Evaluation of Aircraft Landing Overrun Risk

Pasindu H.R
Graduate Student
Department of Civil Engineering
National University of Singapore
[email protected]

Fwa T.F.
Professor
Department of Civil Engineering
National University of Singapore
[email protected]

Ong G.P.
Lecturer
Department of Civil Engineering
National University of Singapore
[email protected]

ABSTRACT

Risk and safety have always been important considerations in civil aviation. According to
historical data, aircraft are subject to a higher accident risk during the landing phase than during
other flight phases. One of the major aircraft accident types during landing is overruns, and each
year a large number of landing overruns occur worldwide. Factors such as long landings, excess
approach speed, slippery runways, significant tailwind etc. increase the risk of a landing overrun.
In this paper several causal factors of landing overrun accidents were analyzed to evaluate its
impact on the landing stopping distance. The research used finite element simulation modeling to
analyze tire-pavement interaction on a wet pavement, and evaluated the skid resistance behavior
of an aircraft tire. This is taken into account for computation of aircraft braking distance during
landing in wet pavement conditions. It can be incorporated with the probabilistic nature of aircraft
operating characteristics such as speed, weight, touchdown position to compute the landing
distance variation in wet runway conditions. It presents a useful tool to assess how each factor
will affect the landing distance and subsequently the overrun risk. A numerical example is
presented to illustrate this procedure for a given aircraft type and runway conditions.

KEY WORDS

Aircraft Landing, Runway overrun, Risk, Skid resistance

INTRODUCTION

Aircraft landing is one of the critical phases of flight, where 25% of the fatal accidents take
place worldwide (Boeing 2008). A study which analyzed the total commercial aircraft accidents
(with major or substantial damage) from 1995-2008 showed that nearly 23% were landing
excursion accidents which constitutes of overruns and veeroff accidents (FSF 2009). Considering
the fact that, the annual overrun accident number has shown an increasing trend over the last few
years (Boeing 2008) and with the expected growth in international air traffic, overrun accidents
could continue to be a major issue of aviation safety and it is imperative that the safety standards
for runway overruns are improved in order to minimize the number of accidents

Overrun occurs when an aircraft landing distance exceeds the runway length available during
a landing ground roll. Several research studies on overrun risk have analyzed accident data and
attempted to identify the main causal factors for overrun accidents. The major risk factors related
to overrun accidents include wet pavement conditions, long landing, fast landings, presence of tail
wind etc. These factors are used to develop regression models to estimate the probability of an
overrun accident (Kirkland et al. 2004). The main obstacle of this method is lack of complete data
on accidents, limited data availability on incidents, and on normal operations.

The second approach adopts analytical methods to evaluate landing distance for different
runway conditions and aircraft characteristics and calculates overrun risk (Biggs et al. 2003,
ESDU 2008). Braking distance computation is a critical component of landing distance
evaluation, which requires estimation of runway friction coefficient especially under wet
pavement conditions. In these methods the parameters estimated from ground vehicle friction
measurements and correlating them to aircraft braking performance, or using empirical models
developed based on aircraft braking performance experimental data. The empirical methods have
limitation in applicability and the correlations between wet runway friction measurement and
aircraft braking performance is not yet well established. Therefore these could be considered as
shortcomings of the existing analytical based approaches. One possible improvement to these
existing methods of evaluating wet pavement aircraft tire frictional performance is a mechanistic
analysis of tire-pavement interaction on surfaces with fluid. It allows better understanding of how
each of the contributing factors such as water depth, aircraft speed, weight etc. would affect the
pavement frictional properties.

This paper proposes a procedure to compute aircraft landing distance under wet pavement
conditions. It incorporates a mechanistic based approach for evaluation of wet pavement
frictional performance of aircraft tire, namely skid resistance using finite element simulation. The
results are incorporated into the calculation of aircraft braking distance calculation. Other
operational parameters such as touchdown position, touchdown speed are also included in the
analysis to compute the overall landing distance. This analysis can be used to assess how runway
conditions such as surface water film thickness, and aircraft operational characteristics such as
touchdown speed, and location affect the landing distance and overrun risk.

