User:Tohline/Appendix/Ramblings/PowerSeriesExpressions
Approximate Power-Series Expressions
|
|---|
| | Tiled Menu | Tables of Content | Banner Video | Tohline Home Page | |
Broadly Used Mathematical Expressions (shown here without proof)
Binomial
|
<math>~(1 \pm x)^n</math> |
<math>~=</math> |
<math>~ 1 ~\pm ~nx + \biggl[\frac{n(n-1)}{2!}\biggr]x^2 ~\pm~ \biggl[\frac{n(n-1)(n-2)}{3!}\biggr]x^3 + \biggl[\frac{n(n-1)(n-2)(n-3)}{4!}\biggr]x^4 ~~\pm ~~ \cdots </math> for <math>~(x^2 < 1)</math> |
|
LaTeX mathematical expressions cut-and-pasted directly from
|
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
As a primary point of reference, note that according to §1.2 of NIST's Digital Library of Mathematical Functions, the binomial theorem states that,
where, for nonnegative integer values of <math>~k</math> and <math>~n</math> and <math>~k \le n</math>, the notation,
Our Example: Setting <math>~a = 1</math> gives,
|
Note, for example, that,
|
<math>~(1+x)^{-1}</math> |
<math>~=</math> |
<math>~1 - x +x^2 - x^3 + x^4 - x^5 + \cdots \, ;</math> |
|
<math>~(1+x)^{-2}</math> |
<math>~=</math> |
<math>~1 - 2x + 3x^2 - 4x^3 + 5x^4 - 6x^5 + \cdots \, ;</math> |
|
<math>~(1+x)^{-3}</math> |
<math>~=</math> |
<math>~1 - 3x + \biggl[ \frac{3\cdot 4 }{ 2} \biggr]x^2 - \biggl[ \frac{ 3\cdot 4 \cdot 5}{ 2\cdot 3} \biggr]x^3 + \biggl[ \frac{3\cdot 4 \cdot 5 \cdot 6 }{2\cdot 3 \cdot 4 } \biggr]x^4 - \biggl[ \frac{3\cdot 4 \cdot 5 \cdot 6 \cdot 7}{2\cdot 3 \cdot 4 \cdot 5 } \biggr]x^5 + \cdots </math> |
|
|
<math>~=</math> |
<math>~1 - 3x + 6x^2 - 10x^3 + 15x^4 - 21x^5 + \cdots \, ;</math> |
|
<math>~(1+x)^{-4}</math> |
<math>~=</math> |
<math>~1 - 4x + \biggl[ \frac{4\cdot 5 }{ 2} \biggr]x^2 - \biggl[ \frac{ 4\cdot 5 \cdot 6}{ 2\cdot 3} \biggr]x^3 + \biggl[ \frac{4\cdot 5 \cdot 6 \cdot 7 }{2\cdot 3 \cdot 4 } \biggr]x^4 - \biggl[ \frac{4\cdot 5 \cdot 6 \cdot 7 \cdot 8}{2\cdot 3 \cdot 4 \cdot 5 } \biggr]x^5 + \cdots </math> |
|
|
<math>~=</math> |
<math>~1 - 4x + 10 x^2 - 20x^3 + 35x^4 - 56x^5 + \cdots \, .</math> |
See also:
Exponential
|
<math>~e^x</math> |
<math>~=</math> |
<math>~ 1 + x + \frac{x^2}{2!} + \frac{x^3}{3!} + \frac{x^4}{4!} + \cdots </math> |
Expressions with Astrophysical Relevance
Polytropic Lane-Emden Function
We seek a power-series expression for the polytropic, Lane-Emden function, <math>~\Theta_\mathrm{H}(\xi)</math> — expanded about the coordinate center, <math>~\xi = 0</math> — that approximately satisfies the Lane-Emden equation,
|
|
<math>~\frac{1}{\xi^2} \frac{d}{d\xi}\biggl( \xi^2 \frac{d\Theta_H}{d\xi} \biggr) = - \Theta_H^n</math> |
A general power-series should be of the form,
|
<math>~\Theta_H</math> |
<math>~=</math> |
<math>~ \theta_0 + a\xi + b\xi^2 + c\xi^3 + d\xi^4 + e\xi^5 + f\xi^6 + g\xi^7 + h\xi^8 + \cdots </math> |
First derivative:
|
<math>~\frac{d\Theta_H}{d\xi}</math> |
<math>~=</math> |
<math>~ a + 2b\xi + 3c\xi^2 + 4d\xi^3 + 5e\xi^4 + 6f\xi^5 + 7g\xi^6 + 8h\xi^7 + \cdots </math> |
Left-hand-side of Lane-Emden equation:
|
<math>~\frac{1}{\xi^2} \frac{d}{d\xi}\biggl( \xi^2 \frac{d\Theta_H}{d\xi} \biggr)</math> |
<math>~=</math> |
<math>~ \frac{2a}{\xi} + 2\cdot 3b + 2^2\cdot 3c\xi + 2^2\cdot 5d\xi^2 + 2\cdot 3\cdot 5e\xi^3 + 2\cdot 3\cdot 7f\xi^4 + 2^3\cdot 7g\xi^5 + 2^3\cdot 3^2h\xi^6 + \cdots </math> |
Right-hand-side of Lane-Emden equation (adopt the normalization, <math>~\theta_0=1</math>, then use the binomial theorem recursively):
|
<math>~\Theta_H^n</math> |
<math>~=</math> |
<math>~ 1 ~+ ~nF + \biggl[\frac{n(n-1)}{2!}\biggr]F^2 ~+~ \biggl[\frac{n(n-1)(n-2)}{3!}\biggr]F^3 + \biggl[\frac{n(n-1)(n-2)(n-3)}{4!}\biggr]F^4 ~~+ ~~ \cdots </math> |
where,
|
<math>~F</math> |
<math>~\equiv</math> |
<math>~ a\xi + b\xi^2 + c\xi^3 + d\xi^4 + e\xi^5 + f\xi^6 + g\xi^7 + h\xi^8 + \cdots </math> |
|
|
<math>~=</math> |
<math>~ a\xi\biggl[1 + \frac{b}{a}\xi + \frac{c}{a}\xi^2 + \frac{d}{a}\xi^3 + \frac{e}{a}\xi^4 + \frac{f}{a}\xi^5 + \frac{g}{a}\xi^6 + \frac{h}{a}\xi^7 + \cdots\biggr] \, . </math> |
First approximation: Assume that <math>~e=f=g=h=0</math>, in which case the LHS contains terms only up through <math>~\xi^2</math>. This means that we must ignore all terms on the RHS that are of higher order than <math>~\xi^2</math>; that is,
|
<math>~\Theta_H^n</math> |
<math>~\approx</math> |
<math>~ 1 ~+ ~nF + \biggl[\frac{n(n-1)}{2!}\biggr]F^2 </math> |
|
|
<math>~\approx</math> |
<math>~ 1 ~+ ~n(a\xi+b\xi^2) + \biggl[\frac{n(n-1)}{2!}\biggr]a^2\xi^2 </math> |
|
|
<math>~\approx</math> |
<math>~ 1 ~+~na\xi + ~\biggl[n b + \frac{n(n-1)a^2}{2}\biggr]\xi^2\, . </math> |
Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of <math>~\xi</math>. Remembering to include a negative sign on the RHS, we find:
| Term | LHS | RHS | Implication |
|
<math>~\xi^{-1}:</math> |
<math>~2a</math> |
<math>~0</math> |
<math>~\Rightarrow ~~~a=0</math> |
|
<math>~\xi^{0}:</math> |
<math>~2\cdot 3 b</math> |
<math>~-1</math> |
<math>~\Rightarrow ~~~b=- \frac{1}{6}</math> |
|
<math>~\xi^{1}:</math> |
<math>~2^2\cdot 3 c</math> |
<math>~-na</math> |
<math>~\Rightarrow ~~~c=0</math> |
|
<math>~\xi^{2}:</math> |
<math>~2^2\cdot 5 d</math> |
<math>~-\biggl[n b + \frac{n(n-1)a^2}{2}\biggr]</math> |
<math>~\Rightarrow ~~~d=+\frac{n}{120}</math> |
By including higher and higher order terms in the series expansion for <math>~\Theta_H</math>, and proceeding along the same line of deductive reasoning, one finds:
- Expressions for the four coefficients, <math>~a, b, c, d</math>, remain unchanged.
