CAPSA 2015 – Peer reviewed published papers

Roughness of Runways and Significance of

Appropriate Specifications and Measurement
Stephen Emery
Kubu Australia Pty Ltd, Perth, Australia

Arno Hefer
Arno Hefer Consulting, Pretoria, South Africa

Emile Horak
KUBU Consultancy Pty Ltd, Centurion, South Africa

Abstract— Most airport runways in southern Africa and and wear) and/ or other factors which may impair the safe
Australia are surfaced with asphalt. Many of the runway operation of the aircraft (cockpit vibrations, excessive g-forces,
surfacing specifications have been adapted from relevant road etc.). There are so many differences between aircraft and
specifications. However roughness is one parameter where vehicles that airport and highway pavement roughness studies
airport and road specifications differ. The International should be treated as different issues [4].
Roughness Index (IRI) is a measure of riding comfort commonly
used and specified for roads. IRI is fundamentally related to A. Road roughness
riding comfort experienced by the motor vehicle passenger. In Road roughness was initially measured as Present
contrast, runway roughness is defined in terms of fatigue on Serviceability Rating (PSR), using a panel of road users who
aircraft components and other factors such as cockpit vibrations subjectively rated the serviceability of various roads. Present
and excessive g-forces. As a control measure of roughness, IRI Serviceability Index (PSI) was then introduced and by the
has been specified on some airport projects which has the
1960s, vehicle measuring devices such as the Mays Meter and
disadvantage in that it is not measuring the elements important
on runway pavements. The paper introduces straight-edge,
PCA Meter were in use [8].
Boeing Bump Index (BBI), aircraft simulation and their use on Routine road roughness measurements have been taken in
various runways in southern Africa and Australia. Emphasis is South Africa since the early 1970s, initially with the PCA
placed on the interpretation and implementation of the latest Roadmeter. The Linear Displacement Integrator (LDI) was
roughness criteria for runways included in ICAO Annexure 14. developed at the CSIR in the late 1970s. This instrument sums
The objectives of this paper are to emphasise the significance of the linear movement between vehicle body and the rear axle,
appropriate roughness specifications on runways and to share much the same as the Mays Meter. The output was converted
experiences with the implementation of more relevant
to the PSI scale through a regression equation. The Quarter-car
specifications, including aspects of roughness measurement and
interpretation. Index (QI) scale was introduced to South Africa in the early
1980s, then by the middle 1990s conversion to the IRI scale for
Keywords—airport, roughness, ICAO, IRI, Boeing Bump Index road roughness commenced in South Africa [16].

I. INTRODUCTION IRI is expressed as the average longitudinal road profile

that represents the vertical response of a hypothetical quarter-
On roads, smoothness/roughness has been defined in terms car traveling at 80 km/h to the measured longitudinal road
of the ride quality experienced by a passenger in a car. profile.
Typically this involves smoothness or riding quality that is
measured with various devices such as non-contact laser B. Runway roughness
profilometers and the values expressed as International Runway roughness is not defined by perceived ride quality
Roughness Index (IRI) units. or passenger discomfort. Although important, passenger
discomfort due to runway surface irregularities is often not a
On airports, smoothness/roughness has been defined as
significant issue since the degree of discomfort is small and the
being free of bumps and irregularities that can impair safe
time of exposure is limited to a few seconds. Further,
operations, cause damage, or increase structural fatigue to an
passenger discomfort often occurs during take-off and landing
aircraft. There is a difference between cars and aircraft. The
operations when engine noise, aerodynamic noise, and/or
primary purpose of an aircraft suspension system is to absorb
horizontal acceleration or deceleration otherwise distract the
energy expended during landing. Aircraft suspension systems
passengers.
have less capacity to dampen the impact of surface
irregularities due to the magnitude of the energy that must be Runway roughness can induce stress on aircraft
addressed during landing. Runway roughness becomes defined components which increases the risk of premature failure due
in terms of fatigue on aircraft components (increased stresses
 CAPSA 2015 – Peer reviewed published papers

