8/18/2019 Helical Fractal
1/40
Division of Physics and Applied Physics,
School of Physical and Mathematical Sciences,
Nanyang Technological University.
April 9th, 29
roducing the Helical Fractal,
Discrete
Versions and
Sa! "ee#$iem, %r!an Ade P&tra.
per Fractals.
8/18/2019 Helical Fractal
2/40
8/18/2019 Helical Fractal
3/40
The helicalised sine c&rve.
No!, ho! a(o&t
helicalising thehelicalised sinec&rve
The helicalised sine c&rveis a c&rve that !rapsaro&nd the sine c&rve.
3ertainly a moreinteresting '#dc&rve.
And helicalising thatonce more
8/18/2019 Helical Fractal
4/40
4elicalised helicalised helicalisedsine c&rve.
This is the c&rve !inds intoa heli, s&ch that theres<ing heli !raps aro&ndanother heli, !hich finallyc&rls aro&nd the sine c&rve.
%mpressive
A close#&p vie!of the innerstr&ct&re ofthe c&rve.
5No, not typo.6
8/18/2019 Helical Fractal
5/40
Some other helicalised c&rves.
para(ola
hyper(olictangent
ellipsestraightline
eponential c&rve
8/18/2019 Helical Fractal
6/40
%ntrod&cing7 The 4elicaliser8 .
• The helicaliser is a set of parametric form&lae s&ch that it /helicalises0 ac&rve.
• To say that a c&rve is /helicalised0 is to mean that the c&rve !o&ld (edescri(ed (y a ne! c&rve that !inds aro&nd the original c&rve, li1e a heli
(eing directed (y the original c&rve.
• ssentially, any c&rve can (e descri(ed (y its parametric e:&ation, i.e.
• 4ence, the helicaliser !o&ld yield the helicalised c&rve, ;&st (y s&(stit&tingthe f&nctions 5t6 , y5t6 , *5t6 ,
8/18/2019 Helical Fractal
7/40
The 4elical =ractal .
• An a(sol&tely self#similar fractal, called the Helical Fractal can (e&nderstood in t!o !ays7
# %t is li1e a c&rve that defines a heli, !hich defines a larger heli,!hich in t&rn defines yet some other (igger heli, ) ) )
# Another !ay to some!hat descri(e it is that the shape of the c&rveis defined (y the c&rve !inding aro&nd it, !here(y that shape of the!inding c&rve is act&ally defined (y the c&rve !inding a(o&t it, !hoseshape is ) ) )
• A formal mathematical definition of the helical fractal !ill (e givenlater, after its highly#la(orio&s derivation is finally accomplished.
8/18/2019 Helical Fractal
8/40
Some very (asic differentialgeometry.
tangent vector, v 5t6
normal, n 5t6
(inormal, b 5t6
c&rve in '#d space
The three &nit vectors v 5t6 , n 5t6 and b 5t6 areal!ays m&t&ally perpendic&lar to each other andthe directions depend on the point on the c&rve.
8/18/2019 Helical Fractal
9/40
Nat&ral parametrisation (y arclength.
• To even start thin1ing a(o&t helicalising a c&rve, it is important to&nderstand the nat&ral parametrisation of a c&rve. This is 1ey too(taining the normal and (inormal.
• To helicalise is to ma1e the c&rve !ind aro&nd in circles. This can (edone (y adding the oscillatory terms / cos >t 0 and / sin >t 0 to thenormal and (inormal.
• -ith these then, it is possi(le to !or1 o&t the helicaliser.
• A c&rve can (e parametrised (y any parameter t . %n partic&lar, thisparameter can (e chosen to (e s , !here s is the arc length. This is1no!n as the nat&ral parametrisation (y arc length.
8/18/2019 Helical Fractal
10/40
Tangent vector, normal and (inormalthro&gh nat&ral parametrisation.
8/18/2019 Helical Fractal
11/40
Messy differentions and an ellipticintegral.
• The tro&(le !ith the helical fractal is that its arc length eval&ation re:&iressolving the elliptic integral of the second 1ind.
# 3ertainly not something that is analytically solva(le, and also
ma1es it very cl&msy in the attempt to define a general form&la.
• Another iss&e is !ith the many differentiations involved.
# Tho&gh differentiations are al!ays possi(le, applying too many
derivatives invites comp&tational pro(lems (eca&se certain
soft!ares li1e MAT$A? are not smart eno&gh to carry o&t sym(ollic differentiations, or other!ise not efficient.
# Nevertheless, n&merical differentiation schemes are of co&rse o&t
there as a last resort aid. 5?UT let
8/18/2019 Helical Fractal
12/40
A simpler !ay approach.
• Act&ally, !hat are needed are not necessarily the normal and the(inormal.
• %n fact, any perpendic&lar t!o vectors sitting in the normal plane !ill dothe ;o(. &st remem(er to divide (y its length to ma1e it a &nit vector.
• Biven the parametric e:&ations of a c&rve, a vector tangent to thec&rve is simply ;&st the derivative.
• A trivial vector that is normal to the tangent vector can easily (e
o(tained (y inspection. The second normal !o&ld then (e the crossprod&ct.
8/18/2019 Helical Fractal
13/40
The !ay to helicalise.
• Then, the oscillatory terms cos5>t6 and sin5>t6 can each (e added tothe t!o normal vectors.
• The process is not diffic< in theory, (&t the possi(ility of act&allysolving the alge(ra is :&estiona(le, (eca&se it involves the inverti(ilityof a '' matri and very long#!inded epressions.
• ?&t the pro(lem is finally solved, tedio&sly . Th&s, there lies thedefinition of the helical fractal, as the limit of repeated iteration of
the helicaliser form&lae.
8/18/2019 Helical Fractal
14/40
The f&ll derivation of the helicaliser.
8/18/2019 Helical Fractal
15/40
Tangent vectors and t!o normals.
8/18/2019 Helical Fractal
16/40
Celationship (et!een the ne! (asis!ith the original ones.
8/18/2019 Helical Fractal
17/40
Celationship (et!een the ne! (asis!ith the original ones.
8/18/2019 Helical Fractal
18/40
The c&rve in terms of the tangentvector and the normals.
8/18/2019 Helical Fractal
19/40
Addition of oscillatory terms.
8/18/2019 Helical Fractal
20/40
Accomplishment.
8/18/2019 Helical Fractal
21/40
=ormal mathematical definition ofthe helical fractal.
• The helicaliser ;&st derived is sort of a r&le that helicalises a c&rve!ith the inp&t of the parametric f&nction and their derivatives.
• ?y ta1ing the *ero#th level of the helical fractal as the straight line,
applying the helicaliser once gives a heli .
• Second application helicalises the heli to prod&ce !hat is called theslin1y .
• The reperc&ssion of many repeated applications !ill then res< in thehelical fractal.
8/18/2019 Helical Fractal
22/40
=ormal mathematical definition ofthe helical fractal.
So here is the long#a!aited definition7
8/18/2019 Helical Fractal
23/40
Presenting7 The 4elical =ractal .
$evel *ero7Straight line.
$evel one74eli. $evel t!o7
Slin1y.
8/18/2019 Helical Fractal
24/40
4elical =ractal7 Up close.
%t is a reminiscent ofitself at all scales definition of a fractal.
8/18/2019 Helical Fractal
25/40
lliptic integral of the second 1ind.
• A thoro&gh analysis on this fractal 5li1e its length and dimension6, especially!ith its mathematical definition in place sho&ld (e carried o&t to give a morein#depth &nderstanding of this fractal.
• 4o!ever as mentioned (efore, the arc length involves an elliptic integral of thesecond 1ind and th&s cannot (e analytically solved.
• This seems to ma1e the eval&ation of its dimension a (it complicated.
• %t has to (e noted that this is an a(sol&tely self#similar fractal.
•?&t !hat is its self#similar dimension, especially that its no! not a simple c&(eor !hatever, (&t a (east of coiling monster that is mind#(oggling
• Please do not offer the idea of eval&ating the 4a&sdoff dimension. Thee:&ation of the c&rve is already complicated eno&gh to !rite o&t on paper@@
8/18/2019 Helical Fractal
26/40
%ntrod&cing7The 3irc&lar =ractal .
• No!, there eists a different fractal, !hich some!hat possesses thesame manifestation as the helical fractal.
• %nstead of defining a c&rve (y coiling aro&nd it, the c&rve can (e
descri(ed (y circles that are normal to it .
• Then starting at level one !ith a circle, the net level !o&ld (e to removethis circle, and replace it !ith > smaller circles that defines the circle.
•
4o! it defines the original circle is (y laying normal to it.
• Pict&res spea1 lo&der than !ords, tho&gh !itho&t emitting any so&nd.4ere goes.
A discrete version,
8/18/2019 Helical Fractal
27/40
Presenting7 The 3irc&lar =ractal .
=irst level7one circle.
Second level7>EF circles, AEF .
Third level7>EFG circles, AEFG .
The 3irc&lar =ractal is the limit !here > more circles smaller (y a factor of A
defines the circles of the previo&s level.
Hf co&rse, the infinite n&m(er of circlesget so small s&ch that its infinitesimal si*e
ma1es it almost invisi(le.
Note that its self#similar dimension is one.5> copies scaled (y a factor A, !here >EAEF .6
8/18/2019 Helical Fractal
28/40
3om(ining together all levels,
• -ell, the Circular Fractal is a fractal. %t has a self#similar dimensionof d E ln>I ln A .
• ?&t ;&st going ;&st slightly more advent&ro&s as to piece &p A$$ levelsof the circ&lar fractal, the res< is the Super Circular Fractal .
%ntrod&cing7 The S&per 3irc&lar =ractal .
8/18/2019 Helical Fractal
29/40
4elical =ractal , A&iliary =ractal ,Discrete 4elical =ractal .
• %n this case, the circ&lar fractal is called an auxiliary fractal , since it is &sedto (&ild &p the s&per circ&lar fractal.
• No!, the circ&lar fractal is essential a discrete version of the helical fractal.4ence it is d&((ed as the
• So here, everything (ecomes o(vio&s. All these three fractals have a self#similar dimension of d E ln>I ln A .
• %n other !ords, the dimension of the fractal highly depends on relationship(et!een the n&m(er of ne! copies as !ell as the scale factor.
Discrete 4elical =ractal.Discrete 4elical =ractal.
8/18/2019 Helical Fractal
30/40
S&per 4elical =ractal
• ?&t !ait. %t seems really nat&ral that the s&per circ&lar fractal eists.There is no do&(t that each smaller circle is an eact copy of thelarger preceding level.
• 4ave another good loo1 at it.
8/18/2019 Helical Fractal
31/40
%ntrod&cing7 The S&per 4elical =ractal .
• Since the circ&lar fractal isthe discrete helical fractal,the helical fractal sho&lditself also (e an a&iliary
fractal.
• Piecing &p all its individ&allevels event&ally forms theS&per 4elical =ractal .
• 4ave a good loo1 at it.
%ndeed@
8/18/2019 Helical Fractal
32/40
Analysis7 Arc length.
8/18/2019 Helical Fractal
33/40
Ma1ing arc length finite.
8/18/2019 Helical Fractal
34/40
The :&antity associated !ith arclength.
Note that from the originalform&la for arc length, it is here!here the po!er n occ&rs s&chthat the nat&ral logarithms areta1en.
8/18/2019 Helical Fractal
35/40
3omparison (et!een ? and d .
8/18/2019 Helical Fractal
36/40
Celationship (et!een ? and d .
8/18/2019 Helical Fractal
37/40
8/18/2019 Helical Fractal
38/40
The fo&r most (ea&tif&l fractals inthis &niverse)4elical =ractal
3irc&lar =ractal
discrete version
piecing &p individ&al levels)
S&per 3irc&lar =ractal
contin&o&s version
individ&al fractal
S&per 4elical =ractal
=o&r different fractals !ith thesame manifestations)
8/18/2019 Helical Fractal
39/40
Ac1no!ledgements.
Dr. =edor D&*hin. Assistant ProfessorAndre! ames Jric1er.
Assistant Professor3he! $oc1 K&e.
8/18/2019 Helical Fractal
40/40
Ceference.
• ?arrett HLNeill, lementary Differential Beometry. P&(lished (ylsevier Academic Press, second edition.