-
UCSD FAA JAMS Paper 2014 1
Impact Damage Formation on Composite Aircraft Structures
Principal Investigator: Hyonny Kim*, Professor
Student Researchers: Zhi Ming Chen, Sara White, Monica Chan,
Sean Luong
Dept. Structural Engineering, Univ. of California San Diego, La
Jolla, CA 92093
Project Description Paper Supporting Presentation Given at
Federal Aviation Administration JAMS 2014 Technical Review
Meeting
25-26 March 2014, University of Washington, Seattle, WA
* corresponding author: [email protected]
Abstract
Ongoing research at UCSD is focused on understanding the key
phenomena governing the formation of damage to composite
structures when
subjected to blunt impact sources. This includes establishing,
physically-accurate
modeling capabilities that are developed based on, and validated
by, impact
experiments conducted on specimens representing several levels
of the building
block. Impact sources of interest are those acting over a wide
area and
potentially affecting multiple structural elements, while
leaving little/no exteriorly
visible signs of the damage. Specifically, impact by ground
service equipment
(GSE) having soft/rubber-covered bumpers, hail ice impact, and
large radius
metal tips are being investigated. Key results are: (i) modeling
capability has
been established allowing prediction of the onset and
propagation of interlaminar
failure modes, including inter-ply delamination and intra-ply
fiber failure, (ii)
successful modeling methodology has been demonstrated for
predicting blunt
impact damage formation in specimens ranging from simple
small-sized
-
UCSD FAA JAMS Paper 2014 2
elements, to large-sized 5-frame panels; models are validated by
experiments,
(iii) acoustic emission measurements have been shown to provide
information
identifying the occurrence of failure events such as the
fracture of shear ties
during a blunt impact event, (iv) four distinct internal core
failure modes were
identified for blunt impact onto honeycomb core sandwich panels,
ranging from
slight core wrinkling (evidence of cell wall buckling) to core
wall fiber fracture, (v)
hail ice impacts on sandwich panels produce significant internal
core damage
even with low dent visibility.
1.0 Introduction
1.1 Motivation
Impact damage remains a major source of damage to composite
aircraft,
including the new generation of composite-intensive aircraft
coming into service.
This is a major concern due to the resilience of the composite
outer skins which
can sustain a high degree of deformation without developing
cracks, even though
the internal sub-structure is damaged. Thus, the traditional
reliance on visual
detection to find damage, which worked well for metal skins that
dent easily, may
be inadequate for composite airframes, particularly those having
skin-stringer
construction. Additionally, sandwich structure exhibits complex
behavior when
subject to blunt impacts, with internal damage to the core
affecting the core shear
and peel strength.
Investigations are ongoing that: (i) establish a clear
understanding of the
development of damage from blunt impacts, and (ii) address the
difficulties in
being able to visually detect and predict the corresponding
damage. The
-
UCSD FAA JAMS Paper 2014 3
research conducted at University of California at San Diego
(UCSD) has been
focused on blunt impact sources that are not well understood. As
summarized in
Figure 1, these include ground service equipment (GSE) and hail
ice.
Furthermore, impacts by widely-ranging radius metal tips are
being used as a
generic impact source, representing GSE corners, railings, and
unknown-
geometry foreign objects. UCSD’s activities in these areas are
closely tied in with
industry and directly address aviation industry-driven needs for
increased
knowledge in these areas, from both experimental and
analytical/computational
viewpoints, ultimately aiding in the development of more
efficient and safe aircraft
structures.
Figure 1. UCSD blunt impact focus: hail ice and ground service
equipment.
Blunt impacts can be defined as impact sources that can affect
large
areas or multiple structural elements, while potentially leaving
little or no
externally visible detectable signs of damage. Blunt impacts
come from a variety
Hail Ice Impact• upward & forward facing
surfaces• low mass, high velocity
Ground Vehicles & Service Equipment• side & lower
facing
surfaces• high mass, low velocity• wide area contact• damage
possible at
locations away from impact
Blunt Impacts• blunt impact
damage (BID) can exist with little or no exteriorvisibility
• sources of interest are those that affect wide area or
multiple structural elements
-
UCSD FAA JAMS Paper 2014 4
of sources, as illustrated in Figure 1, and can involve a wide
range of energy
levels. While the side and lower facing surfaces of the aircraft
are subject to
contact with GSE, all exposed upper and vertical surfaces are
subject to ground
hail impact (terminal velocity + wind gust) and forward-facing
surfaces are
subject to in-flight high velocity hail impacts. These impact
sources are described
in more detail:
i. Impact/contact by ground service equipment (GSE) includes
ground
vehicles, cargo loaders, and any other equipment coming in close
proximity
to a commercial aircraft. GSE is a source of great damage
potential due to
the high mass of the GSE (typically 3,000 to 15,000 kg) and
subsequently
high energies involved. Figure 2 shows the kinetic energy levels
of GSE to
be up to the 103 J range (e.g., 10,000 kg cargo loader moving at
2 mph, or
0.894 m/s, has 4,000 J of kinetic energy). The length scale of
contact for
GSE impact can range from ~20 cm for a belt loader’s single
D-shaped
bumper, to over 2 m for a long cargo loader bumper (both shown
in Figure
2). It should be noted that GSE historically accounts for a
major percentage
of damage occurring to commercial transport aircraft [1], and is
expected to
continue to be a major source regardless of whether the aircraft
is made
from aluminum alloy or carbon/epoxy composite.
ii. Hail ice impacts involve high velocities and thus high
energy levels,
exceeding 1,000 J (see Figure 2), mainly by virtue of typical
aircraft in-flight
speeds (well over 200 m/s). Even higher velocities (energy) are
possible for
rotating components like helicopter rotor and engine fan blades.
Ground hail
-
UCSD FAA JAMS Paper 2014 5
occurs at lower energy levels, with relatively high velocity,
ranging from 20 to
30 m/s, which excites a localized dynamic response in the
impacted
structure. The ice projectile is complex in that it exhibits an
initial elastic-type
response and then severely crushes during the course of impact.
This
crushability results in a larger zone of contact, as shown in
Figure 2, which
significantly reduces the propensity to impart visual damage.
Meanwhile, ice
impact can produce significant internal damage such as
large-area skin-
stringer separation or internal sandwich core crushing, with no
external
visibility, unless penetration is achieved.
iii. Low velocity impacts by metal tips of large radius (up to
76 mm radius) are
being investigated as a representation of impacts by generic
sources, such
as dropped equipment, or contact with GSE features that are not
covered by
rubber bumpers, e.g., railings and rounded corners. The results
of impacts
by large radius tips are contrasted with a 12.7 mm radius (1 in.
diameter) tip
that is most commonly used for creating barely visible impact
damage (BVID)
of a specified dent depth (e.g., 0.3 mm). The 12.7 mm radius tip
damage
source, while commonly employed as part of an airframe
component’s
damage tolerance program, may not be representative of actual
blunt
sources (often of unknown geometry) of damage which can involve
higher
energy levels. Larger radius tips allow higher contact forces to
be applied
before a dent or other externally visible evidence of damage is
left on the
impacted surface.
-
UCSD FAA JAMS Paper 2014 6
Figure 2. Blunt impact energy-damage spectrum.
Due to the quasi-brittle behavior of composites, and the large
elastic strain
energy stored up until failures initiate, extensive sub-surface
damage can
develop when impacts occur at levels just exceeding the amount
needed to
initiate failure [2-4], i.e., when the failure threshold is
exceeded. Additionally,
damage from blunt impacts to internal stiffeners can be
extensive, existing in the
form of separation of stiffeners from the skin and fracture of
the stiffeners. Of
critical concern is whether damage is extensive enough to result
in the structure
losing ultimate (or even limit) load capability.
