function [B66, M, rout] = findthinmpoleraddiffm(rin, PolynomA, PolynomB, L, irho, E0, max_order) %FINDTHINMPOLERADDIFFM % Physical constants used in calculations persistent TWOPI CGAMMA M0C2 LAMBDABAR CER CU if isempty(TWOPI) %Initialize constansts on the first call TWOPI = 2*pi; CGAMMA = 8.846056192e-05; % [m]/[GeV^3] Ref[1] (4.1) M0C2 = 5.10999060e5; % Electron rest mass [eV] LAMBDABAR = 3.86159323e-13; % Compton wavelength/2pi [m] CER = 2.81794092e-15; % Classical electron radius [m] CU = 1.323094366892892; % 55/(24*sqrt(3)) end % Calculate field from polynomial coefficients P1 = i*PolynomA(1:max_order+1)+PolynomB(1:max_order+1); Z1 = cumprod([1, (rin(1)+i*rin(3))*ones(1,max_order)]); S1 = sum(P1.*Z1); Bx = real(S1); By = imag(S1); B2P = B2perp([Bx; By+irho; 0], irho, rin); B3P = B2P^(3/2); p_norm = 1/(1+rin(5)); p_norm2 = p_norm^2; CRAD = CGAMMA*E0^3/(TWOPI*1e27); BB = CU * CER * LAMBDABAR * (E0/M0C2)^5 * L * B3P * (1+rin(5))^4*... (1+rin(1)*irho + (rin(2)^2+rin(4)^2)*p_norm2/2); % Propagate particle rout = rin; % Loss of energy (dp/p) due to radiation rout(5) = rin(5) - CRAD*(1+rin(5))^2*B2P*... (1+rin(1)*irho + (rin(1)^2+rin(3)^2)*p_norm2/2)*L; % Change in transverse momentum due to radiation % Angle does not change but dp/p changes due to radiation % and therefore transverse canonical momentum changes % px = x'*(1+dp/p) % py = y'*(1+dp/p) rout([2 4]) = rin([2 4])*(1+rout(5))/(1+rin(5)); % transverse kick due to magnetic field rout(2) = rout(2) - L*(Bx-(rin(5)-rin(1)*irho)*irho); rout(4) = rout(4) + L*By; % pathlength rout(6) = rout(6) + L*irho*rin(1); % Calculate transfer matrix at rin P2 = i*PolynomA(2:max_order+1)+PolynomB(2:max_order+1); Z2 = cumprod([1, (rin(1)+i*rin(3))*ones(1,max_order-1)]); S2 = sum(P2.*(1:max_order).*Z2); M = eye(6); M(2,1) = -L*real(S2); M(2,3) = L*imag(S2); M(4,1) = L*imag(S2); M(4,3) = L*real(S2); M(2,5) = L*irho; M(2,1) = M(2,1) - L*irho*irho; M(6,1) = L*irho; % Calculate Ohmi's diffusion matrix of a thin multipole element % For elements with straight coordinate system irho = 0 % For curved elements the B polynomial (PolynomB in MATLAB) % MUST NOT include the guide field By0 = irho * E0 /(c*e) B66 = zeros(6); B66(2,2) = BB*rin(2)^2*p_norm2; B66(2,4) = BB*rin(2)*rin(4)*p_norm2; B66(4,2) = B66(2,4); B66(4,4) = BB*rin(4)^2*p_norm2; B66(5,2) = BB*rin(2)*p_norm; B66(2,5) = B66(5,2); B66(5,4) = BB*rin(4)*p_norm; B66(4,5) = B66(5,4); B66(5,5) = BB; function b2 = B2perp(B, irho, rin) % Calculates sqr(|e x B|) , where 'e' is a unit vector in the direction of % velocity. Components of the velocity vector: % ex = xpr; % ey = ypr; % ez = (1+x*irho); E = [rin(2)/(1+rin(5));rin(4)/(1+rin(5));1+rin(1)*irho]; b2 = sum(cross(E/norm(E),B).^2);