Case Study: Electric Vehicle Battery

In this case study, battery design is investigated both in terms of thermal management and mechanical performance. This project requires that we assess a battery pack assembly, which contains a couple of cooling plates, pouch cells, insulator pads, and gap filler material. 

Geometry & Problem

For this project, we used a 3D solid body geometry. Any tool can be used to generate such a solid body model and then can be uploaded to SimScale. 

Simulation Model

After CAD import, simulations require meshing. In this case study, we used automatic meshing.

Physics

In this case study, we conducted a thermomechanical analysis, which allows us to bring thermal stresses into a conduction-based simulation.

Solve

Our cloud-native solution gives us the ability to identify the temperature hot spots as well as the peak mechanical stresses and deformations to understand how our module is performing both thermally, as well as mechanically.

EV Battery Pack Gap Fillers: Project Scope

The model is a lithium-ion battery pack for electric vehicles that consists of four modules. We have two water cooling plates on the top and bottom of the battery module with the gap filler material separating the cooling plates from the pouch cells and the casing itself. 

Gap fillers are critical for thermal management for a number of reasons:

• They provide electric insulation
• They provide vibration damping 
• They have the ability to conduct the heat across the interface between the battery packs and the cooling plates

We want to make sure we utilize the maximum cooling capacity of our cooling plates by selecting the most efficient gap filler material for the model. Additionally, we want to focus not only on optimizing the gap filler selection but also on determining the best thickness of the gap filler.

Gap Filler Materials

The first step of the study is to simulations with three different gap filler materials, each with the same thickness. Each material has a different conductivity value. We want to investigate the effect of using different gap-filling materials, in terms of the maximum cell temperatures that are observed inside the pouch cells. Additionally, we want to detect the amount of mechanical swelling that can take place due to the temperature in the pouch cells. Increased conductivity results in reduced maximum cell temperature, however, change in conductivity had a minor effect on our study, since the dominant impact on reducing temperature was provided by the water cooling plates. Even these minor effects are valuable to note in the early-design stage, however, as the material selection impacts cost in manufacturing.

If we have the optimal selection of gap filler material, we can select the most efficient or least energy-intensive cooling apparatus on the cooling side of the battery cells. Also, the pouch cell principal strain decreases, which is an indication of the actual volumetric swelling of those cells. We can see that we have a higher peak temperature on the left-hand side where we have the gap filler material with low conductivity; when we go over to a high conductivity material, we have both lower temperatures and more uniform temperature distribution as well, critical for battery design. We want to make sure that our cooling capacity is being utilized across the whole module uniformly to avoid cold spots and hot spots. In another cutting plane on the gap filler materials, we see the same trend. 

Gap Filler Thickness

We can use SimScale to optimize the thickness of the battery pack gap filler material, as well. Simulation can help us understand the bearing that gap filler thickness has on the actual temperature distributions and thus predict and manage the mechanical thermal strains within the system. As we increase the thickness of the gap filler material, we see the reverse effect. With a thicker gap filler material, we have a higher pouch cell temperature. That is because we are reducing the effectiveness of the water cooling plates. We are trying to keep the gap filler material as thin as possible, but while still allowing the gap filler material to serve for vibration damping and insulation, as well. Simulation gives engineers the ability to navigate this balancing act. With thermomechanical analysis, we can investigate these physical behaviors and understand what is causing the hot spots, swelling, etc. and test the best way to mitigate them.

Mechanical Performance

Finally, we can assess the mechanical performance from the same simulation. Looking at the lowest conductivity gap filler material set at the lowest thickness, we can see the actual mechanical stress in the steel casing around the battery model itself. Here, we observe maximum mechanical stress to the point of yielding at that joint section. So, in those regions where we have significant stress, we could expect to see some damage and plastic deformation. This asks us to consider resizing the casing thickness or maybe use a more ductile material for the casing, in order to reduce those stresses and increase the lifetime of the product. If we want to go into further detail, beyond a linear static analysis that identifies the areas above yield stress, we can convert this into a non-linear simulation. This will allow us to see the absolute maximum stresses that we would experience during the thermal cycling of the battery module. The same goes for the total strain in the pouch cell. We have roughly 1 to 2 percent of total strain magnitude in the pouch cells due to thermal behavior.

Maintaining battery packs within a specific temperature range is essential for ensuring the performance of the product and the safety of users. With cloud-native engineering simulation, you have the ability to run multiple design iterations in parallel means that simulating a battery pack module from CAD preparation to post-processing can be performed with fast turnaround times.

