Titivating A CNC Plasma Table

[awemawson]

So what about putting the bearing back together 

Well in the end it took over two hours of frustrating eye strain  Several times I was nearly there - down to the last ten or so balls - when something slipped and I'd have to start all over again.

On each side of the slider there are two bearing sized holes parallel to the rail that at their ends point at the tangent point on the rail where bearings are trapped between the rail and the slider. If initially the slider is held in the correct alignment, balls can be pushed down these tubes from one end, and they will emerge between the body of the slider and the rail where the bearing action is happening.

However if the slider is ever so slightly off alignment the balls escape and get jammed in the wrong places 

Eventually I carefully measured the slider that was still intact, and put a small packing piece between the top of the rail and the under side of the loose slider to keep it in vertical alignment, with a tight rubber band around it to keep it together. Then it was a case of getting a few balls between the the rail and slider at the four tangent points to get the side to side alignment. Then 'just' a case of pushing bearings down the holes one by one until the holes and tangent points are full. Then quickly get the end cap on before it falls apart.

I probably repeated that process ten times  In one way the fact that the balls were magnetic was a help in picking them up with a tiny jewellers screw driver. But it was also a disadvantage, as, as you approach 'full tubes' several balls would decide to leap out and stick to the driver 

If I had to do it again, not only would I pack the vertical distance, but I'd make up something to go round the rails to fix the side to side spacing, that could be gradually withdrawn as the tubes filled and the balls emerged at the tangent points.

Anyway eventually it worked, and my headache hopefully will subside over the next few hours 

[Spurry]

You must have the patience of a saint, Andrew. You did well not to lose any of the bearings too. 
From your experiences with these, would it help if the two rails were bolted to a piece of metal to keep the two bar ends in perfect alignment?
Neat result anyway, with those dinky little rails.
Pete

[awemawson]

Thanks Pete. Yes I think that you are right about bolting the rails together temporarily..the second bearing slider transferred ok but it was a  'heart in mouth' moment.

In fact I did lose two ball bearings. The way these work with the balls trapped in a straight line I don't think that it will make any difference, certainly there is no perceivable play or roughness, and comparing with the other one they feel identical.

I will have to transfer two more bearing sliders of the bigger THK 15 size soon, as I won a pair on eBay last night fulfilling my wish to have the main Z slide on four rather than two. All those holes in the recently made plate will now come into their own. 

[awemawson]

Got a bit of time yesterday and this morning to draw up and machine the prototype Floating Z slide. NB this slide will be limited to only about 10 mm travel by stops yet to be incorporated - it just needs enough travel to operate a micro-switch that detects the surface of the metal being cut by the tip of the torch touching and displacing the slide.

Again I incorporated three positions that the bearing sliders can occupy, as just with the full travel' Z slide, I'd like to source two more sliders to give more mechanical rigidity.

Next job is to hack these plates about to fit the limit switches, mounting for the cable guide and plugs and sockets. When that's done I'll incorporate the changes in the Autocad drawings and make the final versions. I also want to take a bit of weight out of them to reduce inertia.

[mattinker]

I can see that your going to entirely rebuild the plasma table to your usual exacting standards!

Glad to see you having fun! Matthew

[RodW]

Looking good.

Personally, I think adding extra blocks is overkill. a pair of them is rated to handle 457 Newtons of force. Divide by 9.81 and you end up with them being able to handle a 46kg weight at the end of  a 1 metre long lever! Your torch will weigh in at around 0.8 to 1.0 kg at a guess. Specs are here
http://docs-asia.electrocomponents.com/webdocs/0f44/0900766b80f448f1.pdf

Once I looked at the ability of my Hiwin HGR15 slides to handle such forces, I just went with one slide for the Z axis mount and the float switch.

[awemawson]

Thanks Matthew and Rod.

Rod, I've no doubt that they'll take the loading, but the existing Z main slide and floating Z slide are both based on single  HGR15 sliders, but there is noticeable wobble at the torch from the magnified microscopic play in the bearings due to the distance from the bearing to the torch. (About 145 mm of beam length)

So by sitting both sliding platforms on (hopefully eventually) four bearing sliders set a reasonable distance apart there is a good amount of triangulation so even a small amount of play will be constrained. It just seems a better engineering solution - due to the slightly heavier construction my beam length is increased to 157 mm (*) but the whole thing is far more rigid 


(* actually that measurement is from the Z axis mounting plane so the actual beam length is about 60 mm less)

[awemawson]

I went to the Plasma Table this evening, not to cut a part, but to sort a PC connectivity issue (not showing on my local network). Now usually I'd go to it, mount a bit of plate, load up a file, cut it and hot tail it out of the room. But this evening as I wasn't rushing to cut a part I noticed just how much crud had built up on the slats, and on the grill under the slats.

Quite amazing - I'd have said I was using it lightly but it looks like it's been going flat out for yonks ! In fact 17th Jan 2017 was the first cut - so not much over a year 

Looks like I'm going to have to dedicate a bit of time to cleaning up, and probably reversing the slats - a bit scary how these things creep up on you !

[awemawson]

A bit of free time today, as a delivery expected for another project didn't happen, so I unleashed myself on cleaning up the plasma table.

Basically just remove all the slats, clean off as much of the dross from them as possible with a hand scraper, run over them with a 36 grit aloxide  sanding disk in an angle grinder, then put them back having swept up the fallings below the slats and in the catch pot.

I was expecting to have to replace some of the slats, but in practice they will be fine for another year or so. I've put them back with the affected ends reversed to maximise life. I've also tried to concentrate the bad bits in the 'far corner' where they will be least used. I suppose that they could be inverted with the dodgy face downwards in future so long as there is sufficient bearing at the three points where they rest on the supports.

Whole job took a good four hours and I'm now desperate for a shower after all that dross sanding.

[kayzed1]

I bet one of those local farmers of yours is doing the night shift...

[RodW]

Andrew, Now you've used it for 12 months, how have you found your downdraft fan? Thats next on the agenda for my super slow plasma build and I'm still struggling between using one like yours or a much more expensive centrifugal fan.

[awemawson]

Rod, it makes an enormous difference . Before I had it, the room filled up with nasty fumes, now I eject them into some waste ground
,

The fan is remarkably powerful and shifts a lot of air, a claimed 3,900 Cu M per hour 

[RodW]

Andrew, thanks. I know about the mess. One reason why I slowed down.  It was bearable cutting 2mm but its nasty with 8mm! I can get a similar fan locally with throughput of 3600 M3/hr.

The centrifugals keep the motor out of the mess but use a lot more power so its hard to get the same flow on single phase power.
I have thought of zoning the table because I've got a few 300 mm air rams someone gave me that need a job to do!

[RodW]

Rod, it makes an enormous difference . Before I had it, the room filled up with nasty fumes, now I eject them into some waste ground
,

The fan is remarkably powerful and shifts a lot of air, a claimed 3,900 Cu M per hour 

Well you motivated me and I joined the club. Hare and Forbes had a 3 day sale so I grabbed one today while it was cheap. They advertised it as 60 m3/min but when I got it home, the instructions said it was 65 m3/min which is 3900 m3/min. So now I need to get my finger out and design the downdraft hopper!

[awemawson]

If I was doing mine again, the hopper would slope on three sides and have the fourth side vertical, allowing the drop catcher door to be near to the edge of the machine and hence far more accessible.

Mine is OK but clearing it out is a kneel on the floor job (which is getting increasingly difficult!) - had it been to one side it would be far easier.

[RodW]

If I was doing mine again, the hopper would slope on three sides and have the fourth side vertical, allowing the drop catcher door to be near to the edge of the machine and hence far more accessible.

Mine is OK but clearing it out is a kneel on the floor job (which is getting increasingly difficult!) - had it been to one side it would be far easier.

Good feedback thanks Andrew. I'm tempted to try a false partition down one side so the air exits the cutting area from one side so the crud stays under the slats (in theory). If I get really ambitious, I could then zone the table using an air ram triggered based on the Y axis position to open and close shutters in the 2 zones. (Only because I have 4 beefy 300mm air rams looking for something to do so I thought I could spare one for this project.) I've played with the rams and they take about 0.6 seconds to extend or retract. So folding in your feedback, the false partition will be wedge shaped and just have a cupboard door for cleaning at the deep end. It actually sounds like a really cool design now with your idea!

 

[mattinker]

Andrew clearly explained some time ago why he didn't want a large volume of water where his tools are, but from what I understand, the water bath method removes a lot of the pollution created by plasma cutting.

Regards, Matthew.

[RodW]

Andrew clearly explained some time ago why he didn't want a large volume of water where his tools are, but from what I understand, the water bath method removes a lot of the pollution created by plasma cutting.

Regards, Matthew.

Everything I've read indicates that the down draft table design is superior to a water tray. Every commercial table I've seen (some up to 90 metres long) are down draft tables.  The problem in cold climates is that you can pump all of the warm air inside the workshop outside. Filtering the expelled air to avoid this is very expensive. Heat loss is not a problem in my climate.

Andrew's recent feedback here after cleaning has been very helpful for me and I have been busy designing a downdraft system with a central duct and an air ram driven shutter to zone the table into two zones. Hopefully the attachment uploads.

EDIT: It did! Fan at one end, hatch for cleaning at the other with a "hat" on top of the  duct to protect from dross with a "fence" each side to stop crud siding down into the duct!

[awemawson]

Nice, but I'd advise to keep it simple as possible. Those bits of dross get everywhere and if you have sliding bits they'll act as abrasives and saw itself apart 

[RodW]

Nice, but I'd advise to keep it simple as possible. Those bits of dross get everywhere and if you have sliding bits they'll act as abrasives and saw itself apart 

Thanks Andrew, It will be a bit of fun trying out the idea. I redesigned this thing for the third time after your last bit of feedback so please don't make me do it again!