The Effect of Glass Plate Thickness and Type and Thickness of the Bonding Interlayer on the Mechanical Behavior of Laminated Glass

doi:10.4236/njgc.2011.12007

Paper Menu >>

Journal Menu >>

New Journal of Glass and Ceramics, 2011, 1, 40-48

doi:10.4236/njgc.2011.12007 Published Online July 2011 (http://www.SciRP.org/journal/njgc)

The Effect of Glass Plate Thickness and Type and

Thickness of the Bonding Interlayer on the

Mechanical Behavior of Laminated Glass

Issam S. Jalham1, Omar Alsaed1

1Industrial Engineering Department, Univers ity of Jordan, Amman, Jordan.

Email: [email protected]u.jo

Received March 8th, 2011; revised April 5th, 2011; accepted April 12th, 2011.

ABSTRACT

In this work the effect of the type of the bonding interlayer (polyvinyl butyral (PVB) or Ethyl Vinyl Acetate (EVA)),

number of bonding layers, and the position and the thickness of the Glass plates on the maximum load capacity and

absorbed energy by laminated glass. Further more, this investigation presents a mathematical model that relates the

maximum force capacity of the glass laminated structure to the glass plate thickness, type and thickness of the interlay-

er regardless the position of the fixed glass plate. Both practical work results and the theoretical model indicate that the

maximum load capacity of laminated glass bonded with either PVB or EVA decreases as the interlayer thickness in-

creases. Moreover, the maximum load capacity for the glasses bonded with EVA is greater than those for the PVB

bonded ones under the same conditions. On the other hand, it was observed that that laminated glass absorbed energy

increases with the increase of the interlayer thickness and the increase of glass plate thickness.

Keywords: Laminated Glass, Polyvinyl Buty ral (PVB), Ethyl Vinyl Ace tate (EVA), Layer, Load Capacity

1. Introduction

Ceramics and glasses, which have strong ionic-covalent

chemical bonds, are very strong and stiff. They are also

resistant to high te mperatures and corrosion, but are brit-

tle and prone to failure at ambient temperatures. In con-

trast, thermoplastic polymers such as polyvinyl butyral,

which have weak secondary bonds between long chain

molecules, exhibit low strength, low stiffness, and a sus-

ceptibility to creep at ambient temperatures. These po-

lymers, ho wever, tend to b e extremely ductile a t ambient

temperatures. When combine glass and polymer to form

a laminated glass, some change in the maximum load

capacity will occur, which depends on both the glass and

polymer type. This led to investigate how the glass

thic kness a nd t he type and n umber of la minate d int erlay-

er affect the maximum load capacity of laminated glass

as well as their effect on the absorbed energy.

2. Literatu re Rev iew

Laminated glass consists of two or more glass plies

bonded together with an elastomeric interlayer, usually

polyvinyl butyral (PVB) or Ethyl Vinyl Acetate (EVA).

Afte r breakage, the interlayer holds the resultant glass

shards in place and, in most cases, the glass remains in

the frame when laminated glass fractures. This post-

breakage characteristic of laminated glass has made it

desirable for use in vehicle windshields for decades be-

cause it makes the occupant safer from glass shards than

other glazing materials.

The shea r modulus st ud ie s we r e c ar r ied o ut by Que net t

[1] and Hooper [2]. Quenett [1] noticed that when the

interlayer thickness decreases, shear modulus increases

and reported that the condition of the interlayer is a con-

trolling factor in static bending and dynamic impact re-

sistance. Hoo per [2] confir med the results of Quenett [1] .

He stated that after testing glass beams in four points

loading with varying temperatures and interlayer hard-

ness, he found that t he shear modul us of the interla yer is

inversely prop ortional to the interla yer thickness and a lso

mentioned that plasticizer contents, ambient tempera-

tures, and load durations are the primary factors control-

ling bending resistance of laminated glass. He attributed

this behavior to the “thermoplastic” nature of the inter-

layer, stating the decreased bending stiffness was the

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Interlayer on the Mechanical

Behavio r of Laminated Glass

primary disadvantage to a r c hitectural laminated glas s.

