Roundabouts Part 3
Roundabouts Part 3
Roundabouts: Part 3
Course# TE4033
[Link]
Ezekiel Enterprises, LLC
301 Mission Dr. Unit 571
New Smyrna Beach, FL 32128
800-433-1487
helpdesk@[Link]
6 6.1
Geometric Design
Introduction 130
6.1.1 Geometric elements 130
6.1.2 Design process 130
6.2 General Design Principles 132
6.2.1 Speeds through the roundabout 132
6.2.2 Design vehicle 142
6.2.3 Nonmotorized design users 144
6.2.4 Alignment of approaches and entries 144
6.3 Geometric Elements 145
6.3.1 Inscribed circle diameter 145
6.3.2 Entry width 147
6.3.3 Circulatory roadway width 149
6.3.4 Central island 150
6.3.5 Entry curves 152
6.3.6 Exit curves 154
6.3.7 Pedestrian crossing location and treatments 155
6.3.8 Splitter islands 157
6.3.9 Stopping sight distance 159
6.3.10 Intersection sight distance 161
6.3.11 Vertical considerations 164
6.3.12 Bicycle provisions 167
6.1 Introduction
Roundabout design Designing the geometry of a roundabout involves choosing between trade-offs of
involves trade-offs among safety and capacity. Roundabouts operate most safely when their geometry forces
safety, operations, traffic to enter and circulate at slow speeds. Horizontal curvature and narrow pave-
and accommodating ment widths are used to produce this reduced-speed environment. Conversely,
large vehicles. the capacity of roundabouts is negatively affected by these low-speed design ele-
ments. As the widths and radii of entry and circulatory roadways are reduced, so
also the capacity of the roundabout is reduced. Furthermore, many of the geomet-
ric parameters are governed by the maneuvering requirements of the largest ve-
hicles expected to travel through the intersection. Thus, designing a roundabout is
a process of determining the optimal balance between safety provisions, opera-
tional performance, and large vehicle accommodation.
Some roundabout features are While the basic form and features of roundabouts are uniform regardless of their
uniform, while others vary location, many of the design techniques and parameters are different, depending
depending on the location and on the speed environment and desired capacity at individual sites. In rural environ-
size of the roundabout. ments where approach speeds are high and bicycle and pedestrian use may be
minimal, the design objectives are significantly different from roundabouts in ur-
ban environments where bicycle and pedestrian safety are a primary concern. Ad-
ditionally, many of the design techniques are substantially different for single-lane
roundabouts than for roundabouts with multiple entry lanes.
This chapter is organized so that the fundamental design principles common among
all roundabout types are presented first. More detailed design considerations spe-
cific to multilane roundabouts, rural roundabouts, and mini-roundabouts are given
in subsequent sections of the chapter.
Exhibit 6-1 provides a review of the basic geometric features and dimensions of a
roundabout. Chapter 1 provided the definitions of these elements.
Roundabout design is an The process of designing roundabouts, more so than other forms of intersections,
iterative process. requires a considerable amount of iteration among geometric layout, operational
analysis, and safety evaluation. As described in Chapters 4 and 5, minor adjust-
ments in geometry can result in significant changes in the safety and/or opera-
tional performance. Thus, the designer often needs to revise and refine the initial
layout attempt to enhance its capacity and safety. It is rare to produce an optimal
geometric design on the first attempt. Exhibit 6-2 provides a graphical flowchart for
the process of designing and evaluating a roundabout.
Roundabout process.
As Potential
Design Option
PLANNING
Evaluate
Appropriateness
OPERATIONS
Preliminary Detailed
Capacity PerformanceAnalysis
Analysis
GEOMETRIC
DESIGN
PerformSafety Audit
Check of Signing,Striping,
SAFETY
Signing, Adjustas
Striping,Lighting, Necessary
Landscaping,and
Construction
Staging
This section describes the fundamental design principles common among all cat-
egories of roundabouts. Guidelines for the design of each geometric element are
provided in the following section. Further guidelines specific to double-lane round-
abouts, rural roundabouts, and mini-roundabouts are given in subsequent sections.
Note that double-lane roundabout design is significantly different from single-lane
roundabout design, and many of the techniques used in single-lane roundabout
design do not directly transfer to double-lane design.
The most critical design objective Because it has profound impacts on safety, achieving appropriate vehicular speeds
is achieving appropriate vehicular through the roundabout is the most critical design objective. A well-designed round-
speeds through the roundabout. about reduces the relative speeds between conflicting traffic streams by requiring
vehicles to negotiate the roundabout along a curved path.
Through movements are usually Exhibits 6-5 and 6-6 illustrate the construction of the fastest vehicle paths at a
the fastest path, but sometimes single-lane roundabout and at a double-lane roundabout, respectively. Exhibit 6-7
right turn paths are more provides an example of an approach at which the right-turn path is more critical
critical. than the through movement.
As shown in Exhibits 6-5 and 6-6, the fastest path for the through movement is a
series of reverse curves (i.e., a curve to the right, followed by a curve to the left,
followed by a curve to the right). When drawing the path, a short length of tangent
should be drawn between consecutive curves to account for the time it takes for
a driver to turn the steering wheel. It may be initially better to draw the path free-
hand, rather than using drafting templates or a computer-aided design (CAD) pro-
gram. The freehand technique may provide a more natural representation of the
way a driver negotiates the roundabout, with smooth transitions connecting curves
and tangents. Having sketched the fastest path, the designer can then measure
the minimum radii using suitable curve templates or by replicating the path in CAD
and using it to determine the radii.
The entry path radius should The design speed of the roundabout is determined from the smallest radius along
not be significantly larger than the fastest allowable path. The smallest radius usually occurs on the circulatory
the circulatory radius. roadway as the vehicle curves to the left around the central island. However, it is
important when designing the roundabout geometry that the radius of the entry
path (i.e., as the vehicle curves to the right through entry geometry) not be signifi-
cantly larger than the circulatory path radius.
Draw the fastest path for all The fastest path should be drawn for all approaches of the roundabout. Because
roundabout approaches. the construction of the fastest path is a subjective process requiring a certain
amount of personal judgment, it may be advisable to obtain a second opinion.
Superelevation values are usually assumed to be +0.02 for entry and exit curves
and -0.02 for curves around the central island. For more details related to
superelevation design, see Section 6.3.11.
Values for side friction factor can be determined in accordance with the AASHTO
relation for curves at intersections (see 1994 AASHTO Figure III-19 (4)). The coeffi-
cient of friction between a vehicle’s tires and the pavement varies with the vehicle’s
speed, as shown in Exhibits 6-8 and 6-9 for metric and U.S. customary units,
respectively.
0.60
0.50
Side Friction Factor
0.40
0.30
0.20
0.10
0.00
5 10 15 20 25 30 35 40
Speed (mph)
Using the appropriate friction factors corresponding to each speed, Exhibits 6-10
and 6-11 present charts in metric and U.S. customary units, respectively, showing
the speed-radius relationship for curves for both a +0.02 superelevation and -0.02
superelevation.
50
40
Speed (km/h)
30
20
10
0
0 20 40 60 80 100 120
Radius (m)
e=+0.02 e=-0.02
40
35
30
Speed (mph)
25
20
15
10
0
0 50 100 150 200 250 300 350 400
Radius (ft)
e=+0.02 e=-0.02
As shown in Exhibit 6-12, five critical path radii must be checked for each ap-
proach. R1 , the entry path radius, is the minimum radius on the fastest through
path prior to the yield line. R2 , the circulating path radius, is the minimum radius
on the fastest through path around the central island. R3 , the exit path radius, is
the minimum radius on the fastest through path into the exit. R4 , the left-turn
path radius, is the minimum radius on the path of the conflicting left-turn move-
ment. R5 , the right-turn path radius, is the minimum radius on the fastest path of
a right-turning vehicle. It is important to note that these vehicular path radii are
not the same as the curb radii. First the basic curb geometry is laid out, and then
the vehicle paths are drawn in accordance with the procedures described in Sec-
tion [Link].
On the fastest path, it is desirable for R1 to be smaller than R2 , which in turn should
be smaller than R3 . This ensures that speeds will be reduced to their lowest level at
the roundabout entry and will thereby reduce the likelihood of loss-of-control crashes.
It also helps to reduce the speed differential between entering and circulating traf-
fic, thereby reducing the entering-circulating vehicle crash rate. However, in some
cases it may not be possible to achieve an R1 value less than R2 within given right-
of-way or topographic constraints. In such cases, it is acceptable for R1 to be greater
than R2 , provided the relative difference in speeds is less than 20 km/h (12 mph)
and preferably less than 10 km/h (6 mph).
The natural path of a vehicle is At single-lane roundabouts, it is relatively simple to reduce the value of R1 . The
the path that a driver would curb radius at the entry can be reduced or the alignment of the approach can be
take in the absence of other shifted further to the left to achieve a slower entry speed (with the potential for
conflicting vehicles. higher exit speeds that may put pedestrians at risk). However, at double-lane round-
abouts, it is generally more difficult as overly small entry curves can cause the
natural path of adjacent traffic streams to overlap. Path overlap happens when the
geometry leads a vehicle in the left approach lane to naturally sweep across the
right approach lane just before the approach line to avoid the central island. It may
also happen within the circulatory roadway when a vehicle entering from the right-
hand lane naturally cuts across the left side of the circulatory roadway close to the
central island. When path overlap occurs at double-lane roundabouts, it may re-
duce capacity and increase crash risk. Therefore, care must be taken when design-
ing double-lane roundabouts to achieve ideal values for R1 , R2, and R3 . Section 6.4
provides further guidance on eliminating path overlap at double-lane roundabouts.
The exit radius, R3 , should not be less than R1 or R2 in order to minimize loss-of-
control crashes. At single-lane roundabouts with pedestrian activity, exit radii may
still be small (the same or slightly larger than R2) in order to minimize exit speeds.
However, at double-lane roundabouts, additional care must be taken to minimize
the likelihood of exiting path overlap. Exit path overlap can occur at the exit when a
vehicle on the left side of the circulatory roadway (next to the central island) exits
into the right-hand exit lane. Where no pedestrians are expected, the exit radii
should be just large enough to minimize the likelihood of exiting path overlap. Where
pedestrians are present, tighter exit curvature may be necessary to ensure suffi-
ciently low speeds at the downstream pedestrian crossing.
Finally, the radius of the fastest possible right-turn path, R5 , is evaluated. Like R1 ,
the right-turn radius should have a design speed at or below the maximum design
speed of the roundabout and no more than 20 km/h (12 mph) above the conflicting
R4 design speed.
Single-Lane Roundabout
30 11 21 54 41
35 13 23 61 43
40 16 25 69 45
45 19 26 73 46
Double-Lane Roundabout
45 15 24 65 44
50 17 25 69 45
55 20 27 78 47
60 23 28 83 48
65 25 29 88 49
70 28 30 93 50
Single-Lane Roundabout
100 35 13 165 25
115 45 14 185 26
130 55 15 205 27
150 65 15 225 28
Double-Lane Roundabout
150 50 15 205 27
165 60 16 225 28
180 65 16 225 28
200 75 17 250 29
215 85 18 275 30
230 90 18 275 30
The design vehicle dictates many Another important factor determining a roundabout’s layout is the need to ac-
of the roundabout’s dimensions. commodate the largest motorized vehicle likely to use the intersection. The turn-
ing path requirements of this vehicle, termed hereafter the design vehicle, will
dictate many of the roundabout’s dimensions. Before beginning the design pro-
cess, the designer must be conscious of the design vehicle and possess the
appropriate vehicle turning templates or a CAD-based vehicle turning path pro-
gram to determine the vehicle’s swept path.
