Vous êtes sur la page 1sur 31

Third Order O.D.E.

's And Cubic Complex Numbers: Possible

Applications By Ronald H. Brady

Table Of Contents

Topic Page No.

Abstract 2

Second and Third Order O.D.E.’s 3

Cubic Complex Numbers 5

The Cubic Complex Exponential Function 11

Norms and Pseudo Norms 22

Kinematics Of Simple Harmonic Motion And Damped

Harmonic Motion 23
Abstract

It is well known that a second order O.D.E. of the form

y’’ + y = 0, where the primes indicate differentiation with

respect to the independent variable (which may be chosen to be

time), may be used, with an appropriate choice of units, to

describe simple harmonic motion in one spatial dimension. If we

consider y to be a function of x, y = y(x), then the solutions

of that equation involves the periodic functions sin(x) and

cos(x). These functions are related, via Euler’ formula,

eix = cos(x) + i*sin(x) .

to the Euler complex exponential function eix . The pure imaginary

unit i, of course, also plays a prominent role in the formula.

The present paper will present a derivation of a 3rd order

O.D.E. that is applicable to damped simple harmonic motion. The

solutions to that 3rd order O.D.E. will be shown to be related to

a generalization of Euler’s formula. This generalization, in

turn, will be based upon a natural generalization of complex

numbers from the plane to a Commutative Algebra in three

dimensions.

The presentation of potentially new identities and

symmetries, rather than formal mathematical rigor, is the

primary purpose of this paper. Therefore the style of


presentation will be informal. All of the indicated derivatives,

of all of the functions under discussion, are assumed to exist.

Second and Third Order O.D.E.’s

Now let us turn our attention to the general homogeneous

second-order ordinary differential equation

y’’ + By’ + Cy = 0 (eq.1)

where the coefficients B and C are constant. In particular

if B = 0 and C = 1, then (eq.1) reduces to

y’’ + y = 0 (eq.2)

As is well known, two (linearly) independent solutions

(eq.2) are cos(x) and sin(x). These functions are as-

sociated with many identities such as

cos2(x) + sin2(x) = 1. (eq.3)

They also appear in the fundamental identity

eix = cos(x) + i*sin(x) (eq.4)

(known as Eulers’ Formula). Equations (eq.3) and (eq.4),

play indispensable roles in math, science,

engineering and theoretical physics. The reader will

recall that since eix is a linear combination of cos(x)

and sin(x), it is also a solution of the ordinary

differential equation (O.D.E), (eq.2).

Let us pose the question: are there 3rd order O.D.E.’s with

constant coefficients such that similar identities are

associated with them? Well why don’t we start our search for the
answer to that question by creating a 3rd order O.D.E by

differentiating (eq.1). Then we will see what we can find out?

Please note that later, when we discuss the kinematics of

simple and damped harmonic motion, we will replace the

independent variable x by t (which will denote time).

It is convenient to write (eq.1) in the following

form

y’’ = - By’ - Cy (eq.5)

Differentiating (eq.5), recalling that B and C are

constants, with respect to x results in

y’’’ = - By’’ – Cy’ (eq.6)

Now substitute the expression for y’’ , from (eq.5)

into (eq.6) and simplify the results. We obtain

y’’’ + (C – B2)y’ – BCy = 0. (eq.6a)

If we now choose units so that

C = B2

then the equation above reduces to

y’’’ – B3y = 0. (eq.7)

Before continuing, let’s note that a more general form

of (eq.2) is the following

y’’ + k2y = 0 (eq.8)

where k is a constant. This equation (eq.8) may be

referred to as the one (spatial) dimensional “un-damped

wave” equation.
When B ≠ 0 then (eq.1) may be thought of as the

one spatial dimensional damped wave equation.

Equation (eq.7) may be thought of as a 3rd order counter-

part of (eq.8).

