Integration of High-Fidelity Analysis Tools in MDO Frameworks for HPC

Backhaus, Thomas; Gottfried, Sebastian; Ilić, Časlav; Merle, Andrei; Stueck, Arthur

doi:10.2514/6.2019-3107

Cited by 3 publications

(5 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3). The main goal of this approach, which is also referred to as Cybermatrix, is to deal with the complexity of real-world MDO and to enable effective involvement of many disciplines [14]. The Cybermatrix MDO protocol is a novel approach towards computationally demanding and collaboration intensive MDO.…”

Section: Overview Of Mdo Architectures and Process Chainsmentioning

confidence: 99%

“…The fuel mass is then sent from each planform back to the optimizer, to start the next base period. This is just one possible implementation pattern, which can be easily reconfigured within the process integration framework called MDO Driver [14]. The particular pattern on Fig.…”

Section: Fig 15 Many Discipline Highly-parallel Approach: Process Imp...mentioning

confidence: 99%

“…While an entirely new multi-disciplinary HPC integration framework was developed in VicToria to address the specific needs of the many-discipline highly-parallel approach [14], the integration environment RCE (Remote Component Environment) [29], [30] was extended in order to model the construction and execution of intertwined MDO processes and sub-processes. RCE allows users to implement multidisciplinary design optimization processes using arbitrary disciplinary tools such as structural solvers or optimizers.…”

Section: F Integration Environmentsmentioning

confidence: 99%

See 2 more Smart Citations

Overview of Collaborative Multi-Fidelity Multidisciplinary Design Optimization Activities in the DLR Project VicToria

Goertz

Abu-Zurayk

Ilić

et al. 2020

Aiaa Aviation 2020 Forum

View full text Add to dashboard Cite

Section: Overview Of Mdo Architectures and Process Chainsmentioning

confidence: 99%

Section: Fig 15 Many Discipline Highly-parallel Approach: Process Imp...mentioning

confidence: 99%

Section: F Integration Environmentsmentioning

confidence: 99%

See 1 more Smart Citation

Overview of Collaborative Multi-Fidelity Multidisciplinary Design Optimization Activities in the DLR Project VicToria

Goertz

Abu-Zurayk

Ilić

et al. 2020

Aiaa Aviation 2020 Forum

View full text Add to dashboard Cite

“…A cybermatrix-compliant HPC integration framework MDO Driver [37], which drove the demonstration cases in Sec. V, was developed concurrently with the work presented here.…”

Section: Parallel Collaboration and Executionmentioning

confidence: 99%

Cybermatrix Protocol: A Novel Approach to Highly Collaborative and Computationally Intensive Multidisciplinary Aircraft Optimization

Ilić

Merle

Abu-Zurayk

et al. 2020

Aiaa Aviation 2020 Forum

View full text Add to dashboard Cite

“…This motivated moving to an all-HPC based solution, called Cybermatrix MDO protocol, in the follow-on DLR project VicToria [30]. The Cybermatrix MDO protocol is a novel approach towards computationally demanding and collaboration intensive MDO and its implementation is described in detail in [31].…”

Section: Fig 12 Rce Implementation Of the Digital-x Mdo Processmentioning

confidence: 99%

Distributed Multidisciplinary Optimization and Collaborative Process Development Using RCE

Boden

Flink

Mischke

et al. 2019

AIAA Aviation 2019 Forum

Self Cite

View full text Add to dashboard Cite

DLR's Remote Component Environment (RCE) is an open-source software environment for defining and executing workflows containing distributed simulation tools by integrating them into a peer-to-peer network. It is being developed primarily by DLR and has been used in various engineering projects, including several aerospace projects dealing with multidisciplinary optimization (MDO) and multidisciplinary analysis (MDA). RCE has several advantages that can help to achieve more reusable multidisciplinary processes. The workflow is composed of built-in and user-defined components. Disciplinary tools are integrated as standalone components, with defined inputs and outputs, and then distributed over the network. While executing the workflow, data dependencies between the components are automatically detected, and a component is executed as soon as all its input data is available. Thus, multiple components can run at the same time. The components of a multidisciplinary process can also be executed in a distributed manner, where the tools are located on different machines with possibly different operating systems. Once configured, the peer-to-peer network is automatically established between the RCE instances running on different machines, making components visible and executable even between instances that are only connected indirectly. The distributed execution capability alleviates tool deployment issues, including those related to the protection of intellectual property. RCE supplies a graphical editor for creation of workflows, using the built-in components to control the data flow. Some built-in components can be used to perform optimization tasks within the workflow, including nested loops, using built-in or user integrated optimization algorithms. We demonstrate the use of some of the key features of RCE for MDA and MDO purposes based on several collaborative DLR projects where distributed simulation tools are being used.

show abstract

Integration of High-Fidelity Analysis Tools in MDO Frameworks for HPC

Cited by 3 publications

References 7 publications

Overview of Collaborative Multi-Fidelity Multidisciplinary Design Optimization Activities in the DLR Project VicToria

Overview of Collaborative Multi-Fidelity Multidisciplinary Design Optimization Activities in the DLR Project VicToria

Cybermatrix Protocol: A Novel Approach to Highly Collaborative and Computationally Intensive Multidisciplinary Aircraft Optimization

Distributed Multidisciplinary Optimization and Collaborative Process Development Using RCE

Contact Info

Product

Resources

About