Hello Guest it is May 30, 2024, 06:57:37 PM

### Author Topic: 5 axis Spinning Lathe: machine and control panel in progress  (Read 31499 times)

0 Members and 1 Guest are viewing this topic.

#### Dan13

• 1,208
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #30 on: February 03, 2010, 01:52:07 AM »
Amazing! How can one jog that accurate on a CNC machine...? More so when you don't feel the reaction forces (or does this system have servos on the MPGs to simulate the load?)

Daniel

#### poppabear

• 2,235
• Briceville, TN, USA
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #31 on: February 03, 2010, 10:13:28 AM »
no, what he does is form a part slowly with the Jog controls, and watches the part. Since as the part is formed, it is under the load it needs to be at to make the part as he wants it. I.e. if it needs more or less he turns the jog wheel more or less for that axis (upto 2 at once). He doesnt have to calculate deflection etc. since he is making the part then. So when the machine makes it, it moves to the same points he pushed it to, and thus will be under the same load for the same part.  He claims that from part to part, the tolorance between them is +/- 0.0002".

To me that is amazing, I personally dont know sqaut about spinning or the science/skill behind it, I just did the control for the machine.

scott
fun times

#### Dan13

• 1,208
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #32 on: February 03, 2010, 10:25:14 AM »
Hi Scott,

Yes, I figured that was what he was doing, and that was why I was amazed by his ability to jog the machine such as to reproduce the form tool shape. The shape is not trivial at all... the curvature is parabolic or circular or something...

As to the accuracy he claims to achieve, I don't know how one can measure to such accuracy on a curved surface. May be he uses a kind of light reflecting measuring system to measure the part (since they reflect light very well), but I wouldn't think so... Also in 10 degrees difference there would be about 4 times more thermal expansion than the claimed accuracy.

Daniel
« Last Edit: February 03, 2010, 10:30:25 AM by Dan13 »

#### BluePinnacle

• 307
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #33 on: February 03, 2010, 02:25:50 PM »
Good work, lighting - lots of work for the right spinner. We do all microwave stuff, generally C-band to Ku band with the odd one at nearly 40GHz. All ours are hand spun though.

If it helps, large PNC machines usually have a roller of a known shape which is moved across the static tool, and the path recorded, then the gap set according to the stock material thickness and the path recreated based on this - Thus arbitary shapes can be reproduced easily. There's more to it than that of course, it's very much a learned skill rather than a fixed process.

Errors on parabolic reflectors are usually measured using photogrammetric modelling: lots of dots and a couple of coded targets are photographed from various angles using a very decent digital camera; all the shots are then fed into a laptop and the target dots' position calculated relative to each other and relative to a 3D model of the dish. the software is called Photomodeler ... it's very very clever.

We can hold 0.6mm RMS surface error on a 3 metre dish

#### Dan13

• 1,208
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #34 on: February 03, 2010, 02:36:49 PM »
Still don't understand it...  When spinning manually you feel the material and you know when it's already touching the form tool, but when jogging a CNC you don't have any feedback whether you reached the form tool. As I see it, you can easily stall the axis by pushing the tool too hard against the form.

The measuring technology is interesting. Thanks for the info.

0.6mm over 3 metre - WOW! Very impressive.

Daniel

#### BluePinnacle

• 307
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #35 on: February 03, 2010, 02:45:43 PM »
My best guess would be to spring load the tool or use a load cell or other indicator.

Large CNC spinning lathes use hydraulic rams with (I think) linear encoders. very expensive, especially if someone plays with it and drives the ram into a rotating tool that weighs as much as a fully loaded pickup truck. It has happened.

#### Dan13

• 1,208
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #36 on: February 03, 2010, 02:51:44 PM »
Yes, also figured that spring loading the tool could help...

Daniel

#### poppabear

• 2,235
• Briceville, TN, USA
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #37 on: February 03, 2010, 10:28:47 PM »
Dan and Blue,

It is my understanding that Russels users his Jedi Powers, and becomes one with the machine, the universe and everything........  his answer to all questions is 42.........  go figure.....

" Master Russel, Spinning..., Much skill he has"   ** Yoda Circa 20199, at the Yenabe spinning conferance.

or as Russel says:  "Hey, I dont need no stinking load cell"

scott
fun times

#### BluePinnacle

• 307
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #38 on: February 04, 2010, 03:20:13 AM »
^^ Yeah, that's the other way to do it

I've come to the conclusion with my two spinners - Who have clocked up around 40 years experience between them - that it's not quite as simple as it looks, never quite obvious in its mechanism, and flexible enough that it is either surprisingly adaptable when required or utterly treacherous. I think they may be Jedi on the quiet.

incidentally I've just worked out that the contact speed of the edge of a big dish blank, when it is trimmed off with a lathe tool, is over 4,200 surface feet per minute, so it's hardly surprising the swarf flies about like it does.

#### josh

• 101
##### Re: 5 axis Spinning Lathe: machine and control panel in progress
« Reply #39 on: March 04, 2010, 02:09:56 AM »
I was wounder if you can use a pendant like the x-box one that has 2 little joystick.
and install it on a lathe and then if there is a way mach can record every move you make with the two
little joystick than you can save it.
then you can run the next part just like the one you recorded;
is this possible.