The reader is probably familiar with many cases

in (math) physics where the units of physical quantities

and the units associated with constants of proportionality

have been chosen for the purpose of simplifying equations

and/or formulas. For example, in quantum field theory it is

often the case that units are chosen so that both c the speed of

light and Plancks’ constant h are set equal to 1 with no loss of

generality.

If we write y(n) to denote the nth derivative of y with

respect to x, then both (eq.7) and (eq.8) may be derived

from

y(n) + Lny = 0 (eq.9)

by letting n = 3 and L = -B we obtain (eq.7). By

setting n = 2 and L = k (or –k) we obtain (eq.8).

The implications of (eq.9), when n is a non-negative

integer different from 2 or 3, will be discussed

in later papers.

Cubic Complex Numbers

In what follows we are motivated by the possibility of

finding identities, associated with (eq.7), that correspond to


the identities that are derived from (eq.2). We propose to

indeed do just that: but we will need the help of a new1

algebraic structure to do so. The name “Cubic Complex Numbers”

(denoted by C3) has been given to this “new” algebraic

structure.

The fundamentals of Cubic Complex Numbers, which can be

shown to be a commutative algebraic ring over the set of ordered

triples of complex numbers, will now be discussed. But first we

will review some basic concepts from ordinary complex numbers.

Recall that the complex “imaginary” unit i may

be defined by

i = (-1)(1/2)

But the multiplicative identity 1 and i may be put into a

one to one correspondence with unit vectors in the plane

as follows:

1 ↔ (1,0)

i ↔ (0,1)

The present author was motivated to write, in the early

1980’s, a more abstract definition of i.

i = (-1,0)(1/2)

so that

i2 = (-1,0) = -1

The reader will recall from elementary abstract algebra

that i is the generator of a cyclical group of order 4. We


may write

i1 = i

i2 = -1

i3 = -i

i4 = 1

Making use of the fact that i2 = -1, as the reader knows

well, we may form the product of two arbitrary complex

variables u = x + iy and v = a + ib, as follows

uv = vu = ax –by + i(ay + bx).

Using the ordered pair notations (corresponding to vectors

in the plane), we may also write

(x, y)(a, b) = (ax – by, ay + bx)

Now we will turn our attention to Cubic Complex Numbers.

We will now provide a preliminary description of the

algebra of Cubic Complex Numbers (to be denoted by C3). The

reader will recall that every field [such as the complex

numbers] is also a ring but not necessarily vice versa. It will

be convenient to denote the ring of complex numbers by C2. The

algebra C3 may be described as the set of ordered triples of

ordinary complex numbers (C2) endowed with the operations of

vector addition and a commutative rule for multiplication (which

will be given later).


Clearly C3 contains the set of 3D vectors (with real

components) as a subset. We will now make the following

identification.

1 <=> (1,0,0).

And now we will define the symbol j by the following:

j = (-1,0,0)^(1/3) (equa.10a)

so that

j3 = (-1,0,0) = -1, j4 = -j, j5 = - j2 and j6 =1. Therefore j is

the generator of a cyclic group of order 6: a fact that will be

exploited later.

For brevity we write2

j = (-1)^(1/3) (eq.10b)

and

j^3 = -1 (eq.11)

The reader will note that j as defined above is different from

any of the three cube roots of -1 <=> (-1,0), because the three

ordinary complex cube roots of -1 are -1 and a pair of

(ordinary) complex conjugate numbers.

The cyclical group (of order six) properties of the powers

of j are readily seen in the following (in which the symbol ^

denotes exponentiation.

j^3 = -1 (eq.12a)

j^4 = -j (eq.12b)
j^5 = -j^2 (eq.12c)

j^6 = j^0 = 1 (eq.12d)

j^1 of course is just j.

An arbitrary cubic complex variable u may be written as

u = a + jb + (j^2)c (eq.12e)

where, in general, a, b and c are (ordinary) complex numbers. If

v = x + jy + (j^2)z (eq.12f)

is another arbitrary cubic complex variable

then we may write the sum and product of u and v as

u + v = a + x + j(b + y) + (j^2)(c + z) (eq.13)

uv = ax - bz- cy + j(ay + bx - cz) + j^2(az + cx + by) (eq.14)

The operation of addition is obviously commutative and

it is not difficult to show that the product is also

commutative.

