Results

Shear Strength—In the majority of specimens, the adhesive joint failed in the compression shear test. Owing to the use of chemically toughened glass for the tests below room temperature, failure of the glass was rare. The results of

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014

Downloaded/printed by

Rochester Institute Of Technology pursuant to License Agreement. No further reproductions authorized.

the tests are given in Table 3. The block shear figures represent the 5 % fractile for a 95 % confidence level on the basis of a logarithmic normal distribution.

For the acrylic adhesive tested, the shear strength is dependent on the tem­perature. However, the degree of this influence is not the same as that for the tests on the pure material specimens. The mean shear strength values are quite similar between —25 and + 50°C. Only at a temperature of + 75°C does the mean shear strength decrease significantly. The results of the test show considerable scatter at a temperature of —25°C, which indicates embrittlement of the mate­rial. The 5 % fractiles for similar mean values are, therefore, far below the val­ues obtained in tests at 0°C, room temperature, and + 50°C. Therefore, limit value considerations, taking into account the different stiffnesses and strengths, are recommended for the structural calculations.

The shear strength of the joint was seriously reduced by storing the speci­mens in water or detergent. One prime reason for this could be the small size of the specimen’s adhesive joint (5 x 20 mm). The surface area is very large when compared with the area of the adhesive joint, which considerably increases the influence of external media. As this ratio is much more favorable with a typical bonded frame corner, an improvement is to be expected here. Tests on aged full-size specimens were, however, regarded as too involved, meaning that for the applications shown above, direct weathering and the direct ingress of clean­ing agents were ruled out by the detailing at the joint. No negative influence can be deduced from the results following climatic cycle tests. Indeed, the mean shear strength values rise. However, greater scatter of the results leads to the 5 % fractiles lying on a similar level as the values for room temperature.

Load-Bearing Capacity—The adhesive joints remained intact in all ten tests. In most cases the outer plies of glass in the posts failed (Fig. 13(a)). In some tests the specimens did not fail at all, which meant that these tests had to be stopped after reaching the upper load limit of the testing machine. The maxi­mum stress was in all cases detected in the tension zone of the frame post at the transition from four to two plies of glass (strain gauges 1 and 2). Consequently, there is a direct relationship between the failure load of the frame corner and

TABLE 3—Results of block shear tests.

Test conditions

Average value (MPa)

Variation coefficient (%)

5 % fractile (MPa)

T = — 25° C, no aging

12.6

27.3

5.5

T = 0°C, no aging

14.8

9.0

11.2

T = +25°C, no aging

13.2

9.9

9.9

T = +50°C, no aging

13.4

10.2

10.0

T = +75°C, no aging

7.3

8.1

5.6

T = +25°C, cleaning agent

5.1

37.9

1.5

T = +25°C, water immersion

4.6

53.8

1.0

T = +25°C, climatic cycle

19.6

17.0

11.7

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 Downloaded/printed by

Rochester Institute Of Technology pursuant to License Agreement. No further reproductions authorized.

TABLE 4—Results of component testing.

Specimen

component

Setting

block

Maximum

load

Fmax (kN)

Glass

stress

rg. max (MPa)

Strain

gauge

no.

Failure/stop criterion

SC 1

Yes

96.4

165.3

2

Limit of testing machine

SC 2

Yes

87.2

148.0

1

Breakage of glass post

SC 3

Yes

83.7

147.5

2

Breakage of glass post

SC 4

Yes

93.5

158.3

2

Breakage of glass post

SC 5

Yes

80.0

137.4

1

Breakage of glass post

SC 6

No

90.9

157.1

1

Breakage of glass post

SC 7

No

85.4

151.1

2

Breakage of glass post

SC 8

No

95.9

165.5

2

Limit of testing machine

SC 9

No

94.9

167.0

1

Limit of testing machine

SC 10

No

90.3

165.6

2

Breakage of glass post

Mean value

89.8

Coefficient of

variation v = 0.0622

5 % fractile

73.8

the maximum achievable tensile bending strength of the glass. Table 4 shows the results of the tests.

The tests were evaluated statistically. For this, in the case of sample compo­nents that remained intact, the maximum load was defined as the failure load. As the measurements exhibit only a small scatter, a normal distribution can be assumed. A logarithmic normal distribution should only be used with coeffi­cients of variation >0.10 [23]. The 5 % fractile (95 % confidence level) of the fail­ure force is 73.8 kN.

The force-deflection curves (Fig. 13(b)), based on the load-bearing tests, reveal a homogeneous picture and are almost linear. All frame corners—with or without setting block—exhibit an equivalent deformation behavior both qualita­tively and quantatively. These results indicate that a setting block does not exert any significant influence on the load-bearing behavior of an intact frame corner at room temperature. Only the measurements obtained from test No. 4 (SC4) are incorrect because of a measuring error. The deformation measurement was distorted by contact between the displacement transducer mounting and the test rig. These values are, therefore, not shown in the diagram.