Strength of the monolithic and laminated glasses tak-

ing in to ac count the geo metr y and thi ckne ss of t he te sted

plates was studied by several researchers. For example,

Pilkington Ltd. [ 3] co mpared monolit hic glas s stre ngth to

the strength of laminated glass specimens made of sheet

and float glass. They found that, at normal temperature,

laminated glass specimens exhibit the same strength as

monolithic glass specimens having the same rectangular

dimensions and glass thicknesses. On the other hand,

Linden et al. [4] conducted a non-destructive test on

monolithic, layered, and laminated glass specimens in-

strumented with strain gages. They concluded that lami-

nated glass strength and monolithic glass strength ap-

peared to be equivalent at normal temperatures; and the

strength of la minated gla ss specimens approached that of

layered glass specimens at elevated temperatures. In ad-

dition, Norville [5] tested two laminated glass specimen

of sizes 38 x 76 and 66 x 66 in. destructively. His de-

structive experimentation also showed that the strength

of laminated glass specimens is the same or greater than

that of monolithic specimens having the same rectangular

dimensions and nominal thicknesses under similar load

conditions.

Keller [6] used novel method to measure the delami-

nating energy in laminated glass in the rele vant dynamic

range. He found that increasing the interlayer thickness

improves the penetration resistance of laminated glass

because more energy can be absorbed in the high speed

delimitation process since the interlayer is simply less

like to tear.

In contrast to the results of the above mentioned re-

searches contradiction was reported in Nagalla et al. [7] ;

Minor and Reznik [8]. Nagalla et al. [7] in their ad-

vanced theoretical work compared layered glass to mo-

nolithic. They discovered that some aspect ratios of the

layered glass experienced lower principal stresses than

monolit hic glass subje cted to unifor m, transver se loading

in some ranges of the loading. They concluded that the

strength factor of 0.6 used by some building codes for

laminated glass may be too low for many window geo-

metries and design pressures.

Minor and Reznik [8] destructively tested three sizes

of laminated glass specimens (33 x 66, 38 x 76, and 66 x

66 in.) with an 0.030 in. interlayer, and compared the

resulting failure pressures to those from tests on mono-

lithic glass specimens having the same rectangular di-

mensions and nominal glass thicknesses. They intro-

duced four variables, which are: glass thickness, glass

type, temperature, and damage to one plate of glass (i.e.,

damage to tension or compression side) . Their te sting led

to the following geral conclusi ons:

1) Laminated glass specimens tested at room tempera-

ture have approximately the same failure pressure as

monolithic glass specimens having the same rectangular

dimensions and nominal glas s thicknesses;

2) As temperature increases laminated glass behavior

migrates towards the layered glass model;

3) Laminated glass specimens having twice the no-

minal glass thickness of monolithic specimens display

strength greater than or equa l to twice t he strengt h of the

monol ithi c specimens.

Some researchers investigated the effect of tempera-

ture on the properties of glass. Linden et al. [9] con-

ducted non-destructive testing on two different plate

geometries. First, they tested the same plate geometry

(60 x 96 x 1/4 in.) as used in the parent report to study

load duration and temperature effects. Second, they

tested a different geometry (55-1/8 x 57-1/8 x 3/8 in.)

with two interlayer thicknesses (0.030 and 0.060 in.) to

study the effects of interlayer thickness on strength and

deflection. They conducted destructive tests on one plate

geometry (60 x 96 x 1/4 in.) at room temperature and at

170°F. Perusal of their data indicates that while load du-

ration and elevated temperatures acting individually re-

duce the s tr uctural rigidity of the laminated glass, the two

factors do not interact, producing a greater combined

reduction in laminated glass strength. Weller [10], Used

experimental study to compare different interlayer mate-

rials in laminated glass in respect to their structural be-

haviour. The material properties above the verification

temperature clearly showed the temperature dependency.