AIRCRAFT LANDING OVERRUN RISK FACTORS

Aircraft landing stopping distance varies significantly depending on the aircraft operational
characteristics such as touchdown speed, location, landing technique, aircraft braking type,
reverse thrust usage and also the airport elevation, runway conditions, wind conditions etc. A
research carried out in the Netherland Aerospace Research Institute (van Es 2010) identified the
most important causal factors for landing overrun based on runway excursion data from 1990 -
2008. The top 6 risk factors are wet or contaminated runway, long landing, speed to high,
incorrect decision to land, hydroplaning, tailwind gear malfunction, which were present in 58.8%,
38.9%, 19.9%, 16.3%, 13.8% and 13.6% of the landing overruns respectively. The significant
impact wet pavement conditions have on overrun accidents were also established from a study on
landing overruns in Canada and the US where 37% of overruns occurred in wet conditions (FSF
2010).

Pavement friction level depends on several factors such as wheel load, speed of the aircraft,
tire pressure, tire type and tread design, pavement surface characteristics and surface water film
thickness. Other factors being constant, pavement friction decreases with the increase in surface
water film thickness (Horne and Leland 1962, Horne 1976). More importantly the reduction in
friction is greater at higher speeds in wet pavement conditions compared to dry pavement
conditions. Compared to highways this is even more crucial in the context of aircraft operations
on runway due to reasons that (a) Vehicles on road can drastically decrease its speed during the
brief periods of heavy rainstorms, however aircrafts have to maintain certain speeds for takeoff
and touchdown speed of exceeding 120 knots = 222 km/h. (b) The drainage length of a runway is
substantially longer than a road section (more than twice). Water film thickness on a surface is
dependent on the drainage path, hence a heavier rainfall concentration can produce deeper water
films on runways. Under normal circumstances an aircraft is expected to touchdown between
1000 to 1400 ft (304.8m to 426.7m) from the threshold. A long landing occurs if this increases by
a considerable margin, e.g. touchdowns of more than 2000-2300 ft (609.6m to 701.04m) from the
threshold or 25-33% of the runway length (van Es et al., 2009). A long landing increases the
required landing distance and consequently available margin in landing distance reduces. An
excess approach speed condition exists when the calibrated air speed at or near the threshold
exceeds Vref + 20 kts. The speed flown at the threshold has a dominant influence on the landing
distance (van Es et al., 2009).

Often several risk factors occur during a landing overrun accident. The presence of multiple
risk factors has a synergistic effect which could significantly increase the probability of a runway
overrun accident. This wet pavement conditions can compound the effects from other operational
factors such as long landing and fast landing on the runway. Therefore it is important to analyze
the aircraft landing performance on wet runway conditions while incorporating the operational
characteristics such as touchdown speed and location, in order to assess how each will affect the
aircraft landing distance and which in turn will influence the potential for overrun accidents.

COMPUTATION OF AIRCRAFT LANDING STOPPING DISTANCE

Aircraft operational characteristics as well as external conditions can vary for each landing,
which will affect the landing distance of an aircraft. Therefore calculation of landing distance
needs to take into account the important factors and their respective uncertainty. For this analysis,
landing distance will be computed for a regional aircraft characteristics of which and the other
parameters used in the calculations are given in Table 1.The values are taken from landing survey
data and manufacturer’s manuals (Ho Sang 1975, Jackson 2001).

TABLE 1: Input Parameters Used for Numerical Example : Aircraft Type : Regional Jet

Main Gear Type : Dual Maximum landing weight = 21319 kg

Landing weight used for analysis 20000 kg
Wing Area 54m2
Air density 1.224 kg/m3
Touchdown speed (m/s) Normal distribution: Mean = 55, Standard Deviation = 7
Touchdown location (m) Normal distribution: Mean = 300, Standard Deviation = 50
Aircraft landing distance calculation includes three main phases as shown in Figure 1,

Touchdown Braking
speed Vt speed Vb V= 0
50ft

Air distance Delay Braking

Threshold

Sa distance Sd distance Sb
Stopping distance SL
Wet stopping distance req.1.92SL

Figure 1: Aircraft landing distance phases

1. Air distance - distance travelled from 50 feet (15.25m) above the runway to the point of
touchdown.
Landing parameter surveys conducted at several airports by FAA (Barnes 1998) gives data for
aircraft touchdown position variation along the runway. Data from such studies can be used to
model the touchdown location distribution for a given aircraft type. Therefore the air distance Sa
in the calculation is represented a random variable f(Sa).

2. Transition distance -The transition distance is the distance travelled from touchdown point to
the onset of braking application. The reduction in speed (k) from the touchdown speed during this
period is around 2m/s for most aircraft types and the delay is estimated to be around 2 seconds
(Trani et al. 1995). This reduction in speed from the touchdown speed Vt can be estimated as
follows,
Vb = Vt - k (1)
where, Vt = aircraft speed at touch down, Vb = aircraft speed at beginning of braking application
k is the speed reduction for the aircraft considered.
From previous landing survey data (Barnes 1998) the aircraft operational characteristics such as
touchdown speed is known to vary. Therefore the distribution of touchdown speed Vt can also be
derived from survey data, and represented as a random variable following a probability
distribution function f(Vt). The delay distance can be computed as follows,
Sd= 0.5 (Vt + Vb) td (2)
where, td= delay time

3. Braking distance - The braking distance is calculated as the distance travelled from the onset
of application of brakes to aircraft stopping. The calculation of braking distance must consider the
effects of the following facts: (a) runway pavement frictional level that determines the tire
braking force available, (b) aerodynamic drag forces acting on the aircraft, (c) use of reverse
thrust, and (d) braking efficiency. For this analysis, it is assumed that no reverse thrust is used,
and the case of locked wheel is considered. Since the locked wheel condition gives rise to the
minimum pavement friction, the computed braking distance will represent the worst scenario and
provide a conservative estimate for the actual braking distance.
 Applying the laws of motion to the aircraft, we can compute the deceleration rate during
braking phase on a runway with zero gradient, with the assumption that both drag and lift
coefficients are independent of speed.
D + µ (Mg -L)= M a (3a)
where, M = mass of the landing aircraft, and the forces acting on the aircraft during ground roll
are defined as follows,
2
Uplift Force, L = 0.5 ρ v A CL (3b)
Drag Force, D = 0.5 ρ v 2 A CD (3c)
where, v = speed of aircraft, ρ = density of air, A = wing area, CD = coefficient of drag, CL =
coefficient of lift, µ = coefficient of friction, g = gravitational acceleration, a = decelerate rate
The braking distance for an aircraft is given by the following equation,

  0.5 ρ [v (t)]2 A (C D - [µ (t)] C L )  

T
S= ∫ Vb -  [µ (t )] g + t  dt (4)
  M  
0  
where, T is the total time during braking phase and Vb= braking velocity
The coefficient of friction is related to skid number, SN by the relationship, µ = 0.01 (SN). As
explained earlier, the skid number depends on the following factors: aircraft speed (v), wheel load
(w), tire pressure (p), surface type (static friction coefficient- SN0 ), water film thickness (tw).
Unlike vehicles, for aircraft the wheel load will vary due to the uplift forces acting on the aircraft
during ground roll. Assuming 95% of the total aircraft weight is on the main gear, the wheel load
w is calculated from the following relationship,
w = 0.95 [(Mg –L)/n] (5)
where, n is the total number of wheels in the aircraft main gear. Similarly, the wheel load of the
nose wheel can be calculated accordingly.
Hence, skid number (SN) can be written as a function of all the above factors,
SN = f ( v, M, L, p, SN0, tw ) (6)
The total stopping distance is given by the summation of results for the three phases during
landing.
S = Sa + Sd + Sb (7)
Braking distance was computed using a numerical step by step integration method. Considering
the probabilistic nature of the input variables representing aircraft operational characteristics, the
final results will also be presented in the form of a frequency distribution.

Aircraft Tire - Wet Pavement Skid Resistance Evaluation

A three-dimensional finite-element skid resistance simulation model based on engineering
mechanics and fluid dynamics theory was developed by the authors for passenger car and truck
tires (Ong and Fwa 2007, Ong and Fwa 2010) was adopted in this study to analyze skid resistance
of aircraft tire. The finite element analysis computer software, ADINA (Adina R&D Inc, 2009) is
used to solve the coupled tire-fluid interaction problem. The model consists of three main
components, namely the pneumatic tire sub-model, the pavement surface sub-model, and the fluid
sub-model. The simulation considers interactions between these sub-models, i.e. tire-pavement
contact modeling, tire-fluid interaction, and fluid-pavement interactions. Figure 2 shows the
finite-element mesh of the tire and the pavement surface. The details of skid resistance analysis
for aircraft tire are presented in authors’ earlier works (Pasindu et al. 2011).
 The skid resistance at any speed v, computed as skid number SN, is given by the following
equation,
F
SN = 100 × x (8)
Fz
where Fx is the horizontal resistance force to motion, and Fz is the vertical wheel load acting on
the tire. The horizontal resistance force Fx is equal to the sum of the traction forces developed at
the tire-pavement contact and the fluid drag forces due to the tire-fluid interaction. The vertical
load Fz is an input parameter and its value remains constant throughout the simulation.
For the purpose of the analysis, a 32 x 8.8 Type VII smooth aircraft tire, with a pressure of 1104
KPa is considered for analysis of skid resistance behavior. The tire size matches approximately to
the main gear tires used by regional aircrafts such as Canadair RJ-200, ATR-42 etc. The static
coefficient of friction SN0 = 50, and the skid resistance is evaluated for a water film thickness of
1mm and 5 mm.

Tire Inflation Pressure on Inside Walls (p)

Tire Rim (Only vertical

Vertical Load displacements are allowed)
(w) on Tire

Fluid Structure
Pressure Outlet at Interaction Surface
Atmospheric Pressure

Fluid Structure Pavement Surface

Interface (Moving wall with
Speed (V)
Water Velocity Inlet with speed V)
speed V
Pavement Surface (Rigid surface)

Figure 2: Finite-element model of aircraft tire and pavement surface

RESULTS

The results of aircraft landing stopping distance calculation for different water film thickness
are presented in Figure 3 calculated following the methodology detailed in the previous section.
The mean landing distance on a pavement with 1mm and 5mm water-film thickness are 933m
and 1012m respectively. As expected the computed landing stopping distance increases by 8.4%
due to the increase in water film thickness, i.e. increases the overrun risk. This is due to increase
in braking distance as a result of decrease in wet-pavement skid resistance at higher water film
thickness. The percentage increase is approximately 20% if the braking distances for the
respective water film thicknesses were compared. The variation of aircraft landing distance with
 touchdown speed and distance is given in Figure 4. It shows that touchdown speed has more
significant impact than touchdown distance on overall landing distance. The calculated results
represent the worst case scenario as it has not factored in additional braking available for the
aircraft by use of reverse thrust and anti-skid braking systems, etc. However, the results clearly
demonstrate the impact wet-pavement conditions as well as aircraft operational characteristics
have on landing distance, and gives an indication as to how overrun risk would increase from
changes in them.

