Transcript CREATE-RF Status - National Defense Industrial Association

CREATE-RF Status
National Defense Industrial Association
February 1, 2011
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-1
Distribution Statement A: Approved for public release; distribution is unlimited.
CREATE-RF Development
Product Description
Product: SENTRI
SENTRI – Scalable Engineering Tools for RF Integration
What it is:
• Computer Aided Engineering Software for DoD Electromagnetic Applications
• Designed for High Accuracy – Full Wave (non-optical) Numerical Methods
‒ Finite Elements
‒ Boundary Integral
‒ Harmonic Expansions
• Designed for Extensibility, Maintainability, and Flexibility
‒ Not All Electromagnetic Applications are the same
‒ Need for Specific and Tailored Methods for Unique Applications
• Designed to Run on Wide Range of Computers
‒ from Engineering Workstations to High Performance Computers
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-2
Distribution Statement A: Approved for public release; distribution is unlimited.
CREATE-RF Requirement Summary
• Antennas on Air, Sea, Ground, And Space Platforms
• Communication, Navigation, Surveillance, Target Recognition,
Electronic Attack, Countermeasure, Etc.
Global Hawk
Computational Electromagnetics Applies to Almost All DoD Systems
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-3
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach Background
Benefits to DoD Acquisition Community
‘First-Pass’ Designs
• Develop High-Fidelity Codes to Model All Design Variables Reducing
Risk That Design Goals Will Not Be Met
Less Prototyping  Faster and More Productive Engineering
• Hardware Prototyping is Time Consuming, Expensive, Provides Little
Insight to Physical Phenomena
New Concept Designs  Higher Performance at Lower Cost
• Methods Allow for the Exploration of Entire Design Space
• Higher Performance Requires New Designs and Materials to be Explored
A Single DoD Vehicle can have More Than 100 Antenna Systems
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-4
Distribution Statement A: Approved for public release; distribution is unlimited.
CREATE-RF Development
Product Description
Challenge of Computational Electromagnetics
• Complex Geometries, Complex Material Application, Multi-Scale
Geometries
• Computationally Expensive for Accurate Full-Wave Analysis
• Electromagnetic Phenomena (Singularities, Resonances, Wide-Band)
Prof. Jin-Fa Lee
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-5
Distribution Statement A: Approved for public release; distribution is unlimited.
CREATE-RF Development
Product Description
Challenge of Computational Electromagnetics
Unknown Count with Notional Aircraft
10
1010
Problem Size (Unknowns)
VHF
UHF
L
S
C
X Ku K Ka
99
10
10
20M
88
10
10
5M
500M
1M
7
10
107
180M
• Length = 15.5 m
50M
200K
• Width = 8.9 m
6
10
106
• Height = 4.1 m
20K
55
10
10
• 100 MHz to 40 GHz
• 10 samples/λ
44
10
10
33
10
10
22
10
10
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-6
3
10
10
4
10
10
Frequency (MHz)
55
10
10
Source: Riverside Research Institute
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Product Roadmap
SENTRI at version 1.5
Antenna Modeling
• Patch, Notch, Horn, Spirals (Applications: Radar, Communication, GPS)
• Phased Array Antennas
• Cavity Backed Antenna (Approximate In-Situ Analysis)
Periodic Structures
• Frequency Selective Surfaces
• Circuit Analog Absorbers
• Metamaterials
• Infrared Filters / Absorbers
Microwave Circuits
• Power Splitting
• Material Measurement
• Filters
• Circulators
Radar Cross-Section Prediction
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-7
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Product Roadmap
Future Releases of SENTRI
Full End-to-End Analysis System
‒
‒
‒
‒
‒
Graphical User Interface
Material Database
Visualization of Solutions
Analysis Traceability
Optimization
Programmable System
‒ Software Release as a Application
Programming Interface for Further
Tailoring by End User
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-8
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Development Strategy
Rapid Design Codes
What Engineers Need
for Their Design Work
High Fidelity Codes
Design Optimization
Techniques Integrated with
Rapid Design / High-Fidelity
Codes
Making Engineers
Productive by
Automating Design Tools
Specialized Codes for
Particular Applications
Big Payoff – Solving
Problems that Others Can’t
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-9
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Development Strategy
Design Space
Rapid
Design
Methods
High
Fidelity
Methods
Optimal
Design
Rapid Design Methods  Narrow Design Space
High Fidelity Methods  Model All Variables, Optimal Design
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-10
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Development Strategy
Rapid Design Codes
What Engineers Need
for Their Design Work
High Fidelity Codes
Design Optimization
Techniques Integrated with
Rapid Design / High-Fidelity
Codes
Making Engineers
Productive by
Automating Design Tools
Specialized Codes for
Particular Applications
Big Payoff – Solving
Problems that Others Can’t
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-11
Distribution Statement A: Approved for public release; distribution is unlimited.
