Gantry table - One rail or Two?

The six month anniversary since I commenced my scratch build plasma table is rapidly aproaching, You guys have been super helpful so far to get my Scratch built plasma to here


And you helped me build a new PC with Running Mint 17.3 and LinuxCNC to drive my Mesa 7i76e card


And I've built the electronics so I can get things to actually move all on my own (after countless hours studying this forum).



So now its time to start the next stage, building the plasma table and adding the gantry X axis.
You see, I used two Hiwin HGR15 linear guide rails on the Y axis


And now I think this was overkill just adding weight and cost to the project. I need to design the table and intend to use Hiwin HGR20 linear rails and carriages for each side of the plasma table to support the gantry. So the question is:

Do I use two rails each side or just one?

Now I know at this point some reader's eyes will glaze over...
My engineering is incredibly rusty not having played with it since the early 1980's but I remember enough to know I need to ensure the table can handle the twisting moments applied to the linear carriages. Hiwin have this nice little diagram of them



For the HGR20 carriages, they say it can handle the following forces.
Mr = 0.27 kN/m, Mp = 0.20 kN.m and My = 0.20 kN/m

Converting back to Newtons, these become 270, 200 and 200 N.

There are no cutting forces on a plasma table that (hopefully) just hovers around just above the material surface so we must just need to worry about the twisting moments that the machine applies when moving.
Using round figures, I know that the Z axis assembly which moves along the Y axis weighs 10kg and LCNC is able of accelerating at 1.0 m/s/s. My guess of the total gantry weight is around 30 kg and I am assuming it will be able to accelerate at the same rate.

Lets assume the gantry is 0.5 m above the X axis rails and that the linear rails will be attached to the table side, not the table top (eg. oriented as per the My diagram above) and I will use 2 carriages per linear rail.

So the z axis gantry head has a force of 10 newtons applied to it (10kg x 1 m/s/s)
And therefore Mr = 10N x 0.5m = 5 Newtons which is well within the 270Nm allowed In a single rail setup, I think this force would be spread across 4 carriages (2 on each side) so each carriage is only subjected to about 1.25 Nm on the MR moment.

So does that sound right?

There appears to be nothing that will exert any Mp forces on the carriages. That implies a twisting moment on the gantry when the carriages are oriented on the side of the table.

So the My moment is going to be a bit bigger as its influenced by the whole weight of the gantry. Assuming the same acceleration applies, we've got 30kg x 1 m/s/s = 30 newtons so therefore Mp = 30 x 0.5 = 15 Newtons which spread over 4 carriages becomes 15/4 = 3.75 Nm per carriage.

So based on these calculations there appears to be no justification for installing 2 rails per side other than to line the suppliers pockets.

Bearing in mind I'm an absolute noob with everything I've done so far and have no experience with this stuff in the real world, the question I have before ordering anything is do I need to spend another $500-$1000 to add a second rail to achieve my goal of commercial quality performance?

Of course my maths may be way out as I've proven by being out by a factor of about 3 when designing the Y axis.

Thank for reading this far (if you made it) and I'd be interested to hear some feedback on this key design decision.

Tommy, thanks for your advice once again.

I am planning on having the rails side mounted on laser cut plates so all of the mounting holes for the rails and gantry will be perfectly aligned. You reminded me about the vertical adjustment which I had thought of earlier and was thinking of having the rail mounting plates resting on bolts at each end that go through nuts or tapped holes welded to the table legs so they can be adjusted before all mounting bolts are tightened up.

I will have a Mesa THCAD as THC.

I will also laser cut the gantry end plates and probably have the edges folded back for extra stiffness.

I was never keen on the idea of a water table so it will be a downdraft table. I am still not sure on what filtering is required. I have added a mains power point to my control box for the fan and have thought of allowing it to be turned on and off by the Mesa card as the machine runs to keep the noise down when not actually required.

I'll look for some 25mm rails.

Well, I found a couple of pieces of HGR25 for less than what I'd been quoted for HGR20. I put my order in but I don't expect to see it until 2017.
I have already ordered some helical gear rack and a couple of nice 5:1 timing belt pinion drive modules for NEMA 34 motors. I'll need to watch the power demands as I was anticipating using the NEMA23 drives I already have. But then again it won't be hard to adapt NEMA34 mounts to suit NEMA23 motors.

I played with my gantry limit switches today and Ive got a little bit of work to do to mount them but it looks like I'll have 1230mm of gantry travel which will easilly accommodate a standard 1200mm wide sheet. The Z axis is a bit less than I really wanted but I'll have 87mm of travel. That will be enough but next time, I will use a 200mm ball screw.