Simulation: A Revolution in Design and Manufacturing – Part 2

As we discussed in part 1 of this short blog series on simulation, the evolution of modelling and simulation provides a bridge between design concepts and product manufacture, with the added benefit of enabling engineering teams to virtually run multiple scenarios to ensure optimized component manufacture. But how does this support the assembly process and what additional benefits will new developments, including the digital twin and AI, bring? In this blog, we will delve into these topics in more detail.

Simulation optimizes assembly processes

Kinematic and motion simulation, or multi-body dynamics (MBD), is a critical area of advancement in machine simulation technology. MBD looks at the way a machine operates, such as how the cam moves, the size and power of the motor needed to turn the cam, and the motion and trajectory of the moving parts within the machine. In practice, this might include analyzing the movement of the cam in a machine that stamps out parts or brings two parts together as a sub-assembly, for example. The analysis would allow the motor to be optimized, so it provides the right level of power and torque. Optimizing the size of an electric motor has many implications, stress on the tool can be minimized, but it can also reduce the cost of operation by reducing the amount of power expended in the process.

The granularity with which MBD simulation software can now operate is astounding. A machine can be modeled to show how the forces applied in individual areas of the machine cause wear or fatigue and how quickly that happens under various operating conditions. This predictive insight means measures can be taken at the earliest stage to prevent failure or extend operating lifetimes.

Software can simulate and analyze the manufacturing process of individual components, as well as how machines operate to manipulate those components during an automated assembly process. The next exciting step in this digital transformation is to simulate everything, including human operators, to see where opportunities for further optimization lie. Discrete event simulation (DES) is such a kind of simulation tool. DES models and predicts behaviors of processes with well-defined activities, or events. It focuses on the end product’s manufacturing process, namely the machines and assembly tools used to automate the production phase, the way operators interact with automation, how machines interact with one another and how the entire flow can be optimized to increase efficiency and productivity.

For example, simulation provides perspective on the way these elements interact at speed and determines the relative throughput of the process. Given that a chain is only as strong as its weakest link, it stands to reason that the line can only move as fast as the slowest process. Simulation enables each part of the process to be analyzed to find the slowest point and then optimized to increase the throughput.

This is particularly beneficial if the line includes machines that are designed and built specifically for automated production lines. Designing machines that make products is increasingly common because it results in productivity gains which ultimately cuts unit price for the primary customer and the customer’s customer down.

Digital twins, AI, and the future of simulation

Simulation software is now being used by companies like Molex to model systems from individual machines to production lines, to the entire factory. Digital transformation and IoT technologies make data exchanging possible between real-world processes and simulation models.

This is the concept known as the digital twin and it creates an exact, virtual replica of the aspects of influence on a given process, including putting humans in the loop.

A digital twin brings additional benefits in multiple ways. Firstly, the digital twin can model the real world, but it can also mimic it, by taking real-world data from the production line and applying it in the virtual world. For example, the data gathered after 10,000 cycles could be applied to a digital twin and then accelerated through extrapolation. This would allow the operators to predict how the production line performs after 20,000 cycles, or when it is likely to need maintenance. Predictive maintenance is a significant benefit of digital twin technology.

Digital twins also predict the effects that changes may cause. Modeling the real world using variable parameters means that if those models show a positive benefit, those parameters can be directly transferred to the real world, with a high level of confidence in the effects of making those parameter changes. Running a digital twin alongside an actual production line will become increasingly common in the future, and it all starts with simulation at the lowest level.

Today, more and more companies are developing manufacturing AI applications. Relying on real-world manufacturing data to train AI models can be expensive and time-consuming.  As its virtual replica, a digital twin can provide a powerful and realistic virtual environment that enables risk-free training and testing of these AI models.

Molex is actively advancing its simulation and modeling capabilities, with a growing portfolio of tools and expertise. We are also building a database to record how those materials behave so that vital foundational parameters can be used in simulation.

Ultimately, simulation drives productivity at every stage of the manufacturing process and cuts waste, time, materials, and cost from the design and manufacturing process. There’s no amount of simulation needed to understand that this benefits everyone!

Predictive Manufacturing Manager