SimScale is a cloud-native simulation and analysis platform accessed via a web browser. Users can instantly access a full-fledged HPC-powered simulation platform from a PC, laptop, or tablet​, ​with access to simulation features, learning resources, and an international community of more than 400K ​engineers​, including architects, urban designers, and engineers.

NVIDIA Omniverse is an extensible platform for virtual collaboration and real-time, physically accurate simulation. Creators, designers, researchers, and engineers can connect tools, assets, and projects to collaborate in a shared virtual space. Developers and software providers can also build and sell Omniverse extensions, applications, connectors, and microservices on the Omniverse platform to expand its functionality.

NVIDIA Omniverse SimScale Converter Extension: Seamlessly Export Scenes and Results Between Tools

The NVIDIA Omniverse SimScale Converter Extension is a powerful tool that allows architects and computational designers and users of the NVIDIA Omniverse tool to seamlessly export scenes from Omniverse to SimScale and bring back the results into Omniverse. This can be a valuable asset for a variety of projects, as it allows users to quickly and easily iterate on designs and test different scenarios.

To use the extension, simply upload your USD prims as models to SimScale. SimScale will then run a computational fluid dynamics (CFD) simulation on your model and return the results back to Omniverse. The results can then be visualized and analyzed in Omniverse, allowing users to make informed decisions about their design. The extension currently supports two types of simulations: pedestrian wind comfort and incompressible LBM (Lattice Boltzmann method). Pedestrian wind comfort simulations can be used to assess how comfortable it would be for pedestrians to walk through a particular area, while incompressible LBM simulations can be used to analyze the flow of fluids around a solid object, in other words building aerodynamics such as evaluating high winds, cornering effects, wind acceleration between spaces, etc.

In the future, the extension is expected to support additional simulation types and CAD formats. This will make it even more versatile and useful for a wider range of projects. Here are some specific examples of how the NVIDIA Omniverse SimScale Converter Extension can be used by architects and users of the NVIDIA Omniverse tool:

• An urban designer might use SimScale and Omniverse to evaluate the wind comfort around an existing building and proposed new development for planning permits.
• An architect could use the extension to simulate the airflow around a building to ensure that it is safe and comfortable for pedestrians.
• A designer could use the extension to simulate the performance of different types of trees, vegetation, landscaping, and street furniture and their impacts on site conditions.

For more information about the NVIDIA Omniverse SimScale Converter Extension, you may find explore the corresponding SimScale Coverter Extension documentation.

The Rise of CFD for Urban Design

Microclimate simulation using computational fluid dynamics (CFD) is a growing requirement for many types of buildings and developments. Complex building physics is needed to design and validate advanced net-zero and well-being requirements of modern building codes and rating systems. It is also necessary in order to supplement or even obviate the need for expensive wind tunnel testing. Using traditional desktop modeling tools requires too much computational resources and time to get meaningful results and prohibits the use of simulation at the early design stages as an iterative design tool.

Architects and engineers can benefit from fast and accurate design simulation using SimScale, accessed from a web browser. With no hardware setup or costs, SimScale allows designers to quickly access powerful simulation capabilities and perform multiple analyses using a single CAD model to understand:

• Pedestrian wind comfort criteria
• Wind safety modeling
• Building aerodynamics for urban design
• The use of mitigative measures such as windscreens, canopies, and vegetation

One of the often-most quoted advantages of using SimScale is the lattice Boltzmann method (LBM) integrated solver, which has extremely robust CAD handling features, meaning the simulation is indifferent to complex CAD (including terrain) and requires no geometry simplification to get the simulations going. It is this very solver that connects to Omniverse giving architects and designers the perfect combination of fast and accurate simulations coupled with compelling and powerful visualizations.

On-Demand Webinar

Watch this on-demand webinar to learn more about how to seamlessly export scenes and simulation results between NVIDIA Omniverse and SimScale and build advanced CFD workflows for your architectural design projects.

The post Wind Simulation in NVIDIA Omniverse™ appeared first on SimScale.

Forced convection cooling is an essential component of modern electronics where high power density devices and enclosures generate heat that needs dissipating. In many cases, natural ventilation-based heat rejection is not always enough, and fans must be used. Fan performance can be complex and based on several variables that need accurate representation in any modeling and analysis. Fan manufacturers provide a fan curve that describes the relationship between static pressure, power demand, speed, and efficiency values per-flow rate. This information is essential for cooling purposes, so we need to model the fan curve when possible. The pressure flow characteristics described by manufacturer fan curves dictate how the volume flow from a fan is affected by pressure drop. Most fan manufacturers provide this data based on standardized testing according to international standards.

Engineers evaluating different fan types and sizes in their device designs must be able to account for the fan curve and trust that it is a true reflection of how the fan might perform in real-world conditions. Creating a digital twin of the fan curve is an excellent way to simulate its performance. The SimScale platform is a cloud-native engineering simulation tool for general purpose flow, thermal and structural analysis. A typical application is the analysis of electronics enclosures for thermal management, specifically, scenario testing multiple cooling strategies. Fan behavior can be simulated in several ways in SimScale. A boundary condition where the fan inlet is placed can have a direct mass flow applied to it as a velocity inlet with appropriate fluid material properties and ambient temperature. Alternatively, a momentum source can be defined to represent the flow from a fan, and, most recently, engineers can now upload fan-curve data in a tabular format directly into the SimScale platform and use it as a boundary condition.

Thermal Performance of a Raspberry Pi Computer

To demonstrate some of the fan modeling features in SimScale, we have simulated a Raspberry Pi computer using a publicly available 3D model of the device taken from GrabCAD. A conjugate heat transfer simulation with forced convection (fan) is used to model the computer using standard 30 mm fans for cooling. We have used manufacturer data for the fan performance, which comes in four different model types; each has a fan-curve providing data on volumetric flow rate and pressure drop (static) at ambient conditions. We have taken the 3D model and simplified it using the CADmode editing features in SimScale. The study is not interested in geometric changes to the enclosure, only in fan and cooling performance.

We can directly upload the manufacturer fan curve data and derive the fan operating points explicitly using simulation. We want to show how the manufacturer’s data can be applied to the former. We can extract the fan specification sheet data into a spreadsheet to upload into SimScale and perform a complete thermal analysis of a Raspberry Pi. In the latter case, we can conduct a flow rate study using simulation to derive the fan operating point, e.g. at what flow rate can we generate the given pressure drop and derive a system resistance curve. Using the parallel simulation capabilities in SimScale, we can run multiple operating points simultaneously and begin to generate the fan curve for evaluating how efficient these fans are for cooling purposes and whether they adequately cool the computer chips in the enclosure. This approach is instrumental when the full fan curve data is not available or to assess in-situ fan performance in the system or device (the manufacturer data is for a simple test setup).

A simplified evaluation might look like this:

• Use the original CAD model and a fan curve as the baseline case
• Derive operating points from simulation
• Derivation of system resistance curve from flow rate study
• Cooling efficiency comparison with different fan models
• Compare to directly uploading manufacturer data

We have used the conjugate heat transfer (CHT) analysis type in SimScale. The CHT analysis type allows for the simulation of heat transfer between solid and fluid domains by exchanging thermal energy at the interfaces between them. Typical applications of this analysis type include heat exchangers, cooling of electronic equipment, and general-purpose cooling and heating systems. A multi-region mesh is required for a CHT simulation to have a clear definition of the interfaces in the computational domain. With the interfaces adequately defined, this is automatically taken care of in SimScale and, in this case, generates a five million cell mesh. For the simulation setup – fan inlet and pressure outlet boundary conditions are used, the air is used for the flow region, and several materials are specified for the chips and electronic components, including Copper, PCBs, Silicon for chips, and Aluminum for heat sinks. Power values represent the CPU (3 watts) and more minor chips (0.25 watts). The air inlet is ambient at 19.85 ℃, and a CSV file is used to upload the fan curve for later use. 

Thermal Simulation of Electronics Cooling

We can visualize heat removal on the chips by looking at surface area average temperatures. The images show up to 392 K on the CPU (max). SimScale will also extract point-specific data and pressure drop across inlets and outlets. Fan inlets show a 4.78 pascal (0.5 mmAq) pressure difference at a flow rate of 7.55e-4 m3/s, and this matches the fan curve data sheet for the baseline model (L). The fan outlets are at 0 gauge pressure as intended. We can easily switch the fan curve data to simulate a more robust fan (H) for comparison. Doing this shows us a 10 pascal (1 mmAq) pressure drop, and the two are compared in the image below. By doing this analysis, we can start generating a system curve for the enclosure (Raspberry Pi) derived from simulating various fan operating points (Orange line in the image below). Running ten simulations of different operating points in parallel, we generated the entire system curve for the Raspberry Pi in two hours. Chip cooling performance based on comparing the two fans is also shown, with the more substantial model H fan better at removing heat.

Fans as Momentum Sources

In some cases, modeling the fan as a momentum source might be needed. Momentum sources can be used to simulate fans, ventilators, propellers, and other similar fluid acceleration devices without having to model the exact geometry of the device. For instance, users might want to model a fan whose dimensions and output velocity are known. With this feature, it is possible to assign average velocities or fan curves to volumes of interest. An external flow domain is needed here to act as the air supply for the CAD opening where the fan would be, and a momentum source is used to define the flow behavior. Users can upload the fan curve, decide which direction the flow is going, and add a geometry primitive cylinder as the source. The external flow domain is used because flow enters the fan as it would do in a reality where the fan inlet is the interface of internal and external boundaries.

Summary 

Simulation is now considered essential to optimizing electronic product design and performance. Using multiple methods to represent complex fan performance, engineers can quickly evaluate the cooling impact of fan types, and using the parallel simulation capabilities in SimScale, a more comprehensive range of scenarios can be evaluated. In summary:

• Fan curves: allows modeling of fans based on fan curves (flow rate to pressure drop relation)
• Fan boundary condition: users can specify a fan inlet or fan outlet as a boundary condition to model fans that are placed at the edge/outside of the enclosure domain.
• Fan Momentum Source: allows modeling of internal fans as a momentum source that are embedded within the model.

Furthermore, in the same platform, engineers can perform virtual shaker table tests, structural analysis, and more specialist fluid flow analysis using the same CAD model and simulation environment. 

To learn more, watch the fan modeling webinar below:

The post Electronics Cooling Using Fans appeared first on SimScale.

SimScale has multiple features useful to engineers designing and developing drones. Salient features include the ability to evaluate:

• Flow
  • Aerodynamic loads under varying flight conditions
  • Airframe-rotor-surrounding interference & impact on the stability
  • Configurational effects on payload, endurance, and efficiency
  • Aerodynamically optimized landing and take-off angles
• Structural
  • Drone drop test 
  • Blade stress analysis
  • Rotor shaft interference fit
  • Vibration analysis
  • Battery pack crush test
• Thermal Management
  • Drone electronics cooling
  • Early-stage battery pack design
  • Battery pack cooling validation

Let us see how some of the above capabilities and features were applied to evaluate the performance of a quadcopter.

Simulating a Quadcopter 

Starting with the airflow and aerodynamic performance, we can set up a CFD case to evaluate the quadcopter in hover mode. This is especially important to investigate as it thoroughly consumes the most power requirements. Drones in hover mode will also be in close proximity to the ground/roof or other objects and might be in confined spaces. Many applications will include drones entering or close to construction sites, mines, and homes for delivering packages and in urban areas where turbulence is more significant. Aerodynamic stability in hover mode is thus critical for the safety and overall performance of the drone and its surroundings. 

We have taken a Parasolid CAD file of a standard quadcopter and made minor adjustments by removing the landing legs and camera mounts. This was done to simplify the geometry. Using the CAD editing feature in SimScale called CADmode, we can create the external fluid domain and further edit the geometry. In the first instance, we created rotating zones for the rotors with a rotational speed of 10,000 rpm. We have selected the Subsonic analysis type in SimScale, which lends itself to rotating machinery applications The Subsonic analysis type includes a robust meshing strategy that produces an automated and robust hexahedral cell mesh, using the Cartesian meshing technique that significantly reduces mesh generation times by order of magnitude. It uses a finite volume-based solver optimized to handle a wide range of flow types. We have applied a globally coarse mesh with refinements around the rotating zones. Engineers and designers who evaluate the drone in hover mode will want to know its thrust performance, rotor-rotor, and rotor-airframe interactional effects. Selecting the analysis type automatically creates the simulation tree, a guided workflow for completing the simulation setup and post-processing the results. A user can now go through and set the material properties and boundary conditions and add rotational speeds to the rotating zones. The rotors are made of a polymer material (PVC); the fluid is air by default. Further time savings can be achieved by exploiting the symmetrical nature of the drone geometry and reducing the model size.

Using a transient simulation to capture time-dependent effects, we simulated for 200 minutes. Users can see velocity around the quadcopter and ascertain pressure to see forces in regions of interest, such as the rotors and joints.

We can run multiple scenarios of differing ground effect heights using further simulations. The ground effect in a multirotor is a change in the thrust generated by the rotors when flying close to the ground due to the interaction of the rotor airflow with the ground surface. It is critical for engineers to address the ground effect for safety and performance correctly. Steady-state simulations of the quadcopter take 20 minutes to run. Using the cloud’s computational power, several simulations can be set up and run in parallel, meaning the total time taken to complete dozens of simulation runs, for example, is still 20 minutes. When the quadcopter is closer to the ground, there is much more pressure on the drone’s underside, and the amount of thrust from the rotors is affected. It must have enough aerodynamic stability and structural integrity to still perform under varying forces, and this feedback is used intricately in the flight control system. Interestingly, we can see some wake air re-entering the rotors using velocity streamlines. This can cause safety problems in hover mode when close to the ground; for example, dust/dirt and other debris might enter the rotors/motors, causing damage.

Drone power requirements change depending on rotor configuration. The number of rotors can impact thrust, drag, and overall performance. Engineers can investigate if a four or six-rotor aircraft might be better and the number of blades on each propeller. In this example, we can generate a force plot from the results and see a force 544N produced. For this study, we have compared the thrust outputs using SimScale to a published academic study showing experimental results and those from a competing CFD tool. The experimental results are shown in red, and the difference is attributed to geometry variations. SimScale results match closely with the competing CFD tool. The advantage of SimScale, however, is that the rate of change of operational thrust points on the curve can be simulated in parallel for each ground effect ratio, saving much time and cost.

Flight Integrity 

The structural forces on a drone can be considerable and change instantly. There are short-term structural effects like when in hover mode or landing; and longer-term material/joint fatigue and stress due to prolonged usage. Vibration analysis is critical to understanding fundamental structural integrity influencing flight stability and safety. Control systems must account for natural frequency and induced vibrational effects. With the right inputs, a flight control system can optimize the stability of a drone by avoiding specific rotor rotational speeds that might excite vibrations and cause adverse effects on the drone. Engineers must correctly apply vibration analysis to design stable drones over various rotational speeds. We have analyzed a propeller to evaluate its natural frequencies at 10,000 rpm and calculate the corresponding eigenfrequencies we wish to avoid. We can then run a harmonic analysis that accounts for these natural frequencies and the material’s damping properties to assess the propellers’ stress and displacement at those eigenfrequencies. For example, we can see high stresses at the rotor base that will cause fatigue over time.

Engineers can also take the entire drone geometry and look at the global eigenfrequencies for the drone that might cause destabilizing movement. In this example, we calculated that 1630 and 6600 rpm would cause harmful effects due to adverse excitation of the drone. A Campbell diagram was used to find these values and represents a system’s response spectrum as a function of its oscillation regime. We have analytically calculated the eigenfrequencies as a function of shaft rotational speed. Essentially we plot the rotational speeds against the eigenfrequencies and read off where rotational speeds intersect the horizontal eigenmodes, which might cause excitation, and have found the intersection points occur at 1630 and 6600 rpm. For example, the rotor materials (PVC) and configuration can be changed to design around these problem points.

Drone Flight Simulation

CFD has been helpful in simulating complex aerodynamics, evaluating drone performance & stability and visualizing airflow behavior. Structural analysis has allowed us to generate a Campbell diagram for vibration analysis, evaluate blade stress and durability and assess drone structural integrity. All these capabilities and many more are available to engineers through a web browser, and existing template projects make it easier to start.

“We realized that some of the simulations of our drone design could be run much quicker and better with SimScale. Also, we used to not have a good process established, but now with SimScale, we have a proper design methodology and a validation tool for all the design modifications and improvements.”

— Marta Marimon, Aeronautical & Flight Mechanics Engineer at VTOL Technologies

This on-demand webinar helps teams involved with drone/UAV research and development learn how to test their products for desired aerodynamic performance and structural integrity with cloud-native CFD simulation:

The post Drone Flight Simulation appeared first on SimScale.

Users benefit from extensive collaboration features built into SimScale that allow entire teams of stakeholders to efficiently work together to solve some of the most significant design challenges needed for a sustainable future. This new SimScale whitepaper on Microclimate Simulation for Urban Design explains how embedding these capabilities in organizations is simple, powerful, and adds significant value to a firm’s service offering.

Robust Solvers

A primary aspect to consider is the robustness of solvers in handling large and complex geometry. A typical CAD model developed by architects and engineers is not easy to import into traditional simulation tools. Most CAD models have too much detail including intersections, holes, gaps, and open shells with complex terrain that require tedious and labor-intensive CAD cleanup when utilizing traditional simulation tools, often stifling the desire for simulation. SimScale has a robust solver that can handle common CAD imperfections. There is no meshing in the traditional sense, meaning that the preprocessing is fast and architects can begin simulating on their as-designed CAD model without having to strip away detail or spend days making cleaner versions for simulation purposes. 

Scalability

Secondly, the GPU-based pacefish® Lattice-Boltzmann Method (LBM) time transient integrated solver achieves considerably higher performance compared to CPU-based solvers resulting in extremely rapid calculations even on large models. It has reduced the simulation times by an order of magnitude for city-scale models and produces animated and visually insightful graphics that are key to understanding complex flow behavior and communicating designs to clients.

Collaboration

In the early stages, designs are flexible and fast-changing; it is essential to operate together with all the stakeholders on a single “source of truth” model. With SimScale, users can bring stakeholders from different teams together to visualize, evaluate, and derive design decisions for further iterations –collaboration features are built into the simulation workflow. Collaborative features in the platform adapt to the specific design practice, they can be leveraged either in the SimScale platform directly, or integrated into the company’s existing workflows via API integrations.

Perkins & Will, a leading global design firm has benefitted from these features on a recent large-scale competition masterplan in a challenging desert environment near a coastal area.

The design brief included several environmental aspects that needed fast and accurate simulations of the microclimate to inform critical design decisions. Accounting for rapidly changing wind conditions and outdoor safety was critical in the design of the overall scheme. Perkins & Will used SimScale for early-stage design iterations to simulate multiple wind directions and climate scenarios, iterate various masterplan CAD models, and design-in mitigation measures.

Microclimate Simulation for Urban Design

A future planet is one in which 80% of the world population will live in densely packed cities with global mega-trends such as urbanization and climate change adding even more stress onto fragile global systems. Simulating microclimate impacts on buildings and cities is critical for achieving true sustainability and comfort targets to move the built environment towards a net-zero carbon trajectory. Early-stage design is where many of the key design decisions are made. Having access to engineering simulation that is intuitive, accurate, and fast is essential when iterating through designs with the wider project team. Cloud-native simulation software, like SimScale, eliminates the need for outdated on-premises hardware and grants full access to high-fidelity engineering for architects and engineers with integrated collaboration features for distributed teams.

Learn how engineers can harness the powerful features in SimScale for evaluating building aerodynamics, pedestrian wind comfort, indoor and outdoor thermal comfort, and structural wind loading on buildings, structures, and entire cities in this whitepaper:

The post Microclimate Simulation and Urban Design appeared first on SimScale.

SimScale supports the most common geometry formats for importing CAD, including Solidworks®, Inventor®, STEP, IGES, STL, and Parasolid®. Third-party CAD connector apps are available for Onshape® and other tools, allowing seamless integration. Furthermore, the native application programming interface (API) in SimScale, with both a Python and C SDK, amplifies the ability to customize CAD workflows and connect or control third-party tools. In this article, we have used a 3D model of a support bracket to introduce the integrated 3D CAD and meshing tools in the SimScale platform. Continue reading and watch our on-demand webinar, CAD Geometry Preparation and Meshing for Engineering Simulation in the Cloud to learn how to source and import 3D CAD models, clean geometry by using internal CAD mode editing features in SimScale, and apply various meshing settings to make the model simulation ready in a few simple steps. With new features continuously added, CAD mode supports operations like scaling, extrude, body and face delete, surface splitting, etc. Engineers can use CAD mode associativity with Onshape and other third-party tools to streamline parametric studies from a baseline CAD model, a powerful feature further enhanced by parallel simulation in the cloud. 

Simulation-Ready CAD 

After CAD upload, additional preparation might be required depending on how the file has been created. SimScale offers a dedicated environment to interact with your model called CAD mode that helps users prepare the model within SimScale without switching to external CAD software. CAD mode supports operations like scaling, extrude, body and face delete, surface splitting, flow volume extraction, etc., with continuously adding new features.

The CAD mode feature in the SimScale platform offers a set of purpose-built CAD editing and simplification tools. For example, the Flow Volume Extraction operation alleviates engineers from laboriously isolating a flow volume as required in other pre-processing systems. Unlike traditional CAD systems, CAD mode focuses on a core set of simple, intuitive, versatile tools ideal for making CAD models simulation-ready. CAD Mode is used to interact with a CAD model, delete, extrude, or scale CAD parts, and perform CAD-related operations directly within the platform. Any gaps or interferences in a 3D model can be automatically identified using CAD mode, which is also used to fix these issues. Preparing, uploading, and adapting your CAD model for analysis is the first step in setting up a simulation. CAD file associativity between varying CAD files is applied automatically in SimScale, maintaining naming conventions for parts/faces from the original CAD model. This means that when swapping CAD files for comparative studies, users do not have to reassign boundary conditions, mesh settings, or result control outputs, making comparing two or more CAD variants of a single product much faster. 

The full list of supported CAD formats:

• SolidWorks (.sldprt, .sldasm)
• Autodesk Inventor (.iam, *.ipt)
• Rhino® 4, 5, 6, and 7 (.3dm)
• CATIA (.CATPart, .CATProduct)
• PTC Creo®  (.prt, .asm)
• Siemens NX (.prt)
• Solid Edge (.par, .asm, .psm)
• Revit® (.rvt)
• Neutral formats:  Parasolid (.x_t, .x_b), ACIS (.sat, .sab), STEP (.stp, .step), IGES (.igs, .iges), STL (.stl)
• CAD Plugins
• SimScale Connector App for Onshape
• SimScale Plugin for Solidworks
• SimScale Integration for Autodesk® Fusion 360
• Grasshopper
• SimScale CAD Mode

Quickly Mesh & Solve Complex Designs

Mesh generation has been a labor-intensive and tedious aspect of traditional CFD software. The SimScale platform is based on a unique meshing technique that ensures the mesh is robust enough to apply to many types of CAD models the first time around. A bespoke meshing technology is harnessed to generate meshes quickly for complex geometries. Engineers can mesh and simulate models with an easy-to-use and intuitive workflow. SimScale strives to make the meshing process as simple and user-friendly as possible. A user should only have to decide on the trade-off between mesh fineness and the required computational resources (number of processors assigned). Due to the robustness and general applicability of the meshing algorithms in SimScale, both automated and manual options are provided. Unlike traditional simulation software, meshing is the second step in the simulation setup after geometry upload.  SimScale delays the meshing action until all other steps are completed so engineers do not waste time waiting for mesh generation. Users can process with simulation parameters or boundary conditions; for example, when they hit simulate, the meshing actually begins.

CAD Preparation and Analysis of a Support Bracket

In this project, we have selected a 3D CAD model of a wall support bracket from the extensive library in GrabCAD. Engineers can access different file formats for the same products depending on their preferences. We then open the support bracket model in Onshape, which integrates with SimScale to import CAD assemblies directly. Any CAD preparation, if needed, can be undertaken in Onshape or SimScale. The model is then imported into SimScale, ready for static structural analysis. We have selected the support bracket to investigate its loading response to a 100 Nm load and evaluate its reinforcement’s performance. To do this, we have created a duplicate of the CAD model and removed the reinforcement. We now have a 3D model of a support bracket with and without reinforcement for a comparative study. Removing the reinforcement can be done in Onshape, before importing into SimScale or, using CAD mode, this minor modification can be completed in SimScale directly.

Analysis-Driven Design Insights 

Once the CAD models are ready, a few steps are needed to set up a simulation. In the first instance, we have not modeled friction, gravity, separation contacts, or the bolt connections:

• We first define the materials for the support bracket. The default metal steel has been selected. Later we can perform a comparative analysis by duplicating this material and customizing its properties with a higher Young’s modulus, for example. Customized and newly added materials appear in the materials library and can be shared across projects and team members. 
• Then we assign structural constraints using a fixed wall boundary condition and a 100 Nm load to the model. In this case, we will output the von Mises (yield) stresses, which is a value used to determine if a given material will yield or fracture. It is used chiefly for ductile materials, such as metals. The von Mises yield criterion states that if the von Mises stress of a material under load is equal to or greater than the yield limit of the same material under simple tension, then the material will give way. 
• Finally, we hit simulate, which is also when the meshing takes place, with default settings. Later we will refine the mesh. 

A simple analysis of the two CAD models (with and without the reinforcement) shows us the value of adding the reinforcement. The images below show the load applied displacement from the brackets’ designed state. The image on the right is the CAD model without reinforcement and has an enormously higher displacement of 70 mm compared to the original model (with reinforcement) displacement of only 2 mm. The structural stresses are also much higher and over a larger surface area in the modified model. We might add bolt connections with pre-load stresses with more time and model the wall for more accurate results.

Mesh Refinement 

The automatically applied default mesh with a level 5 refinement in SimScale gave us a mesh size of 443,000 cells. The default setting is good enough to refine small features and edges without further guidance. Although the results using the default mesh should not be considered mesh independent, they are accurate and sufficient for early-stage analysis and quick insights. The following mesh we have generated is globally refined at level 8 refinement and might be considered a mid-level quality mesh with 678,000 cells. Finally, we apply a more advanced locally refined mesh where we manually select the surfaces and regions of interest, in this case, the gusset/reinforcement. Here we have applied a maximum edge length of 0.0005 m for the mesh giving us a mesh size of 845,000 cells. The images below illustrate the comparative results between the three meshes. 

CAD Cleanup and Meshing for Simulation 

Engineers need tools where they can quickly clean up and prepare their CAD models and assemblies to make them simulation ready. Using the inbuilt features in SimScale and the CAD integrations such as Onshape, users can spend less time on tedious CAD operations and more time simulating their designs and extracting valuable performance data. Engineers are further empowered by the parametric simulation capabilities in SimScale and the practically unlimited computing power in the cloud.

Follow along in a step-by-step demo that introduces our 3D CAD and meshing tools for simulation:

The post CAD Cleanup for Simulation appeared first on SimScale.

Flow Simulation to Shape Optimization 

A GEMÜ 534 globe valve is used for shape optimization, as shown in the image below. The valve exhibits a sharp change of flow direction resulting in higher pressure loss and lower flow rate values compared to a more common angle seat valve body. This valve has a pneumatically operated plastic piston actuator and is available as a shut-off or control valve. The valve is standard in industrial water treatment applications, chemical processing, power plant operations, and mechanical and processing industries. A conventional CAD tool has limitations for optimization, especially for complex geometries where small changes in shape can disproportionately influence outputs such as flow and pressure. Because of these highly sensitive relationships between design change input and performance output, the physical processes governing valve performance are complex and challenging to quantify. The non-orthogonal geometry of the valves also means that many geometric input variables influence their performance. This study aims to investigate the relationship between geometric inputs and valve performance using numerical methods and identify any potential for cost and development time savings.

Most typical bottlenecks in setting up and running a design exploration or optimization process are related to the handling of geometry, as these specialist requirements are less common in the engineering discipline. Geometry variation with traditional CAD systems is often tedious or prone to failure, and it is challenging to consider or even automatically fulfill constraints. Simulation engineers depend on CAD teams to provide geometry (variants) based on manual or ad-hoc simulation analysis. The quality of the CAD model might also not be suitable for simulation (e.g. water tightness, level of detail), which is another common hindrance to spending some time setting up an optimization study. The CAESES tool can be used to overcome these common issues and using the CAESES/SimScale workflow, a simple workflow for this case has been developed as follows:

• Choose a suitable baseline CAD model that is suitable for simulation
• Parameterize the CAD model using CAESES
• Define boundary conditions and simulation setup for the CFD analysis (SimScale)
• Identify the design/parameter space (DoE) that needs to be solved (CAESES)
• Run CFD simulations on selected CAD variants (SimScale)
• Analyze results 
• Reduce the number of CAD variants
• Optimize (CAESES)

The basic simulation setup is a 1 bar pressure inlet with water as the fluid. The outlet pressure is ambient. In this case, fluid flow rates, temperatures, or material properties were not altered, although this would be a simple enough exercise using the parametric simulation features in SimScale. CAD variations using several key geometric dimensions are the focus of this study, and the valve geometry has 16 parameters that have been parameterized using the CAESES tool. The inlet and outlet fluid domains have been extended on either side to allow a developed flow to enter the valve.

In this case, we are looking for significant improvement in the Kv value for the globe valve, a standard measure of valve capacity. The Kv value expresses the amount of flow through a valve at a given valve position with a pressure loss of 1 bar. The unique situation when the valve is fully open is referred to as the Kvs value. When the plug is closed, the flow is zero, and it increases with plug opening in a correlation that depends on the type of valve. This study aims to improve upon the Kv value for this particular globe valve using geometry optimization.

The geometric dimensions are referred to as contours on the CAD model. The following images show two of these contours using a simple line diagram and how changes in those contours affect the 3D CAD model or extracted fluid domain in this case:

• Inlet and Outlet
  • 4 parameters on inlet long contour
  • 2 parameters on inlet short contour
  • 4 parameters on outlet long contour
  • 2 parameters on outlet short contour
• Cross-sections:
  • 2 parameters for inlet ellipse aspect ratio distribution
  • 2 parameters for outlet ellipse aspect ratio distribution

Once CAD preparation and parameterization are done, the model in CAESES can be connected to SimScale using the API and instructed using a Python script. At this point, the parameterized model, which has a generic interface for coupling to external analysis tools, could also be used for different types of physics analysis, including structural and thermal analysis. Defining input and output parameters is also done at this stage to allow simple interpretation of results, which is needed when the extraction of result values is used as objectives or constraints by the optimization algorithm. Geometry and script files are exported from CAESES and loaded into the SimScale platform. The result files generated from SimScale are downloaded to the local workstation after the simulation for further analysis (CAESES runs locally). Using the parameterized geometric contours, the initial design of the experiment starts with 120 designs or shape candidates. In this optimization case, cycling through the first optimization iterations of the workflow, the 120 designs are followed by a response surface optimization and reduced to 50 designs. The baseline Kv of 54.93 has now increased to 58.49, an increase of 6.5%. 

The original 16 parameters have been reduced to eight, and two additional parameters have been added for the long contours and two cross-sections, now giving 12 parameters for the second run. This has been based on flow simulation results (Flow rate and pressure drop to determine the Kv value). Further simulations and optimization led to an increased Kv value of 59.5, an 8.3% improvement over the baseline.

Simulation-Driven Shape Optimize Globe Valve 

Computational fluid dynamics (CFD), combined with shape optimization, is a powerful tool for engineers and designers who want to optimize their product designs. SimScale can be easily connected with specialist third-party tools using the SimScale application programming interface (API). 

Simulation can increase knowledge about a product’s behavior early in the design process and offer insights for improvement, along with quantitative and visual evidence critical in making informed decisions. Simulation-driven Shape Optimization and automated design exploration amplify the benefits of simulation further leading to better product designs with vastly reduced design cycles.  

The critical components of this ideal tool-set are a simulation tool, a driver of the optimization process with suitable DoE and optimization algorithms, and finally, an appropriate CAD tool that can produce different geometry variants based on the feedback from simulation results. This workflow has been used to improve the Kv value of an industrial globe valve by 8.3%. 

Follow along step-by-step in this on-demand webinar that shows how leading companies in the fields of CFD-driven shape optimization are using SimScale to automate their design process:

The post Shape Optimization of a Globe Valve appeared first on SimScale.

Virtual Wind Tunnel Testing in the Cloud

Many engineers are now using SimScale for wind load assessments to calculate mean and fluctuating wind loads on buildings, facade elements, and solar panels, whether rooftop or ground-mounted. The SimScale platform offers a single source of truth for analyzing wind comfort, wind loads, and more, using a fast and accurate lattice Boltzmann method (LBM) solver with robust CAD handling capabilities and automatic meshing. In this example, we take a look at a standard industry case for facade pressure and wind load assessment. SimScales’ virtual wind tunnel application allows for simulating turbulent wind flows in urban environments (Figure 1). Simulated fluctuations in the atmospheric boundary layer (ABL) can be seen in Figure 2. Mean and fluctuating wind loads on a whole building or facade element can now be calculated.

The building simulated is the CAARC building (45m by 30m and 180m in height) in context (with surrounding buildings) placed downstream in the virtual wind tunnel. The solver used is the LBM Pacefish© with a DDES, K-omega SST turbulence model on a mesh with 47 million cells (Figure 4). The setup is modeled at a scale of 1:400. The virtual wind tunnel setup can be seen in Figure 3.

Wind Loading Simulation Results

The main simulation result we are interested in is the wind pressure coefficient (Cp) as visualized in Figure 5 and shown in Figure 6. These results can be visualized in SimScale using the inbuilt post processor or, they can be downloaded for further analysis in spreadsheets or third-party tools such as Paraview. Similar analyses can be performed on arrays of solar panels, facades, structures, bridges, etc. These results are further validated by comparing them to a published wind loading study on the same CAARC building, from which our virtual wind tunnel setup has been duplicated. Comparing the results for the atmospheric boundary layer created in both the published wind tunnel study and the virtual wind tunnel in SimScale, an excellent reproduction is achieved. This means the virtual representation of an advanced wind tunnel setup is adequately simulated; therefore, engineers can have confidence in the results. Figure 6 shows the Cp values for each facade around the building. Each Cp value is extracted using probe points added at simulation setup. The post-processor allows as many probe points as needed and offers engineers more advanced visualization capabilities including 3D slices, streamlines, and transient results for creating animations.

Cloud-Native Wind Loading Studies

Engineers using engineering simulation in the cloud can now perform wind loading studies quickly and accurately, having confidence in their results. Advanced wind tunnel parameters are adequately duplicated and validated using the virtual wind tunnel in SimScale. SimScale customers are leveraging the power of wind loading analyses to optimize buildings and structures. Adrian Smith + Gordon Gill, for example, reduced wind load by 35% on their design of a super tall building using parametric modeling and shape optimization. They also modeled a vertical axis wind turbine integrated into the tower to harvest wind energy to meet 20% of the building base load requirements of 180,000 kWh/annum.

Learn more about SimScale’s virtual wind tunnel application for fast and accurate modeling of turbulent wind flows in urban environments in this on-demand webinar:

The post Virtual Wind Tunnel for Structural Design appeared first on SimScale.

SimScale users have access to hundreds of simulation models and templates in a public library that is a repository of knowledge learned on ventilation design assessments by a global user community. Lessons learned from these projects enable engineers to quickly and accurately predict wind and buoyancy-driven air movement using air quality simulation, and fresh air rates using natural ventilation simulation augmented with mixed-mode solutions. Here we will focus on insights into the performance of mixed-mode ventilation strategies in a classroom by simulating accurate representations of spaces, occupancy, internal gains, and equipment using a 3D CAD model, with multiple design variations for comparative analyses.

Workflow to Simulate Ventilation Equipment 

A 3D model of a classroom is shown in the below images. A parametric model is set up such that the floor area number of occupants can be changed easily. Internal loads (heat gains) are applied to each occupant using data from ASHRAE, LEED, and CIBSE guidelines. Additional heat gains are applied to represent low-power LED lighting and electrical equipment including laptops and a large screen display. Adventitious leakage is applied to the room at 0.1 (air changes per hour) ACH to reflect building infiltration from the outdoors, with an ambient temperature of 15℃. Air is selected from the extensive SimScale materials library for the indoor fluid and the walls have materials applied to them that reflect typical building fabric U-values for new school buildings in the UK. One of the facades has a series of windows with a top-hung style opening and opposite the windows on the inside, are four ventilation diffusers that can be set to supply, extract, or any intermediate combination. They can act as natural ventilation openings (passive) or have a velocity flow rate applied in either direction to represent supply and extract fans. A simple change in the boundary conditions offers the user a high degree of flexibility in evaluating multiple ventilation strategies.

The convective heat transfer (CHT) analysis type is chosen and is ideal for internal room airflows where temperature effects must be captured. CHT allows natural convection (buoyancy and wind-driven flow) and forced convection (from fans or other devices) to be modeled and is considered a robust type of analysis for internal fluid domains, capturing the effects of density and gravity. Additionally, radiation can be simulated as well as a pollutant species represented by applying a diffusion coefficient, using the passive scalar approach. In this case, we are modeling CO2 in parts per million (ppm) as an indicator of indoor air quality. Outputs include internal temperature, velocity distributions, thermal comfort criteria, humidity, radiation loads, and more CO2 concentrations. 

Using Simulation Outcome to Design Productive Learning Spaces  

The classroom has three configurations that have been simulated and the results of various analyses are summarized in the table below:

1. Base case scenario – supply air from horizontal diffusers only. 
2. Configuration 1 – supply air from a diffuser with upward-directed grilles for better air distribution
3. Configuration 2–  same as Configuration 2 and with the top-hung windows open

The base case scenario, although supplied with air, suffers from high heat gains and a build-up of CO2 due to a lack of mixing and ventilation. The diffuser grills are set horizontally at the inlet to the classroom and cold air from the supply tends to move downward due to the ambient conditions and creates an uncomfortable downdraught with poor ventilation efficacy. Configuration 1 and 2 have upward-facing grills at the diffuser, giving better circulation of fresh air inside the room. Similarly, with air quality, the base case configuration has high concentrations of CO2 at the ceiling level due to improper mixing of fresh air. Configuration 1 and 2 give better mixing that drastically reduces CO2 concentration levels.

Concerning indoor thermal comfort, the base case configuration is dominated by buoyancy-driven stack effects that direct temperature and air movement in the space. Due to the pressure distribution in the space and the downward fresh air supply from the horizontal diffusers, there is a stark lack of air movement. Configuration 1 and 2 offer better air circulation and the indoor space is subjected to better mixing of temperature. Configuration 2 offers better air mixing and quality using openable windows that introduce fresh air at lower ambient CO2 levels (400 ppm). Having the openable windows on the opposite side to the fresh air diffusers benefits from cross ventilation. 

According to ASHRAE 55 and ISO 7730, the following are guidelines for indoor thermal comfort:

• For Predicted Mean Vote (PMV):
  • ASHRAE 55 
    1. -0.5 < PMV < +0.5
  • ISO 7730
    1.  -2 < PMV < +0.7 (Limits)
    2. -0.7 < PMV < +0.7 (Existing Buildings)
    3. -0.5 < PMV < +0.5 (New buildings)
• For Percentage of People Dissatisfied (PPD):
  • 5% < PPD < 100% – depending on the calculated PMV
  • To comply with standards → no point in the room should be above 20 % PPD

Marginal changes are achieved in the thermal comfort of occupants between the base case and other configurations using the predicted mean vote (PMV) measure. Our classroom is generally between -0.6 and +0.6 in PMV. Although thermal comfort is not the focus of this exercise, improving the PMV further can be accomplished using better fabric thermal performance values and lowering the supply inlet temperature, for example. Users can also simulate a heat recovery option in the case of extracted flow through a heat exchanger device.

Further Air Quality Insights 

Let’s take a closer look at air quality. High levels of CO2 in classrooms and learning spaces have been linked to decreased cognition and exam scores. This is a widely published area of study. Similar patterns are seen in the productivity of workers in office spaces. Common guidelines for acceptable CO2 concentrations are given below. Classrooms should be less than 1000 ppm in controlled ventilation spaces. Some building regulations allow CO2 concentrations of 1000-1500 ppm where solely natural ventilation is used, although recent trends have favored mixed-mode ventilation in schools to provide better air quality and thermal comfort.

The following images show horizontal slices taken at occupant head height (seated) in the classroom and are colored by CO2 concentration, followed by a vertical slice showing the same. The CO2 source is applied at the head of each occupant. Configuration 2 has the best air quality performance and shows 22% less CO2 in ppm than Configuration 1. Opening windows makes a significant difference by benefiting from fresh air with a low CO2 concentration, better mixing in the room, and cross ventilation. The CFD simulation shows us the spatial distribution of CO2 that can be used for identifying areas of stagnant and stale air. It is also useful to guide the placement of CO2 and other environmental sensors. Spatial level detail from CFD is a major compliment to common energy/thermal modeling tools that designers use which are limited to time-series analysis of one point (node) in the center of the room (thermal zone). Using CFD and thermal models together can give better insights to architects and engineers who are trying to optimize a complex design brief including minimizing energy consumption whilst attaining strict air quality, fresh air, and thermal comfort requirements.

Simulating Ventilation Equipment with SimScale

We have shown how building openings have been parametrically studied to alter internal fresh air supply and air quality thresholds using fluid flow and thermal analysis types in the SimScale platform. With SimScale, users have access to powerful and fast simulation features specifically for analyzing thermal comfort, indoor air quality, and their influences.

“The SimScale platform has allowed us to develop and test scenarios including ventilation flow paths and understanding the dynamics between building fabric, airtightness, window natural ventilation design, and CO2 concentrations in various utilization scenarios. The platform has given us quality, visually digestible output to inform decision-making and to compare results against our monitored data from our extensive post-occupancy studies. Working with SimScale has allowed us to study our learning spaces in a way we never have before. It has helped inform us on which design strategies may provide the biggest impact on performance.”

— Robert White, Technical Associate and CEPH Designer at Architype

Get an even better understanding of CFD simulation for simulating ventilation equipment and airflow with our on-demand webinar with ASHRAE. Our specialists teach techniques that enable engineers to quickly predict wind and buoyancy-driven air movement, air quality, and fresh air rates using high-fidelity engineering simulation:

The post Using CFD to Simulate Ventilation Equipment appeared first on SimScale.

This article presents a design and simulation study of battery cold plate technology for electric vehicles. Engineering simulation is used to perform a fully-coupled conjugate heat transfer analysis of a cold plate for dynamic thermal management. Furthermore, using an advanced Subsonic CFD solver, a design study is performed for evaluating pressure-flow characteristics across the cold plate flow channel. Parallel simulations in the cloud are used for scenario analysis both for geometric variants and multiple coolant flow rates. In this sample case, our simulation workflows show users how to set up and run a complete heat transfer and flow analysis of a cold plate, including pressure drop and temperature at various coolant flow rates. Engineers can follow this example to learn how to quickly complete a parametric design study in SimScale and answer key design questions. 

Reaching Optimum Battery Reliability with CFD and Heat Transfer

SimScale has many analysis types available depending on the application. In this example, two analysis types have been used to analyze battery cold plate optimization for electric vehicle batteries. The Subsonic analysis type in SimScale produces an automated and robust hexahedral cell mesh, using the body-fitted Cartesian meshing technique that significantly reduces mesh generation times by an order of magnitude. The highly parallelized meshing algorithm gives a higher quality mesh requiring much fewer cells to attain comparable accuracy to traditional discretization schemes. This leads to faster convergence and hence, faster simulations.

The Subsonic analysis type is used to simulate both incompressible and compressible flow, with turbulence modeled using the RANS equations and the k-epsilon turbulence model. A powerful feature of this analysis type is the built-in parametric capability for defining velocity inlet boundary conditions. At the simulation setup stage, users can define multiple inlet flow rates at once, that are then simulated simultaneously. The conjugate heat transfer (CHT) analysis type in SimScale enables heat transfer analysis between solid and fluid domains. Typical applications of CHT analysis type include analysis of heat exchangers, cooling of electronic equipment and electronics enclosures, and LED luminaire design. A fluid/solid mesh is required for a CHT simulation with clear definitions of the fluid and solid interfaces, also referred to as contacts. The mesh is automatically generated in SimScale, with local refinement available for advanced users. In the case of the battery cold plate, the subsonic analysis is used to simulate coolant flow through the cold plate channel to ascertain optimal flow rates with regard to pressure drop. The CHT solver is used to demonstrate heat exchanger effectiveness and thermal performance.

Improving Battery Cold Plate Design

The purpose of this study is to maximize the dynamic thermal management of a new battery cold plate design, used to cool electric vehicle battery packs. We will consider two versions of a cold plate heat exchanger. The original design (V1) is a common component widely used in industry and has a single serpentine cooling channel. The design is known to cause hotspots that lead to battery longevity issues. A thermal engineer has come up with a new design (V2) that uses a labyrinth cooling channel with a much-increased surface area for thermal energy exchange. This is an early-stage proof of concept design that requires simulations to validate its predicted performance and compare it with the original version. 

Using SimScale, an engineer can quickly and accurately assess both designs. Both battery cold plate designs have one inlet and one outlet. A velocity inlet is used to define the mass flow rate of the coolant (water) which is the primary heat removal mechanism for extracting heat from the batteries. The battery cold plate is assigned Aluminum from the SimScale materials library, water for the coolant, and a power source is applied to represent heat generated by the battery. When importing the V2 CAD file, associativity between cad files is applied automatically in SimScale, maintaining naming conventions for parts/faces from the V1 CAD model. This means that when swapping CAD files for comparative studies, users don’t have to reassign boundary conditions, mesh settings, or result control outputs, making comparing two or more CAD variants of a single product much faster. 

The heat transfer analysis shows that despite a larger contact surface area in V2, thermal results are much worse compared to the original V1 design, as V2 exhibits higher temperatures across the battery cold plate. The V2 design needs further simulation and optimization. To help resolve these counterintuitive results, the design engineer analyzes the flow in isolation.

Results of the Subsonic fluid flow analysis that takes just four minutes to solve, yields further insight. The flow in V1 maintains a steady flow of 3.5 m/s throughout the single channel. In the V2 design, highly varied coolant velocity is seen throughout the labyrinth channels. In many areas, almost stagnant flow is observed in the areas of peak temperature, hence the cooling channels are not effectively removing heat from those areas which leads to poorer thermal performance. We can use the fluid flow analysis to look purely at flow velocity to better understand pressure drop across the cold plate. The converged pressure drop across any two points can be extracted from SimScale using the results control menu. In this case, for a constant coolant flow rate of 0.13 kg/s, V1 has a 292 KPa pressure drop between inlet and outlet and, V2 is ten times less at 22.7 KPa. This might have some advantages if the thermal performance of V2 was comparable to V1, such as less pump power and hence energy needed to operate the pump. First, we need to try and increase its thermal performance, however. Since we have enough pressure drop to play with, we might try and increase the coolant flow through V2.

Using the parametric inputs in the Subsonic analysis type, we can specify multiple flow rates in the inlet boundary condition to run in parallel. Simscale will recognize that we have a parameterized study setup and all runs (all flow rates) are simulated in parallel in the cloud on SimScale servers. This frees up your local PC/laptop and does not consume machine resources. SimScale will send users an email notification when all the runs have been completed successfully. 

A higher mass flow rate of 0.3 kg/s gives increased flow velocities in V2 and improved thermal performance. The pressure drop across the cooling plate has increased considerably but is still 60% lower than the original V1 design. The new V2 cooling plate design has achieved a lower average temperature, although less uniform, with a higher flow rate but a much lower pressure drop than V1. Additional design iterations would be necessary to improve the temperature distribution and battery reliability further in V2. Engineers looking to optimize this battery cooling plate can parametrize the CAD model, coolant flow rate, and inlet temperatures for more detailed design studies. Swapping materials is also simple by using the extensive SimScale materials library.

Heat Transfer Simulation in the Cloud

Access to simulation in the early product design stages is essential to achieving the required thermal performance of battery cold plate designs, all while reducing physical prototyping costs. With the cloud-native SimScale platform engineers and designers can harness advanced solvers that account for coupled analyses of conduction, convection, and radiation to deliver accurate results across a variety of problem domains and scales. 

Watch our on-demand demo, Cooling Performance Optimization of Electric Vehicle Batteries, to see how you can quickly complete a parametric design study and run a complete thermal analysis of a cooling plate including pressure drop and temperature at various coolant flow rates:

The post Simulating and Optimizing an Electric Vehicle Battery Cold Plate appeared first on SimScale.

This article describes the structural assessment and vibration analysis of an electric motor support bracket to calculate and verify if its critical response frequencies lie within the intended operating range. In addition to the vibration analysis, we show engineers how to perform a load analysis on the electric motor shaft under an applied torque. The design goal of this simulation study is to optimize the electric motor support bracket to ensure that its eigenfrequencies remain outside of the motors’ operating speeds, avoiding possible part damage, bolt loosening, and reduction in clearances between parts and unwanted noise. In many cases, there is a danger that resonance effects might excite vibrations in the support bracket or other components. Meticulously analyzing this uncertainty is needed to reduce the risk of unwanted issues and, simulation at the early stages provides this guarantee. Geometry changes are made to mitigate dangerous vibration at shaft speeds close to the calculated first eigenfrequencies of the support bracket. Note, that the motor in question has a fixed rotational speed. Vibration analysis is even more important when variable speed drive motors are used for example and more in-depth studies are needed in those cases. We can also perform a quick safety factor check of the motor shaft under an applied torque to represent real-world loads and ensure a minimum safety factor of two is validated under normal operating conditions. A factor of two is a good starting point as it means the stresses are half of the yield stress of the shaft material (mild steel), which is getting close to the material endurance and performance limits that affect fatigue and component life. The entire engineering simulation workflow is implemented via a web browser and performed in the cloud.

Fast and Accurate Structural Simulation

The FEA capabilities in SimScale enable engineers to virtually test and predict the behavior of structures and components, and to solve complex engineering problems subjected to static and dynamic loading conditions. SimScale uses scalable numerical methods that can calculate physical variables otherwise very challenging due to complex loading, geometries, or material properties. Another structural analysis software feature, modal (frequency) analysis can help determine the eigenfrequencies (eigenvalues) and eigenmodes (mode shapes) of a structure due to vibration. The results are important parameters to understand and simulate structures and products that are subject to unsteady conditions. The resulting frequencies and deformation modes are dependent on the geometry and material distribution of the structure, with or without the displacement constraints. A static analysis type allows time-invariant calculation of displacements, stresses, and strains in one or multiple solid bodies and the results are a consequence of the applied constraints and loads. SimScale has integrated the Code_Aster solver into the platform for frequency and static analysis simulations. A much broader study can use a multidisciplinary physics approach to simulate and optimize the loading, modal, thermal, and rotating aspects of the electric motor. In this example, we will focus on the vibration analysis of the electric motor support bracket and load stresses on the motor shaft.

Vibration Analysis of an Electric Motor Support Bracket

A simple workflow is required to complete the analysis for both the support bracket and shaft.

Import CAD – Users can import many types of common CAD file formats and use CAD connectors with tools such as Onshape, Solidworks, AutoCAD, and more. The feature CAD mode in SimScale allows users to perform basic CAD operations for editing without leaving the platform. The motor geometry is imported from Onshape. The number of faces increases with support bracket geometry modifications with a total of six bracket CAD variants defined. The motor shaft is also a single volume. 

Simulation set up & mesh for the support bracket – The bracket has a frequency analysis type selected in SimScale. The mesh is generated automatically with 23.5K cells and the element sizing is set to automatic which takes the geometry fineness into account when deciding size. In this case, a moderately fine mesh has been chosen (the user has manual control over mesh settings). A point mass boundary condition is applied to represent the mass/moments of inertia of the whole motor eliminating the need to include the fully detailed motor geometry in the simulation. Our goal is to find the first eigenmode of the support bracket that needs the least energy and is thus the easiest mode to excite. An eigenfrequency plot is generated automatically.  

Simulation set up and mesh for the shaft  – The motor shaft has a static (linear) analysis applied to it and from the materials library, we have selected mild steel as the primary material with linear elastic behavior. In a static analysis, we can define constraints and loads. The shaft has fixed support connected to the motor and the main shaft can be deformed due to remote load forces that represent applied torque using moments around the center of rotation. Deformable here means that the shaft is allowed to deform without applying additional stiffness. Undeformable would mean the shaft was a rigid entity which is not what we want. A centrifugal force is applied to the whole volume of the shaft and a rotational speed is set in radians/second. Automatic mesh settings are again used and a second-order tetrahedral mesh is generated for the shaft with 170K cells.

Post-processing – The SimScale platform’s field results allow us to visualize the deformed shape for each mode showing the displacement magnitude in meters. Statistical data is available to further interrogate the eigenmode number, eigenfrequency, Modal Effective Mass (MEM), Normalized Modal Effective Mass, and cumulative Normalized Modal Effective Mass (CNME). Multiple plots are pre-populated showing the eigenfrequency plot, modal effective mass, and accumulated normalized modal effective mass. From the static analysis, we can evaluate standard outputs such as Von Mises and Cauchy stresses, displacement, strain, and a new output variable that is the safety factor.

Design Insights

Bracket Vibration Analysis

The support bracket vibration assessment shows us that the first eigenfrequency (21.8 Hz) of the original bracket (CAD 1) design is identified as being dangerous as it is too close to the shaft rotation speed equivalent in Hz (19 Hz), causing dangerous resonance effects and possibly leading to severe deformation. This could easily excite the first eigenmode just by running the motor and the excitation could cause resonance in the bracket itself —the displacement would move the motor beyond operating parameters and could cause damage. Although the first eigenmode is the most important to reconcile, different eigenmodes cause more displacement and deformation. They can all be visualized in SimScale. Geometry optimization by defining several CAD variants (1-6) of the support bracket can be simulated in parallel to achieve successful inflation of the first eigenfrequency value, away from the shaft speed. The full results are shown in the table below:

Shaft Safety Factor

The shaft structural integrity is simulated under real-world operating conditions, ensuring that the factor of safety did not drop below 2. Under two loading conditions with a nominal torque of 0.5 Nm and 10 Nm respectively we found the following:

1. Load Case 1 – Nominal Torque of 0.5 Nm
   1. Results show a peak stress value 4.01 MPa.
   2. Results show a minimum factor of safety of 62.35, achieving the design goal.
2. Load Case 2 – Maximum Torque of 10 Nm
   1. Results show a peak stress value 80.19 MPa. 
   2. Results show a minimum factor of safety of 3.118, achieving the design goal.

Engineers must embed structural simulation in the cloud early in their product development processes to ensure that performance of critical design parameters is met using sound engineering analysis. Using parametric analysis, engineers can increase their confidence that the final product will operate and function within expected limits.

The post Vibration Analysis of an Electric Motor Bracket appeared first on SimScale.