Glass Struct. Eng.

(2016) 1:277–288
DOI 10.1007/s40940-016-0025-2

CHALLENGING GLASS PAPER

Investigations on the mechanical strength of enamelled glass

Iris Maniatis · Michael Elstner

Received: 12 January 2016 / Accepted: 24 April 2016 / Published online: 13 May 2016
© Springer International Publishing Switzerland 2016

Abstract Enamelled glass has been used for many the influence of different parameters on the mechanical
years in architectural applications. It is well known that strength. Furthermore the results will be compared with
ceramic frits have a weakening effect on the mechan- the current values of the minimum bending strength
ical strength of thermally toughened glass. This fact given in the product standards.
is addressed in the relevant European product stan-
dards for thermally toughened (with or without heat Keywords Enamelled glass · Bending strength ·
soak test) and heat strengthened glass. In EN 12150- Colour · Manufacturing process
1 the minimum bending strength of thermally tough-
ened enamelled glass is given with σbB = 75N/mm2 1 General
and in EN 1863 for heat strengthened glass the value
is given with σbB = 45N/mm2 . The values are valid 1.1 Definitions
independently of the colour tint and the manufactur-
ing process of the enamel coating. As the experimental An enamel is a glass-like mass which is produced by
data for the current values were determined decades total or partial fusing principally consisting of oxides. It
ago and the background of these investigations is not is applied in one or several layers together with various
available anymore, a recent research project examined supplements to metallic or glass surfaces at temper-
influences of glass type, colour tint and the manufac- atures over 450 ◦ C (Krampe 2013, Technische Regel
turing process of the enamel by performing four point RAL-RG 529 A3 2007). The term “enamel” is used for
bending tests according to EN 1288-3. To compare the the final printed surface. The term “enamelling” is used
results with the non enamelled glass types the glass for the manufacturing process.
samples were taken from a single float glass batch. The
test results show the strength is mainly dependent on
the colour scheme, the thickness of the enamel and the 1.2 Colour scheme
temperature sequence of the fusing process itself. In the
paper some of the results will be presented especially Essentially the dye powder consists of the following
main components: flow system (70–95 %) and pig-
I. Maniatis (B)
ment system (5–30 %). After firing the transparent
Universität der Bundeswehr München, Neubiberg,
Germany and soft glass-flow surrounds the colour pigments con-
e-mail: [email protected] tained in the dye and binds these together permanently.
M. Elstner The adhesion to the glass results from ionic bondings
AGC Interpane, Plattling, Germany of the SiO groups (Krampe 2013). Table 1 lists the most

123
278 I. Maniatis, M. Elstner

Table 1 Components of the

Black Red White
colours black, red and white
as per manufacturer’s Flow system ZnO−B2 O3 −SiO2 ZnO−B2 O3 −SiO2 ZnO−B2 O3 −SiO2
specifications
Pigment system CuO−Cr2 O3 Cd–S–Se TiO2
Particle size (μm) 9 7 9
Density (kg/dm3 ) 3.0 2.9 3.2
Technical Information Ferro Linear thermal expansion 93 × 10−7 93 × 10−7 93 × 10−7
2012 coefficient (1/K)

Table 2 Bending tensile

Glass type Minimum values for bending tensile strength (N/mm2 )
strength for FT and HS
glass FT glass according HS glass according
to EN 12150-1 to EN 1863-1

Float: clear, tinted, coated 120 70

Enamelled float glass 75 45
(enamelled surface under
tensile stress)
Reduction in strength (%) 37.5 35.0

important components for the colours examined in the The experimental data for the current values were
research project (black, red and white). determined decades ago and the scientific background
of these investigations is unknown. Up to now there
have been only a few systematic investigations exam-
ining the reasons for the reduction in strength (see
1.3 Strength Krampe 2013). In (Krampe 2013) following feasible
causes for the stress reduction are given:
From an architectural and manufacturer’s perspective • Differentials in thermal expansion coefficient
important considerations for enamelled glass are: aes- • Fineness of grind
thetics, function (i.e. use as solar-control glass with an • Firing penetration and porosity of the enamel
additional functional coating added to the enamelled • Leaching processes within the boundary layer of
surface), manufacturing process, visual assessment and the glass to the enamel
application. • Changes in thermal conditions during the firing
However for the proper dimensioning of the glaz- process
ing, the characteristic strength values of the glass types In (Krampe 2013) the thermal conditions during the
must be known. The values given in EN 12150-1 (EN thermal toughening process are examined in detail.
12150-1:2000-11) and EN 1863-1 (EN 1863-1:2012-
02) for enamelled thermally toughened safety glass
(fully toughened glass—FT) with or without heat soak 2 Manufacturing
test (HST) and heat strengthened glass (HS) have been
used for many years (see Table 2). The reduced val- 2.1 Colour application process
ues compared with those of non enamelled glass show
that such ceramic frits have a weakening effect on the There exist different processes for applying the colour
mechanical strength of thermally toughened glass. The on the glass surface. These are, essentially: roller coat-
values are independent of the colour tint and the man- ing, screen printing, digital printing, and “curtain” coat-
ufacturing process of the enamel coating. ing. Usually the enamel coating is applied on the “air

