[pippin88]

Totally agree that the relative numbers in each paper are what matter and that we can't take the absolute numbers and compare due to different testing methodologies.

I was also surprised that the table on the Okuma M460 does not pass under the portal/gantry. It is regarded as a very stiff machine in the field. True that there is a saddle (on the gantry rather than the table).

My designs have the table passing under the portal.

The integral crossed bearings look interesting. I'm thinking there is a potential weakness / difficulty: maintaining spacing / preventing any relative movement of the 4 cars.

[peteeng]

Hi Pippin - I spent a bit of time understanding the saddle function in Maximus. Turns out saddles are redundant structural loops and exist for convenience vs function. In Maximus I have spilt the traditional saddle into two to provide room for the Z drive. This has not affected the structure in any way that I can determine. I have used eccentric loads, run non linear to check its not a false linear conclusion etc. So I think the crossed bearings are the next step, this makes the load path more direct. In a saddle the loadpath has to zigzag from bearing thru saddle then back to the other bearing. As long as they are are not extreme $$$ and the mini bridge that is needed for the drive foundation is not too onerous in getting it to work or it becomes difficult to assemble. This is the point that hit me last time (was difficult to assm) so I moved back to the saddle. But I haven't put much energy into this possibility yet just left it in the too be investigated bucket. Cheers Peter

[peteeng]

Hi All - Had a meeting with another sheetmetal company today. Mainly to check that my bending details where the same for their machines tooling. Otherwise I'll have to rework the parts!! I've had parts made for Brevis by them and it was good work but they don't do TIG welding. So I'll sort that when I have the parts. Onward, I have to rejig the RFQ into parts vs assemblies so over the next couple of days I;ll get the RFQ to them.

Put some more time into the Tetrium Mill. I think the major features are rounded out so now I'll have to do some basic engineering on it to check the bearings and structure is OK then its into Round 2 redesigning everything at a detail level. This ones a back burner project but it's quite exciting. I make the mould for the Tetrium casting trial tomorrow. Can't talk much about the material as its within a non disclosure arrangement within another project. But they have said if I start commercial application of the material I can talk about it. So I'll make machine parts and slabs for others to use... thought bubbles..

There are lots of "errors" in this model. It's a visualisation level model to get a feel for dimensions and features So it's like a pick the error game for you... Cheers Peter S

[peteeng]

Hi All - Made a quick dummy model to simulate the bearing loads. The Co of the 20mm car is 2426kgf. The Bamburg document uses 500kgf for the tool load so shall I. The worst loaded bearings are the front column bearings. They are being loaded all the time in plunge and cutting operations. The simulation shows that these are loaded at around 600-700kgf in these conditions which means the bearing won't fail catastrophically but will wear out over time. So I think the main features are good and I'll make the bearings a bit further apart if I can to give them more margin. Let the design rest for a while then come back to it soon, got to get maximus rolling again.... Cheers Peter

[pippin88]

Hi Pippin - I spent a bit of time understanding the saddle function in Maximus. Turns out saddles are redundant structural loops and exist for convenience vs function. In Maximus I have spilt the traditional saddle into two to provide room for the Z drive. This has not affected the structure in any way that I can determine. I have used eccentric loads, run non linear to check its not a false linear conclusion etc. So I think the crossed bearings are the next step, this makes the load path more direct. In a saddle the loadpath has to zigzag from bearing thru saddle then back to the other bearing. As long as they are are not extreme $$$ and the mini bridge that is needed for the drive foundation is not too onerous in getting it to work or it becomes difficult to assemble. This is the point that hit me last time (was difficult to assm) so I moved back to the saddle. But I haven't put much energy into this possibility yet just left it in the too be investigated bucket. Cheers Peter

I did have your Maximus design in mind when thinking about the crossed bearings.

I guess the proper rigid rail mounting takes care of everything. Unless faulty / poor tolerance, the bearings should each only move in one plane - along the rail.

I wonder how much the crossed bearings cost? They have potential to simplify construction quite a bit. Just need a bar between two of them for the ball nut mounting.

EDIT: Appears the another term is "orthogonal linear bearings".
"Crossed linear bearings" sometimes refers to a different type of roller bearing setup.

[pippin88]

The orthogonal linear guideways seem to be relatively uncommon. There are a few on ebay for $$$.

Using a saddle does not mandate a zigzag load path - you can use flange bearings that you bolt from underneath and have the bearings directly aligned.

[peteeng]

Hi Pippin - If you place the flanged bearings on top of each other there's not much space for bolts plus you have to use a small bolt, plus you can't check bolt torque in service without breaking the assembly down, plus its difficult to design it as a bottom up assembly. I did this on Scoot V1 and won't do that again... Its all a very fiddly area of design if you are trying to minimise the size of things. But if you have room alls Ok with a saddle. Peter

[pippin88]

I haven't personally used the flange bearings yet. Interesting to hear your experience.

What size bearings were you using?

I would be looking a 30mm or bigger size guideways.

[mactec54]

The orthogonal linear guideways seem to be relatively uncommon. There are a few on ebay for $$$.

Using a saddle does not mandate a zigzag load path - you can use flange bearings that you bolt from underneath and have the bearings directly aligned.

These Cross Linear configuration seems to be only manufactured in China some of the brand names are good and there price is not bad starting at $68 per linear bearing, the problem I see is the spacing of the 2 rails with fabrication you can do your own and have around the same height or even lower if you are creative

I did a Z axes years ago where there was a conflict with mounting standardLinear Bearings photos below in how I over came the problem, this could of been more compact than it is but this one used a Ballscrew in the center

And yes the Linear Bearings are in pockets for perfect alignment / location, I always do the rails and Bearings in machined pockets, you have perfect alignment when you mount them like this, in most cases the linear Bearing needs to be Ground on ( 1 ) side ( 1 ) side is always Ground the other side is rough, so to have matching Bearing I Grind them all together on the ( 1 ) face then machine the pockets for mounting to match the Bearings

[peteeng]

Hi Pippin - The flanged bearings usually have a greater load rating then the normal ones (edit - seems this depends on brand) . I use flanged as they have bigger bolts (stiffer) and a bigger platform which suits the sheet metal type construction I use, as it provides a bigger base.. You can use the size down bolt from the back. I have done this before. My machines so far have used 20mm flanged for all movement. If you place bearings back to back the hole pattern is very close no big deal but this also means you cannot design using a bottom up assembly philosophy. By bottom up I mean you can build it like a pyramid. Each piece you add then check then continue. If you have to build something on the bench because the bolts go in from behind then check the rails for parallel for instance then slide the assm on you cannot then check what you have done. You have to assume that because you checked the bits the total is correct. This may not be the case. Scoot 1 developed a small wobble and because the bolts where buried I had to remove the entire Z axis assembly to figure out what was happening. The bolts tested to correct torque as I pulled it down but something was not right. So further pulling down showed that the thru bolts had bottomed on the rear bearing and didn't in fact pull up correctly. I firstly cut the bolts to suit but in a production environment you can't cut bolts. So I redesigned the saddle wider so the bearings where not on top of each other so all bolts had rear clearance and could be visually checked and checked in assm order. I lost some travel but the result was better.

The normal bearing size is more compact and would give you a smaller footprint like in the case of the saddle design like Mactec's.

Without getting into international politics the crossed linears (as different from crossed rollers) that I posted are from the PMI company in Taiwan. I used PMI's on my first two machines and found them to be very good. I bought them from Linear Technik in Melbourne. Excellent service on a range of industrial machinery parts.

Back to Maximus - One aspect of Maximus and all my sheet metal machines has been the issue of welding. A rule is if you don't have to weld then don't. The columns on Maximus have to be welded and this could (will) introduce distortion. So I looked at having these 3D printed but cost was huge. Now Tetrium is close I think I'll design these as vac cast as an experiment. I didn't get to build the test mould yesterday as the sanding machine at the Shed needed repair so I spent most of the day fixing that. I'll increase its priority maybe do this at home. Lots to do... Keep making Peter

[peteeng]

Hi Pippin 30mm cars are big? Is your machine big? 30mm flanged PMI brand cars are rated at Co=57.8kN or 5.9 tonne. The normals are same. 20mm are 29.5kN which is 3T. Regards Peter

[pippin88]

The static load rating is totally overkill I agree.

Reasons for big rails for my (hopeful) future build:
1. Maximising stiffness and not eating into my stiffness budget. Going up a size on rails is easy and not that costly.
2. Rail / bearing height and ball screws. Ball nuts are big. This means having to machine channels or mount bearings up on blocks to leave enough room for the ball nut. I have 15mm Hiwins on my router and it was a bit of a pain. Going to a bigger linear bearing gives enough room.

[peteeng]

Hi Pippin - agreed, gaining some room for the ball nut is always an issue. Hadn't thought about it that way before maybe I'll do that on the mill. Thanks Peter

[peteeng]

Hi All and Sundry - So I'm cranking up Maximus again. Getting the dwg stack ready for the new supplier and I've drawn up my first Tetrium part. The rear column. I'll have to do some side by side runs to see what thickness to make it vs the stainless version. I'm aiming at about the same stiffness as titanium 100GPa in the casting but I'm an optimist. I'm sure we will get to aluminium at 70GPa. Next thing is to make a mould... I like this part, the web goes to the edges. In the welded one the welds are back from the edges so load transfer is not as efficient, some simulation to be done. Peter

[sundorlaser2019]

Hello,
We are manufacture of fiber optic or CO2 laser equipment in China,if you need any marker,cutter,clener,please do not hesitate to contact me.
James
whatspp:18600395818
Email:[email protected]

[peteeng]

Hello All out there - Been learning Inventor and it has a topology optimiser. Had a play with the mill column. Peter

[peteeng]

Hi All and sundry - I've sent the dwg and model stack to the new metal shop. I await their quotes. I'm excited about Tetrium, tomorrow I cast the first test block. Means I have more control over shape of the parts and in making the parts. Onward to new things... Here's some blocks I made some time ago. The white one is over 10 years old. It's a tooling resin and microspheres made via infusion. Its white because the resin turns white when it exotherms. Then you know its "cooked" correctly. The felt image is a polyester geotextile stack infused and the fibreglass is as it says. I used to make tooling board for a client and I also used to make thick fibreglass samples for electrical tests and mechanical tests.... I've attached the data sheet for that resin here. Its very low viscosity, probably the lowest I have used 120mPas. No need for vibrating this one, it will flow into a box and around the contents quick sticks. Peter

[peteeng]

Hi All and others - I've made some Tetrium, drilled it, routered it and figured out a few things. The first piece is not as dense as I'd like, it came out at 2500kg/m3 so its stiffness would be well down on what I wanted. I've ordered 25kg of fibres so I can make bigger slabs. So I need to figure out a part and make it. Next time will attend to a better packing. Peter

edit - I got excited so looked at the parts on Maximus that I could convert to Tetrium. But as I looked at bits it became apparent that the sheet metal paradigm was hard to convert to cast parts. I looked at the gantry columns but then I thought if I was casting these I would cast the entire bearing carrier and end plates etc as one. So I think I have to start from scratch and design in the casting possibility space. New project!! That's the Mill project... The sheet metal guys had a couple of Q's today so they are working on the quote yippee. Peter

[peteeng]

Hi All - Quotes are in on the sheetmetal. So now I have to collate all the $$$. Been discussing epoxy granite on other threads so I thought I'd stick this here as well. Its a spreadsheet for calculating resin ratios for grain mixes....Peter

[peteeng]

Good day to all you participants and lurkers out there ---- I started this post on 15/3/2019 and its now 28/10/2019 some seven months. Been a journey. Thanks to all those that participated. The costing is in and this will probably be the last formal entry. But happy to answer ongoing questions. Maximus-HD is now in the design polishing phase. My intent is to sell Scoot, my half sheet machine then build Maximus-HD. Max-HD is mainly 6mm stainless steel. SS is used so it does not have to be painted. I have had trouble with powdercoated finishes mucking up fits and sealing holes. Maximus-M is a lighter version and will be based on laminated 3mm zincanneal (ZNA) material. ZNA is 3x cheaper then SS but the multiple part lasering may work agin us. Until I get it drafted I won't know. In parallel to this I have started developing Tetrium a composite material stiffer then Epoxy Granite that I want to build a 500x250x350Z mill from, then more moulded router parts. I may not go back to metal...

Suppliers comments:
laser and bender - requires a minimum run of 3 units to go. I have two interested parties so far so the first run looks like its a go. If your interested please PM me so I can figure out logistics
BST Motion - They say that long lengths of the 15mm rail do not travel well. So I've had to get them in 1500mm lengths so the long axis will need to be joined. I may change to 20mm so long lengths are OK. This is a big rework but if it's best in the long run then so be it.
All other suppliers have green lights!!

I have attached a price list of all the parts and suppliers for interested lurkers. The breakdown is as follows. For a full sheet router designed to be very stiff:

All mechanical parts $3893 australian dollars not including GST for a local purchase. This does not include nuts and bolts...
If you added a Buildbotic controller this would be $778AUD delivered to your ossie door (I assume its cheaper in USA and Canada). It is fully wired and has enough wire tails and leads to be installed quite easily.
and then add the cost of your bench, so I expect without bench, to build a very good machine for <$4800AUD ($3272USD) plus freight.

You can build a bench from welded steel, construction extrusion, plywood, paper mache whatever you are comfortable with.

You can shop for equivalent parts or use better parts such as better bearings or better ball screws, up to your budget. So to wrap this thread up technically.

Been a very heavy investment in FEA and simulation to determine stiff arrangements and weed out bad ideas and highlight the good ones. Maximus in review has boiled down quite well. Seems the 15mm rails, which I used then flicked then used may be flicked again!! All new machines have small glitches so I shan't be in mass production until Max-No1 has run up some hard hours... So a part by part review and final polish then commit to the build. Hope you enjoyed the ride, may do it again with a mill post soon. One underlying issue is a design load and a design stiffness. For stiff routers I expect a design load at tool of 100kgf is adequate and a minimum stiffness of at least 10um/N to cut aluminium. For a mill 500kgf force at the tool and a stiffness of "pick a number from 10-500 um/N depending on what you want to do. A typical VMC is at least 150um/N. When I get Maximus running and check its stiffness I'll come back and publish it here. Happy to answer any questions and will be watching for posts... Peter I suppose I shall add the build blog here when it starts as well

People struggle for a good gantry. If you use Max's gantry back plate and a 10mm or 12mm thick aluminium front plate that would make a dandy all round gantry....Please note in the image of Maximus some parts are not mirrored. I do not mirror until development end if possible as mirrors are sometimes problematical with BOMs and configured parts.

Edit - I have not been idle on the project while waiting for quotes. I have been testing the geared motors and I've set up a suitable driver system for Maximus. I had a play with running the Y axis on one driver (one driver two motors in parallel). It had a mid range vibration I could not get rid of. When it's one driver one motor and 10:1 gearboxes it runs 0- 15m/min very smoothly. This machine (Brevis2) I am going to set up as a laser engraver with a 10W LED until I use the bits on Maximus.