Aging under Constant Air Humidity Conditions

probability. But a massive influence on the dimension was detectable, whereby aluminosilicate fibers were the only test samples to show no increase of pressure resistance with larger wall thickness. In contrast to aluminosilicate, soda-lime fibers showed a direct influence on dimension and age. The slight increase of pressure resistance could be affected by glass corrosion caused by air moisture at non-loaded conditions. At loaded conditions, glass corrosions could have the reverse effect.

As result it can be summarized:

- Test series were carried out with four different glass materials and two different dimensions. These fibers were stored at T_storage = 35 °C and relative air humidity of ϕ = 55 % for at least one year. The test results were compared with those of new tested fibers. An overview of the development of pressure resistance is given in Table 20.

- Under non loaded storage conditions the effect of aging due to environmental influences was negligible for fibers made of borosilicate.

- The detected deviations were at a scale that these deviations can be attributed to production tolerances. The tested hollow fibers were made of material with the same chemical composition but drawn in different production batches. Changes even in traces may lead to small deviations of test results.

Table 20: Development of burst pressure resistance of glass fibers after aging one year compared new tested fibers by reference to the characteristic pressure P

Wall thickness s = 20 µm

Wall thickness s = 50 µm

Borosilicate Constant Constant

Aluminosilicate Decreased Decreased

Quartz Increased Increased

Soda-lime Increased Increased

brought into focus. Hollow fibers made of borosilicate 3.3 (DURAN) and quartz glass with dimensions of d_o = 400 µm, d_i = 300 µm, s = 50 µm were investigated in test series of new fibers and aged fibers to examine the influence of air moisture. Thereby the tested hollow fibers were taken from the same production batch the test samples of the investigation of different glass types (chapter 6.1.2) were taken from. Hence, any possible influence of different material history was excluded and the results were comparable.

The test samples were glued in stainless steel tubes and stored for 7 weeks in defined air humidity, described in 5.3.2. The test results of borosilicate 3.3 fibers are listed in Table 21. As reference value, the data of new fibers from the same dimensions and batch were used. The new fibers were tested only a few days after delivery.

Table 21: Characteristic test data of borosilicate 3.3 fibers at different air humidity with the dimension d_o = 400 µm, d_i = 300 µm, s = 50 µm

Air humidity φ [%]

Min. burst pressure pmin

[MPa]

Max. burst pressure pmax

[MPa]

Form parameter b

Characteristic pressure P [MPa]

New fiber 61.9 114.2 7.3 89.3

30.9 65.2 138.4 6.8 92.1

75.8 56.0 130.7 6.2 95.2

A consistent behavior of the three test series can be clearly seen. New borosilicate fibers tested at average air moisture of φ = 55 % exhibit the highest form parameter b = 7.3 but also the lowest characteristic and maximum burst pressure. A slight increase of the minimum burst pressure p_min and characteristic pressure P is detectable for hollow fibers stored at φ = 30.9 % relative air humidity. More obvious is the increase of maximum burst pressure p_max by the factor 1.2. Through this, the scattering of measured values is wider, which is detectable by a slightly smaller form parameter b. Also the test results of fibers stored in higher air moisture of φ = 75.8 % exhibit only small deviations from reference values. However, the highest characteristic pressure P indicates a narrow distribution of measured burst pressure in high pressure ranges. Generally it can be seen that the test results are similar and comparable. Nevertheless, the higher characteristic pressures P could be an indication for positive influence of defined air humidity by rounding the edges of defects.

The corresponding distributions of failure probability of the three test series of hollow borosilicate 3.3 fibers are plotted in Figure 37 against the measured burst pressures.

Figure 37: Influence of aging under different humidity on failure probability and pressure resistance of borosilicate 3.3 (DURAN) fibers with the dimensions d_o = 400 µm, d_i = 300 µm, s = 50 µm

The similar development of all graphs is obvious. Due to similar minimum burst pressure values, the initial points of the graphs are close together. The gentle decrease of form parameter b with simultaneous slight increase of characteristic pressure P leads to the displacement of the graphs to the right. However, at pressures above p = 80 MPa a higher pressure resistance of fibers treated with defined constant air humidity is observable. The failure probability F_B at p = 100 MPa of new fibers is about F_B = 90 %.

Fibers stored at φ = 75.8 % air humidity features a failure probability of about FB = 75 % at the same pressure level which is an improvement of about 20 % in pressure resistance.

In addition to borosilicate 3.3 fibers, test samples made of quartz glass were stored and tested in the same way. The characteristic data of these series are summarized in Table 22. Only small differences between the single test series are detectable. It can be seen

that all characteristic values increase under the influence of defined and constant air humidity.

Table 22: Characteristic test data of hollow quartz fibers stored at different air humidity with the dimensions d_o = 400 µm, d_i = 300 µm, s = 50 µm

Air humidity φ [%]

Min. burst pressure pmin

[MPa]

Max. burst pressure pmax

[MPa]

Form parameter b

Characteristic pressure P [MPa]

New fibers 24.4 94.4 4.2 76.6

30.9 38.4 104.5 5.2 81.1

75.8 39.9 109.0 5.6 82.2

The calculated failure probabilities of quartz samples under the influence of air moisture are shown in Figure 38.

Figure 38: Influence of different humidity on failure probability and pressure resistance of hollow quartz glass fibers with the dimensions d_o = 400 µm, d_i = 300 µm, s = 50 µm

The three different curves of quartz glass fibers exhibit nearly the same development.