123
Investigations on strength of enamelled glass 279

surface” of the float glass (that is to say, on the opposite Using the screen printing process the glass edges
side to the tin bath surface). During the thermal tough- generally remain paint-free but can display a slight
ening process of thermally toughened safety glass or bulging or beading of colour in the border area. If appli-
heat strengthened glass the ceramic coating is fused cations with visible edges are planned the manufac-
and firmly bound together with the glass. In Sect. 2.2 turer must be informed beforehand. With this process
these processes are described according to (BF-Bulletin it is possible to carry out multi-coloured printings: for
015 2014). example, a so-called “double screen printing” whereby
two different colours are recognizable on each pane-
2.1.1 Roller coating surface looked at. The manufacturer should always be
consulted on issues of tolerances (e.g. of congruency).
The flat glass plane is passed beneath a ribbed rubber The printing of selected patterned glass panes is also
roller which transfers the enamel paint onto the surface possible but, once again, the manufacturer should be
of the glass. This ensures a regular, even and homoge- consulted about it beforehand.
neous distribution of the paint across the entire surface
of the pane. Typically, the ribbed pattern of the roller 2.1.3 Digital printing
is visible if the pane is examined from up close (on
the enamelled side). Normally, however, these “ribs” The ceramic colour is applied directly onto the glass
are barely visible when the glass is observed from the surface using a process which is comparable to an
front side (looking through the pane). Furthermore, inkjet printer. The thickness of colour application can
a so-called “starry sky” effect (numerous very small vary and is generally thinner than using roller coat-
imperfections) can sometimes emerge in the enamel. ing, screen printing or “curtain” coating processes. It
Due to the nature of this application process a cer- appears, depending on the colour chosen, either opaque
tain “colour overrun” is possible on all edges of the or translucent. A high print resolution of up to 720
pane, this overrun can be slightly corrugated especially dpi is possible at present. Typical for this manufactur-
on longitudinal edges (viewed in the direction of the ing process are minor visible stripes in the direction
movement of the roller unit). The edge surface, how- of printing which are unavoidable due to the produc-
ever, remains, as a rule, paint-free. For this reason, the tion technique. Using the digital printing process the
manufacturer should be consulted beforehand about the glass edges generally remain paint-free but can display
position of installation. Usually roller printed enam- a slight bulging or beading of colour in the border area.
elled glasses are not suitable for vision glass applica- If applications with visible edges are planned the man-
tions, if so these must always be discussed with the ufacturer must be informed beforehand. The printing
manufacturer beforehand. edges are, in the direction of printing, exactly straight
but are, transversal to the print direction, slightly ser-
2.1.2 Screen printing rated. A certain paint spray may also occur along the
printing edges. In the case of dot-, hole- and text-
In contrast to the printing process just described, within motifs the printing edges may display a serration which,
this process the paint can either be applied across the like the mentioned spray, is only visible if the pane is
entire surface or, alternatively, in defined areas. The examined from up close. Above all the digital print-
colour is applied to the glass surface on a horizon- ing process is suitable for complex multi-coloured grid
tal screen printing frame through a narrow-meshed designs or for images. It is less suitable for mono-
screen using a flood bar. The thickness of paint applied chrome printing across the entire surface of a glass
depends on the breadth of the screen’s meshes and on pane.
the thread diameter. However the paint application is
generally thinner than using roller or “curtain” coating 2.1.4 “Curtain” coating
and appears, depending on the colour chosen, either
opaque or translucent. Typical for this manufacturing The glass pane passes horizontally a so-called “curtain”
process are minor stripes both in the direction of print- and its entire surface becomes covered with colour. By
ing and also transversal to this direction, as well as adjusting the quantity of paint poured out and the speed
certain occasionally occurring spots of haze. with which the pane passes through the “curtain”, the

