James S.

Wilbeck1 The Development of a

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

Senior Engineer.

James L. Rand
Substitute Bird Model
Staff Engineer. A comprehensive program was conducted to develop a model synthetic bird for use
Mem. ASME in engine blade impact testing. A hydrodynamic theory of the impact event was used
to aid in determining the bird properties which had to be duplicated in the model.
Division of Engineering Sciences Of the two candidate models studied extensively, it was determined that a projectile
Southwest Research Institute, fabricated from commercial gelatin impregnated with phenolic mircoballoons most
San Antonio, Texas 78284 nearly duplicated the impact loading history of real birds.

Introduction
One of the major hazards to flight safety today is the in- concentrated on developing a fundamental understanding of
flight impact of birds. Aircraft windshields and engine the impact process. Based on this understanding, a general
blading are especially vulnerable to damage. Because of the hydrodynamic theory was developed to predict the impact
importance of this problem, both NASA and the Air Force loads. The selection of a substitute bird model was based on
are conducting extensive programs aimed at improving the those parameters identified in the hydrodynamic theory as
impact tolerance of these components by developing better necessary to simulate the loadings observed from the impact
materials and better structural designs. of actual birds. The development of a hydrodynamic model of
Extensive impact testing is required to evaluate these new bird impact has been reported previously [1, 2]. The purpose
concepts. Each of the major engine manufacturers has a test of this paper is to document the selection and validation of a
facility in which dead birds are impacted against single and substitute bird model. A brief overview of the hydrodynamic
multiple blades for design studies and ingested into entire model will be given to help the reader to better understand the
engines during qualification testing. The use of real birds in selection process.
the testing process lends realism to the test results, but greatly
complicates the testing procedure while raising several Blade Response
questions as to the validity of the results.
One of the major drawbacks to the use of real birds in The response of any structural element, when subjected to
testing is the lack of repeatability. The actual impact loads dynamic loading, is significantly different than its response to
vary from test to test due to the inability to control projectile a static force. It is assumed that the impact of a bird against a
orientation; lack of homogeneity, isotropy, or symmetry; and rotating engine blade will create forces for a finite duration of
variations in bird structure. Secondly, it is very difficult to time which is less than the natural period of the blade in any
conduct scaled tests since small birds, such as starlings, are given mode. Since the blade is a deformable medium, stress
not exact replicas of much larger birds, and it would be waves are generated which tend to mitigate the applied forces
difficult to find a small scale model of a starling. Testing with and spread the effects over a larger area and for a longer
real birds also has obvious sanitary and aesthetic disad- period of time. Under this assumption, engine blading is
vantages. susceptible to failure in two distinctly different modes when
impacted by a bird. Localized failure of the blade may occur
For these reasons, a program was conducted at the Air when the stresses generated are sufficiently high to cause
Force Materials Laboratory (AFML) to develop a synthetic compressive (or shear) failure at the point of contact. In
bird model for use in impact testing. Requirements for such a addition, release waves may cause back surface spall or tensile
model were that the loads generated during impact of the delamination. The second mode is gross structural defor-
model with a deformable structure be similar to those mation which may be sufficient to cause failure at some point
generated during the impact of a real bird, the loads should be other than the point of contact.
highly reproducible, and the model should be easily
manufactured, readily available, and relatively inexpensive. Typical examples of local failure are delamination of the
In order to develop a synthetic bird model which would leading edge (in composite blades), local cracking or
properly replicate the impact loads, initial efforts were cratering, back surface spall, and internal crack nucleation.
Material failures of this nature are governed by two
dominating considerations: the strength of the material under
the existing conditions of stress, strain and strain rate, and the
mechanics of the loading which may produce adverse wave
1
Formerly of the Metals and Ceramics Divisions, Air Force Materials interactions.
Laboratory, Wright-Patterson Air Force Base, Ohio 45433. In order to duplicate local failures of this nature, use of the
Contributed by the Gas Turbine Division and presented at the International actual blade to be evaluated will serve to replicate the material
Gas Turbine Conference and Products Show, Houston, Texas, March 9-12,
1981, of THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS. Manuscript and geometric properties if it is rotated to produce the actual
received at ASME Headquarters, November 26, 1980. Paper No. 81-GT-23. stress state. In order to duplicate the mechanics of loading

