Electro – optic – Magneto Optic and Acousto – Optic effects (1)
ECE 695 –Numerical Simulations of Electro-optic Energy Systems
Transcript of ECE 695 –Numerical Simulations of Electro-optic Energy Systems
![Page 1: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/1.jpg)
ECE 695 – Numerical Simulations
of Electro-optic Energy Systems
Peter Bermel
January 9, 2017
1/9/2017 ECE 695, Prof. Bermel 1
![Page 2: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/2.jpg)
Outline
• Motivation
• My Background and Research
• Topics for This Class
• Goals for This Class
• Assignments
• Grading
1/9/2017 ECE 695, Prof. Bermel 2
![Page 3: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/3.jpg)
Motivation for This Class
• Teach new investigators how to
use computers to achieve their
research goals
• “The purpose of computing is
insight, not numbers!” –
Richard W. Hamming
1/9/2017 ECE 695, Prof. Bermel
RW Hamming (left),
developing error-
correcting codes (AT&T)
3
![Page 4: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/4.jpg)
My Background and Research
• All degrees in Physics
• Master’s degree at Cambridge University: linear photonic bandstructures
• Completed Ph.D. at MIT on active materials in photonic crystals (Advisor: JD Joannopoulos)
• Continued with postdoc on applications in photovoltaics & thermophotovoltaics (Advisor: M Soljacic)
• Current work includes PV, TPV, solar thermal, and quantum optics
1/9/2017 ECE 695, Prof. Bermel 4
![Page 5: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/5.jpg)
1/9/2017 ECE 695, Prof. Bermel 5
![Page 6: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/6.jpg)
1/9/2017 ECE 695, Prof. Bermel 6
![Page 7: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/7.jpg)
Light Management in Photovoltaics
1/9/2017 ECE 695, Prof. Bermel
Si thin film
2 μm
PhC thin film
2 μm
silicon
aluminum
Wafer cell
200 μm
silicon
glass
Absorbed
Lost energy
Absorbed
7
![Page 8: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/8.jpg)
1/9/2017 ECE 695, Prof. Bermel 8
![Page 9: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/9.jpg)
1/9/2017 ECE 695, Prof. Bermel 9
![Page 10: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/10.jpg)
1/9/2017 ECE 695, Prof. Bermel 10
![Page 11: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/11.jpg)
1/9/2017 ECE 695, Prof. Bermel 11
![Page 12: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/12.jpg)
Thermophotovoltaics (TPV) Enables
Unique Energy Systems
1/9/2017 ECE 695, Prof. Bermel
µTPV portable power generator* Solar TPV utility scale electricity†
RTPV for long, remote missions‡
*R. Pilawa-Podgurski et al., APEC 25, 961 (2010); P. Bermel
et al., Opt. Express 18, A314 (2010)
‡ A. Schock et al., Acta Astronaut. 37, 21 (1995); S.-Y. Lin et
al., Appl. Phys. Lett. 83, 380 (2003); D. Wilt et al., AIP Conf.
Proc. 890, 335 (2007)
† M. Castro et al., Solar Energy Mater. Solar Cells 92, 1697
(2008); E. Rephaeli & S. Fan, Opt. Express 17, 15145
(2009)
12
![Page 13: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/13.jpg)
Topics Covered In This Class
1/9/2017 ECE 695, Prof. Bermel 13
![Page 14: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/14.jpg)
Computational Complexity
• Study of the complexity of
algorithms
• Based on Turing machines
• Often, one compares
algorithms for best scaling
in large problems
1/9/2017 ECE 695, Prof. Bermel
Alan Turing (from University
of Calgary Centenary event)
14
![Page 15: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/15.jpg)
Finding Zeros
• Key concept: 1D bracketing
• Bisection – continuously halve intervals
• Brent’s method – adds inverse quadratic interpolation
• Newton-Raphson method –uses tangent
• Laguerre’s method – assume spacing of roots at a and b:
1/9/2017 ECE 695, Prof. Bermel 15
![Page 16: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/16.jpg)
Finding Minima (or Maxima)
• Golden Section Search: local bisection
• Brent’s Method: quadratic fit + fallback
• Downhill Simplex
• Conjugate gradient methods
• Multiple level, single linkage (MLSL): global, derivative
1/9/2017 ECE 695, Prof. Bermel
These and further images from “Numerical
Recipes,“ by WH Press et al.
16
![Page 17: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/17.jpg)
Eigenproblems
• Generalized eigenproblem: �� = ���
• Solution method will depend on properties of
A and B
• Techniques have greatly varying
computational complexity
• Sometimes, full solution is unnecessary
1/9/2017 ECE 695, Prof. Bermel 17
![Page 18: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/18.jpg)
Eigenproblems
• Direct method: solve det(A-λ1)=0
• Similarity transformations: A� Z-1AZ
– Atomic transformations: construct each Z explicitly
– Factorization methods: QR and QL methods
1/9/2017 ECE 695, Prof. Bermel 18
![Page 19: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/19.jpg)
Photonic Crystal Bandstructures
1/9/2017 ECE 695, Prof. Bermel
• Periodic (crystalline)
media
– Periodic atoms:
semiconductors with
electronic bandgaps
– Periodic dielectrics:
photonic crystals with
photonic bandgaps
• Many potential
applications for both
periodic crystalline structures
PBG for diamond structure
1D 2D 3D
Joannopoulos et al., Photonic Crystals (2008)
19
![Page 20: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/20.jpg)
Discrete Fourier Transforms
• DFT defined by: � � =∑ (��)
����(���/��)����
• Naïve approach treats each frequency individually
• Can combine operations together for significant speed-up (e.g., Cooley-Tukey algorithm)
• Specialized algorithms depending on data type
1/9/2017 ECE 695, Prof. Bermel
J.W. Cooley (IEEE
Global History
Network)
20
![Page 21: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/21.jpg)
Cooley-Tukey Algorithm
1/9/2017 ECE 695, Prof. Bermel 21
![Page 22: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/22.jpg)
Transfer Matrices and Rigorous
Coupled Wave Analysis
• Divide space into layers for efficiency
• For uniform layers – transfer matrix approach
• For periodic gratings or similar in certain layers –
RCWA
1/9/2017 ECE 695, Prof. Bermel
From the CAvity Modeling FRamework (CAMFR)
22
![Page 23: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/23.jpg)
Finite-Difference Time Domain
1/9/2017 ECE 695, Prof. Bermel
• Discretize space and time on a Yee
lattice
• “Leapfrog” time evolution of
Maxwell’s equations:
• Implemented in MEEP:
nanohub.org/topics/MEEP
µ
ε
σ
σ
/BH
ED
DJHdt
dD
BJEdt
dB
D
BB
=
=
−−×∇=
−−×∇−=
rrr
rr
23
![Page 24: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/24.jpg)
1/9/2017 ECE 695, Prof. Bermel
Finite-Element Methods
• For 2D or 3D problems, divide space into a mesh
• Solve a wide array of partial differential equations – well
suited for multiphysics problems
24
![Page 25: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/25.jpg)
Ordinary Differential Equations
• Euler method – naïve rearrangement of ODE
• Runge-Kutta methods – match multiple Euler
steps to a higher-order Taylor expansion
• Richardson extrapolation – extrapolate
computed value to 0 step size
• Predictor-corrector methods – store solution
to extrapolate next point, and then correct it
1/9/2017 ECE 695, Prof. Bermel 25
![Page 26: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/26.jpg)
ODE Boundary Value Problems
1/9/2017 ECE 695, Prof. Bermel
Boundary Value MethodShooting Method
26
![Page 27: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/27.jpg)
Partial Differential Equations
• Classes: parabolic, hyperbolic, and elliptic
• Initial value vs. boundary value problems
• Finite difference
• Finite element methods
• Monte-Carlo
• Spectral
• Variational methods
1/9/2017 ECE 695, Prof. Bermel 27
![Page 28: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/28.jpg)
Goals for This Class
• Learn/review key mathematics
• Learn widely-used numerical techniques just discussed
• Become a capable user of software utilizing these techniques
• Appreciate strengths and weaknesses of competing algorithms; learn how to evaluate the results
• Convey your research results to an audience of your new colleagues
1/9/2017 ECE 695, Prof. Bermel 28
![Page 29: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/29.jpg)
Key Policies
• Textbooks:– Salah Obayya, “Computational Photonics”
– JD Joannopoulos et al., “Photonic Crystals”
– WH Press et al., “Numerical Recipes in C”
• Communication:– Course website:
http://web.ics.purdue.edu/~pbermel/ece695/
– nanoHUB group: https://nanohub.org/groups/ece695
– Email: [email protected]
• Full list of policies given on handout; also available under Syllabus on Blackboard
1/9/2017 ECE 695, Prof. Bermel 29
![Page 30: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/30.jpg)
Quizzes
• Will periodically post questions online about
the lecture
• No trick questions – just designed to make
sure you’re keeping up with the material
1/9/2017 ECE 695, Prof. Bermel 30
![Page 31: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/31.jpg)
Class Participation
• Your attendance is important
• Will be grading your involvement, enthusiasm,
and respect for your peers in the class
• Not grading your percentage of correct
answers during class
1/9/2017 ECE 695, Prof. Bermel 31
![Page 32: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/32.jpg)
Homework
• Homework is essential to learn the material
• 10 total homework assignments this semester,
once every 1.5 weeks on average. First one
will be available Jan. 13, and due Jan. 20
• Due at 4:30 pm on the listed dates to
1/9/2017 ECE 695, Prof. Bermel 32
![Page 33: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/33.jpg)
Final Project
• Chance for you to teach the rest of the class about a numerical computing topic based on the class that interests you!
• OK to pick something related to your research as long as it’s new
• I can suggest topics if you’re not sure what to do
• Can ask peers for general advice but all details and presentations should be done by you
1/9/2017 ECE 695, Prof. Bermel 33
![Page 34: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/34.jpg)
Grading
Grading Item Points Date
Quizzes 100 Various
Class Participation 100 All Semester
Homework 100 Every Other Week
Final Project 200 End of Semester
TOTAL 500
• Numerical grades of 60% or above will pass
• Roughly speaking: A’s will be 90% +; B’s 80-89%; C’s 70-79%
• Final letter grades will be assigned at my discretion
1/9/2017 ECE 695, Prof. Bermel 34
![Page 35: ECE 695 –Numerical Simulations of Electro-optic Energy Systems](https://reader031.fdocuments.us/reader031/viewer/2022020702/61faa7ac78c8137dfa7f451b/html5/thumbnails/35.jpg)
Next Class
• Introduction to computational complexity
• Please read Chapter 1 of “Computational Complexity: A Modern Approach” by Arora& Barak
1/9/2017 ECE 695, Prof. Bermel 35