123
280 I. Maniatis, M. Elstner

thickness of the coating can be controlled within a rela- ing which products need to be applied in which per-
tively wide range. In case of a slight uneven pouring lip centages.
stripes of various degrees of thickness in longitudinal
direction (direction of pouring) can be caused. Vision
glass applications must always be discussed with the 2.2 Fusing process
manufacturer beforehand. The “colour overrun” at the
edges is significantly greater than in the case of roller At a furnace temperature of approximately 620–720 ◦ C
coating and can only be avoided through great effort. If (corresponding to glass temperatures of approximately
paint-free visible edges are wanted, this must be spec- 620–660 ◦ C) the ceramic colour is fused into the glass.
ified when placing the order. The total fusing cycle required amounts to approxi-
mately 40 s per millimetre of glass thickness. For proper
fusing an oxidizing atmosphere in the furnace must be
2.1.5 Summary maintained.
Should the enamel prove to be porous after the fusing
Depending on the chosen application process the process, for example due to unfavourable fusing condi-
colours need to have different viscosities. For roller tions, it can lead to a negative influence on the weather
coating, for example, more liquid colours are required resistance, the chemical stability and the mechanical
(i.e. with a higher organic part). Correspondingly each strength of the enamelled glass. In (Technical Informa-
different application procedure will require a differ- tion Ferro 2012) several possible ways of avoiding this
ent drying time. After the application of the colour the are listed:
printed pane has to be dried (at a temperature of approx-
• Improving the drying of the enamel coating before
imately 170–190 ◦ C). Then the pane is thermally tough-
fusing to optimise the burning of the organic
ened and at the same time the enamel coating is fused
medium
onto the surface of the glass.
• Increasing the quantity of energy available for the
As already mentioned depending on the chosen man-
fusing of the colour into the glass by optimising the
ufacturing process different thicknesses of the enamel
fusing conditions (raising of the temperature) or
coating will result (see Fig. 1).
• Lowering the fusing temperature of the glass
The colour systems currently available are capable
enamel by admixture of some transparent flow-
to represent a large number of colours, like for exam-
system
ple those of RAL and NCS chart of colour range. In
order to achieve a specific colour the manufacturers of The enamel produced in this way will be scratch-proof
such colour systems provide recommendations regard- as well as weather, UV and solvent resistant.

3 Quality control

To check the quality of the enamel the following, among

others, test methods are recommended according to
(Technical Information Ferro 2012):
• Gloss value as according to EN ISO 2813
• Surface scratch resistance as according to EN
15771 or ISO 1518
• Un-dried (wet) thickness of the enamel coating or
the degree of porosity of the fused ceramic coating
The evaluation of the visual quality of glass panes
enamelled across the whole or part of their surface can
Fig. 1 Different enamel application processes with average be carried out in accordance with the method described
coating thicknesses (after fusing process) (Krampe 2013) in (BF-Bulletin 015 2014).

123
Investigations on strength of enamelled glass 281

4 Current research In the following sections the results obtained from

the individual series are presented in stress-deformation
4.1 General diagrams. These are divided up according to applica-
tion processes and glass-types. Statistical evaluation
Based on the scientific literature explaining the reduc- was performed using standardised normal distribution.
tion in mechanical strength of enamelled glass in the Evaluation using Weibull distribution produced some-
current research project (Siebert and Maniatis 2015) what lower results and required more spot checks. The
further investigation has been performed. Anderson–Darling test was used in order to determine
Samples of different glass types (HS, FT, FT with which of the two statistical evaluations is more suit-
HST), different colours (red, black and white) and dif- able. Both statistical methods are more or less equally
ferent application processes (screen printing and roller suitable.
coating) were subjected to four point bending tests In addition the crack origins of some samples were
according to EN 1288-3 (EN 1288-3: 2000-09). The examined microscopically in order to analyse the frac-
average coating thickness for screen printed glass was ture mirror.
45 μm and for roller coated glass 75 μm. All speci-
men were sprinkled with corundum before testing to
reduce the dispersion of failure stresses. For evaluation
only specimen were used having their fracture origin in 4.2 Four point bending test results
the area between the bending rollers of the four point
bending test-setup. 4.2.1 Heat strengthened (HS) glass series
The results were compared with un-enamelled panes
of respectively each of the glass-types used. In order to In Figs. 2 and 3 the failure stress values are given for
make possible a comparison between the glass-types each manufacturing process. Also shown are the fail-
and between enamelled and un-enamelled glass, float ure stress values for the un-enamelled reference sam-
glass of a single batch was used. For each series 10 ples. The statistical evaluation of all series is listed in
samples with a nominal glass thickness of 8 mm were Table 3.
tested.

