Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!uwmcsd1!nic.MR.NET!umn-cs!ns!logajan
From: logajan@ns.ns.com (John Logajan x3118)
Newsgroups: sci.electronics
Subject: Re: Lightning rods. Was: Protecting computer equipment
Summary: what's behind it all?
Keywords: lightning, lightning rods, discharges, protection
Message-ID: <432@ns.ns.com>
Date: 13 Jul 88 16:06:40 GMT
References: <1988Jul8.045005.17867@utzoo.uucp> <6179@aw.sei.cmu.edu> <1314@kodak.UUCP>
Organization: Network Systems Corp. Mpls MN
Lines: 33

In article <1314@kodak.UUCP>, ornitz@kodak.UUCP (barry ornitz) writes:
> 
> shown that the sharp point is not optimum but instead an elliptical point
> should be used.
> The article also debunks a number of myths about the "cone of protection"
> and also about the neutralization of space charges

I'm interested in this for reasons unrelated to the current discussion.
Do you remember what was the basis for these statements?
There are three general cases for the field around charged surfaces.
1.) a point charge has a field gradient inversely to the distance squared.
2.) a two dimensional (wire) charge has a gradient inversely to the distance.
3.) a three dimentional (plate) charge has a constant field for
    distances << than the diameter of the plate.
Of course at a great enough distance all these cases begin to look like point
charges.

Anyhow, I would guess for a lightning rod you would want as steep a gradient
as you could get.  The combination of the straight part of the rod might tend
to overwhelm the field around the point.  Perhaps that is why an elliptical
point works better. The tip then is further isolated from the more constant
gradient around the straight section.

I don't know about the cone, but there must be some shape of protection around
the rod, i.e. some area where it is easier to go to the rod than to some other
structure.

I suspect that the rods do tend to neutralize the local space charges, but
since there doesn't have to be a local space charge for lightning to hit, that
factor wouldn't do that much good.

- John M. Logajan @ Network Systems; 7600 Boone Ave; Brooklyn Park, MN 55428 -
- {...rutgers!umn-cs, ...amdahl!bungia, ...uunet!rosevax!bungia} !ns!logajan -