Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!cmcl2!lanl!jp
From: jp@lanl.ARPA (James Potter)
Newsgroups: net.physics
Subject: Re: Limitations on High Electric Fields
Message-ID: <1437@lanl.ARPA>
Date: Mon, 24-Feb-86 11:05:16 EST
Article-I.D.: lanl.1437
Posted: Mon Feb 24 11:05:16 1986
Date-Received: Wed, 26-Feb-86 08:21:34 EST
References: <523@mhuxm.UUCP>
Reply-To: jp@a.UUCP (James Potter)
Organization: Los Alamos National Laboratory
Lines: 23



Electric field breakdown in vacuum is of great interest to accelerator builders
Try looking in specialized journals or proceedings of conferences.
Look for van de Graaff stuff for dc, linear accelerators for rf.  As a "rough"
guide for rf high fields there is a number called the Kilpatrick criterion 
that scales roughly with the square root of frequency.  It is based on some
old experiments with oil diffusion pumps and nowadays is considered a
conservative estimate of maximum surface field.  To set a scale, the
Kilpatrick limit at 500 MHz is approximately 20 MV/m.  Designs using twice
that field level are considered practical with proper preparation of surfaces
and with rf conditioning (gotta burn of them high spots!)  Note that this is
a surface field limitation.  There are those who hold, and have data to back
it up that the frequency doesn't matter, only the gap.        
Don't extrapolate my number, above, to many factors of 2 in frequency without
going to the literature, since the square root scaling is only approximate.
I have a colleague who described a field level as b * V-Kilpatrick, where b is
the "bravery" factor.
If you plan to stick to field levels less than 1 Kilpatrick, just have smooth
and clean surfaces, a good, high quality vacuum and bring the fields up
carefully, with some protection to limit hte damage done by the inevitable
sparking.  It doesn't condition if it doesn't breakdown.  Start with short
pulses and exten to longer ones, even CW, as necessary.