The post EV Battery Pack Gap Fillers: A Thermomechanical Simulation Study appeared first on SimScale.

Electronics Enclosure with Snap-Fitting Cover

The model in this case study is an electronics enclosure with a snap-fitting cover. For these types of enclosures, it is very beneficial to conduct a structural analysis early in the design process to optimize the snapping operation. To gather quality design insights, the outputs of interest from the simulations were peak stress regions which are likely to cause permanent deformation and breakage and also snapping kinematics of the snapping operation itself. Performing a trend analysis facilitated the selection of an appropriate snap and support design.

Cloud-Native Simulation Workflow

The simulation workflow in SimScale, which can be repeated and applied to many different use-cases, starts by uploading a solid body CAD geometry to the platform. By using automatic body meshing, the model is quickly ready for simulation. Though the geometry of this case study was relatively uncomplicated, the physics used within the structural analysis is complex. With SimScale, capturing valuable insights from complex simulations is simpler and easier to share within teams and organizations, even with varying levels of simulation expertise. Below, the workflow for a nonlinear static analysis is represented. As SimScale facilitates a cloud-driven design study, users can leverage parallel computation and solve both a higher number of design iterations and more iterations of increased complexity.

Nonlinear Static Analysis in the Cloud

To understand the snapping kinematics, a quick animation can be created when post-processing the results. With the help of the animation, the movement of the casing can be better understood and the regions where the stress value has built up above the yield stress can be identified. This offers an opportunity to further optimize the design by changing the shape or using another material to minimize stress. 

After acquiring the results of the first simulation, the next step was to run a few more iterations. Based on the first design results, changes and alterations could be made within the geometry to converge upon a better design candidate. The first design change enacted in this study was deleting one of the faces, and creating a filet instead of a sharp edge. As the CAD changes are done in Onshape, a cloud-based tool, there is no need to download the file from Onshape and then upload it to SimScale—all can be transferred with cloud integration between two platforms. The previous simulation template can be applied to the new geometry exactly as done in the previous step, requiring no reassessment of the physical constraints or the topological entities. They are already automatically reassociated with the new CAD model. 

A further variable to experiment with in order to optimize design is testing different materials. This is easily done by selecting a new material from the materials library in the simulation setup and assigning this material to the lid. In a similar manner, many different design strategies can be tried and further improved. Once the first simulation setup is completed, iterating on top of that is straightforward and fast, with the power of the parallel computation. 

Electronics Enclosure Design Insights

After performing the first simulation on the design provided by the CAD engineer, the regions above the yield stress were clearly identified. Another interesting point detected was the fact that the support structure underneath the snap is not carrying any stress. As it does not provide added benefits to the structure, designers further assessed its significance in terms of manufacturing. In the second design, the snap is located without support underneath. The same result as with the first design is derived, proving that some cost could be saved in terms of manufacturing by removing the non-beneficial support element. And, in the last design, the shape of the snap is changed slightly, and also the sharp edge is rounded at the bottom part to have a smoother snapping operation.

Even if the above-yield stress observed on the model was reduced, an overall significant impact was not shown. Here, designers might consider material changes, in addition to shape iterations. Apec, Makrolon 8345, and Stanyl TE300 were tested as alternatives for the lid.

Because Makrolon 8345 is very stiff, it created high stresses and was eliminated as a viable option for this design. Stanyl TE300, on the other hand, produced strong results, significantly reducing the yielded areas.

As designers decide on the best shape and material for a model, prototyping or final validation analysis are a natural next step. In this case study, we included a validation analysis scenario. In the validation step, the CAE engineer might prefer to increase the complexity by checking how much deformation they will end up with by using a nonlinear material model. This can be accomplished by uploading a fully detailed stress-strain curve of the experimental testing of the material. Absolute peak stresses, as well as permanent plastic deformations, can be observed and measures taken to ensure that the single snap-fitting operation will be safe. Additionally, a mesh independence study can be conducted on top of the automatic meshing settings assigned by default to validate mesh independence on the results. 

Nonlinear Static Analysis in the Cloud

Engineering simulation in the cloud gives mechanical and structural engineers more detailed insight compared to physical testing, which is critical during the early stages of design exploration. This case study shows how nonlinear static analysis in early-stage design allowed three different snap-support design candidates to be tested, along with material selection. The accessibility of cloud-native engineering simulation enables designers and engineering teams to leverage parallel computation capabilities and achieve faster design cycles and more robust design insights.

The post Nonlinear Static Analysis: Snap-Fit Assembly appeared first on SimScale.

