The previous 10 rules for CFD consultants weren’t in any particular order but if they had been, then I missed the very first rule, the one I thought would be obvious. But as there’s no such thing as common sense I better state it:
Rule #00 - Charge For It
I know that CFD is fun, but what’s not fun is going bust, so charge for your work. There’s rarely an excuse for doing free work, if you’re generating good marketing material then maybe it could be a freebie. But if a project is important enough that your client won’t let you shout about it, then it’s important enough to be paid for. Free work is rarely a good bait to catch the kind of clients you deserve.
10 Rules for CFD Consultants
These aren’t commandments or the law, just a few rules of thumb that we go by when we’re consulting. You could swap out CFD for FEA, or any other flavour of CAE, it just happens that we do CFD, so that’s how it’s written.
Rule #1 - Thou shall not CFD for CFD’s sake
If you’re not making the topic of your project faster / cheaper / lighter / stronger or just better - you might as well not be doing it, let alone charging for it;
Rule #2 - Tell them how much it’ll cost
Your customer probably isn’t interested how many days you spent on their problem or what that equates to in your arbitrary day rate. They likely have a budget to spend on this and they want it delivered within that budget. “Sorry we haven’t fixed your problem but we have spent all your money” isn’t a good conversation to be having near invoicing time;
Rule #3 - Ask questions…..
……then shut up & listen to the answers. Your client knows more about his problem than you. He probably doesn’t know what the answer is but then neither will you if don’t listen.
Rule #4 - Cheap is not a selling point
Hot on the heels of Rule #2, don’t compete on price - period. If all you’ve got is that you’re cheaper than the next guy then perhaps it’s time for a rethink or to find a contractor position?
Rule #5 - Thou shall not covert thy neighbours hardware
As a wise person once said, “it’s not how big it is, it’s what you do with it that counts”, likewise don’t brag if you’ve got big ‘uns.
Rule #6 - Thou shall not use vanity metrics in marketing material
Just like your count of Twitter followers or Facebook friends tells me nothing about you, other than you like social media. The number of cores, terabytes of RAM, billions of cells, or petabytes of storage you run says little more than you run on computers & I could’ve guessed that bit.
Rule #7 - Your time is more valuable than compute time
If a computer can do it then let a computer do it, they scale much more easily than you. Systematise then automate - simples.
Rule #8 - Don’t blag your clients
If you’re in a client meeting thinking “we can’t model that” - say so. Your client expects that you know what you’re doing, not that you’re going to figure it out at their expense. There’s a good chance that it’s not going to end well otherwise.
Rule #9 - Use the right tools
You love CFD, I get it, but sometimes some things are easier to physically test than they are to model. Know when it’s one of those times & do some physical testing or find someone else who can. Rules 1 & 8 also apply here.
Rule #10 - Pass it on
Be the guy-who-knows-a-guy & build a network of like-minded consultants (not just in your field). When you come across a client you can’t help, you’re too busy or it’s not your bag, then pass them on to someone who can - what goes around comes around.
Well we’ve said it now….
@ufo_cfd - we use it everyday for doing consultancy, but it’s not public yet. We’re hoping to let consulting customers loose in Q4.— CFD Engine (@CFDEngine)
There’s something to be said for announcing things before you do them, it brings with it a layer of accountability that’s missing when you keep it all to yourself.
I can’t imagine we’ll be flinging the doors open on October 1st, but hopefully we’ll get some people rolling before the year is out.
Does that count as a private beta?
I was at Develop3D Live in April this year and was surprised by the spread of opinion in the simulation session regarding cloud simulation. Whether it was using one of the increasing number of cloud simulation platforms (not just CFD) or rolling-your-own on the cloud of your choice there was very little support for cloud simulation from the users in that session. Potentially bad news for us if this reflects the engineering community.
Having chewed it over, here are my top three reasons why this might be the case:
Lack of a champion - this stuff is new & it’s pretty much only vendors who are shouting about how much sense it makes to offload simulation to the cloud. If respected practitioners were also evangelising then maybe more people would be listening.
Lack of a clear definition - I haven’t seen a definitive description of what “the cloud” is & that lack of clarity seems to have led to a lack of meaning. Whilst ever people don’t know what “the cloud” means they won’t be able to relate to it & it’s unusual for people to embrace things they don’t understand.
Vested Interests - very few people come to these sort of debates without baggage, whether that be that they’ve spent thousands on big iron, or years cultivating their way of doing things, sufficient that a contrary opinion just won’t fly, especially if it’s coming from a vendor (these points are becoming recursive ;-)
I think that the majority of naysayers probably haven’t taken the time to try to weigh up the costs and benefits of cloud simulation. The favourite arguments against, security, transfer time etc. are soluble problems rooted in process, practices and procedures. Your existing workflow might not be a perfect fit for the cloud, but it’s pretty rare to find a process or workflow that wouldn’t benefit from a bit of re-imagination. That effort might not lead to the cloud, but it should improve efficiency and that’s got to be good (unless you’re a charge-by-the-hour consultant where efficiencies don’t touch your bottom line ;-)
Cloud simulation isn’t going to be a fit for everyone, but if you don’t take the time to look you won’t see what it could do for you.