Among the three impact sources being studied (see Figure 2),
both the
low velocity metal tips and high velocity ice impact can usually
be associated with
a damage size limit. This damage limit reflects the ability for
these impact
sources to induce localized failure at the point of
impact/contact. Thus,
-
UCSD FAA JAMS Paper 2014 7
maximum-level cases of velocity and energy will typically
produce penetration
damage as a result of this localized response. Penetration
damage has a size
limit roughly equivalent in dimension to the impacting
projectile. For this reason,
the worst-case scenario might not always produce the most
critical mode(s) of
damage.
For the GSE impact, a non-local response has been observed to
be
possible. Specifically, internal damage can occur at locations
away from the point
of impact/contact, e.g., at joints or stress concentrations
along the path of internal
reaction. Thus, the damage size limit is not clearly known and,
depending on the
severity of the impact event, can possibly be larger than the
length scale of the
contact between the projectile and structure. Furthermore, the
GSE contact
speed can be quite varied, even very slow (quasi-static) level
as indicated in
Figure 2, and so the energy-based description might not fully
describe GSE-
aircraft interactions.
1.2 Objectives
The objectives of this research program focuses on impact
damage
formation by a range of sources, including: (i) low velocity
wide-area blunt impact
– vehicle/ground equipment contact with composite aircraft
structure, (ii) high
velocity hail ice impact, and (iii) impact by low velocity
metallic tips of large
radius. A common set of objectives exists for these three
project focuses:
1. Characterize blunt impact threats and the locations where
modes of
damage can occur.
-
UCSD FAA JAMS Paper 2014 8
2. Understand blunt impact damage (BID) formation and visual
detectability,
specifically seeking to:
determine key phenomena and parameters governing BID
formation,
understand how damage is affected by bluntness/contact-area,
identify and predict failure thresholds (useful for design),
establish what conditions relate to development of significant
internal
damage with minimal or no exterior visual detectability, and
provide guidance on the inspection and detection of BID to
internal
structural members.
3. Develop analysis methods, including new modeling capabilities
validated
by tests. This focuses on failure modes and phenomena that are
currently
difficult or even not possible to predict with conventional FEA
approach.
Additionally, guidance will be established on defining how to
analytically
predict whether damage is visually detectable or not.
4. Develop blunt impact testing methodologies, including
defining appropriate
boundary conditions between full-scale panels vs. complete
structure for
GSE impact studies.
1.3 Approach
While each of the project focus areas has unique challenges
related to
their length scales and velocity regimes, a common general
approach to achieve
these objectives is the following:
1. Experiments: impact representative structure/specimens,
specifically
-
UCSD FAA JAMS Paper 2014 9
wide area high energy blunt impact – e.g., from ground
service
equipment,
hail ice impacts – in-flight and ground hail conditions, effect
of
internal stiffeners, internal damage to honeycomb sandwich
core
with little/no dent depth,
low velocity impacts – non-deforming impactor which serves as
a
generic impact source, understanding large radius effects on
damage visibility.
2. Modeling development capable of predicting the various impact
damage
modes, including delamination and just-visible surface cracking
–
nonlinear FEA, analytical simple models, energy balance.
3. Communication of results to industry and collaboration on
relevant
problems/projects via workshops and meetings (at UCSD or at
industry
site, by teleconf/WebEX meeting).
1.4 Expected Outcomes
The expected research outcomes will provide information and
establish
methodologies that will aid engineers in assessing whether an
incident could
have caused damage to a structure, and if so, what sort of
damage mode, extent,
and location such damage would occur. This information assists
with
understanding of what kind of inspection techniques should be
applied to assess
the extent of damage. Furthermore, it is expected that design
engineers can
make use of the research outcomes to: (i) improve the resistance
of composite
-
UCSD FAA JAMS Paper 2014 10
aircraft structures to damage from blunt impacts sources, (ii)
provide critical
information on the mode and extent of seeded damage for use in
damage
tolerance considerations, and (iii) define relationship between
blunt impact threat
and expected damage severity – e.g., specifying safe GSE
operation speeds, or
at for what size hailstones should deeper inspection be
necessary.
Modeling capabilities are being developed to establish clear
guidance on
how to predict blunt impact damage modes and whether the blunt
impact event
will leave visible evidence of any significant damage formation.
This includes
providing the ability to predict the initiation, progression,
and final failure modes
resulting from the three impact sources being studied.
1.5 Research Partners
UCSD’s research on impact of composites is of direct interest to
industry,
as reflected by the research partners. Table 1 summarizes the
research partners
that are involved in UCSD’s FAA-funded research activities.
Large and small
aircraft manufacturers, a small composites-specialty engineering
firm, and a
material supplier are represented.
-
UCSD FAA JAMS Paper 2014 11
Table 1. UCSD research partners on FAA-funded activities.
Company/ Organization
Description/ Expertise
Role in UCSD Project
Boeing OEM – Large Transport Aircraft
Provide guidance and input on blunt impact. Particular focus on
blunt impacts onto panels of stiffened-skin construction. Possibly
supply test panels. Participate in on-site meetings.
Bombardier OEM – Small/Regional Aircraft
Provide guidance and input on hail ice impact, particularly
focused on sandwich panels. Supply test panels for ice impact.
San Diego Composites
Composites Design and Manufacturing
Provide technical advice on the direction of the project,
guidance on the design of the large-scale blunt impact composite
test panels, guidance on the design of tooling for manufacturing
the test panels, and access to large autoclave for curing
panels.
Cytec Engineered Materials
Materials Supplier
Provide technical advice on project directions. Provide guidance
on use of materials. Supply carbon/epoxy prepreg materials to
support fabrication of test specimens at UCSD for both blunt impact
and hail ice studies.
United Airlines
Airline Provide guidance and feedback on project directions
particularly with reference to operator view.
Delta Airlines Airline Provide guidance and feedback on project
directions particularly with reference to operator view.
JCH Consultants
Consultant on Aircraft Safety and Composites
Advise on direction of project, provide guidance on tests, data
reduction, results interpretation and dissemination to the public
and senior-level individuals in industry as well as to military
(Air Force).
Avanti Tech, LLC
Small Business – Nondestructive Evaluation
Coordinated activities focused on measuring ultrasonic guided
wave signals originating from active damage sources, to be used for
damage detection systems.
2.0 Wide Area Blunt Impact Damage 2.1 Background and
Motivation
With the increased use of composites in airframe primary
structural
components (e.g., fuselage and wing), there is a need to better
understand the
damage mechanisms caused by accidental transverse impact
loading,
particularly for high energy levels. The low resistance of
composite laminates to
transverse impact damage is a major weakness of these
materials.
The largest source of damage to a commercial aircraft is caused
by
accidental contact with ground service equipment (GSE) [1]. GSE
can weigh up
-
UCSD FAA JAMS Paper 2014 12
to 15,000 kg and travel at speeds up to 1 m/s while within close
proximity of the
aircraft. The high mass of GSE can result in very high impact
energy upon
collision with the aircraft, and can result in widespread
internal damage. In
addition, rubber bumpers typically outfitted to GSE can reduce
the local surface-
visible damage in the aircraft skin and mask the occurrence of
an internally-
damaging impact event.
Non-destructive evaluation (NDE) and structural health
monitoring (SHM)
methods are of increasing interest for aerospace structures.
Aside from
understanding and predicting damage that occurs from GSE, it is
of interest to
quickly determine what damage may be present after an impact has
occurred by
using methods, such as NDE, that do not involve disassembling
the aircraft for
tedious and costly interior visual-based inspections. Although
accurately
predicting small damage occurring from wide area impacts can be
very difficult
due to complexities in geometry and materials, identifying
large-sized damage
modes is of key interest.