The relaxation times fall with increasing temperature and

the shear stress gets smaller.

Theoretical modeling of the glass behavior was also

carried out by many researchers. Linden et al. [4 ] derived

theoretical results through the finite difference solution

and compared experimental and theoretical results. They

concluded that the theoretical finite difference model for

monolithic and layered glass appeared to be acceptable

for the one glass plate geometry tested. Moreover, Behr

and Kremr [11] used experimental validation of a me-

chanics-based finite element model for architectural la-

minated glass units subjected to low velocity and two

gram projectile impacts. The impact situation models a

scenario commonly observed during severe windstorms.

This study confirmed the ability of an analytical finite

element model to predict accurately the peak strains in

representative architectural laminated glass units as a

function of impact velocity. Correlations between peak

radial strains computed using finite element analysis and

those measured experimentally were close, with the av-

erage difference between analytical predictions and ex-

perimental data being 7.7%.

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Int er layer on the Mechanical

Behavio r of Laminated Glass

Zang et a l. [12] investigation focused on the use of the

3D discrete element method to study the impact fracture

problem of laminated glass. The glass and the (PVB) of

laminated glass plane are discretized to uniform rigid

spherical elements. This investigation showed that the

accuracy of the 3D model and numerical analysis code

are more validated in the elastic range by comparing with

FEM.

Recently, Belies [13] compared (PV B) with st iffe r and

stronger interlayer Sentry Glass Plus (SGP). After brea-

kage of both glass sheets the load decreased to a rela-

tively low level (typically between 2 kN and 3 kN) be-

fore the broken glass pieces and interlayer started again

to buil d up co mpres si ve a nd tensile stre sse s, respe ctively.

Subsequently, the load slightly increased again and after

reaching the maximum, it decreased significantly (to less

then 0. 3 kN). When s ubje cted to in-plane b endi ng (buc k-

ling prevented), the post breakage residual resistance is

relatively poor for both interlayers, as illustrated above.

The residual load-bearing capacity was very limited and

far below the initial glass strength.

It is clear from the above review that the research work

focused on the comparison between the strength of mo-

nolithic and laminated glases and did not take into con-

sideration the bonding interlayer thickness, and the posi-

tion and thickness of the glass plates. Furthermore, the

main bonding material in these studies is PVB. This in-

vesti gatio n d iffe rs fr o m the ab o ve me ntio ned one s in that

it concentrates on how the glass thickness and the type

and number of laminated interlayer affect the maximum

load capacity of laminated glass as well as their effect on

the absorbed energy.

Details for the preparation of the mullite ceramic tile

(900 mm × 1800 mm × 5.5 mm) were reported in [4].

The raw materials were as follows: 50 wt% - 55 wt% fly

ash, 30 wt% - 35 wt% pyrophyllite, 10 wt% - 15 wt%

bauxite and 4 wt% AlF3.

Microcrystal glass was a sort of borosilicate glass and

the composition is sho wn in Table 1.

High purity silica, reagent grade boric acid, zinc oxide,

sodium carbonate and yttrium oxide were used as source

materials and mixed in the above ratios, ball milled and

dri ed. Then, the mixture was gro und in a p latinum cr uci-

ble and kept it at 1500˚C for 3 h. The molten glass trans-

formed from the mixture at high temperature and under-

went water quenching and a course of drying and ball

milling to produce a glass power with an average size

about 1 - 3 μm. These glass powder was distributed un-

iformly by distributor on the surface of mullite ceramic

tile and its thickness was kept at 1.2 mm. Next, the cov-

ered tile was placed in a furnace for a second sintering at

1000˚C - 1200˚C, causing the glass powder to remelt,

nucleate, crystallize and combinesolidly with the ceramic

base. After cooling down to the room temperature, the

large-size ultra-thin mullite glass ceramic tile was pre-

pared finally.