Journal of Engineering for Power OCTOBER 1981, Vol. 103/725

Copyright © 1981 by ASME

 with a substitute bird, it is necessary to produce not only the cylinder with hemispherical ends and a length-to-diameter
identical pressures generated during impact, but also the time ratio of two. The target orientations were both normal and
history of that pressure. This feature of the impact process is oblique to the trajectory of the bird. The test procedures are
important because finite amplitude waves generated by documented more fully in [2] and [4].
contact with the blade will subsequently be reflected from Typical pressure traces recorded by the plate transducer
discontinuities in the blade and may interact with other waves located at the center of impact are given in Fig. I. Several
to cause intense stresses within the blade. distinguishing characteristics can be seen. The pressure is
This type of failure will generally be caused during the dominated by an initial peak pressure followed by a lower,
initial phase of the impact process when very large pressures approximately constant pressure. The pressure profile
are generated for a very short period of time. The initial measured by transducers away from the center of impact
Hugoniot pressure created by the impact of a bird must not displayed similar characteristics, but with a decrease in the
only be duplicated, but the rapid release of this pressure at the amplitude of pressure with radial distance from the center of

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

correct time must also be duplicated. impact. Another outstanding feature of the pressure profile
Gross structural failure occurs when the momentum im- was the observation that the duration of impact, as evidenced
parted to the blade is sufficient to cause blade deflections of by the duration of the pressure trace, was equal to the time
sufficient magnitude to cause bending or tensile failures. Root that it takes for a bird to travel through its length. This can be
failure, tip break off, and blade interference are typical expressed analytically as
examples of this type of response. Studies by various groups,
such as Tsai, et al. [3], have shown that the long time stresses (1)
and deformations in a blade are proportional to the impulse u0
imparted to the blade during impact. where 7 is the duration of impact, L is the length of the bird,
The response of structures to impulsive loading is at least and U0 is the initial impact velocity. An extensive report of
phenomenologically understandable. Since the blade has the pressure measurements is given in [5].
mass, conservation of momentum requires that a change in Several comments about the pressure measurements should
velocity of that mass must occur locally. The integral effect of be added at this point. The lack of axial symmetry and
that local change is the total impulse applied to the blade. In homogeneity of the bird caused large variations in the
the absence of any further external forces, the blade will move measured pressure profiles as seen in Fig. 1. These variations
to equilibrate the local changes in velocity. This may create preclude exact replication or representation of a bird impact.
large tensile stresses near the root due to bending or large It should also be noted the duration of the peak pressure
shear stresses near the tip. appears to have been so short that it is doubtful if the pressure
It is apparent that a substitute bird model should be capable transducers could accurately measure the peak amplitude.
of reproducing not only the total impulse of the actual bird This would explain the differences seen in Fig. 1(a) and 1(b).
impact process, but also the pressure history across the impact
surface. This would be accomplished if pressures generated by
the substitute bird have the same magnitude of initial Formulation of Theory
pressure, time of release of the initial pressure, and the in- Since the pressures measured during the impact tests were
tegral of the pressure over the total time of contact. When so much greater than the reported strength of the bird's flesh
these conditions are duplicated, both local and gross struc- and bone [6], it was felt that the impact process could be
tural failure modes will be apparent if they exist. modeled analytically by a hydrodynamic theory. A theory was
developed in [1] in which the bird was approximated by a
Bird Impact Tests cylinder of fluid. This theory was based on previous studies of
Initial efforts in the AFML program were aimed at water drop and water jet impacts ([7] through [11]).
developing a fundamental understanding of the bird impact At the initial instant of impact of a fluid cylinder impinging
process. An extensive program was conducted in which normally against a rigid target, a shock wave is formed at the
chickens were impacted against rigid targets. Chickens were target-projectile interface. As this shock wave propagates into
used to represent birds because of their widespread use in the projectile, the pressure at the surface is given by the
qualification testing of aircraft parts, including engines. Rigid relationship
steel targets were used to eliminate the effects of target PH=P0USU0 (2)
displacement on the amplitude and duration of the impact
pressures. Tests were conducted at AFML and at Arnold where PH is the Hugoniot (shock) pressure, p0 is the initial
Engineering Development Center (AEDC). density of the fluid, Us is the shock velocity, and U0 is the
initial impact velocity. For impact on a rigid target, the shock
A compressed air gun located at AFML was used to launch velocity (which increases with impact velocity) is always
birds ranging in mass from 60 to 600 g. A much larger greater than the impact velocity. For most fluids, the shock
compressed air gun, located at AEDC, was used to launch velocity is, in fact, much greater than the impact velocity, so
birds ranging in mass from 1 to 4 kg. The launch velocities that
obtained from both guns varied from 50 to 300 m/s. The
targets were flat, rigid plates with pressure transducers em- PH>>PoU02 (3)
bedded flush with the surface. The birds were launched such The importance of this relationship will be seen later.
that they would impact end on, rear end first. This orientation The duration of the shock pressure at the target face
resulted in a projectile which could be approximated by a depends on the shock release process. At the first instant of

