« ZurückWeiter »
meteorites (Edwards, 1953, 1955) in comparison with terrestrial steels seem to exclude terrestrial contamination as the source of hydrogen.
MORE SUPPORTING EVIDENCE
Supporting evidence for the conjectured brittle failure of parent bodies of iron meteorites can be drawn from several areas of research.
One area is the study of preterrestrial deformation effects in meteorites (Axon, 1969; Baldanza and Pialli, 1969). For example, in their study of “dynamically” deformed structures in irons and chondrites, Baldanza and Pialli found extensive evidence that “shear forces of a slow character” acted to distort the phase structure in irons, and concluded that “the pressure was related to a slow dynamic event and temperature was confined to relatively low values.” Such effects are not expected in collisional shock events. Moreover, the recrystallization observed mainly along faults (indicating local loss in ductility, i.e., embrittlement), “shear fractures,” and “deformation due to prolonged stress action” seem to fit well the path leading to hydrogen brittle failure, as does the history formulated for these meteorites (culminating in the production of “split” bodies, when conditions of low temperatures were attained). It is remarkable that the critical range of temperature for failure due to molecular hydrogen is 173 to 373 K (Bernstein, 1970), thus bracketing the values relevant for iron meteorites at 1 AU (~363 K) to 3 AU (~223 K). Moreover, in the presence of atomic hydrogen, brittle failure occurs over a wider range of low temperatures (Nelson, Williams, and Tetelman, 1971), whereas “hydrogen cathode charging” of steels (which is equivalent to solar-wind implantation of hydrogen) is known to cause irreversible brittle failure at levels of 5 to 8 ppm of H2 even at 77 K (Barth and Steigerwald, 1970).
Another significant fact is the presence of very high levels of hydrogen (~8.4 × 10−2 cm3/g) in some gas-rich meteorites (Lord, 1969) as well as proton contents of 4 X 10% to 2 X 1029 per gram found in various chondrites (Chatelain et al., 1970), all being accountable by ~10° equivalent irradiation years at 1 AU in typical solar-wind proton fluxes of ~3 x 108 cm-2-sec-1 (E > 1 keV). Not only is the hydrogen effectively implanted by solar wind into the grain surfaces (possibly prior to their aggregation (Lord, 1969) or while suspended in jetstreams (Arrhenius and Alfvén, 1971; Trulsen, in this book"), but it also is released mostly above -700 K from stones (Lord, 1969), whereas in irons it can be held as “residual hydrogen” to above 1000 K (Johnson and Hill, 1960). In Apollo 11 Moon material (Fireman, D'Amico, and De Felice, 1970), a hydrogen concentration gradient was found to exist in rocks, and the abundant hydrogen content of the lunar soil (1.2 cm3/g) was only in part attributable to solar wind, thus suggesting that some primordial hydrogen was retained.
The solar wind therefore could establish in surface layers a local concentration gradient of hydrogen and an equivalent quasi-static internal stress known to facilitate an eventual fracture of the parent body. It also could establish a trapping layer for hydrogen at irradiation-caused or other defects and dislocations close to the surface. Can the evidence of shock in some iron meteorites (Jaeger and Lipschutz, 1967; Jain and Lipschutz, 1969) rule out the occurrence of noncollisional fragmentation? It seems that it cannot do so because shock of mild to moderate levels appears to be limited to certain groups of iron meteorites (such as the group III octahedrites, which also cluster in cosmic-ray ages at ~650X 10° yr). Other groups, like the hexahedrites, show no evidence of shock; unshocked octahedrites, although randomly distributed among the Ga-Ge groups, also seem to exhibit peaks in their cosmic-ray age distribution at 200 to 500 and 800 to 1000 X 10° yr, suggesting a formation in noncollisional discrete events (Jain and Lipschutz, 1969; Voshage, 1967). Even some group III octahedrites, like Henbury and Cape York, are apparently unaltered; although they are believed to have been mildly shocked (at 130 to 400 kb levels). There are at least two massive, well-studied finds of iron meteorites that appear to have undergone spontaneous brittle failure prior to entering the atmosphere. They are Gibeon (Bethany) in Southwest Africa and Cape York in Greenland. Nininger (1963) remarks that in the case of Gibeon, more than 50 irons totaling 15 X 109 kg were recovered. The fact that all were strongly ablated and that the scatter ellipse covered an area of several hundred square miles indicated that Gibeon meteorites arrived as a “preatmospheric swarm.” Similarly, the giant Greenland irons (specimens weighed 36, 3.5, 3, and 0.4 x 103 kg) were scattered widely over ~250 km2 and seemed to have traveled as a “swarm” along the same orbit without suffering any further fragmentation upon entering the atmosphere. In contrast, in the large fall of Sikhote-Alin, several thousand fragments resulting from atmospheric fragmentation were scattered within less than 2.5 km2 and exhibited a wide range of sizes from grains to several tons. In both Gibeon, whose mechanical properties suggested to Gordon (1970) the formation in a brittle fracture event, and in Cape York, hydrogen was found at levels of ~7 and ~25.5 ppm, respectively (Edwards, 1955). One could thus assume that, in the presence of internal hydrogen and under repeated stress and intense solar-wind bombardment at ~1 AU perihelion approach, brittle failure of the parent body might have occurred and that the pieces were not dispersed considerably from the common orbit in this gentle type of preatmospheric fragmentation. The time of the fragmentations may be indicated by the fairly long cosmic-ray ages of iron meteorites. This type of seemingly spontaneous splitting of a parent body has been well known to occur in comet nuclei at perihelion approach when unusual stresses on compact nuclei could facilitate unstable cracking. Such fragmentation was observed, for example, in the comets Biela (1826), Olinda (1860), Taylor (1916I), which split in two, and the large comet 1882III, which split into six pieces. In these cases, the centrifugal force about the Sun at perihelion approach and/or the intense irradiation by the solar wind may have aided in disrupting the nucleus (Dauvillier, 1963). The propensity of Eros to disrupt and fragment was also noted during its 1931 close approach to Earth (Dauvillier, 1963). In view of the possibly compact, rigid bodylike nature of some comet nuclei (Whipple, 1963) and of the plausibility of a cometary origin for at least some classes of meteorites (see paper by G. W. Wetherill in this volume”), and in the light of a possible evolution of some comets into asteroidal objects (see paper by B. G. Marsden in this volume”), the splitting of comet nuclei into a few large pieces may be highly suggestive of a mechanism for low-energy fracture.”