0.18
1mm
0.15 5mm

0.12
Probability

0.09

0.06

0.03

0.00
400 700 1000 1300 1600 1900 2200

Aircraft Landing Stopping Distance (m)

Figure 3: Aircraft Landing Stopping distance distribution for
water film thickness of 1mm and 5mm

2500 1750

1500
Stopping distance (m)
Stopping distance (m)

2000

1250
1500
1000
1000
750

500 500

0 250
100 200 300 400 500 45 55 65
Touchdown distance (m) Touchdown speed (m/s)
Figure 4: Aircraft Landing Stopping distance variation with touchdown speed and location

CONCLUSION

The paper presents a methodology to compute aircraft landing distance under wet-pavement
conditions. This approach incorporates a mechanistic based analysis and uses finite element
simulation to evaluate aircraft tire wet-pavement skid resistance. The main advantage is that it
can incorporate the effects of the key factors such as water film thickness, wheel load, pressure,
and surface condition into the analysis of skid resistance and braking distance. The analysis also
considers the probabilistic nature of aircraft operational characteristics such as touchdown speed
and weight. It is clear that wet pavement conditions and fast landings in particular present a major
hazard for safe runway operations. This procedure offers an improved understanding of the
factors affecting aircraft landing distance. It presents a useful tool to evaluate the overrun risk
under different runway conditions and aircraft operating characteristics and can be used to
improve the overall safety of aircraft landing operations.

REFERENCE

ADINA R&D Inc. (2009) ADINA Theory and Modeling Guide Volume I: ADINA Solids and
Structures, ADINA R&D Inc. Watertown, Massachusetts.
Barnes, T., DeFiore, T., and Micklos., R. (1998)Video Landing Parameter Survey—
Washington National Airport, Publication DOT/FAA/AR-97/106. Federal Aviation
Administration, U.S. Department of Transportation.
Biggs, D.C., Hamilton, G.B. and Owen, K.D.J. (2003) Benefit-Cost Analysis of Procedures for
Accounting for Runway Friction on Landing, Publication TP 14082E. Transport Canada,
Quebec.
Boeing. (2008) Statistical Summary of Commercial Jets Airline Accidents, Worldwide
Operations 1959 - 2008, Boeing. http://www.boeing.com/news/news/techissues/pdf/
statsum.pdf.
Engineering Sciences Data Unit (ESDU). (2008) Example of risk analysis applied to aircraft
landing distance, ESDU Data Item No. 080005.
Flight safety foundation (FSF). (February 2010). Aero Safety world - Slippery When Wet.
Flight Safety Foundation (FSF). (May 2009). Runway Safety Initiative -Reducing the Risk of
Runway Excursion.
HoSang, V. (1975) A. Field Survey and Analysis of Aircraft Distribution on Airport
Pavements, Report FAA-RD-74-36. Federal Aviation Administration, U.S. Department of
Transportation.
Jackson, P. (2001) Jane’s All World Aircrafts 2001-2002, Jane’s Information Group. McGraw
Hill, New York, 2001.
Kirkland, I.D.L., Caves, R.E., Humphreys, I.M., and Pitfield, D.E. (2004) An Improved
Methodology for Assessing Risk in Aircraft Operations at Airports, Applied to Runway
Overruns. Safety Science, Vol. 42, pp. 891–905.
Ong, G. P., and Fwa, T. F. (2010) Modeling Skid Resistance of Commercial Trucks on
Highways. Journal of Transportation Engineering, Vol. 136, No.6, pp. 510–517
Ong, G. P., and Fwa, T. F. (2010) Wet-Pavement Hydroplaning Risk and Skid Resistance:
Modeling. Journal of Transportation Engineering, Vol. 133, No.10, pp. 590–598.
Pasindu, H.R., Fwa, T.F., Ong, G.P. (2010) Computation of Aircraft Braking Distance,
Transportation Research Board 90th Annual Meeting Compendium of Papers (CD-ROM),
Transportation Research Board, Washington, D.C.
Trani, A. A., B. J. Kim, X. Gu, and C. Zhong. (1995) Runway Exit Designs for Capacity
Improvement Demonstrations- (Phase III), Virginia Polytechnic Institute and State University,
Blacksburg.
Van Es, G.W.H. (2010) A Study on Runway Excursions from European Perspective, NLR-
CR-2010-259. Netherlands, NLR Air Transport Institute.
Van Es, G.W.H., Tritschler, K., and Tauss, M. (2009) Development of Landing Overrun Risk
Index, NLR-TP-2009-280. Netherlands, NLR Air Transport Institute.