- The coefficient is zero for all other terms that contain odd powers of <math>~\xi</math>; specifically, for example, <math>~e = g = 0</math>.
- The coefficients of <math>~\xi^6</math> and <math>~\xi^8</math> are, respectively,
|
<math>~f</math> |
<math>~=</math> |
<math>~- \frac{n}{378}\biggl(\frac{n}{5}-\frac{1}{8} \biggr) \, ;</math> |
|
<math>~h</math> |
<math>~=</math> |
<math>~\frac{n(122n^2 -183n + 70)}{3265920} \, .</math> |
In summary, the desired, approximate power-series expression for the polytropic Lane-Emden function is:
| For Spherically Symmetric Configurations | |||
|---|---|---|---|
|
NOTE: For cylindrically symmetric, rather than spherically symmetric, configurations, the analogous power-series expression appears as equation (15) in the article by J. P. Ostriker (1964, ApJ, 140, 1056) titled, The Equilibrium of Polytropic and Isothermal Cylinders.
Isothermal Lane-Emden Function
Here we seek a power-series expression for the isothermal, Lane-Emden function — expanded about the coordinate center — that approximately satisfies the isothermal Lane-Emden equation; making the variable substitution (sorry for the unnecessary complication!), <math>~\psi(\xi) \leftrightarrow w(r)</math>, the governing ODE is,
|
<math>~\frac{d^2w}{dr^2} +\frac{2}{r} \frac{d w}{dr} </math> |
<math>~=</math> |
<math>~e^{-w} \, . </math> |
A general power-series should be of the form,
|
<math>~w</math> |
<math>~=</math> |
<math>~ w_0 + ar + br^2 + cr^3 + dr^4 + er^5 + fr^6 + gr^7 + hr^8 +\cdots </math> |
Derivatives:
|
<math>~\frac{dw}{dr}</math> |
<math>~=</math> |
<math>~ a + 2br + 3cr^2 + 4dr^3 + 5er^4 + 6fr^5 + 7gr^6 + 8hr^7 +\cdots \, ; </math> |
|
<math>~\frac{d^2w}{dr^2}</math> |
<math>~=</math> |
<math>~ 2b + 2\cdot 3cr + 2^2\cdot 3dr^2 + 2^2\cdot 5er^3 + 2\cdot 3 \cdot 5fr^4 + 2\cdot 3 \cdot 7gr^5 + 2^3\cdot 7hr^6 +\cdots \, . </math> |
Put together, then, the left-hand-side of the isothermal Lane-Emden equation becomes:
|
<math>~\frac{d^2w}{dr^2} +\frac{2}{r} \frac{d w}{dr} </math> |
<math>~=</math> |
<math>~ 2b + 2\cdot 3cr + 2^2\cdot 3dr^2 + 2^2\cdot 5er^3 + 2\cdot 3 \cdot 5fr^4 + 2\cdot 3 \cdot 7gr^5 + 2^3\cdot 7hr^6 + \frac{2}{r}\biggl[ a + 2br + 3cr^2 + 4dr^3 + 5er^4 + 6fr^5 + 7gr^6 + 8hr^7 \biggr] + \cdots </math> |
|
|
<math>~=</math> |
<math>~\frac{2a}{r} + r^0(6b) + r^1(2^2\cdot 3c) + r^2(2^2\cdot 3d + 2^3d) + r^3(2^2\cdot 5e + 2\cdot 5e) + r^4(2\cdot 3\cdot 5 f + 2^2\cdot 3f) + r^5(2\cdot 3\cdot 7 g+ 2\cdot 7g) + r^6(2^3 \cdot 7 h + 2^4 h) + \cdots </math> |
Drawing on the above power-series expression for an exponential function, and adopting the convention that <math>~w_0 = 0</math>, the right-hand-side becomes,
|
<math>~e^{-w}</math> |
<math>~=</math> |
<math>~ e^{0}\cdot e^{-ar} \cdot e^{-br^2} \cdot e^{-cr^3} \cdot e^{-dr^4} \cdot e^{-er^5} \cdot e^{-fr^6} \cdot e^{-gr^7} \cdot e^{-hr^8} \cdots </math> |
|
|
<math>~=</math> |
<math>~ \biggl[ 1 -ar + \frac{a^2r^2}{2!} - \frac{a^3r^3}{3!} + \frac{a^4r^4}{4!} - \frac{a^5r^5}{5!} + \frac{a^6r^6}{6!} + \cdots \biggr] </math> |
|
|
|
<math>~ \times \biggl[ 1 -br^2 + \frac{b^2r^4}{2!} - \frac{b^3r^6}{3!} + \cdots \biggr] \times \biggl[ 1 -cr^3 + \frac{c^2r^6}{2!} + \cdots \biggr] \times \biggl[1 - dr^4\biggr] \times \biggl[1 - er^5\biggr]\times \biggl[1 - fr^6\biggr] </math> |
|
|
<math>~\approx</math> |
<math>~ \biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + \frac{a^4r^4}{24} - \frac{a^5r^5}{5\cdot 24} + \frac{a^6r^6}{30\cdot 24} \biggr] \times \biggl[ 1 -cr^3 + \frac{c^2r^6}{2} -br^2 + bcr^5 + \frac{b^2r^4}{2} - \frac{b^3r^6}{6} \biggr] \times \biggl[1 - dr^4 - er^5 - fr^6\biggr] </math> |
|
|
<math>~\approx</math> |
<math>~\biggl\{ \biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + \frac{a^4r^4}{24} - \frac{a^5r^5}{5\cdot 24} + \frac{a^6r^6}{30\cdot 24} \biggr] - dr^4 \biggl[ 1 -ar + \frac{a^2r^2}{2} \biggr] - er^5 \biggl[ 1 -ar \biggr] - fr^6 \biggr\} </math> |
|
|
|
<math>~ \times \biggl[ 1 -br^2 -cr^3 + \frac{b^2r^4}{2} + bcr^5 + r^6\biggl(\frac{c^2}{2}- \frac{b^3}{6}\biggr) \biggr] </math> |
|
|
<math>~\approx</math> |
<math>~\biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + \frac{a^4r^4}{24} - \frac{a^5r^5}{5\cdot 24} + \frac{a^6r^6}{30\cdot 24} - dr^4 + adr^5 - \frac{a^2d r^6}{2} - er^5 + aer^6 - fr^6 \biggr] </math> |
|
|
|
<math>~ \times \biggl[ 1 -br^2 -cr^3 + \frac{b^2r^4}{2} + bcr^5 + r^6\biggl(\frac{c^2}{2}- \frac{b^3}{6}\biggr) \biggr] </math> |
|
|
<math>~\approx</math> |
