+ All Categories
Home > Documents > Opportunities for Optimization in Ship Design Jay Warren Hull Engineering Newport News Shipbuilding...

Opportunities for Optimization in Ship Design Jay Warren Hull Engineering Newport News Shipbuilding...

Date post: 08-Jan-2018
Category:
Upload: silvia-watts
View: 221 times
Download: 0 times
Share this document with a friend
Description:
© Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 3 of 16 Implementation of Large-Scale Optimization for Ship Design? Must use an incremental approach that identifies the benefits of optimization –Demonstrate and develop design and analysis tools Show the accuracy/effectiveness of existing design/analysis tools Develop/evolve design/analysis tools to handle non-traditional loads –Use existing analysis tools to help the Navy conduct small-scale design sensitivity studies –Closely couple analysis tools to perform large-scale design sensitivity and optimization studies

If you can't read please download the document

Transcript

Opportunities for Optimization in Ship Design Jay Warren Hull Engineering Newport News Shipbuilding Newport News, VA MAD Center Advisory Board Meeting April 25, 2003 Copyright 2003 by Northrop Grumman Newport News All Rights Reserved This document is protected by United States Copyright Law and shall be treated as proprietary to Northrop Grumman Newport News under the Trade Secrets Act and other applicable laws. This document may not be disclosed to, copied by, or used in any manner by others without the prior express, written permission of Northrop Grumman Newport News. For further information, contact Northrop Grumman Newport News Legal Department, 4101 Washington Ave., Newport News, Virginia This document also may contain information that is export-restricted under United States Export Control Laws; therefore, it shall not be exported or otherwise disclosed to foreign persons without an appropriate export license, if required, from the United States Department of State. Distribution within Northrop Grumman Newport News is limited to employees with a genuine need to know. Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 2 of 16 Presentation Overview What approach should be taken to implement large- scale ship design optimization? Where we stand right now What research opportunities are related to analysis and optimization of ships? Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 3 of 16 Implementation of Large-Scale Optimization for Ship Design? Must use an incremental approach that identifies the benefits of optimization Demonstrate and develop design and analysis tools Show the accuracy/effectiveness of existing design/analysis tools Develop/evolve design/analysis tools to handle non-traditional loads Use existing analysis tools to help the Navy conduct small-scale design sensitivity studies Closely couple analysis tools to perform large-scale design sensitivity and optimization studies Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 4 of 16 Implementation of Large-Scale Optimization for Ship Design? Traditional ship design software ( FLAGSHIP, TRIBON, ) could form the core for large-scale optimization efforts Extend traditional software to include good approximations that account for non-traditional loads Use the traditional software to generate a baseline design or to modify an existing design Use detailed finite element analyses to further optimize the baseline design for non-traditional loads Most analyses for non-traditional loads are very expensive Hopefully further optimization will only result in moderate changes from the baseline design Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 5 of 16 Where We Stand Right Now Currently use three types of software Traditional ship design software (FLAGSHIP,TRIBON, ) Handle ~80-90% of ship design but are not intended to predict the effects of non-traditional loads Conventional finite element software Great for solving detailed solid mechanics problems but lack Navy specific capability Specialty software for non-traditional analyses Detailed ship motion (LAMP) Underwater explosions (USA, DYSMAS/GEMINI) Acoustic and Non-acoustic signatures (SARA, SONAX) Hydrodynamics (Navy proprietary) Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 6 of 16 Where We Stand Right Now Have some experience with large finite element models and massively parallel finite element software Full Model 750,000 Elements Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 7 of 16 USS Saratoga after the BAKER explosion, 26 Jul 1946 Research Opportunities: Non-Traditional Analyses The Effects of Nuclear Weapons, S. Glasstone and P.J. Dolan, USDOD & ERDA, 1977, p 250. Airblast Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 8 of 16 Research Opportunities: Non-Traditional Analyses HMAS TORRENS Torpedo Attack Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 9 of 16 Research Opportunities: Non-Traditional Analyses Mine Warfare USS TRIPOLI Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 10 of 16 Research Opportunities: Non-Traditional Analyses USS STARK Missile Attack Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 11 of 16 Research Opportunities: Non-Traditional Analyses USS Theodore Roosevelt (CVN71) Large Standoff UNDEX Loading Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 12 of 16 Research Opportunities: Non-Traditional Analyses Acoustic / Non-Acoustic Signatures Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 13 of 16 Research Opportunities: Non-Traditional Analyses Large-scale analyses that are capable of detailed damage predictions Most ship damage involves ductile failure Need to at least gain some understanding of spalling and fragmentation Rational method for including uncertainty in non- traditional analyses Include uncertainty in loads, geometry, and properties Predict confidence in resulting damage Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 14 of 16 Seaway Loading & Ship Motion USS Enterprise (CVN65) Research Opportunities: Close Coupling of Analysis Codes Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 15 of 16 Research Opportunities: Close Coupling of Analysis Codes Efficient large-scale optimization will require all codes to be closely coupled Efficiency indicates that non-traditional analyses might need to be massively parallel Close coupling implies that all codes should be built around a common framework What type of framework is best? Copyright 2003 by Northrop Grumman Newport News - All Rights Reserved Refer to restrictions on the first page of this document. 04/25/03 16 of 16 Summary Use a multi-step process to implement optimization Research opportunities exist for relaxing some of the current constraints caused by non-traditional loads Use detailed finite element analyses to optimize a baseline design Existing ship design Ship design created from updated versions of traditional ship design and optimization software Detailed finite element results must be produced very efficiently Closely coupled algorithms Massively parallel algorithms


Recommended