tangents and curves which compose the
profile of the road. It is composed of a series of straight-line gradients connected by curves,
normally parabolic in form. The main aim of vertical alignment is to ensure that a continuously
unfolding stretch of the road is presented to motorists so that their anticipation of directional
change and future action is instantaneous and correct [Rogers, 2003; O’Flaherty, 2002].
4.9.1 Major Requirements of Vertical Curves
The two main requirements in the design and construction of vertical curves are:
(i) Provision of adequate visibility, and
(ii) Provision of passenger comfort and safety.
In order to provide adequate visibility, oncoming vehicles or any obstructions in the road must be
seen clearly and in good time to ensure that vehicles travelling at the design speed can stop or
overtake safely. This requirement is achieved by use of sight distances and K-Values.
In order to provide passenger comfort, the effect of the radial force on the vehicle traversing a
vertical curve must be minimized. In crest curve design this effect could cause the vehicle to leave
the road surface (e.g. in hump-back bridges) while in the sag curve the underside of the vehicle
would come into contact with the surface, particularly where the gradients are steep and opposed.
The result is discomfort and danger to passengers travelling can be minimised by:
(i) Restricting the gradients; which has the effect of reducing the radial force;
(ii) Choosing a suitable type and length of curve such that this reduced force is introduced
gradually and uniformly as possible [Uren et al, 1989].
4.9.2 Gradients
The rate of rise or fall of road surface along its length with respect to horizontal distance is termed
as gradient [MoWH&C, 2004].
The use of steep gradients in hilly terrain generally results in lower road construction and
environmental costs. However, it also adds to road user costs through delays, extra fuel costs and
accidents. Gradients of up to about 7% have little effect on the speeds of passenger cars.
Nevertheless, the speeds of commercial vehicles are considerably reduced on long hills with
gradients in excess of 2%. For short distances, gradients of 5% or 6% may have little detrimental
effect on commercial vehicle speeds [O’Flaherty, 2002]. Long, steep, downhill grades are very
dangerous and need careful design, preferably with escape roads (side roads that are designed to
bring out-of-control vehicles to a safe stop) [MoWH&C, 2004]. The Uganda Road Design manual
(2004), suggests maximum gradients as presented in Table 2.16 below:
Table 4.7: Maximum Grades as recommended by MoWH&C

According to British Standards of road design, a minimum longitudinal gradient of 0.5% is needed
to ensure effective drainage of carriageways with kerbs.
4.9.3 Climbing Lanes
The limitation of gradients to a maximum value is not in itself a complete design control, and
therefore an additional climbing lane is normally provided on long uphill climbs. The provision of
a climbing lane is normally considered when the combination of hill severity and traffic volumes
and composition is such that the operational benefits achieved are greater than the additional costs
of constructing an additional lane.
In Uganda, however, climbing lanes are recommended for use if the design truck speed decreases
more than 20 km/h under the truck speed limit, normally 80 km/h in rural conditions. A climbing
lane is inserted into the carriageway by means of entry and exit tapers to the left of the continuous
lane so that slow moving vehicles have to merge into the faster traffic at the termination point as
shown below.

Figure 4.9: Climbing Lane outside the ordinary lane

Source: Uganda Road Design Manual (2004)
4.9.4 Cross falls
A minimum cross fall of 2.5% is normally recommended in the form of either a straight camber
extending from one edge to the other or as one sloped from the centre of the carriageway towards
both edges. The primary aim of these cross falls is to adequately get rid of surface runoff from the
highway pavement.

Figure 4.10: Highway Cross falls

Source: Rogers, 2003
4.9.5 Vertical Curves
A vertical curve provides a smooth transition between successive gradients in the road profile.
When the algebraic difference in gradients, A is positive the curve is called a crest or summit curve
whereas if it is negative the curve is called a sag or valley curve.

Figure 4.11: Typical Vertical Curves

Source: O’Flaherty, 2002
a) Shape of the Curve
Where the ratio of length of curve to radius is less than 1-10, there is no practical difference
between the shapes of a circle, a parabola and an ellipse. Owing to the fact that this condition can
be shown to apply in most of the cases met in practice, a parabolic form of vertical curve is
therefore used to guide vertical curve design [Bannister et al, 1998].
Figure 4.12: Typical Vertical Curves
b) Equation of a Vertical Curve
According to Thagesen (1996), a simple parabola is recommended when modelling vertical curves.
The parabola provides a constant rate of change of curvature, and hence visibility, along its length.
The vertical curve is of the form:

Where; A is the algebraic difference in grade (m - n)

Substituting for k in equation (ii), we get
c) Sight Distances
The length of curve to be used in any given situation depends on the sight distance. It is the distance
of visibility from one side of the curve to the other [Uren et al, 1989].
There are two categories of sight distance namely:
(i) Stopping sight distance (SSD): which is the theoretical forward sight distance required by
a driver in order to stop safely and comfortably when faced with an unexpected hazard on
the carriageway,
(ii) Full overtaking sight distance (FOSD); which is the length of visibility required by
motorists to enable them to safely and comfortably overtake vehicles ahead of them.
When designing vertical curves, it is important to know whether safe overtaking is to be included
in the design. If it is to be included, then the FOSD must be incorporated in the design and if it is
not then SSD must be incorporated. On single carriageways, it is usually necessary to consider
whether to design for overtaking only at crest curves since overtaking is not a problem on dual
carriageways and visibility is usually more than adequate on single carriageways [Uren et al,
1989].

Figure 4.13: Sight distance over crest curves when a) S ≤ L and b) when S > L
Source: O’Flaherty (2002)
d) K-Values
In the past it was necessary to use the appropriate sight distance for the road type and design speed
in question to calculate the minimum length of the vertical curve required. Nowadays, however,
constants which greatly simplify calculations have been provided by the MoWH&C [Uren et al,
1989].
The minimum length of vertical curve Lmin for any given road is obtained from the formula.