3. Materials, Equipment, and Experimental

Procedure

3.1. Material

The materials used in this investigation are float glass

plates, and Polyvinyl Butyral (PVB) and Ethylene Vinyl

Acetate (EVA) as interlayer materials. The maximum

force capacity and the amount of the absorbed energy of

the laminated glass were determined for the input va-

riables that are summarized in Tables 1 -4 below. Figure

1 shows the schematic diagram for the assembly of the

glass plates and interlayer.

3.2. Equipment

Equipment used in this investigation are Glass cutting

machine of BSJ-NL3725 type, Bend testing machine of

OUTOGRAPH AG—1S type, and Charpy testing ma-

chine.

3.3. Experim enta l Pro ced ure

Testing procedure can be summarized as follows:

1) Cutting plates of 40 cm x 30 cm from glass panels

of 4 mm, 6 mm , 8 mm, 10 mm, 12 mm thicknesses. The

sharp cut edges have been broken off or beveled with a

grind ing tool ;

2) Manufacturing of PVB-laminated glass. It compris-

es the washing and drying of individual glass sheets,

laying the PVB film between the two glass sheets by

usin g roller process, and heati ng and pres sing the a ssem-

bly.

An assembly full- surface bond is created in an autoc-

lave using temperatures of about 140°C and pressure of

about 150 psi. The interlayer becomes a viscous at this

temperature and pressure, and any remaining air dis-

solves into the la minate layer;

3) M anufactur ing of EV A laminated glass. It compris-

es the washing and drying of individual glass sheets,

laying the EVA film between the two glass sheets by

using roller process, and the assembly is headed in single

stage lamination process (vacuum with integrated heating

and cooling in the same apparatus);

4) Cutting of the manufactured laminated glass to the

required size by using the cutting machine. For point

bend test, the rectangular sheets dimension is 80 mm x

300 mm while for Charpy test, the rectangular sheets di-

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Interlayer on the Mechanical

Behavio r of Laminated Glass

Table 1. PVB samples for be nding and C ha rpy impact tests (the outer plates and interlaye r thickne ss c hang eable) .

One inter l ayer Four int erlayers Si x int erlayers

Inner plate (mm) Outer plat e (mm) Inner plate (mm) Outer plat e ( mm) Inner plate (mm) Outer plate (mm)

4 4 4 4 4 4

4 6 4 6 4 6

4 8 4 8 4 8

4 10 4 10 4 10

4 12 4 12 4 12

Table 2. PVB sampl es for bending a nd C ha rpy impact tests (the inner plates and interlaye r thickne ss c hang eable) .

One inter l ayer

Four interlayers

Six int erlayer s

Inner plate (mm)

Outer p late (mm)

Inner plate (mm)

Outer p late (mm)

Inner plate (mm)

8 4 8 4 8 4

Table 3. EVA samples for bending and C ha rpy impact tests (the out er plates and interlayer thicknes s changeable).

One inter l ayer One int erlayer One interla y er

Inner plate (mm) Inner plate (mm) Inner plate (mm) Inner plate (mm) Inner plate (mm) Inner plate (mm)

Table 4. EVA samples for bendi ng an d Charpy impact tests (the inner plates and interlayer thickness changeable).

One inter l ayer

Inner plate (mm)

10 4 10 4 10 4

mension is 80 mm x 300 mm.

4. Results and Discussion

As stated before, the maximum force capacity and the

amount of the absorbed energy of the laminated glass

were determined for the input variables that are summa-

rized in Tables 1-4 for the assembly shown in Figure1.

The outer surface is the one in contact with the force

while the inner surface is that locates on the other side

from the force. Results and discussions of the investiga-

tion will be br ie fed in the following sections.

4.1. Load Capacity (Force) and Absorbed

Energy

It is clear from Figure 2 that the higher the thickness

(number) of interlayer, the less the maximum load capac-

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Int er layer on the Mechanical

Behavio r of Laminated Glass

Figure 1. Schematic diagram for the assembly of the glass plates and interlayer. Outer glass is the one in contact with the

applied force.