Nomenclature

k = compressibility U = velocity, m/s

L = projectile length, m p = density, kg/m 3 0 = initial conditions
P = impact pressure, MPa Subscripts s = shock value
T = time, s H = Hugoniot (shock) value ss = steady-state stagnation value

726 / Vol. 103, OCTOBER 1981 Transactions of the ASME

 EXPERIMENTAL
i «•
——r—
DATA m
100
I 250
O - 4.0 kg -
a. Q - 2.0 kg
200 m -
~A - 1.0 kg

essurt
• - 0.5 kg
25 .50 .75 1.00 150 •
~ • - 0.1 kg -
T U„ A
/°

I 30TU
A
100 f • •
-
(a) Impact velocity, U 0 = 225 m/s •

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

50 -
y"o A • •
0 -<£_ _J I i
0 100 200 300

Impact v e l o c i t y , U (m/s)

Fig. 2 Shock pressures measured at the center of impact [1]

pressures, an approximate equation of state was developed to

represent the bird. Although chickens are neither
homogeneous, isotropic, nor axisymmetric, it was felt that
they could be reasonably approximated by a cylinder of
material which was both homogeneous and isotropic. Such a
(a) Impact velocity, U 0 = 197 m/s simplification was required to properly apply one-di-
Fig. 1 Typical pressure data, center transducer, for the normal impact mensional impact theory.
of a bird [1]
In line with the simplified model of a bird, the initial
impact, the high shock pressure exists all along the impact decision was to represent the equation of state of a bird with
surface. At the boundaries of the impact surface, the material that of water. This was a logical first choice since water makes
is bounded on one side by a high pressure and on the other up a very large portion of body mass, it has low strength, and
side by a free surface. This results in an outward acceleration its equation of state is well characterized. Since it was not
of the material, thereby developing a pressure release wave possible to launch a projectile made up entirely of liquid
which propagates inward towards the center of impact. This water, projectiles were fabricated from ice and gelatin.
release wave causes a dramatic decrease in the pressure at the Subsequent impact tests found that these projectiles gave
target face. As a result, the duration of the high shock shock pressures much greater than those obtained during bird
pressure is greatest at the center of impact and is on the order impact. The pressures were so great, in fact, that several
of only tens of microseconds. pressure gages were destroyed. From these results, it was
After the release process is completed, the impact process inferred that pure water was not a good model of a bird.
attains an approximate steady state condition. During this Further insights were gained in laboratory measurements of
steady state, the pressure on the target surface at the center of bird density. An average density of 0.90-0.95 g/cm 3 was
impact is the stagnation pressure, Pw, is given by Bernoulli's found for chickens of various sizes. These measurements of
relationship average density were in contrast to those of Tudor [6] who
found the density of various parts of the chicken body to be
'</'.»• + ''<>> dP U02
(4) approximately l .06 g/cm 3 . From this it was inferred that bird
Jp0 p 2 flesh has a density of approximately 1.06 g/cm 3 , but that a
P0 and U0 are the pressure and velocity of the uniform flow whole chicken, with its voids and cavities, has an average
field some distance away from the impact surface and are density of only 0.90-0.95 g/cm 3 .
approximated by the atmospheric pressure and initial impact The existence of voids in the bird serves not only to lower
velocity. Equation (4) can be solved to give the average density but also to markedly decrease the speed of
propagation of waves. The resulting lower shock velocity
Pss = k Po t V (5) would account for the much lower shock pressures measured
For an incompressible fluid, k = Vi. However, for most for bird impact than for the impact of ice or gelatin.
materials, the density tends to increase with applied pressure, The equation of state for the bird had to be modified to
so that k will have values which may approach 1.0. Com- account for the existence of the voids. This was done by
parison of equations (3) and (5) shows that, for most treating the bird as a cylinder of water with entrapped air.
materials, the shock pressure is much greater than the steady Mixture theory [2] was used to develop a constitutive
state pressure measured at the center of impact (as was the relationship for a homogeneous solution of water and 10-15
case in Fig. 1). Away from the center of impact, the steady percent air (i.e., water with 10-15 percent porosity). By in-
flow pressure decreases with radius. creasing the density of the water in the model to 1.06 g/cm 3 ,
For the impact of a fluid jet against a rigid target, the the resulting mixture had a density of 0.90-0.95 g/cm 3 , which
duration of impact can be approximated by equation (1). corresponds to that measured for chickens.
Inherent in this relationship is the assumption that the fluid An extensive program was conducted to validate the ap-
does not decelerate during impact. This is normally the case propriateness of the derived equation of state [1, 2). Figures
when the impact pressures greatly exceed the strength of the 2 and 3 display the values of the peak shock pressure and the
projectile material (as in bird impact) and the target is ef- average steady-flow stagnation pressure measured at the
fectively rigid (the displacements are small). center of the target during normal bird impact. The predicted
In order to use equations (2) and (4) to predict the impact shock pressures in Fig. 2 agree well with the values measured

Journal of Engineering for Power OCTOBER 1981, Vol. 103/727

 40 i 1 -l r~
EXPERIMENTAL 6 -
DATA
O - 0.06 kg o
/ & 4 -
30 Q.
" A - 0.6 kg
• - 1.0 kg A 2 -
20 - • - 2 . 0 kg
• • 0
A - 4.0 kg

10
T U

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

| I i i

50 100 150 200 250 300 Fig 4 Normal impact of a gelatin projectile with 40 percent porosity,
U0 = 158 m/s [2]
Impact v e l o c i t y , U (m/s)

Fig. 3 Steady flow pressures measured at the center of impact [1 ]

50 1 1 1 1
1 j
for the large birds. However, the results for smaller birds No P o r o s i t y , /a
Eq. ( 4 )
show significant departures from prediction. This
discrepancy, which appears to decrease with increasing bird 40 - I
size was not considered to be a fault of the theory or the
analytical bird model chosen. As mentioned earlier, this was
probably caused by the very short duration of the shock phase
EXPERIMENTAL DATA
/ /°
and the limited response time of the pressure transducers 30 -- D - No Porosity
/
used. O - 40% P o r o s i t y
The experimental data displayed in Fig. 3 were obtained by
measuring the initial amplitude of the steady pressure region
seen in the pressure traces of Fig. 1. Good correlation was ~ 40%
20 - Ja/° Porosity,
obtained between theory and experimental data. It should be
Eq. (4)
noted that large data scatter was observed. The existence of
this variability in measured pressures was one of the major
reasons behind the desire to develop a model substitute bird 10 -
for testing.

Model Development
In the development of a model (synthetic bird), a variety of ***L l i i 1 1
requirements was formulated. In order to accurately 50 100 150 200 250 300
reproduce impact damage using a model, it was required that
Impact velocity, U (m/s)
the loads generated during impact with a deformable structure
be similar to those generated during the impact of a real bird.
Fig. 5 Stagnation pressure measured during the normal impact of
As an aid to design and testing, it was desired that the loads be gelatin projectiles [2]
highly reproducible. For practicality, the model should be
easily manufactured, readily available, and relatively inex-
pensive. simulant, gelatin was found to possess the desired properties
Two separate approaches were taken in developing suitable of a replica model. With its high water content, it was believed
models. One approach was to develop a near replica model in to possess an equation of state similar to that of bird flesh. It
which the synthetic bird was made of a material having an has a density of 1.05-1.06 g/cm 3 and its strength is great
equation of state similar to that of bird flesh, with the same enough to hold its shape but small relative to impact
average density and porosity used earlier to model real birds. pressures. The proper amount of porosity (entrapped air) can
Since the material properties of a replica model would be also be added by the use of phenolic microballoons, which are
similar to those of a real bird, the impact loads should be stirred into the gelatin and water mixture prior to the gelling
similar. A second approach was to develop a nonreplica process.
model. In this development process, emphasis was placed on During initial experimental investigations, projectiles were
developing a model which duplicated the impact loads, not the made from gelatin with no porosity (p = 1.06 g/cm 3 ) and
material properties. large porosity (40 percent porosity, p = 0.64 g/cm 3 ). It was
felt that this large variation would be a good check of the
Replica Model. As mentioned earlier, a chicken can be hydrodynamic theory. Figure 4 shows a typical pressure trace
adequately modeled by a mixture of 85-90 volume percent for the impact of a projectile with 40 percent porosity. The
water and 10-15 percent air, where the water has a slightly coordinates were nondimensionalized to aid in comparison
higher density of 1.06 g/cm 3 . However, no scheme could be with theory. Predicted values of the Hugoniot shock pressure
developed for fabricating a synthetic bird from such a mix- (equation (2)) and the steady-state stagnation pressure
ture. Subsequently, an effort was made to locate another (equation (4)) are also shown.
substitute with properties similar to those of flesh. A large Results from this series of tests are given in Figs. 5 and 6.
number of materials were considered. Synthetic rubbers of The steady-state stagnation pressures are compared with the
various types showed promise but had too great a density. predictions of equation (4) in Fig. 5 and the shock pressures
Finally, commercial gelatin was selected as the best choice. are compared with the predictions of equation (2) in Fig. 6.
Initially considered because of its historical use as a flesh The correlation is seen to be very good in both cases. No

728/Vol. 103, OCTOBER 1981 Transactions of the ASME

 350 1 -1 1•—-J—1 50 1 ' " 1 1 1 1 1 1
1 1—

EXPERIMENTAL DATA
300 -
" — No Porosity,
O - 50% Porosity
Eq. (2) 40 _
- _
250 -

EXPERIMENTAL DATA
200 -
O - 40% Porosity 30 No Porosity, - -
Eq. (4)
150 - / -

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

100 / -
O Jf ^- 50%
A0 Porosity
50 Eq. (4)
rf*^ Porositv, - 10
Eq. (2)'
c_—m*^^*! 1 1 1
50 100 150 200 250 300
•001 1 1 i 1 1
Impact velocity, U (m/s)
50 100 150 200 250 300
Fig. 6 Hugoniot pressures measured during the normal impact of
gelatin projectiles [2] Impact velocity, (m/s)
Fig. 8 Stagnation pressure measured for the normal impact of RTV-
560 projectiles [2]

—1 T- 1—_ , ,... ,_

300 -No Porosity, ~"\ /

Ea.._(4) Eq (2)

