[peteeng]

Hi All - Commenced this mid Oct now its mid Dec, 2 months to get to the starting line!! I'm looking forward to building this in plywood just to see how good that is... Need to get to FE to prove the pudding... I already have the X rails left over from YaG!.. Peter

Re BT30 - Ard - The BT30 seems to be the up and coming size (or its already the go to size) so I have worked with that. I tried to find out about an iso20 size but no info. There's lots of BT30 tooling and eventually an ATC I imagine. Peter

re weight 55kg in high modulus fibreglass or 60kg in CSA (about) see hollow. Yet to be detailed

[pippin88]

Peter, interesting design.

Not sure that it gains you that much over the smaller triangular moving column designs of the big manufacturers?

Consider moving the Y rails in from the edges. About 1/4 in from edges.

[mactec54]

Hi All - Commenced this mid Oct now its mid Dec, 2 months to get to the starting line!! I'm looking forward to building this in plywood just to see how good that is... Need to get to FE to prove the pudding... I already have the X rails left over from YaG!.. Peter

Re BT30 - Ard - The BT30 seems to be the up and coming size (or its already the go to size) so I have worked with that. I tried to find out about an iso20 size but no info. There's lots of BT30 tooling and eventually an ATC I imagine. Peter

re weight 55kg in high modulus fibreglass or 60kg in CSA (about) see hollow. Yet to be detailed

The Z axis would never need to go that far down to the bottom so if you lift that up and add another support rib behind the bottom rail should be a very good Z axis X axis column any where from 50mm to 100mm is the norm from spindle nose to the top of the Table, 100mm Plus is used mostly

Don't forget the Ballscrews has to have a place to fit

[mactec54]

Hi Cat- I think I'll have some large angles machined or a custom jig machined for alignment. The bridge can't come down too far as it gets into the part space...I'm thinking of reducing the Z spec to 350mm vs 450mm at the mo makes things easier. I'll jump the alignment thing when the time comes I've been looking for economy laser alignment devices if anyone knows of such a thing... The 20N/um has been a huge hurdle. maybe unrealistic for a bench machine...Peter

Aligning columns is same issue on routers. Always a bit of a fiddle. Squares aren't big enough, important things are in air. Ultimately its tuned by cutting and measuring and adjusting. Would be good to have the right metrology stuff...

If Datums are machined for Rail and Bearings which is required for linear Rail installs, machine alignment it is very easy, depending on the rail size I machine pockets an average of 3mm to 4mm deep, install the rails and nothing more to do all rail and Bearing pockets are all machined in one setup so alignment is perfect every time

One of the most important parts of a machine build is it's accuracy when putting it together

[peteeng]

Hi Mactec and Pippin - Thanks for the input. First CAD build is very rough. Have to add registers, ballscrew gutters, fluid gutters and dams. Sort nose bottom and top height etc etc. This is the start really... but I'll do some FE work to establish its major dims first... Peter

[peteeng]

Morning all - have you got your coffee in your hand? Tidied Milli-M and got it into FE. The machine mass is 350kg and that has to reduce. But for the first run its not bad. X 16 Y 10 and Z 21 in N/um. The column or headstone is certainly stiff. In the X the base is the culprit. Flat objects aren't torsionally stiff more depth or more cunning needed. The Y, again the base is soft and the Z is OK... The envelope is 600x250 with a Z=425mm may bring the Z down a little. Current aims are part weight max (for lifting) 50kg and all parts meet that except the headstone at 200kg. Just found that the base frame is aluminium so will change that to steel and rerun... more refining needed... Peter

[peteeng]

Hi All - I corrected the base frame to steel and made the base rigid. This bumped all the stiffness's above 30N/um so the base needs attention....Will think through that and update soon. Peter

[peteeng]

Hi All & Sundry - Well Milli-M5 has nearly hit all the notes. Its base is 130mm thick 40GPa material. That gives it some room if I need to use 60GPa material.

Its stats are X 19 Y 20 and Z 33 in N/um. But it weighs 400kg so now's the time to think about what its real size should be... Plus I think the 20mm cars and rails are letting it down a bit. Current table and envelope is 600x250x440mm. Its definitely not a benchtop machine so need to cut it back a bit...

