Geologic and Geotechnical Effects of An Impact Caused by An Airplane Crash
Geologic and Geotechnical Effects of An Impact Caused by An Airplane Crash
www.elsevier.com/locate/enggeo
Abstract
A Turkish Airlines (THY) Boeing 737-400 plane crashed into alluvial soils creating an approximately 13 m deep and 30 m
wide crater near the village of Adatepe, Ceyhan in southern Turkey. Effects of the impact on the soils in and around the crater
were investigated from both the geological and soil mechanics point of view.
The results show that the impact caused severe deformations in the soils in and around the crater. The soils deformed similar
to metamorphic rocks seen at many terrestrial hypervelocity impact craters around the world and became overconsolidated up to
a distance of about 10 m from the crater wall as a result of the impact.
Also, the crash was recorded as a 2.7 magnitude earthquake by a nearby microtremor seismograph which provided both the
location (epicenter) and time of the crash which was not known immediately after the crash.
D 2005 Elsevier B.V. All rights reserved.
Keywords: Airplane crash; Impact; Earthquake; Overconsolidation; Preconsolidation pressure; Overconsolidation ratio
Fig. 1. Location (a) and geologic map (b) of the study area. Geology is adopted from Kozlu (1997).
262 H. Cetin / Engineering Geology 80 (2005) 260–270
Table 1
Some parameters of soils
Properties Soil
A Bk C Mixture
(ABkC)
Natural water 23–28 23–24 26–30 22–30
content (%)
Natural unit 1.74 1.87 1.90 1.86
weight (g/cm3)
Sampling depth (m) 0.4 5.0 8.5 0.4, 5, 8.5
Overburden pressure 0.07 0.93 1.59 0.07, 0.93,
(kg/cm2) 1.59
Liquid limit (%) 48.2 32.1 47.9 43.6
Plastic limit (%) 24.8 18.6 26.1 22.8
Plasticity index (%) 23.4 13.5 21.8 20.8
Specific gravity 2.75 2.70 2.76 2.73
Grain sizes (%)
Gravel 1.5 5.0 0.0 0.0
Sand 14.0 30.5 2.0 16.0
Silt 34.0 25.5 37.0 27.5
Clay 50.5 39.5 61.0 56.5
Soil type (USCS) CL CL CL CL
Description Medium Low Medium Medium
plasticity plasticity plasticity plasticity
clay clay clay clay
Fig. 4. The particle size distribution curves for the three soils and the mixture soil at the crater wall.
The aim of this study was to investigate geologic wide crater. Excavation in the crater revealed that
and geotechnical effects of this plane crash on these there are three different soil units, namely A, Bk and
soils exposed in and around the crash site. C, at the crash site and the impact caused severe
deformations in these three soils in and around the
crater (Figs. 2 and 3). The soils were welded or
2. Methodology mixed together in different amounts within the 1–2
m thick zone of the crater wall. A 20 m long, 8.5 m
When the airplane impacted into the ground it deep trench starting from the edge of the crater was
created an approximately 13 m deep and 30 m excavated in order to study the extent of the defor-
Fig. 5. Acceleration time–history of the impact records recovered from the Kurtkulagi (Ceyhan) station of TUBITAK-MAM (written
communication, TUBITAK-MAM, 1999).
264 H. Cetin / Engineering Geology 80 (2005) 260–270
mation in the soils. Undisturbed block samples were 3.2. Geological effects
taken from the mixed zone (namely ABkC) at the
crater wall (0 m distance) and soils A, Bk and C, at 3.2.1. Seismic
5, 10 and 20 m distances from the crater wall and The plane left Adana at 00:36 h local time and
depths of 0.4, 5 and 8.5 m, respectively, for conso- crashed shortly after take off. Officials both from the
lidation testing. The samples were wrapped with Boeing Company and the Transportation Ministry of
aluminium foil, coated with wax, wrapped in cheese Turkey did not know the exact time of the crash. I
cloth and coated again with wax. They were then indicated to them that I could give the coordinates of
immediately put in coolers, promptly transferred to the crash site to the Marmara Research Center of
the lab, and stored in a controlled humidity room. Scientific and Technical Research Council of Turkey
Some disturbed samples were taken for index prop- (TUBITAK-MAM) and ask if the crash was recorded
erty analyses (grain size and Atterberg limits) needed by their seismograph network of 17 instruments estab-
to classify the soils. lished in 1991, in order to study microearthquake
Consolidation (oedometer) tests were performed activity in the region. By means of this network, the
on the block samples in order to determine their TUBITAK-MAM had provided the most accurate
preconsolidation pressures. The tests were per- epicenter location of the Ms 6.3 Adana (Turkey)
formed following the American Society of Testing
Materials (ASTM) D 2435-90 (1993). Casagrande’s
(1936) method was used to determine the precon-
solidation pressure (maximum effective stress expe-
rienced in the past). The ASTM standard suggests a
loading schedule, which is obtained by doubling the
pressure on the soil to obtain values of approxi-
mately 250, 500, 1000, 2000, 4000, etc., f lb/ft2
(0.12, 0.24, 0.49, 0.98, 1.95, etc., kg/cm2). A
slower loading schedule of between 0.10 and 26
kg/cm2 was followed, to obtain a more precise void
ratio versus log of effective pressure curve, with a
distinct break, which is needed for more accurate
estimation of the preconsolidation stress values. The
consolidation apparatus consisted of a fixed-ring-
doubly-drained oedometer (ELE C-328) with a
ring of 50 mm in diameter and 20 mm in height,
a loading frame (ELE C-320A) and a data acquisi-
tion system.
