What's new
What's new

Going From Printed to Production

So are we saying subtractive manufacturing is merely a crutch for current additive technology hurdles and eventually the crutch will be shed? I can buy that.

Personally I think so. Additive manufacturing, if the techniques can be perfected overtime is simply a more refined way to make objects. It uses what it needs and can build internal geometries. Its just that milling and lathing are workhorses with a very mature industry and technical knowhow so it can do things better in some cases, for now. And it can work with a lot more materials than 3d printer can for now.

But in reality building stuff one layer at a time from the ground up has no material limits either. In fact they can probably have more freedom. When building microchips or nanodevices, you build one layer at a time also with lithography and deposition techniques like CVD etc. And you can pretty much deposit whatever material you want, or even grow diamonds and exotic materials. So once this stuff gets figured out and can be scaled over time it will exceed capabilities of CNC milling and lathing for sure. Its only a matter of when that happens. But right now the hurdles are real. Probably in the next 50 years CNC milling lathing will still be around. We are maybe talking 100 years out when additive will probably be the primary method to make basically anything and everything.

But these techniques will never be phased out. The fact is the designer will choose the method to make the object. Sometimes it cheaper/easier/economical/faster/better to make things with either subtractive or additive manufacturing. Then you will choose that method. You can't choose a method because you like the method more innately.

Just as there are nanoscale precision additive techniques with a FIB or an electron beam deposition system, you can also do precision removal of material with those devices too. In fact the FIB was designed first as a material removal system and the additive deposition sterolithography side was developed later. So you have subnanometer precision additive and subtractive capabilities. Both techniques will be used. Theres no one side winning. Whenever I see these hipsters talk about their toy FDM printers like a Makerbot they go into all the glory like additive manufacturing will eliminate subtractive manufacturing in the future. It won't. But additive will probably be the mainstream method to make things in the future -- that I can buy,
 
Quote: For example, if you designed a part with internal geometries and structures that only a 3D printer can produce, then only a 3D printer can produce this. A machinist cannot make complex internal geometries on a mill or lathe. ………. cncdumm

Not so. Casting processes create internal geometries with coring. A machinist can create complex internal geometries by assembling several parts or employing metal forming techniques. There is nothing new about additive manufacturing. Watch a bird building a nest or a plant growing.

But philosophy and semantics aside a very practical way of turning a 3d solid model into a usable metal part which might otherwise be difficult and expensive to machine is by 3d printing the part with expansion factors tailored to the desired metal in investment wax. This is especially useful for small parts of a size to fit in your hand. But not necessarily. Precision Cast Parts in Portland, OR, has built 5 foot diameter turbofan stators as investment castings where separate wax pieces for each blade are assembled around a large wax ring prior to investing and burnout of the mold. And note that a savvy industrial investment casting source will be able to tell you what they can do by combining water soluble wax cores into conventional casting wax to produce castings with curved internal passageways.
Common commercial investment casting materials are brass and other copper alloys, aluminum, 17-4 ph sst and hard to machine steel and nickel alloys.
I've maintained that additive manufacturing by 3d printing will only mature into a serious commercial process once the secondary operations necessary to achieve optimum material properties of plastics, metals and ceramics are mastered by an economically significant number of source suppliers.
The follow-on concept of 3d printing where optimum material properties are achieved in situ concurrent with the layer building will likely involve a combination of truly new materials and complex machinery working outside the size range that involves easily handling by human scale workers. Specifically micro structures (chips, etc.) and large structures of the type that usually involve construction machinery or cranes.
Also keep in mind that the least expensive methods for making consumer products involve molding of thermo plastic resins and metal forming and cutting of metal mill forms, sheet, wire and extrusions. I see it as doubtful that 3d printing will replace any of these methods as the primary manufacturing method for most economically important products.
Ed Weldon
 
Quote: For example, if you designed a part with internal geometries and structures that only a 3D printer can produce, then only a 3D printer can produce this. A machinist cannot make complex internal geometries on a mill or lathe. ………. cncdumm

Not so. Casting processes create internal geometries with coring. A machinist can create complex internal geometries by assembling several parts or employing metal forming techniques. There is nothing new about additive manufacturing. Watch a bird building a nest or a plant growing.