The choice of design vehicle will vary depending upon the approaching roadway
types and the surrounding land use characteristics. The local or State agency with
jurisdiction of the associated roadways should usually be consulted to identify
the design vehicle at each site. The AASHTO A Policy on Geometric Design of
Highways and Streets provides the dimensions and turning path requirements
for a variety of common highway vehicles (4). Commonly, WB-15 (WB-50) ve-
hicles are the largest vehicles along collectors and arterials. Larger trucks, such
as WB-20 (WB-67) vehicles, may need to be addressed at intersections on inter-
state freeways or State highway systems. Smaller design vehicles may often be
chosen for local street intersections.
Exhibits 6-15 and 6-16 demonstrate the use of a CAD-based computer program
to determine the vehicle’s swept path through the critical turning movements.
Like the motorized design vehicle, the design criteria of nonmotorized potential
roundabout users (bicyclists, pedestrians, skaters, wheelchair users, strollers, etc.)
should be considered when developing many of the geometric elements of a round-
about design. These users span a wide range of ages and abilities that can have a
significant effect on the design of a facility.
The basic design dimensions for various design users are given in Exhibit 6-17 (5).
Bicycles
Length 1.8 m (5.9 ft) Splitter island width at crosswalk
Minimum operating width 1.5 m (4.9 ft) Bike lane width
Lateral clearance on each side 0.6 m (2.0 ft); Shared bicycle-pedestrian path
width
1.0 m (3.3 ft)
to obstructions
Pedestrian (walking)
Width 0.5 m (1.6 ft) Sidewalk width, crosswalk width
Wheelchair
Minimum width 0.75 m (2.5 ft) Sidewalk width, crosswalk width
Operating width 0.90 m (3.0 ft) Sidewalk width, crosswalk width
Person pushing stroller
Length 1.70 m (5.6 ft) Splitter island width at crosswalk
Skaters
Typical operating width 1.8 m (6 ft) Sidewalk width
Source: (5)
Roundabouts are optimally located In general, the roundabout is optimally located when the centerlines of all approach
when all approach centerlines legs pass through the center of the inscribed circle. This location usually allows the
pass through the center of the geometry to be adequately designed so that vehicles will maintain slow speeds
inscribed circle. through both the entries and the exits. The radial alignment also makes the central
island more conspicuous to approaching drivers.
If it is not possible to align the legs through the center point, a slight offset to the
left (i.e., the centerline passes to the left of the roundabout’s center point) is ac-
ceptable. This alignment will still allow sufficient curvature to be achieved at the
entry, which is of supreme importance. In some cases (particularly when the in-
scribed circle is relatively small), it may be beneficial to introduce a slight offset of
the approaches to the left in order to enhance the entry curvature. However, care
must be taken to ensure that such an approach offset does not produce an exces-
sively tangential exit. Especially in urban environments, it is important that the exit
geometry produce a sufficiently curved exit path in order to keep vehicle speeds
low and reduce the risk for pedestrians.
It is almost never acceptable for an approach alignment to be offset to the right of Approach alignment should not
the roundabout’s center point. This alignment brings the approach in at a more be offset to the right of the
tangential angle and reduces the opportunity to provide sufficient entry curvature. roundabout’s center point.
Vehicles will be able to enter the roundabout too fast, resulting in more loss-of-
control crashes and higher crash rates between entering and circulating vehicles.
Exhibit 6-18 illustrates the preferred radial alignment of entries.
In addition, it is desirable to equally space the angles between entries. This pro-
vides optimal separation between successive entries and exits. This results in op-
timal angles of 90 degrees for four-leg roundabouts, 72 degrees for five-leg round-
abouts, and so on. This is consistent with findings of the British accident prediction
models described in Chapter 5.
This section presents specific parameters and guidelines for the design of each
geometric element of a roundabout. The designer must keep in mind, however,
that these components are not independent of each other. The interaction between
the components of the geometry is far more important than the individual pieces.
Care must be taken to ensure that the geometric elements are all compatible with
each other so that the overall safety and capacity objectives are met.
The inscribed circle diameter is the distance across the circle inscribed by the
outer curb (or edge) of the circulatory roadway. As illustrated in Exhibit 6-1, it is the
sum of the central island diameter (which includes the apron, if present) and twice
the circulatory roadway. The inscribed circle diameter is determined by a number
of design objectives. The designer often has to experiment with varying diameters
before determining the optimal size at a given location.
For a single-lane roundabout, At single-lane roundabouts, the size of the inscribed circle is largely dependent
the minimum inscribed circle upon the turning requirements of the design vehicle. The diameter must be large
diameter is 30 m (100 ft) to enough to accommodate the design vehicle while maintaining adequate deflection
accommodate a WB-15 (WB-50) curvature to ensure safe travel speeds for smaller vehicles. However, the circula-
vehicle. tory roadway width, entry and exit widths, entry and exit radii, and entry and exit
angles also play a significant role in accommodating the design vehicle and provid-
ing deflection. Careful selection of these geometric elements may allow a smaller
inscribed circle diameter to be used in constrained locations. In general, the in-
scribed circle diameter should be a minimum of 30 m (100 ft) to accommodate a
WB-15 (WB-50) design vehicle. Smaller roundabouts can be used for some local
street or collector street intersections, where the design vehicle may be a bus or
single-unit truck.
For a double-lane roundabout, At double-lane roundabouts, accommodating the design vehicle is usually not a
the minimum inscribed circle constraint. The size of the roundabout is usually determined either by the need to
diamter is 45 m (150 ft). achieve deflection or by the need to fit the entries and exits around the circumfer-
ence with reasonable entry and exit radii between them. Generally, the inscribed
circle diameter of a double-lane roundabout should be a minimum of 45 m (150 ft).
In general, smaller inscribed diameters are better for overall safety because they
help to maintain lower speeds. In high-speed environments, however, the design
of the approach geometry is more critical than in low-speed environments. Larger
inscribed diameters generally allow for the provision of better approach geometry,
which leads to a decrease in vehicle approach speeds. Larger inscribed diameters
also reduce the angle formed between entering and circulating vehicle paths, thereby
reducing the relative speed between these vehicles and leading to reduced enter-
ing-circulating crash rates (2). Therefore, roundabouts in high-speed environments
may require diameters that are somewhat larger than those recommended for
low-speed environments. Very large diameters (greater than 60 m [200 ft]), how-
ever, should generally not be used because they will have high circulating speeds
and more crashes with greater severity. Exhibit 6-19 provides recommended ranges
of inscribed circle diameters for various site locations.
Entry width is the largest determinant of a roundabout’s capacity. The capacity of Entry width is the largest
an approach is not dependent merely on the number of entering lanes, but on the determinant of a roundabout’s
total width of the entry. In other words, the entry capacity increases steadily with capacity.
incremental increases to the entry width. Therefore, the basic sizes of entries and
circulatory roadways are generally described in terms of width, not number of
lanes. Entries that are of sufficient width to accommodate multiple traffic streams
(at least 6.0 m [20 ft]) are striped to designate separate lanes. However, the circu-
latory roadway is usually not striped, even when more than one lane of traffic is
expected to circulate (for more details related to roadway markings, see Chapter 7).
As shown in Exhibit 6-1, entry width is measured from the point where the yield Entry widths should be kept to
line intersects the left edge of the traveled-way to the right edge of the traveled- a minimum to maximize safety
way, along a line perpendicular to the right curb line. The width of each entry is while achieving capacity and
dictated by the needs of the entering traffic stream. It is based on design traffic performance objectives.
volumes and can be determined in terms of the number of entry lanes by using
Chapter 4 of this guide. The circulatory roadway must be at least as wide as the
widest entry and must maintain a constant width throughout.
When the capacity requirements can only be met by increasing the entry width,
this can be done in two ways:
Exhibit 6-20 and Exhibit 6-21 illustrate these two widening options.
Flare lengths should be As discussed in Chapter 4, flaring is an effective means of increasing capacity
at least 25 m in urban areas and without requiring as much right-of-way as a full lane addition. While increasing the
40 m in rural areas. length of flare increases capacity, it does not increase crash frequency. Conse-
quently, the crash frequency for two approaches with the same entry width will be
essentially the same, whether they have parallel entry lanes or flared entry de-
signs. Entry widths should therefore be minimized and flare lengths maximized to
achieve the desired capacity with minimal effect on crashes. Generally, flare lengths
should be a minimum of 25 m (80 ft) in urban areas and 40 m (130 ft) in rural areas.
However, if right-of-way is constrained, shorter lengths can be used with notice-
able effects on capacity (see Chapter 4).
In some cases, a roundabout designed to accommodate design year traffic vol- Two-phase designs allow for
umes, typically projected 20 years from the present, can result in substantially small initial entry widths that
wider entries and circulatory roadway than needed in the earlier years of operation. can be easily expanded in the
Because safety will be significantly reduced by the increase in entry width, the future when needed to
designer may wish to consider a two-phase design solution. In this case, the first- accommodate greater traffic
phase design would provide the entry width requirements for near-term traffic vol- volumes.
umes with the ability to easily expand the entries and circulatory roadway to ac-
commodate future traffic volumes. The interim solution should be accomplished by
first laying out the ultimate plan, then designing the first phase within the ultimate
curb lines. The interim roundabout is often constructed with the ultimate inscribed
circle diameter, but with a larger central island and splitter islands. At the time
additional capacity is needed, the splitter and central islands can be reduced in size
to provide additional widths at the entries, exits, and circulatory roadway.
The required width of the circulatory roadway is determined from the width of the
entries and the turning requirements of the design vehicle. In general, it should
always be at least as wide as the maximum entry width (up to 120 percent of the
maximum entry width) and should remain constant throughout the roundabout (3).
In some cases (particularly where the inscribed diameter is small or the design Truck aprons generally provide a
vehicle is large) the turning requirements of the design vehicle may dictate that the lower level of operations, but
circulatory roadway be so wide that the amount of deflection necessary to slow may be needed to provide
passenger vehicles is compromised. In such cases, the circulatory roadway width adequate deflection while still
can be reduced and a truck apron, placed behind a mountable curb on the central accommodating the design
island, can be used to accommodate larger vehicles. However, truck aprons gener- vehicle.
ally provide a lower level of operation than standard nonmountable islands. They
are sometimes driven over by four-wheel drive automobiles, may surprise inatten-
tive motorcyclists, and can cause load shifting on trucks. They should, therefore, be
used only when there is no other means of providing adequate deflection while
accommodating the design vehicle.
Exhibit 6-22 provides minimum recommended circulatory roadway widths for two-
lane roundabouts where semi-trailer traffic is relatively infrequent.
As described in Section 6.2.1, the size of the central island plays a key role in
determining the amount of deflection imposed on the through vehicle’s path. How-
ever, its diameter is entirely dependent upon the inscribed circle diameter and the
required circulatory roadway width (see Sections 6.3.1 and 6.3.3, respectively).
