This week’s CFD news includes a wonderful video recording of a presentation on CFD at Imperial College which I highly recommend as your dose of CFD history. There’s a reminder of things happening over at the Cadence CFD blog (that we hope you’ll subscribe to in addition to this one.) We also includes lots of bikes, motorized and otherwise. If you prefer your vehicles with four wheels, shown here is a simulation of a vehicle wading through water as simulated in PreonLab v5.0.
- DEVELOP3D shares their D3D 30, the year’s top tech for product development. Notables [IMO] include Altair SimSolid, Hexagon Recreate, ITI CADfix Viz, and nTopology 3.
- For fans of CFD history with 45 minutes to spare, here’s a video of Dr. Akshai Runchal presenting his personal perspective on The Emergence of CFD at Imperial College.
Geometry & CFD
- The Spatial blog shares 3 keys to translate geometry models from CAD software to the receiving system. The first is units and tolerances. The former can be trickier than you think. Pointwise, for example, is unitless. But on import, especially when the geometry model comes in different files form different sources, we have to be aware of the units in each. Oddly, unit conversion during import can often introduce errors that are related to the latter part of Spatial’s tip: tolerances.
- Bully pulpit: it’s not a CAD model, it’s a geometry model.
- Since attending the IGA Symposium at the USNCCM back in July, I’ve become very interested in immersogeometric methods for CFD.
- The Future of Large-Scale CFD end as follows. “On-prem HPC cannot provide the diversity of options for CFD users to adequately explore the many processors (AMD, Intel, Arm) and accelerators in a way as varied as engineering missions demand. For all these reasons AWS is spearheading the (many) revolutions coming to CFD.”
From that other CFD blog.
As you probably know, Cadence’s CFD blog is another outlet for sharing what’s going on with CFD. Here are some recent posts. Don’t hesitate to subscribe when you’re on the site.
- What Our Interns Did This Summer
- Aerovehicles-4 and AutoCFD-2 Wrap Up
- The Road Toward Next-Generation CFD Software Systems by Charles Hirsch
- How Omnis Addresses Simulation Challenges for Automotive
- What Exactly Is Intern Reading Club? (This post was also shared in on Semiconductor Engineering.)
- Cadence CFD at the SU2 Conference
- Beta CAE released v21.1.4 of their software suite.
- The PHOENICS Summer Newsletter (pdf) includes a summary of what’s new in PHOENICS 2021.
- I don’t usually share market forecasts because I find most of them to be SMH bad and on the verge of begin fraudulent. But this one I’ll share because Digital Engineering has covered it. Manufacturers will spend $2.6 billion on simulation software in 2030. That’s a 7.1% annual growth rate from now until then. Let’s not try to define “manufacturers.” The interesting part of the report is that the growth will be fueled “as the user base of simulation software expands.” In other words, our old friend democratization.
- NASA Glenn seeks an aerospace engineer to do CFD.
- Cadence seeks an Account Technical Executive for CFD.
- TU Eindhoven seeks a PhD candidate for Wind effects on building-integrated photovoltaic systems.
- Tech Soft 3D has several job openings including software engineering.
- owner-racer 5-person yachts, the Persico Fly40.
- elastomeric check valves for medical devices.
- McLaren’s F-1 car.
- driving your car through standing water [which I thought you weren’t supposed to do].
- hydrogen-fueled plant machinery.
- Tokyo Olympic Stadium.
- carbon fiber bicycle wheels.
- marine propulsion systems.
- aero road bikes.
An Inseparable Couple: Form and Color
Of the spousal pair of artists, Josef and Anni Albers, it is Josef who gets most of the recognition. The attention he gets is not unwarranted if for no other reason than his classic book Interaction of Color. Because the creation of colorful imagery is – for better or worse – part and parcel of CFD and computational science, I believe everyone should read Josef’s book to learn how to do it right. But I digress.
Anni’s contributions to art, to the exploration of form and color, carry a significance equal to Josef’s. There is a passionate playfulness in Anni’s work such as TR II shown here in which the inseparability of form and color is on display and within which the distinction between foreground and background is blurred, as paradoxical as that is for the rigidity of the simple triangle.
Read more about the couple’s work on Artsy.
Bonus? How sewers work. [Caution: poop]
Why is in all these plans on turbulence (SRS, AI , ..) nothing like a question whether higher order Lattice Boltzmann, e.g. something in the resolution range of about 0.1 micrometer should bring different results than DNS ? Please assume that a reply with computational costs is not what I mean. I would just try very simple geometric examples just to clear up what happens in the resolution range where CFD should vanish to be valid (to my opinion around micrometers and where Lattice Boltzmann gets more and more important – as in the range already shown.
I am looking forward to fruitful discussions. Could it not be that Lattice Boltzmann effects already get more important at micrometer and somewhat larger ?
Thanks for your question. That is not my area of expertise but I’ll see if I can get you an answer.
I’m not an expert on Lattice Boltzmann, so I won’t comment on that. But you may be interested to read the paper linked below  , which I’ve always found fascinating. It goes beyond LB and considers Molecular Dynamics simulations of turbulent flow, comparing the results to classical CFD (a spectral DNS code for channel flows) and finding very good agreement between the two.
That setup is made for a generic fluid (like you can run a CFD simulation for a generic fluid if you specify everything in terms of Re and other dimensionless groups), but if we take it to be nitrogen gas, the “resolution” of the MD simulation is of the order of 0.3 nanometers.
In general it has been observed in many cases of fluid mechanics (e.g. for capillary pressure in nanotubes, behaviour of nanodrops, etc.) that continuum descriptions of fluids continue to give correct answers way below the lengthscale where you intuitively think they would break down, all the way down to 5-10 nanometers characteristic length. There was a recent paper related to this transition between continuum and non-continuum behaviour in nanoscale fluid mechanics in Ann. Rev. Fluid Mech. that you may also find interesting .
Open Access link: https://arxiv.org/abs/1508.01163
Thanks. You’ve added to my pile of reading 😉
Thanks to the informations . They help very much.
Networking is : You may assume to get informations you are looking for since long.