Thanks: 0
Likes: 0
Needs Pictures: 0
Picture(s) thanks: 0
Results 1 to 12 of 12
Thread: Ballscrew Run-out
-
13th Jan 2011, 06:15 PM #1Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Ballscrew Run-out
I am checking every aspect of the fit as I put my Kondia back together after a pull-down and clean.
The Y-axis ballscrew is cantilevered from the front, has a block of aluminium mounted on the nut, and is unsupported except for the mounting bearing at the front.
The screw runs down 1.35 over 400 mm front to back.
The effect of gravity appears to be minimal as it takes considerable force to lift the free end the 1.35 mm, much more than the weight of the nut/block and half the screw.
The TIR close to the front is 0.05, at 400 mm it is 0.14.
As the original acme leadscrew is 25 dia. and the ballscrew 32 dia, I assume a new bearing has been fitted to the flange mounting bolted to the front of the knee.
This bearing may not be perfectly square and could account for the run-out.
The TIR could also be a poor fit in the bearing.
I propose to ignore the TIR as minimal force corrects it.
For the run-out, I plan to shim the mounting flange until a force equal to the weight of the nut/block plus half the screw brings the back end level with the front.
In use the nut/block is bolted firmly to the carriage that carries the table.
Alternative suggestions or comments are welcomed.
John.
-
13th Jan 2011, 06:45 PM #2Dave J Guest
Your plan to shim to get the screw running true and level should be fine. As for the 0.05mm runout I don't think it will bother you and fixing it would mean a lot of work.
Dave
-
13th Jan 2011, 09:42 PM #3Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Thanks for the re-assurance Dave.
I recently got at auction some Starrett rolls of 1/2 " shimstock, Catalog #666, 0.003, 6, 7 and 8 inches.
The auction representative didn't even know what they were !
The mounting flange is 100 mm high, so two thicknesses of 0.007 inches were added under the top edge of the flange.
The run-out was reduced to 0.12 mm without any weight relief, and a very modest force upwards got easily to no run-out.
Onwards to the next problem, trying to work out the separation between the Y-axis nut/block and the carriage so that I can add the necessary spacer, or remove metal from the existing block.
An opportunity for more inventive instrumentation.
John.
-
13th Jan 2011, 10:19 PM #4Dave J Guest
That shimstock was a good score.
Dave
-
13th Jan 2011, 11:19 PM #5Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Dave,
Yes, one does occasionally get a bargain at an auction.
The carriage is on the knee.
Thankfully, I measure a gap between the nut/block and the underside mounting face on the carriage - 0.07 towards the front and 0.15 towards the back.
Machining will not be necessary, just an appropriate shim pack.
I will track down the difference tomorrow, it could be in the carriage itself.
One disappointing aspect discovered is that there has been a loss of Y-axis traverse when the CNC was incorporated.
The manual says 380 mm, I have got 305.
Cannot fix it as the ballscrew nut occupies so much space compared with the original very neat double acme nut used on the original Bridgeport.
John.
-
13th Jan 2011, 11:42 PM #6Dave J Guest
CNC converted machines often lose saddle travel because of the ball nut size, 75mm is a bit of loss though.
Both my mill and lathe cross slide needed shimming. I think from memory the mill needed about 1.4mm shim on the Y and about 1.0mm on the X. I just made up some shims out of brass sheet I had laying around to bring them up and down as needed.
Dave
-
15th Jan 2011, 12:49 PM #7Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Dave,
Fighting back to regain some Y-axis travel.
In the photo below, the ' X ' show edges I plan to machine back and the ' / ' and ' \ ' show corners to get chamfers on the (nut) block and the protector plate.
I only regain 15 mm in total, but every bit of travel helps.
The problem lays in the ballscrew conversion, it was done with block style nuts instead of the flange type.
The photo below from the US series 'Craig Rebuild CNC' shows a flanged nut for the X-axis.
I think Craig re-used the original 'Dog-bone Acme Nut' casting.
Re-using the dog-bone seems much more logical and it provides a better more-rigid connection.
Another aspect:
My Y-axis ballscrew is supported one end only, and the nut in the middle.
Due to its length, should the X-axis be supported both ends ?
John.
-
15th Jan 2011, 01:55 PM #8Dave J Guest
Hi John,
A lot of full size CNC machines are only supported on one end, I think they call it a floating end so I wouldn't worry about that.
I cut mine out as can be seen in the photo's below. Where the tapered part at the front is was originally cast and I have not found it to be any less rigid because there is still plenty of cast around it.
Dave
-
16th Jan 2011, 10:57 AM #9Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Dave,
Why do you taunt us mortals with pictures of your beautiful work ?
For me, cutting the knee seems so drastic !
The planned minor alterations to the nut/block and protector plate should give me about 320 mm Y-axis traverse.
With the ram racked in as far as possible, the table centreline should traverse 20 mm past the quill centre closer to the column.
The table edge should then be 60 mm from the column.
I think that this will be adequate for my needs.
Any gain from cutting into the knee would only get me 5 mm due to ballscrew length.
I tracked down some Bridgeport specifications that show Y-axis as 12" = 305 mm.
So, I am not so bad afterall.
John.
-
17th Jan 2011, 01:39 PM #10Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Update, the nut/block has been machined and the protector plates filed.
The photo shows (darkly - but I will improve when I get some floodlights working) the centreline of the table 20.5 mm on the column side of the spindle with the ram fully retracted and the traverse at 322 mm.
I am considering horizontal guard plates attached to the nut/block just above the ballscrew and below the knee casting top to reduce the impact of swarf.
Not as good as proper telescoping tubing, but better than nothing.
Anyone tried something like this ?
John.
-
17th Jan 2011, 02:36 PM #11Dave J Guest
I made up new plates for my extended travel. It took some time to get it right, but I used cardboard to make up templates so I knew they all worked together without any gaps before making them out of some sheet metal. I ended up making 4 sheet metal covers up instead of the standard 3 to allow for the extra travel. When I wind the saddle back or forward now the Y axis nut catches the slots in the plates and each plate has 15mm overlap on the next which is plenty.
At the moment I have some vinyl from the top of the column down to the back of the saddle to stop any swarf getting on the back of the knee. I made it about 100mm wider each side and don't get any swarf on their or the DRO scales at all now. In the future I will buy some 3mm rubber to go their from Clark rubber, I think it was $30 for 1/2 meter 1200mm wide.
I would recommend you doing this as well especially for CNC.
Dave
-
17th Jan 2011, 07:38 PM #12Golden Member
- Join Date
- Apr 2009
- Location
- Kingswood
- Posts
- 930
Dave,
I have spent some time reviewing my swarf guarding and I can see the merit of your approach.
If I remember correctly, you said earlier that your protector plates were some sort of coated sheet mild steel.
Can you please provide material and thickness details ?
My two SS protector plates are 2.6 mm thick, stiff and heavy.
The trough they slide in is 5.2 deep.
The clearance above the plates to the saddle is 2.0 mm.
I have ended up with two 5 mm gaps between plates at the rear when the saddle is at the maximum forward position.
The gaps would be covered by the rubber guard as you described, but gaps are gaps.
I think I could fit 4 plates each 1.6 mm thick, but the top plate would only have (theoretically) 0.4 mm to guide it.
How deep is your trough and what is the clearance to the saddle ?
I will sketch out tonight a 3-plate arrangement and see what I get.
I will also look at the possibility of using the existing SS plates with an additional MS plate.
PS:
I got a quote for two bellows, each 500 mm wide x 500 mm extended with end mounts.
$747 if I pick them up in Sydney.
The 3mm rubber from Clarke Rubber seems the way to go.
John.