<math>~\biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + r^4\biggl(\frac{a^4}{24} - d\biggr) + r^5\biggl(ad - e-\frac{a^5}{5\cdot 24}\biggr) + r^6 \biggl(\frac{a^6}{30\cdot 24} - \frac{a^2d}{2} + ae - f \biggr) \biggr] \times \biggl[ 1 -br^2 -cr^3 + \frac{b^2r^4}{2} + bcr^5 + r^6\biggl(\frac{c^2}{2}- \frac{b^3}{6}\biggr) \biggr] </math> |
|
|
<math>~\approx</math> |
<math>~ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + r^4\biggl(\frac{a^4}{24} - d\biggr) + r^5\biggl(ad - e-\frac{a^5}{5\cdot 24}\biggr) + r^6 \biggl(\frac{a^6}{30\cdot 24} - \frac{a^2d}{2} + ae - f \biggr) </math> |
|
|
|
<math>~-br^2\biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} + r^4\biggl(\frac{a^4}{24} - d\biggr) \biggr] -cr^3 \biggl[ 1 -ar + \frac{a^2r^2}{2} - \frac{a^3r^3}{6} \biggr] + \frac{b^2r^4}{2}\biggl[ 1 -ar + \frac{a^2r^2}{2} \biggr] + bcr^5\biggl[1 -ar \biggr] + r^6\biggl(\frac{c^2}{2}- \frac{b^3}{6}\biggr) </math> |
Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of <math>~r</math>. Beginning with the highest order terms, we initially find,
| Term | LHS | RHS | Implication |
|
<math>~r^{-1}:</math> |
<math>~2a</math> |
<math>~0</math> |
<math>~\Rightarrow ~~~a=0</math> |
|
<math>~r^{0}:</math> |
<math>~6b</math> |
<math>~1</math> |
<math>~\Rightarrow ~~~b = + \frac{1}{6}</math> |
|
<math>~r^{1}:</math> |
<math>~2^2\cdot 3c</math> |
<math>~-a</math> |
<math>~\Rightarrow ~~~c = -\frac{a}{2^2\cdot 3} =0</math> |
|
<math>~r^{2}:</math> |
<math>~(2^2\cdot 3d + 2^3d)</math> |
<math>~\frac{a^2}{2} - b</math> |
<math>~\Rightarrow ~~~d = \frac{1}{20}\biggl( \frac{a^2}{2} - b \biggr) = - \frac{1}{120}</math> |
With this initial set of coefficient values in hand, we can rewrite (and significantly simplify) our approximate expression for the RHS, namely,
|
<math>~e^{-w}</math> |
<math>~\approx</math> |
<math>~ 1 -d r^4 -e r^5 -f r^6 -br^2 ( 1 -d r^4 ) + \frac{b^2r^4}{2} - \frac{b^3r^6}{6} </math> |
|
|
<math>~=</math> |
<math>~ 1 -br^2+ r^4 \biggl(\frac{b^2}{2} -d \biggr) -e r^5 +r^6\biggl( bd - \frac{b^3}{6} -f \biggr) \, . </math> |
Continuing, then, with equating terms with like powers on both sides of the equation, we find,
| Term | LHS | RHS | Implication |
|
<math>~r^{3}:</math> |
<math>~30e</math> |
<math>~0</math> |
<math>~\Rightarrow ~~~e=0</math> |
|
<math>~r^{4}:</math> |
<math>~(2\cdot 3\cdot 5 f + 2^2\cdot 3f)</math> |
<math>~\biggl(\frac{b^2}{2} -d \biggr) </math> |
<math>~\Rightarrow ~~~f = \frac{1}{2\cdot 3\cdot 7}\biggl(\frac{1}{2^3\cdot 3^2}+\frac{1}{2^3\cdot 3 \cdot 5}\biggr) = \frac{1}{2\cdot 3^3\cdot 5 \cdot 7}</math> |
|
<math>~r^{5}:</math> |
<math>~(2\cdot 3\cdot 7 g+ 2\cdot 7g)</math> |
<math>~-e</math> |
<math>~\Rightarrow ~~~g = 0</math> |
|
<math>~r^{6}:</math> |
<math>~(2^3 \cdot 7 h + 2^4 h)</math> |
<math>~\biggl( bd - \frac{b^3}{6} -f \biggr)</math> |
<math>~\Rightarrow ~~~ h = -\frac{1}{2^3\cdot 3^2}\biggl( \frac{1}{2^4\cdot 3^2 \cdot 5} + \frac{1}{2^4\cdot 3^4} + \frac{1}{2\cdot 3^3\cdot 5\cdot 7}\biggr) = -\frac{61}{2^{6} \cdot 3^6\cdot 5\cdot 7} </math> |
Result:
| For Spherically Symmetric Configurations | |||
|---|---|---|---|
|
See also:
- Equation (377) from §22 in Chapter IV of C67).
NOTE: For cylindrically symmetric, rather than spherically symmetric, configurations, an analytic expression for the function, <math>~w(r)</math>, is presented as equation (56) in a paper by J. P. Ostriker (1964, ApJ, 140, 1056) titled, The Equilibrium of Polytropic and Isothermal Cylinders.
Displacement Function for Polytropic LAWE
The LAWE for polytropic spheres may be written as,
|
<math>~0 </math> |
<math>~=</math> |
<math>~\frac{d^2x}{d\xi^2} + \biggl[\frac{4}{\xi} - \frac{(n+1)}{\theta} \biggl(- \frac{d\theta}{d\xi}\biggr)\biggr] \frac{dx}{d\xi} + \frac{(n+1)}{\theta}\biggl[\frac{\sigma_c^2}{6\gamma } - \frac{\alpha}{\xi} \biggl(- \frac{d\theta}{d\xi}\biggr)\biggr] x </math> |
|
|
<math>~=</math> |
<math>~\theta \frac{d^2x}{d\xi^2} + \biggl[4\theta - (n+1)\xi \biggl(- \frac{d\theta}{d\xi}\biggr)\biggr] \frac{1}{\xi}\frac{dx}{d\xi} + \frac{(n+1)}{6} \biggl[\frac{\sigma_c^2}{\gamma } - \frac{6\alpha}{\xi} \biggl(- \frac{d\theta}{d\xi}\biggr)\biggr] x \, ,</math> |
where, <math>~\theta(\xi)</math> is the polytropic Lane-Emden function describing the configuration's unperturbed radial density distribution, and <math>~\gamma</math>, <math>~\sigma_c^2</math>, and <math>~\alpha \equiv (3-4/\gamma)</math> are constants. Here we seek a power-series expression for the displacement function, <math>~x(r)</math>, expanded about the center of the configuration, that approximately satisfies this LAWE.
