What's a standard way for building a prototype?

[fuzzybabybunny]

So I'm making some parts in CAD for a flash bracket. The parts are small, like 10cm x 1.5cm. The final material will be something like 6061 Alu, but I'd like to find some way to pump out prototypes beforehand. Are there places like woodshops that take CAD drawings and make the parts? Are there other types of shops/materials that I should consider?

Man, I wish I had a 3D printer

EDIT:

Is there a book or resource that I can read that details how I should design a part for manufacturability? I'm designing parts and I'm not sure if such and such a design would be hard/expensive to machine over another design.

Also, I realized yesterday that I should somehow take into account machining imperfections into my design. Say that I have a 5mm diameter cylinder that fits into a 5mm diameter hole as part of a moving pivot assembly. When it actually goes for manufacturing, the 5mm cylinder could come out being 5.1mm and the hole could be 4.9mm, meaning the pivot would be too tight and won't move. Or the inverse could happen - cylinder is too narrow and hole is too wide, leading to a lot of slop in the pivot. How do I design my pivot to take into account manufacturing imperfections? Should I leave the hole and cylinder at 5mm, or should I define the cylinder as something like 4.9mm and the hole at 5.1mm? I'd imagine I'd also have to take into account the specific machine in a specific shop?

Then there's the "feel" of the part - 5.00mm cylinder and 5.15mm hole would pivot, but how does the motion feel? Is it smooth? Too loose? Too much friction?

 

[Paperdoc]

What you want is a machine shop with NC machining equipment. NC means "numerically controlled". Look in your local yellow pages and call each to verify whether they can do 3-D machining, and in what form do they need input files? Most of the modern ones can take some form of Autocad files that fully describe the 3-dimensional shape.

For initial prototyping just to examine the feel and fit, have it made of some simple material that is cheap and easy to machine. To evaluate actual performance and suitability to end use you'll need it made of something durable enough to stand up to normal use for a while. If you're trying to verify performance under stress loads, have it made from the final material. Along the way, ASK the machine shop for input on how it really should be designed. Firstly, their people may have good ideas you had not considered. And just as importantly, they can give you good advice on how to design so it is easy to make in their equipment, and maybe also on larger mass-production equipment if you plan high production volumes. They may also have good advice on the exact material to use for final production.

By the way, without seeing your actual design it is possible that the shape is relatively simple and does not need complex 3D machining. It also is possible that, if you plan to have thousands made down the road, some critical parts might better be made as cast parts to be assembled with other simpler parts and reduce the complexity and cost of 3D machining.

I work in an Innovation Center where we advise small companies on design and development of products. If you're planning to develop a product for commercial distribution and sale, look for such a place where you live and seek their advice. Many will provide advice free to start, and point you to funding sources, machine shops, etc. for your development process. They also can advise on how to protect your Intellectual Property (IP) rights. In some cases if an actual Patent is advisable (sometimes a good idea, sometimes not) there are secrecy requirements that you need to understand BEFORE talking to anyone else, even a contractor (like a machine shop) working for you.

 

[keird]

Nice input Paperdoc.

 

[KMc]

If you are really serious, google "Rapid Prototyping Services". In particular, Stereolithography (SLA) and Selective Laser Sintering (SLS) are great technologies specifically for this purpose.

 

[KMc]

Oh, and Fused Deposition Modeling (FDM) is also popular.

 

[ManyBeers]

Try here

I used to have a bridgeport mill in my garage and was a machinist for several years in San Joise from mid 70s to early 80s. Why don't you post a screenshot of your part?


Thisis a toolless quick-change hub I designed and built for a 80cc racing go-kart. Feel free to ask me anything about making parts.

License frame Click on image for better quality.

 

[dullard]

I typically just borrow a vertical mill and make them myself. That is far faster and cheaper than any other method. It takes me twice as long to describe my parts to a professional machinist than to just machine it myself. Of course, that option isn't available to everyone.

Barring that, quickparts can usually get you a part quickly and cheaper than many machinists. Although with the depressed car sales, machine shops are getting desperate and may reduce their prices for you.

My final option would be to send CAD drawings to a machine shop. They'll do the job well, but they can be quite slow and pricey. Plus, you might not have access to CAD and will have to pay for that work too.

 

[fuzzybabybunny]

Sweet! Thanks so much for the info!

I'll post a picture of my assembly when I get around to finishing it - it's been 4 days with Solidworks, so I'm still learning, slowly.

 

[fuzzybabybunny]

Is there a book or resource that I can read that details how I should design a part for manufacturability? I'm designing parts and I'm not sure if such and such a design would be hard/expensive to machine over another design.

Also, I realized yesterday that I should somehow take into account machining imperfections into my design. Say that I have a 5mm diameter cylinder that fits into a 5mm diameter hole as part of a moving pivot assembly. When it actually goes for manufacturing, the 5mm cylinder could come out being 5.1mm and the hole could be 4.9mm, meaning the pivot would be too tight and won't move. Or the inverse could happen - cylinder is too narrow and hole is too wide, leading to a lot of slop in the pivot. How do I design my pivot to take into account manufacturing imperfections? Should I leave the hole and cylinder at 5mm, or should I define the cylinder as something like 4.9mm and the hole at 5.1mm? I'd imagine I'd also have to get tolerances on the specific type of machining process used at a specific machining shop?

Then there's the "feel" of the part - 5.00mm cylinder and 5.15mm hole would pivot, but how does the motion feel? Is it smooth? Too loose? Too much friction?

