Engineering Software
Engineering software – it’s getting more bells and whistles every year, and easier and easier to use.
It’s even more accessible now, with subscription pricing becoming the norm, meaning you can own an incredibly powerful suite of software for a reasonable price, for just as long as you need it.
This is great, but I’m afraid that it might come at an unexpected price.
What do I mean?
Not to brag about my age 😉, but when I started my career we did mostly hand calculations from first principles, and when we did use software, it was text-based coding to build and load the structural model, and reams of printout to read and search through for the results.
Engineering Software – How Simple are they to Use Nowadays
In contrast, today, we have software with which a reasonably intelligent teenager would be able to model something and perform an analysis. I’m not sure I would believe the result… but the teenager would!
The software is so simple to use now, here’s a reasonable analogy – a computer with three buttons, each assigned a function:
- BANANA: Generate model
- ROOTS: Create loads and boundary conditions
- PEANUTS: Print results
Press to generate model
Press to create loads and boundary conditions
Press to print results
In such a case, we would not need engineers, but monkeys. After all, it’s relatively easy to train a monkey to first press the button with the banana, then with the roots, and then with the peanuts.
In the world of “the best software that is always right” there is only one way to make a mistake… and that is to press banana, banana, peanuts.
Luckily, things are not as simple (at least not yet!), but I think we can already observe degradation of engineering skills everywhere due to highly complex software reducing complicated tasks to cookie-cutter style objects and templates, containing all the assumptions the originator made but not necessarily passed onto the user.
Check out this steel structure that would have been first modelled on awesome structural engineering software, before being modelled and shop-detailed on amazing CAD software:
Perhaps you’re thinking, this is ridiculous, only qualified structural engineers use structural engineering software. You’d hope so, but yeah nah!
Austrand Introduces graphical input to modelling
In the late eighties, I’d just finished university, and a new structural software called Austrand was released to the market. It was awesome! It copied the concept of visualisation that Steve Jobs introduced to his Macintosh, and introduced graphical input to modelling. I convinced my boss to buy it.
Then, one of our clients, let’s call him Tom – a mechanical engineer – was impressed when I showed him a graphical model (2D of course!) of a structure, and before his eyes changed the column size from “this” to “that”, reduced the deflection and saved tonnes of steel.
What did Tom do? He bought the software! We lost a client, but he didn’t gain a structural engineer! (Later he did – he built himself a very respectable EPCM company, with quite a few structural engineers, and occasionally they consult with us over complex issues).
How to become a 'monkey' of your models
Here’s a thing often forgotten with models. Not just software models by the way, all models (mathematical solutions).
THEY ARE WRONG!
If I built a steel structure and subjected it to a prescribed load, it will respond in a certain way. Let’s say the maximum stress is 200 MPa and it deflects 20 mm.
Did my model predict this correctly?
If I model that structure in some software a certain way, I might get that result. I can change my model a bit to “better reflect real conditions” and I might get closer. I could change my software to FEA and get a different result – perhaps even closer, or why would I try? I could apply a revised code formula that shows my safety factor against collapse has improved.
So what? All the while, changing the model, changing the method, changing the code from an Australian Standard to equivalent Eurocode, the actual structure hasn’t changed. It will continue to obey Mother Nature while we try and find faster and better and more accurate methods to predict reality.
When we understand this, we become the master of our models.
When we don’t, the software is our master and the design is often crap!
A good photographer doesn’t shoot in “Auto” and he isn’t hired based on his brand being Canon or Nikon.
Likewise, we choose our methods and the appropriate tools for their intended purpose.
So what's in our toolbox?
- A calc pad (the same one we put there ages ago!) and pencil
- Spreadsheets
- Microstran (for stick models – like SpaceGass, but better!)
- Staad.Pro (for stick models and FEA)
- Strand7 (for FEA)
- SolidWorks (for CAD & advanced FEA simulation)
- We’ve got other software tools too, and we’re not afraid to use them!
I'm here to help you.
✔️ Do you know what software you need to get the result you need to achieve?
✔️ Do you want to untap the superpower within your software?
✔️ Do you want to drop your licences and get Yenem to build you awesome solutions to your problems?
Like this article?
Subscribe to our newsletter to receive a weekly email from Yenem’s director, Dave Meney.