Therefore, once the inscribed diameter, circulatory roadway width, and initial entry
geometry have been established, the fastest vehicle path must be drawn though
the layout, as described in Section [Link], to determine if the central island size is
adequate. If the fastest path exceeds the design speed, the central island size may
need to be increased, thus increasing the overall inscribed circle diameter. There
may be other methods for increasing deflection without increasing the inscribed
diameter, such as offsetting the approach alignment to the left, reducing the entry
width, or reducing the entry radius. These treatments, however, may preclude the
ability to accommodate the design vehicle.
Where aprons are used, they should be designed so that they are traversable by
trucks, but discourage passenger vehicles from using them. They should generally
be 1 to 4 m (3 to 13 ft) wide and have a cross slope of 3 to 4 percent away from the
central island. To discourage use by passenger vehicles, the outer edge of the
apron should be raised a minimum of 30 mm (1.2 in) above the circulatory road-
way surface (6). The apron should be constructed of colored and/or textured paving
Leeds, MD
Issues regarding landscaping and other treatments within the central island are
discussed in Chapter 7.
As shown in Exhibit 6-1, the entry curves are the set of one or more curves along
the right curb (or edge of pavement) of the entry roadway leading into the circula-
tory roadway. It should not be confused with the entry path curve, defined by the
radius of the fastest vehicular travel path through the entry geometry (R1 on Exhibit
6-12).
The entry curve is designed curvilinearly tangential to the outside edge of the
circulatory roadway. Likewise, the projection of the inside (left) edge of the entry
roadway should be curvilinearly tangential to the central island. Exhibit 6-24 shows
a typical roundabout entrance geometry.
The primary objective in selecting a radius for the entry curve is to achieve the
speed objectives, as described in Section 6.2.1. The entry radius should first pro-
duce an appropriate design speed on the fastest vehicular path. Second, it should
desirably result in an entry path radius (R1) equal to or less than the circulating path
radius (R2) (see Section [Link]).
At rural and suburban locations, consideration should be given to the speed dif-
ferential between the approaches and entries. If the difference is greater than 20
km/h (12 mph), it is desirable to introduce approach curves or some other speed
reduction measures to reduce the speed of approaching traffic prior to the entry
curvature. Further details on rural roundabout design are provided in Section 6.5.
Exit curves usually have larger radii than entry curves to minimize the likelihood of
congestion at the exits. This, however, is balanced by the need to maintain low
speeds at the pedestrian crossing on exit. The exit curve should produce an exit
path radius (R3 in Exhibit 6-12) no smaller than the circulating path radius (R2). If the
exit path radius is smaller than the circulating path radius, vehicles will be traveling
too fast to negotiate the exit geometry and may crash into the splitter island or into
oncoming traffic in the adjacent approach lane. Likewise, the exit path radius should
not be significantly greater than the circulating path radius to ensure low speeds at
the downstream pedestrian crossing.
The exit curve is designed to be curvilinearly tangential to the outside edge of the
circulatory roadway. Likewise, the projection of the inside (left) edge of the exit
roadway should be curvilinearly tangential to the central island. Exhibit 6-25 shows
a typical exit layout for a single-lane roundabout.
In rural locations where there are few pedestrians, exit curvature may be de-
signed with large radii, allowing vehicles to exit quickly and accelerate back to
traveling speed. This, however, should not result in a straight path tangential to
the central island because many locations that are rural today become urban in
the future. Therefore, it is recommended that pedestrian activity be considered at
all exits except where separate pedestrian facilities (paths, etc.) or other restric-
tions eliminate the likelihood of pedestrian activity in the foreseeable future.
Pedestrian crossing locations at roundabouts are a balance among pedestrian Pedestrian crossing locations
convenience, pedestrian safety, and roundabout operations: must balance pedestrian
convenience, pedestrian safety,
• Pedestrian convenience: Pedestrians want crossing locations as close to the
and roundabout operations.
intersection as possible to minimize out-of-direction travel. The further the cross-
ing is from the roundabout, the more likely that pedestrians will choose a shorter
route that may put them in greater danger.
• Pedestrian safety: Both crossing location and crossing distance are important.
Crossing distance should be minimized to reduce exposure of pedestrian-ve-
hicle conflicts. Pedestrian safety may also be compromised at a yield-line cross-
walk because driver attention is directed to the left to look for gaps in the
circulating traffic stream. Crosswalks should be located to take advantage of
the splitter island; crosswalks located too far from the yield line require longer
splitter islands. Crossings should also be located at distances away from the
yield line measured in increments of approximate vehicle length to reduce the
chance that vehicles will be queued across the crosswalk.
• The pedestrian refuge should be designed at street level, rather than elevated
to the height of the splitter island. This eliminates the need for ramps within the
refuge area, which can be cumbersome for wheelchairs.
• Ramps should be provided on each end of the crosswalk to connect the cross-
walk to other crosswalks around the roundabout and to the sidewalk network.
• It is recommended that a detectable warning surface, as recommended in the
Americans with Disabilities Act Accessibility Guidelines (ADAAG) §4.29 (De-
tectable Warnings), be applied to the surface of the refuge within the splitter
Detectable warning surfaces
island as shown in Exhibit 6-26. Note that the specific provision of the ADAAG
should be applied within the
requiring detectable warning surface at locations such as ramps and splitter
pedestrian refuge.
islands (defined in the ADAAG as “hazardous vehicle areas”) has been sus-
pended until July 26, 2001 (ADAAG §4.29.5). Where used, a detectable warning
surface shall meet the following requirements (7):
- The detectable warning surface shall contrast visually with adjoining sur-
faces, either light-on-dark or dark-on-light. The material used to provide
contrast shall be an integral part of the walking surface.
- The detectable warning surface shall begin at the curb line and extend
into the pedestrian refuge area a distance of 600 mm (24 in). This creates
a minimum 600-mm (24-in) clear space between detectable warning sur-
faces for a minimum splitter island width of 1.8 m (6 ft) at the pedestrian
crossing. This is a deviation from the requirements of (suspended) ADAAG
§4.29.5, which requires a 915-mm (36-in) surface width. However, this
deviation is necessary to enable visually impaired pedestrians to distin-
guish the two interfaces with vehicular traffic.
In urban areas, speed tables (flat-top road humps) could be considered for wheel-
chair users, provided that good geometric design has reduced absolute vehicle
speeds to less than 20 km/h (12 mph) near the crossing. Pedestrian crossings
across speed tables must have detectable warning material as described above to
clearly delineate the edge of the street. Speed tables should generally be used
only on streets with approach speeds of 55 km/h (35 mph) or less, as the introduc-
tion of a raised speed table in higher speed environments may increase the likeli-
hood of single-vehicle crashes and is not consistent with the speed consistency
philosophy presented in this document.
Splitter islands (also called separator islands or median islands) should be provided Splitter islands perform
on all roundabouts, except those with very small diameters at which the splitter multiple functions and should
island would obstruct the visibility of the central island. Their purpose is to provide generally be provided.
shelter for pedestrians (including wheelchairs, bicycles, and baby strollers), assist
in controlling speeds, guide traffic into the roundabout, physically separate enter-
ing and exiting traffic streams, and deter wrong-way movements. Additionally, splitter
islands can be used as a place for mounting signs (see Chapter 7).
The splitter island envelope is formed by the entry and exit curves on a leg, as
shown previously in Exhibits 6-24 and 6-25. The total length of the island should
generally be at least 15 m (50 ft) to provide sufficient protection for pedestrians and
to alert approaching drivers to the roundabout geometry. Additionally, the splitter
island should extend beyond the end of the exit curve to prevent exiting traffic from
accidentally crossing into the path of approaching traffic.
Exhibit 6-26 shows the minimum dimensions for a splitter island at a single-
lane roundabout, including the location of the pedestrian crossing as discussed
in Section 6.3.7.
Larger splitter islands enhance While Exhibit 6-26 provides minimum dimensions for splitter islands, there are
safety, but require that the benefits to providing larger islands. Increasing the splitter island width results in
inscribed circle diameter be greater separation between the entering and exiting traffic streams of the same
increased. leg and increases the time for approaching drivers to distinguish between exiting
and circulating vehicles. In this way, larger splitter islands can help reduce confu-
sion for entering motorists. A recent study by the Queensland Department of Main
Roads found that maximizing the width of splitter islands has a significant effect on
minimizing entering/circulating vehicle crash rates (2). However, increasing the width
of the splitter islands generally requires increasing the inscribed circle diameter.
Thus, these safety benefits may be offset by higher construction cost and greater
land impacts.
Standard AASHTO guidelines for island design should be followed for the splitter
island. This includes using larger nose radii at approach corners to maximize island
visibility and offsetting curb lines at the approach ends to create a funneling effect.
The funneling treatment also aids in reducing speeds as vehicles approach the
roundabout. Exhibit 6-27 shows minimum splitter island nose radii and offset di-
mensions from the entry and exit traveled ways.
Stopping sight distance is the distance along a roadway required for a driver to
perceive and react to an object in the roadway and to brake to a complete stop
before reaching that object. Stopping sight distance should be provided at every
point within a roundabout and on each entering and exiting approach.
V2
d = (0.278)(t )(V ) + 0.039 (6-2a, metric)
a
where: d = stopping sight distance, m;
t = perception-brake reaction time, assumed to be 2.5 s;
V = initial speed, km/h; and
a = driver deceleration, assumed to be 3.4 m/s2.
V2
d = (1.468)(t )(V ) + 1.087
a
where: d = stopping sight distance, ft;
t = perception-brake reaction time, assumed to be 2.5 s;
V = initial speed, mph; and
a = driver deceleration, assumed to be 11.2 ft/s2.
Exhibit 6-28 gives recommended stopping sight distances for design, as computed
from the above equations.
10 8.1 10 46.4
20 18.5 15 77.0
30 31.2 20 112.4
40 46.2 25 152.7
50 63.4 30 197.8
60 83.0 35 247.8
70 104.9 40 302.7
80 129.0 45 362.
90 155.5 50 427.2
100 184.2 * 55 496.7
Assumes 2.5 s perception-braking time, 3.4 m/s2 (11.2 ft/s2) driver deceleration
At least three critical types of At roundabouts, three critical types of locations should be checked at a minimum:
locations should be checked for • Approach sight distance (Exhibit 6-29);
stopping sight distance.
• Sight distance on circulatory roadway (Exhibit 6-30); and
• Sight distance to crosswalk on exit (Exhibit 6-31).
Forward sight distance at entry can also be checked; however, this will typically be
satisfied by providing adequate stopping sight distance on the circulatory roadway
itself.
Intersection sight distance is the distance required for a driver without the right of Roundabout entries require
way to perceive and react to the presence of conflicting vehicles. Intersection sight adequate intersection sight
distance is achieved through the establishment of adequate sight lines that allow a distance.
driver to see and safely react to potentially conflicting vehicles. At roundabouts,
the only locations requiring evaluation of intersection sight distance are the en-
tries.
Exhibit 6-32 presents a diagram showing the method for determining intersection
sight distance. As can be seen in the exhibit, the sight distance “triangle” has two
conflicting approaches that must be checked independently. The following two
subsections discuss the calculation of the length of each of the approaching sight
limits.
