From: Arthur W. Brooks
Sent: Monday, September 13, 2004 1:52 PM
To: Wayne T. Reiersen; 'Brad Nelson <nelsonbe@ornl.gov>'; Michael C.
Zarnstorff; 'Williamson, David E.'; 'Freudenberg, Kevin D (KD2)';
'leonard.myatt@myattconsulting.com'
Cc: 'hneilson@pppl.gov'
Subject: Field Errors from Nonlinear Modular Coil Analysis

	

Folks,

Field errors from the nonlinear deformations recently calculated by Kevin for the modular coils have been calculated using the VACISLD Code. The results are summarised below and compared to earlier results found using a linear model from Len Myatt:


					Structural Analysis	
Island Size, % Total Flux	Nonlinear	Linear
    3/5				1.3		2.7
    3/6				1.4		1.0
    3/7				0.6		0.3
  Sum					3.3		4.0

Avg B.n, G				7.4		3.3		
Max B.n, G				54.7		22.1


Both the linear and nonlinear model used the 2.0 T, high beta scenario at t=0.0s from TDS_c08r00_c3 (http://ncsx.pppl.gov/NCSX_Engineering/Requirements/Specs/GRD/Rev1/TDS_XL_C08R00_c3.pdf). The nonlinear model does not include thermal effects, whereas the linear model does include the differential thermal growth from 85K to 105K (actually the displacements from Len's 85K run and 105K run were subtracted). 

Contributions to field errors were calculated for each of the modular coils as data became available from the recent nonlinear analyses (each modular coil was apparently run separately due to the size of the nonlinear models). The results showed significant cancelation of some resonant field errors:

Resonant Component of Bs/Bphi, 1.E-4	
		3,5	3,6
Mod A		 1.00	-0.16
Mod B		-0.78	-0.14
Mod C		-0.12	 0.06
Combined	 0.10	-0.26

The impact on island size was deemed acceptable for the linear model when presented last April. The new predictions of island sizes with the nonlinear model are no worse (although thermal effects could alter this somewhat). There is a significant increase in non resonant error fields which may warrant further evaluation by physics.

Art