250 -
EXPERIMENTAL DATA

T U 200 - O - 50% Porosity -

c
L

Fig. 7 Normal impact of an RTV-560 projectile with 50 percent 150 - -

O
porosity, UQ = 160 m/s
/ 50% Poros ity, *
100 Eq. (2)
shock data were given for nonporous gelatin since the am-
0 ^^"^
plitudes of the pressures were too great to measure.
From an analysis of these data, it was concluded that 50 tyXro -
O ^
gelatin with 15 volume percent microballoons would make a
good replica model of a bird. 0 1 1 i i i
0 50 100 150 200 250 300
Nonreplica Model. An extensive survey was made of
various materials which might effectively duplicate the Impact velocity, U (m/s)
loading history of birds during impact. Greatest consideration r
o
was given to a variety of synthetic rubber compounds because Fig. 9 Hugoniot shock pressure measured during the normal impact
of their ease of fabrication, low cost, availability, and good of RTV -560 projectiles [2]
shelf life.
One of the rubber compounds studied extensively was RTV
560, a room temperature vulcanized rubber manufactured by comparable to that of the porous gelatin studied. A typical
General Electric. Although its density of 1.33 g/cm 3 was low pressure trace is given in Fig. 7. Results from these tests are
relative to most of the other rubber compounds, it was still displayed in Figs. 8 and 9. Again, the correlation with theory
considerably higher than that of a bird. Initial impact tests is good.
confirmed that impact pressures were much higher than those From inspection of these data, it was concluded that a
experienced in bird tests. projectile fabricated from RTV-560 with 50 percent porosity
One of the desirable features of the RTV compounds was would give impact pressures similar to those resulting during
that the density could be decreased in a controllable manner bird impact.
with the addition of blowing agent during mixing and prior to
curing. The result of the addition of the blowing agent was a Model Selection. In the final selection process of a sub-
homogeneous, porous material with lower density and lower stitute bird model, two materials were given strong con-
wave speed. sideration. These were gelatin with 15 percent porosity and
An initial test program was conducted using a projectile RTV-560 with 50 percent porosity. Both materials were
having 50 pecent porosity and a density (p = 0.66 g/cm 3 ) readily available, relatively inexpensive, and could be easily

Journal of Engineering for Power OCTOBER 1981, Vol. 103/729

 50 300 1 1 i
1 r
A- Chicken o - Gelatin
(15% Porosity)
40 - o - G e l a t in 250 -
(15% P o r o s i t y ) D - RTV
(50% Porosity)
RTV 0
30 (50% Porosity) i" 200 " • - 4 kg Chicken
-
"a
A- 1 kg Chicken
AA
20 # 150 -
-
Ct^j A
a

Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/103/4/725/5805082/725_1.pdf by Technische Informationsbibliothek (TIB) user on 07 May 2024

A
10- 100
O
a& o a
50 n a a
50 100 150 200 250 300 * A° °
a A
impact velocity, U (m/s) _L 1 1 1 I, i
o
0 50 LOO 150 200 250 300
Fig. 10 Comparison of experimental stagnation pressure data lor
several candidate materials Impact velocity, U (m/s)

Fig. 11 Comparison of experimental shock pressure data for several

fabricated into projectiles. The RTV-560 had the advantage candidate materials
of long shelf life, whereas the gelatin projectile had a shelf life
of only a few hours unless refrigerated. impact loading history, thereby negating the repeatability of
The microballoon gelatin bird was chosen because it more impact loading which is so desirable. This problem would be
nearly replicated the impact loads of real birds. Direct most prevalent in engine ingestion testing, in which projectile
comparisons of the experimentally obtained steady-state rotation prior to impact is almost unavoidable. For this type
stagnation pressures and Hugoniot shock pressures are given of testing, it is recommended that the projectile be fabricated
in Figs. 10 and 11, respectively. In Fig. 10, there appears to be in a spherical geometry.
no significant difference between the measured stagnation
pressures of birds and the two candidate materials. The major Acknowledgments
difference between the two candidate materials is seen in Fig.
11, where the high porosity of the RTV-560 has resulted in This work was conducted at the Air Force Materials
shock pressures far below those of the gelatin and the birds. Laboratory under Project 7351 and Task 735106. The
Because of the importance of the initial shock pressures in the program manager was Dr. Alan K. Hopkins. The ex-
local damage problem, RTV-560 with 50 percent porosity was perimental data were obtained by the University of Dayton
felt to be an unacceptable bird substitute. Efforts to increase Research Institute on Contract F33615-73-C-5027, under the
the shock pressure by decreasing the porosity resulted in an direction of Dr. John P. Barber.
unacceptable increase in stagnation pressure.

