Shaping of filamentary streamers by the ambient field

IEEE Trans. Dielect. Electr. Insul.

Hallock

2008

Earlier work by Fowler, Davaney, and Hagedorn showed that the morphology of an anode streamer could be modeled as stochastic growth of a branching fractal tree in point-plane geometry. This investigation reproduces the results of that earlier study. Because one of the concerns about the earlier work is that the electric field dependence appeared to be unphysical, the model was modified to operate under assumptions that are consistent with those that have proven useful in earlier investigations. Specifically, linear electric field dependence was assumed and there is an assumed variability in the number density of available electrons. Computations using this assumption also produce the same range of morphologies that has been measured in experiments. In addition, some assessments of sensitivity to other possible variables are made. First, the sharp cutoff in the electric field strength is replaced with a presumably more realistic exponential dependence on energy. Under this assumption, it is also possible to simulate the experimentally observed behavior of anode streamers. It is shown that three possible refinements to the model have small, and likely negligible, effects. The first is using variable streamer step lengths in the calculation rather than the fixed step length used in the earlier work. The second is to assume growth at one point in the streamer makes growth in other parts somewhat less likely. The third is the assumption that the probability of a streamer making the next step in growth is influenced by the distance of the inter-electrode gap that has already been traversed.

Section: Software Validationsupporting

confidence: 76%

Section: Introductionmentioning

confidence: 99%

Modeling the Growth of Streamers During Liquid Breakdown

IEEE Trans. Dielect. Electr. Insul.

Hallock

2008

“…Two different groups of investigators have explored the growth model of high-speed filamentary streamers during high-voltage breakdown in liquid dielectrics [1,2,6]. The key aspect of the simulation using this model was a routine for the stochastic growth of a fractal pattern generated by the solution to Laplace's equation on large 3-D Cartesian grids.…”

Section: Reproduction Of Earlier Resultsmentioning

confidence: 99%

“…Earlier stochastic Laplacian fractal simulation with power-law (4 th -power to linear) of the field strength and threshold (cutoff) voltage [1,2] modeled the 1 st or 2 nd anode mode [3] streamer as stochastic growth of a branching fractal tree in point-plane geometry. While this was a significant accomplishment, results of recent work [4] suggest that a different set of assumptions should be explored with the same theoretical framework.…”

Section: Introductionmentioning

confidence: 99%

Examination of the influence of streamer growth criteria on morphology

2006 IEEE Conference on Electrical Insulation and Dielectric Phenomena

Hallock

2006

Work by Fowler, Daveny, and Hagedom showed that the morphology of an anode streamer could be modeled as stochastic growth of a branching fractal tree in point-plane geometry. In their work, the experimentally observed range of fractal densities, from sparse to bushy, was modeled by using two assumptions. One was that the growth was driven by E n , where E is the local electric field and n is 1, 2, 3, or 4. The other assumption is that there was a threshold (cutoff) electric field strength for streamer growth in a particular direction. This investigation first reproduced the results of the earlier study to demonstrate that the model and its software implementation yielded the previous results. The model was then modified to operate under a different set of assumptions. In this case, only linear electric field dependence was assumed, and the number density of available electrons was used as a parameter to match the observed data. In addition, rather than assume a sharp cutoff of the threshold field strength, the cutoff was assumed to be a function of electron density. Under these assumptions, it was also possible to simulate the experimentally observed behavior of anode streamers. The current assumptions are consistent with those that have proven to be useful in an earlier investigation.

“…There is a discrepancy, however, between the field fluctuation model and the critical volume model that embeds the stochastic behavior in the production of electrons rather than in electric field variations [10]. While it is likely that the higher density of a liquid than a gas leads to variations in the local dielectric constant, there is not sufficient evidence to determine which stochastic process dominates.…”

Section: Critical Volumementioning

confidence: 88%

The incipient mode of streamers in a dielectric liquid as a function of electric field

Reedy

IEEE International Conference on Dielectric Liquids, 2005. ICDL 2005.

Many previous studies of electric breakdowns in dielectric liquids in point-plane geometry have examined the relationships among the breakdown structure (or speed) and the electrode geometry (point radius, gap length) and/or voltage. This paper explores the hypothesis that the electric field, not the geometry per se, is the determining factor in the incipient streamer structure in dielectric liquids. This work extends earlier work and explores the extent to which the processes that determine initiation and growth are similar. This paper makes it explicit that geometry and voltage are indirect factors in the prebreakdown process, important only insofar as they determine the electric field in the liquid.