|

Simulation and Modeling in Product Development and Innovation

Modeling and simulation is the exploration of the complexities of the physical world, virtually. Often used together with high performance computing (HPC) techniques, these technologies are vital tools in engineering, science, and product development. 

Simulation and modeling software plays a major role across many industries to solve complex problems. Many concepts benefit from mathematical exploration, rather than visual or intuitive approaches. 

From 2023 to 2030, the global simulation software market is expected to expand at a compounded annual rate of 13.6%, according to market analysis. Today, simulation and modeling powered by cloud HPC and supercharged by AI/ML delivers more advanced capabilities with greater efficiency for engineers, scientists, and researchers.

The Foundations of Modeling and Simulation

Simulation and modeling, also known as M&S or MODSIM, have different meanings depending on the context. Sometimes, the terms are used interchangeably. However, there is a difference between them. Modeling is a representation of a system that helps with decision-making or predictions. On the other hand, simulation imitates a real experience, thing, or process using modeling. 

Simulations drive innovation in many sectors, from aerospace and automotive to earth and life sciences. In semiconductors and electronic design automation (EDA), simulations verify designs and guide new chip layouts. In manufacturing, simulations help with materials testing and factory optimization. Model-based engineering also supports various design and operating decisions for process development and optimization. 

Visualization of code-based and numerical models is an important part of simulation and modeling, from model building and simulation preparation to testing and production simulation. For example, popular tools like Paraview and NiceDCV visualize (render) complex simulation outputs like structural mechanics, fluid dynamics, and architectural modeling. 

In practical application, computer-aided engineering (CAE) uses simulations to advance product development and allows for testing very early in the process. This reduces the need for actual representations or prototypes, making the journey from concept to design to production quicker and more effective than ever before. 

In the past, CAE was run using on-premises supercomputers, only affordable for well-established companies. Now, cloud HPC has democratized access to simulation. HPC clusters allow for even more advanced simulations with real-time collaboration across distributed teams. They consist of hardware (semiconductors, servers, storage, dedicated network) and software (applications, filing systems, compilers, libraries, debuggers). 

Design engineers rely on various M&S concepts to accelerate product development, troubleshoot designs, and drive quality. There are four common types: discrete events, continuous simulations, agent-based modeling, and system dynamics modeling.

Discrete Event Simulation

Stress test simulation of a compliant mechanism designed for efficiency and durability.
Stress test simulation of a compliant mechanism designed for efficiency and durability.

Discrete event simulation (DES) models a real-world situation in a digital environment, showing a system’s operations as a specific sequence of events. The model is often referred to as a digital twin.  

In this type of simulation and modeling analysis, there is no change in the system design between consecutive events. Rather, a discrete event simulation leaps to the moment of the next event, which is often referred to as “next event time progression.” An alternative is “incremental time progression.” Here the simulation updates events occurring during specific time increments, each requiring analysis. 

There are many kinds of related simulations, including fluid dynamics, thermodynamics, particle fluidics, and electromagnetics. Also, the discrete element method is often used to compute the motion and effect of a copious amount of small particles. The finite element model analyzes the behavior of products under varying loads and boundary conditions. 

DES helps users reach quick, trustworthy decisions in the early stages of product development, making it ideal for logistics, healthcare, and manufacturing industries. It reduces lead times, optimizes layouts, and reduces work in progress.

Continuous Simulation

Weather forecast simulation using WRF model for numerical prediction
Weather forecast simulation using WRF model for numerical prediction.

In a continuous simulation, differential equations set the rate of change so that a system’s state can change constantly over time. With no queue of events, essential components are completed without delay. 

Continuous simulation has numerous applications across various sectors. For example, civil engineers use this simulation type for dam embankment and tunnel construction, while a defense department might use it for missile trajectory analysis. In logistics, continuous simulations analyze passenger flows at airport terminals. 

Even in product development, continuous simulations are essential. They are used to develop everything from electronic circuits to robotics and vehicle suspensions to hydraulics.

Agent-Based Modeling

Agent-based modeling (ABM) simulates the interactions of autonomous agents, whether individual or collective. The goal is to better understand system behavior and what influences it. 

ABM takes advantage of the tools used in game theory and computational sociology while also borrowing from evolutionary programming and multi-agent systems. It’s commonly deployed in scientific disciplines, such as ecology, biology, and the social sciences.

Since the early 21st century, ABM has also been used to develop and validate autonomous driving (AD) systems. For example, Waymo created Carcraft to test algorithms for self-driving vehicles by simulating interactions between the driver, vehicle, and pedestrian. The artificial agents in this example emulate human behavior.

System Dynamics Modeling

System dynamics modeling attempts to understand and predict interactions between a system’s components and their evolution over time. It accounts for the extremely nonlinear nature of complex model based systems engineering set ups, using causal feedback loops, stocks, flows, table functions, and time delays to enhance and inform decision-making. 

Since system dynamics modeling analyzes the impact of alternative policies, it has a wide range of applications. Professionals can use it to study managerial systems, macroeconomics, ecological systems, and population development.

In product development, system dynamics can help investigate resource dependencies. Engineers can also use stock and flow logic in a system dynamics model to evaluate the behavior between various design elements.

Simulation and Modeling: A Recap

Product developers have an array of modeling and simulation tools at their disposal, meaning they can choose between different types of simulations: discrete event, continuous, agent-based, and system dynamics modeling. Yet a combination of these tools might work better since certain real-world cases can be too complex for any single method. 

With a well-conceived cloud HPC network, engineers have everything they need to run multiple intensive simulations, including a choice of on-demand compute power. There is a long history of computer-aided engineering (CAE) using on-premises infrastructure that only large companies could afford. Now, the cloud has democratized access to CAE simulations, expanding access to smaller businesses and startups.

The Future of Modeling and Simulation: The Era of AI

Lid-driven cavity fluid flow simulation used to test the efficiency of physics-informed neural networks
Lid-driven cavity fluid flow simulation used to test the efficiency of physics-informed neural networks.

Modeling and simulation is now headed to a future radically enhanced by artificial intelligence (AI) and machine learning (ML). This is the next phase in the evolution of computational research and development. 

Consider the growing trend to use deep learning surrogates (DLS). When enhanced by the power of AI, DLS automates design optimization to create more precise digital twins. 

Such advances, however, demand even more computing power. In essence, computing power drives M&S sophistication. In turn, advances in modeling and simulation drive the need for more computing power to further innovation efforts.

Learn More 

With Rescale, enjoy instant access to unlimited HPC resources to drive your innovation efforts. You can supercharge product development and digital engineering efforts with the on-demand HPC resources available on the Rescale Platform. And most any R&D application is available instantly through the Rescale Software Catalog.

Author

  • Garrett VanLee

    Garrett VanLee leads Product Marketing at Rescale where he works closely with customers on the cutting edge of innovation across industries. He enjoys sharing customer success stories, research breakthrouths, and best-practices from Rescale engineers, scientists, and IT professionals to help other organizations. Garrett is currently focused on the convergence of supercomputing, HPC, and AI simulation models and how these trends are driving discoveries in science and industry.

Similar Posts