Glass thickness [ mm]

46810 12

Force [N]

400

800

1200

1600

2000

2 INTE RL AYERS

6 INTE RL AYERS

4INTE RL AYERS

Figure 2. Testing the maximum force on (PVB) laminated

glass where the thickness of inner plate was fixed and the

outer plate was fixed and the outer plate and interlayer

were change able.

Glass thickness[mm]

46810 12

Force [N]

400

800

1200

1600

2000

2 INTERLAYERS

6 INERLAYERS

4 INTE RL AYERS

Figure 3. Testing the maximum force on (PVB) laminated

glass where the thickness of outer plate was fixed and the

inner plate was fixed and the inner plate and interlayer

were changeable.

Glass thi c kn e ss [mm]

46810 12

Force [N]

500

1000

1500

2000

2500

2 INTERLAYERS

4 INTERLAYERS

6 INTERLAYERS

Figure 4. Testing the maximum force on (EVA) laminated

glass where the thickness of inner plate was fixed and the

oute r plate an d interlayer were changeable.

Glass thi c kn ess [mm]

46810 12

Force [N]

500

1000

1500

2000

2500

2 INTERLAYERS

4 INTERLAYERS

6 INTERLAYERS

Figure 5. Testing the maximum force on (EVA) laminated

glass where the thickness of outer plate was fixed and the

oute r plate an d interlayer were changeable.

ity of the laminated glass bonded with PVB material for

the fixed thickness of the inner glass plate. This load ca-

pacity is a characteristic strength from Weibull strength

distribution. The same behavior can be observed for the

laminated glass bonded with the same material although

the fixed thickness is the thickness of the outer glass

plate (Figure 3). The same trends also can be observed

for the laminated glass bonded with EVA (Figures 4 and

5). The trend of these results is in agreement with the

shear modulus results reported by Quentt [1], Hooper [2],

and the predictions of Zang et al. [12]. On the other

hand, they contradict with the results of Minor and Rez-

nik [8 ].

Figure 6 shows that the position of the plate of the

fixed thickness does not affect the maximum load capac-

ity and the maximum load capacity for laminated glasses

bonded with EVA is greater than that for the ones

bonded with PVB provided that the same conditions are

maintained.

The absorbed energy shows an opposite effect. For

example, Figure 7 shows that the higher the thickness

(number) of bonding interla yer, the higher the amount o f

the absorbed energy. Moreover, the laminated glass

which is bonded with PVB absorbs more energy than

those bonded with EVA. The trends in these results are in

agreement wit h the results of K e lle r [6].

An int er es ti n g b ehavio r is s hown i n Figure 2 whe n t he

outer thickness of the outer glass is 6 mm. In this case,

the maximum load capacity for the 4 interlayer is less

Plate Thickness

Interlayer Thickness

Outer (Upp e r ) Glass plate

Inner (Lower) Glass plate

Plate Thickness

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Interlayer on the Mechanical

Behavio r of Laminated Glass

than that for the laminated glass bonded with 6 interlay-

Glass thickness[ mm]

46810 12

Force [N]

500

1000

1500

2000

2500

PVB, OUTER FI X ED THICKNE SS

PVB, INNER F I X ED THICKNE SS

EVA, INNER FIX ED THICKNE SS

EVA, OUT ER FIXED THICKNESS

Figure 6. Comparison of the ma x i mu m load c apacit y for the

2 or fixed interlayer thickness, variable bonding material,

and different positio ns of gl ass thicknesses.

Glass thickness [mm]

46810 12

Absorbed energy until fracture [J]

6 INTERLAYERS, PVB

4 INTERLAYERS, PVB

2 INTERLAYERS, PVB

2 INTERLAYERS, EVA

4 INTERLAYERS, EVA

6 INTERLAYERS, EVA

Figure 7. Abso rbed energ y unti l fracture by Char py impact

test when the inner thickness is variable and the bonding

material is PVB and EVA.

ers. Furthermore, the amount of absorbed energy the la-

minated glass of 4 mm thickness and 6 bonding inter

layer of EVA is greater than that for 4 interlayers boded

with PVB for the same thickness. These interactions

worth more investigations in the future.

