[handlewanker]

Looks pretty, I doubt it would fit in my garage......one thing you can be assured of from the beginning, by the time it gets off the ground it'll be obsolete...…...any institution teaching tomorrow's kids today's technology will want to spend as little as possible …..that makes the Chines solution for them a very real prospect......always has been the solution for education suppliers.

BTW, now you have blown the gaff I expect the Chinese, who are very up on their technology too, will already have a model up and running before the ink is dry on this thread...….but I expect you already know that Siemens products are almost totally Chinese component derived, so Trump will love you to bits for going offshore...….by the time this project sees the light of day no doubt Trump will be history.

The end of the road begins when the people you derive your crust of bread from decide that enough is enough and progressively water down your proposals until in the end an EBAY 6040 with a water cooled spindle is their idea of a cheap learning machine even though it isn't a 6 axis...……..3D printing is set to take the World by storm, I'm assured, so perhaps you are designing wagon wheels for tomorrows electric cars.

At the rate technology is evolving, anything made today was already obsolete yesterday.

It has been said that to keep pace with technology, you would need to design a brain to replace itself continuously so that any new design is not an accumulation of known facts but a new approach that is beyond our comprehension...…..the 3D printer is a prime example of comprehending that you could make a product by additive machining before it became common knowledge.
Ian.

[ishi]

Looks pretty, I doubt it would fit in my garage......one thing you can be assured of from the beginning, by the time it gets off the ground it'll be obsolete...…...any institution teaching tomorrow's kids today's technology will want to spend as little as possible …..that makes the Chines solution for them a very real prospect......always has been the solution for education suppliers.

BTW, now you have blown the gaff I expect the Chinese, who are very up on their technology too, will already have a model up and running before the ink is dry on this thread...….but I expect you already know that Siemens products are almost totally Chinese component derived, so Trump will love you to bits for going offshore...….by the time this project sees the light of day no doubt Trump will be history.

The end of the road begins when the people you derive your crust of bread from decide that enough is enough and progressively water down your proposals until in the end an EBAY 6040 with a water cooled spindle is their idea of a cheap learning machine even though it isn't a 6 axis...……..3D printing is set to take the World by storm, I'm assured, so perhaps you are designing wagon wheels for tomorrows electric cars.

At the rate technology is evolving, anything made today was already obsolete yesterday.

It has been said that to keep pace with technology, you would need to design a brain to replace itself continuously so that any new design is not an accumulation of known facts but a new approach that is beyond our comprehension...…..the 3D printer is a prime example of comprehending that you could make a product by additive machining before it became common knowledge.
Ian.

handlewanker,

Pretty is good!

Additive manufacturing still has a long way to go. I personally believe that hybrid (subtractive and additive) is the way to go. Ideally, you want a machine that can do both by simply changing tool.

And as far as politics are concerned, I'm the wrong guy to talk to, at least on this forum.

Cheers!

[mactec54]

Several people have advised that I switch to roller guides, and I think that makes a lot of sense. The best I could find are developed by NSK:

NSK - Roller Guides - High-Performance Seal

No idea what their cost might be. For the X axis, we could go with something like that:

NSK RA251600ANC3-**K43

They are spendy I have some Star Roller Linear Rails ( Bosch Rexroth) that I use o CNC machine rebuilds, and they do make a big difference, Z axes is the most beneficial for the Roller Rails

All the major manufactures make roller linear rails, they are all much the same in quality, NSK make a good product and also very good Ballscrews, I use roller rails when you want an overall better machine, most have more mounting holes as well so if you where using ball linear rails then you will have to add extra mounting holes in your design

Shoulders or pockets are always needed for correct installation for linear rails, I make pockets when I can this eliminates any clamp plates or wedges

[ishi]

mactec54,

Thanks a lot for the advice. I will definitely go for roller rails, and I will check if pockets can be machined on the granite base and column.

[ishi]

Now that the project is squarely focused on the education market, we've decided to build the first prototype in partnership with a local school. We're currently discussing with one that would be perfect for it, but if it does not work out for whatever reason, we'll reach out to other potential candidates. If we can't find one to support the project, we'll take it as an indication that we might have misread the opportunity. But if we find a school to partner with, we'll know that we might be onto something. In other words, a good acid test for the project.

