Roll forming question about step-up

[Ryoko]

TL;DR Summary

A paper I read says that each successive roll station in a roll forming line runs at a slightly higher speed than the previous station. But pictures of a typical roll former don't seem to reflect this. What gives?

A paper I read says that each successive roll station in a roll forming line runs at a slightly higher speed than the previous station to prevent buckling. In order for each successive stage to run slightly faster, either the drive speed would need to increase or the rollers would need to be slightly larger to increase the surface speed at each station. The drive speed appears to be constant due to the gearing mechanism being the same for each station. So the rollers must be increasing diameter. But if that's the case, the axles must be shifted vertically to allow for this difference in roller diameters. However, it looks like all the axles (lower rollers) are lined up perfectly horizontally with no incline visible. So how is the step-up being accomplished or is there no step-up after all?

 

[jrmichler]

TL;DR Summary: A paper I read says that each successive roll station in a roll forming line runs at a slightly higher speed than the previous station. But pictures of a typical roll former don't seem to reflect this. What gives?

A paper I read says that each successive roll station in a roll forming line runs at a slightly higher speed than the previous station to prevent buckling.

How much faster? Is it 0.1%, 1%, or 10%? And how much forming was being done at each station?

Does the paper say anything about the effect of scrubbing when the material being formed wraps around the forming rolls? One roll will tend to hold the material back, the opposite roll will tend to pull the material ahead.

 

[Ryoko]

Actually, the paper wasn't entirely clear. It suggested that the first 3 stations increase the diameter of the driven roller by 0.050" each for working with thinner materials. Latter stations would use smaller increments under the assumption enough form would be present to resist column buckling. The problem is the base diameter of the roller isn't given. Scrubbing was mentioned, but not in the context of pulling the material along. The example discussed used 12 to 20 degrees as the delta bend angle per station.

I can see how scrubbing would accomplish the same effect as increasing the roller diameter.

 

[Baluncore]

Note that iron and aluminium behave differently. Aluminium can be bent once, rapidly through a large angle, before it cools and work hardens.

Roll forming a 'U' cannot be done in a straight set of rolls, as the edges must be longer than the bend. The edges travel in two helical paths, with opposite sense, each through 90°. The centreline of the 'U' must therefore curve upwards, (step up), along a similar circumference to compensate, allowing the edges to take a shorter path without stretch.

If only forming 100 ft, I would use a press brake. The cost of setting up rollers, then testing, would be excessive.

 

[Ryoko]

Note that iron and aluminium behave differently. Aluminium can be bent once, rapidly through a large angle, before it cools and work hardens.

Roll forming a 'U' cannot be done in a straight set of rolls, as the edges must be longer than the bend. The edges travel in two helical paths, with opposite sense, each through 90°. The centreline of the 'U' must therefore curve upwards, (step up), along a similar circumference to compensate, allowing the edges to take a shorter path without stretch.

If only forming 100 ft, I would use a press brake. The cost of setting up rollers, then testing, would be excessive.

The problem is each needed piece is 80" long. That's a pretty big press brake. I made a test brake that could do 12" test strips and it mostly worked. The problem it had was that the material wanted to shift producing uneven flange height from one end to the other. But I agree, a press brake would be far easier to make if the shifting problem could be resolved.

I've had some success with experimental roll forming. The biggest issue was the strips not wanting to go straight. The new design has traps on the rollers to keep the strip centered.

 

[Baluncore]

The problem it had was that the material wanted to shift producing uneven flange height from one end to the other.

Then trim the final dimensions to the required profile after the bend.

 

[Lnewqban]

The problem is each needed piece is 80" long. That's a pretty big press brake. I made a test brake that could do 12" test strips and it mostly worked. The problem it had was that the material wanted to shift producing uneven flange height from one end to the other. But I agree, a press brake would be far easier to make if the shifting problem could be resolved.

Could you use cylindrical rollers that act like the fixed and pivoting dies of a press brake?
I imagine a hybrid between a typical roller and a press brake.

Each stage would have:
1) Two central cylindrical driving rollers pressing the strip with enough force to induce friction for pulling-pushing and keeping the strip centered.
2) Two lateral cylindrical driven rollers each mounted on a hinged support which angle can be progressively adjusted to simultaneously bend each side more and more with each pass.

 

[Ryoko]

I tried something similar to that. The device held the strip in place while interchangeable rollers "massaged" it into shape. It almost worked. But it was prone to twisting the work piece..