Below, we cover two case studies that demonstrate what we’ve learned from over two years of rapid ventilation design assessments. These techniques enable architects and engineers to quickly and accurately predict air movement and air quality, determine HVAC sizing, and assess ventilation strategies.

Case Study 1: Classroom Ventilation Strategy 

In this case study, the aim was to assess the heating and ventilation strategy of a classroom and to ensure that the design complied with the Passivhaus environmental and energy metrics. In various HVAC systems, CFD can be used to determine which parameters will have the largest impact in relation to improving indoor air quality and living conditions. Some of these parameters within this case were U-values for different parts of the room (due to the different materials being used), radiation sources to represent solar gains, occupants, and additional insights the team wanted, such as the quality of air.

Once the base design was simulated, the architects faced two challenges: Firstly, due to the inlet, the air was drafted on top of the occupants in a rough manner, which would create a significant amount of thermal comfort disturbance. And secondly, existing air within the classroom was observed to be of low quality with high CO2 levels. Adjustments would be required to make the design more efficient, with increased air quality and better thermal comfort, but without additional energy loss.

In order to solve these issues, architects came up with different designs to improve the air supply strategies and validate cases. They ran different configurations in parallel to converge on the optimal solution. For each of the design variations, the team compared temperature distribution within the space and assessed air quality by means of CO2 concentration inside the room.

In the first design change of the base case, different HVAC inlet configurations were tested. By utilizing the guide vanes and having a high wall diffuser, inlet air was pushed towards the ceiling. As the air hit the wall it allowed for circulation inside the room and solved the first issue of occupant discomfort.

In the second design change of the base case, the designers focused on solving the low air quality issue inside the room. In order to reduce the amount of CO2, the high wall diffuser setup was combined with various window opening scenarios to measure the impact on ventilation and indoor air quality. In the end, the combination of a top-hung window opening and a high wall diffuser setup was found to increase air quality and reduce CO2 by improving the flow pattern within the room. This scenario satisfied both air quality and thermal comfort criteria.

Case Study 2: External Wind Conditions and Ventilation

The goal of this case study was to assess the effect of external wind conditions on ventilation strategies. By taking into account wind conditions, designers can ensure their study is as close to real-world scenarios as possible. The goal here was to assess ventilation in every room of an apartment building when external wind impacts are considered.

The process to derive the internal conditions for this particular case was started by modeling the building together with its surroundings in a Pedestrian Wind Comfort (PWC) analysis. With a PWC analysis, we are interested in assessing the wind comfort criteria for various wind directions. For this internal case study, the interest was in one particular direction so that it could be applied directly to the building. We focused specifically on observing how the wind from the southwest direction would affect the third floor of the building. Pressure tappings were applied from the external wind study to the windows and, to simulate a life-like scenario as much as possible, internal door leakages were also considered within the setup.

As the base design was simulated, engineers focused on answering two main questions. First, whether the natural ventilation of one window open was sufficient for the building or if forced ventilation was needed. And secondly, if CFD could help in deciding the correct ventilation strategy.

After getting the results of the base case setup, different scenarios with wider window openings were simulated to compare the mean age of air within the room. By increasing the opening angle of the window, a one-sided improvement in the ventilation was observed. But after some duration, no matter how much the window was opened there were no more significant changes in the ventilation inside the building.

Mean age of air plots can give us a good understanding of where the air is stagnant and where there is a good amount of ventilation inside of a building. As the plots were checked, a pattern was observed. Rooms on the corners and the kitchen were well ventilated, whereas the rooms in the middle were not well ventilated.

In light of the results obtained from the  CFD analysis, it became easier to focus on potential reasons for poor ventilation in some rooms and solutions on how to improve it. For example, even though rooms six and seven looked very similar to each other, the conditions of airflow and ventilation were not the same. This was due to prevailing wind conditions.

The animation visualizing the wind results shows a large amount of wind coming from the right-hand side and hitting the side of the building where the kitchen is located. This helped with both ventilation in the kitchen and with the cross-ventilation, as a high degree crossflow was entering from the window and then going out of the other side. On the other side of the building, the air was creating a recirculation zone within rooms five, seven, and eight. This was causing single-sided ventilation with air entering and leaving if the other window was also open. This explained the reason why even though rooms six and seven were similar in structure, room six was not as well ventilated. Room six lay in a region where the recirculation ended and also the interaction between the wind effects from the surrounding buildings started. There was a dead zone created in front of the building and that caused insufficient air to pass through.

