DAMAGE TOLERANCE AND NOTCH SENSITIVITY TEST METHOD DEVELOPMENT FOR SANDWICH COMPOSITES

Daniel O. Adams, Bradley Kuramoto, and Marcus Stanfield

Department of Mechanical Engineering University of Utah

Salt Lake City, UT 84112

ABSTRACT

Although the development of coupon-level test methods for notch sensitivity and damage tolerance of monolithic composites has reached a high level of maturity, less attention has been directed towards the development of standardized test methods for sandwich composites. For monolithic composites, these test methods are commonly performed as part of composite materials characterization programs to provide design allowables associated with notch sensitivity and damage tolerance. Additionally, results from such tests have been used to validate numerical modeling approaches for simulating damage progression in composite laminates.

The objective of this research is to develop and evaluate candidate notch sensitivity and damage tolerance test methodologies for sandwich composites and to propose specific methodologies and configurations for standardization. Two test methodologies have been developed for damage tolerance testing of sandwich composites. The first utilizes an end-loaded Compression After Impact (CAI) configuration. The second methodology consists of a four-point flexure test configuration in which the damaged facesheet of the composite sandwich panel may be loaded in compression or tension. For both test methodologies, research has focused on their limits of applicability, the required sandwich specimen dimensions, and their recommended testing procedures. For notch sensitivity, initial research has focused on the development of test methods for sandwich composites under in-plane loading, with an expected progression to more complex out-of-plane loading cases.

Additionally, finite element analysis with progressive damage modeling is being assessed for use in predicting strength reductions of sandwich composites with impact damage and notches (open holes). Previously developed analysis methodologies for monolithic composite laminates are being extended to sandwich composites using ABAQUS with add-on software NDBILIN. Experimental validation is being aided through the use of Digital Image Correlation (DIC) to provide full-field deformations and strain fields associated with damage progression.

INTRODUCTION

Damage tolerance test method development for monolithic composite laminates

has reached a high level of maturity relative to sandwich composites. There currently

exists a standardized damage resistance test method (ASTM D7136 [1]), as well as a

damage tolerance standard (ASTM D7137 [2]) for monolithic composites. For sandwich

composites, relatively little effort has been placed towards arriving at one or more

standardized damage tolerance test methods for sandwich composites until recently.

However, the need for such a standardized test recently has been identified by the

aviation industry in general, and ASTM Committee D30 on Composites specifically. In

October 2009, a joint SAMPE/ASTM D30 Panel Session on “Damage Resistance and

Damage Tolerance of Sandwich Structures” was held in Wichita, KS. This panel,

organized by the PI of this research project, focused on the current status and future

needs of damage resistance and damage tolerance test methods for sandwich

composites. Considerable interest was identified for standardized damage resistance

and damage tolerance test methods for sandwich composites, and work commenced

towards a damage resistance test method. In 2011, a damage resistance standard

practice for sandwich composites (ASTM D7766 [3]) was published by ASTM

Committee D30. Later in 2011, a second joint SAMPE/ASTM D30 Panel on “Damage

Resistance of Composite Sandwich Structures” was held in Ft. Worth, TX. At this

panel, interest in a subsequent damage tolerance test method for sandwich composites

was reaffirmed. At a subsequent ASTM D30 meeting, the University of Utah (Dr. Dan

Adams) was identified as the lead institution to lead the effort towards the development

of a standardized test method in damage tolerance.

One objective of this research is to identify and develop one or more test

methodologies for assessing the damage tolerance of sandwich composites. The

possibility of having multiple test methods or practices exists in order to accommodate

the largest number of sandwich configurations and loading conditions as well as

intended usages of sandwich composites. Additionally, efforts have focused on

investigating possible methodologies to scale test results from small-scale sandwich

panel tests to larger sandwich structures. Following an initial assessment of candidate

test methodologies, two have been selected for further development: edgewise

compression after impact and four-point flexure after impact. Testing has been

conducted to evaluate these methods, and finite element analysis performed using

ABAQUS/NDBILIN to investigate specimen configurations and predict the residual

strength of the specimens.

