More fixer devel. (bmad-sim#1804)

* More fixer devel.

Author: DavidSagan

bmad/code/closed_orbit_calc.f90

@@ -52,7 +52,7 @@
 !                          the energy adjusted so that vec(5) is the same 
 !                          at the beginning and at the end.
 !                     = 6 True closed orbit.
-!                     Default: Use 4 or 6 depending upon if RF is on or off.
+!                     Default: 4 if RF is off, 6 if RF is on.
 !   direction      -- Integer, optional: Direction of tracking. 
 !                       +1 --> forwad (default), -1 --> backward.
 !                       The closed orbit will be dependent on direction only

bmad/custom/make_mat6_custom.f90

@@ -14,7 +14,8 @@
 !
 ! Output:
 !   ele       -- Ele_struct: Element with transfer matrix.
-!     %mat6     -- 6x6 transfer matrix.
+!     %mat6(6,6)     -- 6x6 transfer matrix (first order part of the transport map).
+!     %vec6(6)       -- zeroth order part of the transport map.
 !   end_orb   -- Coord_struct: Coordinates at the end of element.
 !   err_flag  -- Logical: Set true if there is an error. False otherwise.
 !+

bmad/doc/elements.tex

@@ -2116,11 +2116,15 @@ \section{Fiducial}
 \section{Fixer}
 \label{s:fixer}
 
-<b>This element is in construction and is not yet usable.</b>
+{\bf This element is in construction. Status: Fixer elements can be used but testing is not yet complete.}
 
-A \vn{fixer} element is used to set the particle orbit and/or the Twiss parameters at any point in
-an open geometry lattice branch. The orbit and/or Twiss parameters can then be propagated from the
-fixer element both forward and backwards to the ends of the branch.
+A \vn{fixer} element is used to set the particle orbit and Twiss parameters at any point in an open
+geometry lattice branch. The orbit and Twiss parameters can then be propagated from the fixer
+element both forward and backwards to the ends of the branch. \vn{Fixe} elements are convenient, for
+example, when it is desired to design a beam line with Twiss and/or orbit parameters that are fixed
+at some point that is not the beginning of the line. Another example is when the orbit or Twiss parameters
+have been measured at some point in a beam line and a calculation of the orbit or Twiss parameters at
+other places in the line is desired consistent with the measured values.
 
 Note: For the purposes of this discussion, the term ``Twiss'' parameters include coupling and
 dispersion parameters and ``orbit'' includes both phase space and spin coordinates.
@@ -2171,24 +2175,24 @@ \section{Fixer}
 If the geometry of a branch is \vn{closed} (\sref{s:param}), \vn{fixer} elements have no effect.
 Thus, in the following discussion, it is assumed that the branch geometry is set to \vn{open}.
 Additionally, for this discussion, the term ``\vn{fixer}'' element will include \vn{beginning_ele}
-elements (\sref{s:begin.ele}) since, for the purposes of setting orbit and/or Twiss parameters
-\vn{beginning_ele} elements essentially act as a fixer element at the beginning of a branch.
-
-A fixer element is ``active'' if its \vn{is_on} parameter (\sref{s:is.on}) is True. One and only one
-fixer element in a branch can be active at a given time. The \bmad bookkeeping routines will ensure
-that this rule is enforced and turning on a given fixer will result in all others being turned off.
-Turning off the active fixer will result in the \vn{beginning_ele} element being turned on. The
-default for a fixer element defined in a lattice file is to be non-active except for the
-\vn{beginning_ele} which defaults to being active. When a lattice is read in, for a given branch,
-the last fixer that is set active will be the active fixer and if non are set the \vn{beginning_ele}
-will be active.
-
-Essentially, a fixer elements has two classes of Twiss and orbit parameters. There are the
-``stored'' parameters listed above. These parameters may be set by the User. The other class are the
-``actual'' parameters which cannot be set by the User. For a \vn{fixer} element that is active, the
-actual parameters are set equal to the corresponding stored parameters and these actual parameters
-are then used by \bmad to propagate the Twiss and orbit backwards and forwards through the rest of
-the branch. For fixer elements that are not active, the actual Twiss and orbit is calculated in the
+elements (\sref{s:begin.ele}) since, for the purposes of setting orbit and/or Twiss parameters,
+\vn{beginning_ele} elements essentially act the same as a fixer element.
+
+A fixer element is ``active'' if the value of its \vn{is_on} parameter (\sref{s:is.on}) is True. One
+and only one fixer element in a branch can be active at any given time. The \bmad bookkeeping
+routines will ensure that this rule is enforced and turning on a given fixer will result in all
+others being turned off.  Turning off the active fixer will result in the \vn{beginning_ele} element
+being turned on. The default for a fixer element defined in a lattice file is to be non-active
+except for the \vn{beginning_ele} which defaults to being active. When a lattice is read in, for a
+given branch, the last fixer that is set active will be the active fixer and if no \vn{fixer}
+elements are set active, the \vn{beginning_ele} will be active.
+
+A fixer elements has two classes of Twiss and orbit parameters. There are the ``stored'' parameters
+listed above. These parameters may be set in a lattice file. The other class are the ``actual''
+parameters which cannot be set by the User. For a \vn{fixer} element that is active, the actual
+parameters are set equal to the corresponding stored parameters and these actual parameters are then
+used by \bmad to propagate the Twiss and orbit backwards and forwards through the rest of the
+branch. For fixer elements that are not active, the actual Twiss and orbit is calculated in the
 standard fashion by propagating the Twiss and orbit from the active fixer. The stored Twiss and
 orbit parameter values in a non-active fixer do not affect the Twiss and orbit calculation.
 

