My New Jar Has the Same Capacity but a Different Shape — Why Might Glass Adhesion Change?

Capacity is a volume. Adhesion happens on a surface.

When a candle that behaved in one jar starts spotting in a jar of the same capacity, it feels as though nothing has changed: same wax, same oil, same load, same fill weight. But the wax has a new container around it. A different shape commonly changes four things at once — how much glass each gram of wax touches, how deep the wax sits and therefore how long the centre stays warm compared with the wall, which way the walls lean as the wax contracts, and where the curves, shoulders and heels sit. Any of those can change how and where the wax sets against the glass. None of them is visible on the capacity label.

The single cheapest first test is a split pour: one batch of your normal candle, poured at your normal temperature into three of the old jars and three of the new ones, cooled side by side on the same rack, and photographed at the same checkpoints. If the new jars spot and the old ones do not, the container is implicated — shape, glass or both. If both spot alike, the shape is not your problem and something else moved at the same time; the jar-supplier-change post and the wet-spot diagnosis capstone cover that branch. If the new jars spot, the next step on this page is a zone map: recording exactly where on each jar the spots sit, because shape-driven spots tend to cluster at the shape's features rather than appear at random.

The worked example is a maker moving from a tall straight tumbler they have used for a year to CSI's 200 ml Salsa Glass Jar, filled to the same weighed 170 g — an illustrative figure, because CSI does not publish the Salsa jar's fill weight (200 ml is its brim volume). The candle is 13.6 g of Japanese Cherry Blossom and 156.4 g of Luxury Soy Wax CSI 464 at 8%. CSI's Salsa page describes clear glass with a golden lug cap; it does not publish internal dimensions, glass thickness or a hot-pour limit, so every dimension below is illustrative and every measurement that matters is yours.

Three jars that hold the same candle, measured

Geometry is the one part of this problem you can calculate before you pour anything. Take one candle of 170 g. Using a labelled approximate density of about 0.88 g of wax per ml — wax is lighter than water, the figure varies by wax and temperature, and it is used here only to turn weight into a shape — that candle occupies about 193.2 ml. Pour that volume into three illustrative jars: your old tall straight tumbler with a 6.5 cm internal diameter, a wide low tumbler at 7.5 cm, and a flared jar that is 6.0 cm across the base and 8.0 cm across the mouth. The wax depth, the glass it touches and the open top all change, although the candle is identical.

Read the table across. Moving from the tall tumbler to the wide low one, the wax depth falls by 24.9% and the glass the wax touches falls by 3.2%, while the open top grows by 33.1%. Moving to the flared jar, the glass touched falls by 7.9% and the wall leans out by about 11.3° from vertical. None of these numbers tells you that one shape adheres better — CSI publishes no adhesion data for any jar, and geometry alone predicts nothing about your wax. What the table does is tell you which differences exist, so that when the zone map shows spots in a particular place, you have a mechanism to test rather than a guess.

Two geometric points are worth holding on to. First, a wider jar puts more wax between the centre and the wall at the same depth, so the core can stay warm and keep contracting after the wall face has set. Second, a jar whose walls lean outward gives contracting wax a direction to move away from the glass: as the candle settles downward, a wall that widens upward is no longer pressed against it. A shoulder — a jar that narrows towards the mouth — does the opposite at the top: wax under the shoulder is held from above and can pull down and away from the underside of the curve. These are mechanisms commonly described in candle-making, not measurements of CSI's Salsa jar.

The zone map: where the spots sit tells you what the shape did

A photograph of a spotted jar is evidence; a photograph of eight spotted jars, each scored zone by zone, is a pattern. Divide every jar into five zones you can see through the glass: the shoulder or upper rim band, the upper wall, the middle wall, the lower wall and the base heel where the wall meets the base. Turn the jar to four marked sides — label them A, B, C and D with a marker on the base — and score each zone on each side. That gives twenty cells per jar. A marked acetate sleeve with the zone lines printed on it makes the scoring repeatable from jar to jar.

The point of the map is that it turns "the new jar is worse" into "the new jar is worse in the lower wall and heel on every side", and that second sentence suggests its own next test. It also protects you from blaming the shape when the real cause is the rack position: if the spots move from side A to side C when you rotate the jars on the rack, they follow the room, not the glass. Rotating half the jars by 180° on the second day is a cheap way to check this.