to fatigue. It can cause enough vibrations in the cockpit that pilot could have difficulty manipulating the controls. The
pilots cannot focus on critical instrumentation or could have authors have heard it expressed in phrases like “couldn’t see
difficulty manipulating the controls during take-off or landing. the instruments” or “the stick was shaking so badly that it was
Aircraft response to surface irregularities can even reduce hard to hold”.
braking capacity as the aircraft responds to vertical
acceleration. These factors can occur individually or in D. Aircraft simulation
combination, depending on aircraft response. Airport pavement roughness can be assessed by simulating
the movement of aircraft along the runway or taxiway. The
Runway roughness is therefore often defined in terms of history of aircraft simulation was summarised by Chen and
fatigue on aircraft components (increased stress and wear) and/ Chou [4] who reported that “In 1967, the National Aeronautics
or other factors which may impair the safe operation of the and Space Administration (NASA) established an airport
aircraft (cockpit vibrations, excessive g-forces, etc.), and not in pavement roughness evaluation procedure using an aircraft’s
terms of driver and passenger comfort. vertical acceleration at the cockpit, setting the maximum
The FAA has defined two categories of roughness based on acceptable acceleration to be 0.4 g.” Gerardi [7] “conducted a
the dimensions and frequency of surface deviations [6]. series of studies to develop a rigid-body aircraft model to
simulate the vertical acceleration at the pilot’s station and at the
Single Event Bump. Single event bumps are isolated centre of gravity of aircraft, as well as pavement loading at
events where changes in runway elevation occur over a main and nose gears. That model has degrees of freedom on
relatively short distance of 100 metres or less. This may occur pitch, roll, vertical and horizontal translation and it was verified
as an abrupt vertical lip or as a more gradual deviation from a with the field data gathered from KC-135, B-52, F-4C, and C-
planned runway profile. Depending on the operational speed 141. The model was further implemented to become the
and bump length, an aircraft suspension system may not be commercialized software, APRas (Airport Pavement
able to fully absorb the energy produced when it encounters a Roughness assessment software)”. Elements of APRas have
bump. Aircraft components and occupants feel the impact as a been incorporated in the ProFAA software, where aircraft
shock or sudden jolt. Discrete bumps create impact loading that response can be simulated for a library of representative
can accelerate fatigue damage, as well as rattle equipment, commercial aircraft [5].
crew, and passengers. Basic “straight-edge” analysis can easily
identify single event bumps. Riding the runway in a passenger E. Regulatory control of runway roughness
vehicle might reveal shorter length bumps, but finding longer Historically runway roughness has been dealt with by fairly
length bumps might require a thorough analysis of the runway simple regulations, such as earlier editions of International
profile. Civil Aviation Organisation Annex 14 [11]. Their only
Profile Roughness. Profile roughness is surface profile mandatory requirement in earlier editions was that:
deviations present over a portion of the runway that cause 3.1.22 The surface of a runway shall be constructed without
aircraft to respond in ways that can increase fatigue on aircraft irregularities that would result in loss in friction
components, reduce braking action, impair cockpit operations, characteristics or otherwise adversely affect the take-off or
and/or cause discomfort to passengers. Response depends on landing of an aeroplane. Note 1. Surface irregularities may
aircraft size, weight, and operation speed. Repeated large adversely affect the take-off or landing of an aeroplane by
wavelength bumps can induce harmonics and can accelerate causing excessive bouncing, pitching, vibration, or other
fatigue damage to both the aircraft and the pavement. Repeated difficulties in the control of an aeroplane.
short wavelength bumps can cause heat build-up in
The earlier Annex 14 editions up to and including the 4th
struts/suspension. Even when roughness does not cause
discomfort to passengers, it may still affect the fatigue life of edition (ICAO, 2004) had two suggested roughness
aircraft components or decrease operational safety of the measurements in Attachment A-6 to the Annex:
aircraft. Depending upon aircraft characteristics and operating 5.1 In adopting tolerances for runway surface irregularities,
speed, an aircraft may be excited into harmonic resonance due the following standard of construction is achievable for short
to profile roughness which can increase inertial forces or distances of 3 m and conforms to good engineering practice:
vibrations within the aircraft structure. One example is resonant Except across the crown of a camber or across drainage
response in a dual tandem 4-wheel gear, such as the Airbus channels, the finished surface of the wearing course is to be of
A330 gear, where the pitch mode increases friction in the pivot such regularity that, when tested with a 3 m straight-edge
joint. placed anywhere in any direction on the surface, there is no
deviation greater than 3 mm between the bottom of the
C. Role of pilot feedback
straight-edge and the surface of the pavement anywhere along
Pilot observations and complaints are one of the more the straight-edge.
important factors in determining runway roughness. Although 5.3 The operation of aircraft and differential settlement of
pilot observations do not directly indicate that structural fatigue surface foundations will eventually lead to increases in
of aircraft components is occurring, they are often the first sign surface irregularities. . . . . In general, isolated irregularities
that something is wrong with the runway profile. The
of the order of 2.5 cm to 3 cm over a 45 m distance are
magnitude of observation and discomfort is a different order in
tolerable . . . [there are also maximum limits of 80mm over
aircraft compared to cars. Pilot discomfort is vertical
acceleration so severe that the instruments are blurred or the 45m and temporarily acceptable limits of 130mm over 45m] . .
. If the maximum limits are exceeded, corrective action should
 CAPSA 2015 – Peer reviewed published papers

be undertaken as soon as reasonably practicable to improve files, are converted to a .TXT file using ProVAL, which is then
the ride quality. If the temporarily acceptable limits are converted by the ConvertProfileFormat software to the
exceeded, the portions of the runway that exhibit such ProFAA format which is a .PRO file.
roughness should have corrective measures taken immediately
if aircraft operations are to be continued. TABLE I. ACCEPTANCE CRITERIA FOR IRI ROUGHNESS VALUES ON A
RECENT SOUTHERN AFRICAN RUNWAY
The Annex 14 5th edition [12] continued with these and
introduced Boeing Bump Index at its Fig. A-3 which compared Limit Value
Maximum (%) of 1 km Segment with Roughness
ICAO and FAA standards against a background of the Boeing (Average 100
Worse Than Limit Value
m IRI)
Bump Index; this is not a standard nor is it mandatory. It is the
same as the latest Annex 14 (6th edition) [13]. 1.40 20%

1.60 5%
II. ROUGHNESS MEASURING
2.00 0%
A. Road profile measurement
The roughness measurement equipment in most common
use today on roads is the laser (non-contact) profiler, in which Isolation of runway roughness can only be really
a vehicle is fitted with a laser-based measurement system understood by a surface profile survey. A high-speed run over
consisting of lasers and accelerometers that measure and record the runway in a vehicle is less than an ideal indicator because
the road profile. In South Africa, the laser profilometer is of the differences in wheel base, mass, speed, and suspension
validated on validation sections and calibrated by the between the car and the airplane. Visual observance will
manufacturer. Response type devices are calibrated on normally not reveal a runway roughness problem either,
calibration sections, measured with rod-and level and against a because the bumps are often too long in length or shallow in
contact profiler such as the Dipstick profilometer. Laser depth to appear to the eye.
profilometers, such as the Dynatest RSP or Hawkeye 1000 C. Application of road profile measurements to runways
profilometers, measure at speed and are used on roads. They
have had limited use on runways in some countries, although The IRI is not part of the ICAO regulations, but roughness
this has sometimes proved unsatisfactory because road equipment used to measure it on roads is readily available. This
measurements do not necessarily translate to runway makes the application of IRI to runways tempting, even as an
requirements. indirect value to quantify smoothness, even if it is not
recommended. A recent specification for an airport runway in
B. Runway profile measurement southern Africa is shown in Table 1 to illustrate such an
Runway profile measurement is often done using elevation application. The average 100 m IRI for a “survey lane” was
surveys. At its simplest, a profile elevation survey over the calculated by averaging the 100 m IRI left and right wheel path
runway using a rod and level survey is a quick and valid values as specified and the measurements had to be taken
method. The profile survey should be conducted along the immediately after completion of the runway construction [10].
centreline of the runway over the reported rough areas as a
minimum, but it is preferable to survey the entire runway III. AIRPORT ROUGHNESS MEASUREMENTS
length. Survey lines along the track where the main gear would A. Roughness software
normally be, are very helpful in determining the full extent of
ProFAA is the Federal Aviation Administration’s software
the roughness and airplane response. The main gear tracks are
for computing airport pavement elevation profile roughness
normally about 3 metres either side of the centreline for narrow
indexes [5]. Data analysis performed by the program includes
body passenger aircraft such as the Boeing 737. An additional
the calculation of the indices for Straight-edge (SSI), Boeing
survey track for wide body aircraft such as the Airbus A330 at
Bump Index (BBI) and International Roughness Index (IRI). It
the 6 metre offset may be necessary. It is recommended that the
also does the simulation of aircraft response in a similar
longitudinal survey interval be on a maximum of 3 metre
process as the APRas software.
stations. It can be performed with an ordinary surveyor's level
and rod or by the use of a laser instrument and a rod that The ProVAL software allows one to view and analyze
detects the laser beams. A runway can be surveyed in a few longitudinal pavement profiles in many different ways,
hours. including the common road measurements such as IRI,
profilograph and rolling straight-edge. It takes machine data
The self-contained profile measuring devices are also
files from several profilometers, such as the Dynatest .RSP
proficient in measuring the runway profile quickly. These
files. ProVAL [15] is a US FHWA/LTPP product, and was
include the contact profilers such as the DipStick and SurPRO.
originally released in 2001; the current version is 3.5.
These are typically used to survey a full runway at 3m, 6m,
10m and 20m offsets from the centreline. The commercial software APRas can simulate 15 different
types of commercial aircraft ranging from the Boeing 737-800
Laser profilometers can be used on runways provided they
up to the Boeing 747-400 aircraft as well as a select variety of
are able to output a longitudinal profile in a form that can used
military aircraft. Simulations include take-off, landing,
by runway roughness software such as ProFAA. ProFAA has
constant speed taxi and aborted take-off scenarios [1].
very specific requirements for input data and it is not normally
possible to read directly from the profilometer software.
Instead profilometer machine files, such as the Dynatest .RSP
 CAPSA 2015 – Peer reviewed published papers

B. Airports analysed 9.1032mm. Clearly the limit of 3mm is exceeded at a number

Roughness measurements at two airports were analysed for of localised points along the runway, which is unacceptable.
this paper. Airport M had a new asphalt surfaced runway 2720 Roughness of existing runways in the ICAO system is also
x 45 metres; it was designed to ICAO 4E geometric standards. measured with a straight-edge. This detects surface
The forecast traffic was very light, with less than 100 wide- irregularities resulting from the operation of aircraft and
body departures forecast for the 20 year design life. The differential settlement of surface foundations. ICAO Annex 14
surfacing construction quality was hindered by asphalt plant suggests that isolated irregularities of the order of 25 to 30mm
supply problems and paving issues, and it was done by asphalt over a 45 m distance (under a 45m straight-edge) are tolerable,
paving to a roads standard without the use of a shuttle buggy. with the maximum surface irregularity height (or depth) of 80
Airport B had an existing 2500 x 45 metre runway which was mm over 45m. If the maximum limits are exceeded, corrective
designed to ICAO 4C geometric standards. The pavement had action needs to be undertaken as soon as reasonably practicable
been overloaded for many years by Boeing 737 traffic and the to improve the ride quality.
runway was significantly deformed and rough. The runway
was the subject of some pilot complaints about roughness. At Airport M, for the left and right wheel tracks (3m each
Airport B runway was then rehabilitated and reshaped by a side of centreline), the 45m straight-edge was applied along the
major asphalt overlay involving multiple layers of asphalt. The runway using the ProFAA software, even though this runway
construction quality of the surfacing was to a high standard, was newly constructed. The graph of 3m north of centreline is
and smooth construction was assisted by the use of a shuttle shown in Fig. 2 (3m S CL is similar and is omitted for reasons
buggy (materials transfer vehicle). of space); the Y axis ranges from 0mm to 42.491mm. Clearly
the 25-30mm tolerable limit is exceeded at a few localised
C. Straight-Edge Measurement points along the runway, which is unacceptable for new
Roughness for acceptance of new construction in the ICAO construction. It would however be acceptable for an existing
system is typically measured with a straight-edge. This detects runway which had been trafficked for a number of years.
adjoining asphalt paver runs or concrete slabs that are mis-
2) Airport B
matched in elevation. It also detects light fittings or drain grates
Here, for the left and right wheel tracks (3m each side of
that may be set too low or protrude too high. The ICAO
centreline), the 3m straight-edge was applied along the runway
standard is 3mm under a 3m straight-edge at construction.
using the ProFAA software. The graph of 3m north of
The ProFAA software calculates a Straightedge centreline is shown in Fig. 3 (3m S CL is similar and is omitted
Smoothness Index (SSI), which uses data from any profiling for reasons of space); the upper graph is the rough runway
device and finds the maximum deviation anywhere along that before rehabilitation (Y axis ranges from 0mm to 1.9186mm)
straightedge and plots the absolute value [14]. This is different and the lower graph is the smooth runway after rehabilitation
to the rolling straightedge software where, for example, (Y axis ranges from 0mm to 0.8379mm). In neither case was
ProVAL simply determines the vertical deviation between the the limit of 3mm exceeded, which is acceptable. It does show
centre of the straightedge and the profile. Experience shows the limitation of the 3m straight-edge in not detecting runways
that undulations can have both an upward and downward which have become rough over time due to aircraft traffic
component and Straightedge Smoothness Index better captures generating long wavelength roughness. During the
the bump from both. rehabilitation at Airport B, a 3m rolling straight-edge was used
on the asphalt surfacing layer, run longitudinally along every
1) Airport M paver lane, and no occurrences of 3mm or more were recorded.
At Airport M, for the left and right wheel tracks (3m each This provided a check of SSI against the manual rolling
side of centreline), the 3m straight-edge was applied along the straight edge.
runway using the ProFAA software. The graph of 3m north of
centreline is shown in Fig. 1 (3m S CL is similar and is omitted
for reasons of space); the Y axis ranges from 0mm to

Fig. 1. Airport M - 3m N CL - 3m straight-edge graph

Fig. 2 Airport M - 3m N CL - 45m straight-edge graph

 CAPSA 2015 – Peer reviewed published papers

Fig. 3 Airport B - 3m N CL - 3m straight-edge – upper graph is before/rough and lower graph is after/smooth

Fig. 4 Airport B - 3m N CL - 45m straight-edge graph – upper graph is before/rough and lower graph is after/smooth

Fig. 5 Boeing Bump Index - Roughness Acceptance Criteria

Fig. 6 Airport M - 3m N CL – Boeing Bump Index

At Airport B, for the left and right wheel tracks (3m each runway using the ProFAA software. The graph of 3m north of
side of centreline), the 45m straight-edge was applied along the centreline is shown in Fig. 4 (3m S CL is similar and is omitted
 CAPSA 2015 – Peer reviewed published papers

for reasons of space); the upper graph is the rough runway The BBI for the two new/rehabilitated runways were
before rehabilitation (Y axis ranges from 0mm to 111.18mm) remarkably close at 0.2876 and 0.2612, even though the
and the lower graph is the smooth runway after rehabilitation surfacing on one was built to a higher standard than the other.
(Y axis ranges from 0mm to 29.763mm). This suggests that BBI is a good measure of the intrinsic
smoothness of the runway since, despite the difference in
Clearly the 25-30mm tolerable limit was exceeded for the surfacing quality, both runways were intrinsically new and
rough/before case. The maximum surface irregularity height smooth.
(or depth) of 80 mm over 45m was also exceeded at a few
localised points along the runway, but the temporary limit of E. Aircraft response
130mm was not exceeded. Corrective action was needed to be Aircraft response was simulated using ProFAA. The
undertaken as soon as reasonably practicable to improve the aircraft modelled was a Boeing 727 which is representative of
ride quality, but the runway did not require immediate narrowbody jet airliners; the aircraft was modelled at high
corrective measures. speed (100 knots, which is 180 kph). Key outputs of the
D. Boeing Bump Index simulation are a graph along the runway of the vertical
acceleration at the pilot’s station (cp), and a second graph of
1) Introduction the vertical acceleration at the aircraft’s centre of gravity (cg).
Boeing developed a criteria that describes runway To these was applied a +/- 0.4g limit, which is defined as the
roughness as a single event condition [3]. FAA took the Boeing “threshold of discomfort” as reported by Goldman [9].
bump procedure and created the “Boeing Bump Index” (BBI).
Fig. 5 shows the acceptable, excessive, and unacceptable 1) Airport M
evaluation zones in terms of BBI. When the BBI value is At Airport M, for the left and right wheel tracks (3m each
below 1.0, the Boeing bump criteria is in the acceptable zone. side of centreline), the vertical acceleration at the pilot’s station
Values of BBI greater than 1.0 fall in either the excessive or (G cp) was calculated along the runway. The high speed graph
unacceptable zones. of 3m north of centreline is shown in Fig. 8 (3m S CL is
similar and is omitted for reasons of space). The Y-axis ranges
The Boeing Bump Index quickly identifies roughness that of G cp is from -0.3256g to 0.2361g (a range of 0.56g). This
can produce poor aircraft ride quality, but it should be noted has some evident bumps, but is below the suggested limit of +/-
that there are some fundamental issues that limit its 0.4g and so the runway is acceptable, although this is not
effectiveness. Because this index only evaluates the event's considered a good result for a new runway. Vertical
wavelength and amplitude, the Boeing Bump Index can only acceleration at the aircraft centre of gravity (G cg) was
evaluate one event at a time. Many roughness investigation measured and was always lower than G cp.
projects find that the poor aircraft response is due to several
events in succession; each of which could be found acceptable Airport B
by the Boeing Bump Index. But in reality, the aircraft responds
to the chain of events as a whole and as a result, can produce At Airport B, for the left and right wheel tracks (3m each
side of centreline), the vertical acceleration at the pilot’s station
some very poor ride quality. Unfortunately, the Boeing Bump
Index can declare a runway as acceptable that, in reality, would (G cp) was calculated along the runway. The high speed graph
of 3m north of centreline is shown in Fig. 9 (3m S CL is
fail the aircraft response threshold of +/- 0.4 g (gravitational
forces) of aircraft response [2]. similar and is omitted for reasons of space). The upper graph is
the rough runway before rehabilitation (Y axis ranges from -
2) Airport M 0.6629g to 0.3765g) and the lower graph is the smooth runway
At Airport M, for the left and right wheel tracks (3m each after rehabilitation (Y axis ranges from -0.2306g to 0.1632g).
side of centreline), the Boeing Bump Index was calculated The runway roughness before rehabilitation was well in excess
along the runway using the ProFAA software. The graph of 3m of the suggested limit of +/-0.4g, and this corresponds with the
north of centreline is shown in Fig. 6 (3m S CL is similar and pilot complaints about that runway; the runway roughness was
is omitted for reasons of space); the Y axis ranges from 0 to unacceptable. However after rehabilitation, the roughness was
0.2876. This is smooth, and well below the BBI limits of Fig. below the limits and acceptable. In this respect, aircraft
5, and the runway is acceptable. response simulation can detect complex roughness problems
which BBI would miss.
3) Airport B
At Airport B, for the left and right wheel tracks (3m each The aircraft simulation also picked up the surfacing quality
side of centreline), the Boeing Bump Index was calculated differences between the two airports, with the range of
along the runway using the ProFAA software. The graph of 3m acceleration at Airport M with the lower quality surfacing
north of centreline is shown in Fig. 7 (3m S CL is similar and construction being 0.56g and Airport B with high quality
is omitted for reasons of space); the upper graph is the rough surfacing construction being 0.39g.
runway before rehabilitation (Y axis ranges from 0 to 0.7521)
and the lower graph is the smooth runway after rehabilitation
(Y axis ranges from 0 to 0.2612). The runway before
rehabilitation was known to be rough and pilot complaints
were being received, but is considered acceptable by BBI.
 CAPSA 2015 – Peer reviewed published papers

Fig. 7 Airport B - 3m N CL – Boeing Bump Index – upper graph is before/rough and lower graph is after/smooth

Fig. 8 Airport M - 3m N CL – vertical acceleration at the pilot’s station (cp) for Boeing 727 aircraft at 100 knots

Fig. 9 Airport B - 3m N CL – vertical acceleration at the pilot’s station (Cp) for Boeing 727 aircraft at 100 knots– upper graph is before/rough and lower graph
is after/smooth

