The Soy Pour Window: 5-15 Degrees Separates a Clean Candle from a Cracked One

Three defects, one cause

Every failed batch you have ever sent back carries one of three surface injuries: a hairline crack running down the jar wall, a sunken center two or three millimeters below the rim, and wax that lifts away from the glass after the first burn. Buyers call these three different problems. In the melt pot, they are one problem with three temperatures attached to it.

The difference between a clean pour and a rejected jar is not the wax, the fragrance, or the wick. It is the number on your thermometer in the thirty seconds the wax touches the glass. This article is about that number: the real pouring window for soy containers, how wide it actually is, and what to do when you land outside it.

What pour temperature actually controls

Soy wax shrinks as it cools. Roughly 8–12 percent by volume between melt and solid. That shrinkage is why a candle ever has a center dip. But shrinkage alone does not crack a candle — a shrinking body that stays bonded to the glass pulls inward and settles into a crater. A body that has already lost its bond to the glass pulls inward and tears, and that tear is your crack.

So the variable you are really managing is adhesion, and adhesion is a function of temperature and time. Three regimes, in order of the damage they do:

• Poured too hot (above roughly 140°F / 60°C for most soy-blend bases): the wax floods the glass, cools unevenly, and shrinks hardest in the center. Result: deep sinkage, frosting, sweating in summer, and a bond that is soft at the wall and fails on the first burn. High pour temperature is the quiet cause of “my candle smells great in the jar and weak in the room.”
• Poured inside the window (for a 447-type base, 125–130°F; for a 464-type base, 120–130°F; for a 2408-type base, 130–135°F): the wax wets the glass well enough to bond, stays fluid long enough to fill the wall-to-wick geometry, and shrinks into a shallow, uniform center dip that a second pour fixes.
• Poured too cold (below about 105–110°F / 40–43°C): the wax reaches the glass still partially crystallizing. It cannot wet the surface, the bond never forms, and the shrinking mass cracks along whatever thermal gradient it hit. Cold pours also give you weak adhesion at the wick channel — the classic wax shedding after burn one.

The window is 5–15°F wide. That is why “pour when it looks right” fails in QA: you cannot see the difference between 118°F and 108°F, and one of those two is a rejected jar.

Why the same base pours differently in your shop

The datasheet window assumes a clean room: 65–72°F ambient, wax at full fragrance load, wick already set in a room-temperature jar. Your floor breaks one or all three of those assumptions, and the shift is predictable:

• Warm shop (85°F+ summers): the wax falls through the window fast — often two to four degrees per minute in an open pitcher. A pour that is perfect at the tap is under-temperature by the time it reaches the third jar in the run. Fix: pre-heat the pitcher to just above the pour point, batch in smaller pours, and keep a lid or towel on the pot between jars. This is the single most common reason summer batches crack and winter batches sink.
• Cold jars: a glass jar at 60°F steals heat from the first contact layer and starts it crystallizing against the wall — locally “too cold” even if the bulk pour is in window. Warming jars to 80–85°F (a few minutes in a 90°F oven, or a hot-water rinse and full dry) measurably reduces wall cracks in thin-glass tins and jars.
• Fragrance above 8–10 percent: higher load raises the melt, widens the crystallization spread, and narrows the effective pour window. If your load is at the top of the range, pour at the top of the datasheet window, not the middle.

The execution: a pour you can repeat

Here is the sequence we run on the line for a 350g soy-blend container candle, 8% fragrance:

1. Melt fully to 170–180°F (77–82°C). Undissolved fragrance crystals are a separate defect — pour temperature cannot rescue them.
2. Pull off heat, add fragrance at flashpoint minus 10°F, stir 60 seconds. This cools the bath a few degrees and starts crystallization.
3. Watch the number, not the clock. For a 447-type base, start pouring at 125–130°F. Mark the pitcher: first jar of the run at the high end, last jar at the low end of the window. If the last jar is outside it, your batch is too big for the room — split it.
4. Wick in before the pour, trim to 1/4–3/8 inch. The wick channel is where the bond is weakest. Pouring cold over a set wick is what makes wax drop off after burn one.
5. Let it set 20–30 minutes, then spot-pour the center. Bring the pot back to 125–135°F and add a small pool over the sinkage. A flat or shallow center is a spec; a 3mm crater is a burn-time defect waiting to happen, because the melt pool will hit the wick unevenly.
6. Cure 72 hours minimum for adhesion, 7–14 days before shipping anything at load above 8%. Curing is when the wax lattice locks around the fragrance. Hot throw measured on day two will understate the candle by a wide margin; a 2408-type base at day one can read half its final throw.

Diagnosing what you already have

Take one failed candle from the batch and read it:

• Crack on the wall, bond clean along the crack face: pour too cold, or jar too cold. Move the first pour of the run 5–10°F higher and warm the jars.
• Sunken center 3mm+, frosting, soft wall bond: pour too hot. Drop the start temperature 5–10°F, batch smaller, cover the pot.
• Wax shed from glass after burning: bond failure regardless of appearance — recheck both pour temperature and cure time, then your wick channel geometry. A wick too fine for the size will melt a pool that detaches a marginal bond; too coarse, and the over-pour of flame heat does it for you.
• Clean surface, weak hot throw: not a pour defect at all. That is load and curing time; see the ceiling math in the fragrance load article. Chasing throw with hotter pours is how you trade a weak-throw defect for a sinking defect.

The QA line on the batch sheet

Four fields, every batch, no exceptions: pour-start temperature, pour-end temperature, ambient temperature, and spot-pour time after initial set. When a customer photo shows a crack, those four numbers tell you which jar in the run it was and what to move next. Without them, you are tuning by guesswork and burning batches to learn what a thermometer would have told you for free.

The pour window is the cheapest control you have in the entire process. One degree of data replaces one rejected carton. If your line is producing any of the three defects above, stop changing wax and start writing down the number on the thermometer — then move it five degrees at a time until the defect disappears. It will disappear. That is the whole trick.