5
$\begingroup$

I would like to learn mechanical engineering as a side project. But, I would like to learn in a way similar to programming. Learn by doing.

Of course I understand I can't build things automatically just yet as I have no 3D-printer and they aren't advanced enough for most things (right?)... but Autodesk has simulation software which may do what I need.

Can I learn mechanical engineering by using AutoCAD for real projects?

$\endgroup$
  • 11
    $\begingroup$ A very short answer: No, AutoCAD won't make a mechanical engineer out of you. $\endgroup$ – Algo Jul 26 '15 at 9:15
  • 6
    $\begingroup$ (Not that making webpages with PHP by trial and error will make a software engineer out of you either.) $\endgroup$ – Tobia Tesan Jul 26 '15 at 13:22
  • 5
    $\begingroup$ Computer science is far more than just programming as mechanical engineering is far more than drawing parts. $\endgroup$ – Paul Jul 26 '15 at 18:12
  • 1
    $\begingroup$ Sure you can. Sell the AutoCad license and buy a small metalworking lathe. (There's always LibreCad if you run out of used envelopes... $\endgroup$ – Brian Drummond Aug 6 '15 at 10:05
  • 1
    $\begingroup$ @BrianDrummond I find the original QCAD lots superior. It's also free. $\endgroup$ – Paul Uszak Sep 8 '15 at 1:10
17
$\begingroup$

AutoCAD is a computerised drafting package that lets people draw pictures.

Mechanical engineering is more than just knowing how to draw pictures in two or three dimensions.

Amongst other things, mechanical engineers are required to know about:

  • Strength of materials
  • Energy
  • Power
  • Torque
  • Stresses
  • Bending moments
  • Couples
  • Torsion
  • Efficiency
  • Thermodynamics
  • Heat & heat transfer
  • Psychrometry
  • Fluids & fluid mechanics
  • Hydraulics
  • Pneumatics
  • Tribology
  • Kinematics
  • Statics & dynamics
  • Structural analysis
  • Mechanical cycles like the Carnot cycle
  • Gears & gear teeth design
  • Crank shafts
  • Vibration
  • Manufacturing processes
  • A good knowledge of chemistry, maths and physics is definitely required

You won't learn any of this from AutoCAD or any other drafting package. Such software may help you develop drafting skills.

All forms of engineering need to be studied at universities or colleges to ensure those who study engineering are qualified and competent. It's not something you pick up by reading a book or learning how to use one or more software packages.

$\endgroup$
  • 1
    $\begingroup$ FWIW, there is actually a job that you can sort of learn by playing with AutoCAD: Draftsman. A draftsman is its own job separate from engineers or technicians. Though generally speaking, draftsmen earn less than engineers and are not considered engineers (for example, their signature on a piece of plan or blueprint generally don't carry the same legal weight) $\endgroup$ – slebetman Jul 27 '15 at 5:34
  • $\begingroup$ Eh, I would consider software engineering an exception. $\endgroup$ – apscience Jul 27 '15 at 8:05
  • $\begingroup$ I understant I cannot learn by ONLY using autocad. I mean can I design things AS I'm learning and see if it works. So I would have something as my goal, and then work to learn what I need to learn to build it. Thats what I mean. Sorry if my question wasnt clear. $\endgroup$ – MechLearner99 Jul 27 '15 at 8:07
  • 3
    $\begingroup$ It's worth noting that any particular mechanical engineering job may only require you to know about a subset of the above list of topics, but it will require you to know about some of them in more depth than could be achieved through tinkering alone. $\endgroup$ – Air Jul 28 '15 at 17:57
  • 1
    $\begingroup$ @TobiaTesan: It depends on the branch of mechanical engineering that you work in. If you specialize in car motors & gear boxes you will never use it. If however, you deal with clothes driers, ventilation systems, air conditioners, etc., where the properties of air & water vapour are critical to humans, that's when you use it. Mechanically engineering is very broad. $\endgroup$ – Fred Aug 8 '15 at 11:43
8
$\begingroup$

No. AutoCAD isn't going to teach you mechanical engineering any more than a compiler will teach you software engineering. You have to actually learn some things, then you will pick up experience by doing. Also AutoCAD is a tool for describing mechanical parts in machine-readable form. It's not going to design a part for you, or teach you how. Without a sufficient background in math and physics and other fields, you will have no way of knowing how to synthesize a design that meets a set of specs.

Knowledge and experience are two different things. Without up front knowledge, learning by experience and making mistakes will be very slow, and will never show you the things that you could have learned that you didn't bump into.

There is a reason it takes a four year degree minimum to become a mechanical (or any other type) of engineer. There is much much more to being a engineer than knowing how to run a few tools.

$\endgroup$
5
$\begingroup$

As Fred alluded to in the above post, mechanical engineering is a vast topic. It takes time and commitment to become a mechanical engineer. Also AutoCAD or 3D-printing alone is not mechanical engineering.

AutoCAD and 3D-printers are tools used by mechanical engineers. By the way, there are few good self-taught mechanical engineers. Following a four-year engineering program is the conventional approach to becoming a mechanical engineer. But alternatively there are plenty of free online material to help become a mechanical engineer or learn a specific area of mechanical engineering. Below are few from edx and coursera.


