ORD Bot For Small Scale Production
Posted: Sun Jun 10, 2012 3:35 am
I'm at one of those chicken and egg moments. I want to use a 3D printer for small scale production, but because it hasn't been capable of that yet I haven't experimented with 3D printing so I have less than amateur experience. I'm now wanting to run before I crawl.
I've been watching 3D printing for years, waiting for the right time to jump in. Much as I think it would have been fun to tinker with 3D printing as a hobby, I already had too many hobbies I wasn't pursuing. What's really fun for me is the engineering... that moment when hobby level concepts or laboratory science breaks out into the real world and there is a practical application.
It seems that there has been absolutely explosive growth in just the last few months. It seems that the technology started from nowhere and has been undergoing exponential growth and is now at the knee of the exponential curve and it's really taking off. There is a synergistic combination of electronics, firmware, rigid and low cost three axis frameworks, extruder design and more. For example, there is now a relatively easy to build low cost and high quality ORD Bot, and QU-BD is gearing up to make high quality low cost extruders. Slic3r and other software seems to be updated almost daily.
What I've really wanted to do is use 3D printers for short run manufacturing. I have a small internet based business that I run from my home. I have access to a friend's $30K Dimension 3D printer if I want to make the occasional prototype, and while that's nice, what works for me is having my own 3D printer so I can iterate 3-4 designs in a day. What REALLY works for me is to then gear up and start manufacturing the same products on the same machine for my short run production.
I know most people simply do not see 3D printers in this way. Even in the hobby realm where imaginative people are pushing the state of the art, there seems to be a strong bias toward personal printing - making little things for fun or to solve some little problem. 3D printing has been plagued by poor print quality and very long print times. I feel that's changing. Certainly, the print quality has improved a lot in the last year. I feel that for my market, the print quality is now acceptable to sell to people who aren't buying 3D printed parts to build a 3D printer. To achieve that print quality requires .1 mm to .2 mm layers, so the print times are necessarily long. The more rigid printers like the ORD Bot are capable of moving at higher speeds, so that helps some with the print times.
Right around this point in my mad ravings, I can imagine those who have been printing for a year or two thinking about the hours that are needed to print a part and thinking I'm nuts to be talking about production, but please keep an open mind. My production needs might be ten parts a day. Remember that I like to do the engineering to make things more practical, so I'd be doing some work to overcome the current issues that prevent people from printing parts to sell instead of CNC manufacturing or injection molding. That would be my open source contribution, since I skipped all of the hobby fun.
One obvious idea is multiple print heads. I've been kicking around the idea of a dual head 3D printer that prints two identical parts at once. The extruder stepper motors would be wired together, just as the ORD Bot's Z axis steppers are wired together and operate in locked unison. A dual extruder 3D printer could print at the same resolution but twice as fast. The bed would need to be flat and level and the nozzles would need to be at the same Z height. After seeing the new QU-BD dual extruder, I'm actually thinking of making an ORDish printer with beefier 4040 extrusions and NEMA 23 motors and have a dual extruder on the front and back of the gantry to print four identical parts at once. Yes, to answer the question already posed, I do realize that the extruder spacing determines the maximum part size. My parts are small enough that this would work well. When I design new parts, I'd probably design new carriages to mount the extruders to accommodate the parts I want to build. The 3D printer would still be available for prototype printing by using only one extruder, but would be almost completely devoted to one product or maybe two products of similar size.
Another obvious idea is fully automated operation. I don't want to unload the parts every couple of hours. My initial thought was an improved version of the MakerBot Automated Build Platform, and I did some preliminary design work to that end. Lately, I've been wondering if I can turn off the heat to the build platform to cool the complete part which should create some shear force between the part and the build platform, and then use some sort of ejector bar to slowly tip the part off the build platform and sweep it into a bin at the side of the printer. That's one of those times when some practical 3D printer experience would be very helpful. I don't have that experience yet. Feel free to chime in with comments and suggestions.