Another objective of this research is to investigate the notch sensitivity of

sandwich composites. Although the development and numerical modeling of coupon-

level test methods for notch sensitivity of monolithic composite laminates has reached a

high level of maturity, little research has been performed towards the development of

such test methods and numerical modeling methodologies for sandwich composites.

Two standardized tests currently exist for notched monolithic composites under in-plane

loading: open-hole tension (ASTM D 5766 [4]) and open-hole compression (ASTM D

6484 [5]). However, currently no standardized test methods exist for assessing notch

sensitivity of sandwich composites. A survey of the literature has been conducted to

determine candidate notched sandwich composite test methods. Based on this review,

the initial focus of this research has been on an open-hole edgewise compression test.

Emphasis has been placed on both suitable specimen and hole sizing as well as the

numerical simulation of the test specimen to predict the notched strength. Digital Image

Correlation (DIC) is used during testing to provide full-field deformations and strain

fields.

LITERATURE REVIEW

Research began with a literature review to identify candidate damage tolerance

test methodologies for sandwich composites. Emphasis was focused on both papers

and reports in the open literature as well as common practices in the aerospace industry

and from commercial test labs. From the candidate test configurations identified, two

were selected for investigation and further development. The first methodology utilizes

an end-loaded compression after impact (CAI) test configuration. Second, a four-point

flexure test methodology was identified for evaluating post-impact performance where

the damaged facesheet can be loaded in compression or tension depending on

specime

toleranc

A

tests tha

each ca

open-ho

two ex

compos

compos

hole, eit

methodo

been se

A

While a

D5766 [

introduc

is expec

used pre

T

determin

test wo

monolith

open-ho

en setup.

ce test meth

Additionally,

at could be

andidate tes

ole sandwic

isting stan

ites (ASTM

ites (ASTM

ther through

ology, also

lected as th

Another can

standardiz

[4]), no sta

ce sufficient

cted to be

eviously by

The out-of-

ned using a

ould be pe

hic compos

ole composi

Figure 1. F

Further de

hodologies

a literatur

investigate

st method

ch compres

ndardized

M C364 [6

M D6484 [5

h the entire

used in a

he initial foc

ndidate tes

zed open-h

andardized

t tensile loa

required, a

Tomblin et

plane flexu

a four-point

erformed s

sites (ASTM

ite sandwic

Four-point

evelopment

is currently

re review h

ed for sand

identified.

sion test. T

tests: the

6]) and th

5]). This ca

e sandwich

previous r

cus of this r

t is an in-p

hole tension

in-plane te

ad into the s

as evidence

t al. [7].

ural streng

t flexure te

similarly to

M D7249 [8

ch specimen

t flexure te

t and evalu

y underway.

has identifie

dwich comp

The first c

This candida

edgewise

he open-ho

andidate te

specimen

research inv

research in

plane loade

n test exist

ensile tests

sandwich s

ed by the

gth of notc

st configur

the stand

8]), but wit

n.

est for notc

uation of th

.

ed several

posites. A b

candidate t

ate test wo

e compres

ole compre

st would fe

or through

vestigation

vestigation

ed, open-h

ts for mon

exist for s

specimen, r

open-hole

ched sandw

ration as sh

dardized fo

th modificat

ched sandw

he two ide

potential n

brief descrip

test is an in

ould incorpo

ssion test

ession test

eature a ce

h one faces

by Tombli

n.

ole sandw

olithic com

sandwich c

relatively co

tension tes

wich comp

hown in Fig

our-point f

tions to ac

wich comp

entified dam

notch sensi

ption follow

n-plane loa

orate aspec

for sand

t of mono

entered cir

sheet. This

in et al. [7]

ich tensile

mposites (A

composites.

omplex fixtu

st configur

posites can

gure 1. Su

flexure tes

ccommodat

posites.

mage

itivity

ws for

aded,

cts of

dwich

olithic

rcular

s test

] has

test.