Thus, the graph of fibers stored at φ = 30.9 % superimpose this of test samples stored at φ = 75.8 %. At lower pressures, the graph of new hollow fibers deviates from the other two graphs due to the lower minimum burst pressure. At higher pressure values an approximation is observable.

The investigation of the influence of air moisture under non-loaded conditions with defined and constant air moisture and temperature conditions showed no considerable variations with comparable results within the different glass types. Nevertheless, an influence of the different atmosphere on the hollow fibers was observable especially for borosilicate fibers. The presence of air moisture in defined concentrations leads to the increase of the maximum burst pressure and the characteristic pressure P with simultaneous decrease of form parameter b and nearly constant minimum burst pressure. Water, even in small concentrations, forms a thin film on the surface of glass and leads to the dissolving of alkaline substances out of the glass network [97]. That chemical reaction can affect particular the surface characteristics in flaws [176].

The distribution of defects in or on the fibers seems to have an important role in that behavior. It can be assumed that the low burst pressures within each test series are caused by test samples with bigger or a high number of defects. Such defects led to stress peaks inside the fiber and caused the breakage. Consequently, test samples with higher burst pressures could exhibit smaller defects or, respectively, a lower number of defects. Hydrolytic reactions on the surface could have influenced the characteristics of the flaws especially of the smaller ones. The edges of defects could be rounded by the hydrolytic reaction on the surface and therefore possible stress peaks under inner pressure could be mitigated or removed [176]. Due to the constant air moisture the hydrolytic reaction featured a higher intensity than in alterable moisture [94]. The decreasing failure probability of samples stored at defined concentration at higher internal pressure in comparison to new fibers could be an indication for this assumption.

Therefore borosilicate fibers with small flaws or a low number of defects could withstand higher pressures after being stored in constant air moistures. Bigger flaws were rounded and changed by the chemical reaction as well. Because of the size or the depth a positive influence were not possible.

Also the concentration of the air moisture had a positive influence on the pressure resistance of borosilicate fibers. Test samples stored in a relative air moisture of φ = 75.8 % exhibit lower values of failure probability at higher pressure than those stored at φ = 30.9 %. A higher concentration of water in the atmosphere or often alternating air

moisture benefits the dissolving of alkaline substances out of the network [94]. The edges might be rounded more and therefore the stress distribution at internal pressure is more regular and smoother.

Hydrolytic reaction can take place at all glasses featuring alkaline or earth alkaline substances. Quartz glass is made of pure silica without the addition of any substances.

Consequently hydrolytic reaction could not take place on the surface of tested quartz glass fibers. Nevertheless, a comparable trend of increasing pressure resistance was detectable as seen at tests with the borosilicate 3.3 fibers. All characteristic data features higher values when being stored in defined air moisture before testing. Due to impurities in raw material only in traces the presence of water-soluble substances in quartz glass could be possible. In that case hydrolytic reaction might take place on the surface of tested quartz fiber while being stored. Especially in the range between p = 40 MPa and p = 80 MPa a meaningful increase of pressure resistance could be detected in Figure 38. As on the surface of borosilicate fibers a hydrolytic reaction could lead to the rounding of edges of possible flaws. The result can be the reduction of their decreasing influence on the pressure resistance.

Fibers of both materials borosilicate 3.3 and quartz showed a positive effect of air moisture on their pressure resistance under non-loaded conditions where hydrolytic reactions could originate the rounding of edges of surface flaws [176]. Defined and constant air moisture leads to the increase of the resistance against inner pressure.

Thereby at higher concentration of air moisture that effect was even higher.

Nevertheless, the effect was slightly low. The characteristic values of the test series within one type of glass are still close together. However, the difference of quartz and borosilicate 3.3 fibers was clearly shown. At a defined burst pressure of p_B = 80 MPa fibers made of borosilicate 3.3 exhibit a failure probability of FB = 30 %. The failure probability of quartz fibers at same burst pressure reached a value of F_B = 80 %. Hence, borosilicate fibers showed a significant higher resistance against inner pressure load in comparison to quartz fibers.

The results of this part of the thesis can be summarized as following:

- Borosilicate and quartz glass fibers with the dimensions d_o = 400 µm, d_i = 300 µm, s = 50 µm were tested after storing them for more than 1100 h without inner load in defined air moistures. The test results were compared to measured burst pressures determined with new fibers and are summarized in Table 23.

- Only slight differences under the given circumstances were detected. Hence, the pressure resistance of hollow fibers made of borosilicate or quartz glass was independent on the air moisture under non loaded conditions.

- The high chemical resistance of both glasses excluded possible decreasing effects on the pressure resistance by chemical reactions on the surface. Under non-loaded conditions hydrolytic reactions may even have a positive effect on the resistance against inner pressure by rounding and increasing the radius of surface cracks [94].

Arising stresses inside the material can be distributed more homogeneous and peaks will be reduced.

- Because of the higher burst pressure values independent on air moisture it can be concluded that under given circumstances borosilicate fibers exhibit a higher pressure resistance than those made of quartz.

Table 23: Pressure resistance development of glass fibers aged under the influence of different defined air humidity compared new tested fibers by reference to the characteristic pressure P

Low air humidity φ = 30.9 %

High air humidity φ = 75.8 % Borosilicate Slightly increased Slightly increased

Quartz Increased Increased