Results of Dynamic-Mechanical Material Analysis

Identification of the Complex Mechanical Material Behavior

A precondition to the physically correct identification of the complex mechani­cal material behavior is the determination of the linear-visco-elastic range of materials properties (LVE region). In this region, a controlled loading causes a proportional material response without any irreversible structural deteriora­tion. In addition, this region also provides some first information about the material’s structure.

Figure 6 shows that, even at lowest temperatures down to —60°C, all seal­ants tested in this study are able to accommodate shear deformations of at least 1%. In detail, the sealants exhibit different responses; however, in general, G’ was around ten times larger than G”, which corresponds to a visco-elastic-solid behavior. The material’s stiffness ranking under controlled increasing shear amplitudes at —60°C was as follows: Sealant A > Sealant B > Sealant C. The bearable shear deformation у [%] inside the LVE range for Sealant C is around 9 times higher compared to Sealants A and B. To induce any specimen

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.

deformation the activating shear stress sActivation for Sealants A and B is much higher. Maximum shear stresses TLimit above 5.1 and 5.0 MPa, respectively, cre­ate irreversible structural deterioration in Sealants A and B specimens, leading to break failure. The maximum test shear stress TLimit for Sealant C is 1.4 MPa, however, without any specimen deterioration but leaving the use conditions of this test equipment. Another precondition, to make DMA measurements on cured solid sealants with the solid rectangular fixture feasible, was the determi­nation of the maximum tolerable normal force providing physically exact test conditions, without causing structural effects in the sealant specimen.

Interpretation of the results shown in Fig. 7 for the three sealants studied here suggest that at a temperature of +80°C—which seems to be a suitable esti­mation of the temperature exposure sealants experience during summer [10]— tensile forces of up to 0.7 N do not indicate any structural changes in the seal­ants. Normal forces above this value indicate leaving of the LVE range for Seal­ant C. That’s why we decided to use this normal force as the control value for all sealant tests with solid rectangular fixture (SRF; see Fig. 5) identifying the com­plex material behavior. Normal force values above Fmax. (see Fig. 7) cause slip­ping effects between the specimen and clamping device, which have to be avoided to ensure reproducible test conditions.

Quite important information about the complex mechanical behavior of the sealants can be deduced from subjecting the specimens to the temperature- sweep mode, where the temperature dependency of the storage and loss moduli (i. e., G’ and G") are quantified. In addition, this sweep mode allows one to obtain detailed polymeric material structural information (morphology) with respect to thermal effects.

According to Fig. 8, over the full range of test temperatures, all sealants show a visco-elastic-solid behavior (G’ > G”). Sealants A and B are very similar in mechanical behavior but differences in absolute stiffness are evident. Both sealants exhibit changes in temperature-dependent stiffness between + 60° C and the end of the rubber-like elasticity range, which is below temperatures ranging between —57°C and —61°C. Entropy-elasticity effects with increasing temperatures were not detected inside the rubber-elasticity range of the seal­ants, probably attributed to the test configuration chosen. Nearly complete tem­perature independence of the moduli was evident down to temperatures of —73 °C for Sealant C, and this feature characterizes this product as having the broadest field of application. The ranking of the sealant products investigated in terms of stiffness at temperatures ranging from —60°C to +150°C was as follows: Sealant C > Sealant B > Sealant A. It is interesting to note that the ranking with respect to the stiffness of the different products was unrelated to the filler con­tent (shown in Table 1 and Fig. 8 as mass% of total formulation); this suggests that either the crosslink density is decisive in influencing the temperature – dependent stiffness or different kinds of fillers were used in these products.

A detailed evaluation of the temperature dependency, e. g., of tand (ratio G”/G’), yields information on general phase change or transition temperatures. Resulting from the temperature-sweep investigations we found three character­istic effects by T1 (low transition temperature, indicating lower end of rubber – elasticity region), T2 (high transition temperature, indicating upper end 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.

rubber-elasticity region by melting or, finally, pyrolytic processes), and T3 (sec­ondary transition temperature, indicating internal structure effects (“areas of relaxation") without fundamental changes of material performance), which are typical for the formulated product, exploring relevant changes in material per­formance and which also can be used (as a “fingerprint") in identifying each product. This kind of assessment underlines the potential of this material test approach also for quality-assurance purposes. Resulting from the temperature – sweep test results for Sealants A, B, and C as shown in Figs. 8 and 9, we found indication for temperature-dependent material effects listed in Table 2.