ASTM

. To

uring

ation

n be

uch a

t for

te an

A

bending

[9]. An

subjecte

limited to

modified

the tens

Fig

P

shear (M

compos

test me

sandwic

A three-poin

moment in

edge V-no

ed to both

o a notch lo

d to four-po

ion or com

gure 2. Fo

Parmigiani e

Mode III te

ites. As sho

thodology

ch composit

nt flexure t

n a sandwic

otch was u

static and

ocated on t

oint flexure

pression sid

our-point in

et al. at Or

earing) test

own in Figu

will need t

tes.

test config

ch composi

used in th

fatigue te

the tension

as shown

de or a dou

n-plane flex

regon State

method to

ure 3, this t

to be inves

uration has

te specime

he compos

sting. This

side of the

in Figure 2

uble edge n

xure for no

e University

o investigat

test uses a

stigated fu

s been us

en by placin

ite sandwic

s three poin

e sandwich

2, the edge

notch could

otched san

y [10] has d

te the notc

keyhole no

rther for it

ed to indu

ng the sand

ch specim

nt flexure c

. However,

e notch cou

be used.

ndwich com

developed

ch sensitivit

otch in the

s possible

uce an in-p

dwich on its

en, which

configuratio

if the test w

uld be place

mposites.

an out-of-p

ty in mono

specimen.

applicabili

plane

side

was

on is

were

ed at

plane

olithic

This

ty to

F

RESEAR

T

general

standard

compari

both of t

element

predictin

Sandwi O

sandwic

Figure 3. O

RCH FOCU

The primary

test metho

dization.

son of sand

the propose

t analyses

ng failure st

ch Compo

One of the

ch compos

Out-of-plan

US: SAND

y focus of th

odologies, e

However,

dwich pane

ed test met

have bee

trengths of

osite Edgew

most com

sites is e

ne shear te

WICH COM

he sandwic

edgewise c

a second

els with idea

hodologies

en perform

damaged s

wise Comp

mmonly us

edgewise

est for notc

MPOSITE D

h damage t

compressio

ary focus

alized dam

. Additiona

ed to esta

sandwich pa

pression A

sed method

compressi

ched sandw

DAMAGE T

tolerance e

n and four-

is to pro

mage as we

lly, digital im

ablish prop

anels.

After Impac

ds to asse

on testing

wich comp

TOLERANC

effort is to e

-point flexu

ovide a re

ll as impac

mage corre

per specim

ct

ess damag

g. Typicall

posites.

CE

evaluate the

ure, for pos

sidual stre

t damage u

elation and

men sizing

ge toleranc

ly, end-loa

e two

ssible

ength

using

finite

and

ce of

aded

Compression After Impact (CAI) testing on sandwich structures has been performed

using ASTM D7137 [2], the standardized method for CAI testing of monolithic composite

plates, as a reference. This test method requires that the loading ends of the specimen

are machined to be parallel within tight tolerances, and then loaded edgewise until

failure. The specimen sides are typically supported along their edges to prevent

buckling failure of the sandwich panel. This test methodology has been used previously

on Nomex honeycomb panels as well as stitched and unstitched foam panels with

success [11-14]. On undamaged or barely damaged high-strength specimens, however,

edgewise loading using this test configuration can lead to undesired failure modes such

as end crushing before failure in the gage section [12].

Tomblin et al. [11-12] and Butterfield and Adams [13-14] have used edgewise

compression to assess the damage tolerance of sandwich composites. The sandwich

panels tested in both laboratories were 216 mm (8.5 in.) wide by 267 mm (10.5 in.) in

height. Impactor diameters of 25 mm and 76 mm (1.0 in. and 3.0 in.) were used.