Summary and Conclusions References

A synthetic bird model has been developed which accurately 1 Wilbeck, J. S., and Barber, J. P., "Bird Impact Loading," The Shock
reproduces the impact loads caused by real birds. The model and Vibration Bulletin, Vol. 48, Part 2, Sept. 1978, pp. 115-122.
material is a mixture of commercial gelatin and phenolic 2 Wilbeck, J. S., "Impact Behavior of Low Strength Projectiles," Air
Force Materials Laboratory, Technical Report No. 77-134, July 1978,
microballoons. This material is inexpensive, readily available, 3 Tsai, S. W., Sun, C. T., Hopkins, A. K., Hahn, H. T., and Lee, T. W.,
and the projectile can be easily and quickly fabricated. A "Behavior of Cantilevered Beams under Impact by a Soft Projectile," Air
gelatin projectile also has sufficient strength to withstand Force Materials Laboratory, Technical Report No. 74-94, Nov. 1974.
inertial loads during launch. 4 Barber, I . P., Wilbeck, J. S., and Taylor, H. R., "Bird Impact Forces
and Pressure on Rigid and Compliant Targets," Air Force Flight Dynamics
In this program, a mixture of 85 percent gelatin (by volume) Laboratory, Technical Report 77-60.
and 15 percent microballoons was used to model chickens, 5 Barber, J. P., Taylor, H. R., and Wilbeck, J. S., "Characterization of
which are most commonly used in engine qualification Bird Impacts on a Rigid Plate: Part 1," Air Force Flight Dynamics Laboratory,
testing. Comparison of the impact loads between model and Technical Report 75-5, Jan. 1975.
6 Tudor, A. J., "Bird Ingestion Research at Rolls-Royce," Symposium on
prototype was very good. The data from both types of the Mechanical Reliability of Turbo-Machinery Blading, Derby and District
projectiles also compared favorably with a hydrodynamic College of Technology, Apr. 1968.
theory developed earlier. It was also apparent that the impact- 7 Bowden, F. P., and Brunton, J. H., "The Deformation of Solids by
generated pressure profiles were much more reproducible for Liquid Impact at Supersonic Speeds," Proceedings of the Royal Society of
London, Series A, Vol. 263, Oct. 1961, pp. 433-450.
the impact of the model material than for that of birds. 8 Bowden, F. P., and Field, J. E., "The Brittle Fracture of Solids by Liquid
In the reported test program, the model projectiles were Impact, by Solid Impact, and by Shock," Proceedings of the Royal Society of
fabricated in the form of a right circular cylinder with a London, Series A, Vol. 263, 1964, pp. 331-352.
length-to-diameter ratio of 2. As the program progressed, it 9 Heymann, F. J., "On the Shock Wave Velocity and Impact Pressure in
High-Speed Liquid-Soil Impact," ASME Journal of Basic Engineering, Series
became apparent that a bird would be more accurately D, Vol.90, July 1968, p. 400.
modeled by a cylinder with the same length-to-diameter ratio 10 Heymann, F. J., "High-Speed Impact Between a Liquid Drop and a Solid
but with hemispherical ends. Surface," Journal of Applied Physics, Vol. 40, No. 13, Dec. 1969, pp.
5113-5122.
Such a geometry, though correct physically, requires ex- 11 Glenn, L. A., "On the Dynamics of Hypervelocity Liquid Jet Impact on a
treme care during testing in control of impact orientation. Flat Rigid Surface," Journal of Applied Mathematics and Physics, Vol. 25,
Projectile yaw during impact would result in a variation in the 1974, pp. 383-398.

730 / Vol. 103, OCTOBER 1981 Transactions of the ASME