The question behind the question
“How much load will this tube take” — almost every conversation opens there. What actually decides whether the part works is how much deflection is allowed. In service a carbon tube is rarely ended by strength. Stiffness goes first, then the fittings at each end, then the way the tube is clamped.
Watch the deflection; strength rarely fails first
A boom that carries the load but sags out of position has already failed in service, even though nothing broke. Composite tube is stiff, but stiffness belongs to the section and the lay-up, not to a headline number for the material: two tubes of the same diameter can differ more from each other than either differs from steel.
Sizing starts with how much movement the machine tolerates. Some machines are limited not by deflection but by first-mode frequency — the boom shakes and the control loop shakes with it. That is still stiffness being specified, not strength.
Diameter does more than wall
The same weight of fibre, placed further from the centre, bends less — far better than piling that material into a thicker wall. Double the wall and you double the weight for a proportional gain in stiffness. Move that material to a larger diameter and stiffness rises faster than weight.
If there is room to grow the diameter, grow the diameter first; the wall comes second. If there is no room — bores, bearing seats, a fixed envelope — then wall and lay-up carry it, and the price is weight.
The lay-up decides what the tube is good at
Axial plies carry bending; ±45° plies carry torsion. Same diameter, same wall, a different lay-up is a different part. A tube that suits machine A can be wrong on machine B: the load changed, the lay-up did not. If torsion dominates, say so early — that decision belongs to design and lay-up, not to the order desk.
Once the lay-up is fixed, leave the wall alone. Changing the wall re-opens the lay-up, which means verifying the part again.
Length and support, before wall
Double the span and deflection grows faster than intuition suggests. It is the cheapest change on the drawing, and the last one anyone looks at. Fix where the tube is clamped, how long the free span is and whether the end is guided — then fix the wall. A support in the middle is often worth more than a thicker tube.
Change the span or the support and the section has to be checked again: support first, section second, always in that order.
The joint fails first
Tubes rarely break in the middle. They fail at the ends: loose, crushed, worn. Clamp, bonded sleeve, flange — each puts a local load into the interface, and each asks for a different wall.
Decide how the part is mounted, then the wall. Reverse the order and the tube comes out right in the middle and wrong at both ends. Bonded joints add a layer: with poor surface preparation the real strength is a fraction of the design value, and nothing about it shows from the outside.
Where this logic stops
A short, thick tube does not fail in bending; it fails by local crushing or under pressure. A tube under internal pressure is sized on hoop stress. A tube that needs holes is sized by what the holes do to the fibre. Each has its own rule, and a tube sized for bending is not automatically right in any of them.
Sizes and tolerances live on the product page
This is about the order of decisions, not a table. Diameter, wall, tolerance and lead time sit on the product page and move with stock — see carbon fibre round tube and the tube sizes we make. The product page gives nominal values; your duty is calculated separately. The two layers do not conflict.
What to send
Section, span, load, and the deflection you can allow. Settle those three and the rest follows.
For choosing the process, see Pultruded or roll-wrapped tube.
Need a manufacturability review? Start an RFQ