Project Details
Projekt Print View

True (3D) chondrule size distributions: Basic data for the reconstruction of chondrule formation and asteroid accretion

Applicant Dr. Knut Metzler
Subject Area Mineralogy, Petrology and Geochemistry
Term from 2016 to 2020
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 318792451
 
Final Report Year 2021

Final Report Abstract

The outcome of this project is the first comprehensive and self-consistent data set for the true (3D) chondrule size-frequency distributions in unequilibrated samples of the 11 major chondrite groups (H, L, LL, R, EH, EL, CM, CO, CV, CR, CK). Two different meteorites of each group have been investigated by micro computed X-ray tomography (µ-CT) and by this the 3D sizes of 5065 chondrules have been determined. The 3D data sets are complemented by the distributions of apparent chondrule sizes (2D), which have been measured on BSE images of polished sections. This initially unplanned work led to the additional determination of the 2D size of 6360 chondrules. The long-standing question whether the frequency distributions of 2D chondrule sizes are shifted towards lower or higher values compared to the true (3D) sizes has been resolved by this study. It turned out that 2D sizes in all chondrite groups are always shifted towards lower values to various extents. This important observation has been already published for the case of ordinary chondrites. This empirical result is the opposite of what has been calculated and predicted by other authors, but has been recently confirmed and explained by mathematical modelling in cooperation with Dr. Dominik Hezel, University of Frankfurt. In all groups the mean chondrule sizes in the sample pairs are similar to each other. The largest size variance was observed for the LL chondrites and the CR chondrites, confirmed by the corresponding 3D sizes. Due to the lack of 3D literature data, mainly the 2D findings can be compared to the literature. The literature values of 2D mean chondrule sizes of L, LL, EH, CM, and CK chondrites could be basically confirmed. Considerable differences were found for the H, R, EL, CO, and CR chondrites. In all these cases, but CO chondrites, the 2D values of this study are smaller than the literature values. The new values for CO chondrites (210-240 µm) are considerably larger than the values in the literature (150 µm). This is of importance regarding the CM-CO relationship, since in contrast to the recommended different mean values (CM: 270 µm; CO: 150 µm) the new values for both groups are nearly indistinguishable. This finding appears to be in line with conclusions by, e.g., Torrano et al. (2021), who conclude that COs and CMs are closely related. The new 2D size data for chondrules in both investigated CV chondrites are strikingly different to literature values. While a mean size of 900 µm is given in Friedrich et al. (2015), the new 2D data for CV chondrites are only 420 µm (Vigarano, CVred) and 470 µm for Allende (CVox-A). An important aspect of this finding is that the mean size of chondrules in CV chondrites is less than half the mean size for CK chondrules. This is in strong contrast to literature values, where the same mean 2D size (900 µm) for CV and CK chondrites is given. Furthermore, the size-frequency distributions of both groups are quite different. The new data indicate that both groups are probably not closely related, as frequently proposed in the literature (e.g., Greenwood et al., 2010). From the shape of their distribution curves it becomes clear that a formation of CK chondrites from CV chondrites (and vice versa) by parent body processes can be excluded. An important finding is the positive correlation between the mean chondrule size and the width of the distribution curve. This is observed both in C chondrites and non-C chondrites. If sorting was responsible for the distribution shapes, then the sorting process was more / less efficient in case of chondrite groups with small / large mean chondrule sizes. Another important finding are the positive correlations between mean and minimum chondrule size, as well as mean and maximum chondrule size. If sorting was responsible for this correlation, it appears that this process seems to have cut out a restricted size range (which finally accreted) from a potentially preexisting chondrule population with a larger size range. The positive correlation between 2D chondrule diameters and the thickness of a corresponding dust rim (mantle) in CM chondrites could be confirmed by the 3D measurements. It turned out that in both investigated CM chondrites the dust rim thickness is, on average, ~0.15 times the radius of the corresponding chondrule. The dust mantled objects consist, on average, of ~1/3 chondrule material and 2/3 dust material by volume. The exact volume fractions are remarkably similar in both samples, i.e., 67.1% and 68.2 %. Obviously, the accretion of dust onto chondrule surfaces in both samples was the result of the same physical process. In the course of this study, the true (3D) chondrule size-frequency distributions of cluster chondrites have been determined for the first time. The diameter of 539 chondrules was measured in 2 cluster chondrite clasts of NWA 5205 (LL3.7). The chondrule size-frequency distributions in these samples are distinctly more symmetric than those in other OCs. I found a co-enrichment of chondrule types with a priori small mean sizes (type I, porphyritic) in clasts with overall small mean chondrule sizes. This is considered as the fingerprint of an additional / second size-sorting process, which acted later on these chondrule populations. Finally, five ungrouped C chondrites with very similar chemical and petrologic properties (e.g., chondrule size-frequency distributions) were investigated. This led to the definition and establishment of the new Loongana (CL) group of carbonaceous chondrites.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung