External dimensions are on every drawing. Length, width, height — measured, agreed, quoted, delivered. And yet a surprising number of reusable box programmes hit trouble in the first month for reasons that have nothing to do with the external envelope.
The problems come from three measurements that rarely make it onto the spec sheet, and all three only reveal themselves once boxes are moving along a live line.
1. Internal corner radius versus the part you are putting in
A die-cut or welded box has a corner radius. Small parts, castings, and anything with a square housing will sit proud of a tight radius by a few millimetres — enough that the part rocks, and enough that a robot gripper placing to a fixed Z-height will collide.
The fix is either to specify a radius that suits the largest square-cornered part in the family, or to accept that the part sits slightly high and design the dunnage insert around it. What you cannot do is assume the internal envelope equals the external one minus two wall thicknesses. Do the arithmetic on the actual radius and check the diagonal.
2. Stack height tolerance, not stack height
Nobody orders a box and says “give me ±2 mm on nested stack height”. Everybody notices when it is wrong.
If boxes are stacked ten high on a pallet and each one is 1.5 mm over nominal, the top of the stack is 15 mm above where the pallet wrapper, the rack beam or the container door expects it to be. Multiply that across a standard stack and you are either re-engineering your pallet pattern or leaving expensive air at the top of every container.
Ask for a stack height tolerance and hold the supplier to it over a production batch, not just on samples. Tooling wears and weld fixtures drift, and the tenth box off the line is not always the same height as the thousandth.
3. Handle cut-out depth and what it does to the wall
A hand hole is a hole in the load path. Put it too deep, too close to the top rim, or through a wall that is already carrying the stack load, and the box will crack at the cut-out under stacked weight. It usually shows up as a split running down from one corner of the handle.
Generous radii at the cut-out corners help a lot, as does keeping the opening away from the rim by at least a couple of finger widths. If the box is genuinely heavy when loaded, consider a full-depth grip formed into the wall rather than a cut-out through it — it keeps the section continuous.
The measurements that matter more than dimensions
Once the geometry is right, three application facts drive the rest of the spec, and they are worth stating explicitly to your supplier:
- Loaded weight, worst case. Not average. The one SKU that is 20% heavier than the rest is the one that will split a handle.
- Stacking method and duration. Stacked nested, stacked with lids, or on a rack? Stacked load and rack load are different failure modes, and a box optimised for one is often mediocre at the other.
- What is on the line. Conveyor rollers, chain transfer, a robot gripper, or a person lifting it all day. Automated handling wants dimensional consistency above everything; manual handling wants reachable grips and a weight a person can manage at arm’s length.
A short note on standardising
Most plants end up with too many box sizes, usually because each line ordered its own. Before you finalise a new size, check whether an existing one can be made to work with a different insert. Standardising on a footprint that is a clean divisor of your pallet — euro or industrial, whichever you actually run — pays back quietly for years in pallet utilisation alone.
A corrugated PP turnover box is not a complicated object. But the difference between one that lasts three years on a line and one that is quietly replaced every eight months is usually found in the three numbers nobody wrote down.