Cutting foam with a hot wire, and the cut that is wider than the wire
Kerf in thermal cutting, and how to allow for it in a pattern.
A wire heated hot enough to melt through foam does not simply push the material aside along a line exactly as thin as the wire itself, it carries enough heat that the foam for a short distance on either side also gets hot enough to soften and pull back, so the gap left behind, the kerf, ends up measurably wider than the wire that made it.
What is really going on
A hot wire cuts foam by melting or vaporising the plastic directly in its path rather than by any mechanical sharpness, since the wire itself is usually a plain, smooth strand of metal with no edge at all. Heat spreads outward from the wire by thermal conductivity, and expanded foam, built almost entirely of trapped air inside a thin plastic structure, softens and melts at a fairly low temperature, so the zone hot enough to give way extends a short distance beyond the wire's actual physical width in every direction. The faster the wire is moved through the material, the less time that heat has to spread sideways before the wire has already passed by, so a fast pass leaves a narrower kerf closer to the wire's own thickness, while a slow pass, or a wire run hotter than it needs to be, lets more heat spread outward and leaves a noticeably wider channel than the wire that produced it. A thicker wire compounds the effect, since it holds more heat along its length to begin with and radiates it outward from a wider starting line, so a change in wire gauge shifts the kerf by more than the simple difference in the wire's own diameter would suggest.
The hot-knife comparison
Cutting through a cold block of butter with a knife heated under hot water leaves a channel that is not simply as wide as the blade, it is a touch wider, with a thin margin on either side softened and slightly melted by heat that spread out from the metal before the cut passed through. A cold knife through the same butter leaves a channel that matches the blade's own thickness far more closely, since there is no heat spreading outward to soften anything beyond direct contact. The extra width left by a hot blade in butter comes from exactly the same cause as the extra width a hot wire leaves in foam, heat conducting a short distance into the material before the cut is finished, softening a margin the tool itself never physically touched.
The one number worth remembering
The melted margin a hot wire leaves beyond its own diameter can easily be half again as wide as the wire itself, sometimes wider still depending on how fast the wire is moved and how hot it is running, so a pattern cut to an exact outline with a hot wire will consistently come out smaller than intended unless that margin is allowed for before cutting rather than corrected afterward.
What follows from this
Because the kerf is predictable for a given wire, temperature and cutting speed, a pattern intended for hot-wire cutting is usually drawn slightly oversized, offsetting the outline outward by roughly half the expected kerf width so the finished part, after the melted margin is accounted for, comes out at the intended size rather than undersized by the width of material the wire quietly took with it. Keeping the wire's speed and temperature consistent across a cut matters for the same reason, since a wire that slows down partway through a pass, whether from resistance in the foam or from a shaky hand, spends longer heating that section and leaves a locally wider kerf there than everywhere else, producing a part with an outline that is not consistently offset the way a steady pass would leave it. Guiding the wire along a rigid edge or template rather than freehand helps for exactly this reason, since a steady, unhesitating pass through the foam keeps the exposure time, and therefore the kerf, as close to constant as possible along the whole length of the cut.