Professional Documents
Culture Documents
net/publication/270706800
Proof for the Beal Conjecture and a New Proof for Fermat's Last Theorem
CITATIONS READS
8 2,456
1 author:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Leandro Torres Di Gregorio on 13 May 2016.
Email address:
leandrogregorio@ig.com.br (L. T. Gregorio), leandro.torresdigregorio@gmail.com (L. T. Gregorio)
Abstract: The Beal Conjecture was formulated in 1997 and presented as a generalization of Fermat's Last Theorem,
within the number theory´s field. It states that, for X, Y, Z, n , n and n positive integers with n , n , n > 2, if X +
Y = Z then X, Y, Z must have a common prime factor. This article presents the proof for the Beal Conjecture, obtained
from the correspondences between the real solutions of the equations in the forms A + B = C , δ + γ = α and
X + Y = Z . In addition, a proof for the Fermat's Last Theorem was performed using basic math.
Keywords: Beal Conjecture, Fermat´s Last Theorem, Diophantine Equations, Number Theory, Prime Numbers
simply refereed as Fermat´s equation) and X Y Z equation for real numbers. That is, from the sum of two
(here simply referred as Beal equation) were freely squares it is possible to obtain a third square, which can be
explored, searching for different aspects of their meanings. graphically represented in Figure 3:
The deduction of Beal Conjecture´s proof is divided into
five parts. The first part consists on the presentation of
geometric illustrations that allow a view of the argument to
be adopted. The second part presents the deduction of the
correspondence betweenn the real solutions of
Pythagoras´ equation and Fermat's equation, obtaining the Figure 3: Graphic representation of Pythagoras´ equation for squares
transforms that enable this operation. The third part with sides A, B and C, respectively. Source: author.
contains the deduction of the transforms between the real
solutions of Fermat's equation and the Beal equation, and Assuming that each square has an infinitesimal thickness
the fourth part includes analysis and discussion about the du, the infinitesimal volumes of the elements correspond to:
topic, including some examples. Finally, the fifth part dV% A du , dV' B du and dV, C du . In fact, the
develops the completion of the proof to Beal Conjecture sum of the areas can be converted into volumes by simply
and after that it is presented a new proof for Fermat's Last multiplying the scalar du on the Pythagoras´
Pythagoras equation,
Theorem. resulting in:
The presented language
anguage is simple, sometimes making use
of geometrical illustrations to explore analytical aspects of A du B du C du (1)
the problem, which is believed to be helpful for the Or, equivalently:
solution´s understanding. It is thus intended that the
statements are sufficiently clear and straightforward,
straight dV% dV' dV, (2)
allowing analysis by the scientific community and by non-non
experts enthusiasts on the subject. As illustrated in [5], one can integrate the infinitesimal
volumes along an axis u transverse to plane of the figures
(therefore linearly independent to it, it [6]) until some
2. Proof for the Beal Conjecture arbitrary thickness U, and then have:
2.1. Geometric Illustrations ) ) )
(* dV% (* dV' (* dV, (3)
Using a geometric approach to illustrate this proposition,
) ) )
one can take the figures below, in which are represented (* A du (* B du (* C du (4)
two squares of sides A and B, with areas A and B ,
respectively. The forms are out of scale and possibly out of A U B U C U (5)
proportion what, however, do not imply in loss of the
By making U numerically equal to A, that is, making
mak the
problem´s understanding.
thickness set to all elements be equal to the side of the first
element, there comes that:
A A B A C A (6)
A B A C A (7)
Figure 1: Squares of sides A and B, and areas and , respectively . As one can see, by setting U = A, the cube of side A on
Source: author. the first element
ement was obtained in a direct way. Being C
always greater than A and B, and assuming B different from
By adding the areas of the squares of sides A and B one
A (just for the purpose of visualization, but this is not a
can obtain an area numerically equal to S A B . There
condition at first), the corresponding volumes to the second
are infinite geometric surfaces that may represent the area S,
and third elements will necessarily
necessar be different from their
amongng which some are exemplified in figure 2.
respective cubes, corresponding to tridimensional solids in
the shapes of parallelepipeds.
