Talking Cfd

It’s the last show of Season 04 (time flies when you’re podcasting) & Simon & I are back chatting about the most recent shows, the season just gone & our thoughts/plans for 2020: - Simon reveals his ultra-personalised, CFD-powered, custom surfboard & custom wave idea #futurist

Time for the last insight episode of the season - I hope you've enjoyed them as much as I enjoyed doing them? In today's episode I had the pleasure of talking to Riccardo Rossi about combining his experience in fluid dynamics research with his passion for surfing and the intricacies of applying CFD to the design of surfboards. We barely scratch the surface of this fascinating topic, but Riccardo introduces us to some of the challenges, including: - relating performance in the "predictable" world of CFD with that on the unpredictable ocean wave - using CFD to relate the "feel" of a board to it's fluid dynamics - whether the insight from CFD can make you a better surfer - using CFD as an attention-grabber in product marketing ## Shownotes - Connect with Riccardo on LinkedIn - https://www.linkedin.com/in/riccardo-rossi-a58a692b/ - Find out more about Riccardo's company - Red Fluid @ http://www.red-fluid.com/ and the work they've been doing in this area - Follow them on Facebook (https://www.facebook.com/Red

I've wanted to have a guest on the show to chat about artificial intelligence & machine learning in relation to CFD for a while now. It keeps coming up in other discussions but this is the first founder interview where I get to scratch the surface of this huge topic. Today I’m talking to Lieven Vervecken one of the founders of Belgian CFD startup Diabatix - mixing AI/ML with generative design, optimisation & traditional CFD methods to design a new breed of heat sinks. Lieven fills us in on: • their PhD-to-Business journey & the importance of early contact with (potential) customers • how they operate their business somewhere in between software development house, consultancy & software-as-a-service - taking the best of all worlds • what they mean by the baggage-laden terms, generative design & artificial intelligence • why they’re more than just a fancy optimisation tool Check out the kind of thing we’re talking about here - https://youtu.be/D52fwUTf0GM Shownotes • Check out the Diab

It’s the last episode of 2019 & we’re introducing a new section of the show called “grumpy old CFDistas get a bit bah humbug about stuff” which may not become a regular feature. Some of the things on our radar this episode include… • Cybertruck simulations on LinkedIn • In awe of Patrick’s CFD skills & PreonLab’s renderings • the “importance” of aerodynamics for battery electric vehicles (& the importance of not getting too “rant-y” while podcasting) • AWS clusters vs. SPH & the fight for control of Simon’s to-do list • a few other shiny nuggets from the AWS conference r.e. CFD Links to some things we mentioned • NASA Turbulence Modelling Resource - https://turbmodels.larc.nasa.gov/ • Discussion of Formula1® doing CFD on AWS with a few details - https://www.zdnet.com/article/formula-1-turns-to-aws-to-develop-next-gen-race-car/ • Slightly less enlightening but more entertaining talk from Rob Smedley about the above - https://www.youtube.com/watch?v=eBX7lPk5qmA • Peter DeSantis talking about

In today’s episode we welcome back previous guest, friend of the show and all-round CFD expert, Patrick Hanley. The last time Patrick was on the show we spoke about his experience as an independent CFD software vendor. This time we’re hearing about his adventures designing “flying cars” as part of the mobility startup Terrafugia. We cover a lot of ground, including, but not limited to: • How speed of execution is critical in the sprint to market & how different CFD tools can play into that (including a huge shout-out to Pointwise) • The value of understanding what’s gone before, not only in terms of aircraft development, but also code development & being able to mine those archives to give you an unfair advantage today • The importance of being able to select the right simulation tool for the job, being able to get multiple approaches to agree & the alarm bells that should go off when they don’t Shownotes • Connect with Patrick on LinkedIn - https://www.linkedin.com/in/hanleyinnovations/ * Che

Today I’m chatting with Jens Cornelis, one of the founders of FIFTY2 Technology – CFD software developers based in Freiburg, Germany & home to the particle-based CFD code PreonLab. Jens shares some of the story of FIFTY2 Technology & the development of what became PreonLab, including: • why PreonLab is about more than just pretty pictures (but boy, do they make some pretty pictures) & whether using a CFD package can actually be “fun” • how it was a little “too physical” for Pixar • how they “crossed the chasm” & their journey from academia (University of Freiburg) into business • the benefits of partnering with a distributor & having an industry focus A couple of “must watch” videos to get a better handle on what we’re talking about: • The “wipers & rain” video I mentioned - https://youtu.be/J-tvjQVQ3hQ • The water wading simulation Jens mentioned - https://youtu.be/zku5AJ-_NNI • Another video featuring windscreen wipers, this time dealing with snow - https://youtu.be/YrWQsrBNDBg

Simon & I celebrate 50 episodes of this podcast, share our thoughts on the last 2 shows & disappear down a few tangential rabbit holes. It’s just your typical social episode really, including: • getting it done - having ideas & then executing on them (even without a coffee boost) • different ways to make CFD more accessible (AirShaper & Discovery Live) - a.k.a. “getting your 80%” • how “democratisation” might have got lost in a cupboard with “digital twins” & “the internet of things” • SPH (whatever that is)

In our third insight episode, Kurt Smithgall, formally of S20 Design, takes us into the fast-flowing world of whitewater-park design or, to borrow a phrase from Kurt - “recreational hydraulics.” Kurt gives us an insight into how CFD is helping in the design of whitewater parks & some of the challenges therein - including: • mitigating the risk of building a water park that doesn’t work • dealing with uncertainty in geometry & flow conditions • scale - both in terms of the mesh/physics, but also the length of the transient simulations • the importance of domain expertise & whether you need to be a paddler to do this job • the future - mixing CFD post-pro & virtual reality to “paddle” features before they’ve even been built Shownotes • Connect with Kurt on LinkedIn - https://www.linkedin.com/in/kurt-smithgall-a4802b66/ • Find out more about S20 Design - https://s2odesign.com/ Check out these two video to get a better idea of what we discuss: • Boise River Park - https://bit.ly/boise-whitewate

You may have noticed that I’m a big fan of doing CFD in the cloud & a big fan of interesting CFD companies, so today’s interview is a particular favourite. I’m talking with Wouter Remmerie, founder of AirShaper a cloud CFD platform aiming to make aerodynamic analysis more accessible & affordable. Wouter shares his experiences, thoughts & stories on: • his sports car ambitions & death-defying aerodynamic testing • “black box” CFD for aerodynamics • AirShaper’s unique selling point in a growing cloud CFD market • their LinkedIn videos and digital marketing for CFD Shownotes • Check out the AirShaper website for more information, including: - videos - https://airshaper.com/videos - a sample project - https://app.airshaper.com/projects/sample/fixed-wing-drone - try with your own geometry - https://app.airshaper.com/projects/new • Connect with Wouter on LinkedIn - https://www.linkedin.com/in/wouter-remmerie/

In this week’s round of CFD pub chat Simon & I share our takeaways from the recent episodes, plus what we’ve been up to since last time, including (but not limited to): • fancy optimisation stuff • going (or not going) to conferences • sharing your work & secret customers

In the second of our "insights" series Brian Willis, of specialist engineering consultancy Astrimar, is taking us deep into the world of CFD for oil & gas. Not deep into the technicals, but deep under the ocean & into the mud. Hold onto to your headphones, we’re going downhole for a spot of plugging & abandonment. Brian gives us a flavour of what CFD looks like in this particular niche - what’s tricky & what’s commonplace, including: • the extreme scales that dominate their models • how the levels of uncertainty in the geometry, environment & the flow conditions, impact the kind of simulations & techniques employed • how this sector might be a great fit for the lesser-spotter digital twin • how, despite the technical challenges, CFD is proving a feasible & economical alternative to experimental testing. Particularly when it comes to developing & proving new technologies Shownotes • Connect with Brian on LinkedIn @ https://www.linkedin.com/in/brian-willis-20847343/ or send hi

You wait around for a CFD consultant interview & then two come along at once. This is the second “founder” interview of the season & the second CFD consultant - duff scheduling, my bad. Today I’m chatting with James Linfield, one half of the dynamic duo behind Navier Flow Consultants – a (relatively) new CFD consultancy based in Nottingham, UK. James gives us the benefit of his recent experience & shares his take on: • the transition from employee to “boss” • choosing a niche to focus on • advice for those looking to start a CFD consultancy in today’s market • their approach to the cloud & why they have a stack of machines keeping the office warm • the importance of networking, for everyone, but particularly for those looking to start a new businesses Shownotes Check out Navier Flow Consultants @ https://navier-flow-consultants.com/ Connect with James on LinkedIn @ https://www.linkedin.com/in/james-linfield-51a37175/