But philosophy and semantics aside a very practical way of turning a 3d solid model into a usable metal part which might otherwise be difficult and expensive to machine is by 3d printing the part with expansion factors tailored to the desired metal in investment wax. This is especially useful for small parts of a size to fit in your hand. But not necessarily. Precision Cast Parts in Portland, OR, has built 5 foot diameter turbofan stators as investment castings where separate wax pieces for each blade are assembled around a large wax ring prior to investing and burnout of the mold. And note that a savvy industrial investment casting source will be able to tell you what they can do by combining water soluble wax cores into conventional casting wax to produce castings with curved internal passageways.
Common commercial investment casting materials are brass and other copper alloys, aluminum, 17-4 ph sst and hard to machine steel and nickel alloys.
I've maintained that additive manufacturing by 3d printing will only mature into a serious commercial process once the secondary operations necessary to achieve optimum material properties of plastics, metals and ceramics are mastered by an economically significant number of source suppliers.
The follow-on concept of 3d printing where optimum material properties are achieved in situ concurrent with the layer building will likely involve a combination of truly new materials and complex machinery working outside the size range that involves easily handling by human scale workers. Specifically micro structures (chips, etc.) and large structures of the type that usually involve construction machinery or cranes.
Also keep in mind that the least expensive methods for making consumer products involve molding of thermo plastic resins and metal forming and cutting of metal mill forms, sheet, wire and extrusions. I see it as doubtful that 3d printing will replace any of these methods as the primary manufacturing method for most economically important products.
Ed Weldon

How is it 'not so'? You can create some internal geometries with casting. Didn't say you can't do that. You can use other even more advanced techniques like thermal forming used by Rolls Royce to create internal frame structure in their titanium blades. But there are limits to what you can achieve. Sterotlithography allows you build absolutely any internal structure you want with much higher degrees of complexity than what you can achieve with a mold.

3D printers will not replace everything. It won't be, Oh 3D systems released Mega3Dprinter and now we can throw away all our CNC millls and lathes. 3D printers will simply play a major role in the manufacturing process and a complement to existing techniques for a while. Its not always about making the actual part. Sometimes its about 3d printing the molds. Eventually though 3d printers can replace everything. Why not? On demand manufacturing. Flexible manufacturing. Perhaps things are made to order in the future. Why have a bunch of stock and supply chain setup to mass produce items when you can never really predict market conditions. Its all too common companies over or under estimate demand and thereby create inefficiencies. On demand, just in time manufacturing streamlines this, and what better than have 3D printers build products for customers on demand.

Imagine a factory that is filled with hundreds of thousands of 3D printers, able and flexible enough to be changed over what they're producing to support a new product line in an instance, without the need to retool the plant. This is true flexible manufacturing of the future. An engineer can design the current generation of Ford mustang. Customers who order will get their cars manufactured on demand. No excess supply. Then the following year, when engineers improve the design and release the latest generation Ford mustang, they will update the design files for their 3D printers so new customers who order that will get those parts printed instead. Or heck, if they want to buy the 1998 model, the engineer can upload that also. Not a chance in one's wildest dream, can you even hope to achieve this kind of flexibility and capability with current techniques where manufacturing plants are tooled to mass produce specific models.
 