Where;
K = constant obtained from MoWH&C standards (K = R/100)
R = radius of curvature of the curve (in meters)
A = algebraic difference in grade (%)
There are three categories of K-Values for crest curves (SSD and FOSD crest curve values
obtained from Table 4.8) and one category of K-Values for sags obtained from Table 4.9. The
K-Values obtained are derived from the sight distances as already discussed [MoWH&C, 2004].
Table 4.8: Minimum Radii for Crest Curves as Recommended by MoWH&C

Table 4.9: Minimum Radii for Sag Curves as recommended by MoWH&C

4.9.6 Vertical Crest Curve Design and Sight Distance Requirements

In calculating the minimum lengths of crest curves, two design conditions have to be considered
namely:
(i) Where the sight distance is contained within the length of the vertical curve i.e. the sight
distance is less than the length of curve.
(ii) Where sight distance overlaps onto the tangent sections on either side of the vertical curve.
In this case the sight distance extends beyond the vertical curve [TRL, 1993].
Considering the properties of the parabola:
For S ≤ L;

Where;
Lmin = minimum length of vertical crest curve (m)
S = required sight distance (m)
A = Algebraic difference in gradients
h1 = Driver eye height (m); taken as 1.05m
h2 = Object height (m); taken as 0.26m
For full overtaking sight distance, FOSD, h1 = h2 = 1.05m. The decision to which equation should
be used at a given site can be made by solving either of the equations below;

If e > h1 then equation (4.29) is used and when e < h1, equation (4.30) is used [O’Flaherty, 2002].
4.9.7 Vertical Sag Curve Design and Sight Distance Requirements
a) Based on clearance from structures during day time
In certain situations when a road passes beneath structure such as an over pass or bridge on sag
curves, the primary design criterion for designing the sag curve may be the provision of necessary
clearance in order to maintain the drivers’ line of sight [Rogers, 2003]. The minimum length of
sag curve which meets minimum stopping sight distance requirements is given by;
When SSD ≤ L

Where;
h1 =` Drivers eye height (Usually 1.05m)
h2 = Object height (usually 0.26m)
L = minimum length of sag curve (m)
A = algebraic difference in grades expressed as a decimal.
D = vertical clearance (ideally taken as 5.7m) to the critical edge of the bridge
The critical edge is assumed to be directly over the point of intersection of tangents.
In practice both equations can be considered valid provided that the critical edge is not more than
60m from the point of intersection [O’Flaherty, 2002].
b) Based on night time Conditions
During night time conditions, a critical concern in design of sag curves can be the headlight sight
distance, where the length of the highway illuminated by the cars headlight is the governing
parameter [Rogers, 2003]. The minimum length of sag curve is thus given by:
For SSD ≤ L

Where;
h3 = headlight height (usually 0.6m above the carriageway),
α = angle of upward divergence of light beam (usually 1.0o), and
L, A, and S are as defined previously.
It should however be noted that the above equations (based on night time conditions) are;
 Very sensitive to the assumption of a 1degree upward divergence of the light beam;
 They erroneously assume that headlights can illuminate an object on the carriageway at
long distances and they ignore the fact that many vehicles are driven on dipped lights; and
 The effect of headlamps is reduced on horizontal curves [O’Flaherty, 2002].
c) Based on Motorist Comfort
The minimum length of vertical sag curve is given by:

Where;
V = design speed (km/hr),
A is the algebraic difference in grade (%), and
a = vertical radial acceleration (m/s2) usually taken as 0.3 m/s2 for comfortable design
d) Design Speed and Speed Limit
Design speed is a measure of road quality. It may be defined as the maximum safe speed that can
be maintained at a given section of the road where conditions are so favourable that the design
features of the road govern vehicular movements [MoWH&C, 2004].
The selection of design speeds for road sections of a particular classification is primarily
influenced by;
 Nature of terrain:- whether level, rolling or mountainous; and
 Motorist expectations: - in relation to free speed at which it is safe to drive (in rural areas)
or legal to drive (in urban areas) [O’Flaherty, 2002].
Speed limit on the other hand is the maximum allowable speed on a road.
The normal speed limit on rural roads in Uganda is 80km/hr and that in trading centres, towns and
cities is 50km/hr. Speed limits may be reduce but not increased by local speed limits shown on
regulatory traffic signs [MoWH&C, 2004].
In a nutshell, the design speed should not be lower than the speed limit and should be preferably
10km/hr higher than the speed limit. Short rural sections with design speeds lower than the speed
limit should be treated with warning signs and no overtaking markings [MoWH&C, 2004].
e) Length of Vertical Curve to be used
Normally the value for minimum length of curve obtained from the K-Value is not used. A greater
value is instead chosen. This may be due to the necessity to fit the curve into particular site
conditions and the necessity to fit the vertical alignment of the road to the horizontal alignment (a
process known as phasing of vertical and horizontal alignment) [Uren et al, 1989].
f) Phasing of the Vertical and Horizontal Alignment
Phasing is usually done when designing new roads or improving existing alignments and follows
the procedure below:
(i) Designing or redesigning the horizontal alignment;
(ii) Taking reduced levels at regular intervals along the proposed centreline and plotting a
longitudinal profile;
(iii)Superimposing chosen gradients on the longitudinal section, altering their percentage as
necessary to try to balance out any cut and fill in addition to trying to get the vertical tangent
points to coincide with those of the horizontal curve. It is this last point that often gives the
length of vertical curve in order to avoid the creation of optical illusions in the vertical
plane [Uren et al, 1989].
g) Setting-Out Data
In setting out a vertical curve on ground, the objective is to place large pegs at the required intervals
along the line of the proposed roadway and to nail a cross-piece to each peg at a certain height
(usually 1.0m), above the proposed road level. These pegs are called profiles and the erection of
these profiles is the standard method of setting out proposed levels on any construction site. The
following information is required for any setting out calculations; the length of the curve (which
is dependent on the gradient of the straights and site distance) and the gradients of slopes together
with one change point preferably a point of vertical intersection [Irvine, 1998].
4.9.8 General Controls for Vertical Curve Alignment
The following general controls for vertical alignment should be kept in view while designing the
vertical profile of a highway:
 The grade line selected should be smooth with gradual changes, consistent with the class of
highway and terrain. Numerous breaks and short lengths of grades should be avoided;
 The ‘roller-coaster’ or ‘hidden type’ of profile should be avoided as it is hazardous and
aesthetically unpleasant;
  Undulating grade line, involving substantial lengths of momentum grades, should be
appraised for their effect upon traffic operation. Such profiles permit heavy trucks to operate
at higher overall speeds than when an upgrade is not preceded by a down grade, but may
encourage excessive speeds of trucks with consequent hazard to traffic;
 A broken-back grade line (two vertical curves in the same direction separated by short
section of tangent grade) should generally be avoided;
 On long continuous grades, it may be preferable to place the steepest grades at the bottom
and flatten the grades near the top. Alternatively, long grades may be broken by short
intervals of flatter grades;
 Intersections on grades should be avoided as far as possible. Where unavoidable, the
approach gradients and the gradient through the intersections should be flattened to the
maximum possible extent.
Vertical Curve Examples
Question one
The elevation of an intersection of rising gradient of 1.5% and a falling gradient of 1.0% on a
proposed road is 93.600m AOD. Given that the K-Value for this particular road is 55, the through
stationing of the intersection point is 0 + 671.340 and the vertical curve is to have equal tangent
length. Calculate:
a) The through stationing of the tangent points of the vertical curve if the minimum required length
is to be used.
b) The elevations of the tangent points and the elevations at exact 20m multiples of through
stationing along the curve.
c) The position and level of the highest point on the curve.
Solution
1.0 Data Summary
a) Grades; Initial, m = 1.5%
Final, n = -1.0%
b) Point of Intersection
Stationing of PVI = 0 + 671.340
Elevation of PVI = 93.600m AOD
c) K-value = 55
Question Two
An equal tangent vertical curve is to be constructed between grades of -2.0% (initial) and +1.0%
(final). The PVI (Point of vertical intersection) is at station 11 + 000.000 and elevation 420.000m
AOD. Due to a street crossing, the elevation of the roadway at station 11 + 071.000 must be at
elevation 421.500m. Design the curve assuming it has a shape of the form; y = ax2 + bx + c.
Solution
1.0 Data Summary
a) Type of vertical curve: Equal tangent
b) Grades; Initial, m = -2.0%
Final, n = +1.0%
c) Point of Intersection, PVI
Stationing = 11+000.000
Elevation = 420.000m AOD
d) Point of Interest
Stationing = 11+710.000
Elevation = 421.500m AOD
2.0 Sketch Drawing
Exercise
A 150m long equal tangent vertical curve connecting grades of +1.2% (initial) and -1.08% (final)
crosses a one-meter diameter pipe at right angles. The pipe is located at station 11 + 025.000 and
its centerline is at elevation 1091.6m. The PVI of the vertical curve is at station 11 + 000.000 and
elevation 1095.2m. Using offsets determine the depth, below the surface of the curve, to the top
of the pipe and determine the station of the highest point on the curve.
4.10 Cross-Sectional Elements
4.10.1 General
The cross-sectional elements of a highway design pertain to those features which deal with its
width. They embrace aspects such as road reserve width, carriageway width, central reservation
(median), shoulders, camber, side slopes, horizontal and vertical clearances etc.
Figure 4.14: Single Carriageway Cross-section Elements
Source: Uganda Road Design Manual, 2004