The two-round shape test, jar by jar

Hold everything constant that is not the shape: the same bag of Luxury Soy Wax CSI 464, the same bottle of Japanese Cherry Blossom at 8%, the same wick from the same pack, the same pour temperature, the same cleaning routine and the same rack. Clean every jar of both shapes identically — a wipe with Isopropyl Alcohol Spray and a lint-free cloth, then let them dry completely; isopropyl alcohol is flammable, so keep it away from flame and the melter. CSI's CSI 464 page gives an 80–90°C melt/pour working range. CSI's Salsa page publishes no hot-pour limit, so pick one temperature inside the wax page's range that your current jar already uses, keep it identical for both shapes, and ask the jar supplier — or us on WhatsApp — before pouring at the top of the range.

Three jars per arm is a working practice for a screen, not a statistical rule. Each extra Salsa jar in this example costs about ₹193.38 in CSI materials — the jar at the pack-of-5 rate, 156.4 g of CSI 464 at the 1 kg rate, 13.6 g of Japanese Cherry Blossom at the 100 g rate and an example Eco Thin C1 wick. Going to five jars per arm in round 2 would add ₹773.51. Do it when round 2 gives a close call, not by default.

The zone-map scoring sheet

There is no industry standard for an acceptable wet spot, so the scale is your house standard, written at the top of the sheet before the first pour. The example below grades each zone on each side from 0 to 3, so a perfect jar scores 0 and the worst possible jar scores 60. You can also count squares on a 1 cm acetate grid if you want an area figure; whichever you choose, use it for every shape test that follows so that your records stay comparable.

Add each jar's four side totals for a jar score out of 60, and add the zone columns across all jars of one arm for a zone profile. The jar score answers "which shape is worse?"; the zone profile answers "where?". Write your acceptance rule before pouring — for example, "an arm passes if no jar scores above 4 at 7 days and no zone averages above 1 across the arm". That rule is your labelled choice; no industry figure exists. What to record when comparing adhesion lists the rest of the batch-sheet fields worth keeping.

The priced shape-test kit

Both rounds, priced with listed in-stock CSI packs. Grams use the CSI convention (oil = W × L) with a 5% pour-loss working allowance. Your old tumblers come from your own stock. The Salsa jar is sold in 5s, 50s and 100s; for a test, buy the smallest packs that cover the count.

The wick line is an example only. CSI publishes no wick size for any jar, and a new shape re-opens the wick check as well as adhesion: a wider mouth needs a melt pool that reaches further, and a deeper jar changes how the pool develops as the candle burns down. Run that burn test separately from the adhesion jars — never light an adhesion jar before its 30-day photograph.

What the shape costs in production

If the Salsa jar passes, the materials barely move: the candle is still 156.4 g of CSI 464 and 13.6 g of oil. At the 10 kg CSI 464 rate (₹507.40/kg), the 1 kg Japanese Cherry Blossom rate (₹4.84/g), an example Eco Thin C1 wick at ₹5.90 and the Salsa jar at the 100-pack rate (₹29.50), the material cost is ₹180.61 a candle before labour, packaging, freight and GST. The Salsa jar comes with its golden lug cap, so there is no separate lid line.

Note the step: the pack of 5 works out at ₹40.12 a jar and the 50 and 100 packs at ₹29.50. That is one of the few real per-unit breaks in the jar range, and it is a reason to finish the shape test before ordering 50 — a failed shape bought by the hundred is the expensive outcome. If the test passes only with a change such as a lower fill line, recalculate: a lower fill is a smaller candle at the same jar price, and your price per gram of candle rises.

The decision rule

Read the sheets from the top and stop at the first line that is true. Every line ends in a purchase or a process change.

A wax purchase appears only on the fourth line, after the shape has been shown to be the cause and a process change has failed to fix it. That order matters because a wax change re-opens wick, load, cure and hot-throw validation for every shape you sell. Whether container shape belongs in wax qualification takes the same idea one level up: if you sell several shapes, which ones does a wax need to pass in?

Record the shape as a container change

A shape change is a container change, and the KEEP / CONFIRM / REPEAT / NEW vocabulary makes the scope obvious. Keep this record with the zone sheets; the next time a delivery of the same jar looks different, you will be able to compare like with like.

Keep three numbered jars from the passing arm, unlit, as retained samples, and photograph them again at 30 days. Why one supplier's glass can behave differently, what thicker glass does to the pour and giving every new vessel a small adhesion test are the neighbouring questions; the section hub routes the rest.

The CSI products in this shape test

Frequently asked questions