For an uncompensated closed square cut, estimated kerf is half the difference between its measured opening and cut-out widths. The corresponding geometric path-offset magnitude is half that kerf.
Use measurements along the same axis of the same uncorrected square cut. All dimensions are in millimetres. The prefilled example is illustrative.
Compare the disagreement with your measuring instrument's uncertainty. A large mismatch calls for checking scale, squareness and measurement technique before using an offset.
Calculated locally. Repeat measurements for both axes and the actual material/setup. This estimate does not select cutting parameters or approve machine operation.
Let D be the drawn width, P the cut-out width and H the opening width, measured along the same axis. Assuming equal material removal around a centred toolpath: P = D - k and H = D + k. Therefore full kerf k = (H - P) / 2, and half-kerf offset magnitude = k / 2.
A 20 mm drawing with a 19.8 mm cut-out and a 20.2 mm opening gives 0.2 mm estimated kerf and a 0.1 mm offset magnitude. The separate estimates D - P and H - D are both 0.2 mm. This is an illustrative calculation, not a measured DTU machine result.
The inside and outside estimates should agree within the uncertainty of your measurement and process. A mismatch can reveal measurement difficulty, non-square geometry or a scale error that a single average could hide. This tool reports the difference without inventing a universal acceptable threshold.
It is a geometric estimate. LightBurn notes that tiny dimensional differences can be difficult to measure accurately even with calipers, and provides a Vernier kerf test as an alternative. Repeat measurements on both axes, record the material and setup, and use the operator's validated calibration workflow for close-fitting assemblies.
Full kerf is the width of material removed. A compensating path shift is usually described using half that width. Check whether your software asks for a full cut width or an offset, how it handles inside/outside contours, and whether compensation is already active. Applying compensation in both CAD and cutting software can overcorrect a part.
Use a matching cut-out and opening from the same uncorrected closed cut, with existing kerf compensation disabled for the test. Keep units, axis and measurement positions consistent. Material, focus and other setup changes can change the result. Approved materials and operating parameters must be confirmed with the machine operator.
Next: record a repeatable prototype test, review the laser cutter preparation page, or browse all prototyping tools.