I can imagine objections based on the belief that 3D printing is still a twitchy proposition... in other words, more art than science. I know there's a lot of truth to that and a lot of real world experience to back that up, but I did mention that I'm an engineer and I love the art of engineering and using science to take the twitchy parts out of a process. Most of my career has been spent in the design of custom machinery (lots of electronics and microcontrollers, but also PLCs, servo and stepper motors and all manner of industrial machines), but I also spent some time as a manufacturing engineer where my job required a lot of reducing process variation and improving quality and yield through process improvements and better automation equipment. It seems that most 3D printer owners print one or two of something, and just as they're finished getting acceptable parts, they're finished with that part. That probably leads to the false belief that all 3D printing will be problematic and will require constant process adjustments. I think that running production would allow optimizations to the slicing algorithm, the infill, and pretty much every possible parameter. Do something a hundred times and you'll be good at it.
I may determine that I need to buy my own resin in pellet form and make my own custom filament extruder to get the consistent filament quality I need. Cutting the materials cost to 30% would be an added bonus.
I'm very interested in print quality. It's difficult to polish plastic because it tends to fuzz. I've had good success using a random orbital sander to polish flat plastic plates, particularly black Delrin, to give it a very attractive flat matte velvet finish that is very resistant to fingerprints. I'd like to experiment with various polishing methods to improve the surface finish of printed parts. I have a large rotary drum polisher that's intended to wet polish rocks with abrasive. I use it with small stainless pins and soapy water to polish brass and it does a great job. I'd like to see if I can create a simple batch polishing process that helps remove the last of the layered appearance with no manual labor on each piece.
I've heard the suggestion that I should just get a quick mold made and have an injection molder run my parts. That had been my goal for some time. When the production 3D printing bug began to bite, I forced myself to go online and get a couple of quotes, thinking that the numbers would quickly disabuse me of any crazy ideas of printing my own parts, but the opposite happened. I got an online quote from a Chinese and an American company. Basically, the Chinese company wanted $3200 for the mold and the ABS parts would be 34 cents each. The American company wanted $6400 for the mold and the parts would be $1.33 each. Protomold might have cheaper prices, but probably not too much cheaper. I'd be looking at a decent up front investment of several thousand dollars to get into injection molded parts. It was doable, but the payback would be a long time before I was in the black. For a small business, cash flow is king! I also would be stuck with the die, so there would probably not be any rolling improvements based on customer feedback or Continuous Quality Improvement. I'd also have a fairly long lead time, particularly for the Chinese parts, so my time to market would suffer, and my ability to respond quickly to changes in market demand would not be good. I believe short run 3D printing will be a transformational technology for small companies like mine. I could get my products to market very quickly, I don't need to maintain bulky and expensive inventory, I don't need to deal with the hassles of injection molding and dependence on a vendor, and I'm immediately profitable.
I'm painfully aware of my 3D printing ignorance, and I hate to speculate from a base of ignorance, much less make plans. So far, I'm thinking a lot about small production 3D printing, but my initial goal is simply to build a Hadron ORD Bot and gain some much needed experience. I've been obsessed with this notion lately, so I'm hoping that obsession will lead to a lot of experience in a short time. It also doesn't hurt that I'm not piddling with my 3D printer in my free time. This is now my job!
I'm an electrical engineer with machine design and building experience, and I have a milling machine and lathe in my basement. The question is, can I make a significant contribution to getting 3D printing out of the hobby market and into the small business world? I'm particularly interested in the idea that parts can be printed that can't be made any other way. The infill can be optimized to add strength or rigidity in the direction it's needed, and not have excess material where strength isn't needed. This not only saves the cost of the material, it also saves weight. 3D printed parts can also be printed that have great complexity, about as easily as printing a solid cube of the same mass. I can design complicated parts that wouldn't be cost effective if they were machined or molded.
Anyway, I've been hinting at my short run continuous production plans in several other threads, and it's inevitable that people would have comments about this concept. I didn't want to hijack those other threads, so I thought anyone who might want to speculate along these lines with me and discuss this topic could do so here.