First we note that, near the center, an accurate power-series expression for the polytropic Lane-Emden function is,
|
<math>~\theta</math> |
<math>~=</math> |
<math>~ 1 - \frac{\xi^2}{6} + \frac{n}{120} \xi^4 - \frac{n}{378} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^6 + \cdots </math> |
Hence,
|
<math>~-\frac{d\theta}{d\xi}</math> |
<math>~\approx</math> |
<math>~ \frac{1}{3} \biggl[ \xi - \frac{n}{10} \xi^3 + \frac{n}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^5 \biggr] \, .</math> |
Therefore, near the center of the configuration, the LAWE may be written as,
|
<math>~6~\theta \frac{d^2x}{d\xi^2} + \biggl\{ 12~\theta - (n+1)\xi \biggl[ \xi - \frac{n}{10} \xi^3 + \frac{n}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^5 \biggr] \biggr\} \frac{2}{\xi}\frac{dx}{d\xi}</math> |
<math>~\approx</math> |
<math>~ - (n+1) \biggl\{ \frac{\sigma_c^2}{\gamma } - \frac{2\alpha}{\xi} \biggl[ \xi - \frac{n}{10} \xi^3 + \frac{n}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^5 \biggr] \biggr\} x </math> |
|
<math>\Rightarrow~~~ ~6\biggl[ 1 - \frac{\xi^2}{6} + \frac{n}{120} \xi^4 \biggr] \frac{d^2x}{d\xi^2} + \biggl\{ 12 \biggl[ 1 - \frac{\xi^2}{6} + \frac{n}{120} \xi^4 \biggr] - (n+1)\biggl[ \xi^2 - \frac{n}{10} \xi^4 \biggr] \biggr\} \frac{2}{\xi}\frac{dx}{d\xi}</math> |
<math>~\approx</math> |
<math>~ - (n+1) \biggl\{ \mathfrak{F} + 2\alpha \biggl[ \frac{n}{10} \xi^2 - \frac{n}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^4 \biggr] \biggr\} x </math> |
|
<math>\Rightarrow~~~ ~\biggl( 6 - \xi^2 + \frac{n}{20} \xi^4 \biggr) \frac{d^2x}{d\xi^2} + \biggl[ 12 - (n+3)\xi^2 + \frac{n(n+2)}{10} \xi^4 \biggr] \frac{2}{\xi}\frac{dx}{d\xi}</math> |
<math>~\approx</math> |
<math>~ - (n+1) \biggl[ \mathfrak{F} + \frac{n\alpha}{5} \xi^2 - \frac{2n\alpha}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^4 \biggr] x \, ,</math> |
where, for present purposes, we have kept terms in the series no higher than <math>~\xi^4</math>. Let's now adopt a power-series expression for the displacement function of the form,
|
<math>~x</math> |
<math>~=</math> |
<math>~ 1 + a\xi + b\xi^2 + c\xi^3 + d\xi^4 + e\xi^5 + f\xi^6\cdots </math> |
|
<math>~\Rightarrow ~~~ \frac{1}{\xi}\frac{dx}{d\xi}</math> |
<math>~=</math> |
<math>~ \frac{a}{\xi} + 2b + 3 c\xi + 4d\xi^2 + 5e\xi^3 + 6f\xi^4 +\cdots </math> |
and,
|
<math>~\frac{d^2x}{d\xi^2}</math> |
<math>~=</math> |
<math>~ 2b + 6c\xi + 12d\xi^2 + 20e\xi^3 + 30f\xi^4 + \cdots </math> |
Substituting these expressions into the LAWE gives,
|
<math>~\biggl( 6 - \xi^2 + \frac{n}{20} \xi^4 \biggr) \biggl( 2b + 6c\xi + 12d\xi^2 + 20e\xi^3 + 30f\xi^4 \biggr) + \biggl[ 12 - (n+3)\xi^2 + \frac{n(n+2)}{10} \xi^4 \biggr] \biggl( \frac{2a}{\xi} + 4b + 6 c\xi + 8d\xi^2 + 10e\xi^3 + 12f\xi^4 \biggr)</math> |
<math>~\approx</math> |
<math>~ - (n+1) \biggl[ \mathfrak{F} + \frac{n\alpha}{5} \xi^2 - \frac{2n\alpha}{21} \biggl( \frac{n}{5} - \frac{1}{8} \biggr) \xi^4 \biggr] \biggl( 1 + a\xi + b\xi^2 + c\xi^3 + d\xi^4 \biggr)</math> |
Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of <math>~\xi</math>.
| Term | LHS | RHS | Implication |
|
<math>~\xi^{-1}:</math> |
<math>~24a</math> |
<math>~0</math> |
<math>~\Rightarrow ~~~a=0</math> |
|
<math>~\xi^{0}:</math> |
<math>~(12b + 48b)</math> |
<math>~-(n+1)\mathfrak{F}</math> |
<math>~\Rightarrow ~~~b = - \frac{(n+1)\mathfrak{F}}{60}</math> |
|
<math>~\xi^{1}:</math> |
<math>~[36c+72c-2a(n+3)]</math> |
<math>~-a(n+1)\mathfrak{F}</math> |
<math>~\Rightarrow ~~~108c = 2a(n+3)-a(n+1)\mathfrak{F} \Rightarrow~~c=0</math> |
|
<math>~\xi^{2}:</math> |
<math>~[72d-2b+96d-4b(n+3)]</math> |
<math>~\biggl[-b(n+1)\mathfrak{F}-\frac{n(n+1)\alpha}{5}\biggr]</math> |
<math>~\Rightarrow ~~~d = - (n+1)\biggl\{ \frac{n\alpha +\mathfrak{F}[(4n+14)-(n+1)\mathfrak{F} ]}{10080} \biggr\}</math> |
In summary, the desired, approximate power-series expression for the polytropic displacement function is:
|
Displacement Function for Isothermal LAWE
The LAWE for isothermal spheres may be written as,
|
<math>~\frac{d^2 x}{dr^2} + \biggl[4 - r \biggl(\frac{dw }{dr}\biggr) \biggr] \frac{1}{r}\frac{dx}{dr}</math> |
<math>~=</math> |
<math>~ - \biggl[ \frac{\sigma_c^2}{6\gamma} - \frac{\alpha}{r} \biggl(\frac{dw }{dr}\biggr)\biggr] x \, , </math> |
where, <math>~w(r)</math> is the isothermal Lane-Emden function describing the configuration's unperturbed radial density distribution, and <math>~\gamma</math>, <math>~\sigma_c^2</math>, and <math>~\alpha \equiv (3-4/\gamma)</math> are constants. Here we seek a power-series expression for the displacement function, <math>~x(r)</math>, expanded about the center of the configuration, that approximately satisfies this LAWE.
First we note that, near the center, an accurate power-series expression for the isothermal Lane-Emden function is,
|
<math>~w(r) </math> |
<math>~=</math> |
<math>~\frac{r^2}{6} - \frac{r^4}{120} + \frac{r^6}{1890} - \frac{61 r^8}{1,632,960} + \cdots \, .</math> |
Hence,
|
<math>~\frac{dw}{dr}</math> |
<math>~\approx</math> |
<math>~\frac{r}{3} - \frac{r^3}{30} + \frac{r^5}{315} \, .</math> |
Therefore, near the center of the configuration, the LAWE may be written as,
|
<math>~\frac{d^2 x}{dr^2} + \biggl[4 - \biggl(\frac{r^2}{3} - \frac{r^4}{30} + \frac{r^6}{315}\biggr) \biggr] \frac{1}{r}\frac{dx}{dr}</math> |
<math>~\approx</math> |
<math>~ - \frac{1}{6} \biggl[ \frac{\sigma_c^2}{\gamma} - 2\alpha \biggl(1 - \frac{r^2}{10} + \frac{r^4}{105}\biggr) \biggr] x \, . </math> |
Let's now adopt a power-series expression for the displacement function of the form,
|
<math>~x</math> |
<math>~=</math> |
<math>~ 1 + ar + br^2 + cr^3 + dr^4 + \cdots </math> |
|
<math>~\Rightarrow ~~~ \frac{1}{r}\frac{dx}{dr}</math> |
<math>~=</math> |
<math>~ \frac{a}{r} + 2b + 3 cr + 4dr^2 + \cdots </math> |
and,
|
<math>~\frac{d^2x}{dr^2}</math> |
<math>~=</math> |
<math>~ 2b + 6cr + 12dr^2 + \cdots </math> |
Substituting these expressions into the LAWE gives,
|
<math>~2b + 6cr + 12dr^2 + \biggl[4 - \biggl(\frac{r^2}{3} - \frac{r^4}{30} + \frac{r^6}{315}\biggr) \biggr] \biggl[ \frac{a}{r} + 2b + 3 cr + 4dr^2 \biggr] </math> |
<math>~\approx</math> |
<math>~ - \frac{1}{6} \biggl[ \frac{\sigma_c^2}{\gamma} - 2\alpha \biggl(1 - \frac{r^2}{10} + \frac{r^4}{105}\biggr) \biggr] \biggl( 1 + ar + br^2 + cr^3 + dr^4 \biggr) \, . </math> |
Keeping terms only up through <math>~r^2</math> leads to the following simplification:
|
<math>~ 2b + 6cr + 12dr^2 + 4 \biggl[ \frac{a}{r} + 2b + 3 cr + 4dr^2 \biggr] - \frac{r^2}{3} \biggl[ \frac{a}{r} + 2b \biggr] </math> |