Whereas the Wichita State University researchers [11-12] used honeycomb core

sandwich panels, the subsequent University of Utah investigation [13-14] focused on

unstitched and stitched foam sandwich composites. In both investigations, sandwich

composites with 9.5 mm and 19 mm (0.375 in. and 0.75 in.) thick cores were tested.

Composite sandwich specimens were clamped on the upper and lower loading edges,

and knife-edge supports were used on the specimen sides. The test fixture used for the

University of Utah investigation is shown in Figure 4. In this investigation, strain gages

were placed 25 mm (1.0 in.) from the lower corners in order to align the panels prior to

testing. A hemispherical alignment stage, placed under the test fixture, was used in

conjunction with the corner strain gages and a small preload to provide proper panel

alignment.

Figur

B

authors,

further

necessa

without

external

potting o

material

Addition

the form

boundar

whereas

selection

CAI san

e 4. Comp

Based on th

several a

investigatio

ary to reinf

premature

clamp-typ

of the entir

may be f

nally the ch

m of knife

ry condition

s clamped

n of suppo

ndwich pan

pression Apanel

he review o

aspects of

on prior to

force the c

end failure

e restraint

re end of th

filled or re

oice of side

edge supp

n desired;

supports a

ort condition

nel testing

After Impacls used at t

of the litera

sandwich C

o test stan

core at the

es due to ei

is typically

he sandwic

eplaced wit

e supports

ports or cla

knife edge

are used to

ns are imp

to larger

ct (CAI) testhe Univer

ature as w

CAI testing

ndardizatio

e specimen

ither facesh

y imposed

ch specime

th a highe

must be c

amp-type s

e supports

o minimize

portant whe

size struct

t fixture forsity of Uta

well as from

g have bee

n. For ma

n ends to

heet broom

on the san

en is not de

er strength

considered.

supports, d

are used t

e panel rot

en consider

tures. The

or composiah.

m past exp

en identifie

any core

allow prop

ming or buc

ndwich pan

esirable. R

potting or

Such sup

depending

to replicate

tation at th

ring scaling

e best-suit

ite sandwi

eriences o

d that requ

materials,

per end loa

kling. Sinc

nel at the e

Rather, the

r filler mat

pports may

on the typ

e a pinned

he edges.

g the resul

ted size of

ch

f the

uired

it is

ading

ce an

ends,

core

erial.

take

pe of

joint

The

lts of

f the

sandwich panel as well as the boundary conditions may both be selected to represent

the composite structure of interest. Tomblin et al. [11-12] investigated the effects of

scaling the planar dimensions of sandwich specimen both experimentally and using

finite element analysis. For sandwich configurations with thin facesheets, the residual

strength increased with panel width. However, this increase was not observed for

sandwich configurations with thicker facesheets. Results suggest that CAI-level tests

may be useful for scaling if the width-to-damage ratio is sufficient to avoid edge effects.

This topic is currently under investigation as part of the standardized test method

development.

Testing conducted to evaluate the sandwich edgewise CAI test method used

composite sandwich panels fabricated from 1.5 mm (0.06 in.) thick G11

fiberglass/epoxy and 0.38 mm (0.015 in.) thick carbon/epoxy facesheets bonded to 13

mm (0.5 in.) thick Nomex honeycomb core using FM300 film adhesive. The sandwich

specimens were machined to 216 mm (8.5 in.) in width by 267 mm (10.5 in.) in height,

the same dimensions as used in previous studies [11-14]. To prevent the loaded ends

from crushing prior to failure in the gage section, they were potted using a room

temperature cure epoxy potting compound. In order to minimize the variability in initial

test results due to variations in damage states from impact testing, the use of idealized

damage in the form of circular holes machined in one facesheet was employed rather

than actual drop-weight impacting. By testing these sandwich configurations with and

without the idealized damage, these tests provided an initial comparison of residual

strength between different damage sizes. Following these initial idealized damage

states, actual impact damage was created in composite sandwich panels.

