hanermo,
Those are amazing numbers and your reasons make a lot of sense.
The way I look at it is that time invariant error can be reduced or canceled but time variant error can only be reduced. In other words, if a reference surface is not perfectly flat, the user can still measure the error and compensate or scrape to reduce that error. But if the error is from things like vibration which means that the error varies with time, adding mass will reduce it. This kind of error can never be canceled unless you have a way to predict it.
Adding mass lowers the frequency of vibration and can move you further away from a resonance point.
Rick
Rick.Sparber.org
On Sep 13, 2011, at 12:03 AM, "\"hanermo\" - CNC 6-axis Designs" <gcode.fi@gmail.com> wrote:
> 1,2,3: I built a sturctural steel mill and cnc´d a 7x and a 12x lathe.
>
> Mill is accurate to 0.01 mm, or about twice as good as the original Bp
> M-head I use as spindle.
> 7x lathe is accurate to about 0.008 mm TIR.
> 12x lathe has 0.001 mm mechanical resolution on z-axis, and about 0.002
> mm on the x axis.
> (Lathe z-axis has 0.1 micron theoretical resolution on the 1.5 kW servo,
> with 30.000 counts/rev).
>
> 4: Lapping machine, 90% done. On hold.
> 5: Heavy sheetmetal bender for making shields for the mill, in tests.
> Needs paint and handles. Works. Call it 98%.
> 6: 2 micron T&C grinder, about 20% done. Metal, spindle components, some
> cuts, 3D design fairly advanced. Will be 6-7 axis, and cnc.
>
> I use industrial barstock, and industrial cnc drives (geckodrives,
> GraniteDevices), and industrial motiol control components like linear
> slides, HTD belts, etc. for my builds.
> I am building to industrial quality, and am willing to spend significant
> money.
> I test with digital micrometers, and digital 1 micron DTIs.
>
> The lathe is becoming a 7 axis horizontal machining center.
> Last week, I made the mount for the spindle servo, and on friday got a
> hinge for it.
> Need to mount the hinge this week, yesterday connected most servo cabling.
> Lateh spindle has two drives, the original 1.5 kW industrial induction
> motor, and the 1.5 Hp treadmill servo, driven at 68 V DC with a gecko
> drive G320 (Until I get the granite devices drive hooked up).
>
> Still need to test, then disassemble and paint, and then re-wire in
> conduits, put in energy chain, and enclosures.
>
> All machine tools, hand built, can fairly easily be made to deliver
> their basic mechanical accuracy.
> Mills: 0.02 mm
> Lathes: 0.001 mm
> Lapping: 0.0002 mm
> If you are willing to bolt some 50 kg of barstock to it, most all of
> these machines WILL perform to these numbers.
> Handbuilt machines can easily outperform the hobby cheap chinese tools,
> like a 9x lathe or X1 micro mill.
>
> If well built, following correct practice, they can easily be on a par
> to industrial modern machining centres (think of a HAAS mini milling
> machine, VF1 iirc, at 20.000 $).
> If such performance is desired, they cannot be buitl in alu, or from all
> cheap junkyard components.
> However, they can be built about 15x cheaper for 20% of the performance
> and 80% of the accuracy.
> My milling machine will deliver machining performance about equivalent
> to a modern HAAS.
>
> Accuracy on both is about 0.01 mm.
> Repetability on a modern one will be about 2x-3x better, because they
> use 3000$ ballscrews and I use 300$ ballscrews.
> But, the spindle power on mine is 1 Hp, vs 20-30 Hp on modern machines.
> OTOH, I have 20x more machining envelope (1600 mm squared) than the HASS.
> They can make stuff much faster, and have chip conveyors etc. which I
> dont have (yet!).
>
> The lightweight alu construction is the weakest part of the gingery
> designs, imo, ime.
> Adding some mass in ground flat stock, either barstock or cast iron,
> (very cheap and easy in the us) is a way to easily surpass the
> capacities of cheap chinese hobby tooling, and enter into the realm of
> industrial quality tooling.
> Unfortunately, ground flat stock is not available over here.
>
> The only tools you really truly need are (measuring equipment, to suite)
> a drill and an angle grinder. The rest just make is easier, faster and
> more productive.
>
>> Keith,
>>
>> Right you are. I would only be able to give a ball park figure of
>> accuracy. But I think some are wondering if one of these machines can
>> hold 0.001", 0.01", or 0.1". I plan to cut a test cube.
>>
>> Rick
>>
>
>
>
> [Non-text portions of this message have been removed]
>
>
>
> ------------------------------------
>
>
>
>
http://groups.yahoo.com/group/gingery_machines/app/peoplemap/view/map
No comments:
Post a Comment