Figure 4.15: Dual Carriageway Cross-section Elements

Source: Uganda Road Design Manual, 2004
4.10.2 Road Reserve
The road reserve or right-of-way width is the width of land secured and preserved in public interest
for road development purposes. The road reserve should be adequate to accommodate all the
elements that make up the cross-section of the highway and may reasonably provide for future
development.
In order to prevent ribbon development along highways, it is sometimes necessary to establish
‘control lines’ and ‘building lines’. A ‘control line’ is a line which represents the nearest limits of
future uncontrolled activity in relation to a road. This signifies that though building activity is not
totally banned between the building line and the control line, the nature of buildings permitted here
is controlled. A ‘building line’ on the other hand is a line on either side of the road between which
no building activity is permitted at all.
4.10.3 Carriageway Width
The term “carriageway” is used here to cover the traffic lanes, any auxiliary lanes, and the
shoulders [MoWH&C, 2004]. The width of traffic lanes governs the safety and convenience of
traffic and has a profound influence on the capacity of a road.
The factors that influence capacity of a carriageway are:
(i) The design volume, i.e. the greater the traffic volume the wider the carriageway and,
normally, the greater the number of lanes;
(ii) Vehicle dimensions, i.e. heavy commercial vehicles require wider carriageways to ensure
adequate clearances when passing each other;
(iii)The design speeds, i.e. vehicles travelling at high speed, especially commercial vehicles,
require wider carriageways to ensure safe clearances between passing vehicles;
(iv) The road classification, i.e. the higher the road classification the greater the level of service
(and width of carriageway) expected.
Internationally, it is generally accepted that lane widths should normally be at least 3.5m, although
narrower lanes are often used for economic or environmental reasons on both rural and urban
roads. However, increasing the lane width up to 3.65m on two lane two way rural roads decreases
accident rates [O’Flaherty, 2002].
4.10.4 Central Reservation (Median) Strip
A central reservation strip is the longitudinal space separating dual carriageways. The functions of
the median strip are:
(i) To separate the opposing streams of traffic;
(ii) To minimise head-light glare;
(iii)To include space for safe operation of crossing and turning vehicles at intersections at
grade;
(iv) To provide a stopping area in case of emergencies.
The central reservations on high-speed heavily trafficked rural roads in the United States are
typically 15m to 30m. In Europe they tend to be much narrower (say 4 – 10m) and to be used with
safety barriers. Those in Britain are normally 4.5m wide, and include a crash barrier. In urban areas
they can be as narrow as 1m, but 3m is preferred so that a crossing pedestrian pushing a pram or
wheelchair has space to wait in safety [O’Flaherty, 2002]. On severely restricted arterial streets,
where a narrow separator of 0.6 – 1.2m is feasible, it may be desirable to have few, if any, openings
in median except at intersections.
4.10.5 Shoulders
A shoulder is a portion of the roadway adjacent to the carriageway and is intended for
accommodation of stopped vehicles, emergency use and lateral support of base and surface
courses. The width of the shoulder should be adequate for giving working space around a stopped
vehicle. American practice recommends a 3m width for high type facility and a width of 1.2m -
2.4m for low type facilities. UK practice for rural roads recommends widths ranging from 1.2m to
3.65m depending upon the road type and nature of kerb treatment.
4.10.6 Laybys and bus bays
When economic considerations do not favour the construction of shoulders on rural roads, laybys
should be provided instead, at spacings that are appropriate to the traffic volume. Thus, for well
trafficked and lightly trafficked single carriageways, it is British practice to provide 2.5m and 3m
wide by 30m long laybys at 1.5km and 5.8km intervals, respectively, on either side of the
carriageway, while 3m wide by 100m long laybys are provided at approximately 1km intervals on
each side of dual carriageways. Laybys should be located at sites with good visibility and provided
with tapered hard-strips at either end to assist in the safe deceleration and acceleration of vehicles
using them.
Full bus bays (3.25m by at least 12m, plus 20m end tapers) may be provided at bus stops in urban
areas. However, the appropriateness of this provision is dependent on the traffic volumes on the
road in question.
4.10.7 Kerbs
A kerb (as termed as curb) is a vertical or sloping member along the edge of a pavement or
shoulder, forming part of gutter, strengthening or protecting the edge, and clearly defining the edge
to vehicle operators. Its functions are:
(i) To facilitate and control drainage;
(ii) To strengthen and protect the pavement edge;
(iii)To delineate the pavement edge;
(iv) To present a more finished appearance;
(v) To assist in the orderly development of the roadside.
Kerbs are classified as ‘barrier’ or ‘mountable’. Barrier kerbs are designed to discourage vehicles
from leaving the pavement. The face may be vertical or sloping and the height may range from
15cm to 25cm. Mountable kerbs are those which can be easily crossed by vehicles if required.
They are used at medians and channelizing islands.
4.10.8 Camber
Camber, also known as cross fall, facilitates drainage of the pavement laterally. The pavement can
have a crown or a high point in the middle with slopes downwards towards both edges. This is
favoured on two-lane roads and wider undivided roads. On divided roads, the individual
carriageways may be centrally crowned separately or a unidirectional slope may be provided
across the entire carriageway width. The amount of camber to be provided depends upon the
smoothness of the surface and the intensity of rainfall. In the UK, a value of 2.5% is generally
adopted for design. A cross fall for the shoulders should be generally steeper than for the pavement
by about 0.3 – 0.5% to facilitate quick drainage. The UK practice is to provide 5% slope on the
shoulder [Kadiyali, 2006].
4.10.9 Side slope
According to O’Flaherty (2002), soil mechanics analysis enables the accurate determination of
maximum slopes at which embankments or cuts can safely stand. However, these maximum values
are not always used, especially on low embankments not protected by safety fences. The slopes of
embankments and cut sections depend upon the type of soil and the height of embankment or depth
of cuttings. A flatter slope is conducive for erosion control, but is costly.
Flatter slopes of embankments promote safety of traffic. Ordinarily, 1.5:1 to 2:1 in mild slope
conditions and 2:1 to 3:1 in overwhelming slope conditions will be adequate.
4.11 Intersection Design and Capacity
4.11.1 General
An intersection is defined as the general area where two or more highways join or cross, within
which are included the roadway and roadside design features which facilitate orderly traffic
movements in that area. An intersection leg is that part of any one of the highways radiating from
an intersection which is outside of the area of the intersection.
The importance of intersection design stems from the fact that efficiency of operation, safety,
speed, cost of operation and capacity are directly governed by the design. Since an intersection
involves conflicts between traffic in different directions, its scientific design can control accidents
and delay and can lead to orderly movement of traffic. Intersections represent potentially
dangerous locations from the point of view of traffic safety. It is believed that well over half the
fatal and serious road accidents in built-up areas occur at junctions [Kadiyali, 2006].
The following principles should be considered in a good design:
(i) The number of intersections should be kept to a minimum. If necessary, some minor roads
may be connected with each other before joining a major road;
(ii) The geometric layout should be so selected that hazardous movements by drivers are
eliminated. This can be achieved by various techniques such as channelizing and
staggering;
(iii)The design should permit the driver to discern quickly either from the layout or from traffic
signs about which path he/she should follow and the actions of merging and diverging.
This can be achieved by good layout, traffic islands, signs and carriageway markings. Good
visibility improves safety;
(iv) The layout should follow the natural vehicle paths. Smoothness, in contrast to abrupt and
sharp corners, should guide minor streams of traffic into stopping or slowing down
positions;
(v) The number of conflict points should be minimised by separating some of the many cutting,
merging or diverging movements;
(vi) Vehicles that are forced to wait in order to cross a traffic stream should be provided with
adequate space at the junction.
4.11.2 At-grade and Grade Separated Junctions
An intersection where all roadways join or cross at the same level is known as an at-grade
intersection. An intersection layout which permits crossing man oeuvres at different levels is
known as a grade separated intersection.
The choice between an at-grade and grade separated intersection at a particular site depends upon
various factors such as traffic, economy, safety, aesthetics, delay etc.
Grade-separated junctions generally are more expensive initially, and are justified in certain
situations; which may include:
(i) On high type facilities such as expressways, freeways and motorways;
(ii) Certain at-grade intersections which have reached the maximum capacity and where it is
not possible to improve the capacity further by retaining the at-grade crossing;
(iii)At certain locations which have a proven record of bad accident history when functioning
as at grade junctions;
(iv) At junctions where the traffic volume is heavy and the delays and economic loss caused
justify the provision of grade-separation;
(v) At certain specific topographical situations where it is logical to provide a grade-separated
structure rather than an at-grade intersection, which may involve considerable earthwork
or acquisition of land.
4.11.3 Basic Forms of At-grade Intersections
Intersections can be divided into the basic forms shown in Figure 4.16. From a design aspect these
intersections can also be divided according to whether they are controlled, priority controlled (stop,
Give Way), space-sharing (i.e. roundabouts), time sharing (i.e. traffic-signal controlled), or grade-
separated (including interchanges) [O’Flaherty, 2002].

Figure 4.16: Basic Forms of At-grade Intersections

4.11.4 Overview of the Design Process
The at-grade intersection design process involves data collection of both traffic and site conditions,
the preparation of preliminary designs from which a layout is selected, and the development of the
final design using appropriate design standards.
Traffic data gathered for design purposes normally include:
 Peak period volumes,
 Turning movements and composition for the design year,
  Vehicle operating speeds on the intersecting roads (these are needed for sight-distance/
speed-change lane design) pedestrian and bicycle movements (these affect the
layout/traffic control design),
 Public transport needs (e.g. bus priority measures and bus stop locations affect the
layout/traffic control design),
 Special needs of oversize vehicles (the selected design may have to cope with the
occasional heavily loaded commercial vehicle with a wide turning path),
 Accident experience (if an existing intersection is being upgraded), and parking practices
(especially in built-up areas).
Site data collected typically include:
 Topography,
 Land usage, and related physical features (natural and manufactured), public and private
utility services (above and below ground),
 Items of special interest (e.g. environmental, cultural and historical features),
 Horizontal and vertical alignments of intersecting roads (existing and future), sight
distances (and physical features which limit them), and adjacent (necessary) accesses.
The preliminary design phase is essentially an iterative one. It involves preparing a number of
possible intersection layouts and generally examining each in terms of its operating characteristics
(especially safety and capacity), ease of construction and likely capital cost, and environmental
and local impacts that might affect the design selection.
The most promising of the rough layouts are then selected for further development and analysis
(including road user and vehicle operating costs, if appropriate), refined and examined in greater
detail until that considered most suitable for the intersection is selected for detailed design and
preparation of final construction plans and specifications [O’Flaherty, 2002].
4.12.5 At-grade Intersection Types (from a design perspective)
Different at-grade junction (intersection) types will be appropriate under different circumstances
depending on traffic flows, speeds, and site limitations.
4.13 An Access
According to MoWH&C (2005), an access is defined as the intersection of an unclassified road
with a classified road and shall generally be provided within the road reserve boundary of the
classified road. Access roads are used to connect properties etc. to the road network. Accident risk
increases with the frequency of access roads, so they should, as far as possible, be discouraged on
higher classes of roads. The lay out and location of the access must satisfy the visibility
requirement for "stop” conditions given in Figure 4.17.
Figure 4.17
4.14 A Junction or an Intersection
A junction is the intersection of two or more classified roads on the same surface / at grade. At
grade intersections can be classified in to two main intersection categories based on the type of
control used. For each category, there are a number of intersection types as shown below.
Table 4.10: Types of At-grade Intersections as recommended by MoWH&C

4.14.1 Priority Intersections

Priority intersections will be adequate in most rural situations. Three types of T intersections are
given below:
(i) Unchannelised T-Intersection (A)
The unchannelised design is suitable for intersections where there is a very small amount
of turning traffic. It is the simplest design and has no traffic islands (see Figure 4.18).
(ii) Partly channelized T-Intersection (B)
 The partly channelized design is for intersections with a moderate volume of turning
traffic. It has a traffic island in the minor road arm. In urban areas, the traffic island would
normally be kerbed in order to provide a refuge for pedestrians crossing the road.
(iii)Channelized T-Intersection (C)
The fully channelized design is for intersections with a high volume of turning traffic or
high – speeds. It has traffic islands in both the minor road and the main road.

Figure 4.18: Typical T-Intersections

Source: Uganda Road Design Manual, 2005
The crossroads form of priority intersection must not be used. It has a very high number of
conflict points, and has a much higher accident risk than any other kind of intersection. Existing
crossroads should, where possible, be converted to a staggered intersection, or roundabout, or be
controlled by traffic signals [MoWH&C, 2005].
4.14.2 Control Intersections
Control intersections are mostly used in towns and trading centres. However, roundabouts can be
used in rural areas in intersections between major roads or other intersections with high traffic
volumes. There are two types of control intersections:
Roundabout (D)
Roundabouts are controlled by the rule that all entry traffic must give way to circulating traffic.
The ratio of minor road incoming traffic to the total incoming traffic should preferably be at least
10 to 15%. Roundabouts can be of normal size, i.e. with central island radius 10 m or more, or
small size, i.e. with central island radius less than 10 m (see Figure 4.18).
Signalised Intersection (E)
Signalised intersections have conflicts separated by traffic signals. No conflicts are allowed
between straight through traffic movements.
Typical design of control intersections is shown in Figure 4.19.