2.2 Summary of Previous Results
GSE Blunt impact experiments on large scale fuselage panels have
been
completed up to Phase II (see Figure 3). The last 5-frame
specimen tested,
Frame04, was geometrically identical to Frame03. This panel was
tested twice,
once mounted with composite shear ties (ID: Frame04-1) and again
with stout,
aluminum shear ties (ID: Frame04-2). Comparing the composite
shear tie tests
to the aluminum shear tie test, the overall failure path has
been altered. In the
-
UCSD FAA JAMS Paper 2014 13
composite shear tie panel tests, the initial failure mode would
always be shear tie
delamination and fracturing. Then, the C-frames become free to
rotate or make
contact with the stringers, resulting in C-frame failures from
large deformation or
high contact stress (see Figure 4). In contrast, the Frame04-2
panel sustained
higher impact load without shear tie failure and with minimal
deformation of the
shear ties. Subsequently, the C-frames were more directly
loaded, resulting in
localized shear fracturing of the C-frames as both the initial
and final failure mode
(see Figure 4).
Figure 3. Blunt impact program phases and modeling capability
development.
Basic Elements- Excite Key Failure Modes- Model Correlation
Data
- Understand Damage Formation &Relationship to Bluntness
Parameters
Large Panel- e.g., 5 Bays
- Damage Excitation- Damage Thresholds- Model Correlation
OEMHardware- 1/4 to 1/2Barrel Size
- Vehicle Impacts
Scaling,B.C. EffectsDynamics
Scaling,B.C. Effects
Increasing LengthScale, Complexity,and Specificity
Phase III(Year 3)
Phase II(Year 2)
Phase I(Year 1)
Modeling CapabilityDevelopment & Correlationwith Test are
Key Aspects
at Each Level
High Fidelity Explicit
Dynamic FEA- Complete Simulation- Damage
Progression- Final Damage
State
Elastic Static FEA
- Damage Initiation
Simple Models: Reduced Order 2DOF, Energy & Momentum
Balance
UCSD Developing Methodologies at All Levels
Increasingly Complex Phases of Activity
- gain fundamental understanding at bottom
- use modeling to generalize
-
UCSD FAA JAMS Paper 2014 14
Figure 4. Comparison of failure modes for panels having
composite (Frame03) vs. aluminum (Frame04-2) shear ties.
Significant efforts were made toward FE modeling of the GSE
blunt impact
tests. An Ogden material model was used to simulate the OEM
bumper material
03
Frame03 –
Non‐local failure of frame away from center due to load transfer between stringer‐frame contact & frame rotation. No visible cracks or dent.
Near BCs Near
BCs
04‐2
Frame04‐2 –
Local failure of frame @ center. No shear tie failure, thus no major frame rotation. Minor skin cracks visible away from impact site.
-
UCSD FAA JAMS Paper 2014 15
to obtain accurate impact contact pressure distribution. The
modeling
methodologies were refined to accurately model the deformation,
failure modes,
and failure propagation in the StringerXX experiments,
specifically for specimens
Stringer05 and Stringer06. Finally, Python scripts were
developed to enable
parametric studies of the large-sized FrameXX panel impacts.
Basic models
have been established for Frame03 and Frame04 up to C-frame
failure. These
models provided some insight into damage evolution within the
panels, but were
missing the initial delamination failure modes which
subsequently affect the
overall panel response.
2.3 Recent Results
2.3.1 Lab-Scale Element-Level Experiments
An inverted testing pyramid approach [5] was taken to analyze
the blunt
impact experiment results. The locations of key failures from
the previously-
conducted experiments have first been identified, and then
isolated small-scale
experiments were conducted on the individual elements of the
large panel to
recreate these failure modes. Element and coupon size specimens
were
manufactured, and were tested in loading conditions designed to
simulate the
loading sustained by these components during the blunt impact
experiments.
2.3.1.1 Stringer Element Specimen Tests
Stringer element specimens have been fabricated to examine the
bumper-
to-skin interaction and initiation of skin surface damage (i.e.,
visible surface
cracking) during the StringerXX tests. The stringer element
specimens are 76
-
UCSD FAA JAMS Paper 2014 16
mm wide with skin and stringer components cut from the undamaged
part of
specimen Stringer00. Specimens are potted in epoxy resin and
gripped in an
uniaxial test machine as shown in Figure 5. The stringer element
is indented
with the same 76 mm wide D-shaped rubber bumper as the
StringerXX panel
tests. One stringer element, SE01 has been tested so far. During
the test, the
specimen was quasi-statically indented until initial
skin-visible surface damage
was created. Due to the potting constraint of the specimen (see
Figure 5), the
majority of the deflection occurred within the skin.
Figure 5. Test setup for stringer element indentation.
-
UCSD FAA JAMS Paper 2014 17
High bending stress concentration on the skin near the
stringer-to-skin
joint location led to visible skin surface failure. Figure 6a
shows the visibility of
surface cracks in specimen SE01, and the photomicrograph in
Figure 6b shows
detail of the crack observed at the joint corner. It can be
conclude from these
figures that at the point when barely visible external damage is
created, the
external plies (near surface plies on both sides of the skin)
will experience
through-thickness cracks, and some modest delamination. This
focused
experiment provides critical data for establishing
analysis-based criteria defining
the onset of damage visibility.
Figure 6. (a) SE01 post failure, visible skin crack at the joint
location, and (b) photomicrograph of the SE01 showing the formation
of cracks and delamination
at the skin-stringer joint location.
The SE01 test is simulated using continuum shell (SC8R in
ABAQUS)
elements to establish the ability to predict the formation of
visible cracks. The
quarter symmetry model shown in Figure 7 uses the Hashin-Rotem
failure
criterion to represent in-plane ply failure. The force vs.
actuator displacement
curve of the test and FE model is shown in Figure 8. As can be
seen from the
figure, a nearly 1:1 matchup is found between these curves in
the elastic region.
(a) (b)
-
UCSD FAA JAMS Paper 2014 18
In the experiment, failure initiation was measured at 21.55 kN,
whereas the FE
model failed at 19.04 kN. The failure mode of the FE model
matches the failures
found in the experiments, with cracks forming in the laminate
outer plies, as
shown in Figure 6b and 7.
Figure 7. Quarter-symmetric model of the test.
Figure 8. Load vs. displacement plot of the SE01 test and FE
model.
2.3.1.2 Shear Tie Compression Element Specimens
During Phase II large-panel blunt impact experiments on the
specimen
Frame03 which was configured with composite shear ties, damage
to the shear
0
5
10
15
20
25
0 10 20 30 40 50 60 70
Contact F
orce (k
N)
Actuactor Displacement (mm)
SE01 Test
SE01 FEA Model
-
UCSD FAA JAMS Paper 2014 19
ties always occurred first, thereby absorbing much strain energy
and allowing the
impact loading to distribute over a larger number of structural
elements. The
sequence of failure modes in the shear ties are: delamination in
the curved
region, followed by local crushing and fracturing under
compression loading, and
finally buckling-induced bending failures resulting in complete
fracture of the
shear ties. Subsequent panel deformations are then affected by
these shear tie
failure modes. Hence, it is desirable to understand and
accurately model the
complete damage sequence in the shear ties.
A focused experiment on shear tie behavior under direct
compression has
been conducted to make observations on damage initiation and
evolution which
will support model development. Figures 9 and 10 show the test
photos and the
load vs. displacement plot of this experiment, respectively.
Figure 9a shows the shear tie compression experimental setup
prior to
loading. A 66 mm wide shear tie section was bolted to a curved
aluminum
support base plate, as shown. A top loading plate with V-groove
was used to
apply compression loading on the top of the flange, while
permitting free rotation.