Fig. 10 Airport M - 3m N CL - IRI averaged every 100m

 CAPSA 2015 – Peer reviewed published papers

Fig. 11 Airport M - 3m S CL - IRI averaged every 100m

F. IRI Boeing Bump Index (upper and lower graphs respectively of

At Airport M, for the left and right wheel tracks (3m each Fig. 12). There is no comparable trend or correspondence
side of centreline), the International Roughness Index (IRI) was between the two measures. However the BBI graph does show
calculated using ProVAL 3.5. This was checked using ProFAA three locations where some bigger ‘bumps’ are – two near the
with excellent agreement. Data was collected immediately after left hand side and one at the centre.
construction in 2013 using a Dynatest Road Surface profiler. The two ICAO straight-edge measurements give more
The airport is served by narrow-body aircraft, so the 3m offset insight as to where the bumps really are, and where corrective
wheeltracks were the appropriate ones to measure. The IRI was action should be taken (Fig. 13). The long wavelength ‘bumps’
averaged every 100m and plotted so that the data could be show up on the 45m straight-edge graph and are the same as
evaluated against Table 1, with IRI=1.6 mm/km shown as a red the BBI graph – two near the left hand side and one in the
line (Fig. 10 and Fig. 11). centre. The small discontinuities, possibly due to poor
Using the acceptance criteria of Table 1, the runway was construction such as stop/start asphalt paving, show up at a
smoother (better) than the criteria after the runway end effects number of places on the 3m straight-edge graph.
were discounted (detailed analysis showed that high average Aggregated data, in the form of 100m segment averages,
IRI values were limited to 50m start/stop at the ends). was used to compare IRI and Boeing Bump Index. At Airport
M, for both the left and right wheel tracks (3m each side of
IV. DIRECT COMPARISON OF ROAD AND AIRPORT centreline), IRI and Boeing Bump Index were averaged for
MEASUREMENTS 100m segments and compared (in Fig. 14 and Fig. 15). It is
Chen and Chou [4] compared the aircraft simulation evident that the general trend is similar for both measures; this
outputs of the APRas software (similar to ProFAA) with IRI. is not unexpected since in the simplest of terms, smooth is
They concluded that it was clear that an aircraft’s vertical smooth and rough is rough. However the two measures are still
acceleration and pavement loading responses were not identical not directly comparable, as the left hand side of Fig. 14 clearly
to the responses of IRI, and concluded that IRI is not suitable shows.
for evaluating airport pavement roughness.
In this paper, a direct comparison was made of road and
airport measurements at Airport M. The measures were
compared by plotting the raw data (un-averaged) for IRI and
 CAPSA 2015 – Peer reviewed published papers

Fig. 12 Airport M - 3m N CL - IRI upper graph and BBI lower graph

Fig. 13 Airport M - 3m N CL – 3m straight-edge upper graph and 45m straight-edge lower graph

discontinuities. The averaged IRI data do not show these at all,

and while the BBI pattern shows the large/long wavelength
‘bumps’, the BBI segment averages are so low in value that the
runway appears to be very smooth. Averages are considered
unacceptable for bump analysis.
V. RECOMMENDED AIRPORT ROUGHNESS
SPECIFICATIONS
The assessment of runway roughness is complex and no
single index is sufficient to describe the roughness of runways.
The Boeing Bump Index (BBI) gives an overall measure of the
Fig. 14 Airport M - 3m N CL - comparing IRI and BBI for 100m segment runway’s acceptability, and shows the location of long
averages wavelength ‘bumps’ but not small discontinuities such as poor
paving or light fittings. BBI can only evaluate one event at a
time, and often poor aircraft response is due to several events
in succession; each of which could be found acceptable by the
BBI. But in reality, the aircraft responds to the chain of events
as a whole and as a result, can produce some very poor ride
quality.
Aircraft response simulation of aircraft travelling at high
speed can calculate vertical accelerations and strut forces. The
aircraft simulation gives a good measure of acceptability,
relates to pilot concerns, and shows up long wavelength bumps
as well as surfacing quality differences between the two
airports. The two ICAO straight-edge measurements give
Fig. 15 Airport M - 3m S CL - comparing IRI and BBI for 100m segment insight as to where bumps really are, and where corrective
averages action should be taken. The long wavelength ‘bumps’ show up
on the 45m straight-edge graph and are the same as the BBI
The use of segment or section averages raises the concern graph – two near the left hand side and one in the centre. The
that they serve to mask the effect of bumps, which get lost in small discontinuities, possibly due to poor construction such as
the process. The raw data BBI and straight-edge measurements stop/start asphalt paving, show up at a number of places on the
show where bumps really are, and it is evident in Fig. 12 and 3m straight-edge graph.
Fig. 13 that several large/long wavelength ‘bumps’ really do
exist, as well as a number of small/short wavelength It is recommended that:
 CAPSA 2015 – Peer reviewed published papers

