DIY mill...

[irving2008]

That program looks interesting Irving! I wonder if the chip load calculation in Mach3 would correlate with this program? :cheers:

The program doesnt do chip load or other machining calculations. It is a general purpose framework analysis. The estimation of forces on the cutting tip which give rise to the twisting moments that cause the frame to distort is not an easy one.

A 500W motor (approx 0.75HP output) at 32000rpm gives .17Nm torque, so the max cutting force before the motor stalls on a 4mm cutter is .17/.002 = 85N, however it probably wouldnt get close to that in reality.

Another way to look at it is the cutting force = chip size x chip removal energy = h * w * E

where E = power * tool efficiency / volume removal rate

Aluminium requires about 0.6W per mm3 so a 300W output motor will only acheive a removal rate of 180mm3/sec

E = 300 * .75/1.8e-7= 16.7e6 and therefore force for a .5mm deep cut on a 4mm endmill is around 16.7e6 *.0005 *.002 = 16.7N

[irving2008]

Some more calculations using the 'Framework' program.

Option 1 - 50mmx25mmx2mm aluminium rectangular tube for uprights 300mm high (50mm width in X plane), 1 length across the back at the top (50mm width in Z plane). 2 x 16mm horizontal precision rails centrally mounted at top on uprights vertically spaced 100mm (i.e one at 200mm above table and other 300mm above table).

Option 2 - 100mm x 19mm MDF for uprights, 1 length across back (100mm width in Z plane obviously). Rest as above.

in both cases the spindle mount is modelled by a 300mm long x 90mm x 10mm aluminium plate located centrally in Y plane, and at bottom of Z movement. There is 1.7kg dead load (spindle motor) and 20N force in both X and Y planes on the bottom of spindle mount.

The lattice and load diagrams are shown in the attached picture.

The interesting thing is that the X/Y displacement of the spindle tip under 20N load is, for the Aluminium version, 0.22mm/.003mm and for the MDF version is 0.43mm/.004mm which shows that an MDF version is definitely fine for PCB milling (where the load is much smaller) and can probably do lightweight aluminium work but the aluminium frame is prefered for anything more strenuous. These calculations assume all joints are perfect and non-moving. Another interesting thing is that the MDF version weighs 5.4Kg while the Aluminium one is 3.6kg (both including the 2 x 16mm rails and the 'spindle mount'). Aluminium is definitely the way to go, its considerably lighter and more rigid (but not significantly so - doubling the force doubles the offset but thats mainly in the distortion of the spindle mount and rails rather than the frame).

[Robin Hewitt]

Aluminium requires about 0.6W per mm3 so a 300W output motor will only acheive a removal rate of 180mm3/sec

That's the stuff dreams are made of for me, most I dare cut is around 70 and that involves a degree of tool flexing. I have really irritating software that insists on cutting outlines uphill so I have to leave a generous finishing allowance :D

[irving2008]

That's the stuff dreams are made of for me, most I dare cut is around 70 and that involves a degree of tool flexing. I have really irritating software that insists on cutting outlines uphill so I have to leave a generous finishing allowance :D

on what equipment Robin?

[Robin Hewitt]

on what equipment Robin?

One of these, but about 20 years old

http://www.warco.co.uk/productimages/documents/P61.pdf

[irving2008]

One of these, but about 20 years old

http://www.warco.co.uk/productimages/documents/P61.pdf

I'm surprised you get that much flex.. do you know how much, and is it the tool flexing or the mill? what feed rate etc?

[Robin Hewitt]

I'm surprised you get that much flex.. do you know how much, and is it the tool flexing or the mill? what feed rate etc?

Everything bends, the trick is to minimalise it :D

I don't usually post pics of where it goes not so good, but what the heck...

1: Backlash on a change of direction, looks like a ditch, actually 4 or 5 thou deep

2: Chatter, an echo from the roughing out, should have changed speed.

3: Tool flexing, you can see where the tool tip bent inwards and wasn't completely removed by the finishing pass. You know it's the tool becae it gets deeper towards the tip.

[irving2008]

Everything bends, the trick is to minimalise it :D

I don't usually post pics of where it goes not so good, but what the heck...

1: Backlash on a change of direction, looks like a ditch, actually 4 or 5 thou deep

2: Chatter, an echo from the roughing out, should have changed speed.

3: Tool flexing, you can see where the tool tip bent inwards and wasn't completely removed by the finishing pass. You know it's the tool becae it gets deeper towards the tip.

Thanks for sharing Robin, helps know what I'm looking for... :) btw, #3 are you sure the tool flexed rather than had excessive runout or the work was offline? I'm not sure I understand how the tool tip can bend into the work, unless it was catching because of excessive feed rate