[Zorbit]

I suppose a CF milling machine is like a CF tractor - technically possible but commercially pointless. CF sports cars sell, so CF gantrys for high speed applications might be the future.

[pippin88]

Granite / Rock and sand are cheap...

[Sterob]

.... or 400kg/m3 PU foam... .

Who sells this Peter? Any stuff I have found only has a density of ~40kg/m3.


Would you fill each section separately or cut holes on the insides before welding up and fill all in one hit?

[peteeng]

Hi Sterob - No welding, would be bolted up then filled.

https://dalchem.com.au/product/polyu...dfoam-190.html heres some 190kg/m3 I used to use some 350 foam to make planks as plywood replacement in boats. I'm sure Allnex in Oz will have some. See attachment ERA Polymers they have 400kg foam. CF components are common on expensive machines. I've seen parts on Mazaks that are definitely CF. So for fast expensive machines needing dampness its the go to material...

I need a benchmark machine so I think I'm going to design a plate aluminium one to compare to. Milli-A vs Milli-C so to speak... Peter

[Zorbit]

Peter, given your interest in CNC and experience with CF perhaps it would be worth making pick-and-place machines. They could benefit from the properties of CF.

[peteeng]

Hi Pippin - Yes I've just been down the beach and could have brought a few buckets home for free. but do they do the job? This is the problem, then we have to characterise them so we can design for those materials... Peter

[peteeng]

Hi Zorbit - The industrial robot makers already make their top line fast robots from CF. The machine builders and robot builders do not market the fact that things are made from CF. They market performance not materials like a CF sports car does. CF sells various things no doubt, but someone buying a CNC machine wants a machine that does a job and they don't care if its CF or Kryptonite. If it does the job and is within budget they are in. The hobby Maker market probably can't afford CF as we know it. I'll have to figure out another pathway... Peter

[peteeng]

Hi All - Torsion is the area of interest so I made a 200x200mm column 600mm high in infused FG. I then made a 200x200x16mm thick aluminium tube and they are nearly the same torsional stiffness. I applied 10kNm to the columns.The FG (E=30GPa) weighs 43kg and the Al (E=70GPa) weighs 19.4kg . I'll add a bolted al version to this as well to see what that does. The current model is bonded.... When I started on this some 1.5 years ago I decided to use 16mm Al plate for everything... seems I maybe going back there....I've sent some requests out to CF suppliers for any info on short fibre availability.... Peter

[peteeng]

Hi All - I added 8xM6 bolts down each edge of the column. The result is nearly identical to the bonded joint. I did them up to a C8.8 torque. So I'm confident a bolted plate mill will be stiff enough. Plus I'd used a seeping loctite to set the joints once done. So I think I'll start a plate version Milli-Al to establish a comparison design. Peter

We did a lot of research in the EG thread. Will go back to that and review plus get serious and set up a steel fibre panel for modulus testing....

[Zorbit]

How does the thermal expansion of the FG compare with Ally plate ? I read that basalt fibres are 20% stronger, stiffer, and less prone to expansion than E or S glass, and cheaper than CF.

[peteeng]

Hi Zorbit - Basalt is pretty much the same as Eglass all round. Basalt has a higher melting temp than E-glass. An Eglass laminate ha a expansion coefficient of around 9-12e-6 similiar to steel. Alum is about 26 from memory. Basalt is cheaper then CF but more expensive then Eglass. Eglass fibre E=70GPa there are high modulus (HM) glasses available up to 100GPa which are cheaper then basalt. HM glass has been developed mainly for the wind turbine industry which is moving into CF now the blades are getting huge. Peter

[peteeng]

Hi All- so I got up early and built a plate model. It weighs 155kg so is promising from that point of view. Its stiffness in N/um are X8 Y11 and Z12. Mostly in the base is the deflection. So I shall thicken the base and see how we go. First target is 20N/um or more on all axis.Peter

I thickened the base to 100mm and it improved to X10.5 Y15 and Z26 so I'll have to make the column bigger. Its all about the torque on the column...

[mactec54]

Hi Zorbit - The industrial robot makers already make their top line fast robots from CF. The machine builders and robot builders do not market the fact that things are made from CF. They market performance not materials like a CF sports car does. CF sells various things no doubt, but someone buying a CNC machine wants a machine that does a job and they don't care if its CF or Kryptonite. If it does the job and is within budget they are in. The hobby Maker market probably can't afford CF as we know it. I'll have to figure out another pathway... Peter

Those CF robots are not cutting metal it does not matter what they make the robots out of, it would depend on the job they have to do and the speed they need to move

[peteeng]

Hi All,
I have made the column steel and the figures are N/um X36 Y36 and Z34. The column is approx 300x300x16 steel. This seems to be overkill but its really stiff, So perhaps it's a materials for courses thing. Steel for the column Al for the rest and Tetrium for the arm. I'll have a play with the column geometry a bit more. I updated my screen capture software and it now saves in .mov format so I can publish short movies here now....Peter

[pippin88]

Hi All- so I got up early and built a plate model. It weighs 155kg so is promising from that point of view. Its stiffness in N/um are X8 Y11 and Z12. Mostly in the base is the deflection. So I shall thicken the base and see how we go. First target is 20N/um or more on all axis.Peter

I thickened the base to 100mm and it improved to X10.5 Y15 and Z26 so I'll have to make the column bigger. Its all about the torque on the column...

Peter,

Base should be stiffer if you do the bracing between the plates diagonally rather than along Y direction.

[peteeng]

Hi Pippin - The base is quite good. Most of the deflection is in the column (but see below 50% column 50% base). With a plate build you can't really do a diagrid. I'll look at an aluminium column filled with something next. I don't like steel because it has to be painted. I have had machine parts zinc plated before and thats OK. But zinc is not a quality tactile feel. Anodised aluminium always feels good for some reason.... Peter

I looked at the structural loop stiffnesses. If I hold the arm bearings rigid the arm is X 379N/um so uber stiff. If I hold the column bearings it drops to 48N/um and if I hold the machine feet its 26N/um. So the base and column are about the same stiffness. So a bit of work on the base is warranted. But I think its the X foundation bits. Not sure what to call them vs the base itself. Will have a play with those as well. Peter

[peteeng]

Hi Ralph - What is the approx cross section size of the Mikini column? and height? cast iron is 100GPa modulus. Your about to fill in the back? Regards Peter s

[spumco]

Peter,

Mikini column is 150mm deep x 180mm wide at the upper section. Halfway down it tapers to 350mm at the base (including the ribs).

-Ralph

EDIT - it's about 300 wide between the ribs. Entire back is open with cross-ribbing that's about 1/2 the depth. See earlier photo I posted.

[peteeng]

Thanks Ralph so 250x250 is not unreasonable... And aluminium is not as stiff as cast iron and steel is stiffer then cast iron.... Peter

[spumco]

I think (with no math to prove it) that a 250x250 aluminum box filled with your EG would be stiffer than my open-back cast iron column. BTW, the column sides are about 20mm thick if that helps refine your analysis.