In the end, CFD studies determined that natural convection was not sufficient for all rooms of this building. In most cases, opening the windows increased the ventilation, but for the corridor and some of the rooms, like room six, additional forced ventilation is needed to satisfy the air quality standards.

Cloud-Native Simulation for Ventilation Strategies

As shown in these two case studies, CFD analysis is necessary for understanding and predicting the effectiveness of natural and forced ventilation.  As a next step, CFD analysis can even inform design decisions on the best sizing for HVAC equipment for a particular building or room. This not only helps avoid undersizing or oversizing HVAC equipment but also ensures proper ventilation, thermal comfort, and indoor air quality while optimizing designs for less energy loss. 

Cloud-native CFD analysis enables engineers to solve for internal and external flows, study indoor and outdoor thermal comfort, and scale HVAC device-level simulation results from room-level to building-level and beyond.

The post Ventilation Strategies: Tested with Simulation appeared first on SimScale.

With the growing interest in electric vehicles, competition in the market—and the resulting demand on product performance—have drastically increased. The need to quickly improve highly-efficient electric car components requires a tool that allows engineers to simulate multiphysics early, reduce the need for physical prototyping and costly late-stage design changes. With cloud-native simulation, engineers have access to testing many scenarios in a simpler workflow with shorter turnaround times. In this article, we explore a multiphysics investigation of an EV inverter. Three different physics simulations are performed including a pressure drop analysis, a Conjugate Heat Transfer (CHT) analysis to validate that all components operate in a safe temperature range, and a vibration study—three different physics, one web application, simulated in under 30 minutes. 

EV Inverter: One Model, Three Physics Simulated

To achieve better thermal performance and structural integrity of an EV inverter three simulations are run using the same EV inverter model. Each simulation defines a different analysis but follows the same workflow. Depending on the placement of this equipment, different physical factors need to be taken into account. Oftentimes they are set in vehicles, hence vibrations become a concern. The size of the pump is also essential to ensure the pressure drop across the flow channel is minimized. And lastly, as different electronics components are mounted and operated simultaneously within the EV inverter, temperature and power should be safely managed, as well, to ensure liquid within the device is kept cool. 

Case Study 1: Thermal Management

The first simulation was a thermal management analysis of the module by using SimScale’s CHT solver. This study was used to ensure that all components operate in a safe temperature range. By performing a full-fledged CHT analysis, convective cooling through the flow channel was simulated, followed by conduction cooling within the solid. Temperature distribution both within the fluid and solid parts were of concern, so the CAD model was used together with the flow volume during the CHT simulation. As expected, the electronics components had high temperatures. By visualizing the streamlines through the water channel, the temperature distribution within the flow was also observed. The goal of running the CHT analysis was to assess the quantitative results on the critical components of the EV inverter and determine whether it was operating under safe conditions. If not, the analysis would inform design changes to ensure an effective and safe operation by keeping the temperature within certain limits.

Case Study 2: Pressure Drop Analysis

The second simulation was an incompressible pressure drop analysis to predict the fluid flow inside the flow channel and reduce the overall pressure drop. In order to run pressure drop analysis, the flow channel should be isolated as a separate part. SimScale provides all tools for an end-to-end simulation platform. Prior to setting up the physics of the simulation, the flow volume was extracted by using SimScale’s CAD environment. To assess the pressure drop across the channel an incompressible CFD analysis was performed. Based on the results, the critical points could be determined and changes in the design could be enacted. With the changes in the design, new simulations are run to achieve a much more efficient overall product performance. 

Case Study 3: Vibration Analysis

The third and last simulation was a vibration analysis using SimScale’s frequency analysis tool. This simulation helps to identify the first 20 eigenfrequencies and eigenmodes to prevent unwanted vibrations. The resulting eigenfrequencies and eigenmodes would help the designer evaluate the overall rigidity of the model, and assess whether there might be eigenmodes affecting the operation of the device. Knowing that this piece of equipment would be mounted on a vehicle, eigenfrequency analysis would be needed to ensure that vibrations are not triggered by the operating conditions of the device. As the simulation is run, the physical deformation of the device can be visualized by stepping through each of the eigenfrequencies. The resulting frequencies and modes are dependent on the geometry and material distribution of the structure. Therefore, by performing vibration analysis for different geometry alternatives, the best design can be picked. Thus, unwanted vibrations can be prevented. 

Digital Prototype of an EV Inverter

Traditional product development processes require lab testing and physical prototyping, both of which often reveal design flaws that necessitate further redesign. It’s a multi-step process that can stretch the development schedule indefinitely. By adopting digital prototyping, users can eliminate the unnecessary loops from the design optimization process at an early stage, cutting both time and costs of development as a result.