<math>~\approx</math> |
<math>~ - \frac{\mathfrak{F} }{6} \biggl( 1 + ar + br^2 \biggr) - \frac{\alpha}{3} \biggl(\frac{r^2}{10} \biggr) </math> |
where,
<math>~\mathfrak{F} \equiv \frac{\sigma_c^2}{\gamma} - 2\alpha \, .</math>
Finally, balancing terms of like powers on both sides of the equation leads us to conclude the following:
| Term | LHS | RHS | Implication |
|
<math>~r^{-1}:</math> |
<math>~4a</math> |
<math>~0</math> |
<math>~\Rightarrow ~~~a = 0 </math> |
|
<math>~r^{0}:</math> |
<math>~2b + 8b</math> |
<math>~- \frac{\mathfrak{F}}{6}</math> |
<math>~\Rightarrow ~~~b = - \frac{\mathfrak{F}}{60}</math> |
|
<math>~r^{1}:</math> |
<math>~6c + 12c - \frac{a}{3}</math> |
<math>~-\frac{a\mathfrak{F}}{6}</math> |
<math>~\Rightarrow ~~~c=0</math> |
|
<math>~r^{2}:</math> |
<math>~12d + 16d - \frac{2b}{3}</math> |
<math>~-\frac{\mathfrak{F}b}{6} - \frac{\alpha}{30}</math> |
<math>~\Rightarrow ~~~ 28d = \frac{1}{30}\biggl[ 5b (4- \mathfrak{F} ) - \alpha \biggr] ~ \Rightarrow~ d = \frac{1}{10080}\biggl[ \mathfrak{F}(\mathfrak{F} -4) - 12\alpha \biggr] </math> |
In summary, the desired, approximate power-series expression for the isothermal displacement function is:
|
Taylor Series (Hunter77)
First (Unsuccessful) Try
First:
|
<math>~f_0</math> |
<math>~=</math> |
<math>~ f_3 + (- 3\Delta) f_3^' + \frac{1}{2} (- 3\Delta)^2 f^{}_3 + \frac{1}{6} (- 3\Delta)^3 f_3^{'} + \frac{1}{24}(- 3\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 - (3\Delta) f_3^' + \frac{3^2}{2} (\Delta)^2 f^{}_3 - \frac{3^2}{2} (\Delta)^3 f_3^{'} + \frac{3^3}{2^3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ - \frac{3^2}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ f_3 - f_0 - (3\Delta) f_3^' - \frac{3^2}{2} (\Delta)^3 f_3^{} + \frac{3^3}{2^3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
Note that, replacing the <math>~(\Delta)^3 f_3^{}</math> term with the expression derived in the Second step, below, gives,
|
<math>~ - \frac{3^2}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ f_3 - f_0 - (3\Delta) f_3^' + \frac{3^3}{2^3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ - \frac{3^2}{2} \biggl\{ \biggl[\frac{2^2}{3^2} \biggr] f_0 - f_1 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \biggl[- \frac{5}{6} \biggr] (\Delta)^4 f_3^{iv} \biggr\}\biggl[ -\frac{3}{2} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ f_3 - f_0 - 3 (\Delta) f_3^' + \frac{3^3}{2^3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ 3 f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[\frac{15}{2^2} \biggr] f_3 + \biggl[- \frac{3}{2}\biggr] (\Delta) f_3^' + \biggl[- \frac{3^2\cdot 5}{2^3} \biggr] (\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ 2f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[1 + \frac{15}{2^2} \biggr] f_3 + \biggl[-3 - \frac{3}{2}\biggr] (\Delta) f_3^' + \biggl[\frac{3^3}{2^3}- \frac{3^2\cdot 5}{2^3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ 2f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[\frac{19}{2^2} \biggr] f_3 + \biggl[- \frac{9}{2}\biggr] (\Delta) f_3^' + \biggl[- \frac{9}{4} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
Then, replacing the <math>~(\Delta)^4 f_3^{iv}</math> term with the expression derived in the Third step, below, gives,
|
<math>~ - \frac{3^2}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ 2f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[\frac{19}{2^2} \biggr] f_3 + \biggl[- \frac{9}{2}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[- \frac{9}{4} \biggr] \biggl\{ \biggl[-\frac{1}{3^2} \biggr]f_0 + \biggl[\frac{1}{2} \biggr] f_1 - f_2 + \biggl[ \frac{11}{2\cdot 3^2} \biggr] f_3 + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' \biggr\}\biggl[- 2^2\cdot 3 \biggr] </math> |
|
|
<math>~=</math> |
<math>~ 2f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[\frac{19}{2^2} \biggr] f_3 + \biggl[- \frac{9}{2}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[-3 \biggr]f_0 + \biggl[\frac{3^3 }{2} \biggr] f_1 - 3^3 f_2 + \biggl[ \frac{3 \cdot 11}{2} \biggr] f_3 + \biggl[- 2\cdot 3^2 \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ - f_0 + \biggl[\frac{3^3 }{2} - \frac{3^3}{2^2}\biggr] f_1 - 3^3 f_2 + \biggl[\frac{3 \cdot 11}{2} + \frac{19}{2^2} \biggr] f_3 + \biggl[- 2\cdot 3^2- \frac{9}{2}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ - f_0 + \biggl[\frac{3^3}{2^2}\biggr] f_1 - 3^3 f_2 + \biggl[\frac{5\cdot 17}{2^2} \biggr] f_3 + \biggl[- \frac{3^2\cdot 5}{2}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
Second:
|
<math>~f_1</math> |
<math>~=</math> |
<math>~ f_3 + (- 2\Delta) f_3^' + \frac{1}{2} (- 2\Delta)^2 f^{}_3 + \frac{1}{6} (- 2\Delta)^3 f_3^{'} + \frac{1}{24}(- 2\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 - 2(\Delta) f_3^' + 2 (\Delta)^2 f^{}_3 - \frac{2^2}{3} (\Delta)^3 f_3^{'} + \frac{2}{3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 - 2(\Delta) f_3^' - \frac{2^2}{3} (\Delta)^3 f_3^{} + \frac{2}{3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ - \biggl[\frac{2^2}{3^2} \biggr] \biggl[ f_3 - f_0 - (3\Delta) f_3^' - \frac{3^2}{2} (\Delta)^3 f_3^{} + \frac{3^3}{2^3}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) \biggr] </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} \biggr] f_0 + f_3 \biggl[1-\frac{2^2}{3^2} \biggr] + \biggl[ \frac{2^2}{3^2} (3\Delta) - 2(\Delta) \biggr] f_3^' + \biggl[ \biggl(\frac{2^2}{3^2} \biggr) \frac{3^2}{2} (\Delta)^3 - \frac{2^2}{3} (\Delta)^3 \biggr] f_3^{} </math> |
|
|
|
<math>~ + \biggl[ \frac{2}{3}(\Delta)^4 - \biggl( \frac{2^2}{3^2} \biggr) \frac{3^3}{2^3}(\Delta)^4 \biggr] f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} \biggr] f_0 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \biggl[ \frac{2}{3} \biggr] (\Delta)^3f_3^{} + \biggl[- \frac{5}{6} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ - \biggl[ \frac{2}{3} \biggr] (\Delta)^3f_3^{} </math> |
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} \biggr] f_0 - f_1 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \biggl[- \frac{5}{6} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
Now, replacing the <math>~(\Delta)^4 f_3^{iv}</math> term with the expression derived in the Third step, below, gives,
|
<math>~ - \biggl[ \frac{2}{3} \biggr] (\Delta)^3f_3^{} </math> |
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} \biggr] f_0 - f_1 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[- \frac{5}{6} \biggr] \biggl\{ \biggl[-\frac{1}{3^2} \biggr]f_0 + \biggl[\frac{1}{2} \biggr] f_1 - f_2 + \biggl[ \frac{11}{2\cdot 3^2} \biggr] f_3 + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' \biggr\} \biggl[ -2^2\cdot 3\biggr] </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} \biggr] f_0 - f_1 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[-\frac{2\cdot 5 }{3^2} \biggr]f_0 + \biggl[5\biggr] f_1 + \biggl[- 2\cdot 5 \biggr] f_2 + \biggl[ \frac{5\cdot 11}{3^2} \biggr] f_3 + \biggl[- \frac{2^2\cdot 5}{3}\biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{2^2}{3^2} -\frac{2\cdot 5 }{3^2} \biggr] f_0 + \biggl[4\biggr] f_1 + \biggl[- 2\cdot 5 \biggr] f_2 + \biggl[\frac{5}{3^2} + \frac{5\cdot 11}{3^2}\biggr] f_3 + \biggl[- \frac{2^2\cdot 5}{3} - \frac{2}{3}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ \biggl[-\frac{2}{3} \biggr] f_0 + \biggl[4\biggr] f_1 + \biggl[- 2\cdot 5 \biggr] f_2 + \biggl[\frac{2^2\cdot 5}{3}\biggr] f_3 + \biggl[- \frac{2\cdot 11}{3}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
Third:
|
<math>~f_2</math> |
<math>~=</math> |
<math>~ f_3 + (- \Delta) f_3^' + \frac{1}{2} (- \Delta)^2 f^{}_3 + \frac{1}{6} (- \Delta)^3 f_3^{'} + \frac{1}{24}(- \Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ -1 \biggr](\Delta) f_3^' + \biggl[ \frac{1}{2} \biggr] (\Delta)^2 f^{}_3 + \biggl[ - \frac{1}{2\cdot 3} \biggr] (\Delta)^3 f_3^{'} + \biggl[ \frac{1}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ -1 \biggr](\Delta) f_3^' + \biggl[ \frac{1}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[ \frac{1}{2} \biggr] \biggl\{ 2f_0 + \biggl[- \frac{3^3}{2^2}\biggr] f_1 + \biggl[\frac{19}{2^2} \biggr] f_3 + \biggl[- \frac{9}{2}\biggr] (\Delta) f_3^' + \biggl[- \frac{9}{4} \biggr] (\Delta)^4 f_3^{iv} \biggr\} \biggl[-\frac{2}{3^2}\biggr] </math> |
|
|
|
<math>~ + \biggl[ - \frac{1}{2\cdot 3} \biggr] \biggl\{ \biggl[\frac{2^2}{3^2} \biggr] f_0 - f_1 + f_3 \biggl[\frac{5}{3^2} \biggr] + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \biggl[- \frac{5}{6} \biggr] (\Delta)^4 f_3^{iv} \biggr\} \biggl[-\frac{3}{2}\biggr] </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ -1 \biggr](\Delta) f_3^' + \biggl[ \frac{1}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[ -\frac{2}{3^2} \biggr]f_0 + \biggl[\frac{3}{2^2}\biggr] f_1 + \biggl[-\frac{19}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{1}{2}\biggr] (\Delta) f_3^' + \biggl[\frac{1}{4} \biggr] (\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{1}{2^2} \biggr] f_1 + f_3 \biggl[\frac{5}{2^2\cdot 3^2} \biggr] + \biggl[- \frac{1}{2\cdot 3}\biggr] (\Delta) f_3^' + \biggl[- \frac{5}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ -1 \biggr](\Delta) f_3^' + \biggl[ \frac{1}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{3^2} -\frac{2}{3^2} \biggr]f_0 + \biggl[\frac{3}{2^2}- \frac{1}{2^2} \biggr] f_1 + \biggl[\frac{5}{2^2\cdot 3^2} -\frac{19}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{1}{2}- \frac{1}{2\cdot 3}\biggr] (\Delta) f_3^' + \biggl[\frac{1}{4} - \frac{5}{2^3\cdot 3} \biggr] (\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[-\frac{1}{3^2} \biggr]f_0 + \biggl[\frac{1}{2} \biggr] f_1 + \biggl[ \frac{11}{2\cdot 3^2} \biggr] f_3 + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \biggl[\frac{1}{2^2\cdot 3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ - \biggl[\frac{1}{2^2\cdot 3} \biggr] (\Delta)^4 f_3^{iv} </math> |
<math>~=</math> |
<math>~ \biggl[-\frac{1}{3^2} \biggr]f_0 + \biggl[\frac{1}{2} \biggr] f_1 - f_2 + \biggl[ \frac{11}{2\cdot 3^2} \biggr] f_3 + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
And, finally:
|
<math>~f_4</math> |
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \frac{1}{2} ( \Delta)^2 f^{}_3 + \frac{1}{6} (\Delta)^3 f_3^{'} + \frac{1}{24}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \frac{1}{2} \biggl\{ - f_0 + \biggl[\frac{3^3}{2^2}\biggr] f_1 - 3^3 f_2 + \biggl[\frac{5\cdot 17}{2^2} \biggr] f_3 + \biggl[- \frac{3^2\cdot 5}{2}\biggr] (\Delta) f_3^' \biggr\} \biggl[ - \frac{2}{3^2} \biggr] </math> |
|
|
|
<math>~ + \frac{1}{6} \biggl\{ \biggl[-\frac{2}{3} \biggr] f_0 + \biggl[4\biggr] f_1 + \biggl[- 2\cdot 5 \biggr] f_2 + \biggl[\frac{2^2\cdot 5}{3}\biggr] f_3 + \biggl[- \frac{2\cdot 11}{3}\biggr] (\Delta) f_3^' \biggr\} \biggl[ -\frac{3}{2} \biggr] </math> |
|
|
|
<math>~ + \frac{1}{24}\biggl\{ \biggl[-\frac{1}{3^2} \biggr]f_0 + \biggl[\frac{1}{2} \biggr] f_1 - f_2 + \biggl[ \frac{11}{2\cdot 3^2} \biggr] f_3 + \biggl[- \frac{2}{3}\biggr] (\Delta) f_3^' \biggr\} \biggl[ -2^2\cdot 3 \biggr] </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[ \frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{3}{2^2}\biggr] f_1 +\biggl[ 3 \biggr] f_2 + \biggl[- \frac{5\cdot 17}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{5}{2}\biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{2\cdot 3} \biggr] f_0 + \biggl[-1 \biggr] f_1 + \biggl[ \frac{5}{2} \biggr] f_2 + \biggl[- \frac{5}{3}\biggr] f_3 + \biggl[\frac{11}{2\cdot 3}\biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{2\cdot 3^2} \biggr]f_0 + \biggl[- \frac{1}{2^2} \biggr] f_1 + \biggl[ \frac{1}{2} \biggr] f_2 + \biggl[ -\frac{11}{2^2 \cdot 3^2} \biggr] f_3 + \biggl[\frac{1}{3}\biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[ \frac{1}{3^2} + \frac{1}{2\cdot 3} + \frac{1}{2\cdot 3^2} \biggr] f_0 + \biggl[- \frac{3}{2^2} - 1 - \frac{1}{2^2} \biggr] f_1 +\biggl[ 3 + \frac{5}{2} + \frac{1}{2} \biggr] f_2 + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[1 - \frac{5\cdot 17}{2^2\cdot 3^2} - \frac{5}{3} - \frac{11}{2^2 \cdot 3^2} \biggr] f_3 + \biggl[1 + \frac{5}{2} + \frac{11}{2\cdot 3} + \frac{1}{3} \biggr] (\Delta) f_3^' </math> |
|
|
<math>~=</math> |
<math>~ \biggl[ \frac{1}{3} \biggr] f_0 + \biggl[- 2\biggr] f_1 +\biggl[ 6 \biggr] f_2 + \biggl[- \frac{10}{3} \biggr] f_3 + \biggl[\frac{17}{3} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
Result:
|
Definitely WRONG! |
|||
|---|---|---|---|
|
When I used an Excel spreadsheet to test this out against a parabola, the integration quickly became wildly unstable, strongly suggesting that there is an error in the derivation. My first attempt to uncover this error produced a new coefficient on the <math>~(\Delta) f_3^'</math>, namely,
|
Somewhat Improved |
|||
|---|---|---|---|
|
Although it showed improvement, this expression still blows up. So I have not bothered to revise the original (definitely WRONG!) derivation. Instead, let's start all over and approach it with a more gradual derivation.
Second Try
We will work from the following foundation expression in which <math>~f_4</math> is the variable that we desire to evaluate, and the "known" quantities are: <math>~f_3</math>, <math>~f_3^'</math>, <math>~f_2</math>, <math>~f_1</math>, and <math>~f_0</math>.
|
<math>~f_4</math> |
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \frac{1}{2} ( \Delta)^2 f^{}_3 + \frac{1}{6} (\Delta)^3 f_3^{'} + \frac{1}{24}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
Let's use similar Taylor-series expansions for <math>~f_2</math>, <math>~f_3</math>, etc. in order to eliminate the <math>~f_3^{}</math> term, the <math>~f_3^{'}</math> term, etc.
|
<math>~f_2</math> |
<math>~=</math> |
<math>~ f_3 + (- \Delta) f_3^' + \frac{1}{2} (- \Delta)^2 f^{}_3 + \frac{1}{6} (- \Delta)^3 f_3^{'} + \frac{1}{24}(- \Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~f_1</math> |
<math>~=</math> |
<math>~ f_3 + (- 2\Delta) f_3^' + \frac{1}{2} (- 2\Delta)^2 f^{}_3 + \frac{1}{6} (- 2\Delta)^3 f_3^{'} + \frac{1}{24}(- 2\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~f_0</math> |
<math>~=</math> |
<math>~ f_3 + (- 3\Delta) f_3^' + \frac{1}{2} (- 3\Delta)^2 f^{}_3 + \frac{1}{6} (- 3\Delta)^3 f_3^{'} + \frac{1}{24}(- 3\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
First:
|
<math>~-\frac{1}{2} (- \Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ f_3 + (- \Delta) f_3^' - f_2+ \frac{1}{6} (- \Delta)^3 f_3^{} + \frac{1}{24}(- \Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ \frac{1}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ - f_3 + (\Delta) f_3^' + f_2+ \frac{1}{6} (\Delta)^3 f_3^{} - \frac{1}{24}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ f_4 </math> |
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' - f_3 + (\Delta) f_3^' + f_2 + \mathcal{O}(\Delta^3) </math> |
|
|
<math>~=</math> |
<math>~ f_2 + 2(\Delta) f_3^' + \mathcal{O}(\Delta^3) </math> |
| <math>~\mathcal{O}(\Delta^3)</math> | |||
|
This expression works very well for a parabola.
Second:
|
<math>~f_1</math> |
<math>~=</math> |
<math>~ f_3 + (- 2) \Delta f_3^' + 2 (\Delta)^2 f^{}_3 + \biggl[- \frac{2^3}{6}\biggr] \Delta^3 f_3^{'} + \biggl[ \frac{2^4}{2^3\cdot 3} \biggr] \Delta^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (- 2) \Delta f_3^' + \biggl[- \frac{2^3}{6}\biggr] \Delta^3 f_3^{} + \biggl[ \frac{2^4}{2^3\cdot 3} \biggr] \Delta^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + 2 \biggl\{ - f_3 + (\Delta) f_3^' + f_2+ \frac{1}{2\cdot 3} (\Delta)^3 f_3^{} - \frac{1}{2^3\cdot 3}(\Delta)^4 f_3^{iv} \biggr\} \biggl[ 2 \biggr] </math> |
|
|
<math>~=</math> |
<math>~ f_3\biggl[ 1 - 2^2\biggr] + (2^2 - 2) \Delta f_3^' + 2^2f_2 + \biggl[\frac{2}{3} - \frac{2^3}{6}\biggr] \Delta^3 f_3^{} + \biggl[ \frac{2^4}{2^3\cdot 3} - \frac{1}{2\cdot 3}\biggr] \Delta^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3\biggl[ -3\biggr] + (2) \Delta f_3^' + 2^2f_2 + \biggl[- \frac{2}{3}\biggr] \Delta^3 f_3^{} + \biggl[ \frac{1}{2} \biggr] \Delta^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow~~~ \biggl[\frac{2}{3}\biggr] \Delta^3 f_3^{} </math> |
<math>~=</math> |
<math>~ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' + \biggl[ \frac{1}{2} \biggr] \Delta^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
This also allows us to improve the expression for the <math>~f_3^{}</math> term, as initially derived in the "First" subsection, above. Namely,
|
<math>~ \frac{1}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ f_2 - f_3 + (\Delta) f_3^' - \frac{1}{24}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \frac{1}{6} \biggl\{ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' + \biggl[ \frac{1}{2} \biggr] \Delta^4 f_3^{iv} \biggr\} \biggl[ \frac{3}{2} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ - \frac{1}{4} f_1 + 2f_2 + \biggl[ - \frac{7}{4} \biggr] f_3 + \biggl[ \frac{3}{2} \biggr] (\Delta) f_3^' + \biggl[\frac{1}{2^2\cdot 3} \biggr](\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
Hence, an improved expression for <math>~f_4</math> is,
|
<math>~f_4</math> |
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^4) </math> |
|
|
|
<math>~ + \biggl\{ - \frac{1}{4} f_1 + 2f_2 + \biggl[ - \frac{7}{4} \biggr] f_3 + \biggl[ \frac{3}{2} \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \frac{1}{6} \biggl\{ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' \biggr\} \biggl[ \frac{3}{2} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ - \frac{1}{2} f_1 + 3f_2 - \frac{3}{2} f_3 + 3(\Delta) f_3^' + \mathcal{O}(\Delta^4) </math> |
| <math>~\mathcal{O}(\Delta^4)</math> | |||
|
Third:
|
<math>~f_0</math> |
<math>~=</math> |
<math>~ f_3 + (- 3\Delta) f_3^' + \frac{1}{2} (- 3\Delta)^2 f^{}_3 + \frac{1}{6} (- 3\Delta)^3 f_3^{'} + \frac{1}{24}(- 3\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ - 3 \biggr] (\Delta) f_3^' + \biggl[ \frac{3^3}{2^3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + 3^2 \biggl\{ - \frac{1}{4} f_1 + 2f_2 + \biggl[ - \frac{7}{4} \biggr] f_3 + \biggl[ \frac{3}{2} \biggr] (\Delta) f_3^' + \biggl[\frac{1}{2^2\cdot 3} \biggr](\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
|
<math>~ + \biggl[-\frac{3^3}{2^2} \biggr] \biggl\{ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' + \biggl[ \frac{1}{2} \biggr] \Delta^4 f_3^{iv} \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ f_3 + \biggl[ - 3 \biggr] (\Delta) f_3^' + \biggl[ \frac{3^3}{2^3} \biggr] (\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[- \frac{3^2 }{4} \biggr] f_1 + \biggl[ 2\cdot 3^2 \biggr]f_2 + \biggl[ - \frac{3^2 \cdot 7}{4} \biggr] f_3 + \biggl[ \frac{3^3}{2} \biggr] (\Delta) f_3^' + \biggl[\frac{3}{2^2} \biggr](\Delta)^4 f_3^{iv} \biggr\} </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{3^3}{2^2} \biggr] f_1 + \biggl[-3^3 \biggr] f_2 + \biggl[\frac{3^4}{2^2} \biggr]f_3 + \biggl[- \frac{3^3}{2} \biggr] \Delta f_3^' + \biggl[- \frac{3^3}{2^3} \biggr] \Delta^4 f_3^{iv} \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{3^3}{2^2} - \frac{3^2 }{4} \biggr] f_1 + \biggl[ 2\cdot 3^2 -3^3\biggr]f_2 + \biggl[ 1+ \frac{3^4}{2^2} - \frac{3^2 \cdot 7}{4} \biggr] f_3 + \biggl[ \frac{3^3}{2} - \frac{3^3}{2} -3\biggr] (\Delta) f_3^' + \biggl[\frac{3^3}{2^3} + \frac{3}{2^2} - \frac{3^3}{2^3}\biggr](\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{3^2}{2} \biggr] f_1 + \biggl[ - 3^2 \biggr]f_2 + \biggl[ \frac{11}{2} \biggr] f_3 + \biggl[ -3\biggr] (\Delta) f_3^' + \biggl[\frac{3}{2^2} \biggr](\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
