Building a mini lathe...

[stag53]

Hey all,

I'm a grad student at MIT and am taking an undergrad mechanical design class (2.72) to brush up on some of the analytics... the fun part is that we are designing and building a small lathe for the lab portion. It's a "canned" project in the sense that some of the parts are defined (structural components), but we design our own spindle and some of the other sub-assemblies.

Anyhow, I just finished the spindle today. I posted some pics of the shaft with bearings installed and a cross section of my cad model. I will post some pics of the completed assembly later... and will try to post pics of the rest of the build.

The front bearings are matched duplex angular contacts and a conrad in back. Many of the students are doing tapered rollers... I wanted to do something a little more "trick".The spindle shaft is manually machined from 12L14 steel (not hardened... just wound up being easier and faster). Housing is aluminum. Close tolerances on all... was relieved when I dropped the shaft into the housing and it slid in perfect (thanks to some fenagling with emery cloth... hats off to 5bears for the techniques). I've measured ~.001 runout so far with a dial. Not great... but not too shabby for a 3-day build. Will be doing a Donaldson test tomorrow for a better measurement. For a size perspective, the OD of the angular contacts is 47mm and the shaft is 6.75" long.

I've really enjoyed this site. So much info. I've gotten a lot of ideas and information for work and more recently, school... this is really an amazing resource. Thanks!

Adam S

[stag53]

Here's a pic of the assembled spindle. I'm off to the shop to do some tests!

Adam S

[sergizmo]

Adam,

Did you ever re-measure the runout? If it stays high, you may be able to re-machine the threads that hold the chuck with the whole thing assembled in the case. Part off the old threads, and machine new ones on the shaft. That would ensure almost zero runout. It just depends on how much extra material you have protruding. If it isn't enough and you can't use this approach, there may be other ways of improving it. But either way, a working spindle in 3 days is pretty darn good.

Serge

[stag53]

Runout wound up being closer to .0005" when we did it using the Donaldson Ball Reversal method. Thanks for the suggestion about the chuck threads. Not sure if we have enough real estate left on the shaft... I'm going to look into it.