Testing was conducted using the same fixture as used in previous testing at the

University of Utah (Figure 4). The specimens were preloaded and subsequently aligned

using the hemispherical stage to adjust the strain readings from strain gages bonded

near the bottom corners of both facesheets. Two different idealized damage sizes, 25

mm (1 in.) and 76 mm (3 in.) through holes on the damaged facesheet were compared

to undamaged and impact damaged specimens. The impact damaged specimens were

impacted at an energy level of 3.4J (30 in-lbf) in accordance with ASTM D7766 [3]. As

shown in Figure 5, the damage state had a significant impact on the residual strength,

especially for the sandwich composites with the relatively thin carbon/epoxy facesheets.

The strength reduction of the impacted specimens with carbon/epoxy facesheets was

between that measured in specimens with a 25 mm (1 in.) and 76 mm (3 in.) hole. The

sandwich composites with the thicker G11 facesheets were not affected as significantly,

with less strength reduction due to all forms of damage investigated.

Figure 5. Normalized results of edgewise compression after impact testing.

Sandwich Four-Point Flexure After Impact

The second test method selected for investigating damage tolerance of sandwich

composites is the four-point flexure after impact test. As shown in Figure 6, this test

configuration allows for a constant bending moment and zero shear stress in the central

gage section of the specimen. The four-point flexure configuration allows for the

damaged facesheet to be placed under either compression or tension loading

depending on how the specimen is placed into the fixture (damaged facesheet placed

either upward or downward). In a majority of studies performed to date, the damaged

facesheet has been loaded in compression. The University of Utah test set-up for four-

point flexure testing of sandwich composites is shown in Figure 7.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Undamaged 1" Hole 3" Hole Impacted

Glass/Epoxy

Carbon/Epoxy

Norm

alized Compression Strength 

Fig

 

 

Figure

gure 6. She

e 7. Four p

ear and be

point flexursandwich

ending mom

re test fixtucomposite

ment diagr

ure used foes at the U

ram for fou

or assessinniversity o

ur point fle

ng damageof Utah.

exure test.

e tolerancee of

Based on a review of the literature and past experiences of the authors, a

number of aspects of sandwich four-point flexure testing are under investigation in

anticipation of test standardization. The central test section, or region of the sandwich

panel between the inner supports, must be of sufficient length and width such that the

damage present in the panel is not influenced by the panel edges or the presence of the

inner loading rollers. Similarly the outer spans of the panel must be of sufficient length

to generate an adequate bending moment in the central test section prior to shear

failure in the outer spans or compression failure at the load application points. Shear

failure of the core in the outer spans can be avoided through the use of a higher density

core material, or the filling of the cells of a honeycomb core material. Similarly,

precautions must be taken to avoid crushing failure of the facesheet and/or core at the

loading points. Thus, the sandwich panel and span lengths of the test fixture must be

properly designed to prevent undesirable failure modes from occurring under four-point

flexure loading.

An initial round of testing was conducted to evaluate four-point flexure loading for

assessing damage tolerance of sandwich composites. The sandwich specimens were

sized in accordance with the ASTM D7249 [8] sizing guidelines. The sandwich panels

were fabricated using 0.5 mm (0.02 in.) carbon/epoxy facesheets bonded to 13 mm (0.5

in.) thick Nomex honeycomb core using FM300 film adhesive. The sandwich specimens

were machined to 216 mm (8.5 in.) in width by 813 mm (32 in.) in length, the same

dimensions as used in previous studies [13-14]. Initial tests were performed using a

760 mm (30 in.) outer support span and 229 mm (9 in.) inner loading span. Due to the

relatively low shear strength of the Nomex honeycomb, a higher density honeycomb

core material was spliced in the outer section to prevent core shear failures. Testing

was performed using idealized damage in the form of 25 mm (1 in.) and 76 mm (3 in.)

circular holes machined in one facesheet. Additionally, sandwich panels without

damage were tested to determine strength reductions due to the idealized damage.

