You pull a fresh batch out of the pouring station and everything looks right. Level surface, wick standing true, no sinkholes, no starbursts. Then the next morning the box arrives from a customer with a photo: a clean, straight fracture through the wax, running from the rim down into the jar. It looks like shipping damage. It is not. Nothing moved that candle at all.

Soy wax fractures in the cooling phase, and the fracture is already decided by the time you poured it. Most of the advice online treats cracking as a defect that happens to the candle. It is not. It is a direct readout of what happened between 140 degrees Fahrenheit and room temperature, and every step in that window either adds stress to the wax or removes it. This is the part of the process where there are no rescue moves, because by the time you can see a crack, the forces that made it finished working hours earlier.

The problem: soy is a crystal lattice, not a melt that sets.

Paraffin sets by solidifying as a glassy mass. Soy is a long-chain triglyceride that forms a crystal lattice as it cools. Every soy wax you buy is a blend of fatty acids with different chain lengths, and each one nucleates its own crystals at its own temperature. That is why the same wax, poured the same way, on two different days, can end up smooth on one and crazed on the other. The crystal structure is not set by the wax alone. It is set by the wax and the speed at which it cooled, and the jar is part of that equation.

Here is the physics in one sentence: soy wax shrinks as it solidifies, and that shrinkage is roughly five to seven percent by volume, spread across a cooling range of about forty to fifty degrees Fahrenheit. Five percent is not nothing. In a 3-inch diameter jar, that is a real gap of tenths of a millimeter that the wax wants to close. If the wax is anchored to the glass, it cannot close that gap freely, so the strain builds inside the solid until it either relaxes plastically or fractures. Crazing on the surface is the small strain releasing in small cracks. A straight line through the candle is one big strain releasing at once, usually at a stress concentrator, and the most common stress concentrator in the whole system is the junction where the wax meets the glass.

Where the failure actually starts, in order of how often I see it:

1. The glass anchor, and the pour that was too hot to matter. When wax is at 140 degrees and the glass is at 75, the outer layer against the glass skin-cools fast. That outer skin sets first, and it locks the rest of the candle to the jar. Now the inner mass, still warm, still contracting, has nowhere to shrink into. All the contraction that should have happened freely now happens as internal stress against that frozen rim. The standard fix, pouring at a lower temperature with the wick taped and the jar at room temperature, is not mysticism. It is the difference between the whole mass contracting together and a hot core being held by a cold shell. Pour at 130 to 135 instead of 145 and the outer skin forms more slowly, the whole candle sets as one mass, and the contraction is shared instead of stored.

2. The wick, which is a crack starter wearing a collar. A wick that was inserted into a hot pour and then cooled under stress is a column of non-shrinking material running through the center of a shrinking body. It is a built-in stress riser, exactly like the line in the middle of a cracked ice sheet. If the wick is slightly off-center, worse. If it was inserted a little late, after the outer rim had already started to skin, you have now welded a foreign object into a stressed structure. Check the wick position after the first cooldown, not just before the pour.

3. The jar itself, which most people never think about until the third cracked jar in a month. Thin glass and glass that came from the warehouse cold both set the rim faster and harder. A 3mm wall cools the wax against it roughly twice as fast as a 5mm wall does. If your shop is in a warehouse and you are pouring straight from the rack in winter, your jars are at fifty-five degrees and acting like a cold plate. The fix is not to heat the jars, it is to bring them to the same temperature as the room you are cooling in, so the contraction on the wax side and the glass side happen together. Pre-warm jars to about ninety degrees, pour, and cool at room temperature, never on a counter over a cold tile floor, never in a draft.

The fix list, in the order that actually works:

  1. Slow the set. Pour at 130 to 135, jar at room temperature or pre-warmed to about ninety, and let the candles cool for twenty-four to forty-eight hours undisturbed. The longer the cool, the more time the crystal lattice has to reorganize into a low-stress structure before it locks. This is the same reason you do not pull a candle off the shelf on day one and call it finished.
  2. Do not pour from a hot pour station into a cold room. If the pour station sits at 140 and the room is sixty-eight, every candle takes a thermal shock the moment it leaves the funnel. Set the pour station to your working temperature and hold it there. The difference between a pour at 132 and a pour at 142 is roughly the difference between a candle that sets together and one that sets in layers.
  3. Check your wick collar. If you are using a collar, make sure it is centered and not sitting high in the wax. A collar that is off by even two millimeters puts a lateral load on the wick during the whole contraction, and that load has a preferred direction. Candles cracked in one direction, always the same direction, are telling you your wicks are not centered.
  4. Stop judging a batch from the first ten. The first ten candles out of a pour are the ones that took the longest to cool, sitting against the others while the batch settled. They are the worst samples you have. If you want to know how a batch cracked, open the last ones that came off the line, not the first.
  5. Store them standing up, not on their side, for the first week. A candle on its side lets the wax pull away from the glass on one side and anchor on the other, which is a small but real asymmetry in the final stress state. Standing, they set symmetric. It sounds trivial. It is not, in a batch of five hundred.

The number that separates the two camps: most of the cracking I have seen in production came from a pour temperature at or above 142 with jars below seventy-five, or from a fast cool, under eight hours, in a room that was actively air-conditioned. Both are conditions you can change tomorrow without changing your wax, your fragrance, or your jar. The cracking that is still happening after those two fixes is the cracking that belongs to the wax itself, the blend, the crystal habit, and it is solved by changing wax, not by changing process.

Your wax is not failing. Your cooling is. The fracture is a record of the last forty-eight hours, and it is written in the only language a solid wax speaks, which is where it could not move.