12

On a number of occasions I've broken small plastic parts that are nearly impossible to replace but could easily be 3-D printed. The latest such mishap is the volume knob on the factory-installed radio on my car.

I have little experience in 3D printing, and would like to be able to replace these parts with something very close to the original. Spending hours measuring and designing a replacement part that should be $5 isn't really an option. I need something to scan the broken pieces in 3D and somehow just seal up the seam where it's broken.

Is there a scanning/printing/software system to do this that doesn't require a lot of 3D design experience?

Diesel
  • 1,855
  • 2
  • 10
  • 29
James
  • 249
  • 2
  • 7
  • Hello @James, I noticed your question has been up for a while now. Have any of the answers below been able to solve your question? If so, would you mind accepting the appropriate answer. If not, what is missing so that we may help you further? Also, if you have figured it out on your own, you can always answer and accept your own solution. Thank you. – tbm0115 Feb 27 '17 at 15:59

2 Answers2

12

The easiest way is as you currently do: model the pieces by hand, using (digital) calipers to measure them.

Scanning technology isn't very good, and the models are not of printable quality. Usually, fixing a scan is more work than modeling an item from scratch.

Tom van der Zanden
  • 14,588
  • 2
  • 33
  • 60
  • The better the measuring equipment, the easier it becomes. I've used a CMM at work to "scan" a part with the touch probe. This gets me the curved surfaces that are hard to measure with calipers. Then again not everyone has $100k+ measuring equipment at their disposal lol Try looking for a local machine shop! – tbm0115 Jan 19 '16 at 14:32
  • I'd say this is probably the best answer out of the bunch. For something as simple as a volume knob, even for new CAD users, it isn't a long process. I used this exact approach in my CAD class for students to practice modelling. For a volume knob, I can't imagine it should take more than 45 minutes even for a relatively new users – Diesel Nov 23 '17 at 20:16
  • @Diesel 45 minutes for a $5 piece is still far from ideal. Even if it's the best solution at the moment, it's still too high a barrier for some. – Erhannis May 14 '18 at 15:29
  • @Erhannis, I'm not quite sure what you're getting at. 45 minutes is an estimate of course and if the part is 5$ is too much then cost of printing it is going to be a barrier. My comment was pointing out that modeling is a very easy skill to learn (not to be great at, but learn) and if it is a skill that is learned on free software the OP will then both not have to pay for either an expensive scanner and they'll be able to create whatever they need. If 45 minutes is too much time to learn something, how badly do you really need the thing? – Diesel May 15 '18 at 18:51
  • @Diesel Sorry, I was basically complaining. If I had to pick between spending $5 and spending 45 min designing something mundane, I'd pick the $5. Now, as a learning exercise, it may be very worthwhile. Similarly, if the 45 minutes decreased significantly with practice (say, down to 5 or 10 minutes), then I would be ok with settling for the design route. (In general, I'm just dissatisfied with the state of 3d scanning technology, and wish it were magically perfect, haha.) – Erhannis May 17 '18 at 17:49
3

There is scanning technology: either hardware or software (such software typically works from multiple 2D photographs). It has limitations, but is an active research area, and getting better all the time. "Autodesk 123D Catch" and "3-Sweep" are a couple examples.

One tradeoff between automatic and manual is the complexity of the shape you need to duplicate. Sometimes you can make a much simpler shape that still works, and/or thicken the part so it doesn't break the next time. That can make manual design much more attractive.

TextGeek
  • 3,181
  • 2
  • 12
  • 32