As shown in Figure 8 the two idealized damage states had a significant effect on

the residual strength of the sandwich composites. The 25 mm (1 in.) facesheet hole

produced a 42% reduction in strength and the 76 mm (3 in.) hole produced a 60%

reduction in strength.

When discussing these test results at a recent ASTM Committee D30 on

Composites meeting, it was suggested that to simplify the specimen preparation

procedures (avoiding the splicing of higher density core in the outer sections), a greater

core thickness could be used in the test specimen, even if this thickness was greater

than for the intended application. The rationale was that the damage tolerance of a

sandwich composite under flexure is sensitive to facesheet damage, but insensitive to

the core thickness. Thus, the core thickness can be used as a design variable towards

preventing undesirable failure modes from occurring. However, the core thickness must

be sufficient to prevent damage from occurring in the opposite facesheet during the

impact event.

Figure 8. Normalized four-point flexure after impact strength of idealized damage specimens.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

No Hole 1" Hole 3" Hole

Norm

alized Flexural Strength 

RESEARCH FOCUS: SANDWICH COMPOSITE NOTCH SENSITIVITY

The initial focus of the sandwich notch sensitivity research has focused on the

measurement and prediction of the open-hole compression strength of composite

sandwich specimens. The selected specimen size was 114.3 mm (4.5 in.) in width and

152.4 mm (6.0 in.) in height. The diameter of the through-hole in the center of the

specimen was 19 mm (0.75 in.), producing the same 6:1 specimen width-to-hole

diameter ratio used with monolithic open-hole composite specimens [4-5]. The test

fixture was the same as used in ASTM D7137 [2] for compression-after-impact testing

of monolithic composite specimens Fixed end conditions were used on the top and

bottom ends of the specimen and knife edge supports were used on specimen edges.

The composite sandwich panels were fabricated using 0.5 mm (0.02 in.)

carbon/epoxy facesheets bonded to 13 mm (0.5 in.) Nomex honeycomb core using

FM300 film adhesive. The specimens were machined to their final measurements and

the core region at the specimen ends were potted using a room-temperature cure epoxy

to prevent end crushing during loading. The central through-hole was milled using a

UKAM Universal Metal Bond Core Drill.

Four strain gages were bonded 25 mm (1 in.) from the bottom and side edges for

use in aligning the fixture at a given preload. The front face was painted white and

speckled using black ink for Digital Image Correlation (DIC) measurements. The load

and crosshead displacements were also measured during the test. Initial test results

from the notched sandwich composite specimen show a 32% reduction in compression

strength compared to the unnotched specimen. Axial strains measured using DIC and

bonded strain gages were found to be in good agreement as shown in Figure 9.

Figure 9. Axial strain gage and DIC comparison from open-hole compression testing of composite sandwich specimens.

FINITE ELEMENT MODELING

The failure analysis of notched sandwich specimens was modeled using the

commercial finite element code ABAQUS. NDBILIN, a user-defined nonlinear material

model (UMAT) developed by Materials Sciences Corporation [15, 16] was used for

stiffness degradation based progressive damage modeling. Using this UMAT, stiffness

degradation is prescribed at the lamina level within the composite facesheets and is

shaped by inputting the failure criteria (ex: max stress/strain, Hashin, phase averaged

constituent), the stiffness degradation factors, and stiffness transition factor. A graphical

representation of the material model is shown in Figure 10. NDBILIN allows for two

different material inputs for fiber stiffness degradation and matrix stiffness degradation.

When matrix damage occurs, the stiffness reduction is applied to all matrix modes (22,

33, 12, 13, and 23).

0.00E+00

5.00E‐04

1.00E‐03

1.50E‐03

2.00E‐03

2.50E‐03

3.00E‐03

3.50E‐03

4.00E‐03

4.50E‐03

5.00E‐03

0 20 40 60 80 100 120

Strain (in/in)

DIC Index

DIC Left

Left Gage

DIC Right

Right Gage

In

monolith

quasi-iso

experim

displace

experim

F

specime

compres

during m

F

nitial progr

hic compos

otropic lam

ental resu

ement resu

ental result

Figure 11.