The sum and product operations may also be expressed in

terms of ordered triples as follows:

(a, b, c) + (x, y, z) = (a + x, b + y, c + z)

(a,b,c)(x,y,z) = (ax – bz - cy, ay + bx - cz, az + cx + by)

By definition, if u = a + jb + (j^2)c is a cubic complex

number and if v = x + jy + (j^2)z is also a cubic complex

number, then

u = v if and only if

a = x

b = y
c = z

The additive identity of C3 is (0,0,0) <=> 0. Also for a

given ordered triple of complex numbers (a,b,c),

the additive inverse is (-a, -b, -c). And it can be shown that

the multiplicative inverse of (a,b,c) is given by

(a, b, c)^(-1) = (L/S, -N/S, -M/S) (eq.14a)

where S ≠ 0 and where L, M, N and S are defined by

L = a^2 + bc (eq.14b)

M = ac – b^2 (eq.14c)

N = c^2 + ab (eq.14d)

S = a^3 – b^3 + c^3 + 3abc (eq.14e)

So then every cubic complex number

(a, b, c) ≡ a + jb + (j^2)c

for which S, as defined above does not vanish, has a unique

multiplicative inverse. This can easily be verified as follows:

(a, b, c)*(a, b, c)^(-1) = (a, b, c)*( L/S, -N/S, -M/S)

By using the rules of multiplication, the definitions for L, M,

N and S, and simplifying, it is seen that the right side of the

above equation reduces to (1, 0, 0). Therefore

(a, b, c)*(a, b, c)^(-1) = (1, 0, 0) <=> 1

when S = a^3 – b^3 + c^3 + 3abc ≠ 0, as required.

Using the definitions provided it, can be shown, in a

straight forward fashion, that C3 is a commutative ring with

identity.
The Cubic Complex Exponential Function

We will now make use of the cyclicality (of order six) of

j^n, (where n is an integer) to define the Cubic Complex

Exponential e^(jx). In a later paper we will examine the more

General form of the Cubic Complex Exponential: e^{jx + (j^2)y}.

It will be shown that the Cubic Complex Exponential e^(jx) will

generate solutions of the third order O.D.E.

y’’’ – y = 0,

just as the complex exponential e^ix generates solutions of the

second order O.D.E.

y’’ + y = 0

We have

e^jx = 1 + jx + (1/2)(jx)^2 + (1/6)(jx)^3 + ... + (1/n!)(jx)^n

+ ...

e^jx = 1 + jx + (1/2)(j^2)(x^2) + (1/6)(j^3)(x^3) + ... +

(1/n!)(j^n)(x^n) + ... (eq.15)

Now since j^n will always equal (+/-)1, (+/-)j or

(+/-)j^2, where the symbol +/- denotes “plus or minus”,

the above equation may be written as

e^jx = 1 – (1/3!)(x^3) + (1/6!)(x^6) + ...+ j[x-(1/4!)(x^4)

+(1/7!)(x^7)+...]+j^2[(1/2!)(x^2)-(1/5!)(x^5)

+(1/8!)(x^8)+ ...] (eq.16)

It will be noted that three separate infinite series

are indicated on the right of (eq.16). We may define


them, using the notation F1, F2 and F3, as follows:

F1 = 1 – (1/3!)(x^3) + (1/6!)(x^6) + ... (eq.17a)

F2 = x - (1/4!)(x^4) + (1/7!)(x^7) +... (eq.17b)

F3 = (1/2!)(x^2)-(1/5!)(x^5) +(1/8!)(x^8) + ... (eq.17c)

The reader is reminded that even though the (conventional)

notation + ... appears after the third term, on the right sides

of each of the three equations directly above, in actuality the

terms will alternate in sign.

