What is a process simulation?
Simulation plays a fundamental role in our technical offices: in fact, it would now be unthinkable to design a component, or more generally a product, without the use of dedicated software, from CAD to FEM systems. By using these tools, it is possible to predict the mechanical behavior of components in an extremely accurate way, allowing the designer to understand what will happen in operation. Furthermore, for some types of simulation, considerable computational capacity is required, especially for multi-body and multi-physical simulations, i.e. interactions between different bodies or physical states in which our system operates and interacts (e.g. solid-liquid, phase transition, gas-liquid, etc.)
Similarly, process simulation consists in any form of simulation that aims to predict the behavior of a production system at different levels, on all of them:
- performance: how does the production system behave with reference to key metrics (eg OEE, hourly cost, production capacity)?
- logistics: how do my production flows interact? Where are the bottlenecks?
- multi-physics and multi-body: what are the process parameters to use to meet the product requirements? How do my manipulators interact with the artifact in transit?
- ergonomic: how does the design of the workstation impact the health of the operators?
Unlike what happens in product design, these tools are not yet widely used at the production process level, especially in SMEs, mainly due to the lack of knowledge of some aspects:
- modes: what software and what skills does their use require?
- benefits: why should I invest in these tools?
- costs: they require a certain initial investment (how much?), even if the offer on the market is now becoming more and more convenient.
Finally, process simulation plays a fundamental role in the implementation phase of a Digital Twin, that is, a digital model of a production system that behaves (in real time) like the real one.
Below we will mainly focus on one of the most powerful simulation tools when designing a factory, called Discrete Event Simulation.
The Discrete Event Simulation
In a discrete event simulation, the process is broken down or “discretized” into a sequence of blocks containing a certain number of information, such as:
- cycle time;
- set uo time;
- required manpower;
- required equipment;
- utilities (e.g. electricity, water)
- process variability;
- flow logic (FIFO, LIFO)
- Defect rate
Based on this information, with the simulation it is possible to evaluate the interaction of the various blocks, and consequently predict the key metrics of the process, for example:
- how many pieces can I produce per hour, per shift, per day, per month, per year?
- what is my hourly cost?
- use of factors, e.g. work, facilities, utilities;
- line balancing: how many buffers should I expect and where? Where can I optimize the process and where should I sub-process it?
To get results that reflect reality, it is important that the input data is “quality”. This means that:
- the more reliable the data, the better;
- the greater the amount of data available, the more the input values are statistically reliable;
- it is desirable to have an automatic data collection system, usually using MES (Manufacturing Execution Systems)
In some cases, especially when designing a new plant, it is difficult to have a history to refer to: it is therefore necessary to make initial estimates, and then refine the model by collecting real data once the system becomes operational. For this reason, it is important that all process experts who can provide reliable initial estimates are involved in this phase.
DES: what benefits?
According to Kokareva et al. (01), the main benefits associated with a discrete element simulation include:
- An increase in the productivity of existing plants by up to 20%
- A reduction of investments in the planning phase of new plants up to 30%
- A reduction in inventory and processing times of up to 40%
- Optimization of the size of the production system, including the size of the buffers
- Reduction of investment risk early on in concept
- Increase in the use of production resources
- Line design improvement
DES: what challenges?
Despite the obvious benefits, there are, however, some challenges in implementing a DES:
- The simulation cannot be separated from an in-depth knowledge of the product, of the processes and of the production system in general;
- The use of these tools requires specific skills, which must be acquired internally, or in outsourcing;
- As said, the quality of the output data depends crucially on the reliability of the input data. Without the adequate knowledge of the process and without an adequate initial data collection system, it is very difficult to obtain expendable results.
Where to start?
In this post, we focused on discrete event simulation, the first fundamental tool to design an optimized production system.
For greater levels of detail, it is possible to use other types of process simulation, such as kinematic, multi-physics, ergonomics simulations, to be used according to the specific aspects of the investigation. In fact, it is important to underline that the type of simulation must be functional to the requirements of cost, quality, admissible risk and associated implementation times. Consequently, the project must be managed considering every single aspect in a structured manner.
In this sense, Accialini Training & Consulting is able to provide adequate support in the following implementation phases:
- Identification of the type of simulation best suited to business requests;
- Choice of the best service able to meet customer expectations;
- Planning of activities, from data collection to system optimization based on production metrics;
- Support in the system implementation phase.
For more info, contact us to discuss your needs together in detail.
Moreover, subscribe to our newsletter to stay up to date on our services!
01- Kokareva V.V. et al, Production Processes Management by Simulation in Tecnomatix Plant Simulation, Applied Mechanics and Materials Vol 756 (2015) pp 604-609