This is the first of our new style of shows where I talk to CFD practitioners / experts / “those that do” to find out what CFD looks like in their world. We typically start these episodes with an overview of how CFD gets used in their industry, then zero in on some of the uniquely challenging aspects and dig a little deeper. In this show I’m talking to Stefano Capra of engineering consulting giants Ramboll, about some of the ways CFD is used in the design of the built environment & the challenges faced in that sector. Stefano gives us his insight into what’s challenging, what’s “normal” and what’s on his CFD wish list, including: • why CFD in the built environment is about more than just pedestrian comfort studies • how CFD is becoming increasingly accepted as an aid to the established experimental methods, as opposed to a replacement • how CFD routinely helps save millions of dollars in construction costs (and thousands of tons of CO2 ) by accurately predicting the wind loading on new designs Links

It’s the first episode of Season 04 & we’re back with the usual problem - how to condense 25+ years experience running a simulation business into a single episode. We fail (as usual) but it still makes for an interesting show. In this episode I’m talking with Eric Miller, one of founding team of PADT – an expert simulation consultancy which first opened its doors 25 years ago this year. Eric shares his thoughts on: - migrating from an internal research group at a large company to an autonomous small business - the evolution of their computing from SGI Octane through the Pentium revolution and on into the cloud - whether the current crop of open-source tools make commercial software a harder sell? - where the simulation “low-hanging fruit” has gone & how consultants need to be extracting higher value from their simulation tools - building the kind of business where it’s a pleasure to come to work in the morning Links ------ • PADT – https://www.padtinc.com/ • Connect with Eric on LinkedIn – https:/

A quick heads-up that Season 04 of Talking CFD is inbound, with new episodes starting 24th September 2019. Please make sure you're subscribed in your favourite podcast app & following us on the socials to make sure you don't miss a thing. www.talkingcfd.com

Someone once described these episodes as Simon and I "noodling it out" - you get to listen to us working through our thoughts and (hopefully) coming to some conclusions or at least giving you something to think about. And that description probably best fits this episode. We started out with a provocative-ish question, left over from the NAFEMS episode & expanded from there. We didn't put this one to bed, but instead we opened up the question of what "doing CFD" really means. Question of the day (not the one in the show title) but "Do you DO CFD or is it just a tool in your toolbox?"

We don't get out much. So when one of us does manage to break away from our workstation and get to a conference, it's worth talking about. Simon gives us the lowdown from the NAFEMS UK conference including his key takeaways, what themes stood out & what was missing. Question of the day "which conferences are worth getting out of the office for and why?" Conference Link – https://www.nafems.org/2018/uk/

In this episode we break down some of the NASA CFD Vision 2030 document, to see if our vision of CFD in 2030 looks anything like NASA's (spoiler alert - not so much ;-). We pick out some of what we feel are the most important topics & put them in context of where we (Simon & I - not the industry as a whole) are now & where we think they're going. Plus I expand on why I think this document is actually just a shopping list for a cash-strapped NASA. NASA CFD Vision 2030 Study: A Path to Revolutionary Computational Aerosciences Read the study - https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20140003093.pdf

We've still got our head in the the clouds this week, covering some of the topics we didn't touch in the last episode. Including –but not limited to– addressing your cloud security concerns by having a cloud-in-a-box delivered to your front door (not exactly elastic computing). The data transfer conundrum & why I think that you need to think a little differently when it comes to putting a cloud CFD workflow together. We also go on to talk about the new crop of cloud-only/cloud-first CFD providers. Is it a problem if you don't control the version of the code you're running? How might you deal with vendor lock-in when you never get a copy of the code? This week's question is - "What's stopping you from cloudifying your CFD?" Checkout The Beginner's Guide To OpenFOAM on AWS @ http://learn.cfdengine.com for a cloud-first CFD workflow on Amazon Web Services