[ivicencio]

Hello everyone!

I'm building a large 3D printer (around 2x2x2 meters) using steel profiles and MGN linear rails. I previously built a small CNC where I could machine surfaces precisely. However, due to the large size of this printer, I don't have access to milling equipment to flatten mounting surfaces.

Although this isn't a CNC machine, my question relates to the manufacturing as I need a bit of guidance in this matter:

1) What are common methods used by industry or DIY community to mount MGN rails without precision machining?

2) Are there proven techniques to achieve accurate alignment using basic tools?

3) Can anyone share links or resources about similar challenges?

Any advice or experiences are greatly appreciated!

Thanks!

[joeavaerage]

Hi,

1) What are common methods used by industry or DIY community to mount MGN rails without precision machining?

To my knowledge there is not a solution.

Firstly, there are machines that can do 2m x 2m.....although they are specialist and will not be cheap.....but they are around.

Many hobbyists like the idea of using a thin epoxy and pouring a leveling bed. Cured epoxy has a very low modulus, say 3GPa. If you try to stiffen the epoxy with a filler, then
it will no longer self level. For a very low rigidity machine it might be effective.

If you think scraping is a good idea then think again. It is EXTREMELY slow. If you already have a surface to 0.05mm or better 0.02mm then you can scrape to get better again, with care and patience
and a very expensive straightedge. For any more material removal than that then scraping is just not feasible.

What material are you mounting your rails on?

I made a trammel mill many years ago, and I used a selected length of 50mm x 50mm x 6mm SHS. I found that over 1.5m (the length of the trammel) that is was straight to within about 0.075mm, which
I thought was pretty damned good, and thus I used the section without straightening or machining, and it worked fine. If you chose your metal section carefully and avoid welding or cutting or drilling
which could significantly bow or warp it you might be surprised just how 'straight' a RHS or SHS section can be.

Were you to use an autocollimator you could with an angle grinder get an RHS or SHS to within 0.05mm or better. Autocollimators are just the 'bees knees' for measuring and observing straightness.
This is an example:

https://www.ebay.com/itm/23589544751...Bk9SR6KJ2uy5ZQ

I used a Hilger & Watts autocollimator at university, and for such a simple device is just so deadly accurate. The old-timers knew a thing or two!

Craig

[ivicencio]

Hey, thanks for the reply.

I'm planning on mounting the rails to carbon steel SHS tubing, probably 40x40x3. I didn't know the epoxi technique for this purpose. I think that could work fine, because I don't need a lot of rigidity with this machine.

Did you use regular carbon steel hot rolled profiles for the mill? I don't think the profiles they sell here where I live are that straight, but will have to check.

Thanks a lot for the valuable information!

Ignacio

[joeavaerage]

Hi,
40x40x3 is the sort of stuff you make furniture out of.......not CNC machines, most probably cold drawn electro-galv....its just not adequate.

Get 50 x 50 x 6. Firstly its hot rolled and likely much straighter to start with and secondly if you drill it and tap, there is enough wall thickness for the thread to actually survive rather
than stripping out. Even if you take to it with a grinder and remove even as much as 0.5mm in order to get it straight, it is still thick enough, whereas 3mm was never thick enough
to start with.

We used it for boat trailer beam axles, and would buy several 8m lengths per year. Unless you get a bad bit I would expect way better than 1mm over 2m, and the piece I used was 0.075mm over 1.5m,
about as good as I could measure at the time.

I don't think the profiles they sell here where I live are that straight,

How do you know? Have you actually measured any to see? Get yourself (beg, steal or borrow) a builders laser and go to your local steel merchant and spend an hour or so measuring some material....then
you will know.

Epoxy is 'squishy'. That is to say when the load comes on it will elastically deform, in short, not very good. If this technique is 'good enough' for your machine then I'd be surprised. Further, I have no experience
with machines that are that flimsy and cannot really offer any advice from personal experience. Good luck.

Craig

[peteeng]

Hi IVI - You need to contemplate the accuracy requirement over 2m for printing. If you are trying for 0.1 - 0.2mm layers then this is quite a mission using std steel sections. Craig suggests using a builders laser but they are only accurate to 1mm over a certain distance and will not be accurate enough for this sort of survey. To recap your Q's 1) all sorts of mucking around. Machining is the best way but if you use small light sections they will deflect under machining loads and won't come out straight anyway 2) No - but you could buy a 2m qualified straight edge and use that as a reference. But be prepared for a very large cost for such a tool 3) Eventually you do the best you can with what you have and seems your an optimist so some things will achieve your expectations some will fail. I'd use large aluminium sections they will be more rigid and straighter then steel and deflect less under gravity. I would also bolt the entire machine together no welding as welding will distort your parts irreparably. Use heavy AL channel not hollows so bolt ups can be very true and firm... The MGN series is a very lightweight series and it will conform to its foundation very easily so the foundation has to be true, plus cheap cars will be clearance or zero clearance (ie no preload) and this will create hysteresis in your motion so you need medium or heavy preload bearings in this application... I imagine you are going to use belts for motion and that is another can of worms at 2m long. Commercial printers this large (like any large machine) are very expensive for good reasons... But Good luck, keep at it. Peter

By the way you say you don't need rigidity for this machine, think again, gravity, hysteresis, inertial loads and movement such as vibration all require rigidity for a successful machine. A machine can never be too rigid but very often they are not rigid enough.

[joeavaerage]

Hi,
firstly, you'd only use a laser to an estimate as to whether a piece of steel (or other material) is straight or reject.

Aluminum of the description Peter give is expensive, and quite frankly unnecessary. SHS/RHS steel is very VERY much more readily available and much MUCH cheaper and unless you get
a bad piece it will be damned good, probably better than your expectation. It certainly exceeded my expectation when I built my trammel mill.

Pete describes the surface of a SHS/RHS to be bowed or cupped. That is not my experience, at least within 0.05mm anyway. Even if it is bowed or cupped a few hours light work with a handheld angle grinder,
some ink, a straightedge and a bottle or two (for patience) and you'll get a bloody good result, I'd hope to get 0.05mm or better over 2m....but your mileage may vary.

Where I would buy a piece of aluminum would be for a straightedge. As big and heavy section as you can find, and as long as your machine. If you can get a steel SHS to as good as the aluminum
you'll be doing fine. The next step up is an Autocollimator, with such a device you could get to 0.01mm over 2m with patience.

If you were in New Zealand, or even in Australia I would go you half shares in an Autocollimator. Its not like you use them often, so if you had one between two, three or more guys that would be fine.
They really are incredibly simple and so bang on accurate....once you've used one you'll swear by them thereafter.

Craig