Industry News

Home / News / Industry News / Cosmetic Packaging Dimensional Inspection: Tolerances, AQL and Cap Fit
Industry News

Cosmetic Packaging Dimensional Inspection: Tolerances, AQL and Cap Fit

A lipstick tube and its cap can both pass their outside-diameter checks and still refuse to close. On injection-molded cosmetic packaging, the dimensions that decide whether a component works are rarely the ones printed on the sales drawing: they are the mating diameter, the thread-start position, the neck T-dimension, and the thickness of decoration stacked on top of them.

Cosmetic packaging dimensional inspection exists to catch those interfaces before a filling line does. A disciplined program controls 12 to 18 characteristics on a typical airless bottle, and roughly four of them decide whether the rest of the report is worth reading.

Definition: cosmetic packaging dimensional inspection is the measurement of functional geometry, meaning mating diameters, thread form, hinge alignment, sealing bores and the build-up of decoration, against a fixed tolerance band at the mold, in process and before shipment. Color, gloss and print registration are judged by a separate visual standard and do not belong in the dimensional report.

Cosmetic Packaging Dimensional Inspection Covers Interfaces, Not Looks

Dimensional inspection on cosmetic packaging measures only the geometry that controls assembly, sealing and dispensing; everything a shopper sees is handled by a visual standard.

The measured characteristics fall into four groups, and each group fails differently once the parts reach a filler.

  • Mating and sealing geometry: cap skirt inner diameter, neck outside diameter, plug diameter and the neck T-dimension that positions a pump or reducer.
  • Motion geometry: thread-start angle, hinge knuckle position, cam track depth and the click that tells a user the cap is fully closed.
  • Fluid path geometry: pump bore, dip tube inner diameter, orifice size and valve seat flatness.
  • Decoration build-up: UV coating thickness, hot-stamp depth and vacuum metallizing layers, all of which change a mating dimension after molding is finished.

The last group is the one buyers forget. A cap measured straight off the press can pass, then fail once the same cap is metallized and printed, because 0.03 mm of coating on each side closes 0.06 mm of clearance.

+/- 0.05 mm
Working band on sealed diameters
1.2 to 2.0%
PP shrink that drives cavity correction
AQL 2.5
Major defect limit at general level II

The Five Dimensions That Decide Fit, Leakage and Cap Feel

Five measurements account for most field failures on molded cosmetic packaging, and every one of them sits at an interface between two parts rather than inside a single part.

Critical dimensions on molded cosmetic packaging, the working tolerance and the failure each one controls.
Feature Working tolerance Typical gauge Failure mode
Neck T-dimension, 24/410 +/- 0.08 mm Go / no-go plug gauge Pump sits high, leaks at the shoulder
Cap skirt inner diameter +/- 0.05 mm Pin gauge plus air gauge Cap loosens in transit or will not seat
Thread-start angle +/- 3 degrees Thread-start fixture Cross-threading and torque scatter
Hinge knuckle offset +/- 0.10 mm 200-cycle functional fixture Lid drifts and will not stay closed
Pump sealing bore roundness 0.03 mm CMM, five points per cavity Dosing drift and product seepage
Decoration build-up 0.02 to 0.05 mm Coating thickness gauge Cap binds after metallizing

Where rejects actually originate is a separate question from where the tolerance sits, and the two do not match.

Where dimensional rejects originate, share of 1,480 rejected parts
Thread-start position
31%
Cap skirt inner diameter
24%
Hinge knuckle offset
18%
Pump sealing bore
15%
Decoration build-up
12%
Compiled from reject logs on three lipstick and compact programs. Thread-start position alone accounts for almost one third of rejects.

Thread-start position dominates that list because it is the hardest feature to correct after the tool is cut. A thread that starts two degrees late still rotates onto the neck, but it leaves the cap 0.3 to 0.5 mm high on the shoulder, and the brand sees a gap line on shelf.

Why Cosmetic Packaging Dimensional Inspection Fails at the Assembly Level

Most dimensional failures on cosmetic packaging are stack-up failures: every component sits inside its own tolerance band and the assembly still does not work.

Piece-by-piece check

Each part is measured against its own drawing and released. Clearance is calculated at nominal only. No assembly is built from the extremes.

Worst-case stack

Largest neck plus coating is checked against the smallest cap. Interference appears before the tool is signed off, not after 40,000 parts are molded.

Take a 15 mm cap and neck pair. The cap skirt is specified at 15.80 mm with a plus 0.05 mm tolerance; the neck is specified at 15.72 mm with a plus or minus 0.04 mm tolerance. At nominal, clearance is 0.08 mm, which feels generous. Push the neck to its maximum of 15.76 mm and clearance falls to 0.04 mm. Add a 0.03 mm UV coating on each side of the neck and the parts interfere by 0.02 mm in the worst case, which reaches the customer as a cap that does not sit flush.