As the load is applied, the shear tie undergoes bending and
shearing at the
curved corner of the shear tie. In response, the shear tie
corner flexes in an
opening-moment mode, until it delaminates, as seen in Figure 9b
(opening
moment produces radial tension stress in curved laminate). The
Initial
delamination, developing at an applied load of 7.0 kN, resulted
in the first load
drop (see Figure 10 curve labeled Experimental Data). Additional
delamination
at the radius then continued to accumulate and grow. In Figure
9c, outer ply
-
UCSD FAA JAMS Paper 2014 20
failures are shown to occur at a load level of 4.9 kN (see
displacement of 5 mm
in Figure 10). This ply failure is driven by the highly
localized transverse shear
stresses at the radius region. Crushing failure eventually
softened and flattened
the radius region against the base plate (see Figure 9c). At
this point, the shear
tie flange is subject to direct compression loading which
induces buckling and
subsequent bending-failure at 10.5 kN (see Figure 9d and 13.9 mm
displacement
in Figure 10).
Figure 9. (a) Pristine shear tie compression specimen, (b)
initial delamination at shear tie corner, (c) crushing failure and
flattening of the corner, and (d) buckling-
induced bending failure.
Figure 10. Load vs. displacement plot of the SE01 test and FE
model.
02468
101214161820
0 2 4 6 8 10 12 14 16 18 20
Contact F
orce (k
N)
Crosshead Displacement (mm)
Experimental DataFE 4 LayersFE 6 Layers
-
UCSD FAA JAMS Paper 2014 21
The experiment was modeled in ABAQUS using cohesive surface
interactions for simulating inter-ply delamination. The
Hashin-Rotem failure
criterion for composite ply failure was not used because it
cannot account for
intra-ply transverse shear failure modes (i.e., fracturing of
the fibers under
transverse shear stress). Instead, the Hill failure criterion
for orthotropic
materials was used. This criterion allows for the modeling of
shear tie crushing
and severing after the shear tie has buckled.
Figure 11 shows a series of FE modeling results for the
composite shear
tie undergoing failure, analogous to the sequence shown in
Figure 9. The shear
tie layup consisted of 12 woven carbon fiber plies. The analysis
has been
conducted using four and six layers of elements (i.e., ply
groups) stacked through
the thickness, with groups composed of three or two plies,
respectively. The
ability of continuum shells to predict transverse stresses tends
to improve with
increasing number of elements stacked through the thickness [6].
In addition,
dividing the plies into multiple layers that can deform
separately allows directly
modeling of the interface layer separation with cohesive zone
techniques. As
can be seen from Figure 9d, delamination occurred at multiple
(approximately
four to five) interface layers at the shear tie corner, and each
delamination
contributed to softening of the shear tie. Hence, this
multi-layered modeling
approach was taken to ensure the correct residual stiffness was
captured. The
shear tie residual stiffness would be over-predicted if fewer
interface layers were
used in the model.
-
UCSD FAA JAMS Paper 2014 22
Figure 11. (a).Unloaded shear tie compression model with 4
layers of elements through the thickness, (b) initial delamination
at shear tie corner, (c) crushing
failure and flattening of the corner, and (d) buckling-induced
bending.
The FE analysis results are plotted in Figure 10 together with
the
experimental data. As can be seen from this figure, modeling the
shear ties with
four layers of elements is insufficient to capture the initial
delamination load.
However, the overall deformation response is still correct, as
can be seen in
Figure 11. Increasing the number of layers through the thickness
to six resulted
in better response of the shear tie coupon for the initial
delamination load, as well
as the shear tie crushing behavior up to 6 mm of crosshead
displacement. After
the ply element failure had initiated, the loads oscillate due
to stress singularity
experienced by elements at the corner. When these elements fail,
they are
removed from the simulation which caused load drops in the model
(a result not
seen in the physical experiment). Additional work needs to be
done to overcome
this shortcoming and improve modeling fidelity. Note that in
this model,
physically-accurate properties best representing these composite
materials were
used. No adjustment and “tuning” of properties or parameters was
done to attain
better correlation.
-
UCSD FAA JAMS Paper 2014 23
2.3.2 Large Scale Panel Modeling Development
The Frame03 (and Frame04-1) large-sized panel model has been
updated
using techniques developed from the shear tie compression test
specimens. The
previous 1st-generation Frame03 model yielded an approximate
overall panel
response, but the exact deformation and pattern of failures were
not captured
(see UCSD 2013 JAMS Paper Figure 10). Key failure modes in the
shear ties
that would later influence subsequent panel failures were
missing. Incorporating
the initial shear tie failure modes, namely delamination and
subsequent
softening, improved the modeling accuracy. The updated half
symmetric model
of Frame03 is shown in Figure 12.
Figure 12. Updated Frame03 model: (a) shear tie delamination
occurs, and (b)
shear tie buckling-induced bending failure.
(a)
(b)
-
UCSD FAA JAMS Paper 2014 24
Figure 12a shows the Frame03 model soon after impact when shear
tie
radius delamination initiates, while Figure 12b shows shear tie
failure and large
rotation of the frames. Figure 13 shows the load vs. indentation
plot of the
Frame03 test compared with the FE model results. In the FE
model, the load
increased to 72.3 kN before delamination and crushing of the
shear ties caused
an initial load drop, matching the response seen in the
experimental loading
curve. The shear tie flange made contact with the panel skin,
similar to the
behavior observed in the shear tie compression tests, thus
allowing the load to
build up again to a load of 97.4 kN, at which point the center
shear tie flange
buckled and fractured (see Figure 12b) and then load transfer is
through direct
stringer-frame contacts. As shown in Figure 13, this model is
accurate for up to
37 mm of indentation. Due to stress singularity issues and how
the failure
criterion was implemented, the FE model cannot predict the
failure pattern past
37 mm of indentation. In the model, the load rose to 113 kN
before the 2nd row
of shear ties buckled and failed in bending.
Improvements to the modeling capability are ongoing, following
the
approach of establishing accurate predictions for small
element-sized geometry
with simple loading, before applying to predicting the response
of a complex
built-up structure (e.g., compare shear tie compression in
Figure 11 and full
panel model in Figure 12). The overall plan for modeling
capability development
is shown in Figure 14. Clear definition of modeling
methodologies to accurately
capture important failure modes, based on physically-accurate
modeling of the
experiments, will subsequently be applied to investigating
various new structural
-
UCSD FAA JAMS Paper 2014 25
configurations that have not been experimentally tested. This
includes aspects
such as glancing impact, and the effects of additional internal
structure such as
the floor beams and the joints between the frames and the
floor.
Figure 13. Load vs. indentation plot of the Frame03 test and FE
model.
Figure 14. Modeling capabilities plan.
0
20
40
60
80
100
120
140
0 20 40 60 80 100
Force (kN)
Indentation (mm)
FE Hill Criterion with Cohesive Zone
Experimental (Combined Frame03 & 04)
Glancing ImpactSize, Complex Internal Structure, Geom.,
Joints
Various Impactors & Scenarios (vo)
Models of Generic Curved Panel Specimens- Static-
DynamicExperimental Validation
Capture Key Failure Modes (Major Damage)
Damage Initiation Criteria
Damage Progression
Dynamic Effects
Externally Visible?
EstablishCapabilities
Define Methodologies
fligh
tglo
bal.c
om/F
light
Blo
gger
How to predict this with FEA?
Apply to study and predict response for:
-
UCSD FAA JAMS Paper 2014 26
2.3.3 Frame-to-Floor Joint
In order to better represent the structural configuration of a
fuselage,
future test specimens have been designed to incorporate floor
beams and the
frame-to-floor joints. These components create a high stiffness
region which can
greatly affect the stress distribution within the frame and the
failure modes,
particularly cracking near/in the joint region.
During a meeting with the FAA and Boeing engineers on November
8,
2013 in Seattle, WA, future plans were made regarding plans to
incorporate
frame-floor interactions in the GSE impact studies. Key impact
regions along a
frame were identified and are shown in Figure 15:
Region 1 is the central (blue color) portion of the frame which
is most
compliant and would develop higher deformations and bending
stresses
during a GSE blunt impact.
Region 2 includes the regions near the floor joints (green zones
within 3-
4X frame depths from floor joints) will be more stiff during
impact, and will
develop high beam shear stress within the frames due to the
proximity of
these regions to the highly stiff floor joints which act as
stiff boundary
conditions to the frames.
Region 3 includes the floor joint region (red) which would be
the most stiff
as an impact at these locations would be directly parallel with
the floor
joint. Direct GSE hits to these locations are anticipated to
easily damage
the frame and frame-to-floor joint.
-
UCSD FAA JAMS Paper 2014 27
Figure 15. Impact regions and floor structures.
The frame-to-floor bracket has been designed to fit tightly
within the floor
beam, connecting to the web and flange. The current design is
shown in Figure
16, where the aluminum alloy bracket will connect the floor beam
flange and C-
frame via bolts. The frame-to-floor joint is first being modeled
in FEA (currently
ongoing) prior to development of a detailed specimen and
conducting
experiments to investigate joint-adjacent failure. The model
geometry shown in
Figure 17 will include a portion of the skin and stringers, as
well as shear ties and
frame. This model development is ongoing, and it includes
boundary condition
studies to determine the appropriate rotational and
translational stiffness that
should be applied as boundary conditions to the ends of the
C-frames.
Region 1
Region 2
Region 2
Region 3
Region 3
-
UCSD FAA JAMS Paper 2014 28
Figure 16. Frame-to-floor bracket design.
Figure 17. FEA model geometry of floor structure incorporated in
to frame.
-
UCSD FAA JAMS Paper 2014 29
Failure modes of interest include the frame cracks observed in
previous
UCSD large-panel tests of specimens Frame01, Frame02, Frame03,
and
Frame04. These cracks developed due to high bending and
torsional
deformation of the frames, as well as high stresses developed
when stringers
and frames made contact during blunt impact. Examples of these
frame cracking
failure modes are shown in Figure 18. Additional failure modes
of interest will be
cracks anticipated to develop at the fastener holes of the
frame-to-floor bracket,
and under direct crushing of the frame during impact at Region
3.
Figure 18. Examples of frame flange cracks resulting from blunt
loading.
-
UCSD FAA JAMS Paper 2014 30
2.3.4 Non-Destructive Evaluation Measurements
Supporting the large-scale panel blunt impact tests,
non-destructive
evaluation (NDE) measurements were performed. Each panel was
equipped with
macro fiber composite (MFC) acoustic sensors which were bonded
to the inside
skin surface of the panels. The acoustic emissions resulting
from impact and
proceeding panel failures were recorded. The MFC sensors were
arranged in a
rosette pattern, as shown in Figure 19, with the aim of
determining the incoming
wave direction by methods described by Matt and Lanza di Scalea
[7].
Figure 19. MFC rosettes used to determine a wave source.
The general concept of the MFC rosettes is that the three
sensors placed
next to each other at a known separation angle can be used to
determine the
incoming direction angle of a Lamb wave. This is due to the
varying voltage
response excited depending on the incoming angle with respect to
the orientation
-
UCSD FAA JAMS Paper 2014 31
of the MFC sensor (sensitivity is directional). When a minimum
of two rosettes
(six sensors) are used, each which will provide an incoming wave
angle which
can be projected to intersect, thereby revealing the wave source
location (a
damage event produces waves). This approach is not dependent on
the velocity
of the wave source but instead relies on the directivity of
voltage output of the
MFC sensors, which is a favorable for complex geometries,
anisotropic
laminates, and dispersive materials where determining wave
speeds is difficult.
Specimens Frame 03 and Frame 04 were both equipped with four
rosettes
(12 MFC sensors total) located as shown in Figure 20.
Figure 20. Rosette locations and shear tie labeling on specimens
Frame03 and Frame04 prior to C-frames being assembled to
panels.
-
UCSD FAA JAMS Paper 2014 32
Unfortunately, the acoustic emissions recorded from Frame 04
were
saturated as a result of too small of a window chosen for the
voltage, thus
making much of the data unable to be analyzed. For the Frame 03
test, the
window was large enough to capture all the data from the test
and will be the
focus of discussion from here on. Using the methods described by
Matt and
Lanza di Scalea [7], along with incorporating small time windows
to analyze each
perceived damage event, the predicted damage locations are
plotted in Figure
21.
Figure 21. Predicted damage locations for Frame 03 for entire
test duration by methods of reference [7].
A closer look at the damage developed as a function of time
was
documented using the high speed videos. The data were recorded
with the
same time trigger as the high speed video. Much of the visible
damage occurred
at the shear ties and in the C-frames. These events were then
compared to the
-
UCSD FAA JAMS Paper 2014 33
responses of the sensors at a given location. Table 2, was
constructed using the
high speed video with each row representing the shear ties which
are labeled
according to Figure 20. The colored block indicates visible
cracking and a
colored block with an 'x' denotes major failure of the shear
tie.
Table 2. Time table of shear tie failure based on high speed
video. Shear tie
label in rows with colored block denoting visible cracking and
colored block with an 'x' denotes major failure.
Time (ms) 1
21
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
M
x x
x R
Time (ms) 1
42
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
H
x x
x x
R
Time (ms) 1
63
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
R
x x
no events detected between 184 ‐ 220 ms
Time (ms) 2
21
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
G
x x x x x
I
x x x x x x x
L
x x x
x
N
x x x x
Q
x x
x x x x x x x
S
x x x x
-
UCSD FAA JAMS Paper 2014 34
The spectrogram from the signals, which visually shows the
frequency
content as it varies with time along with a relative amplitude,
is generated from a
selected event and compared with the timing of damage events
visible in the high
speed videos. Shear tie R failure, which is located near
rosettes 2 and 3 (as
shown in Figure 20), was of particular interest because there
were no other
visible events occurring at that time. The spectrograms shown in
Figure 22
compare the results from one sensor near shear tie R with one
far from shear tie
R, to provide some insight on how much information from a shear
tie failure event
is transferred into the skin and transmits along the skin
(through stringers and
rows of shear ties along the way) before being detected by the
MFC sensors.
Figure 22. Spectrogram results from two different sensors (near
and far) during a shear tie R failure.
-
UCSD FAA JAMS Paper 2014 35
This comparison reveals a strong attenuation of the signals for
increasing
distance, with a high concentration of the signal having ~160
kHz frequency
content. Delamination damage and minor-level cracking in shear
tie R initiated at
121 ms and was constantly evolving and growing, and thus rosette
3 constantly
picked up a significant-strength signal. Finally, when major
fiber-failure occurred
at 167 ms, a more intense broadband signal is abruptly visible
in rosette 3 at 167
ms time, and a significant weaker narrow signal also appears in
rosette 1 at this
time.
2.4 Discussion
Focused, small-scale experiments on element-level specimens
were
conducted to gain insight into the key failure features of the
large experiments.
Element-level FE models were then created to accurately predict
these
experiments which help to validate choices made in defining the
models and
demonstrate the capability to predict these failure modes. Once
validated
models were established at the element level, they were
incorporated back to the
larger-scale models.
These high fidelity physics-based models, created at both small
and large
scales, can predict the key failures modes within a narrow error
margin. The
stringer elements were able to predict skin cracking initiation
within a 10% load
error. In the case of the shear tie compression coupons, the
results were in
accurate agreement with the experiments after refining the mesh
through the
thickness of the shear tie to six layers of elements. The
modeling methodologies
-
UCSD FAA JAMS Paper 2014 36
were then transferred to the large StringerXX and FrameXX
models. The results
for StringerXX modeling was shown in the 2013 JAMS report. It is
now shown
that by using a refined shear ties model in the FrameXX models,
the progression
of damage following failure of the shear ties is much better
captured. However, it
is important to note that this work is ongoing and there is
still much to be done to
improve the predictions (e.g., address singularities created
during element
deletion, overcome the force oscillation created when elements
fail).
