Cambering beams...How hard can it be?

The theory is very simple. You induce enough controlled force to deflect the beam through its elastic phase to plastic stage, while keeping within the yield to ultimate margin. Resulting in a mid-ordinate camber dimension as per specified value.

Assuming that the beams to be cambered are all identical in dimension, straightness (in both axis) residual rolling stresses and web to flange squareness...its a basic calculation to provide the amount of deflection and then bingo, away you go.

But of course, in the real world, that is not the case. No two beams are identical or will perform the same under cambering. No matter what you have been told by all parties involved. So lets look at the variables and their potential issues.

Section Dimension...Rolling tolerances alone can make significant variations in the supplied beams. For example we have cambered 10,000's of 457 x 191 x 67kg/m UB's at 16m, often in complete rollings supplied from mill, and have encountered noticeable variation. BSEN 10034 allows the following tolerances...+5/-3mm on section height...+6/-5mm on flange width...+/-1mm on web thickness...+2.5mm/-1.5mm on flange thickness...+/-5mm on web offcentre...+/-4% on section mass. So that's the first variable to contend with.

Straightness...0.1% of length...16mm on a 16m beam. Not always a constant sweep or even in the same area in both axis.

Rolling stresses...this is the amount of force applied to the web/flange root area during rolling and subsequent roller straightening at the mill. Evident by the "crows feet" scale disruption on the web, usually more pronounced on the straightened side of the beam. How far it extends across the web give a fair indication of how much force has been used at the mill. On a 457 x 191 x 67 UB there is a reduced margin between yield and ultimate every time the beam is roller straightened. Most reputable mills will just run the beam once through an RSM and then cold straighten on a gag press in order to maintain the yield/ultimate margin. Those that don't will face the possible failure of the beam due to root cracking, either during subsequent processing, fabrication or galvanising. This happens more than it should.

Web to Flange squareness...2% of flange width up to a maximum of 6.5mm...causes more variable dimension and apllied force calculations as well as producing a likely kink (or set) in the flange. Not always immediately visible (unless you're actively looking for it) but csan become more pronounced and cause severe deflection in the weak axis after galvanising. Some camber these beams with one moment of camber rather than several along the beam which also increases the likely weak axis deflection. In order to avoid the bottom flange looking like a 50p/thrupenny bit it is best practise to ensure maximum rather than minimum moments of camber.

So next time you think that the charges for cambering are unjustified please reconsider the possible remedial/failure costs you risk by pursuing the cheaper, less experienced option.