Determination of Mechanical Values of the Adhesives

In a first approach, the mechanical properties of the selected adhesives are deter­mined by carrying out small-scale tests using standardized specimens (Table 1, IV):

• Determination of tension strength, Young’s modulus, and Poisson’s ra­tio by means of dumbbell specimens according to DIN EN 527 [16].

• Aging of block shear specimen through the immersion test (six-week water bathat60°C [17]).

• Determination of shear modulus, shear strength, and shear stress – sliding behavior by pressure block shear tests of aged and unaged glass specimens according to EN ISO 13445 [18].

• Differential scanning calorimetry to obtain the temperature behavior, including the glass transition temperature.

The tension tests are performed on dumbbell specimens (specimen type 1A) using a universal testing machine with a video extensometer. In a first approach, the tests are carried out as displacement-controlled with rates of 1.0 mm/min.

For the intended shear tests, usually thick steel or aluminum adherents, according to DIN EN 14869-2 [19], are used. Because of the need to test glass substrates, the geometries mentioned in EN 14869 are not applicable; therefore,

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.

within the INNOGLAST project [5] the shear values were determined using a modified block shear test [20] according to EN ISO 13445.

It is important to note that for the evaluation of test results, the failure mode of the tested shear specimens must be similar, which means that the pre­dominant failure mode must be cohesive, occurring in the adhesive layer or near the boundary layer. That is why the use of hybrid steel-glass bonded joints should be consciously discarded for the tests in order to avoid mixed steel-glass bonds, with which the failure cannot always be allocated clearly to the steel or glass surface, potentially making the failure ambiguous.

The immersion tests are performed according to DIN 53287 [17]. The test is defined by a six-week water storage in 60° C demineralized water. After six weeks the specimens are removed from the water bath and cured in a climate cham­ber. The aged specimens are then subjected to shear tests. The results are com­pared to the shear test results of unaged specimens.

Table 2 gives an overview of the important mechanical values of the selected adhesives as determined by the tests mentioned above. Figure 2 shows the change of shear moduli determined in block shear tests on unaged specimens and aged specimens after six weeks of water bath immersion. In the diagram, mean values and 5 % fractiles are displayed. Especially for the high-strength ep­oxy resins and polyurethane adhesives, there is a significant decrease of stiff­ness and strength after aging; in the case of K02, the specimen even showed considerable embrittlement. However, when the silicone K07, which is com­monly used for structural glazing applications, is compared with high – performance polyurethane (K05, K06), it becomes evident that there are mod­ern adhesive systems with adequate stiffness, strength, and ductility. These adhesives offer good aging resistance together with a stiffness and strength at least five times higher than that for silicones. Even if the durability still must be further verified and conferred on building purposes, new possibilities for dimensioning and load transfer by adhesive connections are possible.

TABLE 2—Characteristics and mechanical values of the selected adhesives (strengths and strains are log-normal distributed 5 % fractile values).

Adhesive

K01

K02

K03

K04

K05

K06

K07

Adhesive system

EP

PU

PU

AC

PU

PU

SI

Components

2

2

2

1

2

2

2

Pot life, min

90

10

90

1(UV)

15

30

10

Tension strength, MPa

27.3

9.3

7.9

9.3

4.2

6.3

0.9

Elongation at break, %

3.3

26

0.5

74

30

111

76

Young’s modulus, MPa

1499

231

1906

314

16.5

20

2.4

Unaged shear strength, MPa

18.6

12.9

2.7

6.3

6.7

3.7

1.4

Unaged shear modulus, MPa

242

128

146

63

5.4

3.4

1.1

Glass transition temperature, °C

60

25

43

-61

-83

-90

Note: EP, epoxy resin; PU, polyurethane; AC, acrylates; SI, silicone

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.

Determination of Mechanical Values of the Adhesives

FIG. 2—Change of shear moduli before and after aging as determined by block shear tests.

For the following finite element (FE) modeling of the push-out and compo­nent tests, the standardized tests described above showed restrictions regarding the precision with which they could be used as input values for finite element method. In particular, the block shear tests revealed non-neglectable bending moments due to the single lapped geometry and their eccentricity. The resulting peel stresses significantly affected the shear stress-strain curve, thus preventing simple shear tests.

Therefore, the approach has been modified, resulting in a changed test setup and test control. The tension tests then have been performed with a Zwick Z100 universal testing machine with touching high-precision extensometers. All tests have been performed as strain-controlled with varying strain rates (0.0033/0.033/0.33/3.33 1/Min.) in order to cover the rate dependency, in par­ticular for common facade application (< 1 Hz).

The shear tests have been carried out on modified thick adherent test speci­mens made of steel and glass using the same strain rate as for the tension tests, in order to compare the results of both load types. The resulting characteristic values have been used as input values for Drucker-Prager material laws in the case of stiff, elasto-plastic adhesives (epoxy resins, some polyurethanes) and have been verified with butt-joint tests. With Drucker-Prager material laws, the influence of the hydrostatic state of stress and yielding can be considered. The modeling of visco-elastic adhesives (silicones, most polyurethanes) has been done using Mooney-Rivlin, Arruda-Boyce, Neo-Hookean, or Ogden, and in most cases Arruda-Boyce and Neo-Hookean fit best.

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.

Additionally, the influence of the surface (steel or glass surface) has been taken into account in order to secure this strategy. This modified approach is still ongoing and will be finished by the end of this year [21].