Thursday 26 March 1998
LHC PROJECT SEMINAR
at 14.15 hrs - LHC Auditorium, bld. 30/7th floor
by Michael Schneider / LHC Division

Dynamic effects scaling law for LHC dipole magnets

In order to reach the nominal LHC luminosity of 10^34 cm^(-2) s^(-1) the field quality during particle injection and the initial acceleration must be accurately controlled. As is well known, high field superconducting magnets have a rich spectrum of non-linear, dynamic field imperfections. In particular two effects have a vital importance for beam stability, namely field drift (Decay) at injection as well as Snapback during the first few seconds of acceleration. From previous experience in Tevatron and HERA, these effects are known to depend on powering history. Because of this we measured the dependence of dynamic effects on operation cycle variations in a small number of one meter long model magnets as well as in one ten meter long model magnet. Based on these results we present two models, which are focused on prediction of Snapback dependence on the operation cycle history. The first model is based on analogy to charging and discharging of an RL circuit during the magnet powering sequences. The second model is given by a multi-layer perceptron (neural network), which we trained to predict snapback by applying the operation cycle description on its input neurons. We discuss the restrictions and capabilities of both prediction techniques.

 

Products | Support | Order | Download | Search | Home

Web Site Design and Implementation Copyright © 1997 NeuroDimension, Inc.
Send questions, comments, and feedback to the
webmaster.