The gravestone column is certainly stiff. Maybe 500x200x350 is a good size for a first mill with 25mm rails... to more detailing... Peter

[pippin88]

The limited Y axis travel is a big downfall of most benchtop machines.

Don't give in!

[peteeng]

Hi Pippin - so X500mmxY250 OK?? Happy to build a big one after the benchtop machine but the benchtop must be a benchtop machine.... I'm thinking around 200kg so have to weight watch somehow!! A hare & forbes BF-20LV is X480 Y175 and Z280mm The table is 700mmx180 and it weighs 162kg. Perhaps 200kg is too light for the headstone...

or maybe square?? 300x300? up in the air

[pippin88]

For most I think 500x250 is better than 300x300.

The Y travel is the stiffness problem for a C frame / cantilever machine so 500x250 easy to make stiffer than 300x300.

I have the 485x175 travel. The 175 can be quite limiting, but I think 250 would be enough for most uses.

[peteeng]

Hi Pippin - OK I'll compromise 450mmx225mm and we'll see how the Z goes... 600mm table Peter at least its bigger then a Taig...

[lukahr]

Yesterday i went and spend some time in analytic tolls and did Stress analyse of my build.
But as i'm new to this analytic tools i have some questions
What kind of simulation do you do ?
static load and then look at displacement ?

[ardenum]

Yesterday i went and spend some time in analytic tolls and did Stress analyse of my build.
But as i'm new to this analytic tools i have some questions
What kind of simulation do you do ?
static load and then look at displacement ?

The very basic you can do is apply loads to certain areas and see how much they deflect.

[peteeng]

Hi Luka- Build a dummy spindle and short tool. At the tool point apply 1000N load in the X direction, 1000N in the Y direction and 1000N in the Z direction. Three different loadcases not all at same time...1000N is 100kgf and this is quite a high cutting force but the main thing is to then use the FE to calculate the deflections in each direction. Then normalise the deflection and the load eg if the tool moves 0.123mm then the machine stiffness is 1000/0.123=8130N/mm stiffness. ie it takes a 8130N to move the tool 1mm.

Machine designers talk in N/um where 1um = 0.001mm so the prior example will be 1000/0.123/1000 or 8.13N/um. if you look up specs for bearings and ballscrews they will be specified in N/um stiffness. Use linear static solvers. Stresses are usually so low that they are not considered but you can look at them occasionally. Now you also have tp restrain the model. In early analysis you can restrain the machine by its feet. But once you get it refined make a block on the table say 100x100x100 that represents a vice. and hold a couple of sides of that. Now you are applying the load and its load loop is correct from tool to vice.

Depending on your FE software will depend on how things like sliding cars and bolted connections are dealt with. If you use a design load of 100N you will get the same static stiffness. Its the stiffness your chasing, the stiffer the machine the better. Machines range from 1N/um through to 600N/um for commercial special machines. Typical VMC are 100N/um plus... Good luck Peter

Also make the dummy tool infinitely stiff so it does not bend!! Millis target theoretical (model) stiffness is 20N/um X Y and Z and I expect in reality it will be 10N/um plus... once bolts, friction, sliding, preloads are taken into account.

[lukahr]

What material do you pick to simulate Epoxy granite or polymer concrete ? ...is aluminium best pick ?

[peteeng]

Hi Luka - Aluminium has a stiffness or 70GPa this is well above most EG at 35-40GPa. There are many terms for this material; epoxy granite, polymer concrete, mineral casting, synthetic stone list goes on. All of these are composites and its stiffness is dependent on the materials used and the volume ratios of the materials in it. Even granite itself can vary from 30-80GPa so have to make a test piece and check its stiffness to really know.... Epoxy is about 3.5GPa....But to avoid disappointment use a low value say 30GPa... Peter

[ardenum]

What material do you pick to simulate Epoxy granite or polymer concrete ? ...is aluminium best pick ?

You can probably get away with with your regular solver and just defining new material. Here are some values for commercial products(attachment).

I highly suggest going through the whole EG thread on this forum. There's a crap ton of data there.

[peteeng]

Hi Ard and others - The second article in the above post is quite wrong with its EG modulus. I think they are all out by at least a factor of 10. They quote 4GPa for granite/epoxy, epoxy is at least 3Gpa so this has to be wrong. Peter

[ihavenofish]

why would it be wrong?