Should My Jar QC Track Internal Diameter, Height, Weight, Wall Consistency and Visible Defects Before the Containers Reach the Pouring Line?

Yes, track all five, but not in the same way

Internal diameter, height, weight, wall consistency and visible defects each protect a different part of your candle, so a jar record that leaves one out has a blind spot. But they are not measured the same way or on the same number of jars. Visible defects are a safety matter and are checked on every jar. The four dimensional properties are checked on a sample and compared against a reference. Designing the record around that difference keeps it quick enough that people actually fill it in.

The other design principle is that a record must lead to a decision. Every field should either feed the accept, quarantine, conditional release or reject call, or help you trace a problem later. A field that does neither is clutter, and clutter is what makes people skip the sheet on a busy day.

Wall consistency: the property most records get wrong

Wall consistency is the hardest of the five to measure at the bench, because a standard caliper only reaches the rim. You cannot read the wall half-way down a closed jar without specialist equipment. That is not a reason to leave it out; it is a reason to record the evidence you can actually collect and be honest about its limits.

Three pieces of evidence work together. First, rim wall thickness at four evenly spaced points: the difference between the thickest and thinnest reading, which you can call the wall spread, shows whether the glass was formed evenly at the top. Second, the spread of empty weights across the sample, which reflects variation in glass quantity. Third, a look through the wall against a light: uneven glass often shows as distortion, a thicker band or a visible wedge when you turn the jar.

The record itself: four blocks on one sheet

Keep each shipment on a single sheet or a single spreadsheet tab. It divides naturally into four blocks, and each field earns its place for a reason.

Block 1: identity of the shipment

Block 2: sampled measurements

One row per sampled jar, with columns for carton number, both internal diameter readings, outer height, internal depth, the four rim wall readings, the wall spread and the empty weight. At the foot of each column, the minimum, maximum and average. The extremes matter as much as the average, because the widest or lightest jar is the one that behaves differently.

Block 3: defect tally on every jar

Columns for chips, cracks or suspected cracks, open bubbles, large enclosed blisters, heavy scratching and rocking bases, with counts per carton. Chipped and possibly cracked glass is removed from candle production without exception, so this block also records that those jars were physically separated and how they were disposed of.

Block 4: decision and follow-up

The decision in one word, the reason in one sentence, any conditions attached to a conditional release, any test burns required with their results, the date of any supplier contact, and the signature of the person releasing the shipment to the pouring line. For CSI orders, note the date you raised anything on WhatsApp (+91 7397976926) and a summary of the reply.

Worked example: the third shipment that the trend caught

The real value of a consistent record shows up across shipments. Here is an illustrative history for one jar style, using figures of the kind you would collect yourself.

Shipment C breaches three bands. The internal diameter is 1.6 mm above reference; the weight band is 235 × 0.95 = 223.25 g at the low end, and 221 g falls below it; and the largest wall spread of 1.4 mm is beyond the flag. The defect rate is 13 ÷ 500 = 2.6%, above the escalation point. Any one of these could be noise; all four together, moving in the same direction, point to a change in the glass.

Because the record caught it at goods-in, the decision is calm rather than urgent: quarantine Shipment C, burn-test candles in the widest and lightest jars beside a candle in a Shipment A reference jar, and contact the supplier with the trend table rather than a vague complaint. Without the record, the first sign would have been customer reports of wet walls or uneven burns weeks later.

Keeping the record usable

Store the sheets in date order in one place, and keep a small summary table like the one above that you update after every shipment. A year of trend data for each jar style is enough to set tighter, evidence-based bands and to decide when a supplier has earned a smaller sample.

Paper or spreadsheet both work; what matters is that the layout never changes. If you use a spreadsheet, lock the column order and let it calculate the minimum, maximum, average and wall spread automatically, so the person measuring only types readings. If you use paper, print the same template every time and transfer the averages into the summary table the same day. A record that changes shape from one shipment to the next cannot be compared, which defeats its purpose.

Keep one labelled reference jar from each accepted shipment for as long as you are selling candles made in that batch of glass. If a customer complaint arrives months later, you can measure the actual jar rather than rely on memory.

And use the same sheet for every supplier, including us. CSI does not publish internal diameters, glass thickness, jar weights or tolerances, so the record you build is the specification you have. It should judge every shipment by the same numbers, whoever sent it.

Where this page fits

CSI already has guides that cover parts of this question. This page covers the goods-in check for the question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions