Appendix/Ramblings/PowerSeriesExpressions: Difference between revisions

From JETohlineWiki
Jump to navigation Jump to search
Joel2 (talk | contribs)
Joel2 (talk | contribs)
Line 1,500: Line 1,500:
  <td align="left">
  <td align="left">
<math>
<math>
\frac{r^4}{2^2\cdot 3^2} - r^6 \biggl(\frac{1}{ 3^3\cdot 5}  \biggr)  \, .
\frac{r^4}{2^2\cdot 3^2} - r^6 \biggl(\frac{1}{ 3^3\cdot 5}  \biggr)  \, ;
</math>
  </td>
</tr>
 
<tr>
  <td align="right">
<math>\epsilon^3</math>
  </td>
  <td align="center">
<math>\approx</math>
  </td>
<td align="left">
<math>
-\frac{r^2}{2\cdot 3}\cdot
\frac{r^4}{2^2\cdot 3^2} = - \frac{r^6}{2^3\cdot 3^3}\, .
</math>
</math>
   </td>
   </td>

Revision as of 18:57, 9 October 2026

Approximate Power-Series Expressions

Broadly Used Mathematical Expressions (shown here without proof)

Binomial

(1±x)n

=

1±nx+[n(n−1)2!]x2±[n(n−1)(n−2)3!]x3+[n(n−1)(n−2)(n−3)4!]x4±⋯      for (x2<1)

LaTeX mathematical expressions cut-and-pasted directly from
NIST's Digital Library of Mathematical Functions

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,

(a+b)n

=

an+(n1)an−1b+(n2)an−2b2+…+(nn−1)abn−1+bn,

where, for nonnegative integer values of k and n and k≤n, the notation,

(nk)

=

n!(n−k)!k!=(nn−k).


Our Example:  Setting a=1 gives,

(1+b)n

=

1+(n1)b+(n2)b2+(n3)b3+(n4)b4+…

 

=

1+n!(n−1)!b+n!(n−2)!2!b2+n!(n−3)!3!b3+n!(n−4)!4!b4+…

 

=

1+nb+[n(n−1)2!]b2+[n(n−1)(n−2)3!]b3+[n(n−1)(n−2)(n−3)4!]b4+…


Note, for example, that,

(1+x)−1

=

1−x+x2−x3+x4−x5+⋯;

(1+x)−2

=

1−2x+3x2−4x3+5x4−6x5+⋯;

(1+x)−3

=

1−3x+[3⋅42]x2−[3⋅4⋅52⋅3]x3+[3⋅4⋅5⋅62⋅3⋅4]x4−[3⋅4⋅5⋅6⋅72⋅3⋅4⋅5]x5+⋯

 

=

1−3x+6x2−10x3+15x4−21x5+⋯;

(1+x)−4

=

1−4x+[4⋅52]x2−[4⋅5⋅62⋅3]x3+[4⋅5⋅6⋅72⋅3⋅4]x4−[4⋅5⋅6⋅7⋅82⋅3⋅4⋅5]x5+⋯

 

=

1−4x+10x2−20x3+35x4−56x5+⋯.


See also:

Exponential

ex

=

1+x+x22!+x33!+x44!+x55!+x66!⋯

Expressions with Astrophysical Relevance

Polytropic Lane-Emden Function

Power-Series Derivation

We seek a power-series expression for the polytropic, Lane-Emden function, ΘH(ξ) — expanded about the coordinate center, ξ=0 — that approximately satisfies the Lane-Emden equation,

1ξ2ddξ(ξ2dΘHdξ)=−ΘHn

A general power-series should be of the form,

ΘH

=

θ0+aξ+bξ2+cξ3+dξ4+eξ5+fξ6+gξ7+hξ8+⋯

First derivative:

dΘHdξ

=

a+2bξ+3cξ2+4dξ3+5eξ4+6fξ5+7gξ6+8hξ7+⋯

Left-hand-side of Lane-Emden equation:

1ξ2ddξ(ξ2dΘHdξ)

=

2aξ+2⋅3b+22⋅3cξ+22⋅5dξ2+2⋅3⋅5eξ3+2⋅3⋅7fξ4+23⋅7gξ5+23⋅32hξ6+⋯

Right-hand-side of Lane-Emden equation (adopt the normalization, θ0=1, then use the binomial theorem recursively):

ΘHn

=

1+nF+[n(n−1)2!]F2+[n(n−1)(n−2)3!]F3+[n(n−1)(n−2)(n−3)4!]F4+⋯

where,

F

≡

aξ+bξ2+cξ3+dξ4+eξ5+fξ6+gξ7+hξ8+⋯

 

=

aξ[1+baξ+caξ2+daξ3+eaξ4+faξ5+gaξ6+haξ7+⋯].

First approximation:  Assume that e=f=g=h=0, in which case the LHS contains terms only up through ξ2. This means that we must ignore all terms on the RHS that are of higher order than ξ2; that is,

ΘHn

≈

1+nF+[n(n−1)2!]F2

 

≈

1+n(aξ+bξ2)+[n(n−1)2!]a2ξ2

 

≈

1+naξ+[nb+n(n−1)a22]ξ2.

Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of ξ. Remembering to include a negative sign on the RHS, we find:

Term LHS RHS Implication

ξ−1:

2a

0

⇒a=0

ξ0:

2⋅3b

−1

⇒b=−16

ξ1:

22⋅3c

−na

⇒c=0

ξ2:

22⋅5d

−[nb+n(n−1)a22]

⇒d=+n120

By including higher and higher order terms in the series expansion for ΘH, and proceeding along the same line of deductive reasoning, one finds:

  • Expressions for the four coefficients, a,b,c,d, remain unchanged.
  • The coefficient is zero for all other terms that contain odd powers of ξ; specifically, for example, e=g=0.
  • The coefficients of ξ6 and ξ8 are, respectively,

f

=

−n378(n5−18);

h

=

n(122n2−183n+70)3265920.


In summary, the desired, approximate power-series expression for the polytropic Lane-Emden function is:

For Spherically Symmetric Configurations

ΘH

=

1−ξ26+n120ξ4−n378(n5−18)ξ6+[n(122n2−183n+70)3265920]ξ8+⋯

 

=

1−ξ23!+(n5!)ξ4−n7!(8n−53)ξ6+n9![(122n2−183n+70)9]ξ8+⋯

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.

Theta-Function Examples

When n=0, all of the terms in the power-series expression higher than quadratic go to zero. As a result, we see that

θn=0

=

1−ξ26.


When n=1, the first few terms in the power-series expression are,

θn=1

=

1−ξ26+ξ4120−ξ65040+ξ8362880+⋯

 

=

1−ξ23!+ξ45!−ξ67!+ξ89!+⋯

which is consistent with the power-series expression for (sin⁡ξ)/ξ.


When n=5, the first few terms in the power-series expression are,

θn=5

=

1−ξ23!+(n5!)ξ4−n7!(8n−53)ξ6+n9![(122n2−183n+70)9]ξ8+⋯

 

=

1−ξ23!+ξ44!−ξ66!(523)+ξ88![523⋅723]+⋯

This should be compared with the known analytic solution, which is

(ΘH)n=5

=

[1+ξ23]−1/2.

From the binomial theorem, we start with,

(1+b)m

=

1+mb+[m(m−1)2!]b2+[m(m−1)(m−2)3!]b3+[m(m−1)(m−2)(m−3)4!]b4+…

then set b=ξ2/3 and m=−1/2 to obtain:

θn=5

=

1+m(ξ23)+12![m(m−1)](ξ23)2+13![m(m−1)(m−2)](ξ23)3+14![m(m−1)(m−2)(m−3)](ξ23)4+…

 

=

1+(−12)(ξ23)+12![−12(−12−1)](ξ432)+13![−12(−12−1)(−12−2)](ξ633)+14![−12(−12−1)(−12−2)(−12−3)](ξ834)+…

 

=

1−(ξ22⋅3)+12![34](ξ432)+13![−158](ξ633)+14![(10516)](ξ834)+…

 

=

1−ξ23!+ξ44!−ξ66!(523)+ξ88!(52⋅7232)+…

QED

Isothermal Lane-Emden Function

Derivation

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!), ψ(ξ)↔w(r), the governing ODE is,

d2wdr2+2rdwdr

=

e−w.

A general power-series should be of the form,

w

=

w0+ar+br2+cr3+dr4+er5+fr6+gr7+hr8+⋯

Derivatives:

dwdr

=

a+2br+3cr2+4dr3+5er4+6fr5+7gr6+8hr7+⋯;

d2wdr2

=

2b+2⋅3cr+22⋅3dr2+22⋅5er3+2⋅3⋅5fr4+2⋅3⋅7gr5+23⋅7hr6+⋯.

Put together, then, the left-hand-side of the isothermal Lane-Emden equation becomes:

d2wdr2+2rdwdr

=

2b+2⋅3cr+22⋅3dr2+22⋅5er3+2⋅3⋅5fr4+2⋅3⋅7gr5+23⋅7hr6+2r[a+2br+3cr2+4dr3+5er4+6fr5+7gr6+8hr7]+⋯

 

=

2ar+r0(6b)+r1(22⋅3c)+r2(22⋅3d+23d)+r3(22⋅5e+2⋅5e)+r4(2⋅3⋅5f+22⋅3f)+r5(2⋅3⋅7g+2⋅7g)+r6(23⋅7h+24h)+⋯

Drawing on the above power-series expression for an exponential function, and adopting the convention that w0=0, the right-hand-side becomes,

e−w

=

e0⋅e−ar⋅e−br2⋅e−cr3⋅e−dr4⋅e−er5⋅e−fr6⋅e−gr7⋅e−hr8⋯

 

=

[1−ar+a2r22!−a3r33!+a4r44!−a5r55!+a6r66!+⋯]

 

 

×[1−br2+b2r42!−b3r63!+⋯]×[1−cr3+c2r62!+⋯]×[1−dr4]×[1−er5]×[1−fr6]

 

≈

[1−ar+a2r22−a3r36+a4r424−a5r55⋅24+a6r630⋅24]×[1−cr3+c2r62−br2+bcr5+b2r42−b3r66]×[1−dr4−er5−fr6]

 

≈

{[1−ar+a2r22−a3r36+a4r424−a5r55⋅24+a6r630⋅24]−dr4[1−ar+a2r22]−er5[1−ar]−fr6}

 

 

×[1−br2−cr3+b2r42+bcr5+r6(c22−b36)]

 

≈

[1−ar+a2r22−a3r36+a4r424−a5r55⋅24+a6r630⋅24−dr4+adr5−a2dr62−er5+aer6−fr6]

 

 

×[1−br2−cr3+b2r42+bcr5+r6(c22−b36)]

 

≈

[1−ar+a2r22−a3r36+r4(a424−d)+r5(ad−e−a55⋅24)+r6(a630⋅24−a2d2+ae−f)]×[1−br2−cr3+b2r42+bcr5+r6(c22−b36)]

 

≈

1−ar+a2r22−a3r36+r4(a424−d)+r5(ad−e−a55⋅24)+r6(a630⋅24−a2d2+ae−f)

 

 

−br2[1−ar+a2r22−a3r36+r4(a424−d)]−cr3[1−ar+a2r22−a3r36]+b2r42[1−ar+a2r22]+bcr5[1−ar]+r6(c22−b36)

Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of r. Beginning with the highest order terms, we initially find,

Term LHS RHS Implication

r−1:

2a

0

⇒a=0

r0:

6b

1

⇒b=+16

r1:

22⋅3c

−a

⇒c=−a22⋅3=0

r2:

(22⋅3d+23d)

a22−b

⇒d=120(a22−b)=−1120

With this initial set of coefficient values in hand, we can rewrite (and significantly simplify) our approximate expression for the RHS, namely,

e−w

≈

1−dr4−er5−fr6−br2(1−dr4)+b2r42−b3r66

 

=

1−br2+r4(b22−d)−er5+r6(bd−b36−f).

Continuing, then, with equating terms with like powers on both sides of the equation, we find,

Term LHS RHS Implication

r3:

30e

0

⇒e=0

r4:

(2⋅3⋅5f+22⋅3f)

(b22−d)

⇒f=12⋅3⋅7(123⋅32+123⋅3⋅5)=12⋅33⋅5⋅7

r5:

(2⋅3⋅7g+2⋅7g)

−e

⇒g=0

r6:

(23⋅7h+24h)

(bd−b36−f)

⇒h=−123⋅32(124⋅32⋅5+124⋅34+12⋅33⋅5⋅7)=−6126⋅36⋅5⋅7


Result:

For Spherically Symmetric Configurations

w(r)

=

r26−r4120+r61890−61r81,632,960+⋯.


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, w(r), is presented as equation (56) in a paper by J. P. Ostriker (1964, ApJ, 140, 1056) titled, The Equilibrium of Polytropic and Isothermal Cylinders.

For Later Use

We have just determined that,

e−w

≈

1−br2+r4(b22−d)−er5+r6(bd−b36−f),

where,

a=c=e=g

=

0,

b

=

+12⋅3,

d

=

−123⋅3⋅5,

f

=

+12⋅33⋅5⋅7.

That is to say,

e−w

≈

1−r26+r4(123⋅32+123⋅3⋅5)+r6(−124⋅32⋅5−124⋅34−12⋅33⋅5⋅7)

 

≈

1−r26+r4(523⋅32⋅5+323⋅32⋅5)+r6(−32⋅724⋅34⋅5⋅7−5⋅724⋅34⋅5⋅7−23⋅324⋅34⋅5⋅7)

 

≈

1−r26+r4(132⋅5)−r6(6123⋅34⋅5⋅7).

Therefore, also,

e+w=(e−w)−1

≈

(1+ϵ)−1

 

≈

1−ϵ+ϵ2−ϵ3+ϵ4−ϵ5,

where,

ϵ

≡

−r26+r4(132⋅5)−r6(6123⋅34⋅5⋅7);

ϵ2

=

[−r26+r4(132⋅5)−r6(6123⋅34⋅5⋅7)]⋅[−r26+r4(132⋅5)−r6(6123⋅34⋅5⋅7)]

 

≈

−r26⋅[−r26+r4(132⋅5)]+r4(132⋅5)⋅[−r26]

 

≈

r422⋅32−r6(133⋅5);

ϵ3

≈

−r22⋅3⋅r422⋅32=−r623⋅33.

Displacement Function for Polytropic LAWE

The LAWE for polytropic spheres may be written as,

0

=

d2xdξ2+[4ξ−(n+1)θ(−dθdξ)]dxdξ+(n+1)θ[σc26γ−αξ(−dθdξ)]x

 

=

θd2xdξ2+[4θ−(n+1)ξ(−dθdξ)]1ξdxdξ+(n+1)6[σc2γ−6αξ(−dθdξ)]x,

where, θ(ξ) is the polytropic Lane-Emden function describing the configuration's unperturbed radial density distribution, and γ, σc2, and α≡(3−4/γ) are constants. Here we seek a power-series expression for the displacement function, x(r), 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,

θ

=

1−ξ26+n120ξ4−n378(n5−18)ξ6+⋯

Hence,

−dθdξ

≈

13[ξ−n10ξ3+n21(n5−18)ξ5].

Therefore, near the center of the configuration, the LAWE may be written as,

6θd2xdξ2+{12θ−(n+1)ξ[ξ−n10ξ3+n21(n5−18)ξ5]}2ξdxdξ

≈

−(n+1){σc2γ−2αξ[ξ−n10ξ3+n21(n5−18)ξ5]}x

⇒6[1−ξ26+n120ξ4]d2xdξ2+{12[1−ξ26+n120ξ4]−(n+1)[ξ2−n10ξ4]}2ξdxdξ

≈

−(n+1){𝔉+2α[n10ξ2−n21(n5−18)ξ4]}x

⇒(6−ξ2+n20ξ4)d2xdξ2+[12−(n+3)ξ2+n(n+2)10ξ4]2ξdxdξ

≈

−(n+1)[𝔉+nα5ξ2−2nα21(n5−18)ξ4]x,

where, 𝔉≡(σc2/γ−2α) and, for present purposes, we have kept terms in the series no higher than ξ4.

This is a derivation check ...

[6θ]d2xdξ2+{24[θ]−6(n+1)ξ[−dθdξ]}1ξdxdξ

=

−(n+1){σc2γ−6αξ[−dθdξ]}x

6[1−ξ26+n120ξ4−n378(n5−18)ξ6+⋯]d2xdξ2+{24[1−ξ26+n120ξ4−n378(n5−18)ξ6+⋯]−2(n+1)ξ[ξ−n10ξ3+n21(n5−18)ξ5]}1ξdxdξ

=

−(n+1){σc2γ−2αξ[ξ−n10ξ3+n21(n5−18)ξ5]}x

[6−ξ2+n20ξ4−n63(n5−18)ξ6]d2xdξ2+{12−(n+3)ξ2+[n(n+2)10]ξ4−[n(n+19)7⋅27](8n−540)ξ6}2ξdxdξ

=

−(n+1){𝔉+nα5ξ2−2nα21(8n−540)ξ4}x

Try ⋯

x

=

1+aξ+bξ2+cξ3+dξ4+eξ5+fξ6⋯

dxdξ

=

a+2bξ+3cξ2+4dξ3+5eξ4+6fξ5⋯

⇒2ξ⋅dxdξ

=

2aξ+4b+6cξ+8dξ2+10eξ3+12fξ4⋯

d2xdξ2

=

2b+6cξ+12dξ2+20eξ3+30fξ4+⋯

in which case,

0

≈

[6−ξ2+n20ξ4−n63(n5−18)ξ6][2b+6cξ+12dξ2+20eξ3+30fξ4]

 

 

+{12−(n+3)ξ2+[n(n+2)10]ξ4−[n(n+19)7⋅27](8n−540)ξ6}[2aξ+4b+6cξ+8dξ2+10eξ3+12fξ4]

 

 

+(n+1){𝔉+nα5ξ2−2nα21(8n−540)ξ4}[1+aξ+bξ2+cξ3+dξ4+eξ5+fξ6]

ξ−1:

0=24a

⇒a=0

ξ0:

0=12b+48b+(n+1)𝔉

⇒b=−(n+1)𝔉60

ξ1:

0=36c+72c+a0(n+1)𝔉

⇒c=0

ξ2:

0=72d−2b+96d−4b(n+3)+b(n+1)𝔉+n(n+1)α5=168d+b[−14−4n+(n+1)𝔉]+n(n+1)α5

⇒d=−(n+1){12nα+𝔉[(4n+14)−(n+1)𝔉]10080}


Displacement Finite at Center

Let's adopt a power-series expression for the displacement function of a form that is finite at the center of the configuration, namely,

x

=

1+aξ+bξ2+cξ3+dξ4+eξ5+fξ6⋯

dxdξ

=

a+2bξ+3cξ2+4dξ3+5eξ4+6fξ5⋯

⇒1ξdxdξ

=

aξ+2b+3cξ+4dξ2+5eξ3+6fξ4+⋯

d2xdξ2

=

2b+6cξ+12dξ2+20eξ3+30fξ4+⋯

Substituting these expressions into the LAWE gives,

(6−ξ2+n20ξ4)(2b+6cξ+12dξ2+20eξ3+30fξ4)+[12−(n+3)ξ2+n(n+2)10ξ4](2aξ+4b+6cξ+8dξ2+10eξ3+12fξ4)

≈

−(n+1)[𝔉+nα5ξ2−2nα21(n5−18)ξ4](1+aξ+bξ2+cξ3+dξ4)

Expressions for the various coefficients can now be determined by equating terms on the LHS and RHS that have like powers of ξ.

Term LHS RHS Implication

ξ−1:

24a

0

⇒a=0

ξ0:

(12b+48b)

−(n+1)𝔉

⇒b=−(n+1)𝔉60

ξ1:

[36c+72c−2a(n+3)]

−a(n+1)𝔉

⇒108c=2a(n+3)−a(n+1)𝔉⇒c=0

ξ2:

[72d−2b+96d−4b(n+3)]

[−b(n+1)𝔉−n(n+1)α5]

⇒d=−(n+1){12nα+𝔉[(4n+14)−(n+1)𝔉]10080}

NOTE:  On 9/10/2026, we realized that the first term in the numerator of the expression for the coefficient, d, should be 12nα instead of just nα.

In summary, the desired, approximate power-series expression for the polytropic displacement function is:

x(ξ)

=

1−(n+1)𝔉60ξ2−(n+1){12nα+𝔉[(4n+14)−(n+1)𝔉]10080}ξ4+⋯

Polytropic Displacement-Function Examples

We will only consider situations in which σc2=0 and γ=(n+1)/n, in which case, α=(3−n)/(n+1) and 𝔉=2(n−3)/(n+1).

When n=5 ⇒α=−1/3 and 𝔉=2/3… we have,

x(ξ)

=

1−(660⋅23)ξ2−(6){−20+(2/3)[(34)−4]25⋅32⋅5⋅7}ξ4+⋯

 

=

1−(115)ξ2−0+⋯

Compare this result with the expression for xP|n=5 found in an an accompanying "related thoughts" paragraph.


When n=3 ⇒α=0 and 𝔉=0… we have,

x(ξ)

=

1−(0)ξ2−(6){025⋅32⋅5⋅7}ξ4+⋯

 

=

1+⋯

Compare this result with the expression for xP|n=3 found in an an accompanying "related thoughts" paragraph.


When n=0 ⇒α=+3 and 𝔉=−6… we have,

x(ξ)

=

1+110ξ2+184ξ4+⋯

Sterne (1937) derived analytic expressions for the eigenvectors that describe radial oscillations in uniform-density spheres. The fundamental (j = 0) mode exhibits homologous expansion/contraction, that is to say,

x|Sterne

=

1.

But in his analysis, the accompanying value of 𝔉 is zero; and the accompanying (square of the) dimensionless eigenfrequency is σc2=2(γ−4/3). We cannot adopt the requirement that γ=(n+1)/n because n=0.


Displacement Function for Isothermal LAWE

The LAWE for isothermal spheres may be written as,

d2xdr2+[4−r(dwdr)]1rdxdr

=

−[σc26γ−αr(dwdr)]x,

where, w(r) is the isothermal Lane-Emden function describing the configuration's unperturbed radial density distribution, and γ, σc2, and α≡(3−4/γ) are constants. Here we seek a power-series expression for the displacement function, x(r), 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,

w(r)

=

r26−r4120+r61890−61r81,632,960+⋯.

Hence,

dwdr

≈

r3−r330+r5315.

Therefore, near the center of the configuration, the LAWE may be written as,

d2xdr2+[4−(r23−r430+r6315)]1rdxdr

≈

−16[σc2γ−2α(1−r210+r4105)]x.

Let's now adopt a power-series expression for the displacement function of the form [modified leading term on 9/29/2026],

x

=

x0+ar+br2+cr3+dr4+⋯

⇒dxdr

=

a+2br+3cr2+4dr3+⋯

⇒1rdxdr

=

ar+2b+3cr+4dr2+⋯

and,

d2xdr2

=

2b+6cr+12dr2+⋯

Substituting these expressions into the LAWE gives,

2b+6cr+12dr2+[4−(r23−r430+r6315)][ar+2b+3cr+4dr2]

≈

−16[σc2γ−2α(1−r210+r4105)](x0+ar+br2+cr3+dr4).

Keeping terms only up through r2 leads to the following simplification:

2b+6cr+12dr2+4[ar+2b+3cr+4dr2]−r23[ar+2b]

≈

−16[𝔉+αr25−2αr4105](x0+ar+br2+cr3+dr4)

 

≈

−𝔉6(x0+ar+br2+cr3+dr4)−αr230(x0+ar+br2+cr3+dr4)+αr4315(x0+ar+br2+cr3+dr4)

 

≈

−𝔉6(x0+ar+br2)−αr230(x0).

where,

𝔉≡σc2γ−2α.

Finally, balancing terms of like powers on both sides of the equation leads us to conclude the following:

Term LHS RHS Implication

r−1:

4a

0

⇒a=0

r0:

2b+8b

−𝔉x06

⇒b=−𝔉x060

r1:

6c+12c−a3

−a𝔉6

⇒c=0

r2:

12d+16d−2b3

−𝔉b6−αx030

⇒28d=130[20b−5𝔉b−αx0]⇒d=x010080[𝔉(𝔉−4)−12α]

In summary, the desired, approximate power-series expression for the isothermal displacement function is:

x(r)x0

=

1−𝔉60r2+110080[𝔉(𝔉−4)−12α]r4+⋯

Then, in the particular case where γ=1 and σc2=0 — which implies that α=(3−4/γ)=−1 and 𝔉=(σc2/γ−2α)=+2 — we have,

xY(ξ)

=

1−ξ230+ξ41260+⋯

This is a derivation check ...      (Numerical values in parentheses assume ξ=0.056, or 0.107)

As was discovered by Yabushita, the analytically prescribed isothermal displacement function is,

σc2=0

 and  

xY=1−(1ξe−ψ)dψdξ.

Given that,

ψ(ξ) =

ξ26−ξ4120+ξ61890−61ξ81,632,960+⋯      (5.22585 × 10-4, or 1.907075 × 10-03)⇒eψ=(1.000522, or 1.001908895)

⇒1ξ⋅dψdξ =

13−ξ230+6ξ41890−8⋅61ξ61,632,960+⋯      (3.33229× 10-1, or 0.332952116) ⇒xY=(1 -  0.465010   =   0.34990)     (1−xY)= (xxx, or 0.333587080)

and that,

ψ2 =

[ξ26−ξ4120+ξ61890−61ξ81,632,960]×[ξ26−ξ4120+ξ61890−61ξ81,632,960]+𝒪(ξ10)

  =

[ξ26]×[ξ26−ξ4120+ξ61890]+[−ξ4120]×[ξ26−ξ4120]+[ξ61890]×[ξ26]+𝒪(ξ10)

  =

[ξ46⋅6−ξ66⋅120+ξ86⋅1890]+[−ξ66⋅120+ξ8120⋅120]+[ξ86⋅1890]+𝒪(ξ10)

  =

ξ422⋅32−ξ624⋅32⋅5+ξ822⋅34⋅5⋅7−ξ624⋅32⋅5+ξ826⋅32⋅52+ξ822⋅34⋅5⋅7+𝒪(ξ10)

  =

ξ422⋅32−ξ623⋅32⋅5+ξ82⋅34⋅5⋅7+ξ826⋅32⋅52+𝒪(ξ10)

  =

ξ422⋅32−ξ623⋅32⋅5+ξ826⋅34⋅52⋅7[25⋅5+32⋅7]+𝒪(ξ10)      (2.73095 × 10-07, or 3.636931 × 10-06   vs. from above  2.73095 × 10-7, or 3.636935 × 10-06 )

and that,

ψ3 =

{ξ422⋅32−ξ623⋅32⋅5+ξ826⋅34⋅52⋅7[25⋅5+32⋅7]}×[ξ26−ξ4120+ξ61890−61ξ81,632,960+⋯]+𝒪(ξ10)

  =

ξ623⋅33−ξ822⋅32⋅120−[ξ824⋅33⋅5]+𝒪(ξ10)

  =

ξ623⋅33−ξ825⋅33⋅5−[ξ824⋅33⋅5]+𝒪(ξ10)

  =

ξ623⋅33−ξ825⋅32⋅5+𝒪(ξ10)      (1.42715 × 10-10  , or 6.935893 × 10-09   vs. from above  1.42715 × 10-10  , or 6.935909 × 10-09 )

and that,

ψ4 =

ξ824⋅34+𝒪(ξ10)      (7.46276 × 10-14  , or 1.325760 × 10-11 vs. from above  7.45808 × 10-14  , or 1.322730 × 10-11 )

we appreciate that the exponential function gives,

eψ

=

1+ψ+ψ22!+ψ33!+ψ44!+⋯      (1 + 5.227213 × 10-4    , or 1+1.908894 × 10-03 vs. from above   1.000522 × 100  , or 1+1.908895 × 10-03)

 

=

1+[ξ26−ξ4120+ξ61890−61ξ81,632,960]+12![ξ422⋅32−ξ623⋅32⋅5+ξ826⋅34⋅52⋅7(25⋅5+32⋅7)]

 

 

+13![ξ623⋅33−ξ825⋅32⋅5]+14![ξ824⋅34]+𝒪(ξ10)

 

≈

1+[ξ26−ξ4120+ξ61890]+12![ξ422⋅32−ξ623⋅32⋅5]+13![ξ623⋅33]

 

≈

1+[ξ26−ξ4120+ξ61890]+[ξ423⋅32−ξ624⋅32⋅5]+[ξ624⋅34]

 

≈

1+ξ26+ξ4[123⋅32−123⋅3⋅5]+ξ6[12⋅33⋅5⋅7−124⋅32⋅5+124⋅34]

 

≈

1+ξ26+ξ4[122⋅32⋅5]+ξ6[23⋅324⋅34⋅5⋅7−32⋅724⋅34⋅5⋅7+5⋅724⋅34⋅5⋅7]

 

≈

1+ξ26+ξ4[1180]+ξ6[−445360]      (1 + 5.22721 × 10-4, or 1 + 1.908894 × 10-03)

As a result, the approximate expression for xY is,

xY−1=−[1ξdψdξ]eψ

≈

[−13+ξ22⋅3⋅5−ξ432⋅5⋅7+8⋅61ξ61,632,960]⏞−0.33295212×{1+ξ26+ξ4[122⋅32⋅5]−ξ6[122⋅34⋅5⋅7]}⏟1.001908894≈   -0.333586729

 

≈

−[13]×{1+ξ26+ξ4[122⋅32⋅5]−ξ6[122⋅34⋅5⋅7]}

 

 

+[ξ22⋅3⋅5]×{1+ξ26+ξ4[122⋅32⋅5]}

 

 

−[ξ432⋅5⋅7]×{1+ξ26}+[61ξ6204,120]

 

≈

−13−ξ22⋅32−ξ4[122⋅33⋅5]+ξ6[122⋅35⋅5⋅7]

 

 

+[ξ22⋅3⋅5]+[ξ22⋅3⋅5]ξ26+[ξ22⋅3⋅5][ξ422⋅32⋅5]

 

 

−[ξ432⋅5⋅7]−[ξ62⋅33⋅5⋅7]+[61ξ623⋅35⋅5⋅7]

 

≈

−13+ξ2[−12⋅32]+ξ4[−122⋅33⋅5]+ξ6[122⋅35⋅5⋅7]+ξ2[12⋅3⋅5]+ξ4[122⋅32⋅5]+ξ6[123⋅33⋅52]+ξ4[−132⋅5⋅7]+ξ6[−12⋅33⋅5⋅7]+ξ6[6123⋅35⋅5⋅7]

Maclaurin Spheroid Index Symbols

In our accompanying discussion of the equilibrium properties of models along the Maclaurin spheroid sequence, we find the "Index Symbols" expressions,

A1

=

1e2[sin−1ee−(1−e2)1/2](1−e2)1/2,

A3

=

2e2[(1−e2)−1/2−sin−1ee](1−e2)1/2,

where,

e

≡

[1−(ca)2]1/2       (always positive).