What do you guys think? Interesting idea? Plausible after some work to adopt 3D printing to the low end of the manufacturing market? Has my cheese slipped off its cracker? Am I smoking crack here?
I've been watching 3D printing for years, waiting for the right time to jump in. Much as I think it would have been fun to tinker with 3D printing as a hobby, I already had too many hobbies I wasn't pursuing. What's really fun for me is the engineering... that moment when hobby level concepts or laboratory science breaks out into the real world and there is a practical application.
It seems that there has been absolutely explosive growth in just the last few months. It seems that the technology started from nowhere and has been undergoing exponential growth and is now at the knee of the exponential curve and it's really taking off. There is a synergistic combination of electronics, firmware, rigid and low cost three axis frameworks, extruder design and more. For example, there is now a relatively easy to build low cost and high quality ORD Bot, and QU-BD is gearing up to make high quality low cost extruders. Slic3r and other software seems to be updated almost daily.
What I've really wanted to do is use 3D printers for short run manufacturing. I have a small internet based business that I run from my home. I have access to a friend's $30K Dimension 3D printer if I want to make the occasional prototype, and while that's nice, what works for me is having my own 3D printer so I can iterate 3-4 designs in a day. What REALLY works for me is to then gear up and start manufacturing the same products on the same machine for my short run production.
I know most people simply do not see 3D printers in this way. Even in the hobby realm where imaginative people are pushing the state of the art, there seems to be a strong bias toward personal printing - making little things for fun or to solve some little problem. 3D printing has been plagued by poor print quality and very long print times. I feel that's changing. Certainly, the print quality has improved a lot in the last year. I feel that for my market, the print quality is now acceptable to sell to people who aren't buying 3D printed parts to build a 3D printer. To achieve that print quality requires .1 mm to .2 mm layers, so the print times are necessarily long. The more rigid printers like the ORD Bot are capable of moving at higher speeds, so that helps some with the print times.
Right around this point in my mad ravings, I can imagine those who have been printing for a year or two thinking about the hours that are needed to print a part and thinking I'm nuts to be talking about production, but please keep an open mind. My production needs might be ten parts a day. Remember that I like to do the engineering to make things more practical, so I'd be doing some work to overcome the current issues that prevent people from printing parts to sell instead of CNC manufacturing or injection molding. That would be my open source contribution, since I skipped all of the hobby fun.
One obvious idea is multiple print heads. I've been kicking around the idea of a dual head 3D printer that prints two identical parts at once. The extruder stepper motors would be wired together, just as the ORD Bot's Z axis steppers are wired together and operate in locked unison. A dual extruder 3D printer could print at the same resolution but twice as fast. The bed would need to be flat and level and the nozzles would need to be at the same Z height. After seeing the new QU-BD dual extruder, I'm actually thinking of making an ORDish printer with beefier 4040 extrusions and NEMA 23 motors and have a dual extruder on the front and back of the gantry to print four identical parts at once. Yes, to answer the question already posed, I do realize that the extruder spacing determines the maximum part size. My parts are small enough that this would work well. When I design new parts, I'd probably design new carriages to mount the extruders to accommodate the parts I want to build. The 3D printer would still be available for prototype printing by using only one extruder, but would be almost completely devoted to one product or maybe two products of similar size.
Another obvious idea is fully automated operation. I don't want to unload the parts every couple of hours. My initial thought was an improved version of the MakerBot Automated Build Platform, and I did some preliminary design work to that end. Lately, I've been wondering if I can turn off the heat to the build platform to cool the complete part which should create some shear force between the part and the build platform, and then use some sort of ejector bar to slowly tip the part off the build platform and sweep it into a bin at the side of the printer. That's one of those times when some practical 3D printer experience would be very helpful. I don't have that experience yet. Feel free to chime in with comments and suggestions.