Following th

en, the AB

ssion testin

mechanical

Figure 10.

ressive dam

site open-h

minate of IM

ults from t

ults, show

ts.

ABAQUS

he initial m

BAQUS/NDB

ng of the co

testing usin

NDBILIN s

mage anal

hole tensio

M7/8552 ca

three spec

n in Figu

S/NDBILIN o

modeling of

BILIN mod

omposite sa

ng DIC wer

stiffness de

lysis using

n specime

rbon/epoxy

cimens tes

ure 11, w

open-hole

f the mono

deling meth

andwich spe

re in good a

egradation

g ABAQUS

en. The m

y, and pred

sted. The

ere in go

tension re

olithic com

hodology w

ecimens. F

agreement

n model.

S/NDBILIN

modeled sp

dictions we

e predicte

ood agreem

esults and

mposite ope

was applied

Full-field st

with predic

focused o

pecimen w

re compare

d load ve

ment with

model.

en-hole ten

d to open-

rains meas

cted strain f

on a

as a

ed to

ersus

the

nsion

-hole

sured

fields

from AB

useful in

at the s

that the

in.), cen

loading

Figu

A

agreeme

failure l

stiffness

experim

included

occurrin

capabilit

are curr

for face

analysis

BAQUS/ND

n determinin

pecimen b

114 mm (4

ter hole is s

in composi

ure 12. Co

Although the

ent with ex

oad. As s

s reduction

entally. T

d into the

g in sandw

ties to sim

rently under

esheet/core

s is being ex

DBILIN ana

ng minimum

oundaries

4.5 in.) wid

suitable for

te sandwic

mparison

e predicted

xperimental

shown in

and a con

hese differ

finite elem

wich open-

ulate both

r investigat

debonding

xplored.

lysis as sh

m specimen

and at the

e and 152

r investigati

h specimen

of Digital Ianaly

d strain field

results, th

Figure 13,

nsiderably h

rences sug

ment simula

hole comp

facesheet/

tion. The u

g, whereas

hown in Fig

n sizing req

location o

mm (6.0 in

ng notch se

ns.

mage Corrysis results

ds in the vi

he initial an

the finite

higher com

ggest that a

ations to p

ression tes

/core debon

use of cohe

s the poss

gure 12.

quirements

f the strain

n.) tall spec

ensitivity un

relation (les.

icinity of th

nalysis did

element s

mpression f

additional f

roperly mo

st. As a r

nding as w

esive eleme

sibility of in

These initi

to produce

n gages. R

cimen with

nder in-plan

eft) and fin

e open hol

not accura

simulation

failure load

failure mod

odel the fa

result, addi

well as face

ents is und

ncorporatin

al results w

e far-field st

Results sug

a 19 mm (

ne compres

ite elemen

le were in g

ately predic

predicts fu

than obse

des need t

ailure sequ

tional mod

esheet buc

der investig

g an insta

were

rains

ggest

(0.75

ssion

nt

good

ct the

urther

erved

o be

ence

eling

ckling

ation

ability

Figure 13. Open-hole compression load versus strain comparison.

SUMMARY

The two candidate test configurations under development for damage tolerance

of sandwich composites, edgewise compression after impact and four-point flexure after

impact, appear to be well suited for future ASTM standardization. The current focus of

the sandwich damage tolerance research is the evaluation of progressive damage

modeling approaches for predicting strength reductions in composite sandwich

structures. Additionally, a combined experimental/numerical methodology for scaling of

damage tolerance results from smaller test specimens to larger sandwich structures is

under investigation.