bmad/modules/fixer_mod.f90

@@ -48,6 +48,11 @@ subroutine set_active_fixer(fixer, is_on, orbit)
   return
 endif
 
+if (fixer%value(beta_a_stored$) <= 0 .or. fixer%value(beta_b_stored$) <= 0) then
+  call out_io(s_error$, r_name, 'Fixer element does not have beta_a or beta_b set!. Active fixer not set.')
+  return
+endif
+
 branch => fixer%branch
 on = logic_option(.true., is_on)
 

bmad/modules/radiation_mod.f90

@@ -72,7 +72,7 @@ subroutine track1_radiation (orbit, ele, edge)
 
 real(rp) int_gx, int_gy, this_ran, mc2, int_g2, int_g3, ran6(6)
 real(rp) gamma_0, dE_p, fact_d, fact_f, p_spin, spin_norm(3), norm, rel_p, c_radius
-real(rp) damp_mat(6,6), stoc_mat(6,6), ref_orb(6)
+real(rp) damp_mat(6,6), stoc_mat(6,6), ref_orb(6), rr
 real(rp), parameter :: rad_fluct_const = 55.0_rp * h_bar_planck * c_light / (24.0_rp * sqrt_3)
 real(rp), parameter :: spin_const = 5.0_rp * sqrt_3 * h_bar_planck * c_light / 16.0_rp
 real(rp), parameter :: damp_const = 2.0_rp / 3.0_rp
@@ -109,8 +109,9 @@ subroutine track1_radiation (orbit, ele, edge)
 endif
 
 if (bmad_com%radiation_damping_on) then
-  orbit%vec = orbit%vec + bmad_com%synch_rad_scale * matmul(rad_map%damp_dmat, orbit%vec - rad_map%ref_orb)
-  if (.not. bmad_com%radiation_zero_average) orbit%vec = orbit%vec + bmad_com%synch_rad_scale * rad_map%xfer_damp_vec
+  rr = orbit%time_dir * bmad_com%synch_rad_scale 
+  orbit%vec = orbit%vec + rr * matmul(rad_map%damp_dmat, orbit%vec - rad_map%ref_orb)
+  if (.not. bmad_com%radiation_zero_average) orbit%vec = orbit%vec + rr * rad_map%xfer_damp_vec
 endif
 
 if (bmad_com%radiation_fluctuations_on) then

bsim/code/adjust_optics.f90

@@ -2,9 +2,11 @@
 
 parameter[geometry] = Open
 
-parameter[p0c]                      =  1.0E6*1e6
+parameter[p0c]                      =  1.0E10
 parameter[particle]                 = Positron
 bmad_com[spin_tracking_on] = T
+bmad_com[radiation_damping_on] = T
+bmad_com[spin_tracking_on] = T
 
