New Build - For Your Amusement - MK-2 build

[Jonathan]

Looks great, is there any issue with using sbr / tbr rails on their side - particularly on X and Y axis - cantilevering off them rather than bearing straight down?

Yes - the load rating of the bearing blocks varies depending on the direction of the applied force. In other words if the force is in the direction attempting to pull the bearing off the rail clearly the deflection will be greater. For that reason it's best to mount them so the rail supports are facing each other. This evens out the deflection, so you effectively have medium deflection all round instead of large deflection when cutting one way and small deflection the other way.

[Karl]

For that reason it's best to mount them so the rail supports are facing each other. This evens out the deflection, so you effectively have medium deflection all round instead of large deflection when cutting one way and small deflection the other way.

I don't understand that - not saying you're wrong just that I don't understand. That said, we of course accept that bearings are not operating at their best when they're being pulled off the rail.

Mounting the rails and thereby bearings (on which the z-assembly moves along the y-gantry) top and bottom - with the bottom rail inverted - means (by my understanding) that the load is primarily carried by the top 2 (typically) bearings; the bottom 2 bearings preventing rotation of the z-axis assembly. With the rails/bearings mounted on their side the z-axis load is (again by my way of thinking) carried by all 4 bearings albeit with the force on the side of the bearing. So load-wise is 4 bearings carrying the weight on their side better or worse than 2 carrying the weight conventionally?

With regard to deflection (or moments), when the y-axis gantry moves in the x-direction, I rather assumed that where one pair of z-assembly bearings were effectively being pulled off the rail (tension) the other 2 would be under compression and vice-versa with direction of x+/x- travel.

This is interesting because I was planning to use the same design for my next larger (8' x 4') machine. I found this design (which I borrowed from another forum member's post) to create a low-profile y-gantry with easy mounting of rails and y-axis motor/ballscrew.

Now I'll have to get the pencil, paper and abacus out to calculate some moments; unless someone has an 'app' for that.

PS. Obviously, I accept the rails-on-their-side design must be wrong since most implementations don't use it - I still think it's quite a neat design though.

Thanks

Karl

[routercnc]

Hi Karl,

I think the design you have posted with rails their side would still work OK, especially for hobby use. But to elaborate a bit further . . .

Considering the forces - there are 2 forces acting on the bearings, one from the mass of the bits they are carrying (router, motors etc.), and one from the cutting forces. I've ignored the forces due to accelerations because some early calcs showed these to be small, especially on a hobby machine.

The mass of the bits is huge compared to the cutting forces (especially wood) so I would orient the bearings to take this larger load. In this case I would favour Jonathan's arrangement, top and bottom.

In comparing 4 bearings on their side with 2 bearings on top (and 2 guide bearings on bottom), I would agree that the top/bottom arrangement puts more load on the top 2 bearings vs spreading it over the 4 side bearings, but I wouldn't worry about exceeding the load rating because you are probably still miles away. Instead consider what happens to that load when it is applied to the stiffness of the bearing in that direction. Arranged on their side the large mass of the Z axis is trying to open up the bearings, thereby slightly reducing their preload or grip on the rail. This would be like having very slightly smaller ball bearings inside the unit which would slightly reduce the accuracy of the carraige on the rail. Arranged on top and bottom you are mainly loading the bearing in the stiffest direction, so would expect that to give the best overall results even though it is carrying more load per bearing.

[Karl]

Yep. I can relate to the argument that having the bearings on their side would tend for the load to open them up. I never considered that but obvious when one thinks about it. Oh well, we won't be using the bearings-on-their-side design for the next machine then.

Thanks

Karl

[Karl]

Ref discussion in thread above. Is this bearings on their side I see in this commercial machine?

[Jonathan]

Ref discussion in thread above. Is this bearings on their side I see in this commercial machine?

Yes it is - alas badly designed commercial machines are nothing to be surprised about.

Check out the one in the background...potentially much better.

[routercnc]

Hi Karl,

For any CNC machine design you just need to ask yourself one question: If I apply the required load to the cutting tool will it deflect less than the accuracy I'm after, or vibrate less than the surface finish I'm after?

If the machine meets those requirements by a reasonable margin then its a good optimal machine. Problem is that although the load requirements are fairly well known (5N wood up to 75N steel as a guide) unless you have access to an FEA programme, and can mesh up the CAD and carry out the analysis, you won't really know how much the tool will deflect under this load. Therefore the advice given in this thread is the arrangement which we believe gives the best performance and lowest risk for those type of bearings. Whilst it might be the case that the bearings on their side arrangement would also give sufficient performance for what you want to do (and I have a feeling it would be), for the same cost the top / bottom arrangement should be superior so it makes sense to use that one.

[Philly]

Sorry bring up an old thread, but I like the look of this machine, would this design be possible of light aluminium work?

I want to build one similar with the rails rotated.