Keep at least 0.10 mm of worst-case clearance on any surface that receives decoration, and 0.05 mm on bare plastic. Anything tighter is a press-fit decision, not a tolerance decision.

Setting Up the Inspection Plan: Five Steps From Drawing to Pre-Shipment

A workable inspection plan is built before the first shot is molded, not after the first rejection, and it takes five steps to put in place.

  1. Freeze the drawing and mark the critical characteristics. On an airless bottle that means the neck T-dimension, the pump bore, the cap skirt and the decoration allowance. Between 12 and 18 characteristics is normal for a decorated package.
  2. Build hard gauges before the first shot. A go / no-go plug for the neck, a cap-fit fixture and a thread-start gauge cost less than one rejected lot and give a press operator an answer in seconds.
  3. Run a first-article layout on every cavity. An eight-cavity tool needs eight reports, repeated after 24 hours of running, because mold temperature and shrink settle over the first shift.
  4. Put the four critical dimensions on an in-process schedule. Five parts per cavity every two hours, charted, with a Cpk target of 1.33 before the process is treated as capable.
  5. Inspect before shipment to ISO 2859-1. For a 3,000-piece lot at general inspection level II the sample is 125 pieces; at AQL 2.5 the lot is accepted at 7 defectives and rejected at 8.

The plan only holds if the factory owns the gauges, the CMM and the documentation. Shunchu (Jiangsu) Co., Ltd., the cosmetics manufacturing arm of Sumseng Group, keeps mold design, injection, coating, printing and assembly in one building, which is where stack-up arguments get settled fastest: the person measuring the cap can walk to the press that molded the neck.

Common Dimensional Defects and the Mold Settings Behind Them

Almost every dimensional defect on molded cosmetic packaging traces back to four mold conditions rather than to operator error.

  • Thread-start drift: uneven cooling on the thread side, often a cavity temperature split above 4 C. Caps then cross-thread and torque readings scatter by 0.3 Nm across cavities.
  • Warpage on compact bases: low packing pressure combined with hot ejection. A 0.2 mm bow over a 60 mm base is enough to throw the hinge knuckle out of alignment.
  • Sink on thick walls: a rib-to-wall ratio below 0.6 shows up at 0.02 mm depth once a cap is metallized, because the reflective finish magnifies what bare plastic hides.
  • Flash on sealing lands: short clamp tonnage leaves 0.05 mm of flash that leaks under 0.3 bar of pressure in a vacuum test.

Where a standard mold already exists, this risk drops sharply, because the cavity, thread form and hinge geometry have been proven over thousands of shots. Several catalog programs from OEM and ODM support are built on exactly those proven tools, which is why a first article from a standard mold typically lands inside tolerance on the first pass.

OEM/ODM Magnetic Lipstick Case in Matte BlackOEM/ODM Magnetic Lipstick Case in Matte BlackOEM/ODM lipstick case with matte black and silver finish, magnetic snap closure, ABS+PC body, and standard 11.1 mm bullet fit for beauty packaging projects.View Product → OEM/ODM Water Ripple Round Air Cushion CompactOEM/ODM Water Ripple Round Air Cushion CompactEmpty round air cushion compact with water-ripple transparent lid, mirror, puff compartment, and 20 ml capacity for foundation packaging and custom logo OEM/ODM projects.View Product → OEM/ODM Airless Vacuum Press Pump BottleOEM/ODM Airless Vacuum Press Pump Bottle15 ml cylindrical vacuum press bottle with duckbill pump, PETG bottle and cap, and PP head for hygienic serum or cream dispensing in OEM/ODM projects.View Product →

Cosmetic Packaging Dimensional Inspection: FAQ

What tolerance is normal for injection-molded cosmetic packaging?

Plan on +/- 0.05 mm for sealing and mating diameters, +/- 0.10 mm for visible fit lines such as a cap-to-shoulder gap, and +/- 0.20 mm for non-functional features. ISO 2768-m works as a fallback for untoleranced geometry, but the critical characteristics on the drawing always override the general standard.

How many pieces should be measured before shipment?

Sample to ISO 2859-1 at general inspection level II. A 3,000-piece lot gives a sample of 125 pieces; at AQL 2.5 the lot passes at 7 defectives and fails at 8. Reserve 100% inspection for launch lots, decoration-critical parts, or a documented capability problem on a specific cavity.

Does dimensional inspection replace functional testing?

No. Calipers and CMMs measure static geometry, while a 200-cycle hinge test, a 0.3 bar leak test and a drop test measure behavior. A cap can sit perfectly on the neck and still fail after 30 open-close cycles if the thread flank angle is wrong.

Can a supplier guarantee zero dimensional defects?

No, because injection molding is a variable process and zero defects would mean zero variation. What a supplier can control is capability: a Cpk above 1.33 on every critical characteristic, documented first-article reports per cavity, and an agreed AQL that both sides sign before the tool is cut.

Related products

Related News