6-axis Horizontal Machining Center for Education

[ishi]

I wonder why Gate's MUL Series ATC uses such a big and complex assembly for each and every tool:

MUL Series | Automatic Tool Changers | High Speed CNC Machine Tool Changers | Quick Change Tool Changers | Gate Technologies Inc

Ideally, I would like to use something as simple as a tool fork:

https://www.hsspindles.com/product/hsk-63f-tool-fork/

Of course, this would require some lock mechanism on the double arm, but it feels a lot simpler and cheaper than adding a bulky steel block for each and every tool. Am I missing something here?

UPDATE: I know what I'm missing. For the double arm to reach the tool, the tool must be offset. This is clear on that picture:

http://www.cdsindexers.com/objects/c...r_mul105-2.jpg

It looks like designing this ATC is going to be a lot of fun, and a very significant challenge (as if we did not have enough of it already).

[ishi]

From all the options we've considered, mounting the ATC on top of the machine (see attached pictures) seems to be the one that will make the best use of volumetric constraints, while keeping mechanics as simple as possible. In fact, it seems that we could use Gate's ATC right off the shelf. We'll now work on mounting beams.

[ishi]

We just had a call with RAMPF, and the news are pretty good: if we go with a steel mold, the unit cost of a mineral casting for our machine will be $6,000, and that's including 300 to 400 threaded inserts. That's about half of what we would pay for granite. The main reason for that is the fact that threaded inserts on mineral castings cost $1 each, compared to the $20 that we would have to pay with granite. Of course, initial costs are high: for a machine of the size we're considering, we're looking at $40,000 for the steel mold. If we go for a wood mold that can be used for a single prototype, the cost goes down to $15,000, but the precision is nowhere near as good. In such a context, it's quite likely that we'll bite the bullet and go for a steel mold right away, after a massive amount of planning and prototyping work that is.

In the meantime, we need to refactor the design taking this new process into consideration. A presentation called "CPT Presentation" on the public drive outlines all the things that we need to take into account.

https://drive.google.com/drive/u/3/f...kwBI-?ogsrc=32

Many thanks to all the readers who recommend that we go for mineral casting. I am convinced this is a much better option.

[ishi]

Here is what the machine could look like with the ATC mounted horizontally onto the vertical column, by using beams that would be part of the mineral casting. By the way, it is now assumed that the base and column (and the beams for the ATC) will be part of a single casting. Most importantly, the machine still fits perfectly within a standard shipping container, including all necessary padding and wrapping.

[ishi]

Here is my favorite perspective for the machine.

[ishi]

This one is either really clever or totally stupid: when looking at the machine from underneath (see picture), I got an idea that could dramatically simplify our ATC, by bringing it from 5 axes (yes, the Gate ATC actually needs 5 degrees of motion) down to a single one. That sounds crazy, but bear with me. Imagine that our toolholders where mounted on a fork like this one:

https://www.hsspindles.com/product/hsk-63f-tool-fork/

Now imagine that the tools would be mounted on the same plane as the conveyor's plane, with tips pointing toward the longitudinal axis of the conveyor. With this arrangement, the bottom row of tools in the magazine would have the toolholder sides pointing toward the spindle when the spindle would point up once rotated around its A axis. Now, if we were to have such a configuration, we could use the spindle head's A axis to snap tools in and out of the forks. That would remove the need for a dual arm (which requires 2 degrees of motion), for a tool post pneumatic actuator (another degree of freedom), and a tool post lock (yet another degree of freedom). The only degree of freedom that would be left is the one driving the magazine's conveyor. Therefore, we would need a single servo motor, with direct drive. Simple and effective.

The only problem I can see with this design is that tool length might be limited and that we might need to design a slightly customized fork in order to snap tools in/out through rotation instead of translation, but that seems a lot easier than designing a conventional double arm ATC mechanism. And it would be massively cheaper to manufacture as well. Also, if rotational tool insertion/extraction turns out to be a problem, it should not be too difficult to design a small pneumatic actuator that would clamp/unclamp the tool. That would add a second degree of motion, but 2 is still a lot less than 5...

What I *really* like about this design is that we would be managing to take advantage of our 2-axis spindle head in a really creative manner, dropping a lot of complexity out of conventional designs. In other words, four of the axes of a traditional ATC are being replaced by one of the two axes of the spindle head.

Thoughts?

[ishi]

This one is either really clever or totally stupid: when looking at the machine from underneath (see picture), I got an idea that could dramatically simplify our ATC, by bringing it from 5 axes (yes, the Gate ATC actually needs 5 degrees of motion) down to a single one. That sounds crazy, but bear with me. Imagine that our toolholders where mounted on a fork like this one:

https://www.hsspindles.com/product/hsk-63f-tool-fork/

Now imagine that the tools would be mounted on the same plane as the conveyor's plane, with tips pointing toward the longitudinal axis of the conveyor. With this arrangement, the bottom row of tools in the magazine would have the toolholder sides pointing toward the spindle when the spindle would point up once rotated around its A axis. Now, if we were to have such a configuration, we could use the spindle head's A axis to snap tools in and out of the forks. That would remove the need for a dual arm (which requires 2 degrees of motion), for a tool post pneumatic actuator (another degree of freedom), and a tool post lock (yet another degree of freedom). The only degree of freedom that would be left is the one driving the magazine's conveyor. Therefore, we would need a single servo motor, with direct drive. Simple and effective.

The only problem I can see with this design is that tool length might be limited and that we might need to design a slightly customized fork in order to snap tools in/out through rotation instead of translation, but that seems a lot easier than designing a conventional double arm ATC mechanism. And it would be massively cheaper to manufacture as well.

What I *really* like about this design is that we would be managing to take advantage of our 2-axis spindle head in a really creative manner, dropping a lot of complexity out of conventional designs. In other words, four of the axes of a traditional ATC are being replaced by one of the two axes of the spindle head.

Thoughts?

Another benefit of this design is that it bring the conveyor closer to the column by aligning the links with the axis of the spindle when the spindle head is vertical pointing up (see attached pictures).

[ishi]

I think it's going to work actually...

Here is a first design with 30 tool holders. The legs of the tool posts can be lengthened in order to accommodate longer tools.

***

High res pictures on public drive:

https://drive.google.com/drive/u/3/f...3dQrx?ogsrc=32

[pippin88]

Ishi, I was thinking that you could use the 2 axis spindle directly to grab tools, if it can rotate to look straight up.

You could avoid the tool length issue if the conveyor is orientated the other way - why have the tools point in towards the conveyor? They could move around the edge (similar to a carousel tool changer, but implemented upside down).

[ishi]

Ishi, I was thinking that you could use the 2 axis spindle directly to grab tools, if it can rotate to look straight up.

You could avoid the tool length issue if the conveyor is orientated the other way - why have the tools point in towards the conveyor? They could move around the edge (similar to a carousel tool changer, but implemented upside down).

pippin88,

The spindle *will* grab the tools directly, because it can rotate to look straight up. This is exactly what I am doing in the latest design:

https://drive.google.com/drive/u/3/f...3dQrx?ogsrc=32

And yes, if we put the carousel horizontal instead of vertical, we solve the tool length issue. This will also make the tool holders much more rigid. Brilliant! Thank you pippin88. I'll update the design accordingly.

[ishi]

pippin88,

The spindle *will* grab the tools directly, because it can rotate to look straight up. This is exactly what I am doing in the latest design:

https://drive.google.com/drive/u/3/f...3dQrx?ogsrc=32

And yes, if we put the carousel horizontal instead of vertical, we solve the tool length issue. This will also make the tool holders much more rigid. Brilliant! Thank you pippin88. I'll update the design accordingly.

Here is what the tool magazine will look like. So much better...

[handlewanker]

Hi....would I be right in saying that without the ATC this is a 5 axis and with an ATC it's a 6 axis?
Ian.

[ishi]

Hi....would I be right in saying that without the ATC this is a 5 axis and with an ATC it's a 6 axis?
Ian.

I don't think so. I think you're confusing the axes of the machine and the axes of the ATC. If you look carefully at the pictures shown there, you will see that a typical double arm ATC requires 5 motors/actuators (5 axes), irrespectively of the machine it is attached to (3, 4, 5, 6, n axes machine).

MUL Series | Automatic Tool Changers | High Speed CNC Machine Tool Changers | Quick Change Tool Changers | Gate Technologies Inc

But with our design, we removed 4 out of 5 actuators in the ATC. So, we still have a 6-axis machine (no matter which ATC we use), but our ATC only requires one motor. Of course, that means that our controller requires 7 motor drivers (in fact, many more, because our 3 linear axes on the machine require two drivers each).

This configuration should save $10k to $20k on the ATC alone/

[handlewanker]

Thanks...…..this configuration is quite involved and, just my opinion, very complicated for what is reqd to machine in one hit 5 sides of a work piece....4 sides and the end face etc.....but as that configuration is the way the industry you're aiming at is probably going you're a front runner no doubt with this design.

BTW, I think the conventional design of a vertical mill with a rotary table in a trunnion is simpler and only requires 5 axes to machine a block on all 4 sides and the end face which is what your design is about.....I can't see any other advantages although no doubt there are some.

BTW2....what is the anticipated maximum end mill size for your design?....also what tool shank is in the spindle end?
Ian.

[ishi]

Thanks...…..this configuration is quite involved and, just my opinion, very complicated for what is reqd to machine in one hit 5 sides of a work piece....4 sides and the end face etc.....but as that configuration is the way the industry you're aiming at is probably going you're a front runner no doubt with this design.

BTW, I think the conventional design of a vertical mill with a rotary table in a trunnion is simpler and only requires 5 axes to machine a block on all 4 sides and the end face which is what your design is about.....I can't see any other advantages although no doubt there are some.

BTW2....what is the anticipated maximum end mill size for your design?....also what tool shank is in the spindle end?
Ian.

handlewanker,

This post explains the design of our 6-axis machine:

https://www.cnczone.com/forums/uncat...ml#post2198222

I actually tend to think that it's a simpler configuration than many 5-axis machines using a trunnion table, especially because of the fully-balanced 3+3 design.

The benefits of 6-axis over 5-axis have been demonstrated for quite a while:

6-axis can offer better accuracy than 5-axis alternatives.
6-axis can offer faster feeds and speeks than 5-axis alternatives.
6-axis can improve ergonomics (HMC + VMC in one machine).

There are many different reasons for these benefits, but going through them would require quite a bit of work on the kinematics side of things. A detailed comparison of our configuration against alternative configurations would make for a great masters thesis.

As far as tooling is concerned, we will offer two options: HSK F63 and HSK E40, as can be seen there:

HST310

You'll be able to use pretty much any tool supporting these standards, especially with the new ATC design that won't impose any limit on tool length or diameter (some tools with very large diameter might require that you put them next to small tools on the ATC though, because there is only 105mm between two consecutive links on the conveyor).

[ishi]

The height of the Y-Axis Carriage has been reduced in order to bring it closer to the ATC. This is required in order to mount the conveyor horizontally right on top of the vertical column and have tools picked up with a rotation on the front of the machine instead of the back. A back rotation through the vertical column's hollow spine would have been possible because our A axis supports ±120° rotation, but this felt funky, and if this design has taught me anything, it is that if it does not feel right (or look right), it's probably wrong... Marcel Dassault (the guy who founded the company that makes the Falcon, Mirage, and Rafale aircraft) famously proclaimed that for an airplane to fly well it must be beautiful. I think he was right, and what he said also applies to CNC machines, even though it takes a special kind of person to appreciate the beauty of a CNC machine...

This reduced height will reduce rigidity on the Y-axis a little bit, but I think it's totally manageable, especially with our two motors driving the axis (like the other two linear axes). Deflection should remain minimal.

[ishi]

This one was hard: we had to find a way to mount our ATC conveyor on top of the vertical column while ensuring that our spindle head could properly reach our tools. Well, that's one scenario where using a CAD tool is really handy, because the set of constraints that one has to deal with is quite large. Anyway, here is a first attempt at it. By the look of it, it's going to be a transition step in the design, in the sense that I fully expect things to be streamlined once we figure out ways to improve a few design elements, but this is a good-enough proof of concept for now. As far as I can tell, and assuming that we can prove that rotary tool insertion/extraction won't break our forks or wear them out too fast, we've gotten ourselves an ATC, and it should not require more than a single motor and driver...

As you can see on the picture called "Magazine", the tools ride on the back of a flange on the column's top. This flange is there to extend the length of our Y-axis rail as high up as we can, thereby ensuring that our spindle head can reach the tools in the first place. The picture called "Column" shows how we're taking advantage of our switch to a mineral casting, implementing geometries that would be virtually impossible with regular granite.

[ishi]

I am designing a streamlined Y-Axis carriage in order to extend the reach of the spindle head and increase the thickness of our thinnest mineral casting profiles. Later on in the design process, FEM will tell us whether I've gone too far with this.

[ishi]

I decided to reinforce the back of the Y-Axis Carriage, just in case...

[ishi]

Here is an updated design taking advantage of the streamlined Y-Axis Carriage. The spindle head can easily reach the tools now, while being able to go much lower on the Y-axis as well, which will make it possible to cut underneath a piece around its fixturing post.

***

High res pictures on public drive:

https://drive.google.com/drive/u/3/f...3dQrx?ogsrc=32