It is easily seen that the general or the “nth” term of the

expressions for F1 , F2 and F3 are x3(n-1)/(3(n-1))!,

x(3n-2)/(3n-2)! and x(3n-1)/(3n-1)! respectively.

It can also be shown in a straight forward fashion, using

the ratio test for example, that each of these infinite series

converges (absolutely) for all real values of x.

So then with the aid of equations 17a, 17b and 17c, we may

re-write (eq.16) simply as

e^jx = F1 + jF2 + (j^2)F3 (eq.18a)

or equivalently as

ejx = F1 + jF2 + j2F3 (eq.18b)

We will refer to (eq.18a or 18b) as the Cubic Complex

(version of the) Euler Formula. These are actually

identities because in each equation, by the definitions

set forth, each side is merely an equivalent


representation of the other side.

Using summation notation we may also write (eq.18b) as

ejx = ∑ j(k-1)Fk

where k is an integer and ranges from 1 to 3.

The three functions F1, F2 and F3 will be called the three

fundamental functions (of a single independent variable) of

Cubic Complex Variable (CCV) theory. It should be noted that

similar functions have been defined in connection with the more

general cubic complex exponential ejx + (j^2)y


(or exp(jx + j2y)).

But they are functions of the two independent variables x and y.

The first function F1 (as defined in (eq.17a)) will be

called the principal function of CCV and it will be seen to play

a role similar to the cosine function in ordinary complex

variable theory.

We will now show that the three fundamental functions are

independent solutions of

y’’’ + y = 0. (eq.19)

which may be obtained from (eq.7) by letting B = -1.

Recall from (eq.17a) that

F1 = 1 – (1/3!)(x^3) + (1/6!)(x^6) + ...

differentiating (term by term) with respect to x gives us

F1’ = -[(1/2!)(x^2)-(1/5!)(x^5) +(1/8!)(x^8) + ...]



where the symbol ( ) denotes differentiation (with respect

to x). But the expression inside of the brackets, in the


above equation, is just F3 as defined by (eq.17c). Therefore

we may re-write the equation directly above as

F1’ = - F3 (eq.20).

Differentiation of both sides of (eq.20) results in

F1’’ = - F3´ (eq.21)

To determine F3´ we differentiate the right side of

(eq.17c). We obtain

F3´ = x - (1/4!)(x^4) + (1/7!)(x^7) +...

It is seen that, according to (eq.17b), the right of the

the above equation is F2. So we have

F3´= F2 (eq.22)

From equas.(21) and (22) it immediately follows that

F1’’ = - F2 (eq.23)

Differentiation of both sides of (eq.23) with

respect to x results in

F1’’’ = - F2’ (eq.24)

Differentiation, with respect to x, of the

defining equation for F2, (eq.17b), we obtain

F2´ = 1 – (1/3!)(x^3) + (1/6!)(x^6) + ...

But the infinite series on the right side of the

equation above is by definition F1 . So we have

F2´ = F1 (eq.25)

Substitution of this result into (eq.24) results in

F1’’’ = - F1 (eq.25a)
or equivalently

F1’’’ + F1 = 0 (eq.26)

Now let y = F1 and obtain

y’’’ + y = 0. (eq.27)

Therefore y = F1 , one of the three fundamental

functions, is a particular solution of (eq.19), which

for convenience has also been labeled as (eq.27),

as asserted. In a similar fashion, it can also be

shown that F2 and F3 are also independent

solutions of (eq.19). It will also be noted that

(eq.19) may be obtained from (eq.7) by setting

B = -1.

So then we have shown that the components of the

Cubic Complex exponential

ejx = F1 + jF2 + j2F3

are independent solutions of the third order O.D.E.

y’’’ + y = 0

just as the ordinary complex exponential

eix = cos(x) + i*sin(x)

has components that are solutions of the second order O.D.E.

y’’ + y = 0.