cncdumm - Well, I was responding the statement "A machinist cannot make complex internal geometries on a mill or lathe." And kinda playing with words just to get the thinking on this subject out of the fantasizing box and into the real world.
3d printing with the current technologies is great in the product manufacturing world because it enables you as a project engineer to make your new project most visible for everyone on the development team really fast and make it's complexities easier to understand. 3D solid modeling was a big leap from 2d engineering drawings because you could rotate a nice 3d image on a computer screen. But a real physical prototype is a lot better for you can get the folks that are low on the skill set of visualization on board. Sure you can invite them to the engineering conference room or the lab to see the only prototype. But the sooner you can give them something to look at back in their offices the sooner you'll get them producing real creative work or leadership. And it's the leadership types who are often mildly dyslectic and are much better at working with people than visualizing anything that's abstract. And you usually need them on board early on.
The idea of producing products exactly to the specifications of the customer will have some place. But it will never be a big winner in economic scale. Certainly not in the automotive industry the way you describe it. They tried this 60 years ago and gave it up in favor of producing a limited mix of models and accessory combinations. In 1960 fresh out of engineering school when I ordered my first new car, a Ford, there were enough option variations that I calculated they could theoretically make every one of the near million cars they expected to sell that year different from each other. By the time my kids were born they quit doing that and stocked the showrooms with the combos they thought would sell best. It's still that way and they are not likely to change what with all the liability, safety and recall issues today. Besides a good salesman will tell you that the customer will ultimately be more happy following a good sale of a car he can drive away today. That said I can visualize where dealers would offer you a special driver's seat that fit you perfectly. All you would need to do was let them scan your butt and their 3d printer would make a new dealer installed custom safety seat for $1699. This would be a lifesaver. Custom fitted driver's seats are already the thing in professional auto racing. As soon as the materials issues are worked out I predict they will do that with 3d printing. Where else are people going to want something build exactly to their equipment?
One of those niches is served by Shapeways today. I've placed several orders for scale model train parts with them in the past year and expect to buy more.
There will be lots of those opportunities waiting for the right 3d printing technology to rise up. And like I said before the biggest opportunities for the small businessman is going to be perfecting the secondary operations needed to make useful products or parts made to spec for industrial customers. If you own a jobbing metalworking shop and truly want to grow your business for a good future then get on that bandwagon. And if you are just a working slug machinist you would do well to develop knowledge and skills in that direction to support your career. So if you or your boss had one of his best customers come to him one day with a 3d print of some part he needed to produce but needed some additional work on it what would you say? Be ready to say something better that "I dunno" when faced with questions like: Can you machine one or two critical dimensions to close tolerances? Can you put threaded inserts in it? Can you smooth out the surface and get it to look nice? What kind of coating will stick to it? Is there a better material to 3d print it from?
Maybe your customer needs to produce at least 25 3d printed prototype parts for something that will eventually be injection molded. But their engineering model shop and one 3d printer simply can't handle that big a job. Your business better be ready to cheerfully quote him a price that won't leave your boss with a problem making the next payroll. That's 3d printing in the real world today. Are you ready?
Ed Weldon
 
Ed made good points.
The broohahah of mostly, imho, nonsense, regarding supposed benefits of 3d printing ignores some basic realities.

Accurate 3d printing is expensive. Accurate 3d printing is very slow.
Accurate 3d printing will *always* be very slow, because it grows as the cube of pixel size.

Thus, compared to say a face mill planing alu, at maybe 600 inches a minute, the 3d printer might take 30 hours to make the same 30" long piece, *when* running at say 1 micron resolution.
The idea that 3d printing will be used to "manufacture" stuff is complete nonsense.

It is likely that is can be used to "enhance" and enable and better manufacturing, in some very specific, narrow, instances.

Never will there be "factories" with 100.000 3d printers "making" stuff ..
Because a combination of standard manufacturing will be vastly cheaper, in terms of manufacturing quality/cost/speed.

It may be that more stuff will be made of plastics, sure.
Plastics are machined on standard machines every day, very efficiently and quickly, with near-zero tooling cost.
As the plastics get better, and the tooling gets better, machining speeds rise.

I dont doubt that variable-geometry nozzles, speeds etc. will greatly improve the 3d printers capabilities.
And that perhaps multi-material or multilayer technologies can deliver "perfect" surfaces - similar to todays coatings.

Expecting such to deliver under 1$/kg high-speed results, with the engineering properties of todays materials especially in metals, is silly.

There are lots of apps where 3d printers con probably produce an acceptable result.
Toys, kitchenware, handles, etc.
Its highly unlikely they can do so cheaper than metal and plastic injection molding, especially as mold costs are likely sink to near-zero delta vs materials costs.

(Standard, basic) mold costs will go down, certainly, as more automated processes and better tooling costs (cheaper machines, sw, automated mold sw) sink the complexity and the extra costs melt away due to competition.
So the the 3d printed "factory" will be limited to very high end uses (organs, lab stuff, precision jigs maybe),
very low end uses (50 units of "stuff" for real world tests with funtional pieces of the plastic bits), at much higher costs per unit than molding or machining, but easier small-scale delivery,
and specialist uses like special geometries.

