Page images
PDF
EPUB
[graphic][graphic][merged small][merged small][merged small]
[merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]
[blocks in formation]

Prior to testing, each creep-rupture specimen was heated in air in a tube furnace to the desired temperature and equilibrated at temperature for 16 hr. It was tested at this temperature under constant load in a multi-lever creep-rupture machine equipped with motorized jack to alleviate shock loading. The loading time was 1 to 2 min. Test temperatures were controlled to an accuracy of ± 2 °F. Extension-time data were obtained from electric contact follow-up type extensometers attached to the shoulders of the specimens.

Metallographic examinations were made on selected ruptured specimens to ascertain the effects of notch geometry, temperature, and rupture time on micro- and macrocracking and changes in microstructure.

3. Results and Discussion

Test data obtained in the present investigation are shown in figures 2 through 10. Extension-time curves

were constructed for all the specimens tested to complete fracture. It was observed that the curves for the notched specimens had the same general shape as those usually associated with creep of unnotched specimens. Each exhibited first stage (decreasing creep rate), second stage (constant rate), and third stage (increasing rate). The first stage is not evident in figures 2 and 3 as it lasted only a very short time as compared to the total test time. As might be expected for tests run at the same temperature and stress, the total strain decreased with increase in notch depth and decrease in root radius. However, as indicated in figures 2 and 3, rupture times and reduction of area values could not be predicted from strain-time data.

A number of investigators [4] have predicted that a linear relation exists between stress-rupture time, stresslog rupture time, or log stress-log rupture time. Data obtained in the present investigation were plotted in each manner, and linear relations were observed only over limited ranges of stresses at each temperature except 600 °F. At this temperature, little or no creep was apparent even at stresses in excess of 95 percent of the short-time tensile strength. Difference in behavior may be due in part to (1) the relaxation of stress concentration at the root radius of the notch as a result of plastic yielding, (2) time at temperature, (3) formation of a compound such as TisAl, or (4) the healing of internal voids. Stress-log rupture time curves for some of the Ti-8Al-1Mo-1V specimens are shown in figure 4. In general, the relative positions of the curves are affected by notch geometry more at low temperatures and high stresses than at high tem peratures and low stresses. For example, at 1200 °F and very low stresses, notch geometry appeared to have no

[merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

effect on the stress-rupture time relations. At 800 °F, an inflection point must occur at stresses below those used in these tests or the stress could be zero for extremely long rupture times. Conversely, at 1200 °F, an inflection point must occur for stresses higher than those shown in figure 4. At 1000 F, the inflection points are clearly shown. and the time of their occurrence appears to be affected by notch geometry.

According to elastic theory, root radius has a more dominating influence on the stress concentration at the base of the notch than does the notch depth. The influence of each of these variables on rupture times at different stresses is shown in figures 5 and 6 for the present tests. Although no exact equivalence between notch depth and root radius could be established for the specimens (table 2), each of these variables had a marked effect on the rupture behavior. The effect of each is more apparent at the low temperatures and high stresses than at the high temperatures and low stresses.

Engineering design curves, showing stress-temperature relations to produce rupture of the specimens in 1, 10. 100, and 1000 hr, are presented in figure 7. Within the range of temperatures used in the tests, no inflection points are evident in the 1-hr and 10-hr curves (figs. 7A and 7B). Obviously, these points must occur at higher values of temperature than those used in this investigation or the specimens would have to rupture in 1 or 10 hr at zero stress. As indicated in figures 7C and 7D, the

[merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors]
[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][subsumed][subsumed][subsumed][subsumed][subsumed][merged small][merged small][ocr errors][subsumed][subsumed][subsumed][subsumed][subsumed][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]
[merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

occurrence of these inflection points is affected both by notch geometry and test time.

A number of "parametric" expressions have been developed to evaluate creep-rupture data for specific materials over wide ranges of stresses and temperatures [4]. It was previously shown that creep data for the unnotched specimens of this Ti-8Al-1Mo-IV alloy [6] could be described by a single curve. The time-temperature parameter used was derived by Larson and Miller [8]. As shown in figure 8 for the present tests, a single curve can also describe stress-"parameter" relations for each group of notched specimens; however, no simple translation was available to cause all the curves to coincide. The difference in the shape of these curves from those previously shown [6] is due to the fact that the stress values in figure 8 are plotted on a log instead of a linear basis. This method of analysis is recommended for comparing and storing engineering data obtained over wide ranges of test conditions, even though the theoretical significance of the relations is

suspect.

The relation between reduction of area values and rup

ture-time is shown in figure 9. Rupture time had little or no effect on ductility, as defined by reduction of area values, at 800 °F (fig. 9A), whereas a large effect was observed at 1000 and 1200 °F (figs. 9B and 9C). Although, at the latter two temperatures, the ductility appeared to increase with increase in rupture time, inflection points and a reversal in one ductility-time curve are shown. One point at 1200 °F and 1450 hr was omitted for the unnotched curve as it nearly coincided with the notched data. These indicate the possibility of a change in the mode of deformation. Introduction of a notch tended to lower reduction of area values below those of the unnotched specimens. Moreover, decreasing the root radius caused the reduction of area values to decrease even more. However, increasing the notch depth from 50 to 70 percent had little effect on reduction of area.

The influence of prior thermal-strain history on creeprupture behavior of specimens tested at 1000 °F with a stress of 30 ksi (207 MN/m2) is shown in figure 10. Although the number of tests are limited, several general observations can be made concerning the effect of prestraining. (1) The rate of extension is significantly changed from that of the unstrained material. (2) Reduction of area values were not seriously affected. (3) Rupture times were decreased by prestraining at temperatures above and below 1000 °F. (4) Prestraining at 1200 °F relieved the stress at the root of the notch to such an extent that the rupture time was increased above that of the metal strained at 800 °F (curves 10B and 10C). This occurred even though the amount of prestraining at 1200 °F was greater than that at 800 °F.

4. Metallography

Metallographic examinations were made on a number of selected specimens after rupturing in creep. Little or no difference in microstructure from that observed in short-time tensile tests [3] was apparent for specimens tested at 600 or 800 °F. This was not totally unexpected since it was indicated, in figure 9 of the present paper, that the reduction of area values for specimens tested at 800 °F were independent of rupture time. Photomicrographs of several specimens tested to rupture at 1000 and 1200 °F are shown in figures 11 through 14. Specimens, shown in figures 11 and 12, were lightly etched to remove any surface effects due to metallographic preparation of the samples. At 1000 °F, increasing the root radius from 0.01-in (figs. 11A and 11B) to 0.5-in (fig. 12A) tended to cause an increase in the number of internal cracks that did not link up to become a part of the main fracture surface. At 1200 °F, however, no significant differences in the number of internal cracks were observed for the specimens even though the increase in elastic stress concentration factor (from 1.1 to 3.9) appeared to cause an increase in the tendency toward surface cracking (figs. 11C and 11D and 12B, 12C, and 12D). Increasing the temperature from 1000 to 1200 °F increased the number and size of cracks in regions away from complete fracture. However, the relation between the number or size of cracks and test time was different at 1000 from that at 1200 °F. At 1000 °F, tendency to form

« PreviousContinue »