 

[herm0016]

you will need to tolerence each part so that whome ever makes it knows how precise it needs to be. all of your measurements should have a +- tolerance. http://en.wikipedia.org/wiki/G...ioning_and_tolerancing

gd&t is very powerful. read up on it a bit.

designing for manufacturing is more of an insite and art than a science. design your parts so that you can get them out of a single block or sheet of standard size and design them to be machined or bent so that you reduce wast.

 

[ManyBeers]

Originally posted by: fuzzybabybunny
Is there a book or resource that I can read that details how I should design a part for manufacturability? I'm designing parts and I'm not sure if such and such a design would be hard/expensive to machine over another design.

Also, I realized yesterday that I should somehow take into account machining imperfections into my design. Say that I have a 5mm diameter cylinder that fits into a 5mm diameter hole as part of a moving pivot assembly. When it actually goes for manufacturing, the 5mm cylinder could come out being 5.1mm and the hole could be 4.9mm, meaning the pivot would be too tight and won't move. Or the inverse could happen - cylinder is too narrow and hole is too wide, leading to a lot of slop in the pivot. How do I design my pivot to take into account manufacturing imperfections? Should I leave the hole and cylinder at 5mm, or should I define the cylinder as something like 4.9mm and the hole at 5.1mm? I'd imagine I'd also have to get tolerances on the specific type of machining process used at a specific machining shop?

Then there's the "feel" of the part - 5.00mm cylinder and 5.15mm hole would pivot, but how does the motion feel? Is it smooth? Too loose? Too much friction?

Sounds to me like you could use some mechanical engineering know-how. Barring that you'll just have to make some prototypes and troubleshoot them until you get it right.

 

[fuzzybabybunny]

Originally posted by: herm0016
you will need to tolerence each part so that whome ever makes it knows how precise it needs to be. all of your measurements should have a +- tolerance. http://en.wikipedia.org/wiki/G...ioning_and_tolerancing

gd&t is very powerful. read up on it a bit.

designing for manufacturing is more of an insite and art than a science. design your parts so that you can get them out of a single block or sheet of standard size and design them to be machined or bent so that you reduce wast.

Ahhh... I see. So every single part I make should have a tolerance spec'd into it. The cylinder would be something like 5mm -0.1mm, and the hole would be 5mm +0.1mm. I chose +/- 0.1mm because it seems like +/- 0.1mm is a good average tolerance for CNC machining, correct?

For parts where sizing is not important, say, a hole that's 100mm, would I just leave the spec at 100mm or does this actually mean "100mm with NO tolerance whatsoever?" I'm thinking it's the latter?

 

[ManyBeers]

Originally posted by: fuzzybabybunny

Originally posted by: herm0016
you will need to tolerence each part so that whome ever makes it knows how precise it needs to be. all of your measurements should have a +- tolerance. http://en.wikipedia.org/wiki/G...ioning_and_tolerancing

gd&t is very powerful. read up on it a bit.

designing for manufacturing is more of an insite and art than a science. design your parts so that you can get them out of a single block or sheet of standard size and design them to be machined or bent so that you reduce wast.

Ahhh... I see. So every single part I make should have a tolerance spec'd into it. The cylinder would be something like 5mm -0.1mm, and the hole would be 5mm +0.1mm. I chose +/- 0.1mm because it seems like +/- 0.1mm is a good average tolerance for CNC machining, correct?

For parts where sizing is not important, say, a hole that's 100mm, would I just leave the spec at 100mm or does this actually mean "100mm with NO tolerance whatsoever?" I'm thinking it's the latter?

Machinery Handbook
Pocket Companion about 5-6 down on page. $20.85

 

[herm0016]

but.. if you tolerance it as 5mm +- .1mm for both, they have a good chance of ending up 5.01 for the shaft and 4.09 for the hole and thus would not fit. you would have to go 5.0 +.01 and 4.99 -.01 so that any way it works out it will fit and have a min. difference of .01 and a max of .03 its a really important point for precision work.

 

[Paperdoc]

For fitting parts together you need to specify for each part BOTH the correct size AND the tolerance. The correct sizes are NOT the same - one is deliberately smaller than the other to allow a proper gap in the fit. Usually that gap actually will end up filled with a lubricant if there is any relative movement. In your example of a cylinder fitting into a bored hole, you need to know whether the cylinder slides in the hole frequently, or is a one-time (at assembly time) slip it in and forget it. But suppose it actually is in frequent motion and a lubricant will be present. You will make a specification like the bored hole has an inside diameter of 5.00 mm ± 0.01 mm, and the cylinder outside diameter is 4.99 mm ± 0.01 mm. Note that if BOTH items are made with errors in opposite directions but still within tolerances, you might not get them to fit! On the other hand, the difference in sizes (that is, the designed gap between parts) must be tight enough to avoid sloppy performance.

WARNING! I just made up numbers for the example. I do NOT know whether a gap of 0.01 mm between cylinder and bore diameter is the proper value. And neither do you, I believe. For this kind of knowledge you need an experienced design engineer. From the questions you ask, you are not one of those, just like the rest of us. At some point in your development process you will need to bring in that level of expertise to polish up the design details. However, for purposes of making the first few prototypes that probably is not necessary. In fact, in prototyping you'll find that any good machinist can tell you immediately what gap you should design into your hypothetical cylinder in a bored hole. You just need to know when to do it, and to recognize that "polishing" the design may actually change it (for the better, we all hope!), so don't resist that.