Our aim, here, is to derive a power-series expression for these two index symbols (a) in the case of nearly spherical configurations (e≪1), and (b) in the case of an infinitesimally thin disk (c/a≪1).

Nearly Spherical Configurations

On p. 457 of [CRC], we find that,

sin−1ee

=

1+[12⋅3]e2+[1⋅32⋅4⋅5]e4+[1⋅3⋅52⋅4⋅6⋅7]e6+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9]e8+⋯       for,       (e2<1,−π2<sin−1e<π2).

Also, from the above binomial-theorem expression, we have,

(1−e2)1/2

=

1−12e2+[12(12−1)2!]e4−[12(12−1)(12−2)3!]e6+[12(12−1)(12−2)(12−3)4!]e8−⋯      for (e4<1)

 

=

1−[12]e2−[123]e4−[124]e6−[527]e8−⋯

So we can write,

ℱ1≡(1−e2)1/2[sin−1ee]

=

{1−[12]e2−[123]e4−[124]e6−[527]e8−⋯}×{1+[12⋅3]e2+[1⋅32⋅4⋅5]e4+[1⋅3⋅52⋅4⋅6⋅7]e6+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9]e8+⋯}

 

=

{1+[12⋅3]e2+[1⋅32⋅4⋅5]e4+[1⋅3⋅52⋅4⋅6⋅7]e6+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9]e8}−e22{1+[12⋅3]e2+[1⋅32⋅4⋅5]e4+[1⋅3⋅52⋅4⋅6⋅7]e6}

 

 

−e423{1+[12⋅3]e2+[1⋅32⋅4⋅5]e4}−e624{1+[12⋅3]e2}−[527]e8+𝒪(e10)

 

=

1+e2[12⋅3−12]+e4[1⋅32⋅4⋅5−122⋅3−123]+e6[1⋅3⋅52⋅4⋅6⋅7−12⋅1⋅32⋅4⋅5−123⋅12⋅3−124]

 

 

+e8[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9−12⋅1⋅3⋅52⋅4⋅6⋅7−123⋅1⋅32⋅4⋅5−124⋅12⋅3−527]+𝒪(e10)

 

=

1−e2[13]+e4[32−2⋅5−3⋅523⋅3⋅5]+e6[3⋅52−32⋅7−5⋅7−3⋅5⋅724⋅3⋅5⋅7]+e8[52⋅72−22⋅32⋅52−2⋅33⋅7−22⋅3⋅5⋅7−32⋅52⋅727⋅32⋅5⋅7]+𝒪(e10)

 

=

1−e2[13]−e4[23⋅5]−e6[233⋅5⋅7]−e8[2432⋅5⋅7]+𝒪(e10)

Hence,

e2A1

=

ℱ1−(1−e2)

 

=

e2[23]−e4[23⋅5]−e6[233⋅5⋅7]−e8[2432⋅5⋅7]+𝒪(e10)

⇒A1

=

23−e2[23⋅5]−e4[233⋅5⋅7]−e6[2432⋅5⋅7]+𝒪(e8).

And,

(e22)A3

=

1−ℱ1

 

=

e2[13]+e4[23⋅5]+e6[233⋅5⋅7]+e8[2432⋅5⋅7]+𝒪(e10)

⇒A3

=

23+e2[223⋅5]+e4[243⋅5⋅7]+e6[2532⋅5⋅7]+𝒪(e8).

This looks okay, in the sense that (2A1+A3)=2.

Infinitesimally Thin Axisymmetric Disk

As e→1 — that is, in the case of an infinitesimally thin, axisymmetric disk — the preferred small parameter is,

ca

=

(1−e2)1/2≪1.

Recognizing as well that,

sin−1e

=

cos−1(1−e2)1/2=cos−1(ca)

⇒ℱ2≡sin−1ee3⋅(1−e2)1/2

=

(ca)[cos−1(ca)][1−c2a2]−3/2,

the expressions for the pair of relevant index symbols may be rewritten as,

A1

=

ℱ2−(ca)2(1−c2a2)−1,

A32

=

(1−c2a2)−1−ℱ2.

Pulling again from p. 457 of [CRC], we find that,

cos−1(ca)

=

π2−ca{1+[12⋅3](ca)2+[1⋅32⋅4⋅5](ca)4+[1⋅3⋅52⋅4⋅6⋅7](ca)6+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)8+⋯}       for,       (c2a2<1,0<cos−1(ca)<π).

LAGNIAPPE:

According to the above binomial-theorem expression, we find for (c2/a2<1),

1e=[1−c2a2]−1/2

=

1+12(ca)2+[−12(−12−1)2!](ca)4−[−12(−12−1)(−12−2)3!](ca)6+[−12(−12−1)(−12−2)(−12−3)4!](ca)8−⋯

 

=

1+12(ca)2+[322⋅2!](ca)4+[3⋅523⋅3!](ca)6+[3⋅5⋅724⋅4!](ca)8+⋯

 

=

1+12(ca)2+[323](ca)4+[524](ca)6+[5⋅727](ca)8+⋯

Hence,

sin−1ee=cos−1(ca)[1−c2a2]−1/2

=

π2{1+12(ca)2+[323](ca)4+[524](ca)6+[5⋅727](ca)8+⋯}

 

 

−ca{1+[12⋅3](ca)2+[1⋅32⋅4⋅5](ca)4+[1⋅3⋅52⋅4⋅6⋅7](ca)6+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)8+⋯}

 

 

×{1+12(ca)2+[323](ca)4+[524](ca)6+[5⋅727](ca)8+⋯}

 

=

π2{1+[12](ca)2+[323](ca)4+[524](ca)6+[5⋅727](ca)8}

 

 

−{(ca)+[12⋅3](ca)3+[1⋅32⋅4⋅5](ca)5+[1⋅3⋅52⋅4⋅6⋅7](ca)7+[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)9}

 

 

−12{(ca)3+[12⋅3](ca)5+[1⋅32⋅4⋅5](ca)7+[1⋅3⋅52⋅4⋅6⋅7](ca)9}

 

 

−[323]{(ca)5+[12⋅3](ca)7+[1⋅32⋅4⋅5](ca)9}−[524]{(ca)7+[12⋅3](ca)9}−[5⋅727](ca)9+𝒪(c10a10)

 

=

π2+π2[12](ca)2+π2[323](ca)4+π2[524](ca)6+π2[5⋅727](ca)8

 

 

−(ca)−[12⋅3](ca)3−[1⋅32⋅4⋅5](ca)5−[1⋅3⋅52⋅4⋅6⋅7](ca)7−[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)9

 

 

−12(ca)3−[122⋅3](ca)5−[1⋅322⋅4⋅5](ca)7−[1⋅3⋅522⋅4⋅6⋅7](ca)9

 

 

−[323](ca)5−[323][12⋅3](ca)7−[323][1⋅32⋅4⋅5](ca)9−[524](ca)7−[524][12⋅3](ca)9−[5⋅727](ca)9+𝒪(c10a10).

