When people talk about simulations today, they are usually computer simulations. Here, processes or objects are simulated by the computer and reality is reconstructed virtually. By abstraction, a model is created, on which experiments are carried out. In such digital experiments, individual parameters, possible effects and effects under different influences can be tested virtually. In this way, we can also look back into the past or reach unreachable areas.
But it is not possible to do entirely without experiments, because sometimes simulations provide several different explanations for one phenomenon. Thanks to simulations, however, experiments are much more targeted in these cases.
Computer simulations are the third pillar of science, alongside experiments and theory. They help scientists to recognize connections, uncover gaps in knowledge and understand processes.
Simulations are an important tool for many disciplines. Whether it is meteorology, medicine, mechanical engineering, physics and chemistry, forestry and agriculture, sports science, environmental technology, engineering science or materials research – progress today is hardly conceivable without simulations. However, simulations do not only play an important role in science and technology, but are also becoming increasingly important in business, politics and the social sciences.
Simulations are used when…
… a system is very large or very small.
Due to the size of the universe, experiments can often only be carried out at great expense. Simulations are also often used on a small scale – in the field of atoms and molecules.
…experiments are too expensive.
Real tests, for example on the behaviour of machines, are very expensive and time-consuming. Crash test simulations save resources and a lot of effort, because in reality fewer cars have to be destroyed.
…experiments are too dangerous.
Some experiments are far too dangerous in reality. These include, for example, experiments on core meltdown or on processes in our environment such as the storage of CO2 in the soil, so-called CO2 sequestration.
…experiments are ethically unacceptable.
In some areas, experiments are prohibited for ethical reasons, for example in medicine or because they endanger others. This is why pilots in their training first train virtually on flight simulators before they actually fly.
… the system does not yet exist in reality.
Simulations also provide predictions about systems that do not yet exist. For example, the properties of new materials can be simulated before they are produced in the laboratory.
… processes run very fast or slow.
For example, when galaxies form or die, it takes many millions of years. The explosion of a supernova, on the other hand, happen very quickly. In computer simulation, you can virtually adjust the time.
In which application areas do we use simulations?
Driving and flight simulator | weather forecast | climate simulations | crash tests | simulation of production plants | spread of drugs in the body | spread of tumors in the body | simulation of geological processes | product development | automobile development | big bang | explosion behaviour | development of materials | election forecasts | risk analyses | company analyses | population trends | effects of money market instruments | forecasts in the insurance industry | simulation of political reform plans | simulation of global capital and commodity flows in stock trading | simulation of historical events | development of prostheses | …
Which disciplines use simulations?
Bioengineering | Food chemistry | chemistry | genetics | disease research | nanotechnology | diagnostics | biology | biotechnology | sports science | pharmaceutics | medical technology | machine learning | AI | mechanical engineering | mechatronics | electrical engineering | space technology | media technology | construction | civil engineering | architecture | automotive engineering | history | linguistics | physics | mathematics | computer science | astronomy | landscape ecology | geodesy | climate research | geology | sustainability studies | renewable energies | hydrology | water management | process engineering | …
And this is how it works!
Explaining simulations using the example of the human gait
The human gait is a complex process. This becomes particularly clear when a human being loses the ability to walk. In order for this ability to be restored, for example through highly developed prostheses, researchers need to gain a better understanding of how the gait works. This is where simulations help.
Determine the structures and influences involved. Formulation of interactions. Creation of a mathematical model.
Inserting concrete values into the formulas of the mathematical model. Translating the model into algorithms.
Developing a software for the simulation in order to calculate it on the computer.
Visualize abstract simulation results and data
Evaluation of the simulation results and comparison with the real experiment
Do the simulation results match the reality?
Incorporation of the findings into the process and improvement of the computer simulation
This video shows the results of a simulation which was processed with the help of data visualization. With the help of simulations, different situations can be simulated, which cannot be repeated or varied in the laboratory due to limitations on time and costs. This simulation shows how liquid flows through the porous medium. It moves in a streamlined way around the grey solid squares and flows from left to right.
The experts in the area of simulations
Who are we?
The Cluster of Excellence „Data-Integrated Simulation Science” at the University of Stuttgart has been in existence since 1st January 2019. It is funded by the German Research Foundation (DFG) within the framework of the German excellence strategy until 2026. In our Cluster, which is by the way one of two Clusters of Excellence at the University of Stuttgart, around 150 people from seven faculties of the University work together in interdisciplinary research teams. At the same time, we also take care of the next generation of young scientists within our study program Simulation Technology, our competition PlaNeT SimTech and our graduate school GS SimTech.
What do we do?