Edx


Coursera

$\endgroup$
  • 3
    $\begingroup$ +1 for proposing alternatives ways of learning. Maybe there are not as good as a 4 year degree, but help to people with eager to learn. $\endgroup$ – rpax Jul 26 '15 at 19:21
  • 1
    $\begingroup$ I would say you do not become a mechanical enginengineer in most juristictions without a degree. You can however become competent in some of the subfield like machime designer etc etc. $\endgroup$ – joojaa Jul 26 '15 at 22:34
  • 1
    $\begingroup$ @joojaa, I agree with you, but OP is interesting in learning ME as a side project. So following few free online course should help OP achieve OP's goal $\endgroup$ – Mahendra Gunawardena Jul 27 '15 at 1:08
3
$\begingroup$

One thing that other people don't seem to have touched on is the utility of the simulation packaged in AutoCAD. It's a type of Finite Element Analysis, which is an incredibly useful tool, but one that must be wielded very carefully. You can model a part, load it into an FEA package, apply constraints and forces as the part would see in the real world, and yet there is no guarantee that the results you get from that simulation have any correlation to real world outcomes.

Understanding how to set up and run the simulation takes training of its own, and even with a well-designed simulation, understanding the results takes independent engineering knowledge. The software I primarily work in is Creo, and Creo's FEA package has a warning message that actually tells you that it thinks something is up with the results of the simulation, but it doesn't know exactly what, and that you need to use your engineering knowledge to analyze the results for accuracy.

That being said, don't let any of this (or anything anyone else has said) stop you from playing around in AutoCAD. It doesn't matter how good of a concept an engineer can come up with, if they can't put it on paper or communicate it to someone else, it's worthless. However, if you want to learn by doing, realize that the end product of mechanical engineering is a machine. What you need to be doing is working with machines. Take apart existing things, find things that you can build yourself, and examine all of this stuff to try and understand how things were made and why they were designed they way they were.

$\endgroup$
3
$\begingroup$

As a working professional ME with 18 years in the field, I'll tell you that there's a great deal of difference between what will work in AutoCAD (or any other such software package) and what will work in the real world. That's why we still review designs and build prototypes before going into production.

I can't tell you the number of times I've had to get the designers who draw up stuff and the people who have to make the stuff in the same room to hash out how to make something work. The designers will pull up their model and say something like, "But it works in the model..." while the people from production show them why it doesn't work in the real world. Then, after everyone sees the same problem, the work starts.

If you just want to goof around and make stuff for your 3D printer, that's great. I'd suggest a solid modeling package like Autodesk Inventor that will let you export .stl files directly and do assemblies, etc. It's expensive for commercial use, but they're supportive of people learning on their own, so I think you can get a license for little or no $$. Still, even becoming an expert user of that, or any CAD package won't make you into an engineer.

$\endgroup$
2
$\begingroup$

I think that there's a problem with your analogy. Autocad is to engineering as is programming - to actually running your code with meaningful results. Otherwise it's just drawing. 3D printing is as the name implies, drawing. In three dimensions, rather than two.

I suspect that you're looking at hobby level engineering if you're categorizing it as a "side project". That can be rewarding and complex too. If you're interested in CAD and 3D printing, design and print some widget. Then, get it to work. Repeat process until it does work.

Examples off the top of my drink addled head:-

  • Design some lego, then get it to fit and stick together like the real stuff. Your issues will revolve around matching the scale of the pieces to the repeatability /precision of the printer. Get into that and see how real Lego manage production quality.
  • Design and build a model arch bridge. I mean out of individual printed pieces, including a keystone. Man's been building them for a 100,000 years. You can get into what makes an arch stand up, and whether circular or parabolic shapes are best.
  • Print something like a car powered by a rubber band. Try to incorporate real metal bearings for the axles (buy them dirt cheap off eBay). Or don't and see it it's feasible to print axles that will work without discrete bearings. How will the rubber band turn the wheels, and how can it overcome axle friction? You'll have to investigate how two printed surfaces can slide past each other - printed stuff looks rough to me.

I know that the above examples are trivial, but if you get into it, real engineering techniques can be discovered. After all, it's pretty useless printing a million angels on a pin head, but dwell on the engineering know how that lets you do it...

$\endgroup$
  • $\begingroup$ Lego pieces have tolerances that far outstrip the ones even the best 3d printers can manage. Getting something that snaps as well and snaps as reliably, and over so long scales and different pieces is a very specialized manufacturing and design problem. $\endgroup$ – joojaa Jul 27 '15 at 11:07
  • $\begingroup$ @joojaa If you keep the same size. As I said, it will be a good engineering exercise to figure out by how much you'd have to scale up a printed brick to work reasonably. This is were the learning comes in... $\endgroup$ – Paul Uszak Sep 8 '15 at 1:12
1
$\begingroup$

As a current Mechanical Engineering student I have been working at a company for 3 years mainly doing 3D design. This ranges from 3D modelling of very complex assemblies all the way down to the simple pipes and sheetmetal.

The most you will learn is the way assemblies should be assembled in order, welding standards, sheetmetal standards and possibly the strength of structures.

However, to be a good mechanical Engineering you need to have knowledge in fluid mechanics, thermodynamics etc.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.