Blue Ridge Numerics has been pushing the adoption of Computational Fluid Dynamics as part of the mainstream design process since its inception. Al Dean looks at the latest release to see where the state of the art lies.

Interleaved V9 intelligently assigns mesh sizes based on curvature, geometric gradients, and proximity to neighbouring features. Optimally sized surface and volume mesh ensure faster mesh generation and higher-quality simulation results without any user input.

Blue Ridge Numerics, along with the majority of other CFD vendors, has been banging the ‘mainstream fluid analysis’ gong for many years. Its CFdesign application provides a platform with which users can access advanced fluid and thermal simulation tools whilst benefiting from CAD associativity and the ease of use ethic that the mainstream market demands. While there are all manner of enhancements and other new functionality, what I’m going to do is look through the nine major updates for the 9.0 release – and how they apply to both existing users and those looking to adopt a CFD system as part of their product development workflow.

Speeded by Rules on Parts, Automated Mesh Sizing, and the Accelerant Solver, this automotive fuel injection intake manifold was taken from MCAD to first-pass flow and thermal simulation in less than 20 minutes. Forty-five minutes later high fidelity results were ready for an interactive design review.

The first big news, perhaps most importantly for existing users, is that the calculation solver that underpins the system has been updated. With the help of proprietary CPU optimisation algorithms, the new Accelerant Solver has been developed to make the calculation process much more efficient. Most of these benefits become obvious when you’re working with very large studies. For example, while a smaller study might be completed 40% quicker, if you’re working with extremely complex studies, tests have shown that speed improvements are of the order of 2,000%.

Another key enhancement is those made to the meshing tools, which manifests itself in three key areas. While the move towards mainstream simulation has seen the automatic meshing tool become the favoured method of mesh creation, there are many different takes on the process. Blue Ridge’s approach is on two fronts, which both automate much of the process, but also make the mesh creation process more efficient. Firstly, there are new geometry diagnostic tools that inspect a study and highlight areas it perceives to be difficult to mesh – ensuring that before you start the analysis process, you’re aware of the quality of the data you’re working with and any potential problems. Moving on from this, V9 also sees new tools to automate much of the meshing process. This initially carries out a topological interrogation of your model to select the most appropriate element sizes, but perhaps more critical to the creation of a stable mesh is the refinement process. This looks at the topology of the model and in addition to finding a ‘global’ element size refines the mesh – adding more elements where needed and decreasing the mesh in areas that do not impact the analysis accuracy. For those used to structural analysis, this usually revolves around handling small features, but within the CFD world, this is extended to handle small gaps and geometry’s proximity to other geometry – the reasons for this are to give the system enough elements to accurately predict how fluid will interact in such small voids.

Finally on the meshing front, this release sees the introduction of intelligent tools for handling extruded forms. In many cases, there are parts or geometric features which are prismatic in nature, so it makes sense to take advantage of techniques to optimise their meshing and hence their calculation. CFdesign now identifies regions that are suitable for extruded meshing – where a constant cross section is present – and uses a specific form of element (wedge) to describe those forms. While the applications for this technology are widespread across all industries, it’s particularly useful for parts with high-aspect ratios, like heat sinks, fan blades, and sheet metal enclosures, as it greatly reduces element count (and as such, calculation times).

Also on the automation front, the new Rules on Parts functionality should assist the user in speeding up the set-up, boundary condition definition process. Essentially, this allows the user to generate filter-type rules which inspect a CAD part’s file name upon import and automatically a material definition based on the name of the part. This is the first release of this technology and while some organisations don’t define their part names based on material, if you can categorise your parts to single out specific component sets, you can automate the materials assignment process – this is pretty common within the electronics space, so will be a major benefit to groups working in that field.

At the launch event, it was hinted that this might be extended in subsequent releases to allow the system to interrogate a CAD part’s custom properties to extract material information. This would make the use of this particular bit of technology much more beneficial, particularly if you’re importing PCB parts. In combination with the new PCB characteriser tools, this will allow you to define a standard set of materials that hold the properties of each copper and dielectric (FR-4) layer and then have the system assign these properties to a large number of PCB parts using the rules functionality.

Also on the electronic component front, there is a new compact thermal model tool to allow the definition of two-resistor microchips for heat transfer simulation. Only two parameters are required and once you’ve defined the Theta JB and Theta JC values for a ‘material’, the board junction, case temperatures are calculated and heat transfer between those (such as junction and board, junction and case) can be derived. The system supports BGA (ball grid array), PBGA (Plastic ball grid array), TBGA (taped ball grid array), FC-BGS (flip chip ball grid array), QFP (quad flat pack), PQFP (plastic quad flat pack), NQFP (no-lead quad flat pack) and SOIC/SOP (small-outline/small out-line package) microchip configurations.

Elsewhere on the intelligent part creation front, this release also allows the user to define fans much more intelligently. Whereas in previous releases you would have to import the full explicit geometry for a fan component for use in a study, from this release onwards you can now define it using a standard fan velocity profile. This allows you to incorporate an idealised fan component in to your simulation quickly that hold the correct velocity distribution data, which would traditionally have required some rather intensive rotating region analysis.

The final major addition for this release is Solar Loading. This allows the user to incorporate transient electromagnetic heat transfer from the sun into their analysis studies, which can affect the performance of outdoor security cameras, climate control systems, power utility transformers and such. This feature includes radiation through transparent media and even shows shadowing based on the movement of the sun. Simulation set up is simple: specify time of year, time of day and location on the globe using the database in CFdesign, or assign specific latitude/longitude coordinates.

In conclusion

There are many vendors looking to bring Computational Fluid Dynamics to the mainstream market. This is where the real action is going to be over the next few years and all are looking to reach a point where the average designer or engineer is using fluid and heat simulation tools as part of their daily work process. The concept is that they not only generate full geometric descriptions of new products within their workhorse 3D CAD system, but to also accurately predict their performance. Within that CFD market, there are many different approaches being taken. There are those with a history in traditional, expert-driven CFD codes looking to commoditise their high-end products into easier to use packages by providing mainstream interfaces to their existing code-base. There are also those that have used a similar background but have chosen to develop new systems from scratch for this mainstream market. Then there are those that set-out their stall in the mainstream camp from the very beginning.

Blue Ridge Numerics has always been in this last category and this is reflected in CFdesign. The system is CAD-integrated in a specific sense, in that it reads native CAD data, imports both geometry, assembly structure and now other information (such as material definition), then allows you to work in an environment that’s tailored to the CFD workflow. This differs from other vendor’s approaches which integrate directly into the CAD system interface – this isn’t a criticism or praise, just a statement of fact. What’s most important is that you have the ability to take the master 3D data from your CAD system and carry out the simulation that you need and feed the results back into the product development process.

This release sees a wide range of updates and enhancement made to the core of CFdesign and there are many other updates that we haven’t got the space to discuss here – what you’ve got here is a brief look at the major work done for V9. There are new tools that allow you to take your CAD model and prepare it for the simulation process in a much shorter timeframe. There are new tools to help define fluid flow or heat transfer specific entities (such as PCB components or fans) more quickly and efficiently. And finally, the new Accelerant solver allows you to calculate the results in a much shorter time. What all of these updates mean to your business is that you can get the results out of the system, gain a better understanding of how your current product iteration is going to perform when built and where you need to make changes or improvement – all within a much shorter space of time than has traditionally been possible. And that means two things, higher quality products finalised and into the market quicker and gain competitive advantage – and that, is what good product development technology is all about.