The critical gap for entering the major road is based on the amount of time required
for a vehicle to turn right while requiring the conflicting stream vehicle to slow no
less than 70 percent of initial speed. This is based on research on critical gaps at
stop-controlled intersections, adjusted for yield-controlled conditions (9). The criti-
cal gap value of 6.5 s given in Equation 6-3 is based on the critical gap required for
passenger cars, which are assumed to be the most critical design vehicle for inter-
section sight distance. This assumption holds true for single-unit and combination
truck speeds that are at least 10 km/h (6 mph) and 15 to 20 km/h (9 to 12 mph)
slower than passenger cars, respectively.
20 36.1 10 95.4
25 45.2 15 143.0
30 54.2 20 190.1
35 63.2 25 238.6
40 72.3 30 286.3
In general, it is recommended to provide no more than the minimum required Providing more than the
intersection sight distance on each approach. Excessive intersection sight distance minimum required intersection
can lead to higher vehicle speeds that reduce the safety of the intersection for all sight distance can lead to
road users (vehicles, bicycles, pedestrians). Landscaping can be effective in re- higher speeds that reduce
stricting sight distance to the minimum requirements. intersection safety.
Note that the stopping sight distance on the circulatory roadway (Exhibit 6-30) and
the intersection sight distance to the circulating stream (Exhibit 6-32) imply restric-
tions on the height of the central island, including landscaping and other objects,
within these zones. In the remaining central area of the central island, higher land-
scaping may serve to break the forward vista for through vehicles, thereby contrib-
uting to speed reduction. However, should errant vehicles encroach on the central
island, Chapter 7 provides recommended maximum grades on the central island to
minimize the probability of the vehicles rolling over, causing serious injury.
[Link] Profiles
The vertical design of a roundabout begins with the development of approach road-
way and central island profiles. The development of each profile is an iterative pro-
cess that involves tying the elevations of the approach roadway profiles into a
smooth profile around the central island.
Generally, each approach profile should be designed to the point where the ap-
proach baseline intersects with the central island. A profile for the central island is
then developed which passes through these four points (in the case of a four-
legged roundabout). The approach roadway profiles are then readjusted as neces-
sary to meet the central island profile. The shape of the central island profile is
generally in the form of a sine curve. Examples of how the profile is developed can
be found in Exhibits 6-34, 6-35, and 6-36, which consist of a sample plan, profiles
on each approach, and a profile along the central island, respectively. Note that the
four points where the approach roadway baseline intersects the central island
baseline are identified on the central island profile.
[Link] Superelevation
Negative superelevation (- 2%) As a general practice, a cross slope of 2 percent away from the central island
should generally be used for the should be used for the circulatory roadway. This technique of sloping outward is
circulatory roadway. recommended for four main reasons:
• It promotes safety by raising the elevation of the central island and improving its
visibility;
• It minimizes breaks in the cross slopes of the entrance and exit lanes; and
• It helps drain surface water to the outside of the roundabout (2, 6).
The outward cross slope design means vehicles making through and left-turn move-
ments must negotiate the roundabout at negative superelevation. Excessive nega-
tive superelevation can result in an increase in single-vehicle crashes and loss-of-
load incidents for trucks, particularly if speeds are high. However, in the intersec-
tion environment, drivers will generally expect to travel at slower speeds and will
accept the higher side force caused by reasonable adverse superelevation (10).
Exhibit 6-37 provides a typical section across the circulatory roadway of a round-
about without a truck apron. Exhibit 6-38 provides a typical section for a round-
about with a truck apron. Where truck aprons are used, the slope of the apron
should be 3 to 4 percent; greater slopes may increase the likelihood of loss-of-load
incidents.
[Link] Drainage
With the circulatory roadway sloping away from the central island, inlets will
generally be placed on the outer curbline of the roundabout. However, inlets may
be required along the central island for a roundabout designed on a constant
grade through an intersection. As with any intersection, care should be taken to
ensure that low points and inlets are not placed in crosswalks. If the central
island is large enough, the designer may consider placing inlets in the central
island.
With regard to bicycle treatments, the designer should strive to provide bicy-
clists the choice of proceeding through the roundabout as either a vehicle or a
pedestrian. In general, bicyclists are better served by treating them as vehicles.
However, the best design provides both options to allow cyclists of varying de-
grees of skill to choose their more comfortable method of navigating the round-
about.
To accommodate bicyclists traveling as vehicles, bike lanes should be terminated Terminate bicycle lanes prior to
in advance of the roundabout to encourage cyclists to mix with vehicle traffic. a roundabout.
Under this treatment, it is recommended that bike lanes end 30 m (100 ft) up-
stream of the yield line to allow for merging with vehicles (11). This method is
most successful at smaller roundabouts with speeds below 30 km/h (20 mph),
where bicycle speeds can more closely match vehicle speeds.
To accommodate bicyclists who prefer not to use the circulatory roadway, a wid-
ened sidewalk or a shared bicycle/pedestrian path may be provided physically
separated from the circulatory roadway (not as a bike lane within the circulatory
Ramps leading to a shared roadway). Ramps or other suitable connections can then be provided between this
pathway can be used to sidewalk or path and the bike lanes, shoulders, or road surface on the approaching
accommodate bicyclists and departing roadways. The designer should exercise care in locating and design-
traveling as pedestrians. ing the bicycle ramps so that they are not misconstrued by pedestrians as an un-
marked pedestrian crossing. Nor should the exits from the roadway onto a shared
path allow cyclists to enter the shared path at excessive speeds. Exhibit 6-39 illus-
trates a possible design of this treatment. The reader is encouraged to refer to the
AASHTO Guide for Development of Bicycle Facilities (12) for a more detailed dis-
cussion of the design requirements for bicycle and shared-use path design.
Set back sidewalks 1.5 m (5 ft) Where possible, sidewalks should be set back from the edge of the circulatory
from the circulatory roadway roadway in order to discourage pedestrians from crossing to the central island,
where possible. particularly when an apron is present or a monument on the central island. Equally
important, the design should help pedestrians with visual impairments to recog-
nize that they should not attempt to cross streets from corner to corner but at
designated crossing points. To achieve these goals, the sidewalk should be de-
signed so that pedestrians will be able to clearly find the intended path to the
crosswalks. A recommended set back distance of 1.5 m (5 ft) (minimum 0.6 m [2
ft]) should be used, and the area between the sidewalk and curb can be planted
with low shrubs or grass (see Chapter 7). Exhibit 6-40 shows this technique.
For safety and operational reasons, bus stops should be located as far away from
entries and exits as possible, and never in the circulatory roadway.
• Far-side stops: Bus stops on the far side of a roundabout should be constructed
with pull-outs to minimize queuing into the roundabout. These stops should
be located beyond the pedestrian crossing to improve visibility of pedestrians
to other exiting vehicles.
Right-turn bypass lanes can be In general, right-turn bypass lanes (or right-turn slip lanes) should be avoided, espe-
used in locations with minimal cially in urban areas with bicycle and pedestrian activity. The entries and exits of
pedestrian and bicycle activity bypass lanes can increase conflicts with bicyclists. The generally higher speeds of
to improve capacity when heavy bypass lanes and the lower expectation of drivers to stop increases the risk of
right-turning traffic exists. collisions with pedestrians. However, in locations with minimal pedestrian and bi-
cycle activity, right-turn bypass lanes can be used to improve capacity where there
is heavy right turning traffic.
The provision of a right-turn bypass lane allows right-turning traffic to bypass the
roundabout, providing additional capacity for the through and left-turn movements
at the approach. They are most beneficial when the demand of an approach ex-
ceeds its capacity and a significant proportion of the traffic is turning right. How-
ever, it is important to consider the reversal of traffic patterns during the opposite
peak time period. In some cases, the use of a right-turn bypass lane can avoid the
need to build an additional entry lane and thus a larger roundabout. To determine if
a right-turn bypass lane should be used, the capacity and delay calculations in
Chapter 4 should be performed. Right-turn bypass lanes can also be used in loca-
tions where the geometry for right turns is too tight to allow trucks to turn within
the roundabout.
There are two design options for right-turn bypass lanes. The first option, shown in Right-turn bypass lanes can
Exhibit 6-42, is to carry the bypass lane parallel to the adjacent exit roadway, and merge back into the main exit
then merge it into the main exit lane. Under this option, the bypass lane should be roadway or provide a yield-
carried alongside the main roadway for a sufficient distance to allow vehicles in the controlled entrance onto the
bypass lane and vehicles exiting the roundabout to accelerate to comparable speeds. main exit roadway.
The bypass lane is then merged at a taper rate according to AASHTO guidelines for
the appropriate design speed. The second design option for a right-turn bypass
lane, shown in Exhibit 6-43, is to provide a yield-controlled entrance onto the adja-
cent exit roadway. The first option provides better operational performance than
the second does. However, the second option generally requires less construction
and right-of-way than the first.
The option of providing yield control on a bypass lane is generally better for both
bicyclists and pedestrians and is recommended as the preferred option in urban
areas where pedestrians and bicyclists are prevalent. Acceleration lanes can be
problematic for bicyclists because they end up being to the left of accelerating
motor vehicles. In addition, yield control at the end of a bypass lane tends to slow
motorists down, whereas an acceleration lane at the end of a bypass lane tends to
promote higher speeds.
The radius of the right-turn bypass lane should not be significantly larger than the
radius of the fastest entry path provided at the roundabout. This will ensure vehicle
speeds on the bypass lane are similar to speeds through the roundabout, resulting
in safe merging of the two roadways. Providing a small radius also provides greater
safety for pedestrians who must cross the right-turn slip lane.
Exhibit 6-42. Configuration
of right-turn bypass lane with
acceleration lane.
As discussed in Section 6.2.1, the fastest path through the roundabout is drawn to
ensure the geometry imposes sufficient curvature to achieve a safe design speed.
This path is drawn assuming the roundabout is vacant of all other traffic and the
vehicle cuts across adjacent travel lanes, ignoring all lane markings. In addition to
evaluating the fastest path, at double-lane roundabouts the designer must also
evaluate the natural vehicle paths. This is the path an approaching vehicle will natu-
rally take, assuming there is traffic in all approach lanes, through the roundabout
geometry.
As two traffic streams approach the roundabout in adjacent lanes, they will be
forced to stay in their lanes up to the yield line. At the yield point, vehicles will
continue along their natural trajectory into the circulatory roadway, then curve around
the central island, and curve again into the opposite exit roadway. The speed and
orientation of the vehicle at the yield line determines its natural path. If the natural
path of one lane interferes or overlaps with the natural path of the adjacent lane,
the roundabout will not operate as safely or efficiently as possible.
The key principle in drawing the natural path is to remember that drivers cannot
change the direction of their vehicle instantaneously. Neither can they change their
speed instantaneously. This means that the natural path does not have sudden
changes in curvature; it has transitions between tangents and curves and between
consecutive reversing curves. Secondly, it means that consecutive curves should
be of similar radius. If a second curve has a significantly smaller radius than the
first curve, the driver will be traveling too fast to negotiate the turn and may lose
control of the vehicle. If the radius of one curve is drawn significantly smaller than
the radius of the previous curve, the path should be adjusted.
To identify the natural path of a given design, it may be advisable to sketch the
natural paths over the geometric layout, rather than use a computer drafting program
or manual drafting equipment. In sketching the path, the designer will naturally draw
transitions between consecutive curves and tangents, similar to the way a driver
would negotiate an automobile. Freehand sketching also enables the designer to feel
how changes in one curve affect the radius and orientation of the next curve. In
general, the sketch technique allows the designer to quickly obtain a smooth, natural
path through the geometry that may be more difficult to obtain using a computer.