4.2. Modeling of the Maximum Load Capacity

(Force) and the Absorbed Energy

The maximum load capacity of glass and its absorbed

energy are very important in real life applications. For

example, high rise buildings or some open areas are ex-

posed to a high impact wind forces. To be able to find the

suitable glass to resist the forces and help in absorbing

higher energy, it is of a great importance to select the

suitable glass. As it was noticed before, there is a contra-

dict ion in the result s whe n compar ing the maxi mum loa d

capacity and the amount of absorbed energy. To over-

come this, the modeling took place for the maximum

load capacity and the amount of absorbed energy sepa-

rately depending on the thickness of glass and the thick-

ness of the bonding interlayer regardless the position of

glass plates. The modeling of the interaction of the

maximum load capacity and the amount of absorbed

energ y will be consid ere d in our future invest igation.

The modeling tool used in this investigation was mul-

tiple regressions with the help of minitab software. Four

relationships were determined because the measured re-

sults of failure strength and absorbed energy till failure

upon impact is different due t o the visco -elastic damping

of interlayer. These are:

1) The maximum load capacity as a dependent varia-

ble and thickness of glass and the thickness of the PVB

bonding interlayer as independent variables.

2) The amount of absorbed energy as a dependent va-

riable and thickness of glass and the thickness of the

PVB bonding interlayer as independ ent variab les.

3) The maximum load capacity as a dependent varia-

ble and thickness of glass and the thickness of the EVA

bonding interlayer as independent variables.

4) The amount of absorbed energy as a dependent va-

riable and thickness of glass and the thickness of the

EVA bonding inter layer a s independe nt variables.

The multiple linear regression assumes that the varia-

ble response is a linear function of the model parameters

and there are more than one independent variable in the

model.

The general form of the developed model may be

written:

y xx

αβ γ

=++

(1)

where

y: is dependent variable (Max bending force or Max

absorbed energy);

α, β, γ: are regression coefficients;

x1, x2: are the thickness of glass and the interlayer gl ass

thicknes s es resp ectively.

After running the minitab software, the results can be

summarized as follows:

1) The equation that relates the maximum load capac-

it y ( y) as a dependent variable and thic kness of glas s (x1)

and the thickness of the PVB bonding interlayer (x2) as

independent variables is:

Maximum load capacity(PVC)34817458.3xx=−+ −

(2)

The observations, which were described by this rela-

tionship, are independent random variable as can be seen

on Figure 8(a) as this figure presents the normal percent

prob ability of the resid uals and the p lot points lie a long a

straight line. So, the h ypothesi zed distrib ution ade qua tely

describe data and the model is appropriate. Furthermore,

the model explains about 91.5% of the variability of the

process because the adjusted R-sq = 91.5%. The analysis

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Int er layer on the Mechanical

Behavio r of Laminated Glass

of variance of the process shows that the results are ex-

tre mely si gni fic ant a s t he P -value is about zero. T he ben-

efits of this equation can be seen clearly when applied to

real life cases. To find the suitable laminated glass with

dependent variables x1(thickness of glass) and x2 (the

thickness of the PVB bonding interlayer) that can resist

the external force (wind force as an example), the varia-

ble x1 can be changed as it is the onl y variable that has a

positive sig n.

