Byron ford | cio review | 9 september 2015
The development of aero-systems is changing. In order to meet constantly evolving regulatory, operator and user requirements, modern systems are becoming increasingly complex--and with this complexity comes increased risk to these very high-capital programs.
So, what new processes, technologies and engineering techniques can manufacturers employ to more effectively manage some of this risk early in the lifecycle? Modelling!
The use of advanced modelling and simulation technologies can dramatically reduce the cost and increase efficiency of the development and test phases of these programs--enabling engineers to test and refine complex systems and their interactions synthetically, repeatedly and inexpensively—in some cases before ever machining a component. There are a number of key ways these technologies can be used to the greatest affect.
The challenge for aero-system manufacturers is that these complex systems (a commercial aircraft, spacecraft), and systems-of-systems (network-enabled aircraft), have increasing levels of interaction between their sub-systems/systems, their environment, and as a result can display unanticipated emergent properties of this complex series of interactions. These emergent properties are sometimes designed, but fail to effectively characterize and then mitigate the effects of undesirable outcomes. This can lead to unexpected performance characteristics, requiring significant rework to address—resulting in delays and cost overruns. In addition, not being able to plan and account for this introduces risk (EADS A400M and its delays are a good case in point1).
According to contemporary ‘Systems Engineering’ theory (the benchmark process for developing complex systems), the mitigation of uncertainty earlier in a program will reduce downstream risk by reducing the chance of a system failing to meet requirements, which would potentially lead to expensive rework or in the worst case, program cancellation. Modelling and Simulation can enable this by exposing ‘synthetic’ systems to real-world use-cases (both positive and negative) to test their reaction/emergent properties, in a safe, controlled and highly repeatable manner.
It is the repetitive (and often cheap) characteristic of the synthetic environment that can be very effective; giving the system developer the ability to iterate system changes and expose them to the varying effectors can very quickly help engineers to discern the impact of design changes, environmental changes or indeed sub-system, system failures or unexpected behaviors.
Below are two indicative use-cases for advanced modelling and simulation techniques, highlighting how they were used in different circumstances during the development of complex aero-systems;
A multi-national defence contractor – reducing the need for real-world trials of Unmanned Air Systems (UAS)
The testing of military aircraft, particularly experimental unmanned systems is very expensive; every test cycle requires many hundreds of man-hours to plan and execute, facilities, fuel and support. For these organizations, reducing the need for these trials (even a single one) later in the program by planning for risk was hugely valuable. To that end, a program of ‘synthetic environment’ trials was started much earlier in the program. This approach allowed very detailed monitoring, data collection and analysis of ongoing synthetic trials and they were highly repeatable, risk-free and low cost (everything from flight performance, airframe loading, to communications and radar performance could be tested). This allows the effects of design changes to be assessed individually and as part of a series of complementary changes, ensuring there are no undesirable system behaviours – and reducing the need for some testing later in the project.
A commercial orbital launch-capability provider – mitigating the risk of failed launch trials
The cost of undesirable behaviors in spacecraft is an obvious one, and these systems are some of the most overly designed and scrutinized systems out there. Another characteristic is that real-world testing is immensely expensive (up to hundreds of millions of dollars per launch), particularly if it goes wrong or fails to achieve the necessary objectives. These systems are also highly affected by their environment, which is often unpredictable. The technique of ‘synthetic trial evaluation’ is a powerful tool for mocking-up trials in high-definition simulated environments. This allows the virtualized system-under-test to be exposed to the test environment and conditions, and rehearse the test process, all in a safe and controlled manner. The system was assessed to ensure it performed as expected, and that the measured variables of interest (variables to be instrumented during the real test) are also as expected. This gives greater confidence that when the system-under-test is placed in the real-world scenario, under actual test conditions, it will not fail and the data collected will be useful and as expected, reducing the risk of having to repeat failed or unsatisfactory trials.
It is easy to see how the two different techniques that were applied above offer significant risk reduction to a program – tackling system complexities as early in the lifecycle as possible to reduce the chance of unexpected or undesired performance being encountered when it becomes very costly to address. The cost of an Unmanned Air System (UAS) trial, for example, can be significant and if a single real-world trial can be replaced with a synthetic substitute, or a single failed trial averted, this can often pay for the Modelling and Simulation tools to facilitate this mode of testing.
There are a number of proven Commercial off The Shelf (COTS) tools available in the marketplace to cover the needs of the above use-cases, most of them easily integrated with other commonly used tools and even Hardware-in-the-Loop (HIL) test systems so as component systems are designed and realized, they can be brought together with the synthetic systems for ever evolving and representative testing. Modern computing technology also brings these techniques within reach of both multi-nationals and small manufacturers.
The use of modelling and simulation also extends well beyond the design-proving of system; it can also be effectively applied to capability improvement, sustainment, operational analysis, training and even support real-time operations.
Byron Ford is an aerospace expert at PA Consulting Group