What adhesion actually is, and the three ways it fails

Glass adhesion is how well the set wax stays in contact with the inside of the vessel — across the whole wall, through temperature changes, and through the burn. Three separate tests are hiding in that sentence. The whole wall is a spatial test: is any part of it showing air behind? Through temperature changes is a thermal test: does it hold at 22°C in January and 34°C in May? Through the burn is a dynamic test: does it hold when a melt pool has softened the top two centimetres?

A candle can pass one and fail the others, which is why "poor adhesion" is a useless diagnosis on its own, and why advice in maker groups contradicts itself — two people describing different faults with the same phrase. Name which of the three you have before you change anything.

Two of these are set in the first hour after the pour. The third is set at the moment you decide which sack of wax to open. That ordering matters, because it tells you where to spend your attention: a maker fighting slump by adjusting pour temperature is solving the wrong problem with the wrong tool, and will keep failing for as long as the wrong wax is in the jar.

The six things that drive adhesion, in order

Everything that determines whether wax grips glass falls into six variables. They are not equal. The first accounts for more failures than the other five combined, and the last two are the ones makers most often blame first.

One mechanism sits behind that ranking. Adhesion is decided in the last few seconds of the liquid state at the wax–glass boundary. A hot jar keeps that boundary liquid a little longer, letting the wax flow into the microscopic texture of the glass before it locks up. A cold jar freezes a skin against the wall immediately, and that skin then shrinks as the rest of the candle cools, peeling itself off a surface it never properly wetted. That explains driver two, most of driver three and much of driver five.

Driver one outranks all of it because it changes what the wax is trying to do. Container blends are built so the set wax stays soft and tacky enough to keep contact with a wall. Pillar blends are built the opposite way: harder and more contractile, so the candle pulls in from the mould and slides out. Both are correctly formulated. Only one is trying to stick.

Vessel shape sits at six because it concentrates adhesion failure rather than causing it. A shoulder, an internal step, a heavy moulded base or a visible seam gives the wax somewhere to bridge across instead of press into, so a wax that behaves in a straight-sided jar shows a repeatable mark at the shoulder of a curved one. If the fault lands in the identical position on every unit, stop adjusting temperatures and look at the vessel. CSI lists 80 products in the Containers For Candle Making collection, and straight-sided glass is the easy setting for anyone still stabilising a process.

Grade is the first thing to check, and usually the last thing you need to change

If you take one thing from this page: open the sack and read what it was made for. The wax names on the CSI store are not marketing decoration, they are the specification. Two of the three below are container grades. The third is not, and putting it in a jar is the most common cause of the slump symptom in the diagram above.

Clean glass and a warm jar — the two steps nobody mentions

Wax does not bond chemically to glass. It grips by settling into the microscopic texture of the surface while still liquid, then freezing in place. Anything sitting between the wax and that texture is a release layer. Packing dust, a smear of factory oil, a fingerprint from whoever unpacked the carton — each gives the wax something smooth and slightly greasy to sit on instead of glass.

This is the most under-discussed cause of adhesion faults, and it has a tell: randomly positioned, sharply-bounded marks that do not repeat between units. Cold-glass wet spots are diffuse and cluster. Contamination marks are discrete, sometimes arc-shaped, and appear on one jar in a tray while its neighbour is perfect. If your failures are scattered rather than systematic, wash before you re-engineer.

On warming: the target is not precision, it is the absence of a cold shock. A jar merely warm to the hand beats a jar at room temperature in a December workshop, and a jar near pour temperature is better again. The practical methods:

Cooling is the other half of the same idea: having kept the boundary liquid, do not then chill it. Pour onto wood or cardboard rather than a cold steel bench, leave a hand's width between jars so they share warmth, and keep the tray away from fans, AC vents, open windows and doorways. Never refrigerate to speed setting — rapid cooling maximises the shrinkage stress that produces the top-edge gap. Cover the tray loosely with a carton in a draughty room. The seasonal swing is real: the same recipe poured on a 34°C May afternoon and a 16°C January morning is not the same process, and January is the batch that shows the fault.

Pour temperature sits fourth because it does less for wall contact than makers assume, but it is not nothing. Eco Soy CSI 400 specifies fragrance addition at 80–90°C with two minutes of stirring, and CSI 464 states melt and pour in the same 80–90°C band. Work to the figure stated for your wax, read it off a pen thermometer at ₹354 rather than judging by eye, and change it in small steps. Pouring hotter buys longer at the boundary but deepens shrinkage and worsens the top — a real trade-off, not a single best number. A 1-litre mini wax melter at ₹2,360 holds temperature steady while you fill a tray, which is worth more than another degree of thermometer accuracy.

Adhesion during the burn — and why a warm gap is usually not a fault

This is the situation that generates the most unnecessary customer emails. A candle leaves your workshop with a flawless wall. The customer lights it, and an hour later there is a visible gap around the melt pool, or the wall looks cloudy and detached halfway down. Nothing has gone wrong.

Wax has a far higher coefficient of thermal expansion than glass. As the candle warms — from the flame, a hot room, a courier van in the sun — the wax expands considerably and the glass barely at all, and on the way down the wax contracts more. Each cycle works the boundary a little, so a candle perfectly bonded on day one can show separation after a few burns. It is normal, common to every container candle in every wax, and the appearance usually improves again as the candle cools.

Put that distinction in front of customers before they find it themselves. One line on a candle care sticker — ₹413 for 500, or ₹0.83 a unit — saying natural soy may pull slightly away from the glass as it warms and that this does not affect the burn turns a complaint into a mark of authenticity. Same logic as explaining frosting: an expected characteristic named in advance stops being a defect.

Which raises the fair question: how much does any of this actually matter? Honestly, less than the effort makers put into it — with one exception worth checking.

That last row is why you keep a standard rather than shrugging the whole thing off. It is also rare, and it belongs to the slump category — which means it belongs to the grade question at the top of this page. Fix the grade and the safety-adjacent case largely goes with the cosmetic ones. For everything else, set a written accept/reject line, apply it consistently at packing, and stop re-pouring units that would have sold perfectly well.

Testing adhesion properly — the two-jar method

Opinions about adhesion are worthless because everyone is comparing different candles poured on different days in different weather. The only useful evidence is two jars poured from the same pitcher, differing in exactly one variable. Here is that test for the highest-value variable on this page: the warmed vessel.

One last argument, and it runs against our own interest. Nothing on this page needs you to spend money. The two changes that fix most adhesion complaints are using a container-grade wax — at ₹507.40 per kilogram, CSI 464 is cheaper than the pillar wax many makers are wrongly using — and warming a clean jar before you pour, which costs an oven that is already on. Do not buy a new jar range, additives, or a second wax to blend in. Make those two changes, run the two-jar test to confirm, and only then look at anything that costs money. If your candles are still failing after that, the diagnosis is somewhere in how to verify wax quality before you buy — message us and we will work through it with you rather than sell you something.

Frequently asked questions