2) The equation that relates the amount of absorbed

energy as a dependent variable and thickness of glass (x1)

and the thickness of the PVB bonding interlayer (x2) as

independent variables is:

Amountofabsorbedenergy(PVB)=17.45.121.74

−+ +

(3)

Figure 8 (b) presents the normal per cent probability of

the residuals and shows that the observations are inde-

pendent rando m variable and follow the normal distribu-

tion. Moreover, the model explains about 90.3% of the

Residual

Percent

4003002001000-100-200-300-400

Normal Probability Plot of the Residuals

(response is MAX PVB1)

(a)

Residual

Percent

86420-2-4-6-8

Normal Probability Plot of the Residuals

(response is MAX PVB 2)

(b)

Figure 8. Normal probability plot of residuals of (a) the

maximum load capacity relationship and (b) amount of

absorbed energy for PVB bonding material.

Residual

Percent

3002001000-100-200-300

Normal Probability Plot of the Residuals

(response is MAX EVA 1)

(a)

Residual

Percent

210-1-2

Normal Probability Plot of the Residuals

(response is MAX EVA 2)

(b)

Figure 9. Normal probability plot of residuals of (a) the

maximum load capacity relationship and (b) amount of

absorbed energy for EVA bonding material.

variability of the process because the adjusted R-sq =

90.3%. The analysis of variance of the process shows

that the results are extremely significant as the P-value i s

about zero. To find the suitable laminated glass with de-

pendent variables x1 (thickness of glass) and x2 (the

thic kness of t he PVB b onding interla yer) t hat can a bsorb

the highes t amount of energy unt il fra cture, t he variab les

x1and x2 can be changed.

3) The equation that relates the maximum load capac-

ity as a dependent variable and thickness of glass (x1) and

the thickness of the EVA bonding interlayer (x2) as in-

dependent variables is:

Maximum load capacity (EVA)8818568.3

=−+ −

(4)

Figure 9(a) presents the normal percent probability of

the residuals and shoes that the observations are drawn

from independent variables and the standard deviation

and the variance of both populations are equal as the plot

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Interlayer on the Mechanical

Behavio r of Laminated Glass

points shows that the data follows a normal distribution. Also the model explains about 94.7% of the variability of

Figure 1 0. Failure observed aft er bendi ng test (side view).

Figure 11. Failure obser ved after bending test (top view).

Figure 1 2. Failure after Charpy t es t .

the process because the adjusted R-sq = 94.7%. The

analysis of variance of the process shows that the results

are extremely significant as the P-value is about zero.

4) The equation that relates the amount of absorbed

energy as a dependent variable and thickness of glass (x1)

and the thickness of the EVA bonding interlayer (x2) as

independent variables is:

Amountofabsorbedenergy(EVA)6.712.740.620xx=−+ +

(5)

Figure 9(b) presents the normal percent probability of

the residuals. The plot points show that the process data

followed a normal distribution and the observations are

independent random variable. Moreover, the model ex-

plains about 95.7% of the variability of the process be-

cause the adjusted R-sq = 95.7%. The analysis of vari-

ance of the process shows that the results are extremely

significant as the P-value is about zero.

4.3. Failure Observation

Bending test took place until fracture. Then the fractured

surface was analyzed. It was found that the propagation

of fracture was linear within the glass plate and non li-

near within the bondi ng po lymer a s seen in t he side view

(Figure 10). This difference may be due to the thermop-

lastic nature of the bonding material which was described

by Hooper [2]. The top view in Figure 11 shows the li-

near nature of propagation within the brittle glass AND

Figure 12 shows the failure after Charpy test.

5. Conclusions

Linear

Non-linear

Outer gla ss plate

Inner glass plate

Binding material

The Effect of Glass Plate Thickness and Type and thickness of the Bonding Int er layer on the Mechanical

Behavio r of Laminated Glass

The conclusions that can be drawn from this investiga-

tion are:

1) The higher the thickness of interlayer, the less the

maximum load capacity of the laminated glass bonded

whether wit h PVB or E V A bo nd ing material for the fixed

thicknes s of t he inner glas s plate

2) The position of the plate of the fixed thickness does

not affect the maximum load capacity and the maximum

load capacity for laminated glasses bonded with EVA is

grea ter than t hat for the o nes bo nded with P VB pro vided

that the same conditions are maintained

3) The higher the thickness of bonding interlayer, the

higher the amount of the absorbed energy whether the

laminated glass bonded with PVB or EVA bonding ma-

terial. Moreover, the laminated glass which is bonded

with PVB absorbs more energy than those bonded with

EVA

4) Regression models were developed to calculate the

maximum load capacity and the amount of absorbed

energy separately depending on the thickness of glass

and the thick ness o f the bo nding i nterl ayer regard less the

position of glass plates. Positive variables are taken into

consi deration during calcula t io ns .

