@@ -73,12 +73,13 @@ and then form the new approximation by
7373Heun's method attains second-order accuracy by combining predictor and corrector slopes:
7474
7575.. math ::
76- :align: left
7776
78- k_1 = \sigma \,\mathbf {u}(\mathbf {x}_n,t_n),\\
79- \mathbf {x}^* = \mathbf {x}_n + \Delta t\, k_1 ,\\
80- k_2 = \sigma \,\mathbf {u}(\mathbf {x}^*,t_n + \Delta t),\\
81- \mathbf {x}_{n+1 } = \mathbf {x}_n + \tfrac {\Delta t}{2 }\,(k_1 + k_2 ).
77+ \begin {aligned}
78+ k_1 &= \sigma \,\mathbf {u}(\mathbf {x}_n,t_n),\\
79+ \mathbf {x}^* &= \mathbf {x}_n + \Delta t\, k_1 ,\\
80+ k_2 &= \sigma \,\mathbf {u}(\mathbf {x}^*,t_n + \Delta t),\\
81+ \mathbf {x}_{n+1 } &= \mathbf {x}_n + \tfrac {\Delta t}{2 }\,(k_1 + k_2 ).
82+ \end {aligned}
8283
8384 This scheme yields a global error of order :math: `O(\Delta t^2 )` with two velocity evaluations per step.
8485
@@ -87,7 +88,6 @@ This scheme yields a global error of order :math:`O(\Delta t^2)` with two veloci
8788The classical RK4 method achieves fourth-order accuracy via four slope evaluations at intermediate points:
8889
8990.. math ::
90- :align: left
9191
9292 k_1 = \mathbf {u}(\mathbf {x}_n,t_n),\\
9393 k_2 = \mathbf {u}\!\bigl (\mathbf {x}_n + \tfrac {\Delta t}{2 }k_1 ,\; t_n + \tfrac {\Delta t}{2 }\bigr ),\\
@@ -122,7 +122,7 @@ As for the coefficients for ``RK6`` are more complex to write into equations, th
122122| :math: `b_i` | :math: `1 /12 ` | :math: `0 ` | :math: `0 ` | :math: `0 ` | :math: `5 /12 ` | :math: `5 /12 ` | :math: `1 /12 ` |
123123+-----------------------------+-------------------------------+----------------------------+---------------------------------+------------------------------+-----------------------------+--------------------------+-------------------------+
124124
125- In our computation, the up symbol side is applied, in other words, ``± `` represents ``+ ``, taking :math: `\lambda =+\sqrt {5 }`. With 15 digis are kept, the Butcher table used by the author is shown in the following table.
125+ In our computation, the up symbol side is applied, in other words, ``± `` represents ``+ ``, taking :math: `\lambda =+\sqrt {5 }`. With 15 digis are kept, the explicit Butcher table for `` RK6 `` used by the author is shown in the following table.
126126
127127+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+
128128| :math: `c_i` | :math: `a_{i1 }` | :math: `a_{i2 }` | :math: `a_{i3 }` | :math: `a_{i4 }` | :math: `a_{i5 }` | :math: `a_{i6 }` | :math: `a_{i7 }` |
@@ -141,7 +141,7 @@ In our computation, the up symbol side is applied, in other words, ``±`` repres
141141+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+
142142| 1 | 0.166666666666667 | 0 | 0.0751416197912285 | -3.38770632020821 | 0.52786404500042 | 3.61803398874989 | 0 |
143143+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+
144- | b_i | 0.0833333333333333 | 0 | 0 | 0 | 0.416666666666667 | 0.416666666666667 | 0.0833333333333333 |
144+ | :math: ` b_i` | 0.0833333333333333 | 0 | 0 | 0 | 0.416666666666667 | 0.416666666666667 | 0.0833333333333333 |
145145+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+-------------------------------+
146146
147147
0 commit comments