Here we go again . . . MK4

[routercnc]

Managed to cut another one out this morning:


Another trial fit showing both triangular supports:


Just 2 more to go for the other side.

Here is the current design for the Y/Z axis. You can see the tram plates (small grey rectangular plates) I mentioned earlier which will allow me to tram the spindle relative to the larger green bearing plates underneath.


Initially I will put the WC spindle in the main housing and drive it direct. It might stay this way depending on how much better it is than today.

But if not, the following pictures show the alternative full feature version. The parts will be made to accept either set up.

Main feature is a home made spindle (Tormach TTS drawbar style with R8 collets), driven by the WC spindle via a pulley with different ratios, plus the addition of powered drawbar.


Here is a rear view of the pulley drive from the WC spindle, which is connected via a sliding frame to allow the belt to be tensioned:


This view shows the Z axis with power draw bar. This will give a simple push button tool release. I don't have pneumatics in my workshop so this is electrically driven via a stepper, through a pulley reduction drive, and finally through a ballscrew. Quick calcs show this should have enough power/torque to overcome the pre-load on the belleville washers (part of the drawbar retainer) to release the tool. The whole PDB system floats on sliding bearings so that it squeezes the drawbar against the spindle shaft, rather than loading up the bearings.




Here is an underside picture showing the spindle which bolts in via the lower flange. It is also held in place with the clamps further up the main body, which are also used to hold the WC spindle in direct drive mode:


Here is a cutaway view of the spindle showing housing, bearings, draw bar and collet system. At the top you can see the pulleys for the drive and the belleville washers which hold the tool in place by pulling up the R8 collet, which in turn squeezes the 3/4" shaft of the tool holder. Google Tormach TTS for more info.

(The power drawbar assembly is not shown in this drawing)

Still some finessing to do (e.g. tube connecting AC bearings in nose to upper bearings not in correct alignment etc.), plus still have to convince myself that I am capable of machining the spindle in the correct steel (oversize by ~0.3mm), get it hardened, then grind the bearing journals and the R8 taper to all be parallel and aligned again as they will be distorted. I'll need to make a fixture for my lathe to hold a die grinder or some other idea.
This is the biggest part of the project, the most risky, and the main reason I may just stick with the direct drive set up. But the attraction of big torque boost, the ability to run larger cutters, and the quick tool change keeps me interested.

[Ross77]

Very nice, look forward to seeing the spindle progress.

[Boyan Silyavski]

Now thats the most out of the box z design i have seen recently. No shame in using 8x Hiwin bearing blocks and 4x rails plus a couple of motors apart from the Hf spindle. How would you align all that stuff in place? Each piece must be carefully machined to a tight tolerance.

Ok. I know that you will patiently make it. Obviously not your first machine 2 questions:

1. Wasn't it simpler and better using BT30 spindle and servo motor on a fixed gantry?

2. So much effort in all places but gantry will slide on round open cage bearings? No square supported ones?

[routercnc]

Now thats the most out of the box z design i have seen recently. No shame in using 8x Hiwin bearing blocks and 4x rails plus a couple of motors apart from the Hf spindle. How would you align all that stuff in place? Each piece must be carefully machined to a tight tolerance.

Ok. I know that you will patiently make it. Obviously not your first machine 2 questions:

1. Wasn't it simpler and better using BT30 spindle and servo motor on a fixed gantry?

2. So much effort in all places but gantry will slide on round open cage bearings? No square supported ones?

Thanks Boyan. This is probably about version 5 of the Y/Z axis design with all sorts of combinations tried out, some with single Z ballscrew, some with pulley driven ER straight shafts, and other options. Eventually the desire to have the option of pulley driven spindle (for lower speed, higher torque), much larger cutting bits, and the ability to add a PDB drove me to this design.
I wanted to make the stiffest Z axis I could as it is all won or lost here. The 4 rails and 2 ballscrews is as stiff as it can be (in concept terms) so went with that. I thought a lot about the alignment of the bearings as you cannot just machine everything and bolt it together. The Y axis box structure has slip planes to allow it to be pre-loaded up to the bearings to squeeze them together. I'll let you know if this plan works !

1. I agree fixed gantry would be even stiffer, but I needed to maintain a very large cutting area for wooden panels. If it was just smaller metal parts then you are correct. So this requirement has forced a compromise. Fixed gantry would have a very large working footprint and I don't have that space.
I also looked at lots of spindles including buying off the shelf BT30 etc but they are very expensive with PBD. In the end I decided to allow fitment of direct drive WC spindle (the hole is 80mm diameter), with the option of making a spindle later.

2. As Zeeflyboy has pointed out this has already cost a lot to make, plus there is more cost to come (more Z axis rails, more aluminium plate, more steppers and drivers, pulley hardware etc.) so the ~£200 profile rails on X will have to wait. Agree it is a weak point, and will also give some loss of accuracy. For now they will have multiple blocks per side, not the usual 2, to get the best out of them. But the rail supports which they run on has already been spot drilled ready for profile rail. A simple adapter plate to the gantry will then allow the upgrade.

[Clive S]

Here is a cutaway view of the spindle showing housing, bearings, draw bar and collet system. At the top you can see the pulleys for the drive and the belleville washers which hold the tool in place by pulling up the R8 collet

I to am following this with interest. Have you decided on the size and number of the bellevilles? Also could you give more details on the drive arrangement for the PDB as I am making one for my mill but using an intensifier (small jack) to give a 7:1 advantage with 12 belleville washers. From a post on the Zone by SDM

[routercnc]

I to am following this with interest. Have you decided on the size and number of the bellevilles? Also could you give more details on the drive arrangement for the PDB as I am making one for my mill but using an intensifier (small jack) to give a 7:1 advantage with 12 belleville washers. From a post on the Zone by SDM

Hi Clive,
I'm only 50/50 on making the spindle+PDB as it is quite a risky project, and could well end up just using the WC spindle as direct drive. But in terms of spec for the PDB system I've not finalised on all the details. It does seem quite hard to find definitive numbers on things like tool pre-load.