CFD Consultancy, Tweaked.
As the backbone of CFD Engine gets more complete we’re finding ourselves using it in-house more & more. Therefore we’re announcing CFD Engine Consulting, it’s CFD consultancy but not as you know it.
We want to work with you to create a bespoke CFD solution tailored to your business, based on our CFD Engine platform. Whilst we could volunteer to churn through your CFD runs for you, we think you can do that yourselves given a sufficiently easy to use platform. Instead, we’ll get to know what you need your CFD to do & endeavour to provide that for you through CFD Engine.
We’ll become your CFD team on a nominal one day per week basis over the course of a year. During that time we’ll get to grips with your specific needs, produce the initial CFD for you so that you’re productive immediately, and tailor CFD Engine so that you can eventually take the helm.
This way at the end of the deal you aren’t left with just a bunch of CFD runs, you’ll have a CFD workflow that has all the benefits of CFD Engine (pay-as-you-go, intuitive & scalable) but tweaked to fit you.
We’re only looking for a handful of these relationships, hopefully spread across a few sectors. By limiting the numbers & extending the time we work together we can really get to the nub of your problems, truly understanding what you need & how you work.
We’re sure that we can offer much more value like this, as an extended member of your team, than if we were just a traditional consultant who has both eyes on the clock.
Do you think that our “Consultancy Plus” model might be a fit for you? We’ve had some good interest already, perhaps you should be one of the few?
If you aren’t already signed up then drop your email in the box to the right (or visit the homepage) & then drop us a line & we’ll see if we can work something out.
Where is the value added?
First off, were not dead, we’ve just been busy. Busy doing code-y, behind-the-scenes type stuff that, as much as we tried, we couldn’t make sound exciting enough for a blog post.
On a different note, we’ve been having a few thoughts about our target customers & we think that we have something to offer to existing CFD engineers, particularly CFD consultants or freelancers, that we hadn’t appreciated at the outset.
We think that any appeal to these potential customers will depend on where they think the value is added in the usual “design-mesh-solve-postpro-analyse” cycle? If you’re a CFD consultant/engineer, where do you sit in that cycle & where do you add most value?
We think that the bulk of the value of CFD should come from the analysis/design phases. Couldn’t the remaining phases be automated without detracting from the overall value? Would you see that automation as freeing you up to add more value or taking away some of your treasured meshing time?
What do you think? We would love to know. Leave a comment, tweet us or send us an email - we’re all ears.
We’re trying out Dropbox to make file transfer pretty much frictionless. Don’t worry, we’re not planning on hogging all of your free 2GB, we’re just using it like the transporter room. None of that FTP nonsense, just drop your geometry files in the box & let us “beam them up.” We’ll check them in & email you when they’re ready to go (or if we pick up a problem with them). On the flip side, if you want something from us, we’ll drop it in the box and you’ll have it automagically appear on your machine - thank you Dropbox.
We’ll probably roll our own fancy uploader in the future, but for now this works a charm. Do you use Dropbox already? Is this a convenient option? You can always use our sign up link if you’re a Dropbox newbie ;-)
Who is CFD Engine for?
If you’re wondering whether CFD Engine is for you then perhaps this will help you puzzle it out?
If you end up in the green box & aren’t on the mailing list then it’s probably worth dropping your email in the box in the right sidebar or over at the homepage. Either way, you’ll be letting us know you like the sound of it & we can keep you posted along the way.
If you’re already on the list & you ended up in the green box - congrats, you’re in the right place.
Let us know if you end up in the orange box, we might be able to hook you up anyway.
Less complex doesn’t mean less capable.
Just because some things are complex doesn’t mean they can’t be easy to use, does it?
We’ve come across a bit of a perception that simplifying the user experience (UX) for CFD comes at the expense of capability. I can see where this comes from, but I see no reason why it has to be the case.
CFD is complicated stuff, but so are lots of things, so is mobile telephony but that doesn’t stop my iPhone being intuitive & simple to use.
We aren’t Apple, but we agree with them “simplicity is the ultimate sophistication.”
Why do CFD in the cloud?
CFD eats CPU cycles, there are no two ways about it, the better the simulation you want, the more crunch power you’re going to need. CFD that runs on your desktop may seem like a great idea at the time but sooner or later you’re going to need that desktop to do all the jobs you originally bought it for & there’s only so much coffee you can drink whilst waiting for your latest & greatest simulation to finish.
Offloading your CFD to the cloud makes sense to us for a whole heap of reasons, most of which we’ll leave for another post, but we can pretty much boil it down to these two:
- Price - if you aren’t using it then you aren’t paying for it - and who likes paying for something they don’t use?
- Scalability - forget queueing systems & waiting for simulations to start - whether you need 1, 10 or 100 simulations, just fire up an instance for each simulation & get cracking.
It might not fit with the current CFD paradigm but there’s change coming & we want to be in the leading pack.