In an analogous manner to the areas (figure 2), a volume 2.3. Correspondence between Real Solutions of Fermat's
V can be represented by various tridimensional solids. It is Equation and Beal Equation
well known that the volume contained in a parallelepiped
can also be represented by a cube of an equivalent volume, Taking the Beal equation in the form X + Y = Z
with side equal to L √V. Relying on this artifice, it can one can realize that it is possible to transform it into the
4 4
be obtained a cubic equivalency to the volumes of the Fermat´s equation by making Y = γ and Z = α ,
second and third solids, resulting in: resulting in:
Y = ∏@
=G,? =G, =G, =G,C Once PS is present in X (k 6,S ≥ 0) and in Y (k 5,S ≥ 0),
<A P< =P P … P@ (24)
it can be put at evidence, resulting
=H,? =H, =H, =H,C
Z = ∏@
<A P< =P P … P@ (25) =F, =F,U7 =F,V =F,∞
PS [WP … PS … PT … P∞ X+
Therefore, B and C can be written as: =G, =G,U7 =G,V =G,∞
WP … PS … PT … P∞ X] = Z (36)
3 JG, 3 JG, 3 JG,C
I I …IC
B = 3 7 JF, 3 7 JF, 3 7 JF,C (26) As the content of the brackets is integer (here named M),
I I …IC
it comes that Z = PS M, that is, Z also has the prime
I
3 JH,
I
3 JH,
…IC
3 JH,C factor PS in its factorization. As Z is assumed integer by
C = 3 7 JF, 3 7 JF, 3 7 JF,C (27) initial hypothesis, then Z and Z have the same prime
I I …IC
factors, resulting that PS is also present in Z.
Or equivalently: Doing the same for PT , once PT is present in X (k 6,T ≥ 0)
and in Z (k 8,T ≥ 0), it can be put at evidence, resulting
=G, 0 0 =F, =G, 0 0 =F, =G,∞ 0 0 =F,∞
B =P P … P∞ PT \WP
=H,
… PS
=H,U
… PT
=H,V
… P∞
=H,∞
X−
(28)
P1n1kX,1…Pmn1kX,m…Pvn1kX,v−1…P∞n1kX,∞=Yn2
=H, 0 0 =F, =H, 0 0 =F, =H,∞ 0 0 =F,∞
C =P P … P∞
(37)
(29)
As the content of the brackets is integer (here named N),
Since B and C are rational numbers, three situations it comes that Y = PT N, that is, Y also has the prime
may occur: factor PT in its factorization. As Y is assumed integer by
Situation 1: and are integers; initial hypothesis, then Y and Y have the same prime
Situation 2: or is integer and the other is non- factors, resulting that PT is also present in Y.
integer; As already mentioned, the exception occurs when n = 2,
Situation 3: and are non-integers. when Y and Z do not depend on the variable X anymore.
One can note that, in principle, it seems not to be necessary
2.4.1. Situation 1 Analysis (KL and ML Integers) that n , n > 2 for the rule to be considered valid, but only
Assuming that B and C are integers, all the powers of n > 2 . However, there are cases of known integer
the primes factors must be positive or zero. If not, then the solutions of X, Y, Z coprime in which n > 2 and n = 2
prime factor would be raised to a negative power, going to or n = 2, i.e. 7 + 13 = 2c e 2d + 17 = 71 [2]. This
the denominator and leading to a non-integer result. This aspect will be clarified in section 2.5.
condition can be expressed in the general form as
2.4.2. Situation 2 Analysis (KL or ML Non-Integers)
n k 5,< − n − 2 k 6,< = K 5,< ∴ n k 5,< = K 5,< + n − 2 k 6,< (30) Writing B (non-integer) in the notation form I' + ε, in
which I' represents the whole part of B and ε is the
n k 8,< − n − 2 k 6,< = K 8,< ∴ n k 8,< = K 8,< + n − 2 k 6,< (31)
decimal part, and as X + B = C , it follows that:
As one can note, if K 5,< ≥ 0 for all primes and a certain
X + I' + ε = C (38)
prime factor PS exists in X (k 6,S ≥ 0), then necessarily
k 5,S ≥ 0, that is, PS is also present in Y. The same is valid X + I' + ε = C (39)
for X and Z once K 8,< ≥ 0, that is, if a prime factor PT is
present in X (k 6,T ≥ 0), then necessarily it will be present Since X and I' are integers, then X + I' = I, , in
in Z (k 8,T ≥ 0 . which I, is the whole part of C . This results that C = I, +
ε, that is, if B is a non-integer real, then C also is, and the
Writing X , Y , Z with highlights to the PS and PT
prime factors, one can have: decimal part ε is common to both. Therefore, situation 2 is
not possible, leaving only situation 3.