The determination of the upper phase transition temperature T2 was sup­ported by additional thermo-gravimetric investigations indicating beginning pyrolytic polymer degradation (not presented here). A traceability of the effects measured to materials composition and their interpretation for more detailed material exploration (differentiation between structural or technological effect), respectively, as well as additional proof of plausibility, e. g., in comparison to other thermo-analytical methods, like DSC, according to variation of loading parameters, will be a task of further investigations. Taking into account possible effects by thermal inertia of specimen geometry (e. g., delayed transition temper­atures depending on controlled heating or cooling procedure), a minimized temperature rate was chosen. Nevertheless, it has to be noted that the effects described were strongly connected with controlled cooling test procedures.

Temperature sweep test

2-Ю

l

1,8-10

Г,; C

1,6-10

V

– rfl’J

і

1,4-10 і. „ – o’1

ГГрЛМ Щ

IS

таги

T 1,2-10

tanfs) ,_-1.

ЧП

1

гР»

-2

k

Ti

T

ЗІ В

1*3

A

s

8-Ю

-2

Tf

6-Ю

-2

4-Ю

-2

9-1П

-100 -80 -60 -40 -20 0 20 40 60 80 100 120 °С 160

Temperature T ————— ►

Temperature sweep Sealant A FEV 1 Temperature sweep Sealant В FEV1 Temperature sweep Sealant C FEV1 Tangent of loss angle Tangent of loss angle Tangent of loss angle

FIG. 9—Tangent of loss angle (tanS) as a function of temperature (temperature-sweep mode).

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 2—Material-identifying information by temperature-sweep mode.

Material

Lower transition temperature T1 (°C)

Upper transition temperature T2 (°C)

Secondary transition temperature T3 (°C)

A

-61

+ 200

+ 30

B

-57

+ 200

+ 20

C

-73

+ 250

n/a

Taking into account the matrix of decisive loads to which the sealant is sub­jected in actual service (see Fig. 2), another important aspect is the frequency – dependence of the sealant’s mechanical behavior. In frequency-sweep-mode, conducted at +25°C, we obtained moduli graphs typical of cross-linked poly­mers with steadily low rising slopes between 10-2 and 3 x 101 Hz (not shown as figures). As already reported for silicone products (e. g., Ref [2]), simplified repe­tition of frequency-sweep measurements at various temperatures permits the creation of time-temperature-shift (TTS) master curves using the Williams- Landel-Ferry (WLF) method [11,12] over a substantially extended frequency range for each reference temperature within the service temperature range (see Fig. 10 for an example of a TTS master curve for Sealant A).

Time-Temperature-Shift Mastercurve

Frequency (со aT)

FIG. 10—Time-temperature-shift (TTS) master curves for Sealant A (T: + 80, + 40, + 20, +10, -20, and -40°C).

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 theoretical background to the frequency-temperature analogy (i. e., the time-temperature shift) proposed by Williams, Landel, and Ferry can be found in the textbook reference published by Ferry [12]. Use of the WLF method makes mechanical moduli for every temperature or load rate available, which substantially improves the application of modern design tools like FEM during the design process.

Figure 11 compares the time-temperature-shift (TTS) master curves for the three sealants for a given reference temperature of —20°C. The chart also indi­cates typical frequencies experienced during the service life of SSG sealants, such as static state, wind, seismic, and impact (hurricane debris or explosion blast), and the relevant mechanical behavior of each sealant. Compared to Seal­ants A and B, the frequency-sweep graph of Sealant C indicates a lower fre­quency dependency and higher cross-linking degree (see Fig. 11). To validate these first results, as well as to improve the knowledge of temperature – frequency dependencies and their structural effects, further investigations are necessary.

Based on the methods described here for the identification of complex me­chanical behavior, the shear moduli for every temperature and loading rate within the LVE region can be determined. If, additionally, the Poisson’s ratio of the material is known, a complete moduli transformation of shear modulus (G) to tensile modulus (E) is possible, permitting new opportunities for modern

Time-Temperature-Shift Mastercurves at Tret -20 C

p Sealant C

4 Sealant В Sealant A

10

Impact

wind

Static state

Frequency(coaT) ■

Additional test results

Stiffness Cross­Sealant Long I Short ‘inking

time time degree

A 9* 10* 9М0» low

В 1*10* 6* 10* low

С 5M06 3»10* higher

FIG. 11—Comparison of time-temperature-shift (TTS) master curves (for reference —20°C).

ted); Tue May 6 12:07:08 EDT 2014

Downloaded/printed by

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

section of DMA

material characterisation

load duration [Hz]

NLVE-region by

shear strength

tension strength

FIG. 12—Identification of the mechanical behavior by dynamic-mechanical analysis (DMA) with respect to material response to loading.

design algorithm (see Fig. 12 for the extended information potential obtained by DMA).