Figure 4.19: Typical Designs for Control Intersections

Source: Uganda Road Design Manual, 2005
a) Design Requirements
The design of at-grade junctions must take account of the following basic requirements:
 Safety
 Operational comfort
 Capacity
 Economy
b) Safety and Operational Comfort
A junction is considered safe when it is perceptible, comprehensible and maneuverable.
These three requirements can generally be met by complying with the following guidelines.
 Perception
 The junction should be sited so that the major road approaches are readily visible;
 Early widening of the junction approaches;
 The use of traffic islands in the minor road to emphasize a “yield” or “stop” requirement.
 The use of early and eye-catching traffic signs;
 Optical guidance by landscaping and the use of road furniture, especially where a junction
must be located on a crest curve;
 The provision of visibility splays which ensure unobstructed sight lines to the left and right
along the major road;
 The angle of intersection of the major and minor roads should be between 70 and 110
degrees;
 The use of single lane approaches is preferred on the minor road in order to avoid mutual
sight obstruction from two vehicles waiting next to each other to turn or cross the major
road.
 Comprehension
 The right of way should follow naturally and logically from the junction layout;
 The types of junctions used throughout the whole road network should be as much as
possible similar;
  The provision of optical guidance by the use of clearly visible kerbs, traffic islands, road
markings, road signs and other road furniture.
 Maneuverability
 A1l traffic lanes should be of adequate width for the appropriate vehicle turning
characteristics. To accommodate truck traffic, turning radii shall be 15 meters minimum;
 The edges of traffic lanes should be clearly indicated by road markings;
 Traffic islands and kerbs should not conflict with the natural vehicle paths.
c) Capacity
The operation of uncontrolled junctions depends principally upon the frequency of gaps which
naturally occur between vehicles in the main road flow. These gaps should be of sufficient duration
to permit vehicles from the minor road to merge with, or cross, the major road flow. In consequence
junctions are limited in capacity, but this capacity may be optimized by, for example,
channelisation or the separation of manoeuvres.
d) Economy
An economical junction design generally results from a minimization of the construction,
maintenance and operational costs.
Delay can be an important operational factor and the saving in time otherwise lost may justify a
more expensive, even grade separated, junction.
Loss of lives, personal injuries and damage to vehicles caused by junction-accidents are considered
as operational "costs" and should be taken into account.
The optimum economic return may often be obtained by a phased construction, for example by
constructing initially an at-grade junction which may later become grade separated.
4.14.3 Selection of Intersection Type
a) General
These selection guidelines provided by MoWH&C mainly deal with traffic safety. The ministry
recommends that other important impacts such as capacity/road user costs, environmental issues,
investment and maintenance costs should also be taken into consideration.
The selection is divided into two steps;
(i) Selection of intersection category (priority or control) and
(ii) Selection of intersection type.
It is based on the following assumptions:
 Priority intersections can be safe and give sufficient capacity for certain traffic volumes
and speed limits;
 If a priority intersection is not sufficient for safety and capacity, the major road traffic must
also be controlled.
Depending on location, traffic conditions and speed limits, different types of priority or control
intersection should be selected.
b) Selection of Intersection Category
(i) Based on Safety
The selection of intersection category should mainly be based on safety. The selection can be made
by using diagrams with the relationships between the safety levels and the average annual daily
approaching traffic volumes (AADT in veh/day) based on accident statistics.
The diagrams shown in Figure 4.20 are for T-intersections on 2-lane roads with 50, 80 and 100
km/h speed limit. The diagrams are, as already stated, based on general European experience on
relationships between speed, safety and traffic flows. They are judged reasonable to be used in
Uganda until sufficient local research is available.

Figure 4.20: Selection of Intersection Category based on Safety

Source: Uganda Road Design Manual, 2005
(ii) Based on Capacity
The selection of intersection category based on safety should be checked for capacity. It can be
made by using diagrams with the relationships between the capacity and the approaching traffic
volumes during the design hour (DHV in pcu/design hour). The diagrams shown in Figure 4.21
are for T-intersections on 2-lane roads with 50, 80 and 100 km/h speed limit. The desired level
refers to a degree of saturation (actual traffic flow/capacity) of 0.5. The acceptable level refers to
a degree of saturation of 0.7.
The diagrams are based on Swedish capacity studies with findings similar to other European
countries. It is judged reasonable to be used in Uganda until sufficient Ugandan research is
available.
Figure 4.21: Selection of Intersection Category based on Capacity
Source: Uganda Road Design Manual, 2005
c) Selection of Intersection Type
Priority intersections
The selection of priority intersection type should mainly be based on safety. The selection can be
made by using diagrams with the relationships between the safety levels and the average annual
daily approaching traffic volumes (AADT in veh/day) based on accident statistics. The diagrams
shown in Figure 4.22 are for T-intersections on 2-lane roads with 50, 80 and 100 km/h speed limit.
Crossroads should be avoided. The number of right turners should obviously also impact the
decision. The diagrams are based on general European findings on safety effects of right turn lanes.
It is judged reasonable to be used in Uganda until sufficient Ugandan statistics are available. Note
however they are only a starting point for determining the most appropriate form of intersection.
Figure 4.22: Selection of Priority Intersection type based on Safety
Source: Uganda Road Design Manual, 2005
Partly channelised T-intersections should normally be used if needed to facilitate pedestrian
crossings and also if the minor road island is needed to improve the visibility of the intersection.
Control intersections
Roundabouts are suitable for almost all situations, provided there is enough space. Roundabouts
have been found to be safer than signalized intersections, and are suitable for both low and medium
traffic flows. At very high traffic volumes they tend to become blocked due to drivers failing to
obey the priority rules. Well-designed roundabouts slow traffic down, which can be useful at the
entry to a built-up area, or where there is a significant change in road standard, such as the change
from a dual carriageway to a single carriageway.
Traffic signals are the favoured option in the larger urban areas. Co-ordinated networks of signals
(Area Traffic Control) can bring major improvements in traffic flow and a significant reduction in
delays and stoppages. However, they must be demand-responsive, in order to get the maximum
capacity from each intersection. Observance of traffic signals by Ugandan drivers is reasonably
good, and could be improved through enforcement campaigns.
For some traffic distributions, for example high traffic volumes on the major road, the total delay
can be shorter in a signalised intersection than in a roundabout. The diagram in Figure 8-8 shows
the traffic conditions for which signalised intersections are most suited, based on Kenyan and UK
experience.