earthquake in 1998 (Cetin et al., 1999). Interestingly, impact travelled the 12 km distance in about 3 s
the crash was recorded as a 2.7 magnitude earthquake indicating that the airplane had crashed at about
by one of their seismographs located at the village of 00:47:05 h local time. This indicates that missing
Kurtkulagi about 12 km south of the crash site at crashed airplanes lacking signal devices may be lo-
00:47:08 h local time, the epicenter being the crash cated if there are microtremor seismographs located in
site (Figs. 1 and 5). Seismic waves created by the the nearby region.
Fig. 7. The presumed sequence of events in the formation of the crater at the impact site [modified from Shoemaker, 1960].
266
H. Cetin / Engineering Geology 80 (2005) 260–270
Fig. 8. The void ratio versus log effective stress curves for the samples taken from the mixture soil ABkC at depths of 0.4, 5 and 8.5 m and soils A, Bk and C at depths of 0.4, 5 and
8.5 m, respectively, at various distances (a) 10 and 20 m, (b) 5 m and (c and d) 0 m from the crater wall. Unloading portions were not plotted for clarity.
H. Cetin / Engineering Geology 80 (2005) 260–270 267
Fig. 9. Average preconsolidation stresses in the mixture soil ABkC and soils A, Bk and C at various depths and distances from the crater wall.
Zero distance values were obtained on samples taken from the mixture soil ABkC at the same depths on the crater wall. There was no pure A,
Bk and C soils left at the sampling depths of 0.4, 5 and 8.5 m at the crater wall (zero distance) due to heavy deformation.
268 H. Cetin / Engineering Geology 80 (2005) 260–270
creeping that tends to deform the soil deposit (Voight, weight of deposits that were later eroded, the weight of
1966; Feda, 1978; Cetin, 1997). ice that later melted, or desiccation because of moisture
Depending on the Over Consolidation Ratio content changes. A change in the total stress or pore
(OCR), which is the ratio of the preconsolidation water pressure may also cause overconsolidation, both
stress to the existing vertical effective stress, soils changes altering the effective stress. Sometimes, over-
can be normally consolidated (OCR = 1), overconsoli- consolidation occurs because of alterations of the
dated (OCR N 1) or underconsolidated (OCR b 1) chemical and thermal environments of the soil (Holtz
(Holtz and Kovacs, 1981). and Kovacs, 1981). Shearing or faulting is another
Overconsolidation may be caused by different rea- reason why a soil might become overconsolidated
sons. Most overconsolidated soils were formed by the (Cetin, 1997).The preconsolidation stress can be deter-
Fig. 10. Overconsolidation Ratio (OCR) values in the mixture soil ABkC and soils Bk and C (a) and A (b) at various depths and distances from
the crater wall. Zero distance values were obtained on samples taken from the mixture soil ABkC at the same depths on the crater wall. There
was no pure A, Bk and C soils left at the sampling depths of 0.4, 5 and 8.5 m at the crater wall (zero distance) due to heavy deformation.
H. Cetin / Engineering Geology 80 (2005) 260–270 269
mined using the Casagrande (1936) (most popular) signal devices may be located if there are microtremor
method for curve construction. seismographs located in the nearby region.
The impact caused severe deformations in the soils
A, Bk and C in and around the crater. A 20 m long, Acknowledgements
8.5 m deep trench starting from the crater wall was
excavated in order to study the extent of the deforma- Some of the soil mechanics tests were done in the
tion in the soils. Undisturbed block samples were Soil Mechanics Laboratory of the Civil Engineering
taken from the mixed zone (namely ABkC) at the Department at Cukurova University. The author
crater wall (0 m distance) and soils A, Bk and C, at thanks Mustafa Laman for providing unlimited access
5, 10 and 20 m distances from the crater wall and to the laboratory. I would also like to thank Dr. Lund
depths of 0.4, 5 and 8.5 m, respectively, for consoli- R. Pusch and Terry R. West for their critical review of
dation tests. Void ratio versus log effective stress the manuscript.
curves for the consolidation tests are shown in Fig. 8.
