6-axis Horizontal Machining Center for Education

[ishi]

Here is the latest version of the full enclosure frame. 200kg of aluminum profiles... And to be clear, every single fastener, nut, bolt, and screw is being included in the design. This is not like what we did for the way covers and inside shell static covers, where we did not pay any attention to fastening or structural integrity. Instead, we're designing a ready-to-assemble blueprint here, including complete bill of materials.

What I am really happy about is that we do not have any unconventional face-to-face mating of profiles, contrary to what we were expecting initially. Instead, the inner frame is much smaller than expected, and is there mostly to provide a solid way for attaching the outer frame to the base casting, while adding sufficient clearance for moving panels.

This came together a lot better than I was expecting actually...

And the flush-mounted static panels will give it a stunning look...

***

High-res picture on public drive:

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

[ishi]

Here is the enclosure bolted onto the machine.

Looking at the picture called "Machine Top Orthographic", it's pretty clear that we'll have plenty of space for our electrical and hydraulic cabinets, on the left and right sides of the vertical column. We'll most likely put one on the left and the other on the right.

Also, the flush mounting of our static panels will ensure that any panel can be removed by simply removing a handful of screws, without having to remove any profile. This will make assembly and maintenance a lot easier than if panels were mounted within the profiles' slots.

***

High-res pictures on public drive:

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

[ishi]

Here is another 4+2 6-axis machine: the Zayer XIOS G:

https://www.zayer.com/en/product/bedtype/xiosg/31

The second video shows how complex their ATC is.

Clearly, if we can make our design work, this will be a major improvement over existing platforms.

[ishi]

Here is how the inner and outer frame sub-assemblies are mounted onto each other, using this fastener:

https://product.item24.de/en/product...acket-8-45776/

I would never use it for anything that requires a lot of rigidity, but for our particular application, this is perfect.

[ishi]

The bracket used to mount the outer frame onto the inner frame on the middle section of the inner frame's left side is positioned right on top of the inner frame's vertical profile that is connected to the angle bracket used to mount the inner frame to the base casting. That way, the weight of the outer frame is directly transferred to this profile, itself properly supported by the massive angle brackets bolted onto the base casting.

This kind of design is what I love about modular building systems like item's: they make it super easy to do this type of thing, and make corrections later on if necessary. This level of flexibility is just awesome... Of course, it's a lot more expensive that custom weldments, but for small production runs, I think it's the best option.

[ishi]

Here are the six angle hinge brackets used to mount the outer frame onto the inner frame.

With the brackets that will be used on the front and back, I don't think we need more to get the right level of rigidity. But we can always add some more later on, without making any changes to the core design.

[ishi]

Here is how the large angle brackets are used to mount the inner frame onto the base casting.

[ishi]

Here are the seven angle hinge brackets used to mount the outer frame onto the inner frame.

Because of the wider gap we have on the front in order to contain our rising front panel, we don't have any additional angle hinge brackets on the front.

At this point, I think that we can consider our outer frame properly secured onto the inner frame.

Next, we can start working on the sliding side panels.

[ishi]

Our moving panels will use item's 6 series construction profiles and THK's SRS-WM 7 miniature rails:

https://product.item24.de/en/product...-6-1001042790/
https://www.thkstore.com/products/li...?thk_size=7591

They will nicely complement our 8 series profiles for the following reason: 8 series have a 40mm x 40mm cross-section, while 6 series have a 30mm x 30mm cross section. For their part, the SRS-WM 7 linear guides have a 9mm height (rail plus slider). As a result, an SRS-WM 7 rail/slider assembly mounted on a 6 series profile will have a total height of 39mm, providing a full millimeter of clearance when paired with an 8 series profile (40mm height). I will be sufficient for dealing with assembly imprecisions, but small enough that it won't let any chip fly through.

I am *really* happy about how this enclosure is shaping up...

[ishi]

Here is a sweet one: sliding side panels with safety glass. I am still working on the details of their drivetrains, but the overall geometry is there, and they look awesome. In designing them and the rising front panel, I am trying to address two requirements:

1. Provide maximum visibility into the machine when the machine is closed.
2. Provide maximum accessibility to the part when the machine is open.

This means getting rid of every static beam that could get in the way of some heavy part loading equipment.

By the same token, I am trying to keep the mechanics of it all as simple as possible.

Obviously, the rising front panel will be a lot trickier than the sliding side panels, but one step at a time...

For the sliding side panel, I am going for the following configuration: each panel will ride on a single rail that will be mounted horizontally on the inner frame. Each panel will have 3 or 4 sliders. At the top of each panel, I will add a couple of guide rods that will ride within the groves of profiles that have been added at the top of the outer frame (actually, I just replaced a pair of 40 x 40 profiles by a pair of 80 x 40 ones). That way, no squaring will be needed, but we should still have plenty of rigidity and very little noise. The guide rods might be equipped with some ball bearings to smooth everything out (still trying to figure out the right assembly though).

***

High-res pictures on public drive:

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

[ishi]

I think we've found a proper way to drive our rising front panels: telescoping air cylinders.

https://www.bimba.com/Products-and-C...es-Telescoping

For the 1,060mm of travel that we need, the cylinder is just 683mm long, which leaves us plenty of space for fasteners. We will use two cylinders, one on each side of the panel assembly. The only tricky part to design will be the telescopic covers, but I think that we can figure that one out.

[ishi]

For controlling our pneumatics circuits, we will use the EB80 platform from Metal Work S.p.a.:

EB80 - Metal Work SpA

I really like their modular design.

[ishi]

We will have three moving panels:

- Two sliding side panels (left and right)
- One telescoping rising front panel (made of two horizontal sections)

All panels will use linear guides (rails and sliders) and pneumatic actuators. This will make the design relatively simple and maintenance-free. By far, it's not the cheapest option, but it's in line with the rest of the machine. And it will give me a great opportunity to learn about pneumatics, which is something that I know absolutely nothing about, beyond the little I could glean from building a LEGO Technic excavator 30 years ago...

Each sliding side panel should be fine with a single cylinder, which will be mounted at the bottom of the panel, thereby ensuring that it's almost colinear with the rail's longitudinal axis. Each sliding panel will have 3 or 4 sliders.

The telescoping rising front panel will be made of two sections, a bottom one with a travel of 600mm, and a top one with a travel of 400mm. Each section will have two rails, mounted on the back sides of the panels. As a result, rails and sliders will be hidden from the operator. The rails for the bottom section will be mounted on the two vertical profiles that have been added to the front side of the outer frame. The rails for the top section will be mounted on two vertical profiles that will be added to the bottom section, perfectly aligned with their matching outer frame's profiles. Each section will have two pairs of 3 sliders in order to provide as much rigidity as possible.

The rods of the telescoping cylinders will be attached to the top profile of the top section. The bottom profile of the top section will also have pins that will engage within matching pins attached to the top profile of the bottom section. Therefore, when rising the front panel, the top section will rise first fully before the bottom section rises, itself being raised by the pin of the top section.

In order to prevent any deflection of the telescoping rising panel assembly alongside the Y axis, the top section will have two guide rods that will engage within vertical groves provided by the left and right side panels. Therefore, the side panels must be extended before the front panel can be raised, and the front panel must be lowered before the side panels can be recessed into the enclosure.

This two-step process will slow things down a bit, but the side panels won't have to be recessed very often. Instead, the most frequent operation will be the rising and lowering of the front panel. This should not take more than 4 or 5 seconds. Clearly, it will be slower than a manually-operated sliding front panel, but it will provide much better clearance, and the coolness factor will be hard to beat...

[ishi]

I usually post on cnczone.com, but the site has been down for a couple of days. Is this a mirror site? If so, I'll post updates to this thread on this site until cnczone.com gets back online.

[ishi]

After a long week-end of work, we have a first version of the complete enclosure. It does not have any drivetrain yet, because our priority was to confirm overall geometries and dimensions. But in my totally biased opinion, the result is positively stunning: the visibility into the machine offered by the front and side panels is fantastic, and the unencumbered access permitted by the complete absence of support pillars or beams is unlike anything I have seen on the market.

https://docs.google.com/presentation...05ca24a66_0_50

Clearly, we still have a lot of work to design the drivetrains, guide rods, and fixtures, but I like to believe that we have a working design. The only tricky piece missing is a roof. Some machines go without one, but with 5 or 6 axes, chips are likely to fly off in any direction, and a proper roof would be a much welcome addition.

The challenge in designing it is that we do not have a lot of headroom if we want the machine *with* enclosure to fit within an intermodal container. In fact, we’re pretty much maxed out at that time, both in terms of height and width. But I am sure that we can figure out some clever options.

[ishi]

Here is a first design for the roof. It is made of two panels, a small one attached to the front rising panel, and a larger one attached to the top of the outer enclosure. These two panels, combined with vertical static panels to be added later on will ensure that no chips fly out of the machine's enclosure when the enclosure is closed.

The small panel attached to the front rising panel is interesting, for it serves an additional purpose: shelf (or counter) when in lowered position. In fact, it was designed to be perfectly flush with the table when the table is moved in its forward-most position alongside the Y axis.

In order to improve the overall ergonomics, thin plates have been added at the bottom of each side panel. The goal is to provide a convenient work area when the front rising panel is lowered, while ensuring that tools cannot fall inside the machine, in a place where they could not be easily picked up.

When designing these panels, we're trying to hide as many of the mounting fixtures as possible. Therefore, the orientation of aluminum profiles in relation to their open t-slots is critically important: we want the overall design to be as "seamless" as possible.

Another thing should be pointed out: the color scheme. Black is for the outside and white is for the inside. This is due to the fact that we want the inside of the machine to be as light as possible. Therefore, using white panels on the inside will ensure that whatever light is provided bounces of surfaces and eventually reaches the piece being installed or cut. A white roof will also ensure that as little heat is transferred from the shop's roof lights to the machine.

With a bit of luck, we'll be able to open and close the roof panel by opening and closing the side panels. By using a couple of pins attached to the roof panel, we should be able to drive it without any actuator, but without a solid connection either. This should make up for the fact that our enclosure assembly won't be super rigid. That being said, I expect quite a few trials and errors before we get in right when time comes to go from CAD to prototype.

***

High-res pictures on public drive:

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

[ishi]

The top panel uses three THK SVS-LV blocks for each rail:

https://www.thkstore.com/products/li...-lv-block.html

[ishi]

Each side panel rides on three THK SHS-LV blocks:

https://www.thkstore.com/products/li...-lv-block.html

[Eldon_Joh]

I usually post on cnczone.com, but the site has been down for a couple of days. Is this a mirror site? If so, I'll post updates to this thread on this site until cnczone.com gets back online.

this is actually the parent website as far as i know.

i know you took mine and some other's advice to make the foundation of that machine much thicker, but from where you're at now i think it would be good for you to hollow out from the top down all the useless material internal to the structure of the vertical axis.

The very deep channel between the vertical axis rails has effectively reduced the vertical axis to two independent columns. i would suggest you hollow out those columns from the top down, a tapered cylinder aka cone or a 4 sided pyramid. the taper is optimal for both the mold required for casting and for moment of inertia considerations.

i've lost track of how big this machine is but i would imagine you could remove about 100 pounds of material from the epoxy granite casting. this weight reduction will significantly increase the resonant frequency of the base structure without decreasing stiffness significantly.

if you were to add that material back to the backside of the machine and make the web that connects the two vertical rails together thicker, you would then increase torsional rigidity and this might be worthwhile.

Attachment 400233
basically a cone projected into the structure as so. all of that can be removed with no impact to rigidity at all.

[ishi]

this is actually the parent website as far as i know.

i know you took mine and some other's advice to make the foundation of that machine much thicker, but from where you're at now i think it would be good for you to hollow out from the top down all the useless material internal to the structure of the vertical axis.

The very deep channel between the vertical axis rails has effectively reduced the vertical axis to two independent columns. i would suggest you hollow out those columns from the top down, a tapered cylinder aka cone or a 4 sided pyramid. the taper is optimal for both the mold required for casting and for moment of inertia considerations.

i've lost track of how big this machine is but i would imagine you could remove about 100 pounds of material from the epoxy granite casting. this weight reduction will significantly increase the resonant frequency of the base structure without decreasing stiffness significantly.

if you were to add that material back to the backside of the machine and make the web that connects the two vertical rails together thicker, you would then increase torsional rigidity and this might be worthwhile.

Attachment 400233
basically a cone projected into the structure as so. all of that can be removed with no impact to rigidity at all.

Eldon_Joh,

Thank you so much for this advice. It makes a ton of sense, and there is no way that I could have come up with that one on my own. I really appreciate. I'll work on that over the coming days.

For reference purposes, think of the machine's envelope as a cube which sides are about 2m long.