Study for the prediction of the long-term durability of seismic isolators.

In Japan, where earthquakes are a frequent occurrence, a lot of seismically isolated buildings have already been constructed. The fundamental element used is a seismic rubber isolator. It consists of rubber and steel plates which are alternatively laminated. Because the seismic isolators are used as the base of buildings, it is important that the special properties of the isolators, such as lateral stiffness and shearing deformation capacity, are strictly managed. It is also extremely important that these properties remain stable throughout the period of use of the isolators.

The application of seismic rubber isolators is considered as one of the steps in assuring the reliability and safe operation of the Fast Breeder Reactor Plant, and the Central Research Institute of the Electric Power Industry is currently carrying out research on its applicability as a safety measure. In this case, the durability of seismic isolators must be demonstrated with certainty. It is therefore extremely important to establish a precise test method for the prediction of the long-term durability of seismic isolators.

Changes in the properties of rubber or plastic products during the period of active use are generally evaluated through a thermal acceleration test based on chemical reaction kinetics. The thermal acceleration test is usually conducted for the target products, based on the previous thermal degradation test results of the thin test sheet. However, it has been demonstrated that the heterogeneous degradation proceeds in thick rubber (refs. 1-3). Therefore, for very large products such as seismic isolators, secular changes should be estimated precisely by the following process. First, we conduct a thermal aging test on a sufficient thickness of rubber specimen.

Using those results, it is then necessary to conduct a thermal accelerating test on the products. However, in current practice, the durability of seismic isolators is often evaluated based only on the test results for thin sheets of rubber. Therefore, in order to propose a test method of accurately predicting the durability of seismic isolators, we examined depth-dependent profiles of tensile properties on thick natural rubber blocks after thermal aging at 60-100[degrees]C. From these measurements and the results of the finite element analysis on stress and strain distribution of the seismic isolator with shear deformation (ref. 4), we would like to propose a new evaluation method to determine the durability of seismic isolators. In this article, we describe the results of our measurements of the depth-dependent profiles of properties in thick natural rubber blocks after thermal aging.

Experimental

In this study, we evaluated two kinds of natural rubber compounds. Rubber compound A and rubber compound B are both materials used for seismic rubber isolators. Thermal aging was carried out at five different temperatures, 60[degrees]C, 70[degrees]C, 80[degrees]C, 90[degrees]C and 100[degrees]C. The specimen used for thermal aging was a rubber block, 220 mm x 150 mm x 40 mm size. The rubber block after aging was cut with a slicer to 2 mm incremental depths, and experimental dumbbells were punched out, according to the No. 3 type specified under the JIS K 6301 standard. The tension test was performed on the dumbbells, following the JIS K 6301 standard, and measurements were taken of elongation ([E.sub.b]), tensile strength ([T.sub.b]), and tensile modulus ([M.sub.100]). Finally, we plotted the depth-dependent profiles of properties, using each measured value against various distances from the surface.

Results and discussion

Figures 1-6 show the measured results of depth-dependent profiles of properties for two compounds of natural rubber blocks after various degrees of thermal aging. It was found that the original profile of each property was almost flat. However, as aging progressed, it became highly heterogeneous. In fact, each of the values of [E.sub.b], [T.sub.b] and [M.sub.100] changed in the range from the surface to a certain depth. Such property changes were not found, however, in the center of the block. It is thought that the combination of oxidative degradation and thermal degradation contributed to the strongly heterogeneous profiles of properties. In other words, oxidative degradation was prominent due to a higher rate of oxygen transmission in the area near the surface showing the greater degree of property changes (hereafter called oxidative degraded area); whereas, only thermal degradation made in progress due to low oxygen transmission in the area nearer to the center of the block which showed rather small changes in properties (hereafter called thermal degraded area). The small changes in properties occurring at the initial stage of aging in the interior of the block is thought to have been due to postcuring.

[CHART OMITTED]

The oxidative degraded area, which showed greater property changes, gradually penetrated in the initial stages of thermal aging. But the velocity of the penetration decreased as time progressed, and eventually the penetration came to a stop after a certain period of time. The depth of the oxidative degraded area from the surface (hereafter called the depth of oxidation) varied with the temperature of aging, and penetrated much deeper at lower temperature.

The fact that the depth of oxidation has temperature dependence is an important consideration for the long-term durability of large sized rubber products such as seismic isolators. Therefore, we carried out a quantitative evaluation for the depth of oxidation, which we defined as the depth at which [E.sub.b] and [T.sub.b] both indicated changes of -5%, and at which [M.sub.100] showed changes of +5% compared to the stable level of such properties in the interior of rubber block. Using this definition, we find the values of depth of oxidation for various aging conditions, and expressed it in relation to time and temperature changes. Looking at the depth of oxidation over time, it is significant that the progress of degradation toward the interior of the block gradually slows down with time, and stops altogether after a certain period of time. It is also noteworthy that the lower the aging temperature the greater the depth of oxidation. Then, we looked analytically at the temperature dependence of the depth of oxidation. In the temperature range of 60-100[degrees]C which we employed in this experiment, there was found to be a linear relationship between the logarithm of the depth and the reciprocal of the absolute temperature. Lindley and Teo previously confirmed the relationship at higher temperatures over 100[degrees]C (ref. 5), and the results of our study confirm that the same relationship holds at the relatively lower temperature below 100[degrees]C. From these findings, we could estimate the depth of oxidation of the rubber at normal temperature as 60 mm for rubber compound A and 100 mm for rubber compound B.

There is thought to be a relationship between oxygen uptake and property changes in the degradation of rubber materials (ref. 6). In light of this, it is fair to assume that the temperature dependence of the depth of oxidation was due to a differential balance between the transmission, diffusion and consumption of oxygen at different temperatures. Moreover, the hard skin layer at the surface, which showed the extremely oxidative degradation, may have also contributed to this process by acting as a barrier to oxygen transmission (refs. 1, 3 and 7).

In order to carry out the thermal accelerating test for rubber products based on chemical reaction kinetics, it is essential to know the activation energy for the degradation of the experimental materials. However, it is very difficult to determine the activation energy from the test results of such thick rubber blocks which showed strongly heterogeneous degradation. This is why we arranged the heterogeneous profile into the items of time change at various depths, of 0-2 mm, 2-4 mm, etc.

Since the depth of oxidation was larger at lower temperature, we also found the reverse to be true that there were smaller changes in properties at higher temperature but larger changes at lower temperature in the interior of the rubber block. Because the degradation response is different between the oxidative degraded area which showed larger changes in properties, and the thermal degraded area which showed smaller changes, it is not possible to obtain data on the activation energy under all types of aging conditions. However, taking into consideration that oxidative degradation progresses to considerable depth at normal temperature, it is necessary to calculate the activation energy in the area in which oxidative degradation has progressed. Then, we calculated the activation energy at intervals of various depths from the data collected after thermal aging in the area in which oxidative degradation progressed.

For [E.sub.b] and [T.sub.b], the activation energy was calculated from the times when changes of 90%, 80% and 70% occurred. For [M.sub.100], the activation energy of the interior was also calculated from the time when change of 110% occurred at the depth of 15-17 mm below the surface. This energy was thought to be equal to the activation energy due to post-curing. The results are shown in tables 1 and 2. As is clear from the data in the tables, we verified that the activation energy calculated in the manner described above, maintained an approximately fixed value independent of depth. The internal activation energy for [M.sub.100], which was due to post-curing, was also found to show the same value.

Table 1 - the activation energy at various depths for compound A

                              Activation energy (kJ/mol)
                     0-2mm    2-4mm    4-6mm    6-8mm   15-17mm
               90%     78.7     76.6     90.8
[E.sub.b]      80%     84.1     80.8     83.3
               70%     79.9     88.7
               90%     83.7     78.2     81.2
[T.sub.b]      80%     78.7     80.3     82.0
               70%     82.4     85.8
[M.sub.100]   110%                                         77.0

Table 2 - the activation energy at various depths for compound B

                              Activation energy (kJ/mol)
                     0-2mm    2-4mm    4-6mm    6-8mm   15-17mm
               90%     84.9     95.4
[E.sub.b]      80%     88.7     92.5     86.2     88.3
               70%     92.5     95.0     92.9     91.6
               90%     83.7     89.1
[T.sub.b]      80%     84.1     90.8     90.0     85.8
               70%     86.6     90.0     87.9     90.4
[M.sub.100]   110%                                         84.5

There is thought to be some relationship between changes in properties of rubber due to oxidative degradation and the amount of oxygen absorbed by rubber (ref. 6). The oxygen absorption is thought to be a characteristic response of the material. Therefore, the activation energy was thought to be a characteristic value of the material, independent of depth.

The methods for estimation of long-term durability of seismic isolators and for estimating the profiles of properties in thick rubber block after secular change are both described in a later piece.

Conclusion

In order to propose a precise test method for estimating the durability of seismic isolators, we examined the depth-dependent profiles of tensile properties in thick natural rubber blocks after thermal aging at 60-100[degrees]C.

The results of this study established the following conclusions:

* After thermal aging, the rubber blocks exhibited heterogeneous degradation behavior. These rubber blocks could be divided into two areas, the oxidative degraded area near the surface showing large changes in the properties, and the thermal degraded area in the interior showing small property changes.

* It was established that the depth of oxidation showed temperature dependence, with greater depth of oxidation at lower temperatures. There was also found to be a linear relationship between the logarithm of the depth of oxidation and the reciprocal of the absolute temperature. As a result, the depths of oxidation at normal temperature may be estimated to depths of 6-10 cm.

* Having calculated the activation energy at depths of 2 mm each from the surface in the oxidative degraded area, it was found that the activation energy held a fixed value independent of depth.

References

(1.)R.L. Clough and K.T. Gillen, Polym, Deg. and Stab. 38, 47 (1992).

(2.)F. Bjoke and B. Stenberg, Polym. Test. 5, 245 (1985).

(3.)G.J. van Amerongen, Rubber Chem. Technol. 37, 1065 (1964).

(4.)Y. Fukahori, J. Soc. Rubber Industry Japan 60, 397 (1987).

(5.)P.B. Lindley and S.C. Teo, Plast. Rubber Mater. Appl. 2, 82 (1977).

(6.)E.T. McDonel and J.R. Shelton, J. Chem. Eng. Data 4, 360 (1959).

(7.)B. Stenberg, Y.J. Shur and J.F. Jansson, J. Appl. Polym. Sci. 35, 511 (1979).