Exhibit 6-44 illustrates a sketched natural path of a vehicle through a typical double-
lane roundabout.
Vehicle path overlap occurs when the natural path through the roundabout of one
traffic stream overlaps the path of another. This can happen to varying degrees. It can
reduce capacity, as vehicles will avoid using one or more of the entry lanes. It can
also create safety problems, as the potential for sideswipe and single-vehicle crashes
is increased. The most common type of path overlap is where vehicles in the left lane
on entry are cut off by vehicles in the right lane, as shown in Exhibit 6-45.
One method to avoid path overlap on entry is to start with an inner entry curve that
is curvilinearly tangential to the central island and then draw parallel alignments to
determine the position of the outside edge of each entry lane. These curves can
range from 30 to 60 m (100 to 200 ft) in urban environments and 40 to 80 m (130 to
260 ft) in rural environments. These curves should extend approximately 30 m (100
ft) to provide clear indication of the curvature to the driver. The designer should
check the critical vehicle paths to ensure that speeds are sufficiently low and con-
sistent between vehicle streams. The designer should also ensure that the portion
of the splitter island in front of the crosswalk meets AASHTO recommendations
for minimum size. Exhibit 6-46 demonstrates this method of design.
Another method to reduce entry speeds and avoid path overlap is to use a small-
radius (generally 15 to 30 m [50 to 100 ft]) curve approximately 10 to 15 m (30 to 50
ft) upstream of the yield line. A second, larger-radius curve (or even a tangent) is
then fitted between the first curve and the edge of the circulatory roadway. In this
way, vehicles will still be slowed by the small-radius approach curve, and they will
be directed along a path that is tangential to the central island at the time they
reach the yield line. Exhibit 6-47 demonstrates this alternate method of design.
The entry path radius, R1 , is controlled by the offset between the right curb line on
the entry roadway and the curb line of the central island (on the driver’s left). If the
initial layout produces an entry path radius above the preferred design speed, one
way to reduce it is to gradually shift the approach to the left to increase the offset;
however, this may increased adjacent exit speeds. Another method to reduce the
entry path radius is to move the initial, small-radius entry curve closer to the circu-
latory roadway. This will decrease the length of the second, larger-radius curve and
increase the deflection for entering traffic. However, care must be taken to ensure
this adjustment does not produce overlapping natural paths.
To achieve exit speeds slower than 40 km/h (25 mph), as is often desirable in envi-
ronments with significant pedestrian activity, it may be necessary to tighten the
exit radius. This may improve safety for pedestrians at the possible expense of
increased vehicle-vehicle collisions.
Roundabouts located on rural roads often have special design considerations be-
cause approach speeds are higher than urban or local streets and drivers generally
do not expect to encounter speed interruptions. The primary safety concern in rural
locations is to make drivers aware of the roundabout with ample distance to com-
fortably decelerate to the appropriate speed. This section provides design guide-
lines for providing additional speed-reduction measures on rural roundabout ap-
proaches.
6.5.1 Visibility
Perhaps the most important element affecting safety at rural intersections is the Roundabout visibility is a key
visibility of the intersection itself. Roundabouts are no different from stop-controlled design element at rural
or signalized intersections in this respect except for the presence of curbing along locations.
roadways that are typically not curbed. Therefore, although the number and sever-
ity of multiple-vehicle collisions at roundabouts may decrease (as discussed previ-
ously), the number of single-vehicle crashes may increase. This potential can be
minimized with attention to proper visibility of the roundabout and its approaches.
6.5.2 Curbing
On an open rural highway, changes in the roadway’s cross-section can be an effec- Curbs should be provided at all
tive means to help approaching drivers recognize the need to reduce their speed. rural roundabouts.
Rural highways typically have no outside curbs with wide paved or gravel shoul-
ders. Narrow shoulder widths and curbs on the outside edges of pavement, on the
other hand, generally give drivers a sense they are entering a more urbanized set-
ting, causing them to naturally slow down. Thus, consideration should be given to
reducing shoulder widths and introducing curbs when installing a roundabout on
an open rural highway.
Curbs help to improve delineation and to prevent “corner cutting,” which helps to
ensure low speeds. In this way, curbs help to confine vehicles to the intended
design path. The designer should carefully consider all likely design vehicles, in-
cluding farm equipment, when setting curb locations. Little research has been per-
formed to date regarding the length of curbing required in advance of a rural round-
about. In general, it may be desirable to extend the curbing from the approach for
at least the length of the required deceleration distance to the roundabout.
Another effective cross-section treatment to reduce approach speeds is to use Extended splitter islands are
longer splitter islands on the approaches (10). Splitter islands should generally be recommended at rural
extended upstream of the yield bar to the point at which entering drivers are ex- locations.
pected to begin decelerating comfortably. A minimum length of 60 m (200 ft) is
recommended (10). Exhibit 6-48 provides a diagram of such a splitter island design.
The length of the splitter island may differ depending upon the approach speed.
The AASHTO recommendations for required braking distance with an alert driver
should be applied to determine the ideal splitter island length for rural roundabout
approaches.
The radius of an approach curve (and subsequent vehicular speeds) has a direct
impact on the frequency of crashes at a roundabout. A study in Queensland, Aus-
tralia, has shown that decreasing the radius of an approach curve generally de-
creases the approaching rear-end vehicle crash rate and the entering-circulating
and exiting-circulating vehicle crash rates (see Chapter 5). On the other hand, de-
creasing the radius of an approach curve may increase the single-vehicle crash rate
on the curve, particularly when the required side-friction for the vehicle to maintain
its path is too high. This may encourage drivers to cut across lanes and increase
sideswipe crash rates on the approach curve (2).
One method to achieve speed reduction that reduces crashes at the roundabout
while minimizing single-vehicle crashes is the use of successive curves on ap-
proaches. The study in Queensland, Australia, found that by limiting the change in
85th-percentile speed on successive geometric elements to 20 km/h (12 mph), the
crash rate was reduced. It was found that the use of successive reverse curves
prior to the roundabout approach curve reduced the single-vehicle crash rate and
the sideswipe crash rate on the approach. It is recommended that approach speeds
immediately prior to the entry curves of the roundabout be limited to 60 km/h (37
mph) to minimize high-speed rear-end and entering-circulating vehicle crashes.
Exhibit 6-49 shows a typical rural roundabout design with a succession of three A series of progressively sharper
curves prior to the yield line. As shown in the exhibit, these approach curves should curves on a high-speed
be successively smaller radii in order to minimize the reduction in design speed roundabout approach helps
between successive curves. The aforementioned Queensland study found that slow traffic to an appropriate
shifting the approaching roadway laterally by 7 m (23 ft) usually enables adequate entry speed.
curvature to be obtained while keeping the curve lengths to a minimum. If the
lateral shift is too small, drivers are more likely to cut into the adjacent lane (2).
Equations 6-4 and 6-5 can be used to estimate the operating speed of two-lane
rural roads as a function of degree of curvature. Equation 6-6 can be used similarly
for four-lane rural roads (13).
6.6 Mini-Roundabouts
The central island of a The design of the central island of a mini-roundabout is defined primarily by the
mini-roundabout should be requirement to achieve speed reduction for passenger cars. As discussed previ-
clear and conspicuous. ously in Section 6.2, speed reduction for entering vehicles and speed consistency
with circulating vehicles are important. Therefore, the location and size of the cen-
tral island are dictated by the inside of the swept paths of passenger cars that is
needed to achieve a maximum recommended entry speed of 25 km/h (15 mph).
The central island of a mini-roundabout is typically a minimum of 4 m (13 ft) in
diameter and is fully mountable by large trucks and buses. Composed of asphalt,
concrete, or other paving material, the central island should be domed at a height
of 25 to 30 mm per 1 m diameter (0.3 to 0.36 in per 1 ft diameter), with a maximum
height of 125 mm (5 in) (14). Although fully mountable and relatively small, it is
essential that the central island be clear and conspicuous (14, 15). Chapter 7 pro-
vides a sample signing and striping planing plan for mini-roundabout.
The outer swept path of passenger cars and large vehicles is typically used to
define the location of the yield line and boundary of each splitter island with the
circulatory roadway. Given the small size of a mini-roundabout, the outer swept
path of large vehicles may not be coincident with the inscribed circle of the round-
about, which is defined by the outer curbs. Therefore, the splitter islands and yield
line may extend into the inscribed circle for some approach geometries. On the
other hand, for very small mini-roundabouts, such as the one shown in Exhibit 6-
50, all turning trucks will pass directly over the central island while not encroaching
on the circulating roadway to the left which may have opposing traffic. In these
cases, the yield line and splitter island should be set coincident with the inscribed
6.7 References
7 7.1 Signing
7.1.1
Traffic Design
and Landscaping
Devices 185
7.1.2 Regulatory signs 186
7.1.3 Warning signs 188
7.1.4 Guide signs 191
7.1.5 Urban signing considerations 192
7.1.6 Rural and suburban signing considerations 194
7.1.7 Mini-roundabout signing considerations 195
7.2 Pavement Markings 197
7.2.1 Relationship with the Manual on Uniform Traffic Control
Devices 197
7.2.2 Approach and entry pavement markings 197
7.2.3 Circulatory roadway pavement markings 200
7.2.4 Mini-roundabout pavement markings 201
7.3 Illumination 202
7.3.1 Need for illumination 202
7.3.2 Standards and recommended practices 203
7.3.3 General recommendations 204
7.3.4 Clear zone requirements 205
7.4 Work Zone Traffic Control 205
7.4.1 Pavement markings 205
7.4.2 Signing 205
7.4.3 Lighting 206
This chapter presents guidelines on the design of traffic elements, illumination, Signing, striping, illumination,
and landscaping associated with roundabouts. The design of these elements is and landscaping are the critical
critical in achieving the desired operational and safety features of a roundabout, as finishing touches for an
well as the desired visibility and aesthetics. This chapter is divided into the follow- effectively functioning
ing sections: roundabout.
• Signing;
• Pavement Markings;
• Illumination;
• Work Zone Traffic Control; and
• Landscaping.
7.1 Signing
The overall concept for roundabout signing is similar to general intersection sign-
ing. Proper regulatory control, advance warning, and directional guidance are re-
quired to avoid driver expectancy related problems. Signs should be located where
they have maximum visibility for road users but a minimal likelihood of even mo-
mentarily obscuring pedestrians as well as motorcyclists and bicyclists, who are
the most vulnerable of all roundabout users. Signing needs are different for urban
and rural applications and for different categories of roundabouts.
The Manual on Uniform Traffic Control Devices for Streets and Highways (MUTCD)
(1) and Standard Highway Signs (2), as well as local applicable standards, govern
the design and placement of signs. To the extent possible, this guide has been
prepared in accordance with the 1988 edition of the MUTCD. However, round-
abouts present a number of new signing issues that are not addressed in the 1988
edition. For this reason, a number of new signs or uses for existing signs have
been introduced that are under consideration for inclusion in the next edition of the
MUTCD. Until such signs or uses are formally adopted, these recommendations
should be considered provisional and are subject to MUTCD Section 1A-6, “Manual
Changes, Interpretations and Authority to Experiment.”