It is hereby conjectured that these results may be

generalized to the nth order case involving


y(n) + y = 0

Now we will show that the three fundamental functions, F1 ,F2 and

F3 can be obtained, using Taylor Series, as solutions for initial

value problems. First we will find a particular solution

of

y’’’ + y = 0 (eq.28a)

such that the following conditions are satisfied:

y(0) = 1, (eq.28b)

y’(0)= 0 (eq.28c)

y’’(0) = 0 (eq.28d)

Now let x = 0 in (eq.28a) and obtain

y’’’(0) + y(0)= 0

but y(0)= 1, from (eq.28b) so we have

y’’’(0) + 1 = 0 or

y’’’(0) = -1 (eq.29)

If we now differentiate (eq.28a) with respect to x we

will obtain

y(4) + y’ = 0 (eq.30)

where the notation y(4) is employed to denoted the 4th

derivative of y with respect to x. In general, we will

denote the nth derivative of y with respect to x by

the symbolism y(n) for (positive) integer n ( > 3 ).

By setting x = 0 in (eq.30) we arrive at the following


result

y(4)(0) + y’(0)= 0

but from (eq.28c) we have

y’(0)= 0, therefore we may write y(4)(0) + 0 = 0 or

y(4)(0) = 0 (eq.31)

Successive differentiation of (eq.30) will give us

y(5) + y’’ = 0 (eq.32)

and

y(6) + y’’’ = 0 (eq.33)

Setting x = 0 in (eq.32), we have

y(5)(0) + y’’(0) = 0

now substituting

y’’(0) = 0, from (eq.28d), will lead us to

y(5)(0) = 0 (eq.34)

And finally, in this sequence of calculations,

if we set x = 0 in (eq.33) and then substitute

y’’’(0) = -1 from (eq.29), we arrive at the

following

y(6)(0) = 1 (eq.35)

Now, thanks to the information provided by

equas.(28b,28c,28d,29,31,34 and 35), we have the

values of y and its first six derivatives evaluated

at x = 0. Therefore we can write the first seven


term of a Taylor Series3 expansion of the function

y = y(x) near x = 0. We obtain

y = y(0) + y’(0)x + (1/2)y’’(0)x2 + (1/6)y’’’(0)x3

+(1/24)y(4)(0)x4 + (1/5!)y(5)(0)x5

+ (1/6!)y(6)(0)x6 + ...

Now making the appropriate substitutions and dropping the

zero terms (of course) we arrive at

y = 1 – (1/3!)x3 + (1/6!)x6 + ... (eq.36)

Except for the differences in the notation used for

exponents, it is seen that the rights of (eq.36) and

of (eq.17a) are identical. Accordingly, they must

represent the same (convergent) infinite series and so

we can state that y = F1, as defined by (eq.17a), satisfies

(eq.28a). It can be easily verified that y satisfies

the auxiliary conditions of equas.(28b, c and d).

Following a procedure, similar to that used above, it

can also be shown that y = F2 satisfies the equation

y’’’ + y = 0

such that the following conditions are satisfied:

y(0) = 0, y’(0)= 1 and y’’(0) = 0.

It can also be demonstrated that y = F3 satisfies the same

equation subject to the conditions

y(0) = 0, y’(0) = 0 and y’’(0) = 1.


Let cos(x) and sin(x), which are two well known linearly

independent solutions of (eq.2), be denoted by f1 = cos(x)

and f2 = sin(x). We then have

f1’ = df1/dx = - f2 = - sin(x)

So then the famous identity

cos2(x) + sin2(x) = 1 may be written as

(f1)2 + (f1’)2 = 1 (eq.37a)

It can also be easily shown that

(f2)2 + (f2’)2 = 1 (eq.37b)

We will now derive the corresponding identities

for the third order case which involve concepts

from CCV theory. But first it will be helpful if

we recall some familiar relationships from ordinary

complex variables.