100

95

90

85

80
[N/mm²]

75

HS screen_black
70
HS screen_white
bB

65 HS screen_red
HS_Ref
60

55

50

45

40
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

wmax [mm]

Fig. 2 Results for series ‘HS glass with screen printing on side exposed to tensile bending stress’. (Color figure online)

123
282 I. Maniatis, M. Elstner

100

95

90

85

80
[N/mm²]

75

HS roller_black
70
HS roller_white
bB

HS roller_red
65
HS_Ref
60

55

50

45

40
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
wmax [mm]

Fig. 3 Results for series ‘HS glass with roller coating on side exposed to tensile bending stress’. (Color figure online)

Table 3 Statistical evaluation of the HS glass series

Manufacturing process Location of enamel Black Red White

Screen printing Surface under Normal distribution (confidence level 54.99 69.38 47.27
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 61.63 81.29 59.10
Standard deviation 2.31 4.14 4.12
Roller coating Surface under Normal distribution (confidence level 38.39 58.36 55.32
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 46.66 66.43 59.95
Standard deviation 2.88 2.81 1.38

The results can be summed up as follows: 4.2.2 Thermally toughened (fully toughened—FT)
glass series
• The reference samples show a high dispersion of
failure stresses despite the sprinkling with corun- In Figs. 4 and 5 the failure stress values are given for
dum. each manufacturing process. Also shown are the fail-
• For the reference samples the mean value of the ure stress values for the un-enamelled reference sam-
tensile bending strength is 70.4 MPa and the frac- ples. The statistical evaluation of all series is listed in
tile value 51.9 MPa (confidence level 95 %, fractile Table 4.
5 %). The results can be summed up as follows:
• Enamel on the side exposed to tensile bending • For the reference samples the mean value of the ten-
stress: Except for the “roller coating, black” series, sile bending strength is 145.7 MPa and the fractile
the minimum bending strength of 45 MPa was value 127.2 MPa (confidence level 95 %, fractile
reached for all other colours and manufacturing 5 %). The minimum bending strength of 120 MPa
processes. for FT was reached.

123
Investigations on strength of enamelled glass 283

155

150

145

140

135

130
σbB [N/mm2]

125
FT screen_black

120 FT screen_white
FT screen_red
115 FT_Ref

110

105

100

95

90
30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
wmax [mm]

Fig. 4 Results for series ‘FT glass with screen printing on side exposed to tensile bending stress’. (Color figure online)

155

150

145

140

135

130
σbB [N/mm2]

125
FT roller_black

120 FT roller_white
FT roller_red
115 FT_Ref

110

105

100

95

90
30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
wmax [mm]

Fig. 5 Results for series ‘FT glass with roller coating on side exposed to tensile bending stress’. (Color figure online)

123
284 I. Maniatis, M. Elstner

Table 4 Statistical evaluation of the FT glass series

Manufacturing process Location of enamel Black Red White

Screen printing Surface under Normal distribution (confidence level 105.02 119.34 95.63
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 110.89 133.05 109.34
Standard deviation 2.04 4.77 4.77
Roller coating Surface under Normal distribution (confidence level 85.28 103.36 99.24
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 98.46 113.96 110.08
Standard deviation 4.59 3.69 3.77