=F, =F,U =F,V =F,C
X =P … PS … PT … P@ (32)
2.4.3. Situation 3 Analysis (KL and ML non-integers)
=G, =G,U =G,V =G,C Assuming that B and C are non-integers, there are one
Y =P … PS … PT … P@ (33) or more prime factors with negative powers in their ratio
=H, =H,U =H,V =H,C form. If not, then the prime factors would be raised only to
Z =P … PS … PT … P@ (34) positive powers, leading to B and C integers.
Naming the prime factors with negative powers in B as
Therefore the equation X +Y =Z becomes:
PS∗ and the prime factors with negative powers in C as PT∗ ,
=F, =F,U =F,V =F,∞ there comes that:
WP … PS … PT … P∞ X+
=G, =G,U =G,V =G,∞ n k 5,S∗ − n − 2 k 6,S∗ = K 5,S∗ (40)
WP … PS … PT … P∞ X=
=H, =H,U =H,V =H,∞ n k 8,T∗ − n − 2 k 6,T∗ = K 8,T∗ (41)
P … PS … PT … P∞ (35)
Pure and Applied Mathematics Journal2013; 2(5): 149-155 153
One can note that for K 5,S∗ < 0, PS∗ will be present in 2.4.1 for PS = PS∗ and PT = PT∗ . Thus, to avoid redundancy,
X (k 6,S∗ > 0) and Y (k 5,S∗ > 0) simultaneously, because if it will be not replicated in this section.
one is null, the other will be negative, what is false, since X
2.5. Conclusion to the Beal Conjecture Proof
and Y are integers by initial hypothesis k 6,< ≥ 0 and
k 5,< ≥ 0 . That is, if B is non-integer, there is at least one 2.5.1. Extending the approach
prime factor PS∗ that is common to X and Y. The same In fact, the initial approach adopted the first shape as the
mechanism can be applied to C and PT∗ once K 8,T∗ < 0, reference for the development (figure 4), what led to
resulting that, if C is non-integer, there is at least one A + B A 0 = C A 0 , and consequently X = A ,
prime factor PT∗ that is common to X and Z. 3
Y = √B X 0 , Z = √C X 0 .
3
Writing X , Y , Z with highlights to PS∗ and PT∗ One can also adopt the second element as the basis,
prime factors, one can have: 3
resulting in A B 0 + B = C B 0 , and X = √A B 0 ,
3
X =P
=F,
… PS∗
=F,U∗
… PT∗
=F,V∗
… P@
=F,C
(42) Y = B, Z = √C B 0 . The same applies if one chooses
the third element to be the reference, coming to A C 0 +
3 3
Y =P
=G,
… PS∗
=G,U∗
… PT∗
=G,V∗
… P@
=G,C
(43) B C 0 = C and X = √A C 0 , Y = √B C 0 , Z = C.
As one can see, a rational solution of A + B = C (in
=H, =H,U∗ =H,V∗ =H,C which at least A or B or C is integer) may lead to three
Z =P … PS∗ … PT∗ … P@ (44)
different types of solutions for X + Y = Z :
The development that shows that PS∗ and PT∗ are present
in X, Y and Z is analogous to the one performed in section
Table 1: Types of solutions for h i + j i = k i that can be obtained using the first, second and third elements as unaltered basis, respectively. Source:
author.
The opposite is also true, that is, from one whole presented the solution 3 + 6 = 3x , resulting in:
solution for X + Y = Z , one can obtain three
different rational solutions for A + B = C . I. e. [1]
SOLUTIONS FOR yL + KL = ML yL KL ML
I´ 9 (A=X=3) 72 81
II´ 4,50 36 (B=Y=6) 40,5
III´ 1 8 9 (C=Z=3)
154 Leandro Torres Di Gregorio:
Gregorio Proof for the Beal Conjecture and a New Proof
roof for Fermat's Last Theorem
3. Proof for
or Fermat's Last Theorem
From the proof above one can note that the following
equations rule the transforms Y B X 0 and
0
Z C X (demonstrations are presented using
solution type I to illustrate, but the process is analogous
for solutions types II and III):
Figure 6: Graphic representation for • and • . Source: author.
n k 5,< N n N 2 k 6,< K 5,< (45)
As one can note, the angular coefficients of each
n k 8,< N n N 2 k 6,< K 8,< (46) equation correspond to 1, that is, the angle is 45º for the
parallels, resulting that U U δ and T T δ.