Special geometries are not needed, today.
I am not aware of anything urgently needed today, that must have specialist geomtery, at any cost, where such would provide order-of-magnitude improvements.
 
To be an innovator you have to be forward thinking. We are not discussing the current. Clearly I prefer milling and lathing right now too to make components because they are superior right now. But I have used 3D printers extensively in R&D settings. With that said I still think 3D printing is the future. I can repeat again, the FUTURE. So theres no need to drag on with the discussion about practicality in this present day and age that creeps up from time to time. And yes, I have used 3D printing for fully functional devices but again in R&D settings and not for end use products by say the public.

Also, you can think as poorly about 3D printing as a whole all you want. You can paint them all with one broad stroke. The fact is it is the designers and the engineers who will ultimately dictate which manufacturing technique will be used in the future. 3D printing will move forward because of its capabilities for creating complex geometries. All the technical issues of speed of manufacturing, materials, tolerance so and and so fourth will be improved over time, and engineers will take full advantage of every advancement.

Lets look at some real life examples of mass use of 3D printers. The GE9x and LEAP turboprop engines will feature 3D metal printed fuel mixer and injectors. These will be fully functional production level usage of 3D printing in highly critical high performance applications. The superior performance of the fuel mixer is owed to its complex design, and that was only possible because of 3D printing, quote unquote.

One thing you should be aware of is everything is getting really automated. In the future more so. This is not news. But what may be news to some is the design process itself is also getting automated. One of the areas in active development is called topology optimization. This is where engineers tell an algorithm what the engineering specifications are, and a program comes up with an optimal geometry to meet and exceed the specifications. Human involvement even in the design process for these critical components are to be removed in the future. For example, complex highly efficient fuel mixers and high strength to weight ratio structures can be designed with these advanced techniques that simply wasn't possible with an iterative design process with humans at the helm. How does this tie into 3D printing? The output is or the design of these parts that come out of toplogy optimization routines are often very organic and very complex. But it does the job and it does the job well. If you think about nature and biological things, they are very organic too in structure. They're not straight corners, right angles, nice circles like how humans design things nowadays. They form themselves to suit and meet the needs of the application. This is topology optimization. The only way to make these highly advanced performance components will be with 3D printing. Quite literally, the future will be design engineers, then straight to production without needing input from manufacturing engineers or tool and die machinists. Will everyone use these advanced techinques? No. But I expect some of the top engineering firms will. And I expect companies at the forefront of technology will adapt these techniques also.
 
3d printing, more thoughts

I think you are looking pretty far into the future. There is far more to product development engineering than meeting performance requirements. Topology optimization is a neat descriptive term. But the challenge will always be parameter definition.
The term "design engineers" is really a misnomer. The engineer you have in mind will need to cover all the technologies from concept through design, prototype, testing, manufacturing process engineering, etc. And topology optimization for assemblies and finished product will become much more complex when the entry parameters apply to things like reliability, human interfaces and safety (to name just 3) in multiple environments.
The fuel system application for 3d printing that you mention sounds pretty neat. Computational methods for solving field problems like stress analysis, fluid flow and heat transfer combined with solid modeling have brought us a long way in improving performance of many products. Materials science has a lot of catching up to do. Example: Try finding numerical engineering data on fatigue of thermoplastics.
For the near future I have a lot of hope for developments in metal and ceramic structures as well as composite structures using multiple printheads.
I think anywhere a device needs to have minimum material weight for energy conservation of whatever has to move it, machine or human there is an opportunity for 3d printing. I also think the world of architecture and civil engineering offers lots of opportunities. Imagine printing a modular building wall with all the structure, insulation, utility lines and decorative coatings in one pass through a giant 3d printer. Or a complete drop in bathroom with shower, toilet, sink, cabinets ready to take your choice of faucets, showerhead, shower curtain/door, light fixtures and handles.