Let

X + iY = (a + ib)(x + iy)

then it is easy show that

X = ax – by (eq.38a)

Y = bx + ay. (eq.38b)

If we let a and b be constants and then consider

equas.(38a & b) as a system of linear equations in

x and y, then the matrix of coefficients of that

system may be denoted by M1 where M1 is given by


M1 = a -b (eq.39)

b a

where for typographical convenience we have omitted

the parenthesis that are conventionally used to

enclose the elements of matrices. We have,

writing the determinant of M1,

det(M1) = a2 + b2 (eq.40)

Now the reader will recall, or can easily verify,

that

X2 + Y2 = (a2 + b2)(x2 + y2) (eq.41)

where X and Y are defined as in equas.(38a & b).

It will also be noted that the left side and both

factors on the right side of (eq.41) are (2nd

degree) expressions that have the same algebraic

form as does the right side of (eq.40). Also if

a = cos(θ) and b = sin(θ), then of course

a2 + b2 = 1

and equas.(38a & b) represent an orthogonal coordinate

transformation.

Now for the CCV counterparts of the foregoing, let u

v be defined as in equas.(12e & 12f) above. Then the

product of u and v is given by

uv = ax - bz- cy + j(ay +bx -cz) + j^2(az+cx+by) (eq.42a)

on account of (eq.14).
Since u and v are cubic complex variables, their

product is one also. Let the product uv be, the

cubic complex variable, given by

uv = X + jY + j2Z

then (eq.42a) may be written as

X + jY + j2Z = ax - bz- cy + j(ay +bx -cz) + j^2(az+cx+by)

(eq.42b)
Then therefore, by the definition of the equality

of cubic complex numbers, we have

X = ax - bz- cy

Y = ay + bx - cz

Z = az + cx + by

It is convenient to rearrange the order of the terms

on the right sides of the three above equations as

follows:

X = ax – cy - bz (eq.43a)

Y = bx + ay - cz (eq.43b)

Z = cx + by + az (eq.43c)

Now if we consider equas.(43a, 43b & 43c)

be to be a system of three linear equations in x,

y and z, then the matrix of coefficients, which we

denote by M2 is given by

M2 = a -c -b
b a -c

c b a

the determinant of the matrix M2 is given by

det(M2) = a3 – b3 + c3 + 3abc (eq.44)

After a somewhat tedious calculation, it can be

shown that

X3 - Y3 + Z3+ 3XYZ =

(a3 – b3 + c3 + 3abc)(x3 – y3 + z3 + 3xyz) (eq.45)

where X,Y and Z are defined by equas.(43a,43b &

43c). So then (eq.45) is the CCV theory analog to

(eq.41) which is fundamental in ordinary complex

theory.

It can be shown, in a straight forward fashion,

that

(F1)3 – (F2)3 + (F3)3+ 3(F1)(F2)(F3) = 1 (eq.46)

where F1, F2 and F3 , each a function of x, are defined by

equas.(17a, 17b and 17c). The relationship holds true for all

real x. So (eq.46) corresponds to the identity of (eq.3)

which is associated with complex variable theory.

Norms and Pseudo Norms

A review of eqs.(44) and (45) will reveal that algebraic

expressions of the form a3 – b3 + c3 + 3abc plays an integral

part in CCV theory. In fact, for real a, b and c, the mapping

(a, b, c) ==> a3 – b3 + c3 + 3abc


is the counterpart of the mapping, from ordinary complex

variables

(a, b) ==> a2 + b2

for real components a and b.

We are therefore motivated to defined a “restricted pseudo

norm” N on the set of Cubic Complex Numbers (with real

components a, b and c) as follows:

N(a, b, c) = |a3 – b3 + c3 + 3abc |1/3

so that

[N(a, b, c)]3 = a3 – b3 + c3 + 3abc

The word “restricted” is placed in the definition of N to

indicate that N may not always satisfy the triangle inequality.

However, as it easily may be shown, N satisfies the other

condition(s) for the definition of pseudo norms or semi-norms.

Kinematics Of Simple Harmonic Motion And Damped

Harmonic Motion

With a convenient choice of units, associated with the

relevant physical constants, Simple Harmonic Motion may

be described by the equation

u’’ + u = 0. (eq.47)

Where u = u(t) and (in this section)

the primes indicate differentiation with respect to t.