(continue expression simplification)

sin−1ee=cos−1(ca)[1−c2a2]−1/2

=

π2+π2[12](ca)2+π2[323](ca)4+π2[524](ca)6+π2[5⋅727](ca)8

 

 

−(ca)−[12⋅3](ca)3−[323⋅5](ca)5−[524⋅7](ca)7−[5⋅727⋅32](ca)9

 

 

−12(ca)3−[122⋅3](ca)5−[324⋅5](ca)7−[525⋅7](ca)9

 

 

−[323](ca)5−[124](ca)7−[3226⋅5](ca)9−[524](ca)7−[525⋅3](ca)9−[5⋅727](ca)9+𝒪(c10a10)

 

=

π2+π2[12](ca)2+π2[323](ca)4+π2[524](ca)6+π2[5⋅727](ca)8

 

 

−(ca)−{[12⋅3]+12}(ca)3−{[323⋅5]+[122⋅3]+[323]}(ca)5−{[524⋅7]+[324⋅5]+[124] +[524]}(ca)7

 

 

−{[5⋅727⋅32]+[525⋅7]+[3226⋅5]+[525⋅3]+[5⋅727]}(ca)9+𝒪(c10a10)

 

=

π2+π2[12](ca)2+π2[323](ca)4+π2[524](ca)6+π2[5⋅727](ca)8

 

 

−(ca)−[23](ca)3−[233⋅5](ca)5−[245⋅7](ca)7−[2732⋅5⋅7](ca)9+𝒪(c10a10).


Referring again to the above binomial-theorem expression, we find for (c2/a2<1),

[1−c2a2]−3/2

=

1+32(ca)2+[−32(−32−1)2!](ca)4−[−32(−32−1)(−32−2)3!](ca)6+[−32(−32−1)(−32−2)(−32−3)4!](ca)8−⋯

 

=

1+32(ca)2+[32(32+1)2!](ca)4+[32(32+1)(32+2)3!](ca)6+[32(32+1)(32+2)(32+3)4!](ca)8+⋯

 

=

1+32(ca)2+[3⋅522⋅2!](ca)4+[3⋅5⋅723⋅3!](ca)6+[3⋅5⋅7⋅924⋅4!](ca)8+⋯

 

=

1+32(ca)2+[3⋅523](ca)4+[5⋅724](ca)6+[5⋅7⋅927](ca)8+⋯

We therefore can write,

ℱ2

=

{π2(ca)−(ca)2−[12⋅3](ca)4−[1⋅32⋅4⋅5](ca)6−[1⋅3⋅52⋅4⋅6⋅7](ca)8−[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)10−⋯}

 

 

×{1+32(ca)2+[3⋅523](ca)4+[5⋅724](ca)6+[5⋅7⋅927](ca)8+⋯}

 

=

π2(ca)−(ca)2−[12⋅3](ca)4−[1⋅32⋅4⋅5](ca)6−[1⋅3⋅52⋅4⋅6⋅7](ca)8−[1⋅3⋅5⋅72⋅4⋅6⋅8⋅9](ca)10

 

 

+32(ca)2{π2(ca)−(ca)2−[12⋅3](ca)4−[1⋅32⋅4⋅5](ca)6−[1⋅3⋅52⋅4⋅6⋅7](ca)8}

 

 

+[3⋅523](ca)4{π2(ca)−(ca)2−[12⋅3](ca)4−[1⋅32⋅4⋅5](ca)6}

 

 

+[5⋅724](ca)6{π2(ca)−(ca)2−[12⋅3](ca)4}+[5⋅7⋅927](ca)8{π2(ca)−(ca)2}+𝒪(c10a10)

 

=

π2(ca)−(ca)2−[12⋅3](ca)4−[1⋅32⋅4⋅5](ca)6−[1⋅3⋅52⋅4⋅6⋅7](ca)8+π[322](ca)3−[32](ca)4−[122](ca)6−[3224⋅5](ca)8

 

 

+π[3⋅524](ca)5−[3⋅523](ca)6−[524](ca)8+π[5⋅725](ca)7−[5⋅724](ca)8+π[5⋅7⋅928](ca)9+𝒪(c10a10)

 

=

π2(ca)−(ca)2+π[322](ca)3−[12⋅3+32](ca)4+π[3⋅524](ca)5−[1⋅32⋅4⋅5+122+3⋅523](ca)6

 

 

+π[5⋅725](ca)7−[1⋅3⋅52⋅4⋅6⋅7+3224⋅5+524+5⋅724](ca)8+π[5⋅7⋅928](ca)9+𝒪(c10a10)

 

=

π2(ca)−(ca)2+π[322](ca)3−[53](ca)4+π[3⋅524](ca)5−[115](ca)6+π[5⋅725](ca)7−[935⋅7](ca)8+π[5⋅7⋅928](ca)9+𝒪(c10a10).

Once again from the binomial theorem,

(1−c2a2)−1

=

1+(ca)2+(ca)4+(ca)6+(ca)8+(ca)10+⋯

which gives us,

A1=ℱ2−(ca)2(1−c2a2)−1

=

π2(ca)−(ca)2+π[322](ca)3−[53](ca)4+π[3⋅524](ca)5−[115](ca)6+π[5⋅725](ca)7−[935⋅7](ca)8+π[32⋅5⋅728](ca)9

 

 

−{(ca)2+(ca)4+(ca)6+(ca)8}+𝒪(c10a10)

 

=

π2(ca)−2(ca)2+π[322](ca)3−[83](ca)4+π[3⋅524](ca)5−[245](ca)6+π[5⋅725](ca)7−[275⋅7](ca)8+π[32⋅5⋅728](ca)9+𝒪(c10a10).

And,

A32=(1−c2a2)−1−ℱ2

=

{1+(ca)2+(ca)4+(ca)6+(ca)8+(ca)10+⋯}

 

 

−{π2(ca)−(ca)2+π[322](ca)3−[53](ca)4+π[3⋅524](ca)5−[115](ca)6+π[5⋅725](ca)7−[935⋅7](ca)8+π[32⋅5⋅728](ca)9}+𝒪(c10a10)

 

=

1−π2(ca)+2(ca)2−π[322](ca)3+[83](ca)4−π[3⋅524](ca)5+[245](ca)6−π[5⋅725](ca)7+[275⋅7](ca)8−π[32⋅5⋅728](ca)9}+𝒪(c10a10).

Notice that, to the highest order retained in these expressions, we find as expected that, (A1+A3/2)=1.

Frequency (temporary)

ω02

=

2πGρ[A1−A3(1−e2)]

⇒ΩMc2≡ω02πGρ

=

2(3−2e2)(1−e2)1/2⋅sin−1ee3−6(1−e2)e2,

Taylor Series (Hunter77)

First (Unsuccessful) Try

First:

f0

=

f3+(−3Δ)f3'+12(−3Δ)2f3'′+16(−3Δ)3f3'″+124(−3Δ)4f3iv+𝒪(Δ5)

 

=

f3−(3Δ)f3'+322(Δ)2f3'′−322(Δ)3f3'″+3323(Δ)4f3iv+𝒪(Δ5)

⇒−322(Δ)2f3'′

=

f3−f0−(3Δ)f3'−322(Δ)3f3'″+3323(Δ)4f3iv+𝒪(Δ5)

Note that, replacing the (Δ)3f3'″ term with the expression derived in the Second step, below, gives,

−322(Δ)2f3'′

=

f3−f0−(3Δ)f3'+3323(Δ)4f3iv+𝒪(Δ5)

 

 

−322{[2232]f0−f1+f3[532]+[−23](Δ)f3'+[−56](Δ)4f3iv}[−32]

 

=

f3−f0−3(Δ)f3'+3323(Δ)4f3iv+𝒪(Δ5)

 

 

+{3f0+[−3322]f1+[1522]f3+[−32](Δ)f3'+[−32⋅523](Δ)4f3iv}

 

=

2f0+[−3322]f1+[1+1522]f3+[−3−32](Δ)f3'+[3323−32⋅523](Δ)4f3iv+𝒪(Δ5)

 

