This work demonstrates that an artificial scaffold structure can be designed to exhibit mechanical properties close to the ones of real bone tissue, thus highly reducing the stress-shielding phenomenon. In this study the scan of lumbar vertebra fragment was reproduced to create a numerical 3D model (this model was called the reference bone sample). New nine 3D scaffold samples were designed and their numerical models were created. Using the finite element analysis, a static compression test was performed to assess the effective Young modulus of each tested sample. Also, two important metrics of each sample were assessed: relative density and surface area. Each new designed 3D scaffold sample was analyzed by considering two types of material properties: metal alloy properties (Ti-6Al-4V) and ABS polymer properties. Numerical analysis results of this study confirm that 3D scaffold used to design a periodic structure, either based on interconnected beams (A, B, C, D, E and F units) or made by removing regular shapes from base solid cubes (G, H, I units), can be refined to obtain mechanical properties similar to the ones of trabecular bone tissue. Experimental validation was performed on seven scaffolds (A, B, C, D, E, F and H units) printed from ABS material without any support materials by using Fused Deposition Modeling (FMD) technology. Results of experimental Young modulus of each printed scaffold are also presented and discussed.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
In this paper, we present a novel approach to calculation of discrete wavelet transform (DWT) on modern Graphics Processing Units (GPUs) with CUDA architecture which takes advantage of highly parallel lattice structure. The experimental results obtained for model signals show that the proposed approach allows to obtain several times acceleration compared with sequential calculations carried out on the CPU while taking into account not only time of calculations but also time required for data transfers.
XX
W artykule zaproponowano nowe podejście do obliczania dyskretnego przekształcenia falkowego (DWT) na nowoczesnych procesorach graficznych (GPUs) o architekturze CUDA oparte o wysoce równoległe struktury kratowe. Wyniki badań eksperymentalnych przeprowadzonych na sygnałach modelowych pokazują, że zaproponowane podejście daje możliwość uzyskania kilkukrotnego przyśpieszenia obliczeń w porównaniu do obliczeń sekwencyjnych na CPU, biorąc pod uwagę nie tylko czas obliczeń , ale również czasy przesyłu danych.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.