In the face of fierce global competition, all manufacturers are searching for ways to reduce cost, optimize designs, and deliver them quickly to market. Companies able to achieve these objectives hold a competitive advantage in the marketplace. However, it is not an easy task optimizing engineering design, especially for systems that involve multiphysics. So, what exactly is multiphysics, and why are systems that involve multiphysics challenging to develop? As the name implies, multiphysics involves evaluating the behaviour of a system that is affected by several physical phenomena. The interaction between these phenomena leads to a completely different behaviour than the one resulting if each physical phenomenon would be treated separately from the others. Let us take an example of the aerospace sector to highlight what multiphysics means. Modern aircrafts, especially the commercial ones, run on gas turbine engines. These engines work by a combustion process between fuel and air inside the engine. Air is forced inside the combustion chamber, mixed with fuel and then ignited. This burning of fuel generates thrust for propulsion. Gases that form expand rapidly and are exhausted through the rear of the combustion chambers. In this process, the gas turbine temperature can reach 1000°C. More when there is a need for additional thrust required at the time of take-off. At an altitude of 10 km or higher - which is how high typical commercial airliners fly - the outside temperature of an aero plane can reach between -50°C to -80°C. And the inside temperature of the engines can be 1000°C or more. Add to this air flow, drag and other forces that act on the turbine. All these present a design challenge to engineers. This is but one example of systems involving multiphysics. Automobiles, oil and chemical processing refineries and several other systems involve multi physics analysis at one stage or another. Even the development of electric motors, say that of an electric vehicle (EV), involve detailed consideration of various physical phenomenon like rotor speed, coil material, armature heat dissipation, battery power and storage, etc. Itis virtually impossible to evaluate a solution manually for phenomenon that involves multiphysics. This is where modern day computers and simulation software are useful.
Multiphysics Simulation
The advancements in modern digital computers have brought CAD (computer-aided design) and CAE (computer-aided engineering) to the aid of engineers. Simulation, which is predicting the behaviour of a system using virtual prototyping, has reduced the time to design and market and saving cost by eliminating the need to create a physical prototype at the design stage. The design of a simulation model can help the engineerstremendouslyin building confidence on validating the required technical specifications and in making critical design decisions. Simulation considers inter-dependencies and design parameter variations, and collaboratively allows examining strategic choices for optimization and robustness. Today, it is unimaginable to think of product development – especially complicated ones that involve multi physics – without the help of simulation.
Multiphysics simulation allows companies to conduct various simulations and analysis simultaneously. The biggest advantage for the companies is that, multiphysics simulation solutions offer companies a single platform to run different tests to evaluate the overall product performance, thus helping companies save, time and costs of development while ensuring excellent product performance efficiency.
At the heart of simulation is numerical analysis. Numerical analysis is the branch of Math that searches for effective ways of designing in order to find optimal solutions to complex mathematical problems. There are various methods used in numerical analysis, some of which include Monte Carlo simulation, Euler method, Runge-Kutta method and many others. Each method has advantages and disadvantages, and selecting the best possible method is not straightforward. Engineers need to understand the differences among the calculation methods. Choosing a particular method can result in trade-off among model complexity, accuracy, and computing time. Engineers use a combination of these calculation techniques as the optimal solution to simulate a system.
Irrespective of the method chosen, all numerical analysis method involves solving differential equations. A differential equation is an equation that contains at least one derivative of an unknown function, either an ordinary orpartial. Ordinary differential equations involve one or more ordinary derivatives but do not have partial derivatives. Partial differential equations have two or more independent variables.While first degree differential equations are easy to solve, systems that require multiphysics invariably involve higher degrees (a degree denotes the degree of the highest derivative in the equation) and are difficult to solve.The construction of a differential equation model requires a thorough understanding and description of the process involved. And yes, merely setting up a differential equation model is not enough, it is essential to solve the equation. This is what simulation software does.
Simulation Software
The formulation of a multiphysics problem is highly dependent on the nature of the problem and how the component physics fields are coupled in space and time. In some circumstances, coupling between continuum and discrete models is required to solve a problem. Whenever there is a need to model a system where the state variables vary with time and / or space, differential equations are natural tools for describing its behaviour. However, setting up and solving of differential equations of higher degree manually is extremely tedious. The process of finding useful solutions of a differential equation is much a symbiosis of modelling and mathematics. Choosing a method - whether analytical or numerical – is more complex for multi physics systems. Multi physics problems are mostly non-linear, and most solution methods require solving a series of linear sub problems. To solve a differential equation problem for a multi physics product or system, it is necessary to know facts about its modelling background, its mathematical properties, and its numerical treatment. The last part involves choosing appropriate numerical methods, and this is where simulation software comes into play. The best in class simulation software (like Altair SimLab) provide process oriented multidisciplinary simulation environment to accurately analyze the performance of complex assemblies. As mentioned above, multi physics systems include structural, thermal and fluid dynamics. Good multi physics software can be easily setup using highly automated modelling tasks, helping users to drastically reduce the time spent creating finite element models and interpreting results.
And of course, multi physics simulation software does not only produce accurate results, it displays them in an intuitive and self-explanatory graphical way as well. This allows design engineers to visualize the modelling process in a better way. In short, multiphysics simulation technology enables users to design, analyze, and deliver efficient, optimized designs, and good simulation software goes a long way in providing it.