It has been shown previously that the mechanical properties of iron meteorites (Gordon, 1970) required their production in brittle fracture breakup events. It was suggested above that the excess hydrogen found in iron meteorites (Edwards, 1955) is likely to be the necessary embrittling agent. A survey of the metallurgy of hydrogen embrittlement (Barth and Steigerwald, 1970; Bernstein, 1970; Groeneveld, Fletcher, and Elsea, 1966; Nelson, Williams, and Tetelman, 1971; Tetelman, 1969) indicated that an iron meteorite parent body could suffer delayed brittle fracture under the action of low rate (accumulated or periodic) stresses whose imprint was found in the metal phase structure (Axon, 1969; Baldanza and Pialli, 1969). Such fracture would occur when the internal hydrogen distribution has reached a critical configuration, which could facilitate rapid propagation of cracks. Two large groups of iron meteorites (Gibeon and Cape York), which are thought to have arrived as preatmospheric swarms, were found to contain sufficient amounts of hydrogen to have been produced in brittle fracture (noncollisional) fragmentation of their parent bodies.
Anders, E. 1964, Origin, Age and Composition of Meteorites. Space Sci. Rev. 3, 583.
*See p. 447.
*See p. 413.
7Note added in press: An alternative mechanism for low-energy fracture of iron meteorite parent bodies has just been proposed by H. L. Marcus and P. W. Palmberg (1971) in J. Geophys. Res. 76, 2095.
Baldanza, H., and Pialli, G. 1969, Dynamically Deformed Structures in Some Meteorites.
Powell, B. N. 1969, Petrology and Chemistry of Mesosiderites—I. Textures and
DOHNANYI: Have you had a chance to examine this problem to see if a critical object size exists beyond which the fragments would tend to stick together? It seems that there should be a critical size where gravity is strong enough to keep the fragments together despite other perturbations. BRECHER: It is difficult to appraise such a critical object size without making very particular assumptions about the size, density, fragmentation spectrum, and orbit of the body, as well as about the mode of disruption and the type of forces (self-gravitation, tidal, solar-wind dynamic pressure, torques, etc.) acting at breakup. Order of magnitude estimates for 10° kg sized pieces of meteoritic iron seem to indicate that they may be kept in contact by mutual gravitational attraction, if the perturbations acting on them are less than ~1 un/kg (~107* dyne/g). Compared to the solar gravitation of 6 m M/kg (0.6 dyne/g) at 1 AU, tidal forces exerted by the Sun and Earth, which are ~10-13 and ~10-1° N/kg-m (~10-1° and ~107* dyne/g-cm), respectively, for 1 AU approach, or ~10-10 and ~10-15 N/kg-m (~10-10 and ~10-15 dyne/g-cm) for 0.1 AU approach, may be neglected. Similarly, rotational instability will not prevail over mutual gravitation of such 103 kg sized chunks, if the rotation period of the body was initially larger than 1 hr. This holds for asteroids, whose spin periods range from 2 to ~10 hr. But the solar-wind dynamic pressure, at the present 1 AU flux of kiloelectron volt protons, would suffice to transfer a momentum of ~5 x 10°g-cm/s per unit area, allowing a body with a 1 m” area to acquire a velocity of ~5 m/s after only 10 million yr of “storage” in a geocentric orbit (Arnold, 1965); if iron meteorites were stored in such an orbit during the “500 million yr that have elapsed since breakup, different surface areas of fragments may have led to considerable scatter velocities. Compare the above to the extremely small initial differential velocities of ~2 x 10-7 cm/s inferred for the presumed members of the Cape York “preatmospheric swarm” from the dimension (~25 km) of their scatter ellipse (Nininger, 1963), if breakup occurred 500 million yr ago. It seems that some cohesive forces (cold welding), or the gentle fracture mode, may have allowed some fragments to hold together in spite of dispersive perturbations. For a 10° kg body, the escape velocity is only ~0.2 cm/s and, indeed, it is hard to see how 10° kg sized chunks could have had smaller scatter velocities at breakup. (I thank Dr. Anders for pointing out this fact.) One could only hope that, just as for asteroidal families assumed to form by collisional breakup, extremely long lifetimes