5) The propagation of fracture was linear within the

glas s plate and non linear within the bonding polymer

REFERENCES

[1] R. Quenett, “The Mechanical Behavior of Laminated

Safety Glass under Bending and Impact Stresses,” Ma-

nuskript-Eing, Forgetragen auf dem DVM-Tag, Wurz-

burg, 1967.

[2] J. A. Hooper, “On the Bending of Architectural Lami-

nated Glass,” International Journal of Mechanical

Science, Vol. 15, No. 4, 1973, pp. 309-333.

doi:10.1016/0020-7403(73)90012-X

[3] P ilkington ACI, “A Practical and Theoretical Investiga-

tion into the Strength of Laminated Glasses under Un-

iformly Distributed Loading,” Laboratory Report and

Discussion, Pilkington ACI Operations Pty. Ltd., 1971, p.

206.

[4] M. P. Linden, J. E. Minor, R. A. Behr and C. V. C. Val-

labh an, “Evaluation of laterally Loaded Laminated Glass

Units by Theory and Experiment,” Supplemental Report

No. 1, Glass Research and Testing Laboratory, Texas

Tech University, Lubbock, 1983.

[5] H. S. Norville, “Breakage Tests of Du Pont Laminated

Glass Units,” Glass Research and Testing Laboratory,

Texas Tech University, Lubbock, 1990.

[6] K. Uwe, “Measuring the delaminating Energy in Lami-

nated Safety Glass,” Proceedings of Glass Processing,

Finland, 17-20 June 2005 , pp. 102-104.

[7] S. R. Nagalla, C. V. G. Va llab han, J. E. Minor and H. S.

Norville, “Stresses in Layered Units and Monolithic

Glass Plates,” NTIS Accession No. PB86-142015/AS,

Glass Research and Testing Laboratory, Texas Tech

University, Lubbock, 1985.

[8] J. E. Minor and P. L. Reznik, “Failure Strength of Lami-

nated Glass, ” Journal of Structural Engineering ASCE,

Vol. 116, No. 4, 1990, pp. 1030-1039.

doi:10.1061/(ASCE)0733-9445(1990)116:4(1030)

[9] M. P. Linden, J. E. Minor and C. V. C. Vallabhan,

“Evaluation of Laterally Loaded Laminated Glass Units

by Theory and Experiment,” Glass Research and Testing

Laboratory, Texas Tech University, Lubbock, 1984.

[10] B. Weller, “Experimental Study on Different Interlayer

Materials for Laminated Glass,” Glass Processin g Days,

Finland, 2005, pp. 386-394.

[11] R. A. Behr and P. A. Kremer, “Dynamic St rains in Arc-

hitectural Laminated Glass Subjected to Low Velocity

Impacts From Small Projectiles,” Journal of Materials

Scien ce, Vol. 34, No. 23, 1999, pp . 5749-5756.

doi:10.1023/A:1004702100357

[12] M. Y. Zang, Z. Lei and S. F. Wang, “Investigation of

Impact Fracture Behavior of Automobile Laminated

Glass by 3D Discret e Element Meth od,” Sprin ger Verl ag,

Berlin, 2007.

[13] J. Belis, J. Depauw, D. Call ewa er, D. Deli ncé and R. Van

Impe, “Failure Mechanisms and Residual Capacity of

Annealed Glass/SGP Laminated Beams at Room Tem-

perature,” Journal of Materials Science, Vol. 16, No. 6,

2008, pp. 1866-1875.