Our work focuses on the integration of simulation and data science. This means that we use the ever-increasing amount of data from various sources for our simulations from the very beginning in order to improve their accuracy. This enables us to make our models more reliable and thus make predictions that are more accurate. Scientific fundamentals are always taken into account. The knowledge gained in this way can be used in almost all areas: simulation of the musculoskeletal system, development of prostheses, development of materials, effects of geothermal drilling, development of patient-specific drugs – to name only a few.
For this purpose, the exchange and cooperation with other national and international research institutions and industry are very important.
We conduct research in the field of simulation and data science. It is based on three so called “Visionaly Examples”: Engineered Geosystems, Digital Human Model and Virtual Materials Design. In order to turn these visions into reality, it is necessary to develop basic methods that can be applied to all areas and are thus are applicable across visions.
Our research is organized in interdisciplinary project networks.
Study Program „Simulation Technology“
The interdisciplinary study program „Simulation Technology“ combines the fields of mathematics, engineering, computer science and natural science. Since 2010, students have been able to study the six-semester Bachelor’s course and since 2013 they have been able to continue their studies in the consecutive Master’s course. An excellent basis for a subsequent doctorate. Perhaps even in the GS SimTech, our graduate school.
Student competition Planet Simtech
Since 2015, the SimTech Junior Academy has been organising the PlaNeT SimTech school competition. Solving problems from natural sciences and technology is the motto here. The competition is for those students who are in their last two years of high school and enjoy mathematics, puzzles, engineering problems and tinkering. The team that finds the best solution for our modelling challenging task, such as “How much fuel do you need for a manned Mars mission?” can win up to 500 EUR.
1: Root water uptake by young, growing root systems
This simulation shows the growing root systems of two young plants. The plants take up water from the soil in a plant pot. The soil eventually dries out.
Model: soil and root
Simulation: water transport and root growth
Roots are visualized by brown tube segments. Blue soil corresponds to high water content, brown soil means low water content.
2: Flow through rock with fractures
This simulation shows the pressure-induced flow of oil through a cylindrical rock sample, in which elliptical fractures are contained. The pressure changes caused by the injection lead to the dilation of the fractures.
Model: porous rock and fractures
Simulation: flow and deformation
The rock is depicted in gray, and the blue region shows until which point the oil has been transported through the sample. The arrows show the velocities of oil within the sample and the deformation of the sample is strongly exaggerated in this visualization.
3: Contrast agent perfusion in capillaries and embedding tissue
This simulation shows contrast agent transport in the smallest blood vessels (microcirculation). Blood is simplified as a fluid without particles. The cell content is accounted for by an increased flow resistance. The porous tissue surrounding the blood vessels is described by an average description. This means that individual pores between cells are not visible.
Model: brain tissue and capillaries
Simulation: blood flow and contrast agent transport
The capillary network is taken from a rat brain and represents all blood vessels contained in a 1mm by 1mm by 2mm tissue sample. The contrast agent is visualized as a black cloud. Contrast agent leakage is restricted to the middle of the domain. Leaked contrast agent has orange color.
About a third of the earth’s surface is covered by land. More than a third of the land surface is agricultural area (pastures, acreage) and one third is covered by forests. Plants and crops substantially influence the nutrient balance in near-surface soil layer (in the ‘vadose zone’). They play a crucial role for the local water budget and water exchange between soil and atmosphere. Water transpires—particularly during the day with the help of the sun—from small vents in the plant’s leaves. The transpiration causes a suction effect which drives water into the roots and upwards through a vascular system (the root xylem) all the way to the leaves. Both soil and roots are porous media! The appearance and structure of the root system differs between plant species and changes with environmental factors (e.g. soil water content).
Some plants exude a gel-like substance which alters the hydraulic properties of the soil and enables the plant to take up water from very dry soil. Fine root hairs on the surface of roots are assumed to play an important role for water uptake. Complex root architectures can lead to a local redistribution of the available water. Such process cannot only be observed in experiments but can also be analyzed with computer simulations. For example, the complex interaction between direct evaporation from soil and transpiration from plants can be investigated by using simulations. Here, simulations have a crucial advantage over experiments. Processes can simply be switched on and off. This allows to investigate a process and its effect in isolation as well as in interaction with other processes. On the basis of investigations with detailed models of one or a small number of plants, it can be decided whether such processes can be neglected or must be considered in large-scale simulation such as climate models.
At the University of Stuttgart, we work—together with colleagues at Forschungszentrum Jülich— on the development of novel computer models for water and nutrient transport, root-soil interaction and root growth. These models are innovative tools for the assessment of scientific theories and hypotheses about water transport in soil. However, this is only possible under the premise that these models accurately represent all considered processes. For instance, most state-of-the-art models overestimate root water uptake in dry soils for a given atmospheric pressure. We analyze computer models and create improved model which address known shortcomings.