Figure 4.23: Selection of Control Intersection Type

Source: Uganda Road Design Manual, 2005
If a signalized intersection is considered due to planning conditions or traffic volumes, a capacity
analysis and economic analysis should be made. This should include road construction and
maintenance costs, accident costs, travel time costs, vehicle operating costs and environmental
costs.
4.14.4 Capacity of a T-Junction
The capacity of a T junction is primarily dependent upon the ratio of the flows on the major and
minor roads, the critical (minimum) gap in the main road traffic stream acceptable to entering
traffic and the maximum delay acceptable to minor road vehicles. As traffic builds up on the main
road, headways between vehicles decline, fewer acceptable gaps become available, and delays to
vehicles on the minor road increase accordingly, theoretically to infinity.
Field measurements on single carriageway roads indicate that the critical time gaps accepted by
minor road vehicles at the head of a queue average about 3 seconds for left turn merging with, and
4 to 5 seconds for right turn cutting of, the traffic stream in the nearside lane of the main road.
Empirical research has resulted in predictive capacity equations for T-intersections, which were
derived from traffic flow measurements and from certain broad features of junction layout.
A T-intersection has six separate traffic streams (see Figure 4.24 below), of which the through
streams on the major road (C-A and A-C) and the left-turn stream off the major road (A-B) are
generally assumed to be priority streams and to suffer no delays from other traffic, while the two
minor road streams (B-A and B-C) and the major road right-turn stream (C-B) incur delays due to
their need to give way to higher priority streams. Predictive capacity equations for the three
nonpriority streams are as follows:
Figure 4.24: Selection of Control Intersection Type
Source: O’Flaherty, 2002
 The superscript s (e.g. qs B-A) denotes the flow from the saturated stream i.e. one in which
there is stable queuing.
 The geometric parameters wB-A and wB-C denote the average widths of each of the minor
road approach lanes for waiting vehicles in streams B-A and B-C respectively, measured
over a distance 20m upstream from the Give Way line;
 wC-B denotes the average width of the right-turn (central) lane on the major road, or 2.1m
if there is no explicit provision for right turners in stream C-B.
 The parameters VrB-A and VlB-C denote right and left visibility distances, respectively,
available from the road;
 VrC-B is the visibility available to right-turning vehicles waiting to turn right from the
major road;
 W is the average major road carriageway width at the intersection; in the case of ghost or
raised islands, W excludes the width of the central (turning) lane;
 WCR is the average width of the central reserve lane at the intersection on a dual
carriageway road.
One heavy vehicle is considered equivalent to two (2) pcu for calculation purposes. Capacities are
always positive or zero; if the right-hand side of any equation is negative, the capacity is taken as
zero. The ranges within which the geometric data are considered valid are as follows: w = 2.05-
4.70m, Vr = 17 – 250m, Vl = 22 – 250m, WCR = 1.2 - 9m (dual carriageway sites only), W = 6.4
-
20m.
4.14.5 Design Reference Flow (DRF)
One of the methodologies used to assess the adequacy of the capacity available to a non-priority
traffic stream is the ratio of the design reference flow (DRF) to the capacity called the reference
flow to capacity (RFC) ratio. For the satisfactory operation of any given approach lane it is
generally considered that reference flow to capacity ratio should not exceed 0.85. DRF value
considers the function of the road. The 200th highest hourly flow in the design year is used on
recreational roads, the 50th highest hourly flow on interurban roads and the 30th highest hourly
flow in the design year on urban roads. It would be economically and/or environmentally
undesirable to design for the highest hours in the design year. For an existing intersection the DRF
values are often determined from manual counts (including classifications and turning movements)
of the existing flows which are grossed up to the design year using appropriate factors.
4.15.6 Delay
An estimate of the total 24 hour delay due to congestion, D24x, at an existing T-intersection can
be estimated from the empirically derived equation

Where;
D3 = total intersection delay (h) during the peak three hours, and
P23 = ratio of flow in the peak three hours to the 24-hour flow. The above formula assumes that
delays are inflicted only on minor road vehicles, which have to yield priority to the major road
streams.
T-Junction Example
A new industrial complex is planned to be sited adjacent to an existing priority intersection. The
width of the main carriageway is 8m. The width of the carriageway for traffic movements B-A,
BC and C-B are 3, 3 and 2m respectively. The visibility distances at the drivers’ eye height for the
junction are: VrB-A = 60m, VlB-A = 75m, VrB-C = 60m, VrC-B = 60m. The width of the central
reservation is 2m wide. The design flows (in pcu/hr) are represented in the figure below.
You are required to determine the following:
i) Calculate the capacities of the turning movements; qs
B-A, qs
B-C, qs
C-B, for the priority intersection shown in the figure above.
ii) Asses the arms of the junctions and advise on which arms have sufficient capacity and which
ones do not.
Based on the reference flow capacity ratios obtained, it is apparent that all the arms have exceeded
their capacities and therefore need to be redesigned.
4.14.7 Rotary Intersections (Roundabouts)
A roundabout is a form of channelization intersection in which vehicles are guided onto a one-way
circulatory road about a central island. Entry to the intersection is controlled by Give Way
markings and priority is now given to vehicles circulating (clockwise in Uganda) in the round
about.
The main objective of roundabout design is to secure the safe interchange of traffic between
crossing traffic streams with the minimum delay. The operating efficiency of a roundabout
depends upon entering drivers accepting headway gaps in the circulating traffic stream. Traffic
streams merge and diverge at small angles and low relative speeds. For this reason, accidents
between vehicles in roundabouts rarely have fatal consequences [O’Flaherty, 2002].
a) General Usage of Roundabouts
Roundabouts are most effective as at grade intersections in urban or rural areas that have all or a
number of the following characteristics:
 High proportions and/or volumes of right turning traffic;
 Priority is not given to traffic from any particular road;
 Presence of accidents involving crossing or turning movements;
 Traffic on the minor roads is delayed by the use of ‘Stop’ or ‘Give Way’ signs;
 Where they cause less overall delay to vehicles than traffic signals;
 Where there is a marked change in road standard e.g. from a dual to a single carriageway
road.
Roundabout intersections are not appropriate at the following sites:
 Where there is inadequate space or unfavourable topography that limits a good geometric
design;
 Where traffic flows are unbalanced, e.g. at major/ minor T-Intersections;
 Where they follow a downhill approaches. The approach should be at least a 2% grade and
should be flattened at least 30m to the intersection.
 Where there are heavy volumes of vehicular traffic and where there is heavy cyclist and
pedestrian traffic.
 Between traffic controlled signal intersections which could cause queuing back into the
roundabout exits.
b) Types of Roundabouts
In Uganda there are two types of roundabouts namely:
(i) Normal roundabouts with a centre island radius greater than or equal to 10m. The central
island radius should normally be between 10m and 25m otherwise it becomes difficult to control
speeds for a radius bigger than the above range and puts pedestrians and cyclists at risk. The width
of the circulating carriageway depends on whether it is one or two lane.
(ii) Small roundabouts with a central island less than 10m. The inner central island radius should
be at least 2m.
c) Design Features of Roundabouts
 For small roundabouts, the central island should be approximately 1/3 of the inscribed circle
diameter (1/3D);
 At larger sites the proportion should be >1/3 to limit the circulatory width to a maximum of
 15m;
 The circulatory width around the roundabout should be constant at about 1.0 to1.2 times the
highest entry width subject to the above maximum of 15m;
 Steep downhill gradients should be avoided at roundabout approaches;
 The frequent occurrence of roundabouts on high speed rural roads should be avoided;
  Mini roundabouts must only be used at existing junctions where there are space limitations
and where the 85 percentile approach speed on all approaches is less than 50km/hr;
 Entries should be flares. Single and two lane approaches should become 3 and 4 lanes
respectively at the give way line;
 The entry flare taper should be approximately 1 in 3. Each lane should be 2.5m to 3.5m
wide at the give way line dependent on site conditions. The taper width at the Give Way
line should never be less than 3m. The best entry angle is approximately 30 degrees. Lanes
may be tapered to 2m width (minimum) on the roundabout approaches;
 The entry width of an approach arm at a roundabout is one of the major factors apart from
approach carriageway half width that affects capacity. Flares on the approaches to
roundabouts should be designed in such a way that maximum entry widths are not greater
than 10.5m on single carriageway roads and 15m on dual carriageway approach roads. A
typical flare length on a rural road is 25m. The length can be as low as 5m on urban roads;
 Pedestrian crossing places (including zebras) should normally be within the flared approach
but as far from the Give Way line as pedestrian convenience will allow. This reduces the
road width to be crossed by pedestrians. A central refuge should always be provided
wherever possible. A deflection island may fulfil this function but should be at least 1.2m
wide;
 Pedestrian guard rail should be used, where necessary to control haphazard pedestrian
crossing of the traffic streams. It also improves safety.
d) Capacity of Roundabouts
The capacity of a roundabout as a whole is a function of the capacities of the individual entry arms.
The capacity of each arm is defined as the maximum inflow when the traffic flow at the entry is
sufficient to cause continuous queuing in its approach road.
The main factors influencing entry capacity are the approach half width, and the width and flare
of the entry, while the entry angle and radius also have small but significant effects. The predictive
equation used with all types of single at-grade roundabouts is

Where; Qe = saturation or capacity entry flow (pcu/h);

Qc = circulating flow across the entry (pcu/h);
k = 1 - 0.00347(ϕ-30) – 0.978[(1/r)-0.05];
F = 303x2; where; x2 = v + (e – v)/(1+2S) and S = 1.6(e –v)/l’;
fc = 0.210tD(1 + 0.2x2) where tD = 1 + 0.5/(1 +M) and M = exp[(D-60)/10].
The symbols e, v, l’, S, D, ϕ and r are described in Table 4.11. Qe and Qc are in pcu/h, and one
heavy goods vehicle is assumed equivalent to 2 pcu for computation purposes.

Table 4.11: The Limits of the Parameters used in Roundabout Capacity Equation
From the roundabout equation above, entry capacity decreases as circulation flow increases. The
sharpness of flare, S is a measure of the rate at which extra width is developed in the entry flare.
Small ‘S’ values correspond to long gradual flares and big ones to short severe ones. The angle ϕ
acts as an alternative for the conflict angle between the entering and circulating traffic streams.
The entry radius, r is measured as the minimum radius of curvature of the nearest kerbline at entry.
e) Design Reference Flow (DRF)
When designing a roundabout intersection, the entry angle for each arm of a trial layout is
compared with the hourly flow for the design (DRF). The reference flow to capacity ratio (RFC)
is an indicator of the likely performance of an intersection under the future year traffic loading. If
an RFC ratio of 0.85 occurs, it can be expected that queuing will automatically be avoided in the
design year peak hour in five out of six cases.
Roundabout Example
The table below shows measured turning movements in the AM peak period as recorded in a traffic
survey at a four arm roundabout. The survey was carried out in 2005. The expected rate of traffic
growth is 2%. It is assumed that funding will be readily available and that if any redesign and
reconstruction is needed, the roundabout will be reopened to traffic in the same year the survey
was carried out. The roundabout is being assessed for capacity to carry peak flows in [Link]
geometric parameters for arms A and B are as shown below:

The base year traffic survey carried out in 2005 revealed the following traffic flows in pcu/hr.
The general layout of the roundabout is shown below
Determine the following;
• The design flows for the year 2019
• The approach capacity of arms A and B of the roundabout.
• Establish which of the two arms still has capacity and which one does not.

1.0 Summary of Design Data

a) Traffic growth rate, r = 2%
b) Design life, Y [= (2019-2005)+1] = 15yrs
c) Geometric parameters of Arm A and B as shown in the table
2.0 Sketch Drawing
As shown in the diagram above
3.0 Traffic Assessment

3.1 Design Flow, DF

Where; P = present flows (in pcu/hr);

r = traffic growth rate (in %);
 Y = design life (in years);
DF = Design Flow (a modification of the future traffic flow);
DRF = Design Reference Flow.

The design flows, DF in 2019 are presented in the table below using the above formulae
4.0 Conclusion
Arm C has a RFC ratio of 61% which is less than 70%, implying that queuing on this arm will be
avoided for 39 out of 40 peak hours.
Arm D, on the other hand, has a RFC ratio of 120% which is far greater than 85%, implying that
queuing will occur on this arm of the roundabout in all the peak hours.