I'm away from my home PC at the moment, where all the data is, but from memory I'd put in the same number as the Tormach PDB kit shows (~6?) since I would be using the same R8 collet style. I've read that the load applied using the Tormach PDB kit is OK for general work but could be on the low side and an aggressive cut could pull it out of the taper, so I could need more. They don't want to share the spec details (understandable) for the bellevilles so no more data than this. I can add more than this without a problem.

I've downloaded a calculator to work out loads depending on series or parallel arrangements and also give the total travel until it bottoms out. This is where the tricky bit is to ensure the loads and travel all work out.
I had allowed for 10mm total travel of the drawbar to release the tool, of which 6mm is used to remove the pre-load off the taper, and the remaining 4mm is to push it out and hopefully release the tool.

The mechanical drive system needs a lot of force/torque, so the 7:1 jack you are thinking of is a good start. I currently have a geared down stepper driving a 1605 ballscrew. I did look at a 1204 screw to get more ratio but the axial forces are very high and I didn't want go that route in the end. Again from memory it needed at least 8 Nm to turn the ballscrew. That's all I have for now.

[Clive S]

I've downloaded a calculator to work out loads depending on series or parallel arrangements and also give the total travel until it bottoms out. This is where the tricky bit is to ensure the loads and travel all work out.
I had allowed for 10mm total travel of the drawbar to release the tool, of which 6mm is used to remove the pre-load off the taper, and the remaining 4mm is to push it out and hopefully release the tool.

Thanks for the info. From what I have read (and I am no engineer) a R8 collet requires a holding force of about 2400 lbs to be safe. I am using part No. D2315162 washers from Belleville Springs 2 in P and 6 in S this give a free stack height of 28.5mm and a flat height of 24mm. 75% deflection height is 25.13 mm with a load of 12337N.

The guy on the zone recons 0.5mm will release the collet after the pre-load has gone.
The setup requires about 180Kg on the jack to give me 1260kg

I would be interested in the link for the calculator

[routercnc]

Thanks for the info. From what I have read (and I am no engineer) a R8 collet requires a holding force of about 2400 lbs to be safe. I am using part No. D2315162 washers from Belleville Springs 2 in P and 6 in S this give a free stack height of 28.5mm and a flat height of 24mm. 75% deflection height is 25.13 mm with a load of 12337N.

The guy on the zone recons 0.5mm will release the collet after the pre-load has gone.
The setup requires about 180Kg on the jack to give me 1260kg

I would be interested in the link for the calculator

Hi Clive,
I got the info together last night but no time to post out. Similar to your findings.

R8 needs 2500 lbf (11,000 N or 1,100 kgf) to hold it, maybe a bit more to be safe
Used calculator on this website:
http://www.meadinfo.org/2009/07/bell...ck-design.html

Used washers (())(( to get 15,000 N when flat.

Release distance was about 0.3mm (between holding load and flat load).

Looked at data on this site:
http://www.leespring.co.uk/browse_ca...=&pageNumber=8
part 500-125-1125

This was a quick look, I think it could be refined.

[routercnc]

Mini update -

Cut out the 3rd bracing piece:


Final one under way:


All 4 profile ops done:


Then onto the side ops. I need to machine several side features starting with 2off 6.8 pilot holes (to later take an M8 thread). Bit nervous as I'm trialling fusion 360 for the first time as it can do these pilot holes as spiral bores. Done a few aircuts and all seems OK from a cutting point of view (apart from that G28 as mentioned in another post).

Here is the fixture:


Luckily I had planned ahead (!?) and the first triangular brace I made with the poor surface finish could be used as a support. It will be sacrificial as one of the ops will cut into it.
The vertical angle is held by the precision vice, with a backstop at one end. The part just drops in and is then clamped. I was just about to hit cycle start and I was called away ! So it will have to wait . . . .

[routercnc]

I've been busy with other projects but back on it again and ready for some updates. I've had to re-read the last few posts so see where I'd left off so apologies if I repeat something.
Last time I left you in suspense over the holes in the side of the bracing pieces. Here they are being machined:


All 4 pieces machined:


Two of the holes need to be M8 tapped, and one is a clearance + counterbore. Here is the counterbore being done.


Bit nerve racking for 2 reasons. Firstly the pilot hole was machined blind on the other side, so I had to hope I'd got it all lined up so that the counterbore appeared in line with the hole. As you can see it was all OK:


Secondly, part was a bit close to the chuck. I'd measured it and knew it would clear. Clearance is clearance as they say but still glad it went OK. . . .


Then tapping the M8 holes on the bottom:


All 4 parts done:


_____________________________
Then back to the side panels. They needed holes on the sides which is always a bit more awkward. Got them dialled in:



Machined the holes and noticed that the top edge had a slight climb to it across the width with one edge 0.05 higher than the other. I guess this is reaching the limit of my current machine when cutting out a profile. So I took a very light skim to get it square again. This needs to be square to make the whole box square:


Repeated for the other one, then placed them machined edge side down onto an assumed good reference surface (lathe cross slide) to get them level with each other and then clamped them together:


Then machined the other edge, by taking a minute skim off, and then machined out the holes:


Then tapped the M8 holes:


Final dry fit of everything:


More to follow . . .