No abstract

Results of simulation testsof the influence of the protective and functional coatings on the resistance of polymeric constructional spacecraft materials to the impact of atomic oxygen with fluences up to 3.5 × × 10 22 cm −2 are presented. It was demonstrated that oxygen plasma beams can be used in accelerated tests of carbon-based and polymeric material structures (with the exception of filled and fluorinated hydrocarbons) to evaluate their resistance to the atomic oxygen impact in low Earth orbit. For unprotected materials, a sharp fall of mechanical properties and a deterioration of optical characteristics were observed. The application of protective coatings is shown to reduce this degradation. Nomenclature dF = fluence increment dM = mass loss E = atomic oxygen translation energy f = normalized energy dependence K = Kapton ® index M = material index Y = erosion yield γ = protection efficiency

About 60 samples of various materials exposed to low Earth orbit (LEO) conditions for 997 days on board the orbital space station ‘Mir’ were investigated. The aim was to determine the properties of the contamination layers formed during LEO exposition. The following methods were used: optical and scanning electron microscopy, local x-ray microanalysis, secondary ion mass spectrometry in fast atom bombardment ionization mode, electron spectroscopy for chemical analysis, temperature programmed desorption mass spectrometry, x-ray phase analysis, spectral reflectance, solar absorbance and relative emittance measurements. Two effects were observed for all samples: the formation of contamination deposits and the erosion of the substrate–original surface. The relative contribution of both effects changes depending on the sample type and on exposure conditions. The deposit thickness varies on the sample surface over a very wide range (at least five decimal orders of magnitude), changing from values exceeding 100 μm to values of less than 2 nm (possibly these regions do not contain any deposit at all). The main element of contamination is silicon; the others are potassium and calcium. Evidence of a chemical reaction between the Teflon FEP substrate and the contamination was observed.

No abstract

We initiate a comprehensive experimental study of objective-based hierarchical clustering methods on massive datasets consisting of deep embedding vectors from computer vision and NLP applications. This includes a large variety of image embedding (ImageNet, ImageNetV2, NaBirds), word embedding (Twitter, Wikipedia), and sentence embedding (SST-2) vectors from several popular recent models (e.g. ResNet, ResNext, Inception V3, SBERT). Our study includes datasets with up to 4.5 million entries with embedding dimensions up to 2048.In order to address the challenge of scaling up hierarchical clustering to such large datasets we propose a new practical hierarchical clustering algorithm B++&C. It gives a 5%/20% improvement on average for the popular Moseley-Wang (MW) / Cohen-Addad et al. (CKMM) objectives (normalized) compared to a wide range of classic methods and recent heuristics. We also introduce a theoretical algorithm B2SAT&C which achieves a 0.74-approximation for the CKMM objective in polynomial time. This is the first substantial improvement over the trivial 2/3-approximation achieved by a random binary tree. Prior to this work, the best poly-time approximation of ≈ 2/3 + 0.0004 was due to Charikar et al. (SODA'19).