“A design is never perfect, but there are things you can do from the beginning
that can increase your chances of success dramatically.”
So Iʼm sitting on the floor in my living room watching Sharktank the other day tearing apart my wifeʼs Dyson Ball vacuum cleaner just because I wanted to see how it “worked.” It is very well thought out by the way… I sat back watching the TV show and thought silently, “Some of these ideas are really bad? Man, if I could demo my venturi-shotgun-beer accessory and make it as good as this Dyson Ball, they would definitely invite me on the show after seeing the genius in that idea…”
“Sigh…” I dropped my head and stared blankly at the vacuum held in my hands, appreciating the whole concept of sucking dirt from the floor wondering, “Who’s the guy that thought about sucking dirt from the ground? I totally want to meet him…” The sounds of poor schmucks being abused and intimidated from the TV show in the background brought me mack to the present and lit a fire in me, “Iʼm only giving up 10% of the profits because I am no sucker like all those other idiots…My venturi-shotgun idea is pure genius.”
“If only I had a magic thingy-ma-jigger that could make it for me!?” But wait!…I did see this youtube video of a machine that made a wrench out of thin air? How in the hell did it do that? -10 minutes go by- “Damn this Dyson Ball Cleaner is really well thought out…” “Well, I donʼt understand how it did it, but it did it…it was like a magic wand! And the wrench worked! Iʼll find someone that has that machine like that…”
Sound familiar? That my friends, is the moment that the entire 3D printing industry huddles over their black caldron in the sky like anxious greedy witches, waiting for every new person with a good idea who gets a whiff of their 3D printing brew to sucker you into spending money on a design that is going to either break or never work…just so you can do it again and again until you give up or run out of money.
We all know that the magic wands of today (3D Printers) can make just about anything out of a variety of materials if you can cough up a few hundred to even a couple of thousand dollars to make it happen. Exciting times for the self-made millionaire dreamer, the do-it your-selfer and the unfound genius in us all. With 3D Printing and some free online-CAD tools, we can finally make our dreams come true or at least have a whole lot of fun trying?
Well, not really…here at i3D we are seeing a lot of customers spending their hard earned money on their designs only later to find out their parts do not fit together, they break easily or it simply doesnʼt work how it was intended. This is obviously great for 3D Printer manufacturers and service bureau providers. They want you to screw up and come back make another or use more material. Spend more money!
With the advent of 3D printing, people see an amazing end game, very akin to a magic wand. The magic wand provides hope we can actually develop our ideas into real products on our own and itʼs breaking a lot of hearts out there. People really do just want to spend $50 and expect parts to be perfect and work the first time. Itʼs not a perfect world, even with 3D printing and a perfect design.
A design is never perfect, but there are things you can do from the beginning that can increase your chances of success dramatically.
I put together 5 simple design tips to follow before you spend any money on parts. Iʼm a PE in Mechanical Engineering and founded my own 3D Printing service bureau and engineering consulting service here in Silicon Valley. I know what it takes to develop a product that works…itʼs tough, a plethora of things to consider from the manufacturing processes, to materials to tolerances and more and itʼs all interdependent. It breaks my heart every time I see someone with a good idea really trying to make their dream a reality, but making all the classic mistakes.
Before you send your design in for a quick print, follow these five simple rules. It will help you make more out of your hard earned money.
1. Add fillets (Curved Joints) to all thin walls that join an unsupported surface. Sharp internal corners and edges create weak spots (stress concentrations) where your part will break/crack much more easily. A good rule of thumb for 3D printing, is to make the fillet equal to the thickness of the material. Remember parts are built in layers, so if you can make it bigger, it will help you in the long run. For example. If there is a 1.5 mm thick snap-fit feature. Make the fillet radii 1.5mm. If you go for 3, it will be more than twice as strong as the joint. The bonus on this, is that if you ever get to production injection molding, these fillets help the plastic flow through the part.
- Plastic Injection Molding Design can conflict with this and you should always do a review redesign of your product before you send it off for injection molding. Sometimes it doesn’t make sense to have these joints in a plastic injection molded part as it makes the mold more expensive to manufacture. It is much easier to cut external sharp corners than curved corners (inverse of a fillet).
