The Impact of AI on Engineering

As AI continues to evolve, it is likely to have a profound impact on engineering processes and practices. Many of the boring and repetitive jobs could be replaced by AI workflows. This is because the real benefit of AI is the speed at which results can be generated. However, it is still essential to review, check, and improve the data that is produced. This is why I do not see it as a threat to engineers, yet. As I was writing this blog I tried using ChatGPT to write this article and it was interesting to see how AI would write about AI. The results were disappointing, to be honest, and although it gave me some ideas of the topics to cover, I opted to not use the majority of the text that was generated. Much of the text that was produced used too many adjectives, see the example quotation:

“At the forefront of this transformation is SimScale, a pioneer in cloud-based simulation solutions, in collaboration with NAVASTO, their esteemed partner, offering ML-based models that propel engineering simulation to new heights.”

ChatGPT 

Engineers are however going to have to adapt and adjust to the changing environment or risk falling behind their peers and their competitors. There will be an increased importance for continued learning and experimentation with AI-based tools to ensure that the workforce is up to date with the latest developments. Engineers should not be threatened by AI, as if it is employed carefully and correctly it has the potential to make engineers a lot more effective.

A Step Change in Engineering Design and Simulation

The simulation industry has already been going through a state of rapid change and development with the introduction of truly cloud-based engineering simulation like SimScale. However, AI and ML models have the potential to further change and evolve how simulation is used in the product design cycle (McKinsey).

Traditionally, physics-based simulation has been used late in the design cycle to validate designs and potentially reduce the reliance on physical prototyping. However, in a recent report by McKinsey, they highlight from their surveys of R&D leaders in engineering that the business case for using engineering simulation is moving more towards faster-time-to-market and reduced product cost.

Consider an example of a project to design a water pump to a specific specification for a client:

• Two-month window for pump design.
• First month: Design and CAD geometry based on experience and company precedent.
• Second month: If the CFD team has capacity they simulate the design and determine its performance against the specification.
• There may be a couple of feedback looks required for CAD cleanup operations, etc.
• This leaves limited time in the final week or two for design improvements (one iteration) based on simulation results.

Proposed alternative with AI-based ML model:

• Designer uses existing data for AI prediction of pump performance.
• Early simulation-informed design changes are possible as soon as the designer has a concept CAD model available.
• This opens up the possibility for generative engineering workflows that automatically explore design spades or use automated design approaches.
• Shorter design process duration, reducing project overrun risk and late delivery risk.
• Allows for final simulation validation of the design in a shorter or equivalent time scale.

Predictive AI and Simulation

Predictive AI works by collecting relevant data, using algorithms, and training neural networks to build machine learning models that can predict the outcome of an event or scenario that was not observed in the original data.

For engineering simulation: GNN (Graph Neural Networks) are a powerful method. This is because they can use structured node-based data that is very similar in principle to how a mesh works for engineering simulation with FEA or CFD. However, unlike with engineering simulation, the results that are obtained from a GNN prediction of a simulation are obtained in seconds as opposed to hours with incredible degrees of accuracy.

There are of course challenges associated with predictive AI. The quality of the results that are produced are highly dependent on the quality of the input data that is used for the training. For many organizations, data collection and sorting is one of the biggest challenges when it comes to predictive AI models.

In a similar way to how OpenAI made AI accessible to the masses with ChatGPT, SimScale has partnered with NAVASTO to bring AI-powered simulation predictions to the engineering community without the need to worry about data management. See the release webinar here.

The post AI and the New Era of Engineering Simulation appeared first on SimScale.

Key Takeaways

1. Cloud-native simulation is leading the “new world” of engineering simulation by eliminating inefficiencies.
2. Purchasing and deploying your simulation tool has never been easier thanks to cloud-native simulation’s accessibility and immediate availability.
3. The cloud-native world enables customizable, online, and on-demand training for new users.
4. No restrictions on who can simulate – SimScale enables the democratization of simulation access, leading to a more diverse, efficient, and enriched design process.
5. In cloud-native simulation, the audit process is streamlined, with a transparent and easily accessible record of who took what action and when.

New World vs Old World: Redefining Efficiency

In contrast to the limitations of the past, cloud-native simulation like SimScale empowers engineers to innovate faster and navigate a streamlined and collaborative engineering design space. One way of looking at this is considering the analogy of a gardening hose.

A twisted hose hampers the flow of water and represents the inefficiencies of the “old world” of simulation. Even when this is untangled, a new kink will appear elsewhere, and without transparency, it’s hard to find where this inefficiency is. The free-flowing spray gun represents the “new world” of simulation, where the inefficiencies are gone, and the flow of water is not only streamlined but also in control, illustrating how cloud-native simulation can increase both efficiency and innovation.

In this article, I will show you how replacing legacy simulation tools with state-of-the-art, cloud-native simulation can streamline your team’s ability to design, innovate, and analyze more efficiently and effectively. You can also see a tabulated comparison at the end of the article.

1. Streamlining Purchasing and Deployment

The process of purchasing engineering simulation tools used to be characterized by long and protracted cycles, creating a considerable delay in acquiring the necessary design insight tools. Conversely, the cloud-native world introduces a revolutionary approach, offering affordability and immediate availability at no cost to kickstart your simulation work. This transformation in the purchasing landscape signifies a shift towards efficiency and accessibility.

Similarly, deployment used to suffer from sluggishness and installation bottlenecks with traditional, on-premise simulation tools. However, the cloud-native world presents a paradigm shift with instant deployment that negates the need for time-consuming installations.

Administrators can still wield control over access, immediately making resources available to anyone, on demand. Users can log in instantly, expediting the onboarding process, and the addition of a new user is streamlined to a simple task of entering their email address.

2. Facilitating Early Simulation Usage

Turning our attention to training and early usage, the old world demands large-scale organization for training sessions, often leading to unapproved training and the looming risk of new users making critical mistakes. On the contrary, the cloud-native world has ushered in an era of online and on-demand training, offering a flexible and customizable approach that can be seamlessly integrated into an organization. Support is not a distant concept; it’s ‘live’ and readily available when users need assistance. Time to answer is no longer measured in days or weeks but in minutes and seconds.

At SimScale, for example, the support system consists of real engineers collaborating with your team in real time, lessening the reliance on automated solutions.

3. Fostering Established Simulation Usage

Legacy simulation tools are fundamentally limited to disconnected, siloed teams, isolated projects on local machines, and data that is hidden and individually owned. Peer-to-peer learning is constrained by individual availability, restricting its effectiveness. However, once a company is fully up and running with its new, cloud-native simulation tool, the collaboration between connected teams is by default fostered, and projects are easily shared with all interested parties at the click of a button. Even in cases where projects require restricted access, there’s an option to form private teams. Administrators retain full access, eliminating the risk of valuable data being lost on a hard drive. Users can support each other within projects, and experts can offer guidance seamlessly within the validation process.

The philosophy of not restricting who can simulate is a cornerstone of the cloud-native world. By front-loading simulations early in the design process, SimScale enables everyone to experiment and learn at the initial stages, producing better designs faster than was previously possible. The democratization of simulation access ensures that input is solicited from everyone, leading to a more diverse and enriched ideation process. Templates, pre-defined by seasoned simulation engineers or SimScale’s own, can further empower new users to contribute meaningfully and confidently, knowing they are working within controlled guard rails. This inclusivity means that, with the support and guidance of experts, everyone can engage in hands-on learning and experimentation.

The transparent usage model in the cloud-native world stands in stark contrast to the old world’s opaqueness. License utilization is crystal clear, providing organizations with the ability to discern their actual needs and optimize costs accordingly. Nobody wants to pay for things they aren’t using, nor should they.

Users’ skill levels are transparently visible, facilitating timely support interventions before errors are made or time is wasted. Simulations linked to real-world projects offer insights into the value they added, allowing organizations to evaluate their tools’ and teams’ effectiveness and identify areas for improvement in subsequent design cycles.

The integration of simulation into the design and approval processes represents a significant departure from the old world’s compartmentalization. With legacy tools, simulation sits outside the process, creating a sequential flow from CAD, through simulatable CAD to results, and finally back to PLM. As a result, design reviews and approvals rely on individual simulation engineers for preparation, limiting access to crucial data. The cloud-native world, on the other hand, seamlessly brings simulation into the design process, allowing anyone to interact organically with insightful results.

4. Efficient Approvals and Audits

In the sphere of audits, legacy simulation tools grapple with questions of who did what and when and the whereabouts of critical data. This often results in team-specific tracking methods that lack standardization across the organization. In the cloud-native world, the audit process is streamlined, with a transparent and easily accessible record of who took what action and when. All data is stored in the cloud, providing flexibility in the organization based on organizational preferences.

The ability to link results from a PLM system with a URL ensures traceability, and the locking of results maintains data integrity and contributes to a comprehensive audit trail. Consistent and easily reproducible reports for each simulation speed up the time taken to interpret results and make approvals far simpler.

Summary

Join the New World with Cloud-Native Simulation

As we navigate the evolving landscape of technological advancements, the cloud-native world’s approach to holistically improving every step, from purchasing a design tool to comprehensive design audits, exemplifies a commitment to efficiency, transparency, and collaborative innovation. This transformative shift promises not only streamlined processes but also a paradigm where technology empowers users across all levels to contribute meaningfully and shape the future of their organizations.

With cloud-native simulation, blockages are removed, and the gardening hose is transformed into a powerful and efficient free-flowing spray gun. By dismantling the barriers of the old world, organizations can leverage the full potential of simulation throughout the product development lifecycle. Cloud-native simulation is not merely a technological advancement; it’s a catalyst for innovation, empowering engineers to explore, experiment, and ultimately, revolutionize product development processes. Get in touch with us below for more information on how SimScale can help you integrate cloud-native simulation into your workflow.

The post Redefining Engineering Efficiency with Cloud-Native Simulation appeared first on SimScale.

In this product update, let’s dive into the latest pivotal features introduced by SimScale in the third quarter of 2023.

1. Temperature-dependent material properties for CHT v2.0 & IBM
2. Humidity Sources Modeling
3. Visualization/Computation of Local Mean Radiant Temperature (MRT) Without Solar Load
4. Parametric Mesh Study on IBM Mesh Fineness
5. Multiphase/Subsonic Features
   • Non-Newtonian Fluids Available
   • Time-Dependent Boundary Conditions
   • Probe Points as Result Controls
6. Rotating Machinery – Blade-to-blade Flow Visualization
7. Automated Sweep Meshing for Structural Analyses
8. Orthotropic linear mechanical material properties in the cartesian coordinate system
9. Modal-Based Harmonics
   • Automatically Capture Resonant Response in Modal-based Harmonic Analysis
10. Yeoh Hyperelastic Model
11. Ogden Hyperelastic Model
12. Select Similar Shapes

1. Temperature-dependent material properties for CHT v2.0 & IBM

CHTv2/IBM simulations now allow for the definition of more advanced fluid properties through temperature-dependent tables:

• Specific heat
• Dynamic viscosity
• Kinematic viscosity
• Prandtl number
• Density

This means that our users can more accurately model specific material properties.

2. Humidity Sources Modeling

3D humidity sources are now available as a new advanced concept for humidity modeling. Attention has been paid to the robustness and stability of simulations in the presence of humidity sources. They can be used to model humidifiers.

The specification of the humidity type for fixed value boundary conditions has been added. The possibility of modeling a humidity source on a wall is also included.

3. Visualization/Computation of Local Mean Radiant Temperature (MRT) Without Solar Load

CHTv2 simulations now allow the calculation of the Mean Radiant Temperature field on fluids. This field indicates the temperature due to radiation heat transfer at a given point and can help quantify the radiant heat exchange between a person and their surroundings.

4. Parametric Mesh Study on IBM Mesh Fineness

The automatic mesh fineness slider in IBM (Immersed Boundary Method) now supports parametric runs, allowing for parallelized mesh independence studies.

5. Multiphase/Subsonic Features

5.1. Non-Newtonian Fluids Available

Our users can now model non-Newtonian fluid behavior in the Subsonic solver using the Herschel-Bulkley model.

This captures the correct physics of highly viscous, non-Newtonian fluids like motor oil and blood, combined with advanced CFD capabilities like multiphase and cavitation.

5.2. Time-Dependent Boundary Conditions

Time-dependent boundary conditions for most variables are available for transient analyses in Subsonic. Users may specify Velocity, Flow rates, Pressure, and Temperature at inlets and some outlets as functions of time in the form of a table input.

5.3. Probe Points as Result Controls

Probe points for Subsonic analyses are now available under the Result Controls. The number of parameters written out will vary depending on the type of simulation chosen.

6. Rotating Machinery – Blade-to-blade Flow Visualization

The newly released feature is a post-processing filter called “Rotational” that allows users to analyze flow through a cascade of blades. Mesh, flow vectors, and contours can be visualized on a 2D unwrapped plane of the blades, and the images can be exported.

This feature is currently available for centrifugal-type turbomachines. We will soon be releasing cascade views for axial impellers as well as meridional cut plane visualization.

7. Automated Sweep Meshing for Structural Analyses

Enable the toggle to automatically mesh bodies with continuous cross-sections using prismatic elements.

Prismatic elements such as hexahedral and wedge elements outshine standard tetrahedral elements in terms of accuracy and performance. With this feature, users can automatically benefit from swept meshes without the need for manual refinement.

8. Orthotropic linear mechanical material properties in the cartesian coordinate system

Our users now have the ability to create solid materials with orthotropic linear elastic behavior in which Young’s modulus, Shear modulus, and Poisson’s ratio are defined independently for the three mutually perpendicular cartesian directions. This allows for simple modeling of PCBs and composite structures in which material orthotropy can significantly influence peak stresses and deformations.

9. Modal-Based Harmonics

Supercharge your vibration analysis with Modal-based Harmonic analysis. This feature allows for efficient computation of many excitation frequencies, even for large mesh sizes! The new analysis method combines frequency and harmonic analysis into a single analysis, streamlining workflows and enabling users to automatically capture resonant behavior.

9.1. Automatically Capture Resonant Response in Modal-based Harmonic Analysis

Frequency responses of vibrating systems can now be resolved at high resolution using two new automation options for setting excitation frequencies in Modal-based Harmonic analysis:

• Cluster around modes: Harmonic loads are applied at frequencies clustered around eigenfrequencies.
• Cover spectrum: Harmonic loads are applied at frequencies clustered around and in between eigenfrequencies to fully capture the entire spectrum.

These options provide a super simple and automated process for capturing resonant behavior, accurately allowing users to confidently check peak values such as maximum deflection, acceleration, and stress.

10. Yeoh Hyperelastic Model

A powerful and user-friendly Hyperelastic model, Yeoh has great stability and requires only uniaxial experimental data for adequate fitting. This versatile model can capture up to 700% strains in elastomers.

11. Ogden Hyperelastic Model

This sophisticated hyperelastic model enables accurate modeling of rubbers and biological material at very high strains.

Here’s what you should know about it:

1. Accuracy: The Ogden model boasts accuracy, outshining other hyperelastic models in predicting material deformation.
2. Complexity: We’ve added options for model complexity – 1st, 2nd, or 3rd order. Allowing some flexibility when fitting the model to stress-strain relations with various levels of complexity.
3. Data fitting: To get the best results, you’ll need extensive experimental data, covering all three deformation modes (Uniaxial, Pure shear, Biaxial).

12. Select Similar Shapes

Our users can now Expand face selection by “Tangent faces”, “Same area” or “Same filet radius”. You will find this in the ‘right-click’ menu.

Note: The selection of similar bodies and selection of similar edges will be added at a later date.

Take These New Features for a Spin Yourself

All of these new features are now live and in production on SimScale. They are really just one browser window away from you!

If you wish to try out these new features for yourself and don’t already have a SimScale account, then you can easily sign up here for a trial. Please stay tuned for our next quarterly product update webinar and blog.

The post NEW Features: Temperature-Dependent Material Properties, Humidity Source Modeling, Non-Newtonian Fluids appeared first on SimScale.

Let’s get you up to date with SimScale’s new key features released in Q2 2023.

1. Custom Wind Comfort Criteria/Plots

SimScale already provides today a variety of different Pedestrian Wind Comfort Criteria, like Davenport, Lawson, London LDDC, NEN8100, and more.

Still, this list can never be exhaustive as there are a multitude of locally used and adapted comfort criteria that are either required by local authorities or have proven to be well suited to the specific local conditions.

SimScale enables our users to define their own comfort criteria with custom wind speed ranges and percentage thresholds.

With this new possibility, a range of new comfort criteria can be created. Here are some examples:

• CSTB Wind Comfort Standard
• Auckland Wind Comfort Criterion
• Melbourne Wind Comfort Criterion
• Bristol Wind Comfort Criterion
• Israeli Wind Criteria
• Murakami Wind Comfort Criteria

2. Thermal Resistance Networks for IBM

This feature is a natural extension to the Immersed Boundary solver and is already available for Conjugate Heat Transfer. It provides thermal resistance networks like two-resistor or star resistor models in the simulation setup and allows you to define detailed components like chips or LEDs as customized components. This avoids the necessity for very fine meshes for those often tiny components.

Users can define a thermal resistance network (TRN) by assigning the top surface of a cuboid as a TRN.

Model the chip as a simple cube in a CAD model or replace the detailed 3D model via ‘Simplify’ on SimScale.

3. Multiphase: Surface Tension

With the addition of surface tension, users of the new multiphase module will be able to improve the accuracy of multiphase results for surface tension dominant flows like microgravity sloshing, capillary flows, microfluidics, etc.

4. Ogden Hyperelastic Model

We have added this model to better simulate highly elastic rubber. In the animation below, you can see the movement of two solid parts coming together and separating again. There is a hollow rubber seal between them with significant deformation.

Use Case & Benefits

• Accurately simulate rubbery and biological materials at high strains
• Increasing hyperelastic functionality

5. Cylindrical Hinge Constraint

The Cylindrical hinge constraint boundary condition replicates the behavior of a fixed hinge. The assigned surface is constrained such that only rotational motion around the hinge axis is free.

SimScale can automatically detect the axis of the hinge based on an assigned cylindrical surface, but the boundary condition also allows for a user-defined input.

6. CAD Swap Improvements

When replacing one CAD model with another, it isn’t always clear what worked and what didn’t. With this feature, we add clarity so that users know what was successful and what requires their attention.

7. Parametric Studies

Boundary conditions can now be parametrized to run multiple simulations with a button click. Some examples are:

• Electronics: change inlet flow rates, change the heat load on parts
• AEC: change inlet flow rates to understand the impact on cooling strategies
• Rotating Machinery: change the inlet velocity and rotational velocity and compare designs

8. CAD Extrude Operations

Extrude is similar to move, although it will maintain the same cross-sectional area — often very useful.

This video shows one move operation followed by one extrude operation. Notice how the extrude option maintains the shape of the adjacent surfaces.

9. Distance Measurement

This is a highly requested feature, and I think we have answered nearly all use cases with this first iteration. We now offer the ability to measure the length/area of an entity and also measure the distance between two entities.

Take These New Features for a Spin Yourself

All of these new features are now live and in production on SimScale. They are really just one browser window away from you!

If you wish to try out these new features for yourself and don’t already have a SimScale account, you can easily sign up here for a trial or request a demo below. Please stay tuned for our next quarterly product update webinar and blog.

The post NEW Features: Custom Wind Comfort Criteria, Thermal Resistance Networks, Surface Tension, and More! appeared first on SimScale.

Electromagnetics (EM) simulation is an advanced technique to investigate the performance of electronic devices and systems virtually, minimizing the need for expensive and time-intensive legacy physical prototyping. With cloud-native simulation capabilities, engineers can go even further and eliminate their reliance on expensive hardware and complex installations of software by simply running all their simulations in parallel directly in their favorite web browser — no installation required. This not only accelerates the design cycle but also enables engineers to innovate faster, collaborate more easily in real time, and apply multiple physics simulations all in one place.

A Deeper Look into SimScale’s Electromagnetics Simulation

Electromagnetic fields play a pivotal role in countless technological innovations, from motors and transformers to medical devices and beyond. That’s why it’s crucial for engineers and designers to have access to state-of-the-art analysis tools that enable them to explore, understand, and optimize electromagnetic phenomena with unprecedented precision.EM systems often present challenges of different scales, particularly when it comes to frequency ranges. In our first roll-out, SimScale is offering low-frequency electromagnetics analysis capabilities with a dedicated magnetostatics solver powered by our partner, EMWorks. This will enable various low-frequency applications, such as linear actuators, sensors, and motors.

The SimScale EM solver enables engineers to visualize and analyze various electromagnetic parameters in magnetostatics, including:

• Magnetic flux density
• Magnetic field strength
• Current density
• Linear and non-linear magnetic permeability
• B-H curves
• Permanent magnets
• Inductance matrix
• Coil resistance
• Forces and torques

Thanks to the power of cloud computing, engineers can run as many simulations as needed at the same time and iterate on their designs following the results of their simulations to reach the optimal design.

Simulate Magnetostatics in SimScale

Magnetostatics is a model that describes magnetic fields when currents are temporally constant (stationary) or approximately constant. It has numerous applications in engineering and science that can be used in a wide variety of industries, including automotive, aerospace, consumer products, healthcare, electronics, and more. Of these applications, one can utilize the magnetostatics analysis type to answer various design questions on:

• DC machines
• Electromagnetic brakes and clutches
• Magnetic levitation devices
• MEMS
• Motors and generators
• Permanent magnet motors
• Relays
• Sensors
• Solenoids

In SimScale, engineers can simulate various low-frequency electromagnetics by simply using the electromagnetics solver, as shown in the figure below.

Electromagnetics Simulation Examples in SimScale

Switched Reluctance Motor (SRM)

Switched Reluctance Motors (SRMs) are distinct electric motors operating on the principle of variable magnetic reluctance. Yet, they do suffer from the presence of torque ripples, which result from the abrupt switching of currents during motor operation. These lead to vibrations, noise, and undesirable mechanical stresses.

With SimScale’s electromagnetics solver, engineers can run magnetostatics simulations that provide a comprehensive understanding of the torque generation mechanisms, torque ripple effects, and efficiency of the motor under different operating conditions.

Electromagnetic-Toothed Brake

The electromagnetic-toothed brake is a sophisticated braking mechanism that operates through the manipulation of magnetic forces to control its engagement and disengagement. It shares structural similarities with the conventional power-on brake, but it boasts a distinct advantage in terms of static torque, which stems from the interlocking teeth between the driving and driven components. By incorporating these teeth into its design, the toothed brake achieves a notably higher torque capacity compared to devices of similar size, thus offering precise and efficient control of motion. When the coil is energized (power-on), the toothed brake engages to provide effective braking, making it a valuable tool for halting the rotation of a load when electrical power is applied.

In the image below, we provide a visual representation of electromagnetics simulation results in SimScale, illustrating the magnetic toothed brake in action. The image showcases both the engaged and disengaged states of the brake.

Linear Solenoid (Actuator)

Linear solenoids are electromagnetic devices that generate linear push or pull motion using magnetic fields. By adjusting the number of coil turns, material properties of the parts, or the applied current through the solenoid, engineers can optimize the stroke length of a linear direct-pushing solenoid. In other words, by controlling the magnetic field, engineers can tailor the solenoid’s stroke to suit specific application requirements, such as valves, locks, actuators, and other linear-motion devices.

More Electromagnetics Simulations Coming Soon

Low-frequency electromagnetics is just the beginning for SimScale. In the near future, we plan to introduce additional modules that will enable simulations of AC magnetics, transient magnetics, electrostatics, AC electrics, and high-frequency applications at last.

All these modules contribute to the multiphysics capabilities that SimScale provides, enabling engineers to run all the necessary simulations and analyses to ensure proper testing and validation before the need for any physical prototyping.

SimScale’s EM simulation software is the new kid on the block, but it is a game changer in terms of minimizing go-to-market time and costs for electromagnetic products.

In our effort to enable engineering organizations to deploy simulation broadly while maintaining central control over simulation knowledge and usage, SimScale is integrating electromagnetics into its comprehensive suite of simulation tools, delivered via a single consistent GUI and API.

The post Low-Frequency Electromagnetics Simulation — Now in Your Browser appeared first on SimScale.

This is where SimScale steps in. SimScale is a cloud-native simulation platform that seamlessly integrates with SOLIDWORKS, offering a rich array of advanced engineering simulation capabilities and computational power in the cloud. Leveraging the advantages of cloud computing, SimScale not only ensures a seamless CAD-import workflow for SOLIDWORKS users but also enables access from anywhere directly in their web browsers and allows for parallel simulations, empowering users to run multiple simulations at the same time. But that’s not all. There is a secret sauce, a key advantage all engineers and designers on SOLIDWORKS would benefit from massively.

In this article, we will look through the SOLIDWORKS Simulation and Flow Simulation tools, and then we will explore the seamless and proven CAD-import workflow between SOLIDWORKS and SimScale, highlighting its secret sauce in how it allows users to easily simulate native SOLIDWORKS parts and assemblies.

SOLIDWORKS Simulation and Flow Simulation Tools

SOLIDWORKS provides two robust, in-house simulation tools, SOLIDWORKS Simulation for FEA and SOLIDWORKS Flow Simulation for CFD. These integrated tools help engineers perform structural, thermal, and fluid flow analyses in their familiar SOLIDWORKS environment.

SOLIDWORKS Simulation

Engineers and designers can make use of SOLIDWORKS Simulation to perform structural analysis and predict how a design would behave under particular conditions. Some of its highlights are:

• SOLIDWORKS FEA capabilities for structural analysis
• Linear, non-linear static, and non-linear dynamic capabilities
• Easy-to-use interface for setting up simulations
• Easy integration with CAD models

SOLIDWORKS Flow Simulation

SOLIDWORKS Flow Simulation extends the simulation capabilities into the realm of CFD. With this tool, engineers can analyze the behavior of fluids (liquids and gases) within or around their designs. Some of its highlights include:

• SOLIDWORKS CFD capabilities that utilize the Finite Volume Method (FVM)
• Simulation of fluid flow, heat transfer, and radiation
• Usability in application areas like HVAC and electronics cooling
• Parametric studies for design optimization

The Bridge to SimScale

While SOLIDWORKS offers useful simulation tools, there are scenarios where advanced analysis options, extensive computational resources, or specialized features are required. SimScale bridges this gap by providing access to a cloud-powered, comprehensive suite of simulation solutions that reinforces SOLIDWORKS’ CAD capabilities. In other words, designing in SOLIDWORKS has become significantly more powerful by leveraging the proven workflow with SimScale.

Key Benefits of Using SimScale for Simulating SOLIDWORKS Models

Here are some key reasons why engineers turn to SimScale in order to simulate their SOLIDWORKS CAD models:

• Cloud-Based Power: SimScale leverages cloud computing, eliminating the need for high-end hardware and reducing simulation time. Users can access SimScale instantly from anywhere, and collaboration becomes effortless, as team members can easily work together on the same simulation and design project directly in their web browsers without any special hardware and with unparalleled, real-time, in-app support from simulation experts.
• Advanced Solvers: SimScale offers highly reliable and advanced solvers for FEA, CFD, and thermal analysis, enabling engineers to tackle complex problems efficiently. These include non-linear analysis, modal analysis, multiphase flow, conjugate heat transfer, transient simulations, and more.
• Automation and Optimization: Parametric studies, optimization, and design exploration are simplified in SimScale. Engineers can explore multiple design variations by simulating them in parallel (multiple simulations at the same time) to achieve optimal results. There is no limit to the simulation size, number of parallel simulations, and storage.
• One Platform, Broad Physics: SimScale offers a single platform with broad physics capabilities for both early-stage and late-stage simulations.
• Cost-Effectiveness: Simulating in SimScale minimizes the total cost of ownership, making it economically viable for everyone from single users up to hundreds of seats.

SOLIDWORKS and SimScale Associativity: The Key to a Seamless Workflow

The workflow between SOLIDWORKS and SimScale is designed to be seamless and user-friendly. The standout feature of this workflow is CAD associativity. This means that changes made to the original SOLIDWORKS CAD model automatically propagate to the SimScale simulation setup, eliminating the need for manual updates and ensuring that the simulation always reflects the latest design iteration.

When you design a product in SOLIDWORKS, every modification, no matter how minor, triggers an update in the associated SimScale simulation. This dynamic link ensures that engineers can make design improvements iteratively based on simulation results, leading to a more streamlined and efficient design validation process.

How It Works: Extending Your Design Capabilities with Simulation

The powerful synergy between SOLIDWORKS and SimScale unlocks a world of possibilities for engineers, making it possible to harness the full potential of their designs. With SOLIDWORKS providing high-quality 3D modeling capabilities and SimScale extending these capabilities with robust and comprehensive simulations in the cloud, users can achieve a higher level of design validation, optimization, and collaboration.

Simply, here is how the SOLIDWORKS-SimScale workflow works:

• Native format support: You can save SOLIDWORKS parts and assemblies in their native file formats and directly upload them to SimScale with no translation losses.
• Cloud-native simulation setup: You can set up sophisticated multiphysics simulations in SimScale by leveraging its broad-physics capabilities all on one cloud-based platform.
• Continue to design: You can continue to freely use your local computer for design work while the simulations are running in the cloud.
• Seamless design update: New design versions are associatively imported into SOLIDWORKS, retaining simulation settings for a fast iterative design process.

By seamlessly transitioning between these tools, engineers can bring innovative products to market faster and with greater confidence, all while continuing to design in SOLIDWORKS on their local computers and simulating in SimScale in the cloud. This integrated workflow truly showcases the simulation capabilities of SimScale and enhances the SOLIDWORKS experience for engineers across various industries.

Explore the SOLIDWORKS-SimScale proven workflow and sign up to SimScale to simulate your SOLIDWORKS CAD parts in the cloud.

Explore SOLIDWORKS-SimScale Workflow

The post SolidWorks Simulation: Seamless Proven Workflow with SimScale appeared first on SimScale.

Evidently, this is reflected in the market. The global simulation market has grown massively over recent years thanks to its facilitating engineering organizations to increase top-line growth and simultaneously drive bottom-line savings, all while improving the sustainability of products and processes. In fact, several reports have shown that analysts expect a market rise of 13% CAGR over the next five years, closing in on the USD 30 billion mark by the end of the decade [1-3]. Particularly, the cloud simulation segment has shown explosive growth and is estimated to drive the highest growth in the coming years. This is mainly because cloud-native simulation allows the use of engineering simulation earlier in the design process and broader across physics while reducing the cost and technology footprint.

From CFD analyses of aerodynamics, turbomachinery, and urban microclimates to FEA simulations of structural mechanics in automotive and consumer products, all the way to thermal analyses for indoor environments and electronics cooling, the applications for engineering simulation are numerous and can involve multiple physics. In this article, we dive into cloud simulation and CAE and find out how it strategically impacts business growth and helps optimize financial resources, enabling sustainable growth and profitability while minimizing risks. We explore in numbers how various companies from different industries have benefited from SimScale’s cloud-native simulation platform and how you can, too.

Accelerating Innovation with Cloud-Native Simulation

SimScale’s cloud-native simulation harnesses the power of cloud computing to execute complex engineering simulations, modeling scenarios, and data analysis with unprecedented speed and efficiency. This approach has profound implications for businesses, particularly in the engineering domain, where traditionally time-intensive and resource-demanding processes have been a barrier to rapid innovation.

In essence, it provides engineers with rapid access to the cloud’s high-performance computing resources, enabling them to conduct multiple simulations simultaneously and iterate designs swiftly. This accelerates the development of new products, allowing for the exploration of a broader range of design possibilities and advanced features.

Not only does cloud-native simulation drive innovation, but it also ensures prudent and strategic financial management of the R&D process, as it delivers quantifiable financial benefits that arise from reduced physical testing, optimized resource utilization, faster time-to-market, and more. By reducing the reliance on physical prototypes and enabling virtual testing, cloud simulation promotes cost-effective experimentation, leading to breakthroughs in product performance and functionality. Additionally, the data-driven insights derived from simulations enable informed decision-making, further fueling innovation and pushing the boundaries of what’s possible in engineering and manufacturing.

Fundamentally, cloud-native simulation maximizes business value by enabling earlier, broader, and more intense simulation use while having no IT/Capex footprint. The table below clarifies how cloud-native simulation compares to traditional computer-aided engineering (CAE) use and mere physical testing without simulation.

	No Simulation	Traditional Simulation	Cloud-Native Simulation
Simulation Use	Physical testing only	Late-stage validation	Continuous
Simulation Users	None	Few CAE experts	Designers, Engineers, Experts
Possible Design Space	Narrow	Broad	Significantly broad
R&D Cycle Time	Long	Short	Minimal
IT/Capex Footprint	High	Significant	None

The impact of SimScale’s cloud simulation can be realized on two levels: top-line growth and bottom-line savings. In the following sections, we look deeper into each level and explore case studies of companies from industries like automotive, AEC, electronics, energy, and others that have already benefited substantially from cloud simulation.

Top-Line Growth with Cloud-Native Simulation

SimScale supports business revenue growth by reinforcing strategic approaches to accelerating innovation. This can be exemplified by boosting product performance and expediting time-to-market. Figure 2 displays the quantifiable benefits of many SimScale customers that have boosted their top-line growth with cloud-native simulation. Let’s dig deeper and learn more about some of these case studies.

Higher Product Performance

Using cloud simulation, engineers and designers can conduct complex simulations that were once computationally prohibitive, which enables them to develop and validate advanced product features more efficiently to reach higher product performance. This can result in higher end-customer satisfaction and enables companies to push design boundaries, innovate faster, and gain a competitive edge in the marketplace.

SimScale enables broad simulation across different physics and applications with a full-stack simulation technology that empowers users to boost their products’ performance and validate faster and better using scalable cloud-computing power. With its powerful solvers for fluid dynamics, structural mechanics, thermal analysis, and electromagnetics, SimScale allows for integrated, complex, and large-scale simulations, generating a wealth of data that reinforce decision-making and product development strategies, leading to better opportunities for establishing a stronger presence in the market.

One example is ITW, a global design and engineering firm and a leading global supplier of auto parts. Using SimScale, ITW engineers conducted nonlinear static simulation and analysis to accelerate the development of plastic automotive fastening components, allowing them to minimize the insertion force of their fasteners by up to 85% while saving 10% of their R&D costs.

Another example is Samco, a semiconductor and materials company out of Japan. One of Samco’s main principles is providing durable and highly reliable equipment, and that’s why they opted to use SimScale’s CFD, mechanical, and thermal simulation capabilities. With its ease of use and low cost compared to on-premises-based solutions, SimScale was an attractive choice for Samco’s design team, especially being first-time users. Using SimScale early in the design phase of a vacuum chamber enabled them to make small but rather effective design modifications, increasing their product lifetime by no less than ten times to exceed a million operating cycles.

Faster Time-to-Market

Shorter development cycles facilitated by cloud simulation can lead to earlier product launches, capturing market opportunities ahead of competitors and increasing revenue streams. In fact, one of the most impactful aspects of SimScale’s cloud-native simulation is parallel simulations. Engineers and designers alike can conduct multiple simulations in parallel with no limit on simulation size, number of simulations, or storage. This not only accelerates time-to-market but also increases scalability and eliminates the need for multiple steps and hand-offs between teams.

With SimScale, you can shorten your design cycle time and fix design issues earlier by benefiting from parallel simulations, faster data processing, rapid prototyping, easier collaboration and accessibility, and reduced administrative overhead. Simply, any authorized team member can access the simulation projects anytime, anywhere, directly in their web browser, and run parallel simulations that can cut design cycles from weeks to days and even minutes.

This is exactly what Withings was able to do using SimScale. This consumer electronics company leveraged the SimScale platform to conduct structural analysis of its health monitoring equipment. Running multiple simulations in parallel allowed them to reduce their design-to-prototype cycles by no less than seven times, enabling them to test new designs and push their products to market much faster.

“By using the novel mechanical simulation based in the cloud offered by SimScale, we engineers at Withings have been able to reduce our design-to-prototype cycles from weeks to days. This tool widens our possibilities to test new designs, materials, and techniques and anticipate possible failures, as well as gains in mechanical performance within a few clicks.”

Victor Pimenta – Mechanical Engineer at Withings based in Paris, France

Another company that shortened its design process significantly is Rimac. This visionary electric car manufacturer made use of SimScale’s thermal simulation tool to study the cooling of their battery pack. They easily managed to conduct 30 simulations in parallel, allowing them to save up to 96% in time savings and run ten different parameter variations. This shortened their time-to-result by 20x.

Bottom-Line Savings with Cloud-Native Simulation

SimScale’s cloud-native simulation enables companies to ensure substantial savings on operational costs, improve profit margins, and increase net income by lowering R&D and engineering costs and minimizing costs of goods sold (COGS). These cost-cutting measures and efficiency improvements can have a significant bottom-line impact, enabling companies to enhance their financial health and profitability. Figure 6 showcases how SimScale customers have leveraged cloud simulation to improve their bottom-line savings.

Lower Engineering and R&D Costs

With SimScale, you can replace a large chunk of your physical prototyping efforts, enabling you to reduce the costs of prototyping and hardware significantly. In some cases, you wouldn’t need to run any physical testing until the end of your testing process for final validation because simulation can provide accurate data and results that enable you to run design iterations faster and, evidently, at lower costs. This reduces your R&D turnover and supports the company’s strategic and financial efforts, not to mention the reduction of costs associated with compliance and safety standards.

Johnson Screens, a global supplier of industrial products in the AEC industry, has leveraged SimScale’s cloud-native CFD tool to conduct airflow analyses instead of conducting physical experimentation. As a result, they were able to save up to $15,000 in engineering costs and months of preparation per experiment.

“To get the same results with a physical test, it would take us months and would cost anywhere from $7k to $15k, even for this project, which is actually very small in scope. With SimScale, we could just run a virtual test at the office, which took only 18 minutes.”

Daryn Bertelson – CAE Engineer at Johnson Screens – Aqseptence Group

Lower COGS and Product Costs

Another bottom-line advantage of cloud simulation is the minimization of COGS and costs associated with materials, energy consumption, and warranties. With virtual prototyping, the need for physical prototypes is reduced, saving time, energy, and material costs during the design phase.

One example is Kichler, a residential lighting supplier out of the US. Kichler has opted to adopt an entirely cloud-based engineering software stack, which has resulted in faster product innovation and lower hardware and software costs. Using SimScale, they managed to entirely eliminate prototyping costs and save significantly on material costs.

“The total material cost saving was 44%. We eliminated testing and prototype costs on this project entirely. It usually takes 1-3 weeks of prototyping time and another 1-3 weeks for all the testing normally done. Overall this project was successful and it helped achieve our goal of reducing cost and keeping within the project timeline.”

Josh Levine – Lead Engineer in Value Engineering Department at Kichler Lighting

Another one is Axens, a solution provider for the conversion of oil and biomass to cleaner fuels. By embracing cloud-native simulation with SimScale, they were able to optimize their design and save up to 20% in energy consumption. They also saved about €27,000 in external simulation costs by opting to do the simulation work in-house rather than outsourcing it.

Sustainable Product Growth with Robust R&D Savings

Cloud simulation stands as a transformative force in engineering and manufacturing, offering the promise of accelerated time-to-market, heightened innovation, and cost-efficient operations. By harnessing the power of the cloud with SimScale, organizations can navigate the complex landscape of product development with greater agility, achieving top-line growth through rapid innovation while simultaneously bolstering bottom-line savings through optimized processes and resource utilization.

This dynamic synergy positions companies not only for immediate success but also for sustained, long-term growth, ensuring their competitiveness in an ever-evolving marketplace while maintaining financial prudence and operational efficiency.

The post How Cloud-Native CAE Simulation Boosts Growth & Savings appeared first on SimScale.

Recent advancements have profoundly shifted the legacy product development workflow. Cloud computing has allowed engineers to collaborate in real-time, access advanced simulation capabilities earlier in the design process, obviate costly physical prototyping, and avoid expensive hardware costs. The virtually unlimited computing power and scalability of the cloud mean that deploying these capabilities across an entire distributed engineering organization is now considered a preferred strategy. 

Additional benefits of cloud-native simulation include access to 3D parametric scenarios analyses without any impedance to time and computing resources. The value added is the ability to fully explore design space and disqualify poor design candidates earlier in the development process. Application programming interfaces (APIs) further amplify the toolkit available to electronics designers and enable third-party CAD, analysis, optimization, and parametric design tools to talk to each other.

However, common bottlenecks to simulation have been CAD preparation and the numerical discretization of that model (meshing). Both consume time and manual intervention. The advent of advanced physics solvers and novel meshing techniques, such as the immersed boundary method, means that engineers spend less time making their CAD models simulation-ready and more time on insight-driven design. Skipping the time-intensive CAD preparation also opens up the possibility of doing simulations very early when some components are still in the draft stage and comparing many variants that otherwise would have required repeated CAD simplification efforts.

Immersed Boundary Method in SimScale

The Immersed boundary method is based on a cartesian grid, in which the geometry gets immersed. It is resilient to geometrical details and does not require CAD simplification, even for very complex models. Salient features of this approach include:

• Automatic defeaturing of small geometrical details
• Mesh refinements are physics-based rather than geometry-based
• Perfect hexahedral meshes
• Highly flexible mesh sizing from very coarse to very fine for all levels of CAD complexity

The Challenge with Complex CAD Models

Engineers who want to improve their thermal design by running 3D simulations with conventional tools based on so-called body-fitted meshing are forced to pick the lesser evil of simplifying their CAD model heavily or requiring huge computational resources to simulate.

In any case, they will waste valuable time. Either they spend hours of their limited working time on tedious CAD cleanup and simplification to reduce the model complexity and required computing resources, or they need to put an unreasonable amount thereof to solve the model. While the second approach allows the engineer to continue working on other topics during the calculation, it likely still requires some level of initial CAD cleanup even to get a successful mesh. Additionally, it blocks any advances on their current design for the time the simulation is running, not even speaking about the required hardware costs.

Immersed Boundary Method to the Rescue

The Immersed Boundary method addresses the core of this dilemma. It completely removes the CAD preparation or reduces it to a few minutes at most. At the same time, the physics-driven meshing avoids high mesh resolutions on detailed CAD features that are insignificant to the system’s thermal behavior. Yet, it resolves physically relevant regions like power sources or flow channels to the level the user requires. This level might differ significantly based on the current simulation intent.

Early in the design process, the engineer might be more interested in qualitative insights into the thermal management concepts he is experimenting with. Those simulations often only require coarse mesh resolutions. Body-fitted approaches do not allow this design space as the geometric details always lead to large mesh sizes (see Figure 2 above).

Later in the process, the focus shifts towards quantitative results, such as maximum temperatures on critical components. In order to provide the required accuracy, the mesh resolution required by body-fitted and IBM meshing will be closer. The CAD preparation time, of course, remains as saved time, and the engineer benefits from the underlying solving based on the same high-fidelity finite volume implementation in both cases.

The Main Benefits for Engineers

Summarizing the main benefits for the engineer, immersed boundary method enables you to:

• Simulate early in the design phase when parts of the system are in the concept phase
• Simulate the detailed original model without the need for CAD preparation
• Run extensive design of experiment studies
• Derive accurate critical temperatures during the validation phase

The graphic below shows how an important result quantity e.g. the junction temperature of a chip changes with higher mesh resolutions i.e. higher computational costs for body-fitted and IBM-based simulations.

So what’s the catch? Well, there is no catch. The Immersed boundary method is a perfect fit for thermal engineers dealing by default with complex systems and looking for optimization insights throughout the design process.

Suppose an engineer is looking for a very accurate representation of the flow boundary layer, for example, when designing fan blades or something similar. In that case, it usually makes more sense to capture the physical effect with mesh boundary layers and prepare the CAD for a conventional body-fitted simulation.

A Case Study of an Electric Vehicle Battery Pack

We have shown a case of an electric vehicle battery pack. The design is an example of an air-cooled lithium-ion battery model for an FSAE electric race car, which is utilized as the accumulator to power the car. Effective thermal management of the battery pack is essential to ensure the reliable and safe operation of the battery cells and hence the car.

When the rest of the vehicle design is constantly changing to meet performance objectives, a continuous update in the battery pack is also required to align with these evolving design parameters. Throughout each design iteration, key questions arise regarding the amount of heat generated within the batteries for relevant duty cycles and the necessary air flow rate to maintain the battery pack within its designated temperature range.

Finding accurate answers to these questions using a robust electronics cooling simulation not only increases the confidence in the design but also makes it possible to get those answers faster, even when the model is geometrically very detailed. The provided example demonstrates the significant advantage of the immersed boundary method in handling geometrical details without the need for CAD simplification. As a result, valuable insights can be swiftly obtained from each design iteration, enabling a more efficient and thorough evaluation of the design without compromising accuracy or intricate geometric details.

Comparing Meshing Methods

In this battery pack example, we attempted to simulate both the original and simplified versions of the CAD model using Conjugate Heat Transfer analyses with body-fitted and Immersed Boundary Method solvers. The main advantage of using the cartesian mesh-based Immersed Boundary Method is that it allows us to overcome the inability to mesh the original model using a body-fitted mesh. By immersing the geometry into the cartesian grid, the cartesian mesh automatically defeatures small geometrical details.

Whereas to prepare the model for CHT analysis, we need to invest time in CAD cleaning. This involves removing small details that do not directly impact the problem’s physics and achieving an efficient mesh size without unnecessary refinements. The time devoted to cleaning the small geometric details for the body-fitted mesh, in this case, ~ 2.5 hours, can be instead well-spent on design iterations when using the immersed boundary method.

Benefits of Using the Immersed Boundary Analysis in SimScale

Based on the results of the two different methods, the benefits of using IBM can be summarized as follows:

• Automatic handling of small details: With IBM, the geometry is ready for simulation without the need for CAD cleaning, even when the original model contains numerous small details. In contrast, CHT relies on a body-fitted mesh, and if the unsimplified original model is not simulation-ready, additional effort must be spent on CAD cleaning.
• Computational efficiency: By manually cleaning the small details in the geometry, the model becomes ready for CHT. Even when this is the case, IBM requires significantly fewer computational resources than CHT. With IBM, the mesh size is reduced by 72%, computational resources are reduced by 61%, and the runtime is reduced by 62%. Moreover, the results of the two analysis methods are almost identical (0.002% difference), demonstrating the accuracy of using IBM for design iterations while efficiently utilizing resources.
• Effective design exploration: Even when the original model is simulated using IBM, the computational resources required are still much less compared to the simplified model simulated with CHT. The mesh size of the original model with IBM is reduced by 70% compared to the simplified model with CHT, while both the computational resources and the runtime are reduced by 44%. This highlights that without wasting time cleaning the model, multiple design iterations can be tested without the overhead of fine mesh regions around insignificant details. Consequently, the time required for users to discover the best design is significantly shortened.

The post Simplify Your Thermal Simulation With Immersed Boundary Method appeared first on SimScale.

For the result fields of harmonic simulations, this is true in variables that are linearly proportional to the input loads, such as:

• Displacement components
• Strain components
• Stress components

On the other hand, derived quantities expressed as nonlinear functions of the result fields naturally do not exhibit a harmonic behavior. This applies, for example, to:

• Total displacement
• Principal strains and stresses
• Von Mises stress
• Tresca stress

Moreover, the harmonic quantities are expressed in terms of their peak value (module or magnitude are terms used interchangeably for it) and a phase delay angle. This expression is encoded in a single complex number, often called a “phasor.”

In a practical sense, a harmonic quantity at any point will oscillate between a maximum and a minimum value given directly by the module of the response. Therefore, this value can be directly used in engineering assessments, e.g., to know the maximum deflection, maximum strain, peak bending stress, etc.

In contrast to this direct method, an issue arises when the analyst is required to perform their assessment using one of the nonlinear derived quantities, for example, maximum von Mises stress. In this case, the simple harmonic peak value evaluation is no longer available, and a more involved process is needed in order to provide an equivalent simple measure, a.k.a. the “Maximum over phase.”

Max Over Phase Von Mises Stress in SimScale

A feature to compute the max over phase von Mises stress in FEA harmonic simulation results was recently released in SimScale. This feature leverages the analytic formulation of the von Mises stress as a function of the principal stresses to find its maximum value across one harmonic phase cycle. The method first finds the phase sweep angle at which the peak value occurs and then computes the value. Only the max over phase value is delivered in the result fields.

In this study, this feature is tested qualitatively and quantitatively for the accuracy of the results by essentially comparing them to a brute force phase sweep history expansion. All the details of the approach are provided below.

Test Case

The test model captures a round shaft in a cantilever bending configuration. The bending load rotates around the center axis of the shaft while staying orthogonal to it.

The parameters of the model are:

• Geometry
  • Length = 0.3 m
  • Diameter = 0.05 m
• Material
  • Young’s modulus = 2.05e11 Pa
  • Poisson ratio = 0.28
  • Density = 7850 kg/m3
• Load
  • Peak value = 1000 N
  • Direction = orthogonal to the shaft axis
  • Vibration = 500 rotations per second

FEA Model

The rotating load is modeled as two orthogonal components spaced at a 90° phase angle:

$$ \vec F = \{F_x, F_y, F_z\} $$

$$ F_x = 0 $$

$$ F_y = 1000 @ 0° $$

$$ F_z = 1000 @ 90° $$

The FEA mesh is implemented with a sweep refinement, giving the following statistics:

• Number of nodes = 49,767
• Number of elements = 11,400
  • Tetrahedrals = 0
  • Hexahedrons = 10,600
  • Prisms= 800
  • Pyramids = 0
• Element type = 2nd order

Methodology

For the numerical verification of the method, we start with the computed stress components in complex form. The stress components results are assumed to be correct and are not tested in this exercise, but only the derived quantities of principal stresses and von Mises stress.

The core idea is to perform a ‘phase expansion’ of the stress components’ histories and then compute the derived quantities, such as the von Mises stress, for each point in the history. This ‘brute force’ approach is completely different from the methodology implemented in the platform and should provide quality reference results for comparison.

