TWI Knowledge Summary

Manufacturing - process simulation

Process simulation involves the modelling of physical processes with the aim of studying their effects on the external environment and the objects they are applied to. These processes usually involve the interaction between two or more systems.

Computer software models are good at examining these interactions that are often too complex for people to comprehend or predict unless real life experimentation is used. Unfortunately real life experiments covering a wide range of situations can take a long time, are costly, and can not often be repeated. Process modelling minimises these issues.

There are many examples of processes being simulated in both manufacturing and everyday life, e.g.

Weather forecasting is used to understand the interactions between weather systems and the land to predict future weather patterns.
Products are often modelled during their service life to foresee the detrimental effects of thermal cycling, load cycling, impacts, and weather. These results are then used to improve the design of the product.
Manufacturing processes such as casting and welding are simulated to understand how they effect the chemical composition, shape, residual stresses and material properties of the products being made. These results are then used to reduce the variability in the manufacturing process.

Current status

The modelling of complex interactions at a physical level usually requires many fundamental calculations to be performed. Consequently these 'scenarios' often need to be modelled in significant detail, and validated against previous experience through real life experimentation. Computers that are able to perform such calculations quickly are both scarce and expensive.

Benefits and risk

Process simulation enables the user to predict, when and by how much a system's properties will change. This capability brings many benefits such as increased product quality, product safety, accident prevention, and improved product design. In short, an improved understanding of the product is gained.

Alternative methods for making such predictions sometimes do not exist or require costly and long real-life experimentation. The risk in building reliable models depends upon the availability of data against which to validate the model.

Areas of development

Companies continue to develop advanced software for increasingly less expense. At the same time, extensive and widespread research is being performed to develop new applications. All applications require not only a deep understanding of the physics, but data on material properties under different environment conditions (e.g. temperature). This is often not available or incomplete. Work is being carried out to determine values for these properties to make models more accurate.

For manufacturing industries, computational efficiency and model building time are critical issues, as they determine the number of possible design iterations and experiments that can be achieved. Developments are occurring in hardware and software to accelerate the speed at which these activities can be carried out.

Further information

If you have a query on this subject please email TWI's Manufacturing Support Department. manufacturing.support@twi.co.uk