The NDE results from the Frame 03 test using methods described
by Matt
and Lanza di Scalea [7] show discrepancies between predicted
damage
locations and those found visually on the specimen and observed
via the high
speed video. This is likely due to the complexities of the test
specimen along
with the inability of this method to differentiate between
multiple acoustic
emission sources occurring simultaneously. Slight variations in
the sensor
sensitivity could also strongly affect the location prediction
results. Further
development is needed for reliable damage detect\ion, especially
for wide area
impacts which cause a number of discrete damage events at
various locations.
The spectrogram results along with cataloging the failure events
are used to
show that the sensors near the shear tie failures exhibit a
higher amplitude
response in comparison with the sensors further away from the
failure event.
This indicates that acoustic waves are capable of being
transferred from the
shear tie to the skin, even though the shear tie and skin are
not bonded together.
-
UCSD FAA JAMS Paper 2014 37
2.5 GSE Blunt Impact Conclusions
To gain insight into the large scale blunt impact experiments,
an inverted
building block pyramid approach to testing was used.
Small-scale
experiments were conducted on element-level specimens extracted
from
the large-sized panel specimens to directly study the individual
failure
modes and to provide data for model development and
validation.
The methodology presented allowed for the in-depth analysis of
failure
initiation and propagation in the experiments, resulting in the
development
of accurate finite element models that are based on actual
physical
observations and test data.
Acoustic emissions from shear tie failures are able to be
identified using
sensors bonded to the skin of a panel, even if the shear ties
are only
bolted to the skin (i.e., not tightly acoustically coupled).
3.0 Blunt Impact Damage to Sandwich Panels 3.1 Background and
Motivation
Composite sandwich panels are being used more widely in
various
aerospace structures because of their lightweight yet stiff
nature. However, their
impact characteristics are not well understood, particularly
when impacted by
blunt sources creating internal damage with little/no exterior
visibility (no dent, no
cracking). An in-service aircraft may encounter impacts from a
number of
projectiles such as blunt impact from ground service vehicles
and runway debris,
-
UCSD FAA JAMS Paper 2014 38
tool/equipment drops, or hail impact while the plane is on the
ground or in flight.
The varying types of projectiles and range in speeds at which
impacts may occur
can result in different damage modes such as fiber failure,
matrix cracking,
delamination, facesheet disbonding, core crushing, and the
interaction between
any of these damage modes. Furthermore, such damage may grow
and
eventually lead to sudden large-scale failure due to repeated
loading cycles,
such as pressure loading/unloading process during repetitive
ground-air-ground
flight cycles. Often, such damage may initiate at a very small
scale (mm) or
below the facesheet surface and cannot be visually detected from
the exterior.
Therefore, it is important to understand what type of damage to
expect after
impact has occurred to best determine what actions need to be
taken.
3.2. Summary of Previous Results
This sandwich impact research activity started in early 2013.
Since no
significant results were available by March 2013 to be reported
in the last 2013
JAMS paper, no summary of previous sandwich impact results is
included here.
3.3 Recent Results
3.3.1 Low Velocity Blunt Impact by Large Radius Metal Tips
Aircraft routinely experience impact damage from maintenance
related
sources, from contact with GSE, or dropped tools and equipment.
In the case of
composite sandwich panels, external damage is not always present
or visible,
making internal damage difficult to detect. This impact study
focuses on low
velocity impacts from a wide range of radius tips to determine
the correlation
-
UCSD FAA JAMS Paper 2014 39
between internal and external damage as a function of impactor
radius. Due to
the lightweight nature of the sandwich specimens, this
investigation is particularly
interested in impacts of low energies (2 to 14 J) in which
external damage is
barely visible.
The test specimens in this investigation were composite sandwich
panels
with a honeycomb core obtained from an Airbus A320 rudder. Four
impactor tips
of different radii, as shown in Figure 23, were mounted to the
pendulum impactor
shown in Figure 24. The test panel was held in a 165 mm square
window and
impacted with several tip radii and kinetic energy level
combinations, as
summarized in Table 3.
Table 3. Number of tests conducted at each radius/energy
combination.
Figure 23. Impact tips of radii 12.7, 25.4, 50.8, and 76.2 mm
(left to right).
-
UCSD FAA JAMS Paper 2014 40
Figure 24. Pendulum impactor test set-up.
-
UCSD FAA JAMS Paper 2014 41
3.3.1.1 External Damage and Data
At each impact, the maximum impact force and impact energy
are
recorded. The dent is measured within five minutes after impact
to obtain the
immediate dent profile and maximum dent depth. Because sandwich
panels tend
to experience relaxation over time, the dent is measured again
after 24 hours to
document a relaxed dent profile. A representative example is
shown in Figure 25.
Figure 25. Dent profile of test site impacted with R12.7 at 4
J.
The visibility of the dent is then characterized based on a
parameter
defined as the maximum depth D divided by the dent span S. The
span is
determined as the span of the dent at one-third of the depth
below the original
surface plane. Because visual detectability of damage is
dependent on both span
and depth, this “visibility” parameter is more indicative of its
visual-detectability
than depth alone. The visibility, plotted in Figure 26 as a
function of impacting
kinetic energy, shows increasing visibility with kinetic energy,
but no differences
-
UCSD FAA JAMS Paper 2014 42
between tip radii is apparent – i.e., the different tip radii
tend to produce visibility
(depth/span) values falling on a common curve over the range of
energies tested.
Figure 26. Measured kinetic energy vs. visibility for lower
energies.
A plot of the relationship between measured kinetic energies
and
maximum dent depth can be seen in Figure 27. The correlation
between dent
depth and kinetic energy is fairly linear up to 6J, and very
similar for all four tip
radii in the lower energy range (up to 6J). At higher energies
of 10 to 14 J, the
dent depths vary considerably for different radii, suggesting
the depths increase
at different rates with increasing impact energy. In most cases,
the panel
experienced only dent damage. However, the uppermost four
outlying data
points are instances where the facesheet was also cracked, which
only occurred
with tip radius 12.7 mm (blue) and 25.4 mm (red).
-
UCSD FAA JAMS Paper 2014 43
Figure 27. Measure kinetic energy vs. maximum dent depth;
circled points indicate facesheet cracking.
3.3.1.2 Internal Damage and Data
Each test site was sectioned (cut through impact site) to
observe the
damage to the honeycomb core. In general, the damage tends to
exist at a depth
which plateaus to fixed level, which is visible in Figure 28. In
this figure, the
damage progression can be observed for panels impacted with tip
radius 50.8
mm at energy levels of 4, 6, 10, and 14 J. Although the depth
reaches a
maximum level, the internal damage span, defined as the
measureable width of
the damage parallel to the facesheet, increases with increasing
kinetic energy.
Face Sheet Cracked
-
UCSD FAA JAMS Paper 2014 44
4 J
6 J
10 J
14 J
Figure 28. Sectioned view of damage progression for R50.8 mm tip
impact with increasing impact energy: 4, 6, 10, and 14J from top to
bottom.
The cross sectional inspections (e.g., in Figure 28) and
additional optical
microscopy (Figure 29) were used to identify four distinct modes
of internal
damage to the Nomex honeycomb core. These modes are identified
as:
Mode A: Slight wrinkling of cell walls (see Figure 29a).
Mode B: Visible wrinkling of cell walls (see Figure 29b).
Mode C: Buckling of cell walls (see Figure 29c).
Mode D: Fracture/Bursting of cell walls (see Figure 29d).