Research into the notch sensitivity of sandwich composites has focused initially

on in-plane compression loading of sandwich composites with centrally located open

holes. Progressive-damage finite element analysis has been performed using ABAQUS

with a user-defined nonlinear material model, NDBILIN. This modeling approach, used

previously for monolithic composite laminates, appears to be a promising approach for

sandwich composites. Experimental validation is being aided through the use of Digital

Image Correlation (DIC) to provide full-field deformations and strain fields associated

with damage progression.

0

500

1000

1500

2000

2500

3000

3500

0 0.002 0.004 0.006 0.008

Nyy (lbf/in)

Strain (in/in)

Experimental

ABAQUS/NDBILIN

REFERENCES

1. ASTM D7136, “Standard Test Method for Measuring the Damage Resistance of a

Fiber-Reinforced Polymer Matrix Composite to a Drop-Weight Impact Event,” ASTM International, West Conshohocken, PA, 2007.

2. ASTM D7137, “Standard Test Method for Compressive Residual Strength Properties of Damaged Polymer Matrix Composite Plates,” ASTM International, West Conshohocken, PA, 2007.

3. ASTM D7766, “Standard Practice for Damage Resistance Testing of Sandwich

Constructions,” ASTM International, West Conshohocken, PA, 2011.

4. ASTM D5766, “Standard Test Method for Open-Hole Tensile Strength of Polymer Matrix Composite Laminates," ASTM International, West Conshohocken, PA, 2011.

5. ASTM D6484, “Standard Test Method for Open-Hole Compressive Strength of

Polymer Matrix Composite Laminates," ASTM International, West Conshohocken, 2009.

6. ASTM C364, “Standard Test Method for Edgewise Compressive Strength of

Sandwich Constructions," ASTM International, West Conshohocken, PA, 2007.

7. Tomblin, J.S., Raju, K.S., Walker, T., and Acosta, J. F., "Damage Tolerance of Composite Sandwich Airframe Structures - Additional Results," DOT/FAA/AR-05/33, 2005.

8. ASTM D7249, “Standard Test Method for Facing Properties of Sandwich

Constructions by Long Beam Flexure," ASTM International, West Conshohocken, PA, 2012.

9. Shafiq, B. and Quispitupa, A., "Fatigue Characteristics of Foam Core Sandwich

Composites," International Journal of Fatigue, Vol. 28, pp. 96-102, 2006.

10. Parmigiani, J.P., Atanasov, N., and Hyder, I., “Failure of Notched Laminates Under Out-of-Plane Bending,” proceedings of the 2014 FAA JAMS Technical Review Meeting, Seattle, WA, March 25-26 2014.

11. Tomblin, J.S., Raju, K.S., Liew, J., and Smith, B.L., “Impact Damage

Characterization and Damage Tolerance of Composite Sandwich Airframe Structures,” Final Report, DOT/FAA/AR-00/44, 2001.

12. Tomblin, J.S., Raju, K.S., and Arosteguy, G., “Damage Resistance and Tolerance of

Composite Sandwich Panels – Scaling Effects,” Final Report, DOT/FAA/AR-03/75, 2004.

13. Butterfield, J.M and Adams, D.O., “Effects of Stitching on the Compression After

Impact Strength of Sandwich Composites”, Proceedings of the 49th International SAMPE Symposium and Exhibition, Long Beach, CA, May 2004.

14. Butterfield, J.M., “Damage Tolerance of Stitched Carbon Composite Sandwich

Structures,” M.S. Thesis, Department of Mechanical Engineering, University of Utah, 2006.

15. Schutte, J., Meeker, J.C., and Clay, S., "Improved Methodologies for Damage

Tolerant Composite Structure Qualification." 53rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference 20th AIAA/ASME/AHS Adaptive Structures Conference 14th AIAA. 2012.

16. Hayduke, D. and Meeker, J., "NDBILIN-AF: Sub-Cell Level Damage UMAT Users

Manual V2.0", Materials Sciences Corporation, Horsham, PA, December 2012.