- Boeing Bump Index be used to give an overall [15] ProVAL (2012) ProVAL Users Guide - Profile Viewing and Analysis
measure of the runway’s acceptability, although it is noted that Software. Manual Revision: 1.2 (3.4). The Transtec Group, Austin,
Texas.
it can give false positives.
[16] Visser A, Kannemeyer L, and Tekie S (1998) Experience with
- Aircraft response simulation must also be done to roughness measurements for pavement management in Southern Africa.
confirm the acceptability of the roughness to pilots, and over- 4th Int. Conf. on Managing Pavements, Durban, South Africa.
ride the false positives of the Boeing Bump Index.
- ICAO 3m and 45m straight-edge tools are used to find
the location of bumps, detect whether they are long wavelength
bumps or short discontinuities, and confirm the BBI and
aircraft simulations.
- Segment averages should not be used for runways
because averaging serves to mask the effect of bumps.
- International Roughness Index (IRI) should not be
used for runways.
- Probabilistic measures not be used as acceptance
criteria (such as shown in Table 1) since they assume a certain
statistical distribution (usually the normal distribution) and it is
evident from the raw data graphs in this paper that this
distribution does not exist for any roughness measure.
REFERENCES
[1] APR (2014a) Aircraft Simulation
http://www.aprconsultants.com/Aircraft-Simulation.html Accessed 20
March 2015.
[2] APR (2014b) Boeing Bump Index
http://www.aprconsultants.com/Boeing-Bump-Index.html. Accessed 20
March 2015.
[3] Boeing (2002) Runway Roughness Measurement Quantification and
Application – The Boeing Method. Pubn D6-81746. Airport Technology
Organization, Boeing Aircraft Company, Seattle.
[4] Chen, Y and Chou, C (2004) Assessment of aircraft's vertical responses
to develop the roughness evaluation index for airport pavement. FAA
Worldwide Airport Technology Transfer Conference, Atlantic City,
USA.
[5] FAA (2009a) ProFAA - the Federal Aviation Administration computer
program for roughness index analysis. User’s Manual (draft). Airport
Technology, Research and Development Branch, FAA. Washington.
[6] FAA (2009b) Guidelines and Procedures for Measuring Airfield
Runway Roughness. Advisory Circular AC150/5380-9. Washington.
[7] Gerardi, A (1977) Digital simulation of flexible aircraft response to
symmetrical and asymmetrical runway roughness. Technical Report
AFFDL-TR-77-37. Air Force Flight Dynamics Laboratory, Wright-
Patterson Air Force Base, Ohio.
[8] Gillespie, T (1992) Everything You Always Wanted to Know about the
IRI, But Were Afraid to Ask! Road Profile Users Group Meeting,
Lincoln, Nebraska. September 22-24.
[9] Goldman D and Von Gierke H. (1977) Shock and Vibration Handbook
Volume¬, Chapter 44. McGraw-Hill, New York.
[10] Horak E, Emery S and Morton, B (2010) Roughening up a wet
pavement surface. 4th SARF/IRF Regional Conference for Africa.
Somerset West, South Africa. 11–13 Oct.
[11] ICAO (2004) Annex 14 Aerodromes. Vol.1. Aerodrome design and
operations. 4th ed. International Civil Aviation Organisation, Montreal.
[12] ICAO (2009) Annex 14 Aerodromes. Vol.1. Aerodrome design and
operations. 5th ed. International Civil Aviation Organisation, Montreal.
[13] ICAO (2013) Annex 14 Aerodromes. Vol.1. Aerodrome design and
operations. 6th ed. International Civil Aviation Organisation, Montreal
[14] IPRF (2007) Airfield Concrete Pavement Smoothness – a Handbook.
Report IPRF-01-G-002-02-4. Innovative Pavement Research
Foundation, Skokie, IL.