Last updated: 2026-09-05 · 5 min read

Dimensional Drift Lot to Lot in Injection Molding — What Causes It, How Do You Measure It Across Production Batches, and What Process Controls Keep Critical Dimensions Within Tolerance?

Dimensional drift is the systematic shift in part dimensions between production lots — caused by resin batch variation (shrinkage ±0.1%), ambient temperature changes (mold temp shifts ±2°C across seasons), and mold wear (gate erosion over 50,000+ shots). Typical drift: ±0.05–0.15mm per 10,000 shots for amorphous resins, ±0.10–0.25mm for semi-crystalline. Statistical process control (SPC) with CMM measurement every 500–1,000 shots detects drift before parts go out of tolerance.

  • Root causes of dimensional drift: (1) Resin batch variation — each lot has slightly different shrinkage (MFI varies ±5–15%, shrinkage varies ±0.05%). A resin with nominal shrinkage 0.5% may range 0.45–0.55% across batches — on a 100mm part: 0.05mm dimensional shift. (2) Ambient temperature — mold cooling water temperature varies with season (15°C winter vs. 30°C summer). A 5°C shift in mold temp changes part shrinkage by 0.02–0.05%. (3) Mold wear — gate erosion enlarges the gate over 50,000+ shots, changing fill balance and packing. A worn gate reduces holding effectiveness → part shrinks more → dimensions decrease.
  • Measurement method: CMM (coordinate measuring machine) measures 5–10 critical dimensions on 3–5 parts sampled every 500–1,000 shots. The data is plotted on an X-bar/R chart (SPC). Control limits: ±3σ from the mean. If a measurement falls outside control limits: the process has shifted and investigation is needed. Trend analysis: if 7 consecutive points trend upward or downward: the process is drifting (even if still within control limits). Early detection allows parameter adjustment before parts go out of tolerance.
  • Process controls to minimize drift: (1) Lock molding parameters after T1 approval — any change requires documented approval (material, temperature, pressure, time). (2) Use resin from a single supplier and single grade — switching suppliers introduces shrinkage variation. (3) Monitor mold temperature with a data logger — alert if temp deviates ±2°C from setpoint. (4) Schedule mold maintenance every 25,000–50,000 shots (clean vents, check gate, polish cavity). (5) Measure first-off and last-off parts from each production run — compare to baseline to detect long-term drift.

What Are the Three Main Sources of Dimensional Drift and How Do You Control Each?

Source Mechanism Typical Drift Control Method
Resin batch variation Shrinkage varies ±0.05% per lot ±0.05mm per 100mm Single supplier, incoming shrinkage check
Ambient temperature Mold