arXiv subjects
Publications and source records attributed to Haruo Honjo.
An autoregressive model with a power-law type memory kernel is studied as a stochastic process that exhibits a self-affine-fractal-like behavior for a small time scale. We find numerically that the root-mean-square displacement for the time interval increases with a power law for small time but saturates at sufficiently large time. The exponent changes with the power exponent of the memory kernel.
We present a simple method for incorporating the surface tension effect into an iterative conformal mapping model of two-dimensional diffusion-limited aggregation. A curvature-dependent growth probability is introduced and the curvature is given by utilizing the branch points of a conformal map. The resulting cluster exhibits a crossover from compact to fractal growth. In the fractal growth regime, it is confirmed, by the conformal map technique, that the fractal dimension of its area and perimeter length coincide.
We apply multifractal analysis to an experimentally obtained quasi-two-dimensional crystal with fourfold symmetry, in order to characterize the sidebranch structure of a dendritic pattern. In our analysis, the stem of the dendritic pattern is regarded as a one-dimensional support on which a measure is defined and the measure is identified with the area, perimeter length, and growth rate distributions. It is found that these distributions have multifractality and the results for the area and perimeter length distributions, in the competitive growth regime of sidebranches, are phenomenologically understood as a simple partitioning process.
Scaling structure of the growth rate distribution on the interface of a dendritic pattern is investigated. The distribution is evaluated for an ${\rm NH_4Cl}$ quasi-two-dimensional crystal by numerically solving the Laplace equation with the boundary condition taking account of the surface tension effect. It is found that the distribution has multifractality and the surface tension effect is almost ineffective in the unscreened large growth region. The values of the minimum singular exponent and the fractal dimension are smaller than those for the diffusion-limited aggregation pattern. The Makarov's theorem, the information dimension equals one, and the Turkevich-Scher conjecture between the fractal dimension and the minimum singularity exponent hold.
Three dimensional dendrites are studied with a coupled map lattice model. We study the fractal dimensions, the f(α) spectrum, the size distribution of sidebranches, and the envelope formed by sidebranches.
This study experimentally examined the explosive fragmentation of thin ceramic tubes using pulsed power. A thin ceramic tube was threaded on a thin copper wire, and high voltage was applied to the wire using a pulsed power generator. This melted the wire and the resulting vapor put pressure on the ceramic tube, causing it to fragment. We examined the statistical properties of the fragment mass distribution. The cumulative fragment mass distribution obeyed the double exponential or power-law with exponential decay. Both distributions agreed well with the experimental data. We also found that the weighted mean fragment mass was scaled by the multiplicity. This result was similar to impact fragmentation, except for the crossover point. Finally, we obtained universal scaling for fragmentation, which is applicable to both impact and explosive fragmentation.
We report the binomial multiplicative model for low impact energy fragmentation. Impact fragmentation experiments were performed for low impact energy region, and it was found that the weighted mean mass is scaled by the pseudo control parameter multiplicity. We revealed that the power of this scaling is a non-integer (fractal) value and has a multi-scaling property. This multi-scaling can be interpreted by a binomial multiplicative (simple biased cascade) model. Although the model cannot explain the power-law of fragment-mass cumulative distribution in fully fragmented states, it can produce the multi-scaling exponents that agree with experimental results well.
We performed vertical and horizontal sandwich 2D brittle fragmentation experiments. The weighted mean fragment mass was scaled using the multiplicity $μ$. The scaling exponent crossed over at $\log μ_c \simeq -1.4$. In the small $μ(\llμ_c)$ regime, the binomial multiplicative (BM) model was suitable and the fragment mass distribution obeyed log-normal form. However, in the large $μ(\ggμ_c)$ regime, in which a clear power-law cumulative fragment mass distribution was observed, it was impossible to describe the scaling exponent using the BM model. We also found that the scaling exponent of the cumulative fragment mass distribution depended on the manner of impact (loading conditions): it was 0.5 in the vertical sandwich experiment, and approximately 1.0 in the horizontal sandwich experiment.
We argue differences between the Kardar-Parisi-Zhang with Quenched disorder (KPZQ) and the Ballistic Deposition with Power-law noise (BDP) models, using the multiaffine analysis method. The KPZQ and the BDP models show mono-affinity and multiaffinity, respectively. This difference results from the different distribution types of neighbor-height differences in growth paths. Exponential and power-law distributions are observed in the KPZQ and the BDP, respectively. In addition, we point out the difference of profiles directly, i.e., although the surface profiles of both models and the growth path of the BDP model are rough, the growth path of the KPZQ model is smooth.
We investigated two-dimensional brittle fragmentation with a flat impact experimentally, focusing on the low impact energy region near the fragmentation-critical point. We found that the universality class of fragmentation transition disagreed with that of percolation. However, the weighted mean mass of the fragments could be scaled using the pseudo-control parameter multiplicity. The data for highly fragmented samples included a cumulative fragment mass distribution that clearly obeyed a power-law. The exponent of this power-law was 0.5 and it was independent of sample size. The fragment mass distributions in this regime seemed to collapse into a unified scaling function using weighted mean fragment mass scaling. We also examined the behavior of higher order moments of the fragment mass distributions, and obtained multi-scaling exponents that agreed with those of the simple biased cascade model.