I am into “industrial design” and general engineering and going forward would like to invest into my hobby as a professional field. Since I am hobbyist I am locked out of many university circles where question like this are just known.
So my question, I am looking for a software that can aid me in much of the design, problem solving and iteration process.
I am wondering if there is software (A virtual testing environment ) that would allow for me test for scenarios like this and many others such as;
Potential collisions of moving parts (gears that need to rotate in relation to one another)
How fast a moving mechanical part needs to complete its cycle for another part to do its job
Heat Tolerance of certain materials (Plastic, wood, glass)
Deciding between materials that are lightweight but also prone to bending or shearing on per project basis
So lets say I was (am not) designing a magnetic snap charging cable like this;
And one of the design goals is for the magnet that holds the cable to the phone to be strong enough for the phone to dangle from the cable without it detaching;
I would need to arrive at a optimal balance between a lightweight magnet and one strong enough to counter the phones weight and velocity.
A physical workshop is always ideal but I’ve been thinking a hybrid method would also be great in terms of reduced costs of machinery and an increased precision.
Ideally, I am looking for single software that I can do the modelling, testing, calculations etc in. The few friends I could ask pointed me to SolidWorks. Is SolidWorks the right tool for me then? If not what other software should I be looking into?
Any help or even just pointing me in the right direction would be greatly appreciated!
Unfortunately, I doubt there are too many shortcuts that can easily replace solid study.
As an example, I use Skyciv for engineering beam analysis at Uni. In order to use Skyciv, you have to have an understanding of engineering of structures or how do you know what the results are and how to interpret them?
SolidWorks is also used at Uni. As a Uni student, you get free use of the tool but it’s used in conjunction with lectures, tutorials and lab exercises.
While simulation packages likely exist for all the things you were after, my experience mirrors Gerard’s, they need a good deal of training and surrounding knowledge, and simulation tools are good for getting quick estimates to things that require a lot of maths, but are no replacement for real testing.
Others here have worked with MATLAB quite a bit, so perhaps they can chime in on what it does and doesn’t do.
This can be handled to an extent by assemblies in parametric CAD software to my knowledge, but my own designs haven’t progressed to needing that yet so I can’t speak too much on that.
This is super broad, as different materials have different heat properties based on a huge matrix of factors. For example, Nylon, as a semi-crystalline polymer, requires annealing to achieve higher resistance to heat, and how much heat it can take before bending depends on the load it’s under. Materials will have differing specs between manufacturers, even between batches of material, so at best you’d get a rough estimate from simulation.
CAD software with addons can do the basics of this, allowing you to test a type of material, but won’t be exact. For example, which direction are the fibers in your composite facing when your part is formed.
Basically, what I’m trying to say is that material science and mechanical engineering are hard, and there is a reason that those experienced in those fields are paid a lot of money. There is no software solution to replace all the decisions that the neural networks in their head are trained to do, but through your own testing in your own field, you can get there.
To add to the excellent discussion that’s already growing(this is what has worked for my personal projects, in the engineering space I imagine its quite different):
In the design space, I ALWAYS start on paper, a little project notebook to store brainstorming, rough sketches and what the end product has to accomplish.
Much of the problem solving that you do is down to imagination, for example, if you cant think of a great solution, take 10 minutes to do something else and restart/reset parts of your brain.
The iteration process is different from project to project, a versioning tool such as GitHub keeps track of where you are up to - before you make it to that point though you should have a rough idea of what you are after
Moving parts - CAD programs usually allow for a ‘section’ to be created, this lets you see inside of an assembly
The others are able to be calculated by hand, gears levers etc all rely on some ratio there’s lots of reading around but higher-level CAD lets you setup dynamic systems that you can interact with (
James went over Heat Tolerance and Material Characteristics - Engineering Toolbox has a table with a few common ones - Engineering Materials
The realm of simulation and modelling can get quite deep and there are always so many variables - it would be quite expensive to have a one-size-fits-all simulation package, using some maths and a computer you can make your own with special tools - Octave is an open source version of MATLAB with some great tools. LTSpice is great for electrical sims
Making you own using differential equations and the relations to the system isnt off the cards (related reading - Bond Graphs, a great visual way to model, Linear modelling
Gerard has an excellent point about interpreting results - you really have to know what you are looking for!
Liam.
Wow you guys sure are a friendly bunch. I really was not expecting this kind of response so I am thankful to all of you!
Liam and James posts was exactly what I was hoping to find
Now that I understand I was getting ahead of myself. I will take it slow and start building up experience and expose myself to more things.
Many of you Mentioned terms that I used as a starting point to learn more on what I did not know about so that was great too.
The link @Liam sent through with bond graphs also lets you do inter-domain systems, a lets you build a model say using an electrical voltage as an inpuit and getting some amount of force out (though a gearbox, taking into account frictional losses).
There are SOOOOOOOO many tools used by different people - and this would only be a small segment.
PS: I can vouch for Octave and MATLAB - Excel or Google Sheets can also let you get away with smaller calculations for static systems.
