Physical system simulation can make a substantial impact in reducing design and operation costs whileaccelerating the time to market for new products. Physical system simulation is an elusive concept –yet on your journey to virtualize the product design process, there is no way around it. In the first blog of the Success with Simulationseries,PieterDermont, Modelon’s Senior Business Development Director, covers what physical system simulation is, what it is not(hint, it is not“systems engineering”), and what it can do for your organization.
Physical System Simulation Offers the Right Fidelity toMake Decisions on the System Level
In the world of simulation, one always deals with atrade-off between the detailin the modelandthecomputational time. A model is a description of the physical world in more or less detail, and the more detail that needs to be simulated, the morecomputationally expensive the task will be. Computationally expensive modelswill take long to simulate,andwhile that may be necessary in some cases, it’s oftena better approach to include theright amount of detail forachievefaster simulation results. Faster simulation results mean that anengineer can update the design (and subsequently the model)more often, thenexperiment and iterate faster throughoutthe product design cycle.
For instance, some simulation methods allow you to capture a high level of detail, while being relatively computationally expensive. They are often used for component-level simulation in 2D or 3D and are often limited to interactions of one or two physical domains. Computational Fluid Dynamicsand Finite Element Analysis are methods that typically fall within this category. Computational Fluid Dynamics isfocused on fluid flow only, possibly with interactions of the chemical or thermal physical domain.Though, for simulating systems these methods are computing cost prohibitive. Instead, for system simulation, the focus is predominantly on the 0D or 1D effects.
For the simulation of a system, and to achieve meaningful and timely engineering insights on a system level,the question then becomes:what details or physical phenomena should be included in the model?
Physical system simulation is focused on representing a system, and thus sacrifices component detail in exchange for faster performance and the ability to capture the complete system.This does not mean that system simulation is less accurate or less predictive -all paradigms require their due diligence in terms of parametrization, calibration, and validation. Rather it means that system simulation is focused on capturing the physicalphenomena that are relevant on a system level.
For example,inFigure 1, theliquified air energy storage system model includes the right level of detail relevant to a whole system analysis. This model can track the thermodynamic state of the fluids as they are changed by each cycle component, according to the transformation the fluid undergoes in eachcomponent. As such, decisions on the system level can be made.
Predictive CapabilitiesUsingPhysical System Simulation
Physical system simulation offers features to accurately describe physics. When an engineer is required to find the right description of her/his physical system,it is crucial to useasimulationparadigm that represents the’physics’conveniently. This physical description is possible thanks to two key capabilities of physical system modeling:
- First-principleequationsare the fundamental laws that govern our physical world, such as the conservation of mass, energy,and momentum. These laws also govern the behavior of physical system models.
- Empirical relationshipsdescribephenomena such as pressure losses, heat transfer,and frictionthat cannot be conveniently modeled using first-principles at the system level. These empirical relationships are determinedby experimental research and often parameter fitting practices.
The combination of these two gives physical system simulation its predictive capabilities, and a leg up compared to purely data-based approaches for simulating physical systems. This foundation permitsbuilding models that accurately represent avariety of physical domains, such mechanical, fluid, heat transfer, electrical, chemical, and/or magnetic physical phenomena. Andthuscapturesthe interactionsof thedomains –the creation ofmulti-physics system models.
A good physical system simulationtoolwill bring features thatare relevant to simulatingphysical systems, as opposed to any system. Such features include the ability to work with units for physical quantities, group quantitiesand models, reuse first-principles, andselect the relevant empirical relationshipsconveniently. Complexity grows rapidly and exponentially, and those features become paramount to the handling of the model. To contrast this,usingwell-established general-purposeor scientific scriptinglanguages to representaphysical systemrapidly adds unnecessary complexity to the description of the model. Complexity is detrimental to the speed and the quality of the product development cycle.Adesign engineercan be prone to human error, whereas the correct physical system simulationtoolcan help combat thisand speedupproduct design iterations.
The model in Figure 2 shows physical system simulationthat’sfocused on the predictive physical phenomena of an aircraft on the system level, covering the variousphysical domains relevant to each aircraft subsystem. The behavior of each subsystem is governed by first-principle equations and empirical relationships. Modelon library content offersa collection of numerically robust, performant and re-usable first principle-equations and empirical relationships, packaged in documented component models. Not having to describe the physics yourself represents a significant time and cost gain in your model-based design journey.
Figure 2:This model represents an aircraft’s true multi-physics system (from left to right): the airframe (mechanics), the power plant (traditionally, a mechanical-thermo-fluid system), and the auxiliary systems (fluids, electrical, etc.).
A technology geared towards physical system simulationbringsthe appropriate level of fidelity to make meaningful engineering decisions on systemsand speaks the language of physics which helps managemodel complexity. Withinthe same model of a system, mechanical, fluid, heat transfer, electrical, chemical, and/ormagnetic physical phenomena can be described effortlessly.
Physical System Simulation vs. Systems Engineering
Physical system simulation can be contrasted with systems engineering. Systems engineering isa paradigm that is primarily concerned with how different sub-systems and components interact without necessarily having any predictive capabilities of the behavior of each component or subsystem, and thus focuses on the requirementsfor successful interactions.
While there are no strict guidelines on the use of any simulation paradigm over the other for a given simulation scenario, and there are many corner cases; each paradigm extends over a range that trades off complexity and the engineer’s mission. There is no inherent preference for a given amount of detail in a model, rather there is the desire to pick the right modeling paradigm and model complexity for each task. Physical system simulation is paramount in making design decisions early in the design process, to support with architecture and optimization; and later to validate verification tasks. Physical system simulation is also an excellent candidate for operational improvements for assets and to develop digital twins. Keep an eye out for more physical system simulation content includingwhen to use physical system simulation, where it fits into your design process, and more. CTA –Contact our experts toseephysical system simulation inaction!
Contact our experts to see physical system simulation in action!