Working with a school for the first prototype will give us access to the following:

- Space
- Plenty of the right kind of electrical power
- Air conditioning
- Metrology equipment
- Lifting tools
- Machines to make some parts with
- Other machines to look at and learn from
- CAD/CAM software
- Expertise
- Extra brains
- Extra pairs of hands

Without a doubt, that would be a lot better than my lonely garage with its 110V electrical circuit...

If we're successful with the first prototype, we'll do a crowdfunding campaign for the first production batch, aimed at getting 25 machines in 25 schools. This would allow us to reduce cost for many parts, but also and most importantly to get a community going. In turn, this community could work on some really interesting projects.

For example, one idea could be to organize a competition among participating schools for the best design of an Automatic Tool Changer for the machine. In the first version, we'll have to go with an off-the-shelf ATC, probably from a Taiwanese supplier. But it would be great to build one that could be made primarily from parts that could be machined on the machine itself.

I am meeting the head of my local school's CAD/CAM/CNC department sometime next week. He used to make parts for NASA. Wish me luck!

[mactec54]

mactec54,

Thanks a lot for the advice. I will definitely go for roller rails, and I will check if pockets can be machined on the granite base and column.

The pockets have to be accurately machined to suit the rail that you are going to use, so you would have to have the rail measurements in your drawings, before the granite machining was done

[skrubol]

mactec54,

Thanks a lot for the advice. I will definitely go for roller rails, and I will check if pockets can be machined on the granite base and column.

From what I've seen of granite based machines (not much, but a few, mostly PCB drilling machines,) the rails are held straight by 2 sets of opposing clamping screws rather than the more common one set of clamping screws holding the rail against a reference surface. Much more effort to install, but should be able to be made as straight as you can measure.

[ishi]

From what I've seen of granite based machines (not much, but a few, mostly PCB drilling machines,) the rails are held straight by 2 sets of opposing clamping screws rather than the more common one set of clamping screws holding the rail against a reference surface. Much more effort to install, but should be able to be made as straight as you can measure.

I will ask the granite base manufacturer.

And to be clear: if granite does not work, we'll go for epoxy granite. I now feel confident enough to do that as well. We would work with a supplier like this one:

Home Page | BaseTek

I'll design everything assuming that we go with granite stone, but keeping in mind that we might have to switch to epoxy granite.

[ishi]

A cold email to HSD's CEO lead to a nice follow-up from their worldwide head of sales. Bad news: the 2-axis head with spindle will cost about $40K. Good news: they did not say no to my request for one sample followed by a small 10 to 20 units batch.

Making progress...

UPDATE: Just talked to HSD USA. They will sell us a single sample for our prototype. We will get all technical documents by tomorrow.

Happy dance...

[ishi]

Just landed back from Tokyo and managed to get going on Fusion360. Here are the two granite pieces, within a 64" cube (to get a bit more room to play with). Shoulders have been added for the rails and inserts for clamping screws will be added in the final design.

[ishi]

I'll add on this post some pictures as the machine is being designed.

The X-axis carriage will be made out of a single iron casting but will be designed a bit later. Just trying to high-level components together first in order to validate the high-level design.

[alannext]

I find this thread very interesting, and very perplexing. I unfortunately just don't get where you are going with all this. Can you give some more insight on the perceived education market? Who is this targeted at, university mechanical engineering depts or ? If so, I would think they would want to design their own machine as a project. If it is to teach machining principles, I think all would prefer a market proven machine that could lead to actual experience that could be applied to the real world, not a one-off machine. Not trying to be dismissive, I love the idea of a challenging project for it's own sake, but help me understand the scaleability to education, especially at the price point you have gotten to...

[ishi]

I find this thread very interesting, and very perplexing. I unfortunately just don't get where you are going with all this. Can you give some more insight on the perceived education market? Who is this targeted at, university mechanical engineering depts or ? If so, I would think they would want to design their own machine as a project. If it is to teach machining principles, I think all would prefer a market proven machine that could lead to actual experience that could be applied to the real world, not a one-off machine. Not trying to be dismissive, I love the idea of a challenging project for it's own sake, but help me understand the scaleability to education, especially at the price point you have gotten to...

alannext,

You can think of the thread and the entire project as an experiment.

The experiment started a year ago when I read this book:

https://pearl-hifi.com/06_Lit_Archiv...l_Accuracy.pdf

I simply fell in love with high-precision machines and decided that I would build my own one day.

But it took me a year to mature the project in the background, without actually working on it. Just letting it sit there until the right time came.

Three weeks ago, the opportunity presented itself to actually start working on the project, and I jumped right in. But I wanted to do it in such a way that the process would make sense to me. For example, I'm not worried about working with a granite supplier for the base, or designing small iron castings. But designing an entire frame out of weldments or iron castings is something that I am not ready for yet. So the project dealt with this unusual set of constraints and went a certain direction. All the while, I got great feedback from this incredible community, and I kept pushing the envelope, trying to see what could be done. This landed us where we are now.

Why do I think that there is a market for this in the education sector? Because there is huge demand for 5-axis machinists in the US and Europe, but not enough students are trained to use them. And when they get some training, it's usually done on simulators, which quite frankly are useless for beginners. Much like you don't learn how to fly a plane by playing Flight Simulator. I know, because I'm a pilot, and I spent a lot of time on Flight Simulator when I was a kid, and I am convinced that it did nothing to make a good pilot out of me. You need to be in the machine, up in the air, on your own for it to really sink in.

I had the same feeling when I used a CNC machine for the first time on my own (ShopBot Buddy). It was intimidating. But the joy of making it work without messing up was immense. And now I am totally comfortable using it. But no amount of YouTube watching, manual reading, and one-on-one coaching replaced this direct experience.

The problem with 5-axis machining is that 5-axis machines are big and expensive, and schools don't let their students spend a lot of time on them, and they don't do it early enough. And if you go into the rarefied space of 6-axis machining, the problem is even worse. Are there even schools out there with 6-axis machines that students have access to.

Therefore, my idea, and it's just an hypothesis at this point, is that schools would benefit from having access to lower cost kits that they could assemble and maintain on their own. Of course, they could do the whole design on their own like I'm doing (and do a much better job at it by the way), but it would be a significant undertaking, and everybody is busy. Also, they would not benefit from doing volume purchases of parts like we would if we were to sell batches of 10 to 20 kits at a time. For example, some of the components like the 2-axis head that I am planning to use are not always available for single unit purchases. Similarly, doing one casting at a time is a really expensive value proposition. But if you're going to do 10 or 20, things start to make sense all of a sudden.

Another thing to consider (and that was wisely pointed out to me on this forum) is that a project like that takes a long time. It's not necessarily a lot of work (even though it is), but it takes a long time from a calendar standpoint. I don't expect to spend thousands of hours working on this (I hope I'm right on that one), but I fully expect to work on it for a year or two, and to get help from many different people at different stages of the project. For that reason, it's a difficult project to undertake for a school, because it would have to be a multi-year endeavor, which would not work for schools that have open enrollment, like most community colleges in the US.

Coming back to 5 or 6 axis, there is another thing that I learned as a pilot: you start becoming a "real" pilot only when you start flying by reference to instruments (IFR, for Instrument Flight Rules). That's what allows you to fly in the clouds, without any visual references. Only when you learn this type of flying do you start to really understand your instrument panel, how the different instruments work together, how you're supposed to navigate in the sky with help from Air Traffic Control, and what will really keep you away from danger.

Well, I'm not yet a 5-axis machinist, but everything I'm seeing and hearing tells me that the very same applies to CNC machining. Anyone can use a 3-axis router. In fact, you can build one for $100 and start cutting paper with a laser. But does that make you a machinist? I don't think so. Only when you upgrade to something like a Haas VM-1 do you start to understand what's really going on. But that's only 3 axes. Once you add the rotary table, you can do a few more things, but it's only when you add that 5-axis that complex kinematics enter into the picture. And that's when you become a modern machinist. From there, the 6th axis is just an incremental step. The big step is going from 4 to 5 axes, and that's what I want to help make more mainstream.

Of course, I could stick to a 5 axis design, by building a conventional 3-axis VMC and adding a 2-axis trunnion table. But the cost of building such a machine would be roughly the same as the cost of building a true 6-axis machine like the one we are designing right now. So we might as well go for 6 axes.

Another thing that took me a long time to appreciate is the benefits of an HMC compared to a VMC. For many applications, horizontal is actually much better, for lots and lots of excellent reasons. Unfortunately, this design remains much less popular than it's vertical cousin, and horizontal machines are usually reserved for much larger envelopes. That's something that I would like to change, but change is hard, which is why a 2-axis head makes a ton of sense, because with such a head, you can use your machine as a VMC as well (albeit with a slightly reduced envelope on the Z axis). Why would you want to do that? To use vacuum tables for example. Therefore, I want a single machine that you can put a student in front of, then tell him/her to figure out the best way to build a part with the machine. That's when the student will have to ask him/herself the really interesting questions, trying to figure out fixturing, tooling, toolpaths, etc. This is even more interesting once you add the rotary table, which you can now use for turning as well. In other words, you have a truly universal machine that can teach you most of what a CNC machinist must learn.

Of course, the minute you use the word "universal", you immediately conjure all kind of analogies like Swiss Army Knife or Jack-of-all-trades-master-of-none. No self-respecting blade lover will ever carry a Swiss Army Knife, and no self-respecting machinist will run intensive production batches on a universal machine. Usually, you're much better off building a custom machine, and that is something that students must learn as well. This is something that took me a really long time to appreciate. I never could figure out why the market for CNC machines was so fragmented. We don't really see it in the US that much because this market is so dominated by Haas, but once you go to Europe or Asia, you realize how many different players there are. But there are good reasons for it: depending on your material, your cutting envelope, your requirements for precision, your production volume, and a host of other parameters, you need very different machines. Well, I want students to develop an appreciation for that as well. I want them to use this universal 6-axis machine in order to learn the basics, then push its envelope as far as they can, to finally understand where the limits are, and when they would be better served switching to a conventional 3-axis VMC, or even a manual lathe.

In the future, most machines will have 6 axes or more, will support hybrid processes (subtractive + additive), and will be interconnected with robots. But to get there, we need better machinists and better software. That, I am convinced of. And this project is a very modest (yet crazy ambitious) attempt at helping the community make a few steps in that direction.

I hope this helps. And if it does not, it's probably that I am totally hopeless, but it does not really matter, because I'm having way too much fun doing this kind of thing.

Cheers!

[ishi]

I find this thread very interesting, and very perplexing. I unfortunately just don't get where you are going with all this. Can you give some more insight on the perceived education market? Who is this targeted at, university mechanical engineering depts or ? If so, I would think they would want to design their own machine as a project. If it is to teach machining principles, I think all would prefer a market proven machine that could lead to actual experience that could be applied to the real world, not a one-off machine. Not trying to be dismissive, I love the idea of a challenging project for it's own sake, but help me understand the scaleability to education, especially at the price point you have gotten to...

alannext,

Another thing that might help you better understand this project is to better understand me personally. I am not particularly good at anything, but I can learn really fast, and I love learning new things. Therefore, from time to time, I pick a new subject, and I try to become an expert at it. At first, I know nothing, and I come across as a total newbie, full of stupid questions and naive ideas. 99% of what I'm saying and writing at first makes no sense at all, but a tiny little bit might. And sometimes, something magical happens: by coming at it from the outside, I stumble upon something that had been missed by the experts, for whatever reason (usually some systemic causes, which tend to create a certain form of myopic atrophy). When I am fortunate enough to make such a find, I dig deeper. Sometimes, it leads nowhere, but I end the journey a bit smarter and wiser than when I started. And sometimes it works and really cool things come out of it.

I did that two and a half years ago with electronics. I knew nothing about it, but I decided to teach myself. To do so, I gave myself the challenge of building the most advanced music synthesizer ever built. That was a real challenge, because I'm neither a EE major (I'm a CS major) nor a musician. But I took the challenge on anyway. This became this:

Ishizeno Instruments

I have yet to build the first one, but the ADC is complete and the DAC is half way there. The synthesizer in the middle is the easy part. Now, what's really interesting with that project is that it allowed me to meet some people who really know what they are talking about. One of them was a guy in Switzerland who is one of the best designers of FPGA-based systems. The kind of guy who has a $500k frequency analyzer on his bench and a couple more on the side just in case the first and second one would break... Well, we started a company together, and last month we released the first version of our product, which is quite possibly the most power-efficient super-computer for big data analytics ever built. What was my contribution to the field? Realizing that the major problem with all existing systems was the lack of bandwidth between compute and storage. Of course, everybody in the field has known that fact since the 50's, but my contribution was to figure out a high-level design that could solve the problem, and so we did. Of course, I'm not the one who actually designed all the electronics for this computer (unlike the ADC and DAC that I have designed for my music box), but no project of real value today can be done by a single individual anyway. In much the same way, I am not trying to build all the components of my machine either. I'm not going to make iron castings in my garden. I'm not going to roll wires around magnets to make my own motors. And I'm not going to write my own CNC control software. I wish I could do all this, and I'm sure I could do any of this if I really wanted to, but there is no point in that exercise. Instead, what interests me is the integration, especially when this integration is made possible only by bringing together different people who have different perspectives on a common subject. That's when the real magic happens. And that's precisely what I'm seeing unfold on this wonderful forum for this particular project.

Thank you so much for your patience and participation.

Best regards
Ismael

[alannext]

appreciate the response Ishi, if you can allow me to float an alternate approach to consider. - If you are really building it for education, I would think work envelope and precision are much less important than overall cost. Perhaps even going so far as to build a machine purpose-built for HDU foam or brass, so it can conceivably be much less expensive (not sure on that but...) This would allow the experimentation of the complex issues of 6 axis machining without worrying about production speed and quality. I have no experience with 4 or 5 axis machining, but wonder if fusion360 supports 5 axis yet? I think that is important as well. As far as 6 axis I'm not clear what tools you even use on the CAM side.

On the other hand, if you are building it for your own use and satisfaction, by all means go high end. I'm a sucker for the best of the best so am with you 100 percent on that, just think you might not be realistic about the education market. I would be happy to be wrong!

[ishi]

Here is the envelope for the Z-axis table. It's 150mm thick, which is the same as the X-axis carriage. The carriage needs that much because of the servo motors, which I want to be fully enclosed. The table needs that (and possibly more) for the rotary table's direct drive torque motor. I'm currently looking for low-profile models, but it's not easy. It's quite likely that we'll have to source it from a supplier other than Siemens. Now on to the Y axis carriage.

[ishi]

appreciate the response Ishi, if you can allow me to float an alternate approach to consider. - If you are really building it for education, I would think work envelope and precision are much less important than overall cost. Perhaps even going so far as to build a machine purpose-built for HDU foam or brass, so it can conceivably be much less expensive (not sure on that but...) This would allow the experimentation of the complex issues of 6 axis machining without worrying about production speed and quality. I have no experience with 4 or 5 axis machining, but wonder if fusion360 supports 5 axis yet? I think that is important as well. As far as 6 axis I'm not clear what tools you even use on the CAM side.

On the other hand, if you are building it for your own use and satisfaction, by all means go high end. I'm a sucker for the best of the best so am with you 100 percent on that, just think you might not be realistic about the education market. I would be happy to be wrong!

alannext,

If I was trying to make a "real" business out of this, I would go for your suggestion, but I'm not. I'm doing several things at once, one of them being to build a statement machine, a bit like an audiophile would with super high-end amps, pre-amps, cables, speakers, etc. The only difference is that I'm not trying to pimp the machine for the bling, I'm doing it for an important reason: accuracy. That's because the kind of parts that I want to make require accuracy. Take a look at the parts made by these guys, I want my machine to be able to make parts exactly like these:

Discommon

Also, the education market is very diverse. Many schools could not afford the machine that I am building, but many can. My school for example has about $5M worth of equipment in its lab, and I know schools that have 2 or 3 times as much, so a $150k or $175k machine is not going to make a big difference for them. In fact, if the machine really works as intended (and with the kind of brainpower that is being gathered around this project, there is no reason to think that it would not), I could see some schools getting several machines to make a full classroom out of it, much like my former school just did with 5 or 6 Haas centers.

As far as software is concerned, you're absolutely right: 6-axis does not really work today, but that's the whole point. If you want it to ever work, you need to start making machines that have 6 axes. Otherwise, you're faced with a catch-22 problem that you can never escape from.

And the more I'm understanding how this machine could be used. I mean, think about it: horizontal, vertical, and turning all in one machine, and all in one setup? How positively cool is that?

[ishi]

Here is the design with head carriage. Of course, without an actual head and spindle, it does not look like much, but the overall proportions are there. Right away, we can tell that it's a pretty dense machine, as in lots of things will be packed within a relatively small volume. This is due to the fact that the head I am using is only 30% smaller diameter-wise than heads that are ten times heavier and are used on machines that are 40 times larger from a volumic standpoint.

Why is that? Because no matter what envelope you have, you need a minimum amount of torque to cut hard materials. As a result, even for relatively small envelopes like mine, there is a minimum size to the spindle you need, hence to its head. In other words, things are not linear, and this is a very important thing to consider. Certain things are, like travel times, but others are not. For example, weight usually goes to the cube of lengths, and price tends to follow the same equation. In fact, 30 years ago, CNC machines used to be sold by the ton. But for things like spindle length, it follows some kind of logarithmic curve, which really puts us at a disadvantage on the low-end of the spectrum. But that's what we have to deal with.

Personally, I find these inter-relations to be absolutely fascinating. So many inter-dependent constraints that you have to deal with when designing such a system! It really blows the mind...

NOTE: The final design will probably use 6 roller guides for the Y axis instead of 4, but NSK's CAD tool only provides support for 1 or 2 guides on a rail, and I got a bit lazy...

[handlewanker]

Hiiiiyyy…..I hope you are carrying enough oxygen when you climb your mount Everest….once you reach the top there's no going back...….as you say, the view going up will be better than the final ascent.....but blizzards happen with monotonous regularity going up mount Everest..

Yesssss…..crowd funding is a solution to the finance problem but you do need to be offering something to encourage the contributors to part with real dough at the end of the trail.....a crate of whisky, beer or a box of chocolates won't cut it.

Usually if you're crowd funding a project you give out some of the final results at the end of the time.....this would normally be actual finished products from the crowdfunded endeavour......I might have misinterpreted this somehow.

How much would you need to ensure the project continues without a slow down due to a funds dry up?......a mill......maybe a mill and a bit...……..if you promised everyone a Haas for a $100K input you would not make any profit this century.
Ian.

[mactec54]

alannext,

If I was trying to make a "real" business out of this, I would go for your suggestion, but I'm not. I'm doing several things at once, one of them being to build a statement machine, a bit like an audiophile would with super high-end amps, pre-amps, cables, speakers, etc. The only difference is that I'm not trying to pimp the machine for the bling, I'm doing it for an important reason: accuracy. That's because the kind of parts that I want to make require accuracy. Take a look at the parts made by these guys, I want my machine to be able to make parts exactly like these:

Discommon

Also, the education market is very diverse. Many schools could not afford the machine that I am building, but many can. My school for example has about $5M worth of equipment in its lab, and I know schools that have 2 or 3 times as much, so a $150k or $175k machine is not going to make a big difference for them. In fact, if the machine really works as intended (and with the kind of brainpower that is being gathered around this project, there is no reason to think that it would not), I could see some schools getting several machines to make a full classroom out of it, much like my former school just did with 5 or 6 Haas centers.

As far as software is concerned, you're absolutely right: 6-axis does not really work today, but that's the whole point. If you want it to ever work, you need to start making machines that have 6 axes. Otherwise, you're faced with a catch-22 problem that you can never escape from.

And the more I'm understanding how this machine could be used. I mean, think about it: horizontal, vertical, and turning all in one machine, and all in one setup? How positively cool is that?

Not trying to discourage you from building your machine, but I think you need to scale back some for your first build

The whiskey flask is not about a machine, any good machinist could do a project like this with regular machines, this is more about design than machining I think you have the two things mixed up, you don't need a precision and supper accurate machine to make parts like this, any regular CNC machine can do parts like this, it just will depend who is programming and running the machine, even a none CNC machine, manual machine can make parts like this, it's more to do with the operator than the machine

They already build machines that do both horizontal / vertical machining, Mill Turn machines use 7 to 12 axes is not uncommon for some years now, so 6 axes and more, is a norm for machines like this

Take a look at this baby machine, and see where they started, https://makezine.com/2015/06/23/5-ax...less-thank-5k/