I can imagine objections based on the belief that 3D printing is still a twitchy proposition... in other words, more art than science. I know there's a lot of truth to that and a lot of real world experience to back that up, but I did mention that I'm an engineer and I love the art of engineering and using science to take the twitchy parts out of a process. Most of my career has been spent in the design of custom machinery (lots of electronics and microcontrollers, but also PLCs, servo and stepper motors and all manner of industrial machines), but I also spent some time as a manufacturing engineer where my job required a lot of reducing process variation and improving quality and yield through process improvements and better automation equipment. It seems that most 3D printer owners print one or two of something, and just as they're finished getting acceptable parts, they're finished with that part. That probably leads to the false belief that all 3D printing will be problematic and will require constant process adjustments. I think that running production would allow optimizations to the slicing algorithm, the infill, and pretty much every possible parameter. Do something a hundred times and you'll be good at it.
I may determine that I need to buy my own resin in pellet form and make my own custom filament extruder to get the consistent filament quality I need. Cutting the materials cost to 30% would be an added bonus.
I'm very interested in print quality. It's difficult to polish plastic because it tends to fuzz. I've had good success using a random orbital sander to polish flat plastic plates, particularly black Delrin, to give it a very attractive flat matte velvet finish that is very resistant to fingerprints. I'd like to experiment with various polishing methods to improve the surface finish of printed parts. I have a large rotary drum polisher that's intended to wet polish rocks with abrasive. I use it with small stainless pins and soapy water to polish brass and it does a great job. I'd like to see if I can create a simple batch polishing process that helps remove the last of the layered appearance with no manual labor on each piece.
I've heard the suggestion that I should just get a quick mold made and have an injection molder run my parts. That had been my goal for some time. When the production 3D printing bug began to bite, I forced myself to go online and get a couple of quotes, thinking that the numbers would quickly disabuse me of any crazy ideas of printing my own parts, but the opposite happened. I got an online quote from a Chinese and an American company. Basically, the Chinese company wanted $3200 for the mold and the ABS parts would be 34 cents each. The American company wanted $6400 for the mold and the parts would be $1.33 each. Protomold might have cheaper prices, but probably not too much cheaper. I'd be looking at a decent up front investment of several thousand dollars to get into injection molded parts. It was doable, but the payback would be a long time before I was in the black. For a small business, cash flow is king! I also would be stuck with the die, so there would probably not be any rolling improvements based on customer feedback or Continuous Quality Improvement. I'd also have a fairly long lead time, particularly for the Chinese parts, so my time to market would suffer, and my ability to respond quickly to changes in market demand would not be good. I believe short run 3D printing will be a transformational technology for small companies like mine. I could get my products to market very quickly, I don't need to maintain bulky and expensive inventory, I don't need to deal with the hassles of injection molding and dependence on a vendor, and I'm immediately profitable.
I'm painfully aware of my 3D printing ignorance, and I hate to speculate from a base of ignorance, much less make plans. So far, I'm thinking a lot about small production 3D printing, but my initial goal is simply to build a Hadron ORD Bot and gain some much needed experience. I've been obsessed with this notion lately, so I'm hoping that obsession will lead to a lot of experience in a short time. It also doesn't hurt that I'm not piddling with my 3D printer in my free time. This is now my job!
I'm an electrical engineer with machine design and building experience, and I have a milling machine and lathe in my basement. The question is, can I make a significant contribution to getting 3D printing out of the hobby market and into the small business world? I'm particularly interested in the idea that parts can be printed that can't be made any other way. The infill can be optimized to add strength or rigidity in the direction it's needed, and not have excess material where strength isn't needed. This not only saves the cost of the material, it also saves weight. 3D printed parts can also be printed that have great complexity, about as easily as printing a solid cube of the same mass. I can design complicated parts that wouldn't be cost effective if they were machined or molded.
Anyway, I've been hinting at my short run continuous production plans in several other threads, and it's inevitable that people would have comments about this concept. I didn't want to hijack those other threads, so I thought anyone who might want to speculate along these lines with me and discuss this topic could do so here.
What do you guys think? Interesting idea? Plausible after some work to adopt 3D printing to the low end of the manufacturing market? Has my cheese slipped off its cracker? Am I smoking crack here?