- However in 3D printing world where parts are made in thousands of layers, itʼs a great way to make more parts functional for testing by reducing the stress in the joints.
- You need to make a judgement call. Does the feature flex a lot? Yes? Add a fillet. Is the thin wall unsupported? Yes? Add a fillet. and so on…
2. Donʼt make parts very thin for the first prototype. The first time itʼs all about Form Fit and Function testing. Most likely, you will suffer form the world wide designer epidemic, “Parts look bigger on screen than in my hand…” Is it big enough, thick enough, fit with the other parts and do what I want it do do? Obviously the thinner they are the less material there will be and the more money you can save, but your part will break or not flex how you intended or worse it will warp with temperature fluctuations. A good safe thickness is 2mm if your are designing plastic injection molded parts. You can test your design and later reduce it to 1.5 mm or even 1 mm for production. What are you testing with your design or checking?
3. Design clearance into your parts and assemblies. Everyone forgets to do this and prints with parts surface to surface contact in assemblies. If you have a shaft you want to rotate or slide in a hole, donʼt make the shaft diameter equal to the hole diameter, it wonʼt rotate. There has to be clearance between all surfaces for them to fit together nicely and on top of that there are tolerances.
- For holes/shafts under a 1” diameter, a radial clearance = .005” min will work. This is safe considering most 3D printing equipment canʼt build more accurately than +/-.005” and there are two parts. The statistical probability of both parts being at the far extremes (one big the other small by .005” is very low); this sweet- spot will get the job done without having to obsess over it.
- For assemblies, it can become very complex and hard to understand clearances and tolerances; each part has a tolerance, they stack on top of each other and can get out of control real fast. Most degreed engineers struggle at first to understand and it requires a detailed analysis. If there are more than 6 parts, itʼs best to take a statistical approach. But experience has taught us a lot and that we donʼt need to do analyze in most cases. Simply design in a “gap” between parts that mate together. Be conservative and make your life easy…if your design is complex with many parts, get some professional help. Itʼs a shame to spend $1500 on 9 different parts and they donʼt fit together! A good gap to work into your assemblies is .01”-.02” between mating surfaces. It may sound sloppy, but do the math.
- If 9 one inch thick bricks for your model stack on top of each other, then 9 parts multiplied by .005” =.036” in the worst case. That means the stack will be 9.036” tall. If it was intended to line up with the roof of another model 9 inches tall, it wonʼt. There is an enormous amount of explanation left out to keep it simple in this article but also why it is so cool how the Romans figured out how to build the Colosseo and avoid this issue. Tolerance stacking depends on the “flow” or “stack” of features and maybe one day soon we can cover this in another article.
4. Pick the right process and material for what you are trying to do. The very first prototype you make is only a “proof of concept”, itʼs not an engineering functional prototype. You may wish it to be an engineering functional prototype, but 9.9 times out of 10, your design is going to need re-work, there is going to be something you didnʼt think of and you will have made a mistake somewhere. So donʼt waste your time and money to pay for a more expensive process and material when you donʼt need too.
- First prototype = Proof of Concept (Use affordable 3D Printing like fast polyjet, FDM or ZPrint) anything more is overkill.
- Second prototype = functional testing (Use SLA if you can, second choice is Polyjet) Why? SLA and Polyjet materials mimic properties of real plastic(ABS, PP, Polycarbonate, etc). Polyjet is amazing and can build very precise models; it is a good alternative but be aware there may be problems at elevated temperatures and require a post-cure; ask if they offer that.
5. Seek out professional advice before you 3D print your first prototype. Seriously, engineers are paid big bucks to do what you are trying to do and they still get it wrong every day. A little help can go a long way to make your dream a reality. Itʼs important not to waste time making a part that may never be able to be manufactured in any other process than 3D printing. With 3D printing you can make anything “once” and that will be the center of the next topic as well as how to take shortcuts and save hundreds of dollars on 3D printing.