Wednesday, 3 July 2019

Bearing retainer plate - CAM and machining

What is it?

This bracket retains the ball bearing at the bottom of the ballscrew. Modelled in a fetching chestnut brown here:


CAM for the bearing retainer bracket:

Found a bit of mystery metal that's about the correct size. Looks like cold rolled steel, about 1/4" thick. Should work nicely.

The CAM seems simple enough. Face off, drill and chamfer the fixings, machine the cavity, then contour the external profile, leaving support tabs which I will remove later. Finally, chamfer the hole and circumference.




This is what it should end up looking like:



Machining the bearing retainer:
The new Korloy APMT1604 PC5300 inserts arrived from China this week, c/o Aliexpress. Yes, they are a massive improvement on the Mitsubishi inserts. It illustrates clearly that there is a critical difference in the geometry, although it's not blindingly obvious at first glance when you compare them side by side. 

Yet again I've caught myself out by being too aggressive with stepdown and feedrate when facing off strip / thin bar. It managed to make a half decent job but squealed like a stuck pig and left a "suboptimal" surface finish. It's not rocket science but seems to be taking me a while to learn this simple lesson. It'll be OK for now.



The other issue which isn't obvious here is that the 8mm cutter slipped when machining the central pocket. The reason for that is fairly simple - the cutter doesn't have the Weldon flat on its shank, yet I insisted on using it in a side holder with grub screw. I've got away with it previously but here I was doing some fast plunges which caused the slippage. It required me to remeasure the tool length offset and repeat the operation (twice!). This is the final re-run, which at last gave the required result for the contour operation. I then had to repeat the operation for the central bore, which you can see didn't quite end up as a through hole.




The other issue highlighted here was that I had a rather aggressive plunge feedrate. And a rather aggressive speed and feed programmed for the cutter. Like 5000rpm, 800mm/min (feed) and 500mm/min (plunge). It was the plunge wot did it. When you add tabs to a 2D contour, the cutter plunges back in after passing the tab. Lucky I didn't bugger the cutter and / or work. And lucky it was a centre cutting end mill.

Here it is, ready for cutting the support tabs. Looks OK - apart from the chatter marks on the surface.





Seems to be as intended.



And yes, the pulley boss clears the central bore.



Some design tweaks and finish machining for the Bridgeport Z axis assembly

What's wrong, Fatty?
Most of the key compts are now pretty much complete. Overall I'm reasonably happy with the design concept and the relative simplicity(?) and ease of manufacture. However, some additional finessing / fettling will be required before I'm happy - and indeed before the parts will go together.

Thrust bearing:
The original design had a deep groove bearing (radial and axial thrust) at each end of the ballscrew. However, that isn't an approved means of supporting a ballscrew, since one end needs to be floating unless you want an indeterminate preload, possibly including zero preload (= backlash territory). Furthermore, there's really no benefit in having a bearing at the top of the ballscrew. I ended up shimming the bottom bearing on the first assembly but that's crap and fiddly.

For the revised scheme, I went with one bearing - at the bottom of the ballscrew. Obviously this needs to be a dual row bearing, so it can take bidirectional axial loads as well as the radial loads from the belt drive. Stupidly, I left the specification(?) of the bearing to The Stupid Fat Bloke. He simply carried over the existing Chinesium "5201" deep groove bearing. 

The 5201 bearing is a 12mm ID / 32mm OD bearing with an O/A length of 16mm. That's way OTT for what I want / need.


The 4201 is very similar (same ID and OD) but is shorter at 14mm. 
I like the look of that better and have ordered some SKF branded examples from Bearingboys

That of course requires some modification to the design of the motor bracket.....

Remachining the motor bracket bearing housing:
So now I have to shorten the length of the bearing bore by ~2mm. However, I machined the bore to a nominal 32mm and although it came out very nicely, The Stupid Fat Bloke didn't bother to actually check if the bearing would fit in the hole. Details, eh. 

The bearing can just about be forced in with the application of a lot of force / blows but that's not the way to do things. I need to machine a few 10s of microns from the bore. I suspect some of the problem is due to slight noncircularity(?).

In fact it's quite easy to create a new setup in Fusion to allow you to go back in and finesse the final dimensions / modify the design - as long as this only involves material removal of course.

In order to be able to drift the bearing out, it makes sense to reduce the diameter of the shoulder, so I'll do this at the same time.

In the new setup, select the final CAD model as the stock, then define a convenient local origin. In my case, I simply picked up the axis of the bore and the internal flat surface of the housing using the Renishaw probe. The operation doesn't look as if it should be removing any material but in fact it works fine. 




Bearing retainer bracket:
Having shortened the length of the bearing by 2mm, I can increase the thickness of the retainer bracket accordingly. It could do with beefing up somewhat and this is an ideal opportunity to do so. That also requires me to be back in and increase the depth of the cavity in the housing that receives the bracket. The process is similar - modify the CAD model, then apparently cut air. In this operation, I went for a 3D Adaptive to rough out the bulk oif the material, then a 2D Contour to provide a nice finish:


Picking up the part origin using the Renishaw is deadly accurate in comparison with the tolerances I actually get from the machining operation itself.


Went nicely. Sorted.

Chinesium Leeb (rebound) hardness tester

Another mouse accident: I bought one of these  portable Leeb hardness testers  from AliExpress. Arrived fairly promptly but of course TNT ...