-
UCSD FAA JAMS Paper 2014 45
(a) Mode A (b) Mode B
(c) Mode C (d) Mode D
Figure 29. Sandwich core damage modes; impact applied to
facesheet located at
top of each photo.
Further correlation of damage modes with impact energy, peak
forces,
and external damage visibility is currently still in
progress.
3.3.2 Ground Hail Impact
Large-sized hail impact is a realistically-occurring event for
aircraft. The
size of hail may reach up to 75 mm and the speed of impact can
range from
terminal velocity (~25 m/s) to in-flight speeds (over 200 m/s).
The test specimens
-
UCSD FAA JAMS Paper 2014 46
utilized for this research were composite sandwich panels
obtained from an
Airbus A320 rudder that was removed from service. Due to the
limited number of
test specimens available from this part, a selected range of
testing conditions
were explored in order to focus on damage morphology for given
threat levels
(velocity) and glancing incidence angles, when subject to
impacts by 50.8 mm ice
spheres (representing hailstone).
The goal of these tests is to determine the condition at which
there is little
to no visible surface damage, while still having internal core
damage present.
While the results are specific to this specific lightweight core
sandwich panel
configuration, the methodology and approach are generically
applicable to a
broader range of the sandwich panel design space.
The panels were oriented at glancing angles of 90 degrees
(normal
impact), 40 degrees, and 25 degrees relative to the path of the
incoming ice
sphere. All impacts were conducted using 50.8 mm diameter ice
spheres
projected at nominal velocities of 25 m/s and 50 m/s. An
additional parameter of
interest was the facesheet thickness which was either ~1.19 mm
or ~1.87 mm
(these will be referred to as thin specimens and thick
specimens, respectively).
Recent data are summarized with all pertinent damage
measurements and
observations in Table 4.
-
UCSD FAA JAMS Paper 2014 47
Table 4. Summary of recent results with damage observations.
After each specimen is impacted, a dent profile is measured
within 15
minutes, and the peak dent is obtained from that profile. As
Table 4 shows, most
of the thin facesheet specimens suffered visible surface damage
as well as
significant internal damage. Figure 30 shows the relationship
between impact
energy and peak dent depth of the thin facesheet specimens. The
points that are
located on the x-axis represent the tests that fractured the
skin of the specimen,
so no useful peak dent data were measured from those tests. Over
the range of
energy tested, the data show that there is a linear relationship
between impact
energy and peak dent depth for each glancing angle, until a
certain threshold is
reached, at which point the skin fractures. As expected, the
dent depth increases
as a function of impacting kinetic energy at a higher rate for
40 degrees than at
-
UCSD FAA JAMS Paper 2014 48
25 degrees. The 90 degree impacts caused such high damage with
relatively low
energy levels (see Tests 001 and 002 in Table 4) that no more
specimens were
devoted to that condition.
Figure 30. Impact energy vs. peak dent depth of thin
specimens.
Figure 31 shows an example of the level of internal damage
resulting from
a 25 m/s impact at 40 degrees (panel sectioned through impact
site). Other
details about the impact can be found in Table 4. Even though
the dent is not
clearly visible, there is still a large amount of wrinkling
visible in the core near the
impact-side facesheet (upper side in photo).
Cratered
-
UCSD FAA JAMS Paper 2014 49
Figure 31. Sectioned view of thin panel impacted at 23.67 m/s
and 40 deg (Test 005).
An example of severe core damage is shown in Figure 32. This
panel was
impacted at 50 m/s (nominal) velocity at a 40 degree glancing
angle. The energy
level was 59 J, resulting in a significant-sized and
clearly-visible crater.
Figure 32. Sectioned view of thin panel impacted at 44.76 m/s
and 40 deg (Test 006).
For comparison, the same testing conditions were replicated on
the thicker
facesheet specimens. Many of the thick facesheet specimens did
not have
-
UCSD FAA JAMS Paper 2014 50
readily visible dents, and the peak dents were essentially
negligible. However,
these specimens have not yet been sectioned, so the internal
damage state of
the core has not been revealed. More work is needed to draw
qualitative
conclusions from these specimens. Figure 33 compares the
relationship
between impact energy and peak dent depth for the thick
facesheet specimens.
Figure 33. Impact energy vs. peak dent depth of thick
specimens.
At similar energy levels, the thicker facesheet specimens show
lower dent
levels than their thin facesheet counterparts, as expected. The
two outlying (low
dent-level) points for the 25 degree impacts represent two tests
in which the ice
crushed before impact or broke apart upon impact, and thus less
energy was
absorbed by the panel. Similar situations occurred for the 25
m/s (nominal)
speed tests at 40 degrees, but the 50 m/s (nominal) speed test
created
significant visible damage (crater) to the panel.
Cratered
-
UCSD FAA JAMS Paper 2014 51
3.4 Discussion
3.4.1 Low Velocity Blunt Impact
There exists a correlation between dent depth and impact kinetic
energy
separated by tip radius. Most of the dent depths seem to overlap
at lower energy
levels (i.e., no radius dependence), with the exception of
cracked facesheet
cases. When higher energy levels are considered, it is clear
that resulting dent
depths separate based on tip radii. In the case of a sharp tip
of radius 12.7 mm,
cracking of the facesheet occurred at 4 J. Tip radius 25.4 mm
did not crack the
facesheet until 14 J. All other tip radii only created a shallow
dent at similar
energy levels, with a maximum dent of 1.58 mm.
The majority of dents had maximum depth below 1.0 mm. While
visibility
of these dents increased with energy, there is no clear
indication that it is
dependent on tip radii. It is therefore important to note that
these shallow dents
are difficult to detect without careful visual inspection. Some
dents are only
visible when carefully observed under oblique lighting. Dent
depth relaxation may
further obscure the ability to visually detect damage.
In all cases, internal damage was present when the specimens
were
sectioned. While the analysis of this damage is still in
progress, relationships
between damage modes, damage extent, dent depths, tip radius,
and kinetic
energy are being sought.
3.4.2 Ground Hail Impact
For ice sphere impacts, these results reflect that lightweight
sandwich
panels (rudder) have low resistance to ice impacts. The presence
of any visible
-
UCSD FAA JAMS Paper 2014 52
dent suggests there is going to be significant internal core
damage as well. The
thick facesheet specimens do not show much visible surface
damage but it is
anticipated that they will have significant internal damage. It
is possible that since
these specimens are stiffer, the level of damage could be more
variable due to
the potential dependence on the state of the hail ice on impact.
Often, the hail
may break upon impact and impart less energy onto the panel than
if it were to
stay intact and rebound from the surface, thus imparting the
full impulse on the
panel. It is also important to note that the peak dents are not
highly accurate
measurements due to the variability in the measurement process,
particularly at
small depth levels (below 0.05 mm). The slightest bump or
inconsistency of the
panel surface will offset the results, and thus the dent data
are only significant
within a 0.02-0.05 mm range of measurement accuracy.
3.5 Sandwich Panel Impact Conclusions
3.5.1 Low Velocity Blunt Impact
Dent depth increases linearly with increasing energy levels for
lower
energy levels up to 6 J. The rate of increase changes with
different tip
radii, and is more evident in higher energy levels beyond 6
J.
Blunter impacts (larger radius) produce more shallow dents that
exhibit
more relaxation over time. Percent relaxation tends to be low,
however,
and reaches final state within 24 hours.
Composite sandwich panels with thin facesheets offer little
resistance to
impact damage. Energy levels as low as 2 J produce both
internal
damage with all impact tip radii investigated.
-
UCSD FAA JAMS Paper 2014 53
3.5.2 Ground Hail Impact
Any visible/measurable dent on the surface corresponds to
significant core
damage, which ranges from core wrinkling to core fracture.
Barely-measureable dents have been found to produce significant
core
damage.