Credits: University of Stuttgart / Timo Koch
Fractures are features which are commonly found in geological materials, and they can have a strong impact on their hydraulic and mechanical properties. For example, highly conductive fractures in a low-permeable rock can act as preferential flowpaths along which rapid fluid flow can occur. Besides that, the fractures provide interfacial areas for the transfer of mass and/or heat between the fractures and the surrounding rock. Several geotechnical engineering applications make use of this fact, as for example geothermal energy or unconventional shale gas production techniques.
The motivation for the development of the model presented here originates from a project related to underground radioactive waste storage. One approach is storing the waste inside tunnel systems that are excavated within low-permeable clay formations, which act as barriers for flow and transport of potentially radioactive components over large time scales. Transport away from the emplacement tunnels could be driven by the pressure increase that is expected to occur due to the release of hydrogen gas as a consequence of the anaerobic corrosion of the metal canisters. Besides this, the excavation of the tunnels leads to the creation of fractures in the near vicinity of the emplacement tunnels.
The scientific question that arises is how the fractures present in the surroundings of the tunnels influence the hydraulic properties of the clay rock. Clay rock is relatively soft, which is why the current hypothesis is that the dilation of the present fractures can help to reduce the pressure build-up inside the emplacement tunnels. To this end, experimental studies on cylindrical rock samples taken from the surrounding clay rock are envisaged, with which the dilation of the fractures as function of the pressure increase is to be quantified.
The model that is presented here was designed to provide a tool with the help of which the experimental results can be better interpreted, and which allows for studying the hydraulic properties of a number of synthetically generated rock samples and fracture networks. In experimental studies, this would involve very large technical and financial efforts. The model considers a poroelastic rock matrix, that is, the interaction between the flow through and the deformation of the rock sample is taken into account, and the rock is described by means of a linear elastic material law. Besides this, flow along the fractures is considered, where the fractures are modeled as two-dimensional planes as the fracture apertures are typically very small in comparison the extent of the samples. The aperture is then a variable defined on the fracture planes and is a function of the deformation of the medium. This way, the influence of the deformation on the hydraulic properties of the fractures, and in turn on the hydraulic properties of the entire sample, is captured.
Credits: University of Stuttgart / Dennis Gläser
Magnetic resonance imaging (MRI) is an immensely important and versatile imaging technique used for medical imaging. The technique is based on the nuclear magnetic resonance of hydrogen atoms to radio frequency signals in strong magnetic fields. MRI avoids damaging radiation (e.g. X-rays) and is usually considered non-invasive. MRI of the brain is, for example, used in the diagnosis and therapy monitoring of brain tumors, in the analysis of neurodegenerative diseases such as Alzheimer’s or Parkinson’s disease, and diseases of the central nervous system, such as multiple sclerosis. In a variant of MRI called perfusion MRI, a contrast agent is injected into the blood stream and a sequence of MR images is taken to observe the fate of this contrast agent.
Why is it important that the brain is a porous medium?
Like most biological soft tissues, brain tissue consists of a mixture of cells, fibres, and fluid within the cells and the interstitial space (pore space). Cells are supplied by blood vessels with oxygen and nutrients. The blood vessels also consist of cells and blood is a mixture of fluid and various cells. This complex tissue architecture complicates the interpretation of the MRI images.
How can computer simulations help?
Computer models are the basis for the image post-processing done for perfusion MRI data. The model simulates both contrast agent perfusion and the resulting MRI signal and compares the result with the data. Hereby, certain properties of the brain tissue can be inferred. For example, the blood volume fraction can be estimated—an important biomarker for tumors, but equally important to assess the possible damage after a stroke. Simple simulations only take seconds and thus immediately provide important additional informations to medical doctors for decision making.
At the University of Stuttgart, researchers develop novel simulation techniques with the goal of extracting further information from MRI data. Using computer simulations, they try to better understand how contrast agent spreads in brain tissue and how different properties of the tissue influence MRT data.
Credits: University of Stuttgart / Timo Koch
Simulation or the question “What if?”
If you look up the meaning of the word “simulation” in the German Duden, the first entry is the explanation “to pretend” – i.e. “to act as if”. However, if you read on, simulation can have a different meaning: imitation. It stands for “what if?”. It is about reproducing processes and facts on the basis of models or imitating them realistically. For centuries, complex phenomena have been described by models in the sciences. Simulations enable us to understand important aspects of the systems described in this way, to predict changes and to decide how to control such systems.