Candling Design in Egg Cartons
Defining Candling Properly
Two people can measure candling and get different numbers without either being wrong, because the definition was never fixed.
Fix the point, the instrument and the timing first.
A candling window placed for convenience costs strength elsewhere.
Where Candling Assumptions Break
The common assumption is that last job setting will work here. The furnish moved, so the setting did.
The second is that a visual check is a measurement. It is a screening tool.
On candling the same symptom traces to one of three places: the mould, the furnish, or the drying profile. Checking in that order costs less than checking in parallel.
Reading the Candling Data
Record the setting before the change and the result after a full run. A change without a baseline cannot be reversed with confidence.
Look for the pattern before the average; a mean that hides one bad position is how a real defect survives review.
Record position as well as value.
The Practical Band for Candling
A band is more useful than a point. Hold inside it and stop chasing the centre, which moves with the weather.
Set the band from the tolerance the customer specified.
Stopping the Candling Adjustment Loop
Stop when the measurement holds for a shift and the next change would cost more than the variation it removes.
Continuing past that is how a stable line becomes one adjusted every hour.
Where to Start With Egg carton candling design
Start with the measurement, not with the setting. Take five parts from the current run, measure the same point on each, and write the numbers down before changing anything. That record is the only baseline you will have, and without it every later adjustment is a guess dressed as a decision.
Then change one input and wait for a full cycle plus the transport delay before looking again. Corrections stacked inside that delay overshoot every time, and the resulting part usually sits further from target than the one you started with.
Finally, ask whether the variation you are chasing is inside the tolerance the customer actually specified. Lines that are tuned past that point consume attention that would be better spent on the next defect down the list.
Questions Buyers Ask
What is a normal range for candling?
The practical band depends on the part, but the rule is the same: hold it through a full shift including the first hour after changeover.
How do I tell a candling problem from a furnish problem?
Check the tool or board first, then the material, then the process. A defect on every unit is process; on one position it is tooling.
How much does it cost to fix candling?
Most fixes are process or tooling time rather than capital. That order matters.
When should I stop tuning candling?
Stop when the measurement holds for a full run and the next change would cost more than the variation it removes.
Next Step
Send the drawing, the target and the current settings, and the list can be narrowed to two or three likely causes before any hardware change is discussed.
See also see the related detail.
See also see the related detail.
See also see the related detail.
See also see the related detail.
See also read the full note.