Me? I'm trying to decide whether to buy a 3d printer or build my own and just buy the printhead and software/steppers/electronics. And then decide what's the best compromise between function and ease of use for the solid modeler software I'll need to create models. What I really want is a 2nd printhead that will print investment casting wax.
Also I'm waiting for reasonably priced scanners to show up. 3 years ago I saw a demo of a NextEngine. Really cool. But at $5K with the software it's still an order of magnitude out of my price range.
Uhhhh... did I ramble too much and get off topic? Sorry guys, Ed Weldon
 
I think you are looking pretty far into the future. There is far more to product development engineering than meeting performance requirements. Topology optimization is a neat descriptive term. But the challenge will always be parameter definition.
The term "design engineers" is really a misnomer. The engineer you have in mind will need to cover all the technologies from concept through design, prototype, testing, manufacturing process engineering, etc. And topology optimization for assemblies and finished product will become much more complex when the entry parameters apply to things like reliability, human interfaces and safety (to name just 3) in multiple environments.
The fuel system application for 3d printing that you mention sounds pretty neat. Computational methods for solving field problems like stress analysis, fluid flow and heat transfer combined with solid modeling have brought us a long way in improving performance of many products. Materials science has a lot of catching up to do. Example: Try finding numerical engineering data on fatigue of thermoplastics.
For the near future I have a lot of hope for developments in metal and ceramic structures as well as composite structures using multiple printheads.
I think anywhere a device needs to have minimum material weight for energy conservation of whatever has to move it, machine or human there is an opportunity for 3d printing. I also think the world of architecture and civil engineering offers lots of opportunities. Imagine printing a modular building wall with all the structure, insulation, utility lines and decorative coatings in one pass through a giant 3d printer. Or a complete drop in bathroom with shower, toilet, sink, cabinets ready to take your choice of faucets, showerhead, shower curtain/door, light fixtures and handles.

Me? I'm trying to decide whether to buy a 3d printer or build my own and just buy the printhead and software/steppers/electronics. And then decide what's the best compromise between function and ease of use for the solid modeler software I'll need to create models. What I really want is a 2nd printhead that will print investment casting wax.
Also I'm waiting for reasonably priced scanners to show up. 3 years ago I saw a demo of a NextEngine. Really cool. But at $5K with the software it's still an order of magnitude out of my price range.
Uhhhh... did I ramble too much and get off topic? Sorry guys, Ed Weldon

Topology optimization isn't just a fancy term. Its being applied and used in the real world for real applications right now. You define the allowable work space, you set the requirements which you will have to already know such as knowing the in-use applied loads in the case of a structural application, and then you solve for an optimal computational solution to achieve certain performance metrics like achieving a target maximum displacement at certain points or achieve a certain factor of safety while minimizing the weight required etc. There's nothing inherently complex or unknown here which is why its being applied already. Heck, there are commercial solvers for this that companies can buy and use turnkey and doesn't require manual coding like it was when this field first started. It is a much more rigorous design technique than the way its currently done with human CAD operators drawing things based on creativity and some engineering principles and then running FEA and then building prototypes and testing and repeating the cycle over and over until you arrive at a final product. TO is rigorous and efficient for use in product design and is an excellent starting point for further refinement. TO ultimately spits out a 3D model, and just like any other product development process, you can then run it through other computational or real world tests on it or do modifications to the basic design as you go along. Anyway, this isn't about T/O. This is just an example for why 3D printing will be used because this is the way products will be designed in the future. Its a natural evolution and advancement of mankind and its beautiful.

The example is to show that TO will often spit out complex shapes with internal geometries that will meet the end goals of the design problem. But that such shape is complex and organic. You will not be creating these objects, either in an R&D setting or for final manufacturing for end use, using mills and lathes anymore. YOu will just 3D print them straight off. Those highly efficient mixers used on the GE9x have probably gone through topology optimization routines to come up with their complex geometry. And they are 3D printed. Its a complex organic structure. Its just more efficient to 3D print it because you simply can't cast that.

Plus, the attitude among the engineers/designers is to do things themselves from bottom up. A 3D printer gives them that capability as they don't need to worry about manufacturability and run it through a machinist. They design and they print and they collect the part. Done.

