Bottleneck hunting

Hey there 👋

It’s Robin from CFD Engine & it’s a short one this week – I’m hunting bottlenecks & I could do with your help.

Which bits of your current CFD workflow slow you down? Which bits do you dread, but can’t avoid? Which bits would you worry about, if your next project was 10x bigger? Or if you had five new projects all start at the same time?

I’m keen to hear your thoughts. I know my bottlenecks, but I’m wondering if there are any universal ones that span different CFD workflows.

Drop me a note.

In my case…

I’ve tweaked my process to let me crank through cases as a solo-consultant. But that process needs a refresh – the cloud CFD platforms are eating my lunch when it comes to time-to-first-result & I can’t have that 😜

I’ve put a lot of effort into reducing my time-to-next-result but I’ve been ignoring the time taken to start a new project & build a baseline model.

It’s always been a time suck & (for me) it still is. It includes lots of messing around in CAD & lots of OpenFOAM template tweaking.

It’s mostly a bunch of once-per-project tasks, which aren’t particularly taxing by themselves (& that’s why I’ve never automated them) but they increase the time a client has to wait for their first data set & I’d like to shrink it.

So I’m currently noodling on what I’d need to do, what bottlenecks I’d need to remove, to be able to turn around a new project baseline in 24hrs? 🤔

Everyone else…

Which made me wonder – what CFD bottlenecks do others have & which ones span our different processes?

What takes too much of your time? What takes too much of your computer’s time? What tasks still need manual intervention?

How are they split between one-off (once-per-project) tasks & those that have to get done for every. single. simulation.

They might not come to mind straight-away. They might have faded into the background, after all they’re just “stuff you have to do”.

But try this thought experiment – what would you worry about if your next project had 10x more runs? Or you had to do it in half the time? Or you had to start 5 new projects all at once?

If you had the chance to automate something before any of that happened, what would it be?

Mine would be something geometry-clean-up related (hence my recent wrapping obsession) & some project template automation – things that could significantly reduce my time-to-first-result.

I also wouldn’t mind spending less time wrangling spreadsheets & preparing reports, but that can wait for now 🫣

Drop me a note with yours – where are the bottlenecks & pain points in your CFD process?

I’m keen to see what common ground exists across our different CFD domains – maybe there’s something we can attack together 🤷‍♂️

Until next week, stay safe 🙏