• Enamel on the side exposed to tensile bending (duration, temperature) which can have an influence
stress: For all colours and manufacturing processes on the bending strengths. In order to determine the
a minimum bending strength higher than 75 MPa influence of the individual parameters, the colour com-
was reached. positions and the fusing parameters would need to be
known. On this basis the thermal processes could be
4.2.3 Thermally toughened and heat soaked numerically simulated and optimised.
(FT-HST) glass series The samples with thinner enamel coating (screen
printing, approximately 45 μm coating thickness) tend
In Figs. 6 and 7 the failure stress values are given for to have greater strengths than thicker coatings (roller
each manufacturing process. Also shown are the failure coating, approximately 75 μm coating thickness). To
stress values for the un-enamelled reference samples. give a more precise statement, coatings of greater thick-
The statistical evaluation of all series is listed in Table 5. ness would need to be tested.
The results can be summed up as follows: The tensile bending strengths ascertained for enam-
elled glass panes are, in some cases, significantly
• For the reference samples the mean value of the ten- greater than the minimum bending strengths specified
sile bending strength is 134.4 MPa and the fractile in the product standards (enamel on the side exposed to
value 126.5 MPa (confidence level 95 %, fractile tensile bending stress), especially the enamel coatings
5 %). The minimum bending strength of 120 MPa produced by screen printing.
for FT was reached. However to use these greater bending strengths as
• Enamel on the side exposed to tensile bending characteristic strength values for dimensioning of the
stress: For all colours and manufacturing processes glazing more precise parameters would need to be
a minimum bending strength higher than 75 MPa established regarding, for example, manufacturing and
was reached. fusing processes, colour tint and colour composition
and coating thickness. The reproducibility of these
4.2.4 Summary of results parameters would also have to be proven, e.g. within
the framework of a National or European Technical
The colours investigated in the present research project Approval.
(black, red, white) and the application processes (screen
printing, rolling coating) lead to different tensile bend-
ing strengths. A result which is to be traced back essen- 4.3 Fracture mirror analysis
tially to the different colour compositions of the enam-
els (different size of the colour grains, viscosity of the 4.3.1 General
dyes, percentage of flow system, which vary depending
on the manufacturing process) as well as the different The structure and appearance of the fracture surfaces
thicknesses of their coatings. These factors lead to dif- allow for statements on the nature and size of the frac-
ferent parameters in the fusing or tempering process ture as well as on the resulting failure stresses. For this

123
Investigations on strength of enamelled glass 285

155

150

145

140

135

130

125
σbB [N/mm2]

120 FT-HST screen_black

115 FT-HST screen_red

FT-HST screen_white
110
FT- HST_Ref
105

100

95

90

85

80
28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
wmax [mm]

Fig. 6 Results for series ‘FT-HST glass with screen printing on side exposed to tensile bending stress’. (Color figure online)

155

150

145

140

135

130

125
σbB [N/mm2]

120
FT-HST_roller_black
115 FT-HST roller_red
110 FT-HST roller_white
FT- HST_Ref
105

100

95

90

85

80
28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
wmax [mm]

Fig. 7 Results for series ‘FT-HST glass with roller coating on side exposed to tensile bending stress’. (Color figure online)

123
286 I. Maniatis, M. Elstner

Table 5 Statistical evaluation of the FT-HST glass series

Manufacturing process Location of enamel Black Red White

Screen printing Surface under Normal distribution (confidence level 100.33 113.59 98.08
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 105.33 121.58 104.94
Standard deviation 1.74 2.78 2.38
Roller coating Surface under Normal distribution (confidence level 79.06 81.23 88.11
tension 95 %, 5 % fractile) (MPa)
Mean value (MPa) 93.01 94.92 98.01
Standard deviation 4.85 4.76 3.44

Fig. 8 Typical appearance

of a fracture mirror crack origin

mirror

mist

hackle

purpose a closer look at the structure of the fracture

surface is required and light microscopic studies were
performed.
At the starting point of a crack a so-called fracture
mirror is created. A typical appearance of such a frac-
ture mirror shows Fig. 8.
If the smooth fracture mirror is clearly developed, it
is a sign of slow crack propagation. Correspondingly
if the crack propagation is fast the shape is less dis-
tinctive. In principle the fracture mirror radius can be
measured and thus the impacting energy that leads to
the failure stress can be calculated. However the mea-
surements may have relatively large variations (Wag-
ner 2012), since the radii cannot be measured very Fig. 9 Roller coated HS glass sample with uneven coating (red).
accurately and are more or less subjective. In addition (Color figure online)
derived relationships between fracture mirror radius
and failure stresses, for example by (Shand 1965), are 4.3.2 Approach
valid only for annealed float glass, as constant stresses
are assumed over the cross section in the area of the A light optical microscope with double magnification
crack. was used to examine the samples. For the investigations

