PDA

View Full Version : New Member and New Build design check



Procato
29-04-2015, 06:22 PM
Hi all

Been lurking for years and finally, after many false starts and many redesigns later, have begun a build only to redesign it again. So in an effort to finalise the design and gain some knowledge and tips from the experienced- here's is what I have so far. As you can see, it is an almalgamation of builds and advice gathered on(Stolen from) this website.

I plan to have mill 3d hardwood, softwood and foam sculptures and when the bed is raised I would like it to mill aluminium. it's 1400mmx1000mmx900.

The frame has been mig welded together using 80x80x3.5 and some 50x50x2.5

I had bought some supported round rails a couple of years ago when I first started thinking about building this but unfortunately work got in the way of that build. I had made the decision to use these rails on this build but then, having got this far with the welding, I thought if I'm going to do this, I ought to do it properly. So out with the round rails and in with linear guides and hopefully a substantially 'beefed up' gantry and Z-azis.

So any advice will be very welcome.

I'd appreciate it if any changes to the frame were additive rather than subtractive - I can't face starting from scratch again!
http://www.mycncuk.com/attachment.php?attachmentid=15217&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15218&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15219&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15220&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15221&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15222&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15223&stc=1


As you see I haven't modelled in any fixings yet and the strange shape of the outside plate on the y-axis is to accommodate another piece of box section if it is needed in the future.

toomast
02-05-2015, 06:13 AM
Hello sir!

Gantry certainly looks beefy! What weight do you estimate for it?
It looks like it could handle steel...and little bit OTT for mostly wood? :)
Frame is simple and sturdy.
How will you solve the bed design and work-piece fixing?
I notice 4th axis?

Boyan Silyavski
02-05-2015, 11:00 AM
I see a new contestant for the sturdiest build here :triumphant:

The only weak spot that i see at first glance are the corner table connectors. I would additionally brace then in the other direction / L. And this is additive :-)http://www.mycncuk.com/attachment.php?attachmentid=15238&stc=1

Procato
02-05-2015, 11:36 AM
Hello toomast

Many thanks for your reply.

Now you've got me worried, after some quick guestimation in CAD i think it's going to weight in at about 95+kg - is this too heavy?

Re; bed design -The lack of fixings and fixing holes is misleading - I hope to utilise the same method as Jonathon's, after the rest of the machine is built, I plan to attach a large 20mm-ish aluminium plate to the steel frame of the bed then mill it flat, then mill the slots and finally drill and tap it so that I can screw down aluminium strips to form the T-slots. I realise the t-slots on the outer edges won't be possible with this method so I will probably leave them unmachined.

I think I will be mainly routing highly detailed wooden figurines up to lifesize (in 2 parts) which will be held in the rotary axis on one end and free rotating tailstock at the other. both of these will be bolted down to the bed via the T-slots.

The 4th axis is a doughty drive which I bought a few years ago but I've also got a bargain harmonic drive off ebay the other week which is bigger and stronger and will turn that into the 4th axis. I hope then to mount the dougthy Drive onto the Z-axis to form a 5th axis when needed.

Procato
02-05-2015, 12:06 PM
Hi Silyavski

Ha yes as you can probably see I'm hoping to use many of your ideas/designs in my build.

I'm also hoping to be able to mill aluminium moulds for resin infusion of carbon fibre hence the attempt at very rigid design.

Thanks for pointing out this weakness- do you mean something like this?
http://www.mycncuk.com/attachment.php?attachmentid=15239&stc=1

Boyan Silyavski
02-05-2015, 12:39 PM
Yes, exactly that's what i meant.

It would be interesting to see the z axis. Where is the Z ball screw? I can not see it

Procato
02-05-2015, 01:35 PM
http://www.mycncuk.com/attachment.php?attachmentid=15240&stc=1http://www.mycncuk.com/attachment.php?attachmentid=15241&stc=1
The Z-axis has been causing me some concern. Do you think there could be a racking issue?
I had thought about putting the ballscrew in the centre of Z-axis pillar with a motor on top but that means either putting a slot nearly all the way down the z-axis pillar which I thought may weaken it too much?
Or making the z-axis twice as long, keep the same amount of travel and put the slot in the top half - this makes a ridiculously long axis.
Or I could add another ballscrew to the other side.
What are your thoughts?

routercnc
02-05-2015, 10:02 PM
Hi Procato,

Interesting design. I drew something a bit like that a while back but didn't like the loss of X travel and obstructed access to the spindle for bit change. But I was working with upgrading an existing machine so if these factors are already considering in the geometry of your clean sheet design then it should be OK.

I should point out that the Z axis box 'front plate', i.e. one closest to the ballscrew needs to be lapped the other way to the side plate. You are trapping precision height linear rails against each other and they will want to be a certain distance from each other. Unless you can machine the Z axis side plates to very fine tolerance, then lapping them the other way is better as you can use slightly oversized bolt holes coming in from the sides and lightly clamp the assembly before nipping them up.