A particular case is when n n n n, resulting
This way, one can write that:
in X Y Z (Fermat´s equation). Rewriting the sum
of exponents there comes that: ) €‚
€ ƒ €‚
n 2 2 PU T (53)
) ƒ
•G,?
zG,? 0 =F,?
F
=F,? 0
nk 5,< N n N 2 k 6,< K 5,< P n 2 (47) What brings to:
=G,? 0=F,?
F =F,? 0=G,?
zH,? 0 =F,?
F
=F,? 0
•H,?
k 6,< N k 5,< k 6,< N k 8,< P k 5,< k 8,< (54)
nk 8,< N n N 2 k 6,< K 8,< P n 2 (48)
=H,? 0=F,?
F =F,? 0=H,?
As one can note, if k 5,<< k 8,< for all prime factors P<
Now one can make: then B C , what is impossible since A 0 by initial
zG,?
hypothesis. Therefore, as there can be no integer solution
U {k 6,< | k 6,< N (49) for X Y Z , the Fermat´s Last Theorem is proved.
(FLT), various incompatibilities become apparent in the to any ~2 ‡tˆ •2 without necessarily ~2 = •2 .
problem structure, so that more than one path can lead to
the final argument. This is expected, since we started
from the assumption that there are integer solutions to References
the equation in the form X + j = Z , for t > 2. As
indeed these solutions do not exist, it is natural that the [1] R. D. Mauldin, “A generalization of Fermat´s last
proof by contradiction (“reductio ad absurdum”) reveals theorem: the Beal conjecture and prize problem”, in
Notices of the AMS, v. 44, n.11, 1997, pp. 1436-1437.
itself in the form of one or more incompatibilities.
In a first approach, from the equations 47 and 48 one [2] H. Darmon and A. Granville, “On the equations ZS =
can write to any "n": F x, y and Ax • + By Ž = Cz • ”, in Bull. London Math.
Soc., 27, 1995, pp. 513-543.
t − 2 k 6,< = tk 5,< − K 5,< = tk 8,< − K 8,< (55)
[3] K. Rubin and A. Silverberg, “A report on Wiles´
Dividing by n, one can have: Cambridge lectures”, in Bulletin (New Series) of the
American Mathematical Society, v. 31, n. 1, 1994, pp. 15-
KY,i KZ,i
kY,i − = kZ,i − = …i (56) 38.
t t
[4] I. Stewart, Almanaque das curiosidades matemáticas,
When plotting the graphs of Figures 5 and 6 it is 1945, translation by Diego Alfaro, technical review by
noticed that the distance between the roots of ~ and ~ Samuel Jurkiewicz, Rio de Janeiro: Zahar, 2009.
and the distance between the roots of • and • have the
[5] Al Shenk, Cálculo e geometria analítica, translation by
same form of equation 56, so that one can write: Anna Amália Feijó Barroso,Rio de Janeiro: Campus,
KY,i KZ,i 1983-1984.
kY,i − = kZ,i − =… (57)
2 2
[6] J. L. Boldrini, S. I. R. Costa, V. L. Figueiredo, H. G.
Where … is illustrated in Figures 5 and 6 and is valid Wetzler, Álgebra linear, 3 ed., São Paulo: Harper & Row
for the i-th prime factor in analysis. From then one can do Brasil, 1980.
go to the equations 53 and 54, completing the proof with [7] M. L. Crispino, Variedades lineares e hiperplanos, Rio de
the argument already presented in the original paper. One Janeiro: Editora Ciência Moderna Ltda, 2008.
can also note that the incompatibility vanishes when
[8] K. Michio, Hiperespaço, translation by Maria Luiza X. de
t = 2 → …i = 0 (equations 55 and 56), causing A. Borges, technical review by Walter Maciel, Rio de
~ = ~ and • = • , resulting in Janeiro: Rocco, 2000.
) ƒ
1= = (58) [9] Euclides, Os elementos, translation and introduction by
) ƒ
Irineu Bicudo, São Paulo: Unesp, 2009.
to any ~2 and •2 without necessarily ~2 = •2 . [10] G. H. Edmund Landau, Teoria elementar dos números,
Another approach for FLT proof´s final argument can Rio de Janeiro: Editora Ciência Moderna, 2002.
also be taken from equations 47 and 48 in the form: