Expansion mounts for linear rails

[Argentum]

I’ve been reading about the issue of rail deformation as a result of differential expansion between steel and aluminium.

I’m trying to work out why it isn’t possible to make a mounting bar that goes inside the extrusion and the tail bolts to, which can be sized to prevent the rail moving up and down or in and out, but which can allow lengthwise expansion.

If it were tight in the middle so the rail were fully clamped there, but looser towards the ends it should be stable. Think of the mounting piece having a slight dip of a couple of thou in the middle.

I’m assuming this isn’t possible, because it’s obvious and nobody’s doing it but I’m not sure why.

Are the tolerances of aluminium extrusions just not tight enough for this to work without binding up?

Thanks.

 

[WhoppingPochard]

Great question! In principle of course if you could fully constrain expansion to one dimension, that would be a perfect solution -- like an expansion joint in a bridge.

In practice, it's just difficult to constrain it well enough in that one dimension without relaxing constraint in the other dimensions, and backers work well enough!

 

[Argentum]

Thank you.

 

[ChrisA]

I’ve been reading about the issue of rail deformation as a result of differential expansion between steel and aluminium.

I’m trying to work out why it isn’t possible to make a mounting bar that goes inside the extrusion and the tail bolts to, which can be sized to prevent the rail moving up and down or in and out, but which can allow lengthwise expansion.

If it were tight in the middle so the rail were fully clamped there, but looser towards the ends it should be stable. Think of the mounting piece having a slight dip of a couple of thou in the middle.

I’m assuming this isn’t possible, because it’s obvious and nobody’s doing it but I’m not sure why.

Are the tolerances of aluminium extrusions just not tight enough for this to work without binding up?

Thanks.

I think you could develop some hard PFTE or PEEK backing plates that would be a custom fit from the 2020 extrusion's v-slot to the 9 or 12-mm rails. These would add less mass than steel backing plates. They would allow the steel and aluminum parts to slide pat each other. The backing plates use a brute force approach that apparently works and is very cheap, well at least steel is cheap. Ti works as well and adds less mass. But these would be expensive custom parts, expensive because of the very low volume and the high tolerances needed.