The introduction of cloud-native multiphysics simulation into the R&D process has empowered engineers to accurately and efficiently simulate hundreds of simulations, including a wide range of complex physical phenomena. As SimScale was born in the cloud, it’s scalable by nature, meaning computational power can scale up or down depending on project demand and multiple design options can be explored across multiple teams. This gives the design team an edge over traditional CAE tools. Providing designers and engineers access to multiphysics simulation allows teams all over the world to push the boundaries of design.

Get a deep dive in this on-demand demo, where we walk you through simulating the three different physics discussed in our EV inverter case:

Watch Demo On-Demand

The post Multiphysics Simulation of an EV Inverter appeared first on SimScale.

Case Study: Optimizing Design for Improved Performance

In this project, SimScale, Onshape, and Datadvance collaborated to optimize the geometry of an insulated-gate bipolar transistor (IGBT) cooling plate design. The objective of this study was to minimize the average surface temperature for maximum thermodynamic efficiency while also minimizing the pressure drop or drag power on the cooling fluid in order to increase the fluid dynamic efficiency of the cooling plate. The project serves as a model for how shape optimization can be used in conjunction with thermo-fluid dynamics to optimize the design.

The cooling plate consists of a milled aluminum block with vertical pins to increase the surface area between the cooling fluid and the IGBT hot surface. The geometric variables were selected to ensure that any final design could be manufactured using approximately the same tooling and manufacturing methods:

• Pin diameter: 2 mm – 6 mm
• Pin row offset: 0 – 0.5 row width
• Pin rows in X: 6, 7 or 8
• Pin rows in Y: 3, 4 or 5

The flow rate was varied in an achievable range for the cooling pump:

• Flow rate: 1.2 l/min – 3 l/min 

The IGBT geometry was modeled in Onshape and the geometric variables were parameterized so that they could be driven by pSeven, Datadvance’s optimization software through the Onshape Python Application Programming Interface (API).

If a single design is investigated, the workflow begins by generating the geometry in Onshape, with a specified parameter set. For that specific design, Computational Fluid Dynamics (CFD) simulation is conducted in the cloud using  SimScale. As relevant quantitative results are extracted from the SimScale simulation, the next step is to set up a surrogate-based optimization in pSeven. The parameters which need to be altered to minimize temperature and pressure drop, hence optimizing the design, can be intelligently decided. Instead of following this procedure one by one for each parameter variation, a feedback loop for the CAD parameters between the optimized design obtained from pSeven and Onshape within this project was established. The API was used to automate the parameter variation and provide programmatic access to preprocessing, simulating, and postprocessing.

Cloud-Native Design Space Exploration

API gives analysts the ability to automate the workflows, explore parameter spaces automatically, and optimize the shape by connecting to the CAD authoring and optimization tools. Within this parameterization study, the API connected the geometric optimization tool to Onshape’s cloud-based design tool and to SimScale’s cloud-native multiphysics simulation engine. pSeven Enterprise was used to build and run the optimization workflow. Each time a particular geometry and simulation parameters were sent to the API, SimScale simulated the model and returned outputs for that instance, including temperature distribution, pressure, and efficiency. Data was then fed into the pSeven surrogate optimization model. As the workflow is scalable, many hundreds of geometric scenarios can be modeled. The physics outputs for each scenario were tracked using a Pareto front to converge on the local minima (the optimized solution) where the average surface temperature and the pressure drop were the key parameters.

The results of the workflow showed that the highest pin density, with pin offset, gave the minimum surface temperature. However, the minimum pressure drop was obtained with the lowest diameter pins and no pin offset. With this set of results, and a maximum surface temperature design target, the design team would be able to select the design that would have the greatest pump efficiency while achieving the required cooling properties.

“Serverless” Design Studies

This project demonstrated how end-to-end design space exploration can be fully performed by using only cloud-native components. By removing the problems caused by the complexity of setup and maintenance of the required software and hardware stack, full design space exploration can be achieved, even in a  ‘serverless’ workflow. Cloud-native components also enable the use of optimization in scenarios and organizations, where previously it was technically and economically not feasible, creating a fully accessible, fully realized design space exploration in new applications.

Learn how connecting Datadvance’s cloud-native low code platform pSeven Enterprise to SimScale’s multiphysics cloud simulation engine using API allows a drastic speedup of simulation and optimization procedures in this on-demand webinar:

Watch Now

The post Design Space Exploration and Performance Optimization in the Cloud appeared first on SimScale.