[Sneap] Information request concerning High pressure safety system in electrostatic accelerators

[Doug Will]

Augusto,

I do not know of a risk analysis but have historical evidence that the 
risk is real.  The original Van de Graaff system at the Nuclear Physics 
Lab of the Univ of Washington consisted of two FN Van de Graaffs in 
series: an injector FN with a neutral "ion" source at its low energy 
(LE) end and negative terminal (containing an alkali metal e- adder 
canal); followed by a tandem FN with positive terminal (this latter, 
still in use).

On Sept 20th, 1969, a Septier lens failed in the terminal of the 
injector tandem.  Large, low pressure relief flanges on both ends of the 
injector FN opened and protected valuable ion sources situated at both 
ends.  This injector FN also had a pumping tube parallel to the "low 
energy" neutral beam drift tube leading from the neutral "ion" source at 
the LE end up to the terminal.  This pumping tube was similar to a beam 
tube but with larger openings in the metal electrodes. Outside the tank 
at the low energy end was a large (2 meter tall) diffusion pump attached 
by a ~30 cm diameter bellows to the pumping tube.  There was NO PRESSURE 
RELIEF FLANGE on this pumping station.  When the Septier lens ruptured 
inside the tank, the large bellows expanded  to its limit (~1 meter)
and tore completely apart from one of its end flanges, toppling the 
diffusion pump stand over on the floor.  Most of the tank gas then 
vented through this ruptured bellows end (with quite a roar).

The underlying cause was poor design. Careful analysis showed the glass 
insulators of the Septier lens were under about 1000 pounds (4400 
Newtons) of tension.  The kovar seal at one end of the Septier lens 
separated over most of its circumference from both its stainless steel 
flange and from its glass insulator.

While I was not at the Nuclear Physics Lab in 1969, I have in my 
possession our 1970 annual report containing a five page report
(p. 14ff) of the incident written by several of my mentors, C.E.Linder,
F.H.Schmidt, and W.G.Weitkamp.  If there is interest, I can scan that 
report and attach it to a direct email reply.  Interested parties should 
send a request including your email address.

Doug Will
Research Engineer, U of Washington
Center for Experimental Nuclear Physics and Astrophysics
diwill at u.washington.edu

Chris Westerfeldt wrote:
> 
> 
> 
> ------------------------------------------------------------------------
> 
> Greetings Augusto,
>     We also built our own gravity/spring sealed “blow off valves” and 
> installed them at the exit of our
> K-3000 and also both ends of our FN.  We feel that a catastrophic 
> failure of accelerator tube is unlikely but
> it is possible to develop a leak at a failing glue joint in a tube – or 
> more likely at the rf ion source in a single-ended machine
> that would trip the vacuum pump interlocks thereby causing the pump gate 
> valve to close.  These valves would limit the fairly
> slow pressure build up in the beamline to a couple of psi.  We made the 
> valve plates as large as possible toward this end.
>  Regards,
> - Chris
> -- 
> 
> Chris R. Westerfeldt
> Research Scientist / T.U.N.L. Radiation Safety Manager
> Duke University Physics Department &
> Triangle Universities Nuclear Laboratory
> Science Drive, Box 90308
> Durham, NC  27708-0308
> Tel: (919) 660-2600        
> Fax: (919) 660-2634
> Email: Cwest at Tunl.Duke.Edu
> 
> 
> ------------------------------------------------------------------------
> *From: *Augusto Lombardi <augusto.lombardi at lnl.infn.it>
> *Reply-To: *Symposium of Northeastern Accelerator Personnel 
> <sneap at tunl.duke.edu>
> *Date: *Wed, 18 Mar 2009 09:40:40 +0100
> *To: *<sneap at tunl.duke.edu>
> *Subject: *[Sneap] Information request concerning High pressure safety 
> system in electrostatic accelerators
> 
> 
> 
> 
> ------------------------------------------------------------------------
>     Dear Sneapers,
> 
> I am Augusto Lombardi from Legnaro and some of you know me already.
> I am now responsible for the operation of the two single stage machine 
> in Legnaro the AN 2000 and the CN 7 MeV.
> 
> The questions I have for the community is:
> 
>    1.     There are any risk analysis and/or records concerning the high
>       pressure hazard due to the column breakdown or any other accident
>       that put in comunication the high pressure vessel with the beam
>       vacuum pipe? (more than 10 bar vessels versus vacuum)
>    2.
>    3.     There are any safety systems for this kind of accident?  (we
>       do have break disks and spring loaded valves but ....)
> 
> 
>     Thank you all for the help.
>     greetings
>     Augusto.
> 
> ------------------------------------------------------------------------
> _______________________________________________
> Sneap mailing list
> Sneap at daytona.tunl.duke.edu
> http://daytona.tunl.duke.edu/mailman/listinfo/sneap
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Sneap mailing list
> Sneap at daytona.tunl.duke.edu
> http://daytona.tunl.duke.edu/mailman/listinfo/sneap