Appendix C
Effective trapped fraction for Circular Concentric FS

Evaluation of K(ϵ) if the small ϵ approximation

We can rewrite

               ∫
        --1----  ξ0 ---dξ′0----
I (ξ0) = (1+ ϵ)     λ2,- 1,0(ξ′)
                ξ0T         0
(C.1)

with, for ξ0 > ξ0T ,

      (  )  ∫ 2π dθ  ξ′
λ2,-1,0 ξ′0 =      -----′-
              0  2π Ψξ0
(C.2)

where

      --------------
 ′  ∘       (    ′2)
ξ =   1-  Ψ 1 - ξ0
(C.3)

In the limit ϵ 1, we have

Ψ ---1+-ϵ--
1 + ϵcosθ
= (1 + ϵ)[              ( 2)]
 1 - ϵcosθ + O  ϵ
= 1 + ϵ(1- cosθ ) + O(  )
 ϵ3 (C.4)

so that

ξ = ∘ ----------------------------(------)
  1 - [1+ ϵ(1 - cosθ)+ O (ϵ2)] 1- ξ′02
= ξ0∘ --------------(------)---------
       (1 - cosθ) 1 - ξ′20
  1- ϵ -------ξ′2--------+ O (ϵ2)
               0

In fact, the series expansion of the square root must be kept to all orders for the term in ϵ, because the ratio ϵ∕ξ02 1 for ξ0ξ0T . This gives

       {     ∞                [            (    ′2) ]n        }
ξ′ = ξ′ 1 - ∑  ---(2n---2)!---  ϵ(1--cosθ)--1--ξ0--  +  O (ϵ2)
     0         22n- 1(n- 1)!n!          ξ′20
            n=1

where we used

(1 + x)α = 1 + n=1α-(α---1)⋅⋅⋅(α--n-+-1)
          n!xn (C.5)
(1 + x)12 = 1 + n=11-
2(      )
  1-
  2 - 1⋅⋅⋅(         )
 1-
 2 - n + 1xn-
n! (C.6)
= 1 + 1-
2x + n=2    n-1
(--1)----
  2n(1 ) (3) ⋅⋅⋅(2n - 3) n
 x-
n! (C.7)
= 1 + 1-
2x + n=2(--1)n-1(2n---3)!
22(n- 1)(n - 2)!n!xn (C.8)
= 1 + n=2     n-1
(--1)---(2n---2)!
 22n-1 (n - 1)!n!xn (C.9)

In addition,

-1
Ψ = -----------1-----------
1 + ϵ(1- cosθ )+ O (ϵ2)
= 1 - ϵ(1- cosθ) + O(ϵ2) (C.10)

We get

λ2,-1,0(  )
 ξ′0 = 02π dθ
---
2 π{                                    [                    ] }
                   ∞∑     (2n - 2)!     ϵ(1- cosθ) (1- ξ′2)  n
  1 - ϵ(1- cosθ) -    -2n-1----------  ---------′2------0--
                   n=12     (n- 1)!n!          ξ0 (C.11)
= 02π-dθ
2 π{                 (     ′2)
  1 - 1ϵ-(1---cosθ)-1-+-ξ0-
      2         ξ0′2
   ∞                [            (    ′2)]n}
- ∑  ----(2n---2)!---  ϵ(1--cosθ-)-1--ξ0--
     22n- 1(n- 1)!n!          ξ′20
  n=2 (C.12)
= 1 - n=1λ 2,-1,0(n)(  )
 ξ′0 + O( )
 ϵ2 (C.13)

with

λ2,-1,0(1)( ′)
 ξ0 = 02πdθ
2π-ϵ
2-          (     )
(1- cosθ) 1 + ξ′02
-------ξ′2--------
        0 (C.14)
=   ϵ
--′2
2ξ0(      )
 1 + ξ′20 (C.15)

and, for n 2,

λ2,-1,0(n2)(  )
 ξ′0 = 02πdθ
---
2π  (2n- 2)!ϵn
-2n-1----------
2    (n - 1)!n![                   ]
  (1- cosθ )(1- ξ′2)
  ---------′2-----0--
          ξ0n (C.16)
=   (2n- 2 )!ϵn
-2n-1----------
2    (n - 1)!n![         ]
  (1- ξ′2)
  ----′20--
     ξ0n 02π dθ
---
2 π(1 - cosθ)n (C.17)
=   (2n- 2 )!ϵn
-2n-1----------
2    (n - 1)!n![ (1- ξ′2)]
  ----′20--
     ξ0n 02π dθ
---
2 π[      (θ )]
 2 sin2  --
        2n (C.18)
=           n
-(2n---2)!ϵ----
2n-1(n - 1)!n![(1 - ξ′2)]
 ----′2-0-
    ξ0n 0πdθ-
πsin2nθ (C.19)

We can transform, for n 2,

Y n = 0πdθ-
πsin2nθ
= 0πdθ-
πsinθ sin2n-1θ
= (2n - 1) 0πdθ
---
πcos2θ sin2(n- 1)θ
= (2n - 1)[∫                 ∫          ]
   πd-θsin2(n-1)θ -   π dθsin2θ
  0  π              0  π
= (2n - 1)[Y n-1 - Y n]

so that

      (2n - 1)
Y n = --------Yn- 1
         2n

with

      1
Y 1 = --
      2

so that

Y n = (2n---1)
   2n(2(n---1)--1)-
   2(n - 1)⋅⋅⋅(6---1)
   6(4---1)
   41-
2
=       (2n)!
----------------2
[2n(2n - 2)⋅⋅⋅2]
= --(2n)!-
22n(n!)2

and we find

λ2,-1,0(n2)(  )
 ξ′0 =   (2n )!(2n - 2)!ϵn
-3n--1-----------3
2     (n- 1)!(n!)[         ]
  (1- ξ′2)
  ---′20--
    ξ0n (C.20)
=     [(2n)!]2 ϵn
---------------4
(2n - 1)23n (n!)[        ]
 (1-  ξ′2)
 ----′20--
    ξ0n (C.21)

We obtain

I(ξ )
 0 = ---1---
(1+ ϵ) ξ0Tξ0      ′
---dξ0----
λ2,-1,0(ξ′0)
= ---1---
(1+ ϵ) ξ0Tξ0 0[    ∑∞  [∑∞        (  )]m     (  )]
  1+          λ(n2),-1,0 ξ′0    + O  ϵ2
     m=1  n=1
= ξ0Tξ0 0[        ∑∞  [ ∞∑           ]m         ]
  1- ϵ +         λ(n)  (ξ′)   + O (ϵ2)
         m=1  n=1 2,-1,0  0
= ξ0 -√ --
  2ϵ12 + m=1I(m )(ξ0) + O(ϵ)

with

                  [             ]m
          ∫ ξ0  ′  ∑∞  (n)  ( ′)
I(m)(ξ0) =    dξ0     λ2,-1,0 ξ0
           ξ0T     n=1

where we used

-1---
1- x = 1 + m=1xm (C.22)
ξ0T = ∘  -2ϵ--
   -----
   1+ ϵ = √ --
  2ϵ[    ϵ     (  )]
 1 - --+ O  ϵ2
     2 (C.23)
-1-
ξ0T = √1--
 2ϵ[             ]
 1+  ϵ+  O (ϵ2)
     2 (C.24)

We have

                 [                                   [(      )]n ]m
          ∫ ξ0  ′   ϵ  (    ′2)  ∑∞     [(2n )!]2ϵn      1 - ξ′20
I(m )(ξ0) =     dξ0  2ξ′2- 1+ ξ0  +     ---------3n----4- ---ξ′2---
           ξ0T        0           n=2 (2n- 1) 2  (n!)       0

Clearly, the only term of order ϵ-12 comes from the ξ0′-2 contribution integrated and taken on ξ0T . We need to keep only

I(m)(ξ0) = ξ0Tξ0 0[ ∞                  [   ]n]
 ∑   ---[(2n)!]2ϵn---- -1-
     (2n- 1 )23n(n!)4  ξ′02
 n=1m + O(ϵ)
= ξ0Tξ0 0[        [   ]  ]
 ∑∞  n     1--n
     ϵ Cn  ξ′2
 n=1        0m + O(ϵ)
= ξ0Tξ0 0 i1=1 i2=1⋅⋅⋅ im=1C i1Ci2⋅⋅⋅Cimϵi1+i2+⋅⋅⋅+im [   ]
  1--
  ξ′20i1+i2+⋅⋅⋅+im + O(ϵ)
= i1=1 i2=1⋅⋅⋅ im=1C i1Ci2⋅⋅⋅Cimϵi1+i2+⋅⋅⋅+im ξ0Tξ0 0[ 1 ]
  -′2-
  ξ0i1+i2+⋅⋅⋅+im + O(ϵ)
= i1=1 i2=1⋅⋅⋅ im=1Ci1Ci2 ⋅⋅⋅Cim ϵi1+i2+⋅⋅⋅+im
------------------------
 2(i1 + i2 + ⋅⋅⋅+ im )- 1[                ]
        - 1
 -′2(i1+i2+-⋅⋅⋅+im)-1
 ξ0ξ0Tξ0 + O(ϵ)
= √ --
  2ϵ12 i1=1 i2=1⋅⋅⋅ im=1-----------Ci1Ci2 ⋅⋅⋅Cim-----------
2i1+i2+⋅⋅⋅+im [2(i1 + i2 + ⋅⋅⋅+ im) - 1] + O(ϵ)

with

                2
C  = -----[(2n-)!]-----
 n   (2n - 1)23n (n!)4

and finally we get

           √ --  [    ∑∞  ∑∞  ∑∞     ∑∞                                     ]
I (ξ0) = ξ0- 2ϵ1∕2  1-             ⋅⋅⋅     -----------Ci1Ci2 ⋅⋅⋅Cim----------- +O  (ϵ)
                      m=1 i1=1 i2=1   im=1 2i1+i2+⋅⋅⋅+im [2(i1 + i2 + ⋅⋅⋅+ im) - 1]

-11 0σξ0H(|ξ |- ξ  )
  0    0TI(|ξ |)
  0
= 2 ξ0T1 0ξ02 - ξ 0ϵ12√ --
  2
×[    ∞   ∞   ∞      ∞                                     ]
 1-  ∑  ∑   ∑   ⋅⋅⋅∑   -----------Ci1Ci2 ⋅⋅⋅Cim-----------
                       2i1+i2+⋅⋅⋅+im [2(i1 + i2 + ⋅⋅⋅+ im )- 1]
    m=1 i1=1i2=1    im=1 + O(ϵ)
= 2-
3 - ϵ12√ --
  2[                                                         ]
     ∑∞  ∑∞  ∞∑     ∑∞  ------------Ci1Ci2-⋅⋅⋅Cim-----------
 1 -            ⋅⋅⋅    2i1+i2+ ⋅⋅⋅+im [2 (i1 + i2 + ⋅⋅⋅+ im)- 1]
     m=1 i1=1i2=1   im=1 + O(ϵ)

and therefore

|-----------------------------------------------------------------------------------|
|        √ --[    ∞∑  ∑∞  ∑∞     ∑∞                                     ]     (   )  |
K  (ϵ) = 3--2  1-             ⋅⋅⋅    -----------Ci1Ci2 ⋅⋅⋅Cim-----------  + O  ϵ1∕2   |
|         2      m=1 i1=1i2=1    im=1 2i1+i2+⋅⋅⋅+im [2(i1 + i2 + ⋅⋅⋅+ im )- 1]          |
-------------------------------------------------------------------------------------
(C.25)

with

|----------------------|
|                2     |
|Cn = -----[(2n-)!]----4 |
------(2n---1)23n-(n!)--

For example, C1 = 12, C2 = 316, C3 = 532, ⋅⋅⋅

m
i1
i2
correction
1
1
-
14
1
2
-
164
1
3
-
1256
1
4
-
2516384
1
5
-
1722737
2
1
1
148
2
1
2
31280 × 2
2
2
2
928672
2
1
3
57168 × 2
2
2
3
549152 × 2
2
3
3
6173015

The sum of all these coefficient gives K = 1.486, which is already much better that 9√ --
  28 = 1.591


PIC


Figure C.1: Bootstrap current coefficient κ as a function of the highest terms M and N kept in the series.


We compute the coefficient K for several values of the highest terms M and N in the series (C.25). The results are shown on Fig. C.1. We first see that K converges for high M and N, towards a value that seems close to 1.46. The best combination seems to be for M = N, as both high M and high N are needed for convergence. Therefore, we take M = N and calculate K as a function of M. Results are shown on Fig. C.2. We can see that K indeed converges, and that the asymptotic limit is about K = 1.46. For M = N = 6, we find K = 1.467.


PIC


Figure C.2: Bootstrap current coefficient κ as a function of the highest terms M and N kept in the series, for M = N