Milli a new composite mill kit

Thank you for the response. I always learn something from you smart folks.

I like your market definition. Now I am just trying to figure out which track I am in. LOL.

I was recently in the market for a 3D printer. I found several on the web that were nice but none matched all the criteria I was looking for. So, then I started looking for kits. Many of the kits were designed such that the buyer sourced the linear rails and hot end. Almost all of them were from China. This is what made me think of a buyer sourced base for your mill.

In the end I built my printer from scratch from my own design. It probably cost me more than the kits would have but ultimately I have exactly what I wanted.

I like your idea of a kit mill. I have been looking for a larger replacement for my G0704 conversion. @stef110 made a nice EG machine here https://www.cnczone.com/forums/verti...rum-posts.html that I like but yours might have more possibilities.

Anyhow, I'm keeping an eye out for your kit.

Hi Max - Thanks for participating. Add a small description of what you would want and it will be incorporated in to the mix. I'm about to review Milli and the more info available the better the outcome will be. Regards Peter

GOOOD MORNING out there! Its time to review Milli :_

Background - Some 1.5 years ago I was asked to look at a benchtop mill design. Since then two others have asked the same question. The usual hiccup is the high Z asked for and the long width. These mills developed but the gantry design with a high Z was not stiff enough and became gangly. The other machine became a moving column mill. Both where to be plate aluminium machines. But contracting complex aluminium parts is expensive plus the freight costs of large parts overseas turned out to be very expensive. The development stalled and the designs were parked. A short time ago I reopened the work and decided to combine the two mills into a lifting gantry design again using plate aluminium. But I have lots of capability in composites and this would allow geometry optimisation and part consolidation. So Milli was kicked off as a lifting gantry composite build. During the structural development the lifting gantry was killed off due to complexity and size. The design has boiled down to the moving gantry. Lots of effort in the motor spindle area has uncovered good info. I'm torn between the 86/90mm frame and the 110mm frame. Using a BT-30 power head sets the tone and size of the arm and this flows back to the machine base. The 110mm size results in a large machine. So that's a decision that needs to be made NOW. Other features will be a long X axis. Trying to be a benchtop machine means taking advantage of width is important. The 3 rail, 2 motor table achieves this. My thoughts are to go the 80mm frame and make it smaller. Being my first mill it needs to be simple....

The other aspect is that I decided to make the parts myself so I'm building a router to make moulds. I was doing that anyway but making a machine to make machines is sort of what it is about for me....Regards Peter

asian
90mm 750W $427AUD 1000W $405AUD includes drivers and cable also available from DMM
110mm 800W $502 1.5kw $513AUD 1.8kW $527AUD includes driver and cable

Hmmmmm

Hi all you CNC people out there - I have been playing around with an idea for a honeycomb machine base. The issue with hexagonal honeycomb is getting it to fit into a rectangular space. You have to use half cells or infills and to do all of this parametrically is a pain. So I decided to go with a square "cell" and a squircular hole. This then can be arrayed out in cell modules very easily. So I have a 100mm square cell and No9 will be 1000mm deep and 1200mm wide to start with. The base will have 6mm or 10mm aluminium plate each side and the honeycomb will be Tetrium-A. Now I can start rebuilding Milli No9.... Peter

Hi All - I have played with a diagrid core. I can do this parametrically using "thin" extrusions. I modelled a test base, 112mm thick. I placed a 100kg load in the middle. The solid FG base deflected 6um and the diagrid with 6mm skins and Tetrium-A core deflected 11um so nearly half the stiffness. The solid base weighs 242kg the sandwich weighs 106kg. I'll make the sandwich thicker to compensate... Peter

I made the sandwich 140mm thick with 10mm skins and it behaves the same as the 112mm 6um so perhaps I start there.

I contemplated the sandwich and it gets away from the original casting concept plus its complex, need to keep it simple. So its FG thick infused then backfilled with Tetrium then maybe capped with FG. I'd better make a test coupon of Tetrium for modulus testing to figure that out.

Hi All - I started modifying No8 to get a visualisation going for the direction of No9. So now I can see a few things so will start again.... I also have to decide on the table size 350x800 or 400x800??? 350x700? Will look at a few mills to decide...Peter

820x240 HF-30
400x650 DEM 4000 VMC
F80 TC 800x260mm X400 Y250 Z400 small VMC spindle nose 90-490mm to table

Morning All and Sundry - I've started populating Milli No9. I split the base into two parts for the visualisation and ease of managing the geometry. This model is to sort out the new spaces and new ideas. The table is 350x700 and I'm tending back to the 90mm frame as the 110mm frame creates a big machine. Even a 700x350 table is big for a Maker machine....Peter

Why does the 110mm motor frame dictate a big machine? Just the weight? Power?

Hi Pippin - Mainly the packaging size, makes the arm bigger, which makes the column bigger etc. Plus the MT30 powerhead is big as well. My original idea was a benchtop machine but this is getting big and heavy for a bench. I was thinking two people could pick it up and move it. That's now unlikely... I'm still floating a bit on the motor size. So I'll detail out on the 90mm frame and see where that goes. I'll have to look for gearboxes or build a speed cassette I think, more stuff & more complexity to cope with!! Early days and lots of spirals at the moment....Peter

Hi All - I looked at the Tormach PCCNC440. It has a 600W BLDC motor with a 2 speed pulley system. Manual change pulley. 10krpm top speed doesn't spec a bottom speed. Its table is 457x160mm and it weighs 272kg. cast iron.... X254mm Y159mm and Z254mm so Millis a bit bigger then this.... Peter

Evening All - Put visual Milli No9 together quite quick to see what we have. The table is 700x350 and the tool can reach all of the table. Currently the nose of the spindle can touch the table and rise 500mm. Probably shorten that a touch and raise the nose starting height to about 75mm above the table. Its looking better balanced than the last one. Things to fix - motor mounts, clearance for the belts. Milli is moving on all axis smoothly now, the rebuild or rebirth has done her good. I'd like to make it a bit space aged look, not sure what that is maybe isogriding or diagriding surfaces or covering it in gold tinsel. I hear the International Space station has a had a leak for two years and it accelerated lately. They isolated it in the Russian sector and used floating tea leaves to find the leak. Prior attempts with high tech detectors hadn't found it. They then used 5 min epoxy to plug the hole temporarily... No Dutchmen around with a spare finger? Then their oxygen system failed... poor Russians had to borrow some O2 from the NASA side... its been in space for 20 years due for a rebuild Peter

Hi All got Milli No9 into FE. Used 30GPa for the infused FG parts, aluminium for the cars and steel for the steel parts. Put 1000N at the nose and it moved 18um. So 1000/18=55N/um which is really good. In reality maybe half that so 27N/um is good for me. Main deflection is in the base so will make it a little thicker or embed some steel in it Get further down the track for that one. I picked up more scrap CF today so maybe make the first moulded one from CF. In the X dirn it moved 70um so is less stiff in that dirn. Will look at that a bit closer. In the Z it moved 65um. I'm happy with the footprint so may make the column and base a little bit bigger/stiffer... Peter

Hi All - Been mucking about with the column trying to make it stiffer. I've put a steel tube inside, an aluminium one and I;'ve made it hollow. The FG solid does as well as the steel tube reinforced one. I'll have to make the tube bigger its currently 100mm dia and 10mm thick. Can get a 125x5 in there may look at that or leave it solid, its easier to laminate that way.... Peter

My lack of experience, but won't a solid fibreglass piece of that size be a lot of work to make?

Hi Pippin - I use a vacuum infusion process. The fibre is laid in dry, a vacuum bag applied and deep vacuum created. This vacuum is then used to draw in the resin and saturate the fibres. The actual construction will be fibreglass about 25mm thick, then that cavity will be filled with dry aluminium or steel grit and then vac infused again. The Al or steel fibres will be about the same stiffness as the long fibre material. So I'm just assuming the whole thing will be 30GPa at the moment until I get the details nailed down better. Then I shall model it more accurately. Can do the same with sand or granite. Dry fill the cavity and use vacuum to fill the part. I have done 100mm thick slabs this way in FG, sawdust, glass spheres, talc,,,, Its on my list to make some panels of AL and steel fibres for modulus testing.... Peter

If I use a 125x5 steel tube I only get 7% torsionally stiffer then the solid cast column. The deflection table is the FE results. I'm trying to get the X axis down. An X load puts the column in torsion. The table results have a 100x10mm tube down the middle. Points at geometry wins!! So make it bigger...

I've attached an image of a gudgeon I made for a yacht some time ago. Its 25mm thick FG made on male form as the inside dimension was the important one. It's about the size of a large loaf of bread.

Hi all - Here's Milli being sternly torqued to. I think making the base thicker will improve the other directions... Peter

The FE model weighs 283kg so turning into a heavy object.

:) lost item but I'll put this one in...

I have been using 1000N as the design load for some time then converting that to N/um for comparisons. Today I realised I should just use a forced deflection of 0.001mm and the FE figures out the load to do this. No more conversions. So then I get a good picture when I make a part "rigid". eg here I've made the machine base rigid and you can see the X axis does not change, makes big delta in Y and a small delta in Z. cheers Peter
better do some more commercial work now....

Hi Y'All - I've rationalised the envelope and shortened the arm. Made the column a little bit bigger in girth but not enough. Really pleased with the Y and Z stiffness. X 12N/um Y 30N/um and Z 27N/um . If I make the column rigid the X axis stiffens to 31N/um so thats the top figure. So to double it I have to use a material at 60GPa or aluminium at 70GPa. I looked up some test figures on carbon and it easily gets to 50GPa... but I don't think people would pay for a CF column. May get a quote on CF material to see what it is anyway. I changed the bae thickness from 100mm to 120mm its made a large difference.... If aluminium the column would weigh 77kgs and be 18N/um so maybe steel!! and 24N/um but that would weigh 218kg !! So geometry it has to be must be BIGGER.

Hi All - I made the column 20% bigger that usually doubles the stiffness. I'll run the model tomorrow... ZZZZ time. Peter

The spindle nose can touch the table and the Z is 425mm so its quite high...

Good Morning All & Sundry - I spent a bit of time this morning looking at the column. I made it 270x240 solid aluminium and it didn't really improve much. This got me to wondering as that's a serious chunk of metal for a small mill. So I started making various parts rigid to see what was happening and found that the 20mm rails were deflecting quite a bit. So I've decided that: 1) Its still too big so will reduce the 350Y to 300Y. I like the length (X) so that remains. 2) Will go to 25mm or 30mm rail. Current envelope 700x350x450Z

Will look at the spindle motor mount because that's going to drive the arm width. I'd like to scale back the arm and column. So the current stats are X24 N/um Y60N/um Z38 N/um. Huge Y stiffness. So this is the 5th version of No9 and we now go to No10.... Tally Ho - Peter

I looked back thru the versions and there are at least 15 distinct versions and a few side trips....