Two variations of this methodology were implemented in this exercise:

1. Direct computation of the VMS history from the stress components history
2. Computation of the VMS history passing through the principal stresses

Following is a detailed explanation of the equations and methods implemented.

Phase Sweep Expansion

The ‘phase (sweep) expansion’ for one variable is performed by using the definitions of the harmonic functions and the relations to the complex form, using Euler’s formula. For instance, for a harmonic variable x, we have:

$$ x(t) = A cos(\omega t +\phi) $$

$$ x(t) = Re\{Ae^{j(\omega t +\phi)}\} $$

Note: using the cosine definition of the harmonic variable is important because the phase angle directly relates to the ‘phase of max’ or the sweep angle at which the peak value of the function happens. In the case of a sine function, the phase angle relates to the zero-crossing, which is not as useful.

The complex variable form for the variable contains the information of the amplitude and phase angle (this type of variable is also known as a “phasor”):

$$ X = Ae^{j\phi} $$

This is the quantity that is found in the harmonic FEA results for the linearly proportional fields (components of displacement, strain, and stress). It is presented as a set of two variables, and can take two different shapes:

• Magnitude and phase:

$$ Magnitude = A $$

$$ Phase = \phi $$

• Real and imaginary components:

$$ Real \ Part = Re\{X\} = A cos(\phi) $$

$$ Imaginary \ Part = Im \{X\} = A sin(\phi)$$

In the latter case, the complex variable is expressed as:

$$ X = Re\{X\} + jIm\{X\} $$

Thus, to find the time history of the variable starting from the complex form we can simply perform the operation:

$$ x(t) = Re\{Xe^{j\omega t}\} $$

This expression is furtherly simplified to abstract away the frequency of oscillation, by expressing the independent variable in terms of the ‘phase sweep angle’:

$$ \theta = \omega t \in [0, 2π] $$

Then the time history of the variable is presented as a function of this angle instead of time.

Derived Quantities

After having the phase sweep expansion of the linear variables, the computation of the history of a derived quantity is done simply by applying the corresponding formula at each angle.

Let’s take, for instance, the maximum total displacement at a given point. The procedure to compute its phase expansion would be as follows:

1. Obtain the complex results for the displacement components at the desired location \(X\), \(Y\), \(Z\)
2. Perform the phase sweep expansion for the displacement components and obtain the curves \(x(\theta)\), \(y(\theta)\), \(z(\theta)\) with \(\theta \in [0, 2] \)
3. For each angle in the phase sweep history, apply the corresponding formula \(\omega(\theta) = \sqrt{x(\theta)^2 + y(\theta)^2 + z(\theta)^2} \)
4. Finally, examine the phase sweep history of the total displacement to extract the maximum value and, optionally, the phase angle at which it occurs (“phase of max”).

Notice that the total displacement is not a linear function of the displacement components, and therefore it is not a harmonic function by itself and can not be expressed in the complex form. Notice how its frequency of oscillation is double the frequency of oscillation of the displacement components.

A similar approach is followed to obtain other derived quantities, such as:

• Principal stresses: at each point in the phase sweep history of the stress components, the stress tensor is built, and the eigenvalue problem is solved for it, obtaining the phase expansion of the principal stresses.
• Von Mises stress: can be computed starting from the phase expansion of the stress components or from the phase expansion of the principal stresses. For both cases, simple formulas are available which can be applied point-by-point in the history:

$$ \sigma_{eq} = \sqrt{\frac{1}{2}[(\sigma_{xx} – \sigma_{yy})^2 + (\sigma_{yy} – \sigma_{zz})^2 + (\sigma_{zz} – \sigma_{xx})^2 + 6(\sigma_{xy}^2 + \sigma_{xz}^2 + \sigma_{yz}^2)]} $$

$$ \sigma_{eq} = \sqrt{\frac{1}{2}[(\sigma_1 – \sigma_2)^2 + (\sigma_2 – \sigma_3)^2 + (\sigma_3 – \sigma_1)^2]} $$

Results

SimScale

The result distribution and max over phase value for the von Mises stress are shown in the following figure.

Maximum value of Von Mises stress, max over phase = 54.64 MPa

Competitor Software

The same model was implemented in a competitor FEA tool, which also offers the ‘Max Over Phase’ calculation feature. The implemented model has the following statistics:

• Number of nodes = 37,094
• Number of elements = 8 550
  • Tetrahedrals = 0
  • Hexahedrons (Hex20) = 8,450
  • Prisms (Wedge15) = 100
  • Pyramids = 0
• Element type = 2nd order

The max over phase distribution shows the same behavior as SimScale, with the following numeral results:

Maximum Von Mises stress, max over phase = 54.599 MPa

Deviation with respect to the SimScale results = 0.08 %

Numerical Verification

By applying the verification methodology at the point reported for the maximum value of the von Mises stress (max over phase), we obtain the following results.

First, we see the phase sweep expansion of the stress components, appreciating the harmonic functions, peak values, phase angles, etc.

From here, we can compute the history of the von Mises stress by applying the formula at each point of the phase sweep:

We can appreciate the doubling of the frequency due to the squaring operations in the formula. Examination of the curve yields a maximum value of 55.4335 MPa.

At this same mesh node, the SimScale results show a max over phase value of 54.6369 MPa. The computed deviation at this point is 1.44%.

If we apply the same procedure for all of the mesh nodes in the results, the Normalized Mean Absolute Difference error measurement gives a weighted deviation of 3.9% (here the larger values are emphasized). This error measure is necessary to weigh down the close-to-zero results, for which the relative deviations are always very high.

The NMAD is computed as:

$$ NMAD = 100 \frac{\lt \vert e – p \vert \gt}{max(\lt \vert e \vert \gt, \lt \vert p \vert \gt)} $$

Where:

• The result values from the simulation e
• There reference computed values p
• The average operator (over all the components)
• The absolute value operator |x|

In order to investigate the cause of the discrepancy, we can note that the SimScale methodology is to compute the von Mises stress starting from the principal stresses. Moreover, the principal stresses are computed from a stress tensor formed by the complex stress components.

That is, a complex-valued stress tensor, for which the eigenvalue problem is solved. The result of this computation is a set of three complex eigenvalues that are, in turn, interpreted as the phasor representation of harmonic principal stresses. A phase expansion can be performed for them, for instance.

It is important to emphasize how this methodology assumes that the principal stresses are indeed harmonic functions. Upon further analysis, this does not seem to be a correct assumption, primarily because the eigenvalue computation is not a linear operation on the stress components. Furthermore, a numerical verification is also performed.

In order to test this hypothesis numerically, a comparison was made between the two approaches for computing the phase expansion of the principal stresses:

1. “Method A” – This is the method implemented in SimScale, where the principal stresses are assumed to be harmonic functions, and their complex representations are obtained by solving the eigenvalue problem for the complex stress tensor. The phase expansion is then performed on the complex principal stresses.
2. “Method B” – This method makes no assumptions about the principal stresses, and works by first obtaining the phase history of the stress components, then forming a real stress tensor at each point in the history. The eigenvalue problem is solved at each point to yield the history of the principal stresses.

The results of the two methods are shown below:

At first glance, it is apparent that the two results are similar. But a closer inspection shows how the actual eigenvalues are not harmonic functions (method B, the dashed lines in the plot), which is expected. The curves are close at the peaks but diverge at the zero crossings.

The real history of the principal stresses seems to be kind of an ‘envelope’ of the assumed harmonic, simplified principal stresses.

Nonetheless, the assumption of the principal stresses as harmonic functions seems to properly capture relevant information, such as the peak values and the phase of max.

Also, it can be seen from the figure that a small deviation is found between each harmonic principal stress and the reference curves (method B). The deviations at the peaks are:

• Deviation in P1 = 0.7%
• Deviation in P2 = 0.5%
• Deviation in P3 = 2.2%

Finally, the von Mises stress computed from method A and B principal stresses are compared, in order to measure the deviations between them:

The curves display the same trend, with the phase of max coinciding on both. But the computation of the von Mises stress seems to aggregate the deviations from the computation of the principal stress (aka the assumption of harmonic shape) to showcase a larger deviation.

As related to the results from the analysis of the principal stress methods, the deviations are larger at the valleys and smaller at the peaks.

The values at the peak are:

• VMS max over phase, method A = 54.6369 MPa
• VMS max over phase, method B = 55.4335 MPa
• Deviation = 1.44%

The value for the von Mises stress using method A coincides with the reported results from the SimScale simulation, which confirms that the assumptions and methodologies are correctly captured in this analysis.

The results show that the methodology implemented in SimScale suffers from the limitations of assuming a harmonic shape for the principal stresses and the corresponding impact on the precision of the results. On the other hand, this method is efficient and fast, which offers a profitable tradeoff between computation time and accuracy.

This lack of precision is minimized when the focus of the analysis is to obtain the max over phase, or the peak value of the history because, at this point, the minimum error is expected to happen if the results of this analysis can be generalized. On the same trend, if the interest would shift to deliver the phase sweep history or the minimum values, then the imprecisions of the method would be of greater impact.

Conclusions

A test methodology is presented for the validation of the results given by SimScale’s implementation of the “Von Mises Stress – Max Over Phase” feature.

The methodology works by implementing two alternative ways of computing the max over phase values starting from the harmonic stress components and comparing the final numerical results.

The methodology was implemented on the test case of a round cantilever beam under the action of a rotating bending force. The case was solved using SimScale FEA Harmonic simulation to obtain the stress results.

The comparison of results shows that the SimScale maximum von Mises stress (max over phase) has a deviation of 1.4% with respect to the reference methodology.

The comparison of results with the same model implemented in the competitor software shows a difference of 0.08% in the maximum value, which seems to indicate that a similar methodology is implemented in that product. The qualitative distribution of the results is also comparable.

The computation of the principal stresses was also tested, which revealed that the probable cause of the imprecision in the SimScale methodology lies in the (arguably mistaken) assumption that the principal stresses are harmonic functions.

It is also found that, although imprecise, the SimScale methodology is fast and resource-efficient. It properly captures information such as the “phase of max” and a close “max over phase”, with an error of 1.4% on the peak value of the von Mises stress and of 2.2% on the peak value of the third principal stress.

The post Max Over Phase in Harmonic Response Analysis appeared first on SimScale.

In this blog post, we want to get you up to date with all of the new key features released in Q1 2023. Let’s dive in!

1. Improved Wind Data for PWC Analysis
2. Humidity Modeling
3. Realizable Turbulence Model
4. Solids included in solar radiation
5. Joule Heating
6. Immersed Boundary Method (IBM) external flow domain flexibility
7. Simplify/heal bodies with Surface Wrapping
8. Multiphase
9. Relative Velocity
10. Boundary condition Visualization Inside 3D Viewer
11. Export result statistics to CSV
12. Teams and Permissions
13. Bilinear elastoplastic material model
14. “Max over Phase” von Mises stress result field for harmonic analysis

1. Improved Wind Data for PWC Analysis

Improved accuracy and global coverage with the new ERA5T dataset, which replaces the previous NEMS30 dataset. Also, the new modal is provided via our connected wind data service partner, meteoblue, and is the most accurate dataset available for wind data. Additionally, we now have the ability to derive seasonal wind roses from the new dataset, which will be coming in future months.

2. Humidity Modeling

Humidity plays a big part in thermal comfort analyses, and SimScale can now account for it.
Humidity modeling can be hugely important to internal thermal comfort studies for identifying where condensation might occur as well as for analyzing indoor environments where tightly controlled humidity levels are critical, such as concert halls, storage facilities, or indoor farming.

3. Realizable Turbulence Model

For urban wind applications, this turbulence model is declared as the preferred one by several best practice guidelines (COST Action 732) as well as wind engineering guidelines, such as the City of London (CoL) Wind Microclimate Guidelines (ref), if a steady-state CFD simulation is run.
The realizable k-epsilon model is now available for the Incompressible analysis type on SimScale within the top-level analysis type settings.

Use Case & Benefits

Urban pollutant dispersion analysis using the Incompressible analysis type on SimScale for enhanced result accuracy compared to the standard k-epsilon turbulence model.

4. Solids included in solar radiation

It is now possible to model solar loads in CHT analyses with models that have both fluids and solids included.

Use Case & Benefits

• Solar radiation can play a large factor in thermal comfort, and the ability to model it with solids included increases the overall simulation accuracy.
• Solid walls at the boundaries of the flow region don’t need to be modeled with specific boundary conditions defining the conductivity and material thickness but can instead be assigned the specific material, and their thermal properties will be correctly accounted for
• Solids inside the fluid domain simply need the correct material assigned.

The current limitation is that the solids in a CHT analysis with solar load can not be semi-transparent, so windows and facade glazings need to be modeled with an appropriate boundary condition.

5. Joule Heating

Q1 sees the release of Joule heating. This works with Direct Current (DC) applications and is integrated into the CHT and IBM analysis types. Applying Joule Heating to a part, or parts will cause them to heat up realistically.

6. Immersed Boundary Method (IBM) external flow domain flexibility

IBM (Immersed Boundary Method) now allows for different external flow domain positions. This means that we can position the test unit on the floor, wall, ceiling, or suspended in the middle.

Use Case & Benefits

Useful for:

• Lighting, as it can be positioned anywhere
• All electronic assemblies as they are often designed with multiple orientations and installation positions in mind

7. Simplify/heal bodies with Surface Wrapping

Parts are sometimes too complex to work with and so can be simplified with Simscale’s CAD tools. This can currently be found under ‘surface wrap’.

• Faulty models can cause meshing problems, and fixing/simplifying parts in advance should avoid this
• Sheet bodies can be difficult to mesh – wrapping them and forming a solid can solve this too

8. Multiphase

One of our highest-requested features is now in production!

This feature introduces a proprietary multiphase capability within SimScale, with industry-validated methods for high accuracy and fast simulation turnaround for rotating equipment, hydraulics, and industrial equipment simulations.

Benefits

• Volume-of-Fluid algorithm with proprietary high-order reconstruction scheme that captures sharp interfaces well
• Comprehensive physics, including heat transfer and surface tension
• Handling of realistic fluid and material properties
• Binary tree-based meshing and automatic local time stepping for proven stability for complicated geometries

Use-Cases

• All types of turbomachinery, rotating equipment & flow control simulations
• Hydraulic engineering / AEC applications (reservoir, dam gate, etc.)
• Industrial mixers, aeration tanks, tank filling simulations
• Marine applications (static ship hydrodynamics, propulsion systems)

9. Relative Velocity

For correct visualization of rotating equipment flow simulations, it is important to show velocity relative to the rotating blades.

Visualizing the relative velocity field in a turbomachine is crucial as it gives designers insight into the nature of flow within the machine. It is used for creating velocity triangles, which help designers estimate the early-stage performance of the rotating geometry.

Highly demanded by our turbomachinery customers, SimScale will now compute and render relative velocity fields through the rotating regions. This field can be visualized as streamlines, vectors, contours, iso-surfaces, or iso-volumes, and will provide our users greater insight into the flow around rotating components.

10. Boundary condition Visualization Inside 3D Viewer

Boundary conditions are now shown inside SimScale! This has been a long time coming, with (believe it or not) years of effort to prepare everything in the background. Now that it is live, we will continue to iterate on it. If you are actively using SimScale, you will see this evolve over the next quarters.

11. Export result statistics to CSV

The ‘Statistics’ panel can now export all of the data points into a CSV file for external processing. This can be hugely useful with models that contain multiple fluid channels like the one shown below.

Use Cases

• Large organizations that need to control access to content internally
• Small organizations that need to organize content more efficiently

Benefits

• Granular access control
• Intuitive data segregation
• Control of sharing of Team content
• All self-managed

12. Teams and Permissions

Members of teams can now have varying levels of access to content contained within a team, and administrators can manage these settings in their dashboard.

Each member can have view, copy, or edit permissions of the teams they belong to. Each team also includes a setting to control with whom content can be shared (to/from a team):

• No sharing
• Share within the Team
• Share with the organization
• Share with anyone

13. Bilinear elastoplastic material model

Users can now apply bi-linear material behavior via a dedicated interface. Young’s Modulus, Yield stress, Ultimate stress, and strain are simple to define. This change improves non-linear simulation robustness.

14. “Max over Phase” von Mises stress result field for harmonic analysis

Von Mises stress is now available for each frequency in harmonic analyses. This means that as well as identifying resonant frequencies, engineers can also ensure that the structure remains within maximum stress safety limits.

Take These New Features for a Spin Yourself 

All of these new features are now live on SimScale. They are really just one browser window away from you! If you wish to try out these new features for yourself and don’t already have a SimScale account then you can easily sign up here for a trial. Please stay tuned for our next quarterly product update.

The post NEW Features: Multiphase, Joule Heating, Humidity Modeling, Boundary Condition Visualization, and More! appeared first on SimScale.