With 50.8 mm diameter ice spheres, speeds as low as 50 m/s are
capable
of fracturing the skin and leaving highly-visible facesheet
cracking and
crater in the sandwich panel. Thicker skins allow higher impact
energies
before fracturing the facesheet.
4.0 Benefits to Aviation UCSD’s research focused on impacts
threats which can produce non-
visible damage have several key benefits to aviation. These are
summarized
below:
4.1 GSE Blunt Impact
Understanding of damage produced from wide-area GSE impact
events
through experimental data and finite element analysis. The
results can
bring awareness to blunt impact damage modes and allow for
prediction of
where damage will likely take place.
Demonstrated inverted building block pyramid approach to
coordinated
tests that support development of accurate finite element
modeling
capabilities. Large-scale experiments first identified failure
locations and
-
UCSD FAA JAMS Paper 2014 54
modes. Then, small-sized element-level specimens were used to
provide
in-depth data supporting analyses and modeling methodology
development.
Established FEA models that were capable of predicting:
o onset of cracks that lead to large-scale damage and
degradation of
material/structural integrity, as well as correlating the
visibility to the
internal damage produced, and
o large scale structure deformation, failure behavior, and loss
of
global stiffness of substructure due to these failures.
Capabilities to model and predict damage of fuselage structure
based on
impact location. GSE speed and mass are key parameters to define
safe
GSE operations, and also provide guidance on what conditions
would
merit further inspection for damage.
Established modeling capabilities can also be used for design of
new
structures that may be more able to resist damage from
commonly-
occurring impact sources
Using NDE methods to determine whether or not large damage
has
occurred, to help in the decision on whether an aircraft should
be removed
from service for further (invasive) inspection.
4.2 Blunt Impact Damage to Sandwich Panels
Increase understanding of the damage modes and governing
mechanisms
in sandwich composite honeycomb structures when subjected to
blunt
impact conditions.
-
UCSD FAA JAMS Paper 2014 55
Understanding of how internal damage develops provides insight
into
improving the impact resistance of composite sandwich
panels.
Establishing a correlation between the extent of core damage and
visible
external damage allows for a prediction of damage based on
external
observations.
5.0 Future Work and Follow-On Activity
The project activities summarized herein are ongoing. Future
planned and
recommended activities are described below. In all areas,
education/training and
dissemination of results via reports, workshops, and seminars is
a key topic.
5.1 GSE Blunt Impact
• Develop FEA models that incorporate aircraft floor joints and
floor beams to
better represent fuselage structure. Various impact locations
will be explored
to determine how this affects development of damage.
• Studies to be performed incorporating different stringer and
shear tie
geometries to establish how these components’ configurations
affect blunt
impact damage formation and progression.
• Explore NDE methods for finding major damage, including frame
cracks and
shear tie failures. Methods may involve using a series of
sensors at various
locations where one sensor sends a signal and the others
receive.
• Perform physical quarter-barrel or half-barrel fuselage test
with specimen
having floor structures, to validate FEA modeling
capabilities.
-
UCSD FAA JAMS Paper 2014 56
– full or ½ barrel – needs to include internal floors, joints,
and other
structure
• relationship made to previous lab-tests – work out methodology
to
account for effect of various threat types and contact
locations, as
well as boundary condition and panel size effects
• internal joints affecting load paths – e.g., proximity of
impact to
passenger and cargo floor or locations of high stiffness
transition –
i.e., Regions 1, 2, and 3
– impact with actual GSE vehicle (or rolling-mass
representative) –
incoming energy, allow to decelerate
– glancing impact effects
• investigate how contact zone moving across surface affects
1)
damage produced and 2) damage visibility (more surface
shear,
scuffing marks?)
• modeling of the surface shear effects on damage formation –
must
account for friction
• Blunt Impact on Other Structure Types
– metal-composite hybrid – could metal frames yield and “pull”
skin in,
leaving more visible dent?
– all-metal construction (can serve as baseline particularly for
dent
formation)
– aged metal structures from in-service fleet or retired
aircraft – can blunt
impact increase widespread fatigue damage (WFD) state?
-
UCSD FAA JAMS Paper 2014 57
– sandwich construction
– non-fuselage locations – e.g., lower wing and empennage
surfaces
Continued developments to establish high fidelity FEA modeling
capability
– accurately predict damage initiation, progressive failure
process,
damage extent, energy absorption
– accounting for interlaminar failures within context of
laminated shell
element modeling approach to accommodate large-structure
modeling,
also account for how interlaminar failures interact with
in-plane failures
predicted by classical lamination
Define generally-applicable visibility metrics and failure
criterion compatible
with FEA – focus on when crack formation is visible
Develop and refine reduced order models
– estimate damage onset for wide parameter range: GSE mass,
velocity,
impact location
– relate test results to GSE field operations
Education/Training: dissemination of results, workshops.
5.2 Blunt Impact Damage to Sandwich Panels
Section test specimens and relate observations of internal core
damage
depth and span to external visibility.
Relate extent of core damage to impact sources and external
damage
formation.
-
UCSD FAA JAMS Paper 2014 58
Establish prediction capability within explicit FEA simulation
framework to
predict blunt impact induced damage modes, size, and severity,
with
relationship extended to post-impact residual strength
reduction.
Detailed photography and microscopy of core damage modes and
morphology.
Conduct post-impact facesheet peel/fracture tests, and correlate
results with
FEA. Relate to core damage modes and morphology.
Investigate effect of multi-hit and impact adjacency
Compression after impact testing of the panels tested - what is
the residual
strength of panels that have experienced the types of damage
that have been
observed?
Use photogrammetry to further characterize the surface dents
and
deformations of the panels.
Investigate the effect of layup orientation on impact damage
thresholds – how
does layup affect final delamination morphology?
Investigate impact onto stiffened skin panels and sandwich
panels.
5.3 Low Velocity Blunt Impact of Monolithic Laminates
Residual strength of damaged panels. Is there a significant
strength
reduction between different damage modes?
Expand experimental testing to modeling and prediction of
damage.
Explore efficient and effective NDE methods to determine core
damage.
-
UCSD FAA JAMS Paper 2014 59
7.0 References 1. International Air Transportation Association
2005, “Ground Damage
Prevention Programme Targets 10% Cost Reduction,” Industry
Times, Edition 7, September, Article 4.
2. Kim, H. and Kedward, K. T., “Modeling Hail Ice Impacts and
Predicting Impact Damage Initiation in Composite Structures,” AIAA
Journal, Vol. 38, No. 7, 2000, pp. 1278-1288.
3. Kim, H., Kedward, K.T., and Welch, D.A., “Experimental
Investigation of High Velocity Ice Impacts on Woven Carbon/Epoxy
Composite Panels,” Composites Part A, Vol. 34, No. 1, 2003, pp.
25-41.
4. Rhymer, J., Kim, H., and Roach, D., “The Damage Resistance of
Quasi-Isotropic Carbon/Epoxy Composite Tape Laminates Impacted by
High Velocity Ice.” Composites Part A: Applied Science and
Manufacturing, DOI: 10.1016/j.compositesa.2012.02.017. Available
online 3 March 2012.
5. "Military Handbook 17 - MIL-HDBK-17-3F: Composite Materials
Handbook, Volume 3 - Polymer Matrix Composites Materials Usage,
Design, and Analysis, Chapter 7," U.S. Department of Defense. June
2002, Chapter 7.
6. Dassault Systèmes Simulia Corp., Providence, RI, USA, “ABAQUS
v.6.11 Analysis User’s Manual,” Section 31.5, 2011.
7. H. M. Matt and F. Lanza Di Scalea, “Macro-fiber composite
piezoelectric rosettes for acoustic source location in complex
structures,” Smart Mater. Struct., vol. 16, no. 4, pp. 1489–1499,
Aug. 2007.