The following signs and applications recommended below are subject to these
conditions:
• Use of YIELD signs on more than one approach to an intersection (Section [Link]);
• Long chevron plate (Section [Link]);
• Roundabout Ahead sign (Section [Link]);
• Advance diagrammatic guide signs (Section [Link]); and
• Exit guide signs (Section [Link]).
A number of regulatory signs are appropriate for roundabouts and are described
below.
For small splitter islands, a Type 1 object marker may be substituted for the KEEP
RIGHT sign. This may reduce sign clutter and improve the visibility of the YIELD
sign.
Exhibit 7-3. KEEP RIGHT sign
(R4-7).
• Exit lane requirements. In general, the number of exit lanes provided should be
the minimum required to handle the expected exit volume. This may not corre-
spond with the number of entry lanes on the opposite side of the roundabout
that would use the exit as through vehicles (see Chapter 4).
• The rules of the road. Drivers have a reasonable expectation that multiple through
lanes entering a roundabout will have an equal number of receiving lanes on
exit on the far side of the roundabout (see Chapter 2).
Lane-use control signs are generally not required where the number of receiving
lanes for through vehicles on exit matches the number of entry lanes, as shown in
Exhibit 7-4. Lane-use control signs should be used only for the following condi-
tions:
• Where only a single exit lane is provided to receive two lanes of vehicles mak-
ing through movements, lane-use designations should be made to indicate that
an entry lane drops as a turning movement (see Exhibit 7-4). This does not in-
clude cases where an approach is flared from one to two lanes at the round-
about.
• Where left- or right-turning traffic demand dictates the need for more than one
left-turn lane or more than one right-turn lane for capacity reasons (see Exhibit 7-5).
The use of a left-turn-only lane designation as shown in the exhibits may be initially
confusing to drivers. This type of designation has worked successfully in other
countries, and there is no evidence to suggest that it will not work in the United
States. However, given the general unfamiliarity of roundabouts to drivers in the
United States at this time, it is recommended that double-lane roundabouts be
designed to avoid the use of lane-use control signs wherever possible, at least until
drivers become more accustomed to driving roundabouts.
A number of warning signs are appropriate for roundabouts and are described be-
low. The amount of warning a motorist needs is related to the intersection setting
and the vehicular speeds on approach roadways. The specific placement of warn-
ing signs is governed by the applicable sections of the MUTCD.
It is also recommended that an advisory speed plate (W13-1) be used with this
sign, as shown in Exhibit 7-7. The speed given on the advisory speed plate should
be no higher than the design speed of the circulatory roadway, as determined in
Chapter 6.
An alternative to the Circular Intersection sign, called a Roundabout Ahead sign, has
been proposed and is shown in Exhibit 7-8. The rationale for this sign is given in
Appendix C. At a minimum it is recommended that the Roundabout Ahead sign be
used in place of the Circular Intersection sign at mini-roundabouts (see Section 7.1.7).
The use of Pedestrian Crossing signs is dependent on the specific laws of the
governing state. If the crosswalk at a roundabout is not considered to be part of
the intersection and is instead considered a marked midblock crossing, Pedestrian
Crossing signs are required. Where installed, Pedestrian Crossing signs should be
located in such a way to not obstruct view of the YIELD sign.
Guide signs are important in providing drivers with proper navigational informa-
tion. This is especially true at roundabouts where out-of-direction travel may disori-
ent unfamiliar drivers. A number of guide signs are appropriate for roundabouts
and are described below.
Leeds, MD Taneytown, MD
The designer needs to balance the The amount of signing required at individual locations is largely based on engineer-
need for adequate signing ing judgment. However, in practice, the designer can usually use fewer and smaller
with the tendency to use signs in urban settings than in rural settings. This is true because drivers are gener-
too many signs. ally traveling at lower vehicular speeds and have higher levels of familiarity at urban
intersections. Therefore, in many urban settings the advance destination guide signs
can be eliminated. However, some indication of street names should be included
in the form of exit guide signs or standard street name signs. Another consider-
ation in urban settings is the use of minimum amounts of signing to avoid sign
clutter. A sample signing plan for an urban application is shown in Exhibit 7-15.
Rural signing needs to be more Rural and suburban conditions are characterized by higher approach speeds. Route
conspicuous than urban signing guidance tends to be focused more on destinations and numbered routes rather
due to higher approach speeds. than street names. A sample signing plan for a rural application is shown in Exhibit
7-16.
In cases where high speeds are expected (in excess of 80 km/h [50 mph]) and the These speed reduction treatments
normal signage and geometric features are not expected to produce the desired can apply to all intersection
reduction in vehicle speeds, the following measures may also be considered (ex- types, not just roundabouts.
amples of some of these treatments are given in Exhibit 7-17):
Due to their small size and unique features, mini-roundabouts require a somewhat
different signing treatment than the larger urban roundabouts. The principal differ-
ences in signing at mini-roundabouts as compared to other urban roundabouts are
the following:
• The central island is fully mountable. Therefore, no ONE WAY signs, Large Ar-
row signs, or chevrons can be located there. It is recommended that the direc-
tion of circulation be positively indicated through the use of pavement mark-
ings, as discussed in Section 7.2.4.
• The splitter islands are either painted or are fully mountable. Therefore, KEEP
RIGHT signs are not appropriate for mini-roundabouts.
• Typically, advance directional guide signs and exit guide signs are unnecessary,
given the size of the mini-roundabout and the nature of the approach roadways
(generally low-speed local streets). However, standard street name signs (D3)
should be used.
Typical pavement markings for roundabouts consist of delineating the entries and
the circulatory roadway.
As with signing, the MUTCD (1) and applicable local standards govern the design
and placement of pavement markings. Roundabouts present a number of new
pavement marking issues that are not addressed in the 1988 edition of the MUTCD.
For this reason, a number of new pavement markings or uses for existing pave-
ment markings have been introduced that are under consideration for inclusion in
the next edition of the MUTCD. Until such pavement markings or uses are formally
adopted, these recommendations should be considered provisional and are sub-
ject to MUTCD Section 1A-6, “Manual Changes, Interpretations and Authority to
Experiment.”
The following pavement markings and applications recommended below are sub-
ject to these conditions:
Approach and entry pavement markings consist of yield lines, pavement word and
symbol markings, and channelization markings. In addition, multilane approaches
require special attention to pavement markings. The following sections discuss
these in more detail.
The MUTCD currently provides no standard for yield lines. The recommended yield
line pavement marking is a broken line treatment consisting of 400-mm (16-in)
wide stripes with 1-m (3-ft) segments and 1-m (3-ft) gaps. This type of yield line is
the simplest to install.
Alternatively, several European countries use a yield line marking consisting of a “Shark’s teeth” provide more visual
series of white triangles (known as “shark’s teeth”). These markings tend to be “punch” but require a new template
more visible to approaching drivers. Exhibit 7-19 presents examples of broken line for installation.
and “shark’s teeth” yield line applications. The “shark’s teeth” ahead of the broken
line has been recommended for adoption in the next edition of the MUTCD.
Alternatively, some European countries paint a symbolic yield sign upstream of the
yield line. This treatment has the advantage of being symbolic; however, such a
treatment has not seen widespread use in the United States to date.
Raised pavement markers are Raised pavement markers are generally recommended for supplementing pavement
useful supplements to markings. These have the benefit of additional visibility at night and in inclement
pavement markings. weather. However, they increase maintenance costs and can be troublesome in
areas requiring frequent snow removal. In addition, raised pavement markers should
not be used in the path of travel of bicycles.
For small splitter islands (in area less than 7 m2 [75 ft2), the island may consist of
pavement markings only. However, where possible, curbed splitter islands should
be used.
A crosswalk marking using a series of lines parallel to the flow of traffic (known as
a “zebra crosswalk”) is recommended. These lines should be approximately 0.3 m
to 0.6 m (12 in to 24 in) wide, spaced 0.3 m to 1.0 m (12 in to 36 in) apart, and span
the width of the crosswalk (similar to the recommendations in MUTCD §3B-18).
Crosswalk markings should be installed across both the entrance and exit of each
leg and across any right turn bypass lanes. The crosswalk should be aligned with
the ramps and pedestrian refuge in the splitter island and have markings that are
generally perpendicular to the flow of vehicular traffic.
The zebra crosswalk has a number of advantages over the traditional transverse
crosswalk marking in roundabout applications:
• Because the crosswalk at a roundabout is set back from the yield line, the zebra
crosswalk provides a higher degree of visibility.
• The zebra crosswalk is also less likely to be confused with the yield line than a
transverse crosswalk.
• Although the initial cost is somewhat higher, the zebra crossing may require
less maintenance due to the ability to space the markings to avoid vehicle tire
tracks.
Circulatory pavement markings In general, lane lines should not be striped within the circulatory roadway, regard-
are generally not recommended. less of the width of the circulatory roadway. Circulatory lane lines can be mislead-
ing in that they may provide drivers a false sense of security.
Bike lanes within the In addition, bike lane markings within the circulatory roadway are not recommended.
roundabout are The additional width of a bike lane within the circulatory roadway increases vehicu-
not recommended. lar speed and increases the probability of motor vehicle-cyclist crashes. Bicyclists
should circulate with other vehicles, travel through the roundabout as a pedestrian
on the sidewalk, or use a separate shared-use pedestrian and bicycle facility where
provided.
• At a minimum, the edges of the mountable central island and splitter islands
should be painted to improve their visibility.
7.3 Illumination
For a roundabout to operate satisfactorily, a driver must be able to enter the round-
about, move through the circulating traffic, and separate from the circulating stream
in a safe and efficient manner. To accomplish this, a driver must be able to perceive
the general layout and operation of the intersection in time to make the appropriate
maneuvers. Adequate lighting should therefore be provided at all roundabouts.
Exhibit 7-22 shows an example of an illuminated roundabout at night.
Loveland, CO
The need for illumination varies somewhat based on the location in which the round-
about is located.
• AASHTO, An Information Guide for Roadway Lighting (6). This is the basic guide
for highway lighting. It includes information on warranting conditions and de-
sign criteria.
• IES RP-8: The American National Standard Practice for Roadway Lighting (8).
This Recommended Practice, published by the Illuminating Engineering Soci-
ety, provides standards for average-maintained illuminance, luminance, and small
target visibility, as well as uniformity of lighting. Recommended illumination
levels for streets with various classifications and in various areas are given in
Exhibit 7-23.
Definitions:
Commercial A business area of a municipality where ordinarily there are many pedestrians during
night hours. This definition applies to densely developed business areas outside, as well
as within, the central part of a municipality. The area contains land use which attracts a
relatively heavy volume of nighttime vehicular and/or pedestrian traffic on a frequent
basis.
Intermediate Those areas of a municipality often with moderately heavy nighttime pedestrian activity
such as in blocks having libraries, community recreation centers, large apartment build-
ings, industrial buildings, or neighborhood retail stores.
Residential A residential development, or a mixture of residential and small commercial establish-
ments, with few pedestrians at night.
Note: Values in table assume typical asphalt roadway surface (pavement classification R2 or R3). Consult
the IES document for other pavement surfaces.
Source: Illuminating Engineering Society RP-8 (8)
The primary goal of illumination is to ensure perception of the approach and mutual
visibility among the various categories of users. To achieve this, the following fea-
tures are recommended:
• Good illumination should be provided on the approach nose of the splitter is-
lands, at all conflict areas where traffic is entering the circulating stream, and at
all places where the traffic streams separate to exit the roundabout.