I can't see in any of the sketches but assume you have a third linear profile rail at the front for support? Should work very well for stiffness, but you also need to be careful on the order of assembly of the Z and Y. The front box section probably needs to be attached last once everything behind it is set. If this is your first build you need to think about the imperfections of machining and building and make sure things can be built in the right order with pre-loads or free running checks made as you progress. Try to build it in your mind.

As for Z axis racking I wouldn't be too concerned about it based on the design and what you want to do with it.

Looks like you've got twin Y axis drive with a twin pulley on the ballscrew nearest the stepper motor. So it looks like you want to link a short belt to the first ballscrew, and then have another belt between each ballscrew. That's no problem but make sure you can tension them. The stepper pulley belt is easy - just add slots which it looks like you have done. But the belt between the ballscrews will need an idler pulley and it looks like it's a tight fit between the side plates. Have you provisioned for one? If so make sure it can take up the required slack as belts come in fixed sizes (although I believe you can have custom ones made).

Sven
03-05-2015, 08:18 AM
I've been told that the linear bearings and rails are stiffer than anything you could bolt them too.

If that is the case then you'd be better off combining all beams in the gantry in to one structure.
I think you now support the Z on two sides for stiffness, but you actually lose torsional stiffness this way.

Also, it will be extremely hard to line up all parts to run smoothly, with multiple carriages on multiple carriers.
All in all, though it looks very nice, I feel it is overly complex but will under perform.

Procato
03-05-2015, 08:15 PM
routercnc - many thanks for your suggestions. It took a while for my brain to get around your notes about the rails on the z-axis but I now I realise what you say and it makes total sense. My original plan was to machine the z box so that when all is put together there is approx 1mm gap between the 3rd rail guides of the z-column and the 'backplate'. I would hold the guides roughly to the 'backplate' with loose screws then inject aluminium epoxy putty into the 1mm gap. I would tighten the screw once the putty had set. This method would I hope would mean I wouldn't need the high tolerance machining that would be otherwise required. Not sure what to do yet so will mull it over.

The order of assembly is also great advice, again some thinking needs to be done.

Re the tensioner - You've caught me being lazy, I will add it in the cad drawing soon but I just wanted feedback before I got into the details. There's quite a big gap between the end plates 80mm.. I plan to have something simple, probably a slot running on the inside end wall of the gantry with a couple of bearings on a screw to tension the belt.



Sven; Many thanks for your input, I have gone to and fro with this thought for a while and I guess what I really need to do is learn the simulation part of the CAD program to get a definitive answer.
I get the idea that unless you build the extra rail perfectly you just end up introducing error and not taking it away but there is this constant niggling at the back of my brain that says when that long Z-column is at full extension, the leverage on two rails will be very high especially in the centre of travel. I can't help thinking 3 rails has to be stronger but that is just a gut feeling and nothing more. It would be very interesting to other peoples thought as well.

Boyan Silyavski
03-05-2015, 11:28 PM
Hi,
what are the exact planned distances in mm for the different color lines, see picture below.

Red line bearing spacing distance seems too small for that enormously long Z, doesn't matter that the Z is 3 rail supported. Best would be if is same distance like violet line.
Meaning 250mm is not right here. ~350mm is more sound. Have in mind that mostly and only Lower bearing blocks take the force, so should be up to the task of coping with 900mm extended axis acting as a lever. Proving that with a calculation would be useful. The upper bearing block acts only as lever arm so basically is doing nothing apart from levering, hence the distance here is extremely important.


Why so long Z? Longest reasonably Z IMHO is 500mm. Unless you plan to shape only egg statues :joker:. The z assembly will intefere, so i dont see what the benefit will be of so long axis.


http://www.mycncuk.com/attachment.php?attachmentid=15252&stc=1

Procato
04-05-2015, 10:23 AM
Blue 318mm
red 240mm
violet 260mm

The travel on the Z-axis is 350mm which is I think is about correct to mill/sculpt human torso(with separate arms) which is the work envelope that I'm trying to achieve. So I might have exaggerated its length in an earlier post!
http://www.mycncuk.com/attachment.php?attachmentid=15255&stc=1
So with the 350mm z-axis travel do you still think that this machine would benefit if I increased the distance between the Y-axis rails to 350mm?
Also do you think that the design benefits from the 3rd rails on both the y-axis and z-axis?

Boyan Silyavski
04-05-2015, 11:01 AM
Difficult to say.

On my current build both red and violet distances are around 350mm and even without 3rd gantry profile and 3rd rail on z all looks like it will reap aluminum like butter , i would say steel will not be a problem at all with small cutters.

So in your case for me it seems logical to extend the red distance a bit. Though the third rail a kind of compensates for that.

I second the previous advice given to you. Make sure piece by piece its possible to fix all together and align it/ tram it/ . Its a complex assembly so dry test all in software.