123
Investigations on strength of enamelled glass 287

4.3.3 Image evaluation

• The glass surfaces of the samples located under

the enamel coating were plane and no preliminary
damages could be seen.
• Partly the enamel coating of some roller coated
specimen was uneven (see Fig. 9).
• The stripes pattern of the roller was visible at some
roller coated samples (see Fig. 10).
• Some samples had chipping at the surface of the
crack origin and were therefore not examinable.
• The higher the experimentally determined bending
stresses the more distinctive is the “hackle” area and
Fig. 10 Roller coated HS glass sample with stripe pattern the smaller the fracture mirror radius (see Fig. 11).
• It was not possible to determine failure stresses out
of the images which corresponded with the exper-
imentally determined values.
the destroyed specimen of the four point bending tests
were used. Therefore, the fragments of the specimen
with recognizable fracture surfaces were used. Only 5 Conclusion
fragments of the HS glass series could be used for the
examinations because the fragments of the specimen In (Krampe 2013) and in the present research (Siebert
from FT glass were too small to locate the fracture ori- and Maniatis 2015) it was shown that different factors
gin exactly. In general the fracture origin of all speci- such as the colour composition, the application process
men was in the area between the bending rollers of the or the thickness of the enamel result in different reduc-
four point bending test-setup. tions of the tensile bending strength values compared
The light was induced by different light sources on to non-enamelled glass. But in some cases these tensile
the sample. The adjustment of the light is of great bending strength values are significantly greater than
importance as the image sharpness and the visible the minimum bending strengths specified in the product
points are influenced by it. standards (enamel on the side exposed to tensile bend-

hackle

Sample: HS glass, screen printed, red Sample: HS glass, screen printed, white
Tensile bending stress experimetally determined: 86.7 MPa Tensile bending stress experimetally determined: 58.1 MPa

Fig. 11 Fracture mirror of different samples

123
288 I. Maniatis, M. Elstner

ing stress), especially the enamel coatings produced by References

screen printing. However to use these greater bending
strengths as characteristic values for dimensioning of BF-Bulletin 015/2013 “Guideline for assessing the visual quality
of enamelled glass”
the glazing more precise parameters would need to be EN 1288-3: 2000-09 Glass in Building – Determination of the
established and the reproducibility of these parameters Tensile Strength of Glass – Part 3: Test With Sample Sup-
would also have to be guaranteed. Nevertheless the val- ported on Two Sides (Four-Point Bending Procedure)
ues which have been adopted and applied for decades EN 1863-1: 2012-02 Glass in Building – Heat-Strengthened
Soda-Lime-Silicate Glass – Part 1: Definition and Descrip-
already in the product standards for enamelled fully
tion
toughened glass and enamelled heat soak tested fully EN 12150-1: 2000-11 Glass in Building – Thermally-Toughened
toughened glass lie on the safe side and can still be used Soda-Lime-Silicate Safety Glass – Part 1: Definition and
for design. Description
Krampe, P.: Zur Festigkeit emaillierter Gläser. Dissertation,
The samples of the heat strengthened glass series Technische Universität Dresden (2013)
showed a high dispersion of failure stresses. To get Shand, E.B.: Strength of glass—the Griffith method revised. J
reliable statements the number of samples would need Am Ceram Soc 48, 43–49 (1965)
to be increased. Siebert, G., Maniatis, I.: Untersuchungen zur Biegezugfes-
tigkeiten von emaillierten Gläsern, Forschungsbericht – Nr.
The analysis of the mirror fracture is quite extensive b-01-14-24, Universität der Bundeswehr (2015)
and only possible if the glass fragments are big enough Technical Information “Flat Glass” from the Ferro Corporation,
to identify the mirror fracture. Therefore only the HS 09/2012
glass samples could be examined. If there are any dam- Technische Regel RAL-RG 529 A3:2007-07: Email(le)
und emaillierte Erzeugnisse –Begriffsbestimmungen/Bezei
ages (chipping) in the area of the crack origin the mirror chnungsvorschriften, Deutsches Institut für Gütesicherung
fracture is damaged and not visible anymore. In princi- und Kennzeichnung e.V
ple the fracture mirror radius can be measured and thus Wagner, Ekkehard: Glasschäden, Oberflächenbeschädigungen,
the impacting energy that leads to the failure stress can Glasbrüche in Theorie und Praxis. Fraunhofer IRB Verlag,
Stuttgart (2012)
be calculated. Since the radii could not be measured Weller, B., Reich, S.: Glasbau-Praxis – Band 1: Grundlagen. Kon-
very accurately the measurements had large variations. struktion und Bemessung. 3. überarbeitete und erweiterte
It was not possible to determine failure stresses out of Auflage. Beuth Verlag, Berlin (2013)
the images which corresponded with the experimen-
tally determined values. Basically it could be observed
that the higher the experimentally determined bending
stresses the more distinctive is the “hackle” area and
the smaller the fracture mirror radius. To get reliable
results more samples have to be evaluated.

123