Some Observations on the Practical Use of Modeling and Simulation Trends in Application of Mathematical Modeling and Simulation in Industry, Research and Innovation

Some Observations on the Practical Use of Modeling and Simulation

The advances in basic knowledge and model-based process engineering methodologies are resulting with an increasing demand for models. The observations given here are commentaries and considerations about some aspects of modeling with the focus on:

  • reliability of models and simulations,
  • role of the industry as final user of modeling and simulation research,
  • role of modeling and simulation in innovations,
  • role of modeling in technology transfer and knowledge management,
  • role of the universities in modeling and simulation development.

Reliability of Models and Simulations

Correctness, reliability and applicability of models are very important. For most engineering purposes, the models must have a broad range of applicability and they must be validated. If the models are not based on these principles, their range of applicability is usually very narrow, and they cannot be extrapolated. In many modeling and simulation applications in the process industry, kinetic data and thermodynamic property methods are the most likely sources of error. Errors often occur when and because the models are used outside the scope of their applicability. With the advent and availability of cheap computer power, process modeling has increased in sophistication, and has, at the same time, come within the reach of people who previously were deterred by complex mathematics and computer programming.

Simulators are usually made of a huge number of models, and the user has to choose the right ones for the desired purpose. Making correct calculations is not usually trivial and requires a certain amount of expertise, training, process engineering background and knowledge of sometimes very complex phenomena.

The problem with commercial simulators is that, since the simulations can be carried out fairly easily, choosing the wrong models can also be quite easy. Choosing a bad model can result in totally incorrect results. Moreover, with commercial simulators, there is no access to the source code and the user cannot be sure that the calculations are made correctly. The existing commercial flowsheeting packages are very comprehensive and efficient, but the possibility of misuse and misinterpretation of simulation results is high. In CFD and molecular modeling,
the results are often only qualitative. The methods can still be useful, since the results are applied to pre-screen the possible experiments, the synthesis routes and to visualize a particular phenomenon.

The Role of Industry as Final User of Modelling and Simulation

This role is not clear, except in the cases of big companies which have their own research and development divisions. In this case, the R&D company division has specialized teams for modeling and simulation implementation. The properly developed models and simulators are then frequently used, as we have already shown, during the life-cycle of all the particular processes or fabrications that give the company its profile. At the same time, each big company’s R&D division can be an important vendor of professional software.

The small companies that are highly specialized in modeling and simulation, operate as independent software creators and vendors for one or more company’s R&D division. The use of modeling and simulation in small and medium size manufacturing companies is quite limited. Since small manufacturing companies and university researchers do not cooperate much, awareness and knowledge about modern Computer Aided Process Engineering tools are also limited. There are of course exceptions among manufacturing companies. Some small and medium size engineering and consulting companies are active users of modeling and simulation tools, which allows them to better justify the solutions they propose to their clients.

Modeling and Simulation in Innovations

Modeling and simulation are usually regarded as support tools in innovative work. They allow fast and easy testing of innovations.

The use of simulators also builds a good basis for understanding complex phenomena and their interactions.

In addition, it also builds a good basis for innovative thinking. It is indeed quite important to understand what the simulators really do and what the limitations of the models are. As a consequence, access to source codes is the key to the innovative use of models and simulators. Many commercial programs are usually stuck in old thinking and well-established models, and then, the in-house-made simulators are quite often better innovative tools.

Molecular modeling can be used, for example, in screening potential drug molecules or synthesis methods in order to reduce their number. The existing molecular modeling technology is already so good that there are real benefits in using it. Molecular modeling can be a very efficient and invaluable innovative tool for the industry. The terms “artificial intelligence” and “expert systems” are based on existing knowledge. The computers are not creative, which means that these tools cannot be innovative. However, they can be used as tools in innovative development work. While most of the modeling and simulation methods are just tools, in innovative work, process synthesis can be regarded as an innovation generator, i.e. it can find novel solutions by itself.

Role of Modelling in Technology Transfer and Knowledge Management

Models are not only made for specific problem solving. They are also important as databases and knowledge management or technology transfer tools. For example, an in-house-made flowsheet simulator is typically a huge set of models containing the most important unit operation models, reactor models, physical property models, thermodynamics models and solver models from the literature as well as the models developed in the company over the years or even decades. Ideally, an inhouse-made simulator is a well-organized and well-documented historical database of models and data. A model is also a technology transfer tool through process development and process life cycle. The problem is that the models developed in earlier stages are no longer used in manufacturing. The people in charge of control write simple models for control purposes and the useful models from earlier stages are simply forgotten. Ideally, the models developed in earlier stages should be used and evaluated in manufacturing, and they should provide information to the research stage conceptual design stage and detailed design stage. One reason for “forgetting” the model during the process life cycle is that the simulators are not integrated. Different tools are used in each process life cycle stage. However, simulators with integrated steady-state simulation, dynamic simulation and control and operator-training tools are already being developed.

The problem is that the manufacturing people are not always willing to use the models, even though the advantages are clear and the models are made very easy to use.

Role of the Universities in Modelling and Simulation Development

The importance of modeling and simulation for industrial use is generally promoted, in each factory, by the youngest engineers. The importance of computer-aided tools to the factory level is best understood when the application of modeling and simulation has a history. The importance of modeling and simulation is not understood so well in the sectors that do not use computer-aided tools.

Technical universities have a key role in the education of engineers as well as in research and development. In fact, the universities’ education role is absolutely fundamental for the future development of the industry.

Indeed, in the future, the work of a process engineer will be more and more concerned with modeling and computation. Moreover, the work will be all the more demanding so that process engineers will need to have an enormous amount of knowledge not only of physics and chemistry, but also of numerical computation, modeling and programming.

Reference: T.G.Dobre, J.S.Marcano: Chemical Engineering: Modeling, Simulation and Similitude