A particular solution of this equation is

u = cos(t), so that u’ = -sin(t), therefore


(u)2 + (u’)2 = (cos(t))2 + (-sin(t))2

= (cos(t))2 + (sin(t))2 = 1

so that

(u)2 + (u’)2 = 1 (eq.48)

It is easily seen that the same result would have been

obtained if we had set u(t) equal to sin(t) which is the other

independent solution of (eq.47).

So then, for the case of Simple Harmonic Motion, it can be

stated that (eq.48) is an identity that involves the position

u = u(t) and the velocity u’ of an object that is in

Simple Harmonic Motion in one spatial dimension.

Let’s also make note of the fact that conventionally

the arguments of the sine and cosine functions, in

connection with simple harmonic motion, are often

written as ωt + β where ω is the angular frequency

and β is a dimensionless constant (initial angular

displacement). A typical representation of a function

describing simple harmonic motion is f(t) = A*cos(ωt + β)

where A is the amplitude. For convenience we are choosing

units that will set A = 1 and initial conditions such that

ω = 1 and β = zero. These specifications should not, in any way,

affect the generality of the principles discussed and proposed.

It will be recalled that previously we let y = y(x) and y’


denote dy/dx, etc., and demonstrated how the second order

equation

y’’ + By’ + Cy = 0

could be transformed into the third

order equation

y’’’ + y = 0.

By a similar procedure the equation

u’’ + Bu’ + Cu = 0 (eq.49)

can transformed into

u’’’ + u = 0. (eq.50)

where now u = u(t) and u’ = du/dt, etc.,

It will be recalled that the second order Equation

(eq.49) may be interpreted as representing Damped

Harmonic Motion in one spatial dimension.

However, a relationship of form of (eq.48) does

not identically hold true for the function (and its

1st order derivative) that is a solution of the 2nd order

equation for damped harmonic motion. But,

alas, when (eq.49) is transformed into the form of (eq.50), such

a relationship does indeed exist.

For notational convenience replace x by t in

equas.(17a, 17b and 17c) and obtain

F1 = 1 – (1/3!)(t^3) + (1/6!)(t^6) + ... (eq.51a)

F2 = t - (1/4!)(t^4) + (1/7!)(t^7) +... (eq.51b)


F3 = (1/2!)(t^2)-(1/5!)(t^5) +(1/8!)(t^8) + ... (eq.51c)

where F1, F2 and F3 are now denoting functions of t.

For convenience we will now recall (eq.20), (eq.23)

and (eq.25a) where the primes will now indicate

differentiation with respect to t.

F1’ = - F3 (eq.20)

F1’’ = - F2 (eq.23)

F1’’’ = - F1 (eq.25a)

Now since we are denoting the displacement variable,

of a physical system undergoing damped harmonic motion,

by u =u(t), we desire to find “differential identities”,

involving u and its derivatives, that correspond to the

“differential identities,

(f1)2 + (f1’)2 = 1

(f2)2 + (f2’)2 = 1

that are associated with simple harmonic motion and that were

recalled from eqs. 37a & 37b.

Motivated by the algebraic form of the right side of

(eq.44) we are led to define the differential operator L(u) as

follows:

L(u) = (u’’)3 - (u’)3 + (u)3 + 3(u’’)(u’)(u) (eq.52)

Now let u = F1 . We may easily calculate u’ and u’’ from (eq.20) &

(eq.23) that were re-written above. We have


u’ = F1’ = - F3

u’’ = F1’’ = - F2

so that

u’ = - F3

u’’ = - F2

Substituting u = F1 and

u’ = - F3

u’’ = - F2

into (eq.52), we obtain

L(F1) = (-F2)3 - (-F3)3 + (F1)3 + 3(-F2)(-F3)(F1)

or

L(F1) = (F1)3 - (F2)3 + (F3)3 + 3(F1)(F2)(F3) (eq.53)

But the right side of (eq.53), as a consequence of

(eq.46), is equal to unity. So we have

L(F1) = 1 (eq.54a)

Now we will calculate L(F2). We may do this by

now setting u = F2 and then calculate u’ and u’’

by making use of eqs. 20, 23 and 25a above. We obtain

u’ = F1

u’’ = - F3

Substituting the above results (along with u = F2) into

(eq.52) will give us

L(F2) = - ((F1)3 - (F2)3 + (F3)3 + 3(F1)(F2)(F3))


and as a consequence of (eq.46) we have

L(F2) = - 1 (eq.54b)

And if we set u = F3 , then by a process similar to

that above it can be shown that

L(F3) = 1 (eq.54c)

Actually eqs.54a, 54b & 54c may be combined by writing

(L(Fn))2 = 1 (eq.55)

for n = 1, 2 or 3 and where the differential operator L

is defined by (eq.52).

It is interesting to note that a corresponding

differential operator may be defined for the Simple

Harmonic Motion case. Let f = f(t) and let the differential

operator P be defined by

P(f) = (f)2 + (f’)2 (eq.56)

where f’ denotes df/dt. Then eq.37a & 37b may be

combined by writing

P(fn) = 1 (eq.57)

where n = 1 or 2 and f1 = cos(t) and f2 = sin(t).

Thus (eq.55) holds identically true, when Fn = Fn(t)

(for n = 1, 2 or 3), are the functions defined by eqs.51a, 51b

and 51c respectively. In general the arguments of each of these

functions, which are solutions of (eq.50) and are realistic

candidates for representing the position of an object that is

executing damped simple harmonic motion in one spatial


dimension, is ωt + β. But as is customarily the case, convenient

choices of the initial conditions and the units of the constants

of proportionality have been adopted. More specifically we have

chosen to set ω = 1 and β = 0.

Returning our attention to (eq.52), the starting point for

the identities expressed in (eq.55), the instantaneous

acceleration, velocity and position are given by u’’, u’ and u

respectively. It is strikingly noteworthy to observe that (u’)3,

the cube of u’, which is also associated with the dissipative

force of friction, [see (eq.49), the 2nd order O.D.E. governing

damped simple harmonic motion], is preceded by a negative sign

in (eq.52). This fact adds weight to the assertion that (eq.55),

which is defined by way of (eq.52), is the damped simple

harmonic motion counterpart of (eq.48): which is a fundamental

identity associated with Simple Harmonic Motion.

While more details of the present theory will be presented

in later papers, it should be noted that these concepts can be

generalized to higher orders and degrees. The author has also

developed “canonical form” techniques for the mathematical

treatment of anharmonic wave phenomena: also to be presented in

subsequent papers.

Some readers may consider the discussion, pertaining to the

potential application of CCV theory to damped simple harmonic

motion, as pure mathematical speculation or mathematical


fiction. The author has coined the phrase “mathematical

phiction” to characterize that view point. Nevertheless, it must

be remembered that much of the science fiction of yesterday is

responsible for many of the engineering wonders of today.

Let’s hope that an effort will be made for a more rapid

review of worthy science innovations and/or math phiction

formulations in order to more quickly determine the merits of

potentially productive conceptualizations.

Footnotes:

1.) Manuscripts explaining the fundamentals of Cubic Complex

Numbers were informally distributed (for comment and review)

among the grad students and faculty of Northwestern University

and the University of Minnesota between 1986 and 1993.

2.)

The author has also generalized the concept of the cubic

complex unit to the nth order. The nth order complex unit

jn been defined by

jn = (-1,0,0,...)^(1/n)

where (1,0,0,...) is an n dimensional unit vector.

3.)
The author outlined in an informally distributed 1980 paper, a

generalization of the method used above for finding Taylor

Series solutions of nth order ordinary differential equations:

subjected to a set of n auxiliary conditions.

Vous aimerez peut-être aussi