Do you mean you want a 3D printer for use in your business or for a hobby activity? Not all 3D printers are based on extrusion of something from a print head. Anyway if you're talking about hobby items, you can look at reprap, where parts can be had for $500 if you just want something working. Same with 3D scanners. Heck, you don't even need a scanner. You can just use your phone to take pictures and stitch them together to generate a 3D model with photogrammetry techniques. There are software that assists in this endeavor also.
 
Please point to 2 examples of succesfully doing this.

No such sw exists, and the process is fraught with problems.
A large point cloud with some important non-obvious errors is the end result, requiring all sorts of work to get it done.

The sw approaches are useless for *engineering and manufacturing*, although they can often create pretty toys.
Engineering tolerances are 2-4 orders of magnitude more accurate than can be done by such techniques.

As such, comments like taking photos with camera, in an engineering and manufacturing forum, reduce the credibility of the message.

The westminster cathedral was scanned with a 3d scanner..
BUT .. several months of manual, expert, expensive work went into this.

Please dont quote fluff and nonsense on a manufacturing forum, where many of us actually understand sw, and could write it from scratch, if someone paid for it.
I dont doubt all sorts of "organic" shapes may become useful.
Just like the fuel nozzles used as an example.
Std manufacturing techniques can be used to make them - weather one technique or another is better depend on a number of factors.

Additive manufacturing is not a better technique in and of itself.
It has some benefits, and several important drawbacks.
It may actually have some use, as an addition to std manufacturing.

For example, making small details to be assembled into std parts, may make sense, where the details are prohibitively hard or expensive to make with traditional techniques.

You can just use your phone to take pictures and stitch them together to generate a 3D model with photogrammetry techniques. There are software that assists in this endeavor also.
 
"Plus, the attitude among the engineers/designers is to do things themselves from bottom up. A 3D printer gives them that capability as they don't need to worry about manufacturability and run it through a machinist. They design and they print and they collect the part. Done." ........ cncdumm

This is an old story (just change the name of the process) except for the last word ....... perhaps the phrase "which doesn't fit" is more appropriate.

Thanks, hanermo, for affirming a position I share.

About my stuff. I'm a retired mechanical design engineer. 30 years of that time I straddled the "wall" trying to catch the designs as they flew by and either give them a soft landing or rebuild them. After 40 some years I resisted being turned into a 3dCAD draftsman and retired. But I never lost interest in how to make things. That's really my hobby now that space for building a model train layout eludes me and my aging body makes the race car project in my shop look less practical every year. So when I finish my current lathe building project the next one is looking like 3d printing. cncdumm's mention of the reprap 3d printers is a much welcome confirmation of my suspicion that they are a good place to start. Ed Weldon
 
Please point to 2 examples of succesfully doing this.

No such sw exists, and the process is fraught with problems.
A large point cloud with some important non-obvious errors is the end result, requiring all sorts of work to get it done.

The sw approaches are useless for *engineering and manufacturing*, although they can often create pretty toys.
Engineering tolerances are 2-4 orders of magnitude more accurate than can be done by such techniques.

As such, comments like taking photos with camera, in an engineering and manufacturing forum, reduce the credibility of the message.

The westminster cathedral was scanned with a 3d scanner..
BUT .. several months of manual, expert, expensive work went into this.

Please dont quote fluff and nonsense on a manufacturing forum, where many of us actually understand sw, and could write it from scratch, if someone paid for it.
I dont doubt all sorts of "organic" shapes may become useful.
Just like the fuel nozzles used as an example.
Std manufacturing techniques can be used to make them - weather one technique or another is better depend on a number of factors.

Additive manufacturing is not a better technique in and of itself.
It has some benefits, and several important drawbacks.
It may actually have some use, as an addition to std manufacturing.

For example, making small details to be assembled into std parts, may make sense, where the details are prohibitively hard or expensive to make with traditional techniques.

lol. This is not made up. Just because you're not familiar with some techniques doesn't make it untrue. And you're not being very clear what you're asking. Its also best to use the full form before acronyms. "sw" is software?

