Upload
philippa-owens
View
216
Download
0
Tags:
Embed Size (px)
Citation preview
Simple Nonlinear ModelsSuggest Variable Star
UniversalityJohn F. Lindner, Wooster College
Presented by John G. LearnedUniversity of Hawai’i at Mānoa
Collaboration
John F. LindnerThe College of Wooster
Vivek KoharNorth Carolina State University
Behnam KiaNorth Carolina State University
Michael HippkeInstitute for Data AnalysisGermany
John G. LearnedUniversity of Hawai’iat Mānoa
William L. DittoNorth Carolina State University
Multi-Frequency Stars
P. Moskalik, “Multi-mode oscillations in classical cepheids and RR Lyrae-type stars”,
Proceedings of the International Astronomical Union 9 (S301) 249 (2013).
Petersen Diagram
Petersen Diagram Rescaled
Spectral Distribution
Strobe signal at primary period and plot its values versus time
modulo secondary periodto form the Poincaré section
If the function
represents the section,is it smooth?
Expand function & its derivativesin Fourier series
For smooth sections, expect Fourier coefficients to decay exponentially, so all the derivatives also decay
For nonsmooth sections, expect Fourier coefficients to decay slower, as a power law, so that some derivatives diverge
Invert to get
Since an averaged spectrum decreases with mode number or frequency
reinterpret
to be the number of super threshold spectral peaks
So rich, rough spectrahave power law spectral distributions
Stellar Analysis
KIC 5520878 Normalized Flux Sample
Lomb-Scargle Periodogram
Spectral Distribution
Gutenberg-Richter law forvolcanic Canary Islands earthquake distribution
Some Number Theory
Golden Ratio
Slow convergence suggests maximally irrational
Liouville Number
Example of nearly rational irrational
Model 1:Finite Spring
Network
Natural Frequencies
Model 2:Hierarchical
Spring Network
Natural Frequencies
Model 3:Asymmetric Quartic
Oscillator
Asymmetric Quartic Potential Energy
Sinusoidal Forcing
Drive Frequency a Golden Ratio Above Natural Frequency
Model 4:Pressure vs. Gravity
Oscillator
Adiabatic Simplification
Pressure & Volume
Potential Energy
Force
Sinusoidal Forcing
Model
Data
Model 5:Autonomous Flow
generalized Lorenz convection flow caused bythermal & gravity gradients plus vibration
Adjust parameters so that
Model 6:Twist Map
Twist map is circles for vanishing push
Push perturbation has vanishing mean
Least resonant golden shift remains
Insights from Helioseismology
Helioseismology and asteroseismology have observed manyseismic spectral peaks in the sun and other nonvariable stars,
which correspond to thousands of normal modes
Yet, despite preliminary analysis, we have not discoveredpower law scaling in the solar oscillation spectrum
Stochasticity and turbulence dominate the pressure waves inthe sun that produce its standing wave normal modes
In contrast, a varying opacity feedback mechanism inside avariable star creates its regular pulsations
In golden stars, interactions with this pulsating mode maydissipate all other modes except those a golden ratio away
Discussion
The simple nonlinear models suggest the importance of considering simple explanations
0: The golden ratio itself has unique and remarkableproperties; as the irrational number least well approximated by
rational numbers, it is the least “resonant” number
1: A finite network model of identical springs and masses hastwo normal modes whose frequency ratio is golden
2: An infinite network hierarchy can be mass terminated in twoways to naturally generate two modes whose frequency ratio is
golden, while a realistic truncation of the model generates aratio near golden, as observed in the golden stars
Summary
3: A simple asymmetric nonlinear oscillator produces a richspectrum with a power-law spectral distribution
4: A more realistic oscillator model of pressure counteringgravity exhibits a recognizable but stylized golden
star attractor
5: An unforced Lorenz-like convection flow also produces asingular spectrum with a power-law spectral distribution,provided its parameters are tuned so that a golden ratio
characterizes its orbit
6: An ensemble of twist maps naturally evolve to a goldenstate, because golden shifts are least resonant with any
oscillatory perturbation
Summary
The Feigenbaum constant delta ~4.67,which characterizes the period doubling route to chaos,
has been observed in many diverse experiments
Does the golden ratio ~1.62,or equivalently the inverse golden ratio ~0.62,
play a similar role?
Or does the mysterious factor of ~0.62, which characterizes many multifrequency stars,
merely result from nonradial stellar oscillation modes?
Simplicity vs. Complexity
Some natural dynamical patterns result from universal features common to even simple models
Other patterns are peculiar to particular physical details
Is the frequency distribution of variable stars universal or particular?
Universality vs. Particularity
Thanks for Listening