Development Approach
Development Strategy
Rapid Design Codes
What Engineers Need
for Their Design Work
High Fidelity Codes
Design Optimization
Techniques Integrated with
Rapid Design / High-Fidelity
Codes
Specialized Codes for
Particular Applications
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-12
Making Engineers
Productive by
Automating Design Tools
Big Payoff  Solving
Problems that Others Can’t
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
Microwave Circuit Analysis
Sanity Check - Code to Code Comparison: SENTRI vs. CST Microwave Studio
High Dielectric Coupled-Ring Filter
ε = 38
Air
(both)
Dielectric
Coax
ε = 2.4
Coax
ε = 2.4
Normal H on
symmetry plane.
Successfully Benchmarked w/ Independent Software Vendor Tools
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-13
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
Metamaterials
Sanity Check – SENTRI vs. Published Results
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-14
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
Metamaterials
Sanity Check – SENTRI vs. Published Results
SENTRI
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-15
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
Printed Circuit Antennas
SENTRI vs. Commercial Codes
Radiating
slotline mode.
λ/4 reflector
y
Radial stub.
x
Exponential tapered
slot, cut into metal.
Near field is post-processed, could be
used as illumination for other problems.
- 10
- 20
- 30
Legend
(v1)
- 40
Return Loss (S11) dB
0
Stripline feed.
Solution
domain.
(v1.5)
4
5
6
7
8
Frequency, GHz
Successfully Benchmarked w/ Independent Software Vendor Tools
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-16
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
Antenna Patch Array
A large printed array is an antenna-type problem that also benefits.
Gain (left) & field structure (right).
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-17
Distribution Statement A: Approved for public release; distribution is unlimited.
Application Examples
In-Situ Antenna Analysis
5
0
-5
-15 -10
Metal
Isolated
In-situ
0
-50 -40 -30 -20 -10
-180
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-18
E-plane
10
Metal
Gain, dBi
12’, 40 λ @ 3.3GHz, 78K / 96K FE/BI unknowns
ε = 1 (air)
Gain, dBi
ε = 1.8 (radome)
Isolated
In-situ
10
Lofted horn on notional missile.
H-plane
-120
-60
0
60
Angle off boresight, degrees.
Distribution Statement A: Approved for public release; distribution is unlimited.
120
180
Analysis Scalability
Running Large, High Fidelity Models
Big Computers Not
the Total Solution
Big Computers +
Algorithms that Scale
with Problem Size
Time to Solution
O(N3)
P
P
P
Desired endpoint:
asymptotically
competitive algorithms,
scaled well to next-gen
HPC platforms
Problem Size
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-19
Distribution Statement A: Approved for public release; distribution is unlimited.
Conclusion
• Computational Electromagnetic Applications are Challenging
but with High Performance Computers and Advanced Numerical
Methods they can be Solved
• CREATE-RF Software Development
‒ For Use by DoD Government and US Defense Industry Engineers
‒ Allows for Specific Tailoring for Special Applications
‒ Tech-Transfer Mechanism for Next-Gen of Gov’t Engineers
• CREATE-RF Transfer to Military Industry
‒ Complete Analysis Package
‒ Runs on Workstations and High Performance Computers
‒ Programmable System for End User Extensibility
CREATE RF-John D’Angelo-NDIA
03/08/2011 Page-20
Distribution Statement A: Approved for public release; distribution is unlimited.