Lighting from the central island • It is preferable to light the roundabout from the outside in towards the center.
causes vehicles to be backlit and This improves the visibility of the central island and the visibility of circulating
thus less visible. vehicles to vehicles approaching to the roundabout. Ground-level lighting within
the central island that shines upwards towards objects in the central island can
improve their visibility.
The reader should refer to the AASHTO Roadside Design Guide for a more detailed
discussion of clear zone requirements (9).
During the construction of a roundabout it is essential that the intended travel path
be clearly identified. This may be accomplished through pavement markings, sign-
ing, delineation, channelizing devices, and guidance from police and/or construc-
tion personnel, depending on the size and complexity of the roundabout. Care
should be taken to minimize the channelizing devices so that the motorist, bicy-
clist, and pedestrian has a clear indication of the required travel path. Each installa-
tion should be evaluated separately, as a definitive guideline for the installation of
roundabouts is beyond the scope of this guide. Refer to Part 6 of the MUTCD for
requirements regarding work zone traffic control.
The pavement markings used in work zones should be the same layout and dimen-
sion as those used for the final installation. Because of the confusion of a work
area and the change in traffic patterns, additional pavement markings may be used
to clearly show the intended direction of travel. In some cases when pavement
markings cannot be placed, channelizing devices should be used to establish the
travel path.
7.4.2 Signing
Construction signing for a
The signing in work zones should consist of all necessary signing for the efficient roundabout should follow the
movement of traffic through the work area, preconstruction signing advising the pub- MUTCD standard.
lic of the planned construction, and any regulatory and warning signs necessary for
the movement of traffic outside of the immediate work area. The permanent round-
about signing should be installed where practicable during the first construction
stage so that it is available when the roundabout is operable. Permanent signing
that cannot be installed initially should be placed on temporary supports in the
proposed location until permanent installation can be completed.
7.4.3 Lighting
Permanent lighting, as described in Section 7.3, should be used to light the work
area. If lighting will not be used, pavement markings, as described in Section 7.2,
should be used.
Construction staging should be As is the case with any construction project, before any work can begin, all traffic
considered during the siting of control devices should be installed as indicated in the traffic control plan or recom-
the roundabout, especially if it mended typical details. This traffic control shall remain in place as long as it applies
must be built under traffic. and then be removed when the message no longer applies to the condition.
Prior to work that would change the traffic patterns to that of a roundabout, certain
peripheral items may be completed. This would include permanent signing (cov-
ered), lighting, and some pavement markings. These items, if installed prior to the
construction of the central island and splitter islands, would expedite the opening
of the roundabout and provide additional safety during construction.
4. Construct splitter islands and delineate the central island. At this point the signs
should be uncovered and the intersection should operate as a roundabout.
It is important to educate the public whenever there is a change in traffic patterns. Public education during
It is especially important for a roundabout because a roundabout will be new to construction is as important as
most motorists. The techniques discussed in Chapter 2 can be applied during the the public education effort
construction period. The following are some specific suggestions to help alleviate during the planning process.
initial driver confusion.
• Prepare news releases/handouts detailing what the motorist can expect be-
fore, during, and after construction;
• Install signing during and after construction that warns of changed traffic patterns.
7.5 Landscaping
7.5.1 Advantages
Landscaping in the central island, in splitter islands (where appropriate), and along Landscaping is one of the
the approaches can benefit both public safety and community enhancement. distinguishing features that
gives roundabouts an aesthetic
The landscaping of the roundabout and approaches should: advantage over traditional
intersections.
• Make the central island more conspicuous;
• Avoid obscuring the form of the roundabout or the signing to the driver;
• Clearly indicate to the driver that they cannot pass straight through the intersec-
tion;
• Help blind and visually impaired pedestrians locate sidewalks and crosswalks.
The central island landscaping can enhance the safety of the intersection by mak-
ing the intersection a focal point and by lowering speeds. Plant material should be
selected so that sight distance (discussed in Chapter 6) is maintained, including
consideration of future maintenance requirements to ensure adequate sight dis-
tance for the life of the project. Large, fixed landscaping (trees, rocks, etc.) should
be avoided in areas vulnerable to vehicle runoff. In northern areas, the salt toler-
ance of any plant material should be considered, as well as snow storage and
removal practices. In addition, landscaping that requires watering may increase
the likelihood of wet and potentially slippery pavement. Exhibit 7-24 shows the
recommended placement of landscaping within the central island.
The slope of the central island should not exceed 6:1 per the requirements of the
AASHTO Roadside Design Guide (9).
Avoid items in the central island Where truck aprons are used in conjunction with a streetscape project, the pave-
that might tempt people to take ment should be consistent with other streetscape elements. However, the mate-
a closer look. rial used for the apron should be different than the material used for the sidewalks
so that pedestrians are not encouraged to cross the circulatory roadway. Street
furniture that may attract pedestrian traffic to the central island, such as benches or
monuments with small text, must be avoided. If fountains or monuments are be-
ing considered for the central island, they must be designed in a way that will
enable proper viewing from the perimeter of the roundabout. In addition, they must
be located and designed to minimize the possibility of impact from an errant vehicle.
In general, unless the splitter islands are very large or long, they should not contain
trees, planters, or light poles. Care must be taken with the landscaping to avoid
obstructing sight distance, as the splitter islands are usually located within the
critical sight triangles (see Chapter 6).
7.5.4 Maintenance
A realistic maintenance program should be considered in the design of the land- Ensure that whatever
scape features of a roundabout. It may be unrealistic to expect a typical highway landscaping is installed, it will
agency to maintain a complex planting plan. Formal agreements may be struck be maintained.
with local civic groups and garden clubs for maintenance where possible. Liability
issues should be considered in writing these agreements. Where there is no inter-
est in maintaining the proposed enhancements, the landscape design should con-
sist of simple plant materials or hardscape items that require little or no mainte-
nance.
7.6 References
8
8.1
System Considerations
Roundabouts have been considered as isolated intersections in most other inter- This chapter considers
national roundabout guides and publications. However, roundabouts may need to roundabouts as they relate to
fit into a network of intersections, with the traffic control functions of a roundabout other elements of the
supporting the function of nearby intersections and vice versa. The purpose of this transportation system,
chapter is to provide some guidance on potentially difficult, but not uncommon, including other intersections.
circumstances or constraints.
Many countries whose initial design and driver experience was with isolated round-
abouts have since extended their application to transportation system design and
operation. This chapter addresses the appropriate use of roundabouts in a roadway
network context and the benefits obtained. Since the design of each roundabout
should generally follow the principles of isolated roundabout design, the discus-
sion is at a conceptual and operational level and generally complements the plan-
ning of isolated roundabouts discussed in Chapter 3. In many cases, site-specific
issues will determine the appropriate roundabout design elements.
Building upon this understanding, the next sections address design and perfor-
mance of two closely spaced roundabouts and the specific application to round-
about interchanges. This is followed by issues pertaining to the use of roundabouts
on an arterial or network that may include or replace coordinated signalized inter-
sections. Finally, the role of microscopic simulation models in assisting with analy-
sis of these system effects is reviewed.
Roundabouts should not be Roundabouts operate effectively only when there are sufficient longer and accept-
planned for metering or able gaps between vehicles in the circulatory lanes. If there is a heavy movement
signalization unless of circulating drivers, then entering drivers at the next downstream entry may not
unexpected demand dictates be able to enter. This situation occurs most commonly during the peak periods, and
this need after installation. the performance of the roundabout can be greatly improved with entrance metering.
Nearby intersections or Another method of metering is the use, with appropriate timing, of a nearby up-
pedestrian crossing signals can stream signalized intersection or a signalized pedestrian crossing on the subject
also meter traffic, but not as approach road. Unlike pure entry metering, such controls may stop vehicles from
effectively as direct entering and leaving the roundabout, so expected queue lengths on the round-
entrance metering. about exits between the metering signal and the circulatory roadway should be
compared with the proposed queuing space.
Full signalization that includes control of circulating traffic at junctions with major Full signalization of the
entrances is possible at large-diameter multilane traffic circles or rotaries that have circulatory roadway requires
adequate storage space on the circulatory roadway. The double-lane roundabout careful coordination and
dimensions resulting from the design criteria recommended in this guide may pre- vehicle progression.
clude such possibilities. As stated previously, full signalization should in any case
only be considered as a retrofit alternative resulting from unanticipated traffic de-
mands. Other feasible alternatives should also be considered, such as flaring criti-
cal approaches, along with the associated widening of the circulatory roadway;
converting a large-diameter rotary to a more compact modern roundabout form; or
converting to a conventional signalized intersection. This guide recommends that
signalizing roundabouts to improve capacity be considered only when it is the most
cost-effective solution.
Traffic signals at fully signalized rotaries should be timed carefully to prevent queu-
ing on the circulatory roadway by ensuring adequate traffic progression of circulat-
ing traffic and especially critical movements. Introducing continuous or part-time
signals on the circulatory roadway requires careful design of geometry, signs, lane
markings, and signal timing settings, and literature on this specific topic should be
consulted (1, 2).
There are essentially two ways in which rails can interact with a roundabout, as
shown in Exhibit 8-1:
• Through the center; or
• Across one leg in close proximity to the roundabout.
In either case, traffic must not be forced to stop on the tracks. A new intersection
should not be designed with railroad tracks passing through the center of it. How-
ever, on occasions, the rail line passes through an existing intersection area. The
traffic engineer might be faced with a decision whether to change the intersection
type to a roundabout or to grade-separate the crossing.
Closing only the leg with the • Method A: Closure only at rail crossing. This method prohibits vehicles from
rail crossing may work if crossing the rails but still allows vehicles to enter and leave the circulatory road-
queues are not anticipated to way. This method allows for many of the movements through the roundabout to
back into the continue to run free, if a queue does not build to the point of impeding circula-
circulatory roadway. tion within the roundabout. A queuing analysis should be performed using the
expected volume crossing the rails and the expected duration of rail crossing to
determine the likelihood that this blockage will occur. In general, this method
works better than Method B if there is sufficient separation between the round-
about and the rail crossing. If blockage is anticipated, the designer should choose
Method B.
• Method B: Closure at rail crossing and at most entries to the roundabout. This
method closes all entries to the roundabout except for the entry nearest the rail
crossing. This allows any vehicles in the roundabout to clear prior to the arrival
of the train. In addition, a gate needs to be provided on the approach to the rail
crossing exiting the roundabout to protect against possible U-turns in the round-
about. This causes increased queuing on all approaches but is generally safer
than Method A when there is insufficient storage capacity between the round-
about and rail crossing.
France (5)
Closely spaced roundabouts Closely spaced roundabouts may improve safety by “calming” the traffic on the
may have a traffic calming major road. Drivers may be reluctant to accelerate to the expected speed on the
effect on the major road. arterial if they are also required to slow again for the next close roundabout. This
may benefit nearby residents.
Note that when conventional T-intersections are used, Option A is less preferable
than Option B due to the need to provide interior storage space for left turns in
Option A. Therefore, roundabouts may be a satisfactory solution for cases like
Option A.
Freeway ramp junctions with arterial roads are potential candidates for roundabout
intersection treatment. This is especially so if the subject interchange typically has
a high proportion of left-turn flows from the off-ramps and to the on-ramps during
certain peak periods, combined with limited queue storage space on the bridge
crossing, off-ramps, or arterial approaches. In such circumstances, roundabouts
operating within their capacity are particularly amenable to solving these problems
when compared with other forms of intersection control.
There are two basic types of roundabout interchanges. The first is a large diameter
roundabout centered over or under a freeway. The ramps connect directly into the
roundabout, as do the legs from the crossroad. This is shown in Exhibit 8-5.
This type of interchange requires two bridges. If the roundabout is above the free-
way as shown in Exhibit 8-5, then the bridges may be curved. Alternatively, if the
freeway goes over the roundabout then up to four bridges may be required. The
number of bridges will depend on the optimum span of the type of structure com-
pared with the inscribed diameter of the roundabout island and on whether the
one bridge is used for both freeway directions or whether there is one bridge for
each direction. The road cross-section will also influence the design decision. Ex-
hibit 8-6 shows an example from the United Kingdom. The designer should decide
if the expected speeds of vehicles at larger roundabouts are acceptable.
The second basic type uses a roundabout at each side of the freeway and is a
specific application of closely spaced roundabouts discussed in the previous sec-
tion. A bridge is used for the crossroad over the freeway or for a freeway to cross
over the minor road. Again, two bridges may be used when the freeway crosses
over the minor road.
One-bridge roundabout This interchange form has been used successfully in some cases to defer the need
interchanges have been to widen bridges. Unlike signalized ramps that may require exclusive left-turn lanes
successfully used to defer the across the bridge and extra queue storage, this type of roundabout interchange
need for bridge widening. exhibits very little queuing between the intersections since these movements are
almost unopposed. Therefore, the approach lanes across the bridge can be mini-
mized.
The actual roundabouts can have two different shapes or configurations. The first
configuration is a conventional one with circular central islands. This type of con-
figuration is recommended when it is desirable to allow U-turns at each round-
about or to provide access to legs other than the cross street and ramps. Examples
from the United Kingdom and France are shown in Exhibit 8-7.
United Kingdom
France
Raindrop central islands make The second configuration uses raindrop-shaped central islands that preclude some
wrong-way movements more turns at the roundabout. This configuration is best used when ramps (and not front-
difficult, but require navigating age roads) intersect at the roundabout. A raindrop central island can be considered
two roundabouts to to be a circular shape blocked at one end. In this configuration, a driver wanting to
make a U-turn. make a U-turn has to drive around both raindrop-shaped central islands. This con-
figuration has an additional advantage in that it makes wrong-way turns into the
off-ramps more difficult. On the other hand, drivers do not have to yield when
approaching from the connecting roadway between the two roundabouts. If the
roundabout is designed poorly, drivers may be traveling faster than they should to
negotiate the next roundabout safely. The designer should analyze relative speeds
to evaluate this alternative. On balance, if the length of the connecting road is
short, this design may offer safety advantages. Exhibit 8-8 provides an example of
this type of interchange configuration.
Roundabouts produce more The benefits and costs associated with this type of interchange also follow those
random headways on ramps for a single roundabout. A potential benefit of roundabout interchanges is that the
than signalized intersections, queue length on the off-ramps may be less than at a signalized intersection. In
resulting in smoother merging almost all cases, if the roundabout would operate below capacity, the performance
behavior on the freeway. of the on-ramp is likely to be better than if the interchange is signalized. The head-
way between vehicles leaving the roundabout along the on-ramp is more random
than when signalized intersections are used. This more random ramp traffic allows
for smoother merging behavior on the freeway and a slightly higher performance
at the freeway merge area compared with platooned ramp traffic from a signalized
intersection.
The traffic at any entry is the same for both configurations. The entry capacity is
the same and the circulating flow is the same for the large single roundabout (Ex-
hibit 8-6) and for the second configuration of the two teardrop roundabout system
(Exhibit 8-8). Note that the raindrop form may be considered and analyzed as a
single large roundabout as in the circular roundabout interchange, but with a
“pinched” waistline across or under one bridge rather than two. The relative perfor-
mance of these systems will only be affected by the geometry of the roundabouts
and islands. The system with the two circular roundabouts will have a slightly differ-
ent performance depending upon the number of U-turns.
The design parameters are not restrained by any requirement here. They are only
constrained by the physical space available to the designer and the configuration
selected. The raindrop form can be useful if grades are a design issue since they
remove a potential cross-slope constraint on the missing circulatory road segments.
If there are more roads intersecting with the interchange than the single cross
road, then two independent circular roundabouts are likely to be the best solution.
Signalized intersections close The performance of a roundabout is affected by its proximity to signalized intersec-
to roundabouts produce gaps tions. If a signalized intersection is very close to the roundabout, it causes vehicles
in traffic that can be used by to enter the roundabout in closely spaced platoons; more importantly, it results in
minor street traffic to enter the regular periods when no vehicles enter. These latter periods provide an excellent
major street. opportunity for traffic on the next downstream entry to enter. Since the critical gap
is larger than the follow-up time, a roundabout becomes more efficient when the
vehicles are handled as packets of vehicles rather than as isolated vehicles.
When the signalized intersection is some distance from the roundabout, then the
vehicles’ arrival patterns have fewer closely spaced platoons. Platoons tend to dis-
perse as they move down the road. The performance of a roundabout will be re-
duced under these circumstances when compared with a close upstream signal. If
arrival speeds are moderate, then few longer gaps allow more drivers to enter a
roundabout than a larger number of shorter gaps. If arrival speeds are low, then
there are more opportunities for priority-sharing (where entering and circulating
vehicles alternate) and priority-reversals (where the circulating vehicles tend to
yield to entering vehicles) between entering and circulating traffic streams, and the
influence of platoon dispersal is not as marked.
Traffic leaving a roundabout tends to be more random than if another type of inter-
section control were used. A roundabout may therefore affect the performance of
other unsignalized intersections or driveways more than if the intersection was
signalized. However, as this traffic travels further along the road downstream of
the roundabout, the faster vehicles catch up to the slower vehicles and the propor-
tion of platooning increases.
Even one circulating vehicle in If a roundabout is used in a network of coordinated signalized intersections, then it
a roundabout will result in a may be difficult to maintain the closely packed platoons required. If a tightly packed
platoon breaking down. platoon approached a roundabout, it could proceed through the roundabout as long
as there was no circulating traffic or traffic upstream from the left. Only one circu-
lating vehicle would result in the platoon breaking down. Hence, the use of round-
abouts in a coordinated signalized network needs to be evaluated carefully. One
possibility for operating roundabouts within a signal network is to signalize the
major approaches of the roundabout and coordinate them with adjacent upstream
and downstream signalized intersections.
Because roundabouts accommodate U-turns more easily than do signals, they may Roundabouts as an access
also be useful as an access management tool. Left-turn exits from driveways onto management tool.
an arterial which may currently experience long delays and require two-stage left-
turn movements could be replaced with a simpler right turn, followed by a U-turn at
the next roundabout.
The ultimate manifestation of roundabouts in a system context is to use them in Roundabouts may require
lieu of signalized intersections. Some European cities such as Nantes, France, and more right-of-way at
some Australian cities have implemented such a policy. It is generally recognized intersections, but may also
that intersections (or nodes), not road segments (or links), are typically the bottle- allow fewer lanes (and less
necks in urban roadway networks. A focus on maximizing intersection capacity right-of-way) between
rather than widening streets may therefore be appropriate. Efficient, signalized intersections.
intersections, however, usually require that exclusive turn lanes be provided, with
sufficient storage to avoid queue spillback into through lanes and adjacent inter-
sections. In contrast, roundabouts may require more right-of-way at the nodes, but
this may be offset by not requiring as many basic lanes on the approaches, relative
to signalized arterials. This concept is demonstrated in Exhibit 8-10.
Analysis tools, such as those provided in Chapter 4, should be used to evaluate the
arterial or network. These may be supplemented by appropriate use of microscopic
simulation models as discussed next. Supplemental techniques to increase the
capacity of critical approaches may be considered if necessary, such as bypass
lanes, flaring of approaches and tapering of exits, and signalization of some round-
about approaches.
Microscopic simulation of traffic has become a valuable aid in assessing the system
performance of traffic flows on networks, as recognized by the Highway Capacity
Manual 2000 (7). Analysis of many of the treatments discussed in this chapter may
benefit from the use of appropriate simulation models used in conjunction with ana-
lytic models of isolated roundabouts discussed in Chapter 4. These effects include
more realistic modeling of arrival and departure profiles, time-varying traffic patterns,
measurement of delay, spatial extent and interaction of queues, fuel consumption,
emissions, and noise. However, the user must carefully select the appropriate mod-
els and calibrate the model for a particular use, either against field data, or other
validated analytic models. It would also be advisable to check with others to see if
there have been any problems associated with the use of the model.
Microscopic simulation models are numerous and new ones are being developed,
while existing models are upgraded frequently. Each model may have particular
strengths and weaknesses. Therefore, when selecting a model, analysts should con-
sider the following:
• What are the model input requirements, are they sufficient, and how can they be
provided or estimated?
• What outputs does the model provide in animated, graphical, or tabular form?
• What special features of the model are pertinent to the problem being addressed?
• Does the user manual for the simulation model specifically address modeling a
roundabout?
When a simulation model is used, the analyst is advised to use the results to make
relative comparisons of the differences between results from changing conditions,
and not to conclude that the absolute values found from the model are equivalent to
field results. It is also advisable to perform a sensitivity analysis by changing selected
parameters over a range and comparing the results. If a particular parameter is found
Simulation results are best to affect the outcomes significantly, then more attention should be paid to accu-
used for relative comparisons, rate representation and calibration of this parameter. Finally, the analyst should
rather than relying on absolute check differences in results from using different random number seeds. If the dif-
values produced by the model. ferences are large, then the simulation time should be increased substantially.
CORSIM Urban streets, freeways FHWA has been investigating modifications that may be required for CORSIM to
adequately model controls such as stop and yield control at roundabouts through
gap acceptance logic. In this research, roundabouts have been coded as a circle
of four yield-controlled T-intersections. The effect of upstream signals on each
approach and their relative offsets has also been reported (8).
Integration Urban streets, freeways Integration has documented gap acceptance logic for permitted movements at
signal-, yield-, and stop-controlled intersections. As with CORSIM, Integration
requires coding a roundabout simply as a series of short links and nodes with
yield control on the entrances.
Simtraffic Urban streets Simtraffic is a simulation model closely tied to the signal timing software package
Synchro. Simtraffic has the capability to model unsignalized intersections and
thus may be suitable for modeling roundabouts. However, no publications to date
have demonstrated the accuracy of Simtraffic in modeling roundabout operations.
Paramics Urban streets, freeways Paramics has been used in the United Kingdom and internationally for a wide
range of simulation projects. It has been specifically compared with ARCADY in
evaluating roundabouts (9). The model has a coding feature to automatically code
a roundabout intersection at a generic node, which may then be edited. The
model has been used in the United Kingdom for a number of actual roundabout
evaluations. The model specifically employs a steering logic on the circulatory
roadway to track a vehicle from an entry vector to a target exit vector (10).
VISSIM Urban streets, transit networks VISSIM is widely used in Germany for modeling urban road and transit networks,
including roundabouts. Roundabout examples are provided with the software,
including explicit modeling of transit and pedestrians. Modeling a roundabout
requires detailed coding of link connectors, control, and gap acceptance
parameters (11).
8.7 References