 beginning[beta_a]       =  1.1E1
 beginning[alpha_a]      = -1.7
@@ -42,7 +44,7 @@ particle_start[spin_z] = 0.5
 S1: Sextupole, L =  1.0E-2
 D: Drift, L =  2.9E-1
 Q1: Quadrupole, L =  5.0E-1, TILT =  5.0E-2, K1 =  5.9E-1, hkick = 1e-3*f, vkick = 2e-3*f
-B: SBend, L =  3.0, G =  1.3E-1, E1 =  7.0E-2, E2 =  7.0E-2
+B: SBend, L = 3.0, G =  1.3E-1, E2 = 7.0E-1
 Q2: Quadrupole, L =  5.0E-1, K1 = -5.0E-1
 S2: Sextupole, L =  1.0E-2
 Lc: LCavity, L =  1.1, VOLTAGE =  1.0E6*1e-10, phi0 = 0.1,
@@ -58,6 +60,6 @@ F2: Fixer
 ! Lattice lines
 
 
-RING: line = ( S1, D, Q1, D, Lc, B, D, Q2, D, F1, S2, Lc, F2, B)
+RING: line = (S1, D, Q1, D, Lc, B, D, Q2, D, F1, S2, Lc, F2, B)
 
 use, RING

regression_tests/fixer_test/fixer_test.bmad

@@ -154,14 +154,17 @@ subroutine orbit_delta_write(indx, ix_ele, orb1, orb2)
 
 str = run_str(indx) // '-E' // int_str(ix_ele) // '-Orbit-'
 
-write (1, '(2a, es16.8)') quote(trim(str) // 'X'),  ' ABS 1E-10', orb1%vec(1) - orb2%vec(1)
-write (1, '(2a, es16.8)') quote(trim(str) // 'PX'), ' ABS 1E-10', orb1%vec(2) - orb2%vec(2)
-write (1, '(2a, es16.8)') quote(trim(str) // 'Y'), ' ABS 1E-10', orb1%vec(3) - orb2%vec(3)
-write (1, '(2a, es16.8)') quote(trim(str) // 'PY'), ' ABS 1E-10', orb1%vec(4) - orb2%vec(4)
-write (1, '(2a, es16.8)') quote(trim(str) // 'Z'), ' ABS 1E-10', orb1%vec(5) - orb2%vec(5)
-write (1, '(2a, es16.8)') quote(trim(str) // 'PZ'), ' ABS 1E-10', orb1%vec(6) - orb2%vec(6)
-write (1, '(2a, es16.8)') quote(trim(str) // 'T'), ' ABS 1E-10', orb1%t - orb2%t
+write (1, '(2a, es16.8)') quote(trim(str) // 'X'),   ' ABS 1E-10', orb1%vec(1) - orb2%vec(1)
+write (1, '(2a, es16.8)') quote(trim(str) // 'PX'),  ' ABS 1E-10', orb1%vec(2) - orb2%vec(2)
+write (1, '(2a, es16.8)') quote(trim(str) // 'Y'),   ' ABS 1E-10', orb1%vec(3) - orb2%vec(3)
+write (1, '(2a, es16.8)') quote(trim(str) // 'PY'),  ' ABS 1E-10', orb1%vec(4) - orb2%vec(4)
+write (1, '(2a, es16.8)') quote(trim(str) // 'Z'),   ' ABS 1E-10', orb1%vec(5) - orb2%vec(5)
+write (1, '(2a, es16.8)') quote(trim(str) // 'PZ'),  ' ABS 1E-10', orb1%vec(6) - orb2%vec(6)
+write (1, '(2a, es16.8)') quote(trim(str) // 'T'),   ' ABS 1E-10', orb1%t - orb2%t
 write (1, '(2a, es16.8)') quote(trim(str) // 'P0C'), ' ABS 1E-10', orb1%p0c - orb2%p0c
+write (1, '(2a, es16.8)') quote(trim(str) // 'Spin-X'), ' ABS 1E-10', orb1%spin(1) - orb2%spin(1)
+write (1, '(2a, es16.8)') quote(trim(str) // 'Spin-Y'), ' ABS 1E-10', orb1%spin(2) - orb2%spin(2)
+write (1, '(2a, es16.8)') quote(trim(str) // 'Spin-Z'), ' ABS 1E-10', orb1%spin(3) - orb2%spin(3)
 
 end subroutine orbit_delta_write