Context
Phased array antennas are advanced antenna systems made of many small individual elements that work together to steer radio wave beams electronically without any physical movement. They are typically designed for a specific mission under hard constraints: cost, weight, volume, and available prime power.
At Northrop Grumman Mission Systems the analysis that answered those constraints lived in separate places. Individual performance models covered weight, power, performance, and reliability, while descriptive models held architecture and redundancy, and results moved between them by meeting, phone call, and email. Determining the performance of a single candidate configuration typically took several days. At that price a program evaluates a handful of candidates and then defends the one it started with.
Value Thinking
Define Instead of asking whether one design met the requirement, the team asked what each increment of performance costs in weight, prime power, and dollars. Prioritize Program-specific objectives, constraints, and input parameters define a design of experiments. Inputs include frequency, bandwidth, array grid, amplifier power per element, and required signal-to-noise ratio; outputs include link margin, beamwidth, size, weight, prime power, and cost. Align Thousands of configurations are evaluated and plotted as a cost-versus-performance Pareto front, which puts the commercial constraint into the design space as a variable. What reaches the customer is then a best-value choice out of a priced set, with the trade between array cost and weight visible instead of asserted.
Architect for Flow
Structure This only works if the models compose. The discipline models, covering aperture, amplifier, communications link, power, size and weight, and cost, were built as modular and reconfigurable blocks with standardized digital interfaces. The systems architecture model was connected to the analysis workflows so that requirements and analysis refer to the same definitions. Along the way, the interface between disciplines stopped being a conversation, which removed both the days of latency and the transcription errors that hand-carried results had introduced. Design A parameter change now propagates through the coupled models, and Verify requirement checks run inside the workflow rather than after it.
Close the Loop
The loop closes between Business and Engineering, and it closes through System Stewardship. Cost, weight, and power constraints are structured into the architecture model as parameters and requirements, and the results return as a priced set of options the business can choose between. The feedback here is entirely model-side. It shortens the distance between a commercial question and a quantitative answer.
Outcome
Evaluating a single configuration fell from several days to seconds or minutes, depending on the complexity of the individual models. Once an answer is that cheap, exploration replaces selection: a program can price the whole space instead of justifying one point in it, and the trade-off discussion with the customer becomes quantitative. The reported effects beyond speed are better technical communication, better design quality, and reduced design and execution risk.
References
Sources for this case differ in kind. The 2018 conference presentation and the 2017 SPIE paper are engineer-authored and cleared for public release. The 2021 case study is vendor-authored marketing material: its numbers are the vendor’s framing of a customer’s results, and the further integration it describes as future work is a plan. Where the two disagree, this text follows the primary presentation.
- John Hodge, “Multidisciplinary System Design Optimization Using Model-Based Engineering to Support Phased Array Antenna Architectural Trades,” Phoenix Integration International Users’ Conference, April 2018. Primary source, presented by the Northrop Grumman engineer who led the work and cleared for public release (#18-0280). Source of the days-to-seconds-or-minutes comparison, of the description of the legacy non-connected state, and of the stated objective of an integrated, modular, and reconfigurable cost-performance model. The original page now redirects to Ansys; use the Internet Archive copy.
- John Hodge, Kerron Duncan, Madeline Zimmerman, Rob Drupp, Mike Manno, Donald Barrett, Amelia Smith, “An integrated radar model solution for mission level performance and cost trades,” Proc. SPIE 10206, Disruptive Technologies in Sensors and Sensor Systems, 102060P, 2 May 2017. DOI 10.1117/12.2261754. The only peer-reviewed publication in this lineage. It documents the 2016 internal research and development precursor, in which RF, power, control, size, weight, thermal, and cost models for an AESA radar were connected through standard interfaces and linked to the Air Force’s AFSIM mission framework. It is the source for the claim that the interfaces between discipline models were standardized rather than merely scripted.
- “ModelCenter + Northrop Grumman,” Phoenix Integration / Ansys case study, 2021. Vendor case study. Source of the trade-study input and output lists, the Pareto front and best-value description, the “checking requirements, performing design trade-offs” quotation, and the claim of broader MBSE adoption across the company. It states a reduction from hours to minutes, where the 2018 presentation says days; it also lists a unified digital twin, CAD integration, and deeper PLM integration as future work, which should not be read as delivered capability. Download the Ansys Case Study.
- U.S. Department of Defense, “Digital Engineering Strategy,” June 2018. The policy document quoted at length in the vendor case study as the motivation for the program. It is context for why a defense supplier pursued this, not evidence about the antenna work itself.
- Northrop Grumman, “Ditching the Documents” and “Finding a Single Source of Truth with Model-Based Systems Engineering,” corporate digital transformation pages. Company-authored background on the wider digital engineering push. They contain no quantitative claims and are cited only to show that the approach was adopted beyond this one program.
English
German