<math>~\Rightarrow ~~~ - \biggl[\frac{3}{2^2} \biggr](\Delta)^4 f_3^{iv} </math> |
<math>~=</math> |
<math>~ - f_0 + \biggl[\frac{3^2}{2} \biggr] f_1 + \biggl[ - 3^2 \biggr]f_2 + \biggl[ \frac{11}{2} \biggr] f_3 + \biggl[ -3\biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
Hence,
|
<math>~ \frac{1}{2} (\Delta)^2 f^{}_3 </math> |
<math>~=</math> |
<math>~ - \frac{1}{4} f_1 + 2f_2 + \biggl[ - \frac{7}{4} \biggr] f_3 + \biggl[ \frac{3}{2} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[\frac{1}{2^2\cdot 3} \biggr]\biggl\{ - f_0 + \biggl[\frac{3^2}{2}\biggr] f_1 + \biggl[ - 3^2 \biggr]f_2 + \biggl[ \frac{11}{2} \biggr] f_3 + \biggl[ -3\biggr] (\Delta) f_3^' \biggr\} \biggl[ - \frac{2^2}{3} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ - \frac{1}{4} f_1 + 2f_2 + \biggl[ - \frac{7}{4} \biggr] f_3 + \biggl[ \frac{3}{2} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{1}{2} \biggr] f_1 + f_2 + \biggl[- \frac{11}{2 \cdot 3^2} \biggr] f_3 + \biggl[\frac{1}{3} \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{1}{2} - \frac{1}{4} \biggr] f_1 + 3 f_2 + \biggl[- \frac{11}{2 \cdot 3^2} - \frac{7}{4} \biggr] f_3 + \biggl[\frac{1}{3} + \frac{3}{2} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{3}{4} \biggr] f_1 + 3 f_2 + \biggl[- \frac{5\cdot 17}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{11}{2\cdot 3} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
And,
|
<math>~ \biggl[\frac{2}{3}\biggr] \Delta^3 f_3^{} </math> |
<math>~=</math> |
<math>~ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl[ \frac{1}{2} \biggr] \biggl\{ - f_0 + \biggl[\frac{3^2}{2} \biggr] f_1 + \biggl[ - 3^2 \biggr]f_2 + \biggl[ \frac{11}{2} \biggr] f_3 + \biggl[ -3\biggr] (\Delta) f_3^' \biggr\} \biggl[ - \frac{2^2}{3} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ - f_1 + 2^2f_2 -3 f_3 + 2 \Delta f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[ \frac{2}{3} \biggr] f_0 + \biggl[- 3 \biggr] f_1 + \biggl[ 2\cdot 3 \biggr] f_2 + \biggl[- \frac{11}{3} \biggr] f_3 + \biggl[ 2 \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[ \frac{2}{3} \biggr] f_0 + \biggl[- 4 \biggr] f_1 + \biggl[ 2\cdot 5 \biggr] f_2 + \biggl[- \frac{2^2 \cdot 5}{3} \biggr] f_3 + \biggl[ 4 \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
Finally, then:
|
<math>~f_4</math> |
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \frac{1}{2} ( \Delta)^2 f^{}_3 + \frac{1}{6} (\Delta)^3 f_3^{'} + \frac{1}{24}(\Delta)^4 f_3^{iv} + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \frac{1}{2} \biggl\{ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{3}{4} \biggr] f_1 + 3 f_2 + \biggl[- \frac{5\cdot 17}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{11}{2\cdot 3} \biggr] (\Delta) f_3^' \biggr\}\biggl[ 2 \biggr] </math> |
|
|
|
<math>~ + \frac{1}{2\cdot 3} \biggl\{ \biggl[ \frac{2}{3} \biggr] f_0 + \biggl[- 4 \biggr] f_1 + \biggl[ 2\cdot 5 \biggr] f_2 + \biggl[- \frac{2^2 \cdot 5}{3} \biggr] f_3 + \biggl[ 4 \biggr] (\Delta) f_3^' \biggr\}\biggl[ \frac{3}{2} \biggr] </math> |
|
|
|
<math>~ + \frac{1}{2^3\cdot 3} \biggl\{ - f_0 + \biggl[\frac{3^2}{2} \biggr] f_1 + \biggl[ - 3^2 \biggr]f_2 + \biggl[ \frac{11}{2} \biggr] f_3 + \biggl[ -3\biggr] (\Delta) f_3^' \biggr\}\biggl[- \frac{2^2}{3} \biggr] </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{3}{4} \biggr] f_1 + 3 f_2 + \biggl[- \frac{5\cdot 17}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{11}{2\cdot 3} \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \frac{1}{2^2} \biggl\{ \biggl[ \frac{2}{3} \biggr] f_0 + \biggl[- 4 \biggr] f_1 + \biggl[ 2\cdot 5 \biggr] f_2 + \biggl[- \frac{2^2 \cdot 5}{3} \biggr] f_3 + \biggl[ 4 \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \frac{1}{2\cdot 3^2} \biggl\{ f_0 + \biggl[ - \frac{3^2}{2} \biggr] f_1 + \biggl[ 3^2 \biggr]f_2 + \biggl[- \frac{11}{2} \biggr] f_3 + \biggl[ 3\biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ f_3 + (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
|
<math>~ + \biggl\{ \biggl[\frac{1}{3^2} \biggr] f_0 + \biggl[- \frac{3}{4} \biggr] f_1 + 3 f_2 + \biggl[- \frac{5\cdot 17}{2^2\cdot 3^2} \biggr] f_3 + \biggl[\frac{11}{2\cdot 3} \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ + \biggl\{ \biggl[ \frac{1}{2\cdot 3} \biggr] f_0 + \biggl[- 1 \biggr] f_1 + \biggl[ \frac{5}{2} \biggr] f_2 + \biggl[- \frac{5}{3} \biggr] f_3 + \biggl[ 1 \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
|
<math>~ \biggl\{ \biggl[ \frac{1}{2\cdot 3^2} \biggr] f_0 + \biggl[ - \frac{1}{2^2} \biggr] f_1 + \biggl[ \frac{1}{2} \biggr]f_2 + \biggl[- \frac{11}{2^2\cdot 3^2} \biggr] f_3 + \biggl[ \frac{1}{2\cdot 3} \biggr] (\Delta) f_3^' \biggr\} </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{1}{3^2} + \frac{1}{2\cdot 3} + \frac{1}{2\cdot 3^2} \biggr] f_0 + \biggl[- 1 - \frac{3}{4} - \frac{1}{2^2} \biggr] f_1 + \biggl[ 3 + \frac{5}{2} + \frac{1}{2} \biggr] f_2 + \biggl[1 - \frac{5\cdot 17}{2^2\cdot 3^2} - \frac{5}{3} - \frac{11}{2^2\cdot 3^2} \biggr] f_3 + \biggl[2 + \frac{11}{2\cdot 3} + \frac{1}{2\cdot 3} \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
|
|
<math>~=</math> |
<math>~ \biggl[\frac{1}{3} \biggr] f_0 + \biggl[- 2\biggr] f_1 + \biggl[ 6 \biggr] f_2 + \biggl[- \frac{2\cdot 5}{3} \biggr] f_3 + \biggl[4 \biggr] (\Delta) f_3^' + \mathcal{O}(\Delta^5) </math> |
| <math>~\mathcal{O}(\Delta^5)</math> | |||
|
|
|
|---|
|
© 2014 - 2021 by Joel E. Tohline |