=

2f0+[−3322]f1+[1922]f3+[−92](Δ)f3'+[−94](Δ)4f3iv+𝒪(Δ5)

Then, replacing the (Δ)4f3iv term with the expression derived in the Third step, below, gives,

−322(Δ)2f3'′

=

2f0+[−3322]f1+[1922]f3+[−92](Δ)f3'+𝒪(Δ5)

 

 

+[−94]{[−132]f0+[12]f1−f2+[112⋅32]f3+[−23](Δ)f3'}[−22⋅3]

 

=

2f0+[−3322]f1+[1922]f3+[−92](Δ)f3'+𝒪(Δ5)

 

 

+{[−3]f0+[332]f1−33f2+[3⋅112]f3+[−2⋅32](Δ)f3'}

 

=

−f0+[332−3322]f1−33f2+[3⋅112+1922]f3+[−2⋅32−92](Δ)f3'+𝒪(Δ5)

 

=

−f0+[3322]f1−33f2+[5⋅1722]f3+[−32⋅52](Δ)f3'+𝒪(Δ5)


Second:

f1

=

f3+(−2Δ)f3'+12(−2Δ)2f3'′+16(−2Δ)3f3'″+124(−2Δ)4f3iv+𝒪(Δ5)

 

=

f3−2(Δ)f3'+2(Δ)2f3'′−223(Δ)3f3'″+23(Δ)4f3iv+𝒪(Δ5)

 

=

f3−2(Δ)f3'−223(Δ)3f3'″+23(Δ)4f3iv+𝒪(Δ5)

 

 

−[2232][f3−f0−(3Δ)f3'−322(Δ)3f3'″+3323(Δ)4f3iv+𝒪(Δ5)]

 

=

[2232]f0+f3[1−2232]+[2232(3Δ)−2(Δ)]f3'+[(2232)322(Δ)3−223(Δ)3]f3'″

 

 

+[23(Δ)4−(2232)3323(Δ)4]f3iv+𝒪(Δ5)

 

=

[2232]f0+f3[532]+[−23](Δ)f3'+[23](Δ)3f3'″+[−56](Δ)4f3iv+𝒪(Δ5)

⇒−[23](Δ)3f3'″

=

[2232]f0−f1+f3[532]+[−23](Δ)f3'+[−56](Δ)4f3iv+𝒪(Δ5)

Now, replacing the (Δ)4f3iv term with the expression derived in the Third step, below, gives,

−[23](Δ)3f3'″

=

[2232]f0−f1+f3[532]+[−23](Δ)f3'+𝒪(Δ5)

 

 

+[−56]{[−132]f0+[12]f1−f2+[112⋅32]f3+[−23](Δ)f3'}[−22⋅3]

 

=

[2232]f0−f1+f3[532]+[−23](Δ)f3'+𝒪(Δ5)

 

 

+{[−2⋅532]f0+[5]f1+[−2⋅5]f2+[5⋅1132]f3+[−22⋅53](Δ)f3'}

 

=

[2232−2⋅532]f0+[4]f1+[−2⋅5]f2+[532+5⋅1132]f3+[−22⋅53−23](Δ)f3'+𝒪(Δ5)

 

=

[−23]f0+[4]f1+[−2⋅5]f2+[22⋅53]f3+[−2⋅113](Δ)f3'+𝒪(Δ5)


Third:

f2

=

f3+(−Δ)f3'+12(−Δ)2f3'′+16(−Δ)3f3'″+124(−Δ)4f3iv+𝒪(Δ5)

 

=

f3+[−1](Δ)f3'+[12](Δ)2f3'′+[−12⋅3](Δ)3f3'″+[123⋅3](Δ)4f3iv+𝒪(Δ5)

 

=

f3+[−1](Δ)f3'+[123⋅3](Δ)4f3iv+𝒪(Δ5)

 

 

+[12]{2f0+[−3322]f1+[1922]f3+[−92](Δ)f3'+[−94](Δ)4f3iv}[−232]

 

 

+[−12⋅3]{[2232]f0−f1+f3[532]+[−23](Δ)f3'+[−56](Δ)4f3iv}[−32]

 

=

f3+[−1](Δ)f3'+[123⋅3](Δ)4f3iv+𝒪(Δ5)

 

 

+{[−232]f0+[322]f1+[−1922⋅32]f3+[12](Δ)f3'+[14](Δ)4f3iv}

 

 

+{[132]f0+[−122]f1+f3[522⋅32]+[−12⋅3](Δ)f3'+[−523⋅3](Δ)4f3iv}

 

=

f3+[−1](Δ)f3'+[123⋅3](Δ)4f3iv+𝒪(Δ5)

 

 

+{[132−232]f0+[322−122]f1+[522⋅32−1922⋅32]f3+[12−12⋅3](Δ)f3'+[14−523⋅3](Δ)4f3iv}

 

=

[−132]f0+[12]f1+[112⋅32]f3+[−23](Δ)f3'+[122⋅3](Δ)4f3iv+𝒪(Δ5)

⇒−[122⋅3](Δ)4f3iv

=

[−132]f0+[12]f1−f2+[112⋅32]f3+[−23](Δ)f3'+𝒪(Δ5)

And, finally:

f4

=

f3+(Δ)f3'+12(Δ)2f3'′+16(Δ)3f3'″+124(Δ)4f3iv+𝒪(Δ5)

 

=

f3+(Δ)f3'+𝒪(Δ5)

 

 

+12{−f0+[3322]f1−33f2+[5⋅1722]f3+[−32⋅52](Δ)f3'}[−232]

 

 

+16{[−23]f0+[4]f1+[−2⋅5]f2+[22⋅53]f3+[−2⋅113](Δ)f3'}[−32]

 

 

+124{[−132]f0+[12]f1−f2+[112⋅32]f3+[−23](Δ)f3'}[−22⋅3]

 

=

f3+(Δ)f3'+𝒪(Δ5)

 

 

+{[132]f0+[−322]f1+[3]f2+[−5⋅1722⋅32]f3+[52](Δ)f3'}

 

 

+{[12⋅3]f0+[−1]f1+[52]f2+[−53]f3+[112⋅3](Δ)f3'}

 

 

+{[12⋅32]f0+[−122]f1+[12]f2+[−1122⋅32]f3+[13](Δ)f3'}

 

=

[132+12⋅3+12⋅32]f0+[−322−1−122]f1+[3+52+12]f2+𝒪(Δ5)

 

 

+[1−5⋅1722⋅32−53−1122⋅32]f3+[1+52+112⋅3+13](Δ)f3'

 

=

[13]f0+[−2]f1+[6]f2+[−103]f3+[173](Δ)f3'+𝒪(Δ5)

Result:

Definitely WRONG!

f4

=

13f0−2f1+6f2−103f3+173(Δ)f3'+𝒪(Δ5).

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 (Δ)f3', namely,

Somewhat Improved

f4

=

13f0−2f1+6f2−103f3+4(Δ)f3'+𝒪(Δ5).

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 f4 is the variable that we desire to evaluate, and the "known" quantities are:   f3, f3', f2, f1, and f0.

f4

=

f3+(Δ)f3'+12(Δ)2f3'′+16(Δ)3f3'″+124(Δ)4f3iv+𝒪(Δ5)

Let's use similar Taylor-series expansions for f2, f3, etc. in order to eliminate the f3'′ term, the f3'″ term, etc.

f2

=

f3+(−Δ)f3'+12(−Δ)2f3'′+16(−Δ)3f3'″+124(−Δ)4f3iv+𝒪(Δ5)

f1

=

f3+(−2Δ)f3'+12(−2Δ)2f3'′+16(−2Δ)3f3'″+124(−2Δ)4f3iv+𝒪(Δ5)

f0

=

f3+(−3Δ)f3'+12(−3Δ)2f3'′+16(−3Δ)3f3'″+124(−3Δ)4f3iv+𝒪(Δ5)


First:

−12(−Δ)2f3'′

=

f3+(−Δ)f3'−f2+16(−Δ)3f3'″+124(−Δ)4f3iv+𝒪(Δ5)

⇒12(Δ)2f3'′

=

−f3+(Δ)f3'+f2+16(Δ)3f3'″−124(Δ)4f3iv+𝒪(Δ5)

⇒f4

=

f3+(Δ)f3'−f3+(Δ)f3'+f2+𝒪(Δ3)

 

=

f2+2(Δ)f3'+𝒪(Δ3)

𝒪(Δ3)

f4

=

f2+2(Δ)f3'+𝒪(Δ3)

This expression works very well for a parabola.


Second:

f1

=

f3+(−2)Δf3'+2(Δ)2f3'′+[−236]Δ3f3'″+[2423⋅3]Δ4f3iv+𝒪(Δ5)

 

=

f3+(−2)Δf3'+[−236]Δ3f3'″+[2423⋅3]Δ4f3iv+𝒪(Δ5)

 

 

+2{−f3+(Δ)f3'+f2+12⋅3(Δ)3f3'″−123⋅3(Δ)4f3iv}[2]

 

=

f3[1−22]+(22−2)Δf3'+22f2+[23−236]Δ3f3'″+[2423⋅3−12⋅3]Δ4f3iv+𝒪(Δ5)

 

=

f3[−3]+(2)Δf3'+22f2+[−23]Δ3f3'″+[12]Δ4f3iv+𝒪(Δ5)

⇒[23]Δ3f3'″

=

−f1+22f2−3f3+2Δf3'+[12]Δ4f3iv+𝒪(Δ5)

This also allows us to improve the expression for the f3'′ term, as initially derived in the "First" subsection, above. Namely,

12(Δ)2f3'′

=

f2−f3+(Δ)f3'−124(Δ)4f3iv+𝒪(Δ5)

 

 

+16{−f1+22f2−3f3+2Δf3'+[12]Δ4f3iv}[32]

 

=

−14f1+2f2+[−74]f3+[32](Δ)f3'+[122⋅3](Δ)4f3iv+𝒪(Δ5)

Hence, an improved expression for f4 is,

f4

=

f3+(Δ)f3'+𝒪(Δ4)

 

 

+{−14f1+2f2+[−74]f3+[32](Δ)f3'}

 

 

+16{−f1+22f2−3f3+2Δf3'}[32]

 

=

−12f1+3f2−32f3+3(Δ)f3'+𝒪(Δ4)


𝒪(Δ4)

f4

=

−12f1+3f2−32f3+3(Δ)f3'+𝒪(Δ4)



Third:

f0

=

f3+(−3Δ)f3'+12(−3Δ)2f3'′+16(−3Δ)3f3'″+124(−3Δ)4f3iv+𝒪(Δ5)

 

=

f3+[−3](Δ)f3'+[3323](Δ)4f3iv+𝒪(Δ5)

 

 

+32{−14f1+2f2+[−74]f3+[32](Δ)f3'+[122⋅3](Δ)4f3iv}

 

 

+[−3322]{−f1+22f2−3f3+2Δf3'+[12]Δ4f3iv}

 

=

f3+[−3](Δ)f3'+[3323](Δ)4f3iv+𝒪(Δ5)

 

 

+{[−324]f1+[2⋅32]f2+[−32⋅74]f3+[332](Δ)f3'+[322](Δ)4f3iv}

 

 

+{[3322]f1+[−33]f2+[3422]f3+[−332]Δf3'+[−3323]Δ4f3iv}

 

=

[3322−324]f1+[2⋅32−33]f2+[1+3422−32⋅74]f3+[332−332−3](Δ)f3'+[3323+322−3323](Δ)4f3iv+𝒪(Δ5)

 

=

[322]f1+[−32]f2+[112]f3+[−3](Δ)f3'+[322](Δ)4f3iv+𝒪(Δ5)

⇒−[322](Δ)4f3iv

=

−f0+[322]f1+[−32]f2+[112]f3+[−3](Δ)f3'+𝒪(Δ5)

Hence,

12(Δ)2f3'′

=

−14f1+2f2+[−74]f3+[32](Δ)f3'+𝒪(Δ5)

 

 

+[122⋅3]{−f0+[322]f1+[−32]f2+[112]f3+[−3](Δ)f3'}[−223]

 

=

−14f1+2f2+[−74]f3+[32](Δ)f3'+𝒪(Δ5)

 

 

+{[132]f0+[−12]f1+f2+[−112⋅32]f3+[13](Δ)f3'}

 

=

[132]f0+[−12−14]f1+3f2+[−112⋅32−74]f3+[13+32](Δ)f3'+𝒪(Δ5)

 

=

[132]f0+[−34]f1+3f2+[−5⋅1722⋅32]f3+[112⋅3](Δ)f3'+𝒪(Δ5)

And,

[23]Δ3f3'″

=

−f1+22f2−3f3+2Δf3'+𝒪(Δ5)

 

 

+[12]{−f0+[322]f1+[−32]f2+[112]f3+[−3](Δ)f3'}[−223]

 

=

−f1+22f2−3f3+2Δf3'+𝒪(Δ5)

 

 

+{[23]f0+[−3]f1+[2⋅3]f2+[−113]f3+[2](Δ)f3'}

 

=

[23]f0+[−4]f1+[2⋅5]f2+[−22⋅53]f3+[4](Δ)f3'+𝒪(Δ5)


Finally, then:

f4

=

f3+(Δ)f3'+12(Δ)2f3'′+16(Δ)3f3'″+124(Δ)4f3iv+𝒪(Δ5)

 

=

f3+(Δ)f3'+𝒪(Δ5)

 

 

+12{[132]f0+[−34]f1+3f2+[−5⋅1722⋅32]f3+[112⋅3](Δ)f3'}[2]

 

 

+12⋅3{[23]f0+[−4]f1+[2⋅5]f2+[−22⋅53]f3+[4](Δ)f3'}[32]

 

 

+123⋅3{−f0+[322]f1+[−32]f2+[112]f3+[−3](Δ)f3'}[−223]

 

=

f3+(Δ)f3'+𝒪(Δ5)

 

 

+{[132]f0+[−34]f1+3f2+[−5⋅1722⋅32]f3+[112⋅3](Δ)f3'}

 

 

+122{[23]f0+[−4]f1+[2⋅5]f2+[−22⋅53]f3+[4](Δ)f3'}

 

 

+12⋅32{f0+[−322]f1+[32]f2+[−112]f3+[3](Δ)f3'}

 

=

f3+(Δ)f3'+𝒪(Δ5)

 

 

+{[132]f0+[−34]f1+3f2+[−5⋅1722⋅32]f3+[112⋅3](Δ)f3'}

 

 

+{[12⋅3]f0+[−1]f1+[52]f2+[−53]f3+[1](Δ)f3'}

 

 

{[12⋅32]f0+[−122]f1+[12]f2+[−1122⋅32]f3+[12⋅3](Δ)f3'}

 

=

[132+12⋅3+12⋅32]f0+[−1−34−122]f1+[3+52+12]f2+[1−5⋅1722⋅32−53−1122⋅32]f3+[2+112⋅3+12⋅3](Δ)f3'+𝒪(Δ5)

 

=

[13]f0+[−2]f1+[6]f2+[−2⋅53]f3+[4](Δ)f3'+𝒪(Δ5)


𝒪(Δ5)

f4

=

13f0−2f1+6f2−2⋅53f3+4(Δ)f3'+𝒪(Δ5)


Tiled Menu

Appendices: | VisTrailsEquations | VisTrailsVariables | References | Ramblings | VisTrailsImages | myphys.lsu | ADS |