Fig. 9 illustrates the average effective stresses (pre-
References
consolidation pressures) recorded in the soils A, Bk
and C at the sampling depths of 0.4, 5 and 8.5 m and ASTM D 2435-90, 1993. Standard test method for one-dimensional
distances of 0, 5, 10 and 20 m from the crater, consolidation properties of soils. Annual Book of ASTM Stan-
respectively. The preconsolidation stresses and, as a dards, vol. 04.08. American Society for Testing and Materials,
result, the Over Consolidation Ratio (OCR) values Philadelphia, PA, pp. 313 – 324.
decrease away from the crater indicating the soils Casagrande, A., 1932. The structure of clay and its importance in
foundation engineering. Proc. Contributions to Soil Mech.,
were anisotropically overconsolidated during the im- 1925–1940. Boston Society of Civil Engineers, Boston, MA,
pact (Figs. 9 and 10). The soils Bk and C which were pp. 72 – 112.
originally normally consolidated (OCR = 1) became Casagrande, A., 1936. The determination of the pre-consolida-
overconsolidated. Soil A was previously overconsoli- tion load and its practical significance. Proc. 1st. Int. Conf. on
dated (OCR = 8), probably as a result of desiccation Soil Mech. and Found, Eng., vol. 3, pp. 60 – 64. Cambridge,
Massachusetts.
since it is near the ground surface. It became even Cetin, H., 1997. How did the Meers fault scarp form? Paleoearth-
more overconsolidated (OCR = 36.3 at zero distance) quake or aseismic creep? A soil mechanical perspective. Engi-
which extended to a distance of about 10 m from the neering Geology 47, 289 – 310.
crater wall as a result of the impact (Fig. 10). Cetin, H., Demirtaş, R., Güneyli, H., Yetiş, C., 1999. The pre-
liminary report on the Adana, Turkey, earthquake of June 27,
1998. Association of Engineering Geologists (AEG) News 42
(1), 4 – 11.
4. Conclusions Feda, J., 1978. Stress in Subsoil and Methods of Final Settle-
ment Calculation. Elsevier Publishing Co., Amsterdam, The
When the Turkish Airlines (THY) Boeing 737-400 Netherlands.
plane crashed into the alluvial soils near the village of Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological
and genetic sequences of carbonate accumulation in desert soils.
Adatepe, Ceyhan in southern Turkey, it created an Soil Science 101, 347 – 360.
approximately 13 m deep and 30 m wide crater and Grieve, R.A.F., Pesonen, L.J., 1992. The terrestrial impact cratering
caused severe deformations in the soils in and around record. Tectonophysics 216, 1 – 30.
the crater similar to the deformations seen at many Holtz, R.D., Kovacs, W.D., 1981. An Introduction to Geotechnical
Engineering. Prentice-Hall, Eaglewood Cliffs, NJ.
terrestrial hypervelocity impact craters around the
Kieffer, S.W., 1971. Shock metamorphism of the Coconino sand-
world. The soils became overconsolidated up to the stone at Meteor Crater, Arizona. Journal of Geophysics 76,
distance of about 10 m from the crater wall as a result 5449 – 5473.
of the impact. Kozlu, H., 1997. Doğu Akdeniz bölgesinde yeralan Neojen basen-
Also, the impact was recorded as a 2.7 magnitude lerinin (İskenderun-Misis-AndVrVn) tektono-stratigrafi birimleri
earthquake by a nearby microtremor seismograph ve bunlarVn tektonik gelişimi (in Turkish). PhD Thesis (unpub-
lished), Çukurova University, Department of Geology, Adana,
which provided both the time and location (epicenter) p. 189.
of the crash. This was not known immediately after the Milton, D.J., 1968. Structural geology of the Henbury meteorite
crash, indicating that missing crashed airplanes lacking craters, Northern Territory, Australia. Geological Survey Profes-
270 H. Cetin / Engineering Geology 80 (2005) 260–270
sional Paper, 599-C. U.S. Department of the Interior, United phism in the basement of complex terrestrial impact craters:
States Government Printing Office, Washington, p. C17. implications for the cratering process. In: Boden, A., Eriksson,
Shoemaker, E.M., 1960. Penetration mechanics of high velocity K.G. (Eds.), Deep Drilling in Crystalline Bedrock. Springer,
meteorites, illustrated by Meteor Crater, Arizona. Report 21st New York, pp. 277 – 297.
Int. Geol. Congr., Session Norden, Part XVIII, pp. 418 – 434. Voight, B., 1966. Interpretation of in situ measurements. Proc. Int.
Stöffler, D., Bischoff, L., Oskierski, W., Weist, B., 1988. Struc- Soc. Rock Mech., Lisbon, Portugal, vol. 3, pp. 332 – 348.
tural deformation, breccia formation, and shock metamor-