And what are you asking? Photogrammetry software or Topology Optimization software available commercially? I know people from grad school or some companies write their own software for both situations in most cases but here are examples of commercial offerings by these companies that one could also buy and use turnkey:

Topology Optimization: COMSOL and ANSYS
Photogrammetry: Geodetic Systems and Autodesk (either standalone software or you can stitch photos together yourself with 3DS max as I've tried also)

No need to get all "gungho" here saying you're so technically advanced and skilled that you can write any software from scratch. You don't need to prove anything to us. You're not even being challenged here. Why take this discussion up a notch and get personal? Capiche?

Again, just because you are unfaimilar, doesn't make these techniques untrue. I really don't expect some oldschool guy from the industry to know it. I expect them to practice what they did know and use it till their own retirement so his replacement can practice better techniques and advance a company forward. And both these unrelated techniques, toplogy optimization and photogrammetry, are used widely in the industry in various capacities.

Regarding 3D scanning by photogrammetry, this isn't anything new. I can tell by your reaction you seem a little shell shocked at the idea some photos can make a 3D object. But to pop your bubble, yes its been around for ages and ages. Its been used by a wide variety of industry for ages. From land survey, to solving crime in forensics. Heck.. you have an iphone? Download 123D App by Autodesk and start having a blast right away with photogrammetry. Yes, even YOU can do it RIGHT NOW. All turnkey. ANd yes, models derived from photogrammetry can be cruder than other point measurement techniques but your model can be scaled back to real life objects to get good dimensions. Again, it depends on one's needs for precision and accuracy. Is photogrammetry going to the world's best method for 3D scanning? No. Did anybody say that here either? No. So you can get off the high horse. But it does sound like Ed Weldon is more of a hobbyist here, and I don't see why a hobbyist needs precision 3D scanners. He's going to scan houshold items and prinit it off his FDM plastic printer. Its not aerospace parts hes banging out here. The suggestion was for a hobbyist, not someone trying to get point clouds of a mechanical component that is best suited for CMM. Yes yes you can keep yapping on about this and that but geez tell us something we don't know. Thats not even what I'm saying. Get real here please.

Finally, I will add that your kind of attitude and approach to new ideas, is exactly why its hard to move forward sometimes in general. People with you kind of attitude to things just like to challenge things even while you have no clear understanding of something or fail to see potential in something. Like 3D printing. Like how you challenge the idea you can do photogrammetry, which is already well established. Just being able to challenge something doesn't add anything useful. It doesn't move people along. Its just ignorance that slows progress most of the time. I can totally see why you're adverse to 3D printing also. You just seem to be the kind.
 
Oh, and I thought I might add. Although this is getting seriously off topic at this point.

There is no inherent limitation to photogrammetry as a technique for creating precise 3D models. Its all geometric based on the data presented in a 2D photograph. So you are really limited by camera resolution and noise (such as clarity and background images). If such were to be improved upon such as camera resolution, then you can in t heory improve part scanning resolution dramatically. Let me just add that in the R&D setting, photogrammetry is often used in unison with other laser scanning techniques to derive the model where you capture a generic outline and you then measure the critical features to obtain critical dimensioning. Thats all that matters. What do you think a casting is in most cases? People only machine out critical components also and leave the rest of the casting crusty as-cast. Same idea.

But I will end by saying I am sure you're a skeptic anyway. It seems to be what you're all about and nothing, no words, certainly not words on a silly post will change that. I can only point you to familiarize yourself with the kind of resolution one can achieve with other software techniques based on using 2D data to extract 3D information. And I did some of this in graduate school. Tomographic reconstruction. It can be with X ray tomography, electron tomography or neutron tomography. These are techniques of using 2D data slices to recreate 3D structures including all of its internal features. A CT scan is something you are more familiar with. But as for resolution? Did you know viral proteins on the order of angstroms scan be 3D scanned with electron tomography on a transmission electron microscope (TEM)? They do this routinely in research. Angstrom level resolution on 2D data slices and nanometer scale 3D resolution. You ponder about that for a second as I'm sure the smallest unit you have to worry about is maybe 1um. If we have no problem getting 3D protein structures by using 2D data, then don't for one second doubt capabilities of using 2D data on other size scales to extract 3D information. All software too. But I am sure you will doubt it anyway so this is just an FYI. It seems to be what many people do most of the time.... question things they don't know.

Sometimes you need to think outside the box. You can't be rigid in life. There are often more than one way of doing things. You might be used to "mechanical" things that you can touch and feel, and think that other techniques are inferior or even impossible. But all that is, is it says more about you than life or the technical art itself. You can question something till you are blue in the face. But if it can be done, it can be done. The universe does not bend to a mortal's distaste for something. A mortal should take advantage of the laws of the universe to their advantage and take full advantage at that.
 
Grimace. The expression on the face of anyone with experience in coding, crashing a numerical controlled machine tool or just working on a project with software engineers. All too often these days we come up with creative ideas and think that we can "throw them over the wall" and there will be an army of software developers there to catch and turn them into working reality in no time flat. Just like we see in current day entertainment fiction. I know, I've noisily advocated or confidently predicted similar impractical solutions myself.
And please think some before you create simple boxes to fit "hobbyists". Out here on the left coast there are a good number of hobbyists who retired early. They could afford to keep doing the things that made their retirement stake at their own pace rather than that of some wacky CEO or money grubbing venture capitalist.
And challenging new ideas? This is the very essence of creative work. Only by challenging each inspiration do we find and accurately plot a course to an uncertain future.
This old man has trouble falling asleep these days. Instead of counting sheep I try to visualize solutions to some problem encountered in the previous day's activity. More often than not the good solution shows upon awakening rather than before I sleep. .................. Ed Weldon
 
Grimace. The expression on the face of anyone with experience in coding, crashing a numerical controlled machine tool or just working on a project with software engineers. All too often these days we come up with creative ideas and think that we can "throw them over the wall" and there will be an army of software developers there to catch and turn them into working reality in no time flat. Just like we see in current day entertainment fiction. I know, I've noisily advocated or confidently predicted similar impractical solutions myself.
And please think some before you create simple boxes to fit "hobbyists". Out here on the left coast there are a good number of hobbyists who retired early. They could afford to keep doing the things that made their retirement stake at their own pace rather than that of some wacky CEO or money grubbing venture capitalist.
And challenging new ideas? This is the very essence of creative work. Only by challenging each inspiration do we find and accurately plot a course to an uncertain future.
This old man has trouble falling asleep these days. Instead of counting sheep I try to visualize solutions to some problem encountered in the previous day's activity. More often than not the good solution shows upon awakening rather than before I sleep. .................. Ed Weldon

hanermo was not challenging new ideas. Hes challenging established technologies. There is a difference. He is questioning an entire section of the technical sciences and questions whether software can do precision part scanning. He is wrong. There is nothing to discuss. End of story. We can 3D scan virus particles and proteins with sub nanometer resolution and it relies heavily on software reconstruction of 2D data to do this. End of discussion already. His comeback might include terms such as "cost prohibitive" or "size of part that can be scanned" so and and so fourth but that just amounts to not willing to admit ultimately one was wrong.

As for your retirement hobby, good luck with that. Find something challenging and fun. Maybe you will invent something that can be useful to the world too as you have all the time to yourself devoted to your hobby. If you want to dip your feet into 3D printing, again those reprap printers are pretty cheap. ANd you have to build it yourself so it will make a fun project. You can also build 3D structured light laser scanners for cheap. I've seen software for that also, so all you really need is a cylindrical glass rod and a laser to create a laser scan line and then use a webcam and the software in a controlled background environment. Or, what is really the most neat in my opinion is simply using your camera to create 3D objects with photogrammetry as mentioned earlier. If you have an iphone, download the app 123D Catch app by autodesk and you can get going fairly quickly.
 
Just wait until someone hooks up a computer controlled wirefeed welder to the printer. Then they will have something that will really print parts.
 
By the way, have any of you built machining, inspection or assembly fixtures by 3d printing of them?
Ed Weldon

Dang, thread-jacking much? I would love to hear the answer to this question but don't want it lost in here. I've seen 3D printed fixtures...for a 3D scanning setup.
 
What would it take to turn this into a production piece?

Check out Senvol. They evaluate the cost of additive production versus traditional production methods. They basically sell a comparison algorithm as a service. Cool company.
 








 
Back
Top