# Thread: X axis design - a different approach

1. ## X axis design - a different approach

Hi All

I've been pondering over this design innovation for a long time and welcome feedback, response.

We all know that X axis has two parallel linear guides set to the same height along their full length. Lets call the original X and the other X'.

Traditionally both X and X' are positioned and then locked down.

My thought is that only X needs to be locked down and X' allowed to float so that the gantry actually determines the position of X' in terms of its distance from X. Providing X and X' are level along their full length, it isn't actually important that X' is locked down.

The advantage is that for setup only X needs to be rigidly connected to the frame. All X' needs is to be raised or lowered to ensure it is level with X and can slide in and out on a flat plate or skate bearings.

It also means that only X has to be absolutely straight (from both a plan and sideways point of view) . X' only need to be straight from a sideways point of view (i.e. level along its full length)

Another advantage is that it becomes a stressless solution.

True, X' needs to be stopped from lifting but potentially the gantry weight might be adequate.

It does place the full sideways load on X but that is an issue of design strength, rigidity. I would say that stengthening X is a minor issue compared with the long term effect of introduced stresses incurred by attempting to make both X and X' totally and absolutely parallel.

At one point I thought that another advantage would be of removing the effects of thermal expansion (in other words the gantry expanding / contracting in length) but I am now of the belief that providing the same material is used throughout the frame and the gantry then it becomes a non issue. If for example you have used wood for the frame and aluminium for the gantry then this could then be another advatage of this approach.

Has anyone else thought along these lines?

Andy

2. How do you lock it down for height, but let it slide side to side?

I just see added complexity with no benefit.

3. .

What do you want your machine to do/cut?

I doubt you could do milling/routing, you might be able to stick a pen in "Z-axis" for maybe drawing/pattern designs.

If you tried to use a router with a router bit, I'm sure it would rack "Y-axis" (side load).

Plus If your (X') wasn't supported to the rail/table on both sides of the rail, It would lift (X') off the rail/table, when the "Z-axis" (Router bit) was forced into the part you would be cutting.

Again, I think about all you could do is drawing/pattern designs, with a setup like your suggesting, maybe a spring loaded pen in "Z-axis"

.

4. Like Switcher says, the leverage of the cutter when it is positioned out near the X' rail is going to be extreme. This means you would need a very long seperation of dual bushing supports on the X rail itself, to prevent cocking and jamming of that axis. I'm suggesting the bearing seperation would have to be roughly as wide as the travel in the Y axis.

5. Andy,
There are advantages to building a machine the way you propose. The one rail that provides both horizontal and vertical support can be a heavier rail, and the other one that provides only vertical support can be lighter. It would greatly decrease the alignment problems. This thread has some excellent insight from an experienced machine designer about this. http://www.cnczone.com/forums/showthread.php?t=6178 (One lead screw off-centered..Will it work?)
He talks about commercial machines designed this way. He suggested using a vertical flat plate to mount the offside bearing. The vertical plate would provide good vertical support, but would be able to twist to accomodate horizontal misallignment of the rails.

Jeff

6. Sounds 'somewhat' similar to how inkjet printers are setup these days.

7. And scanners = there was a member here on the 'Zone a few days/weeks back that was going to do a similar approach - his idea was to use as a painting/drawing xy(slight z) system on canvases. The X fixed per "normal methods" and X' was to be supported by a roller. Seems like it ought to work well for low loads on Z - like a vinyl cutter though!
Best of Luck - Jim

8. I wouldn't be so quick to relegate this approach to loadless applications like printers and plotters. I have a 25mm IKO rail, and it has a dynamic load rating of about 2000 lbs. Even with bearings spaced just 1/4 of the length of the Y axis, this would equate to a load of 500 lbs at the far end of the Y axis.

9. Hi All

A few answers first:

"How do you lock it down for height, but let it slide side to side?"

I have secured a stainless tube onto a 40*40mm*4mm beam (see X rail with tube att.). The beam is allowed to slide in one plane only that is close to or further away from the other fixed rail (see floating x' guides att.).

To explain the four blocks, the bottom two are bolted to the upright and have one skate bearing on each allowing the top two blocks to roll on them. The top two blocks are attached to the horizontal with one skate bearing on each ensuring the horizontal beam can only move in the one plane. (It is just possible to see the outline of a skate bearing sticking out of the bottom right hand side of the top left block)

If you tried to use a router with a router bit, I'm sure it would rack "Y-axis" (side load).

Both the X beam and the X' beam have timing belts on them secured at both ends. The pulley and motor are on the one side of the gantry with a drive rod thrown across to the other side thereby ensuring no racking. Because the floating X' beam is only allowed to float in the one plane, there is full and equal force on both sides

Plus If your (X') wasn't supported to the rail/table on both sides of the rail, It would lift (X') off the rail/table, when the "Z-axis" (Router bit) was forced into the part you would be cutting.

Agreed and well spotted. X' would have to be stopped from lifting but easy to achieve with extra skate bearings.

I believe this approach means that alignment is a lot easier which makes for a machine that can be stored and reassambled very quickly. One of my aims was to create a machine that could be stored away and brought out easily or even put in a trailer and be re-located.

I believe there is no compromise with respect to load capability. It is also an unstressed machine. I'm not sure if vibration would be better or worse for it.

Andy

10. It might be a challenge to copy for a DIY machine but a fully cantilevered Y axis is available on some woodworking machines. The X axis ways are at the back with a large vertical spacing and the carriage that rides on these and carries the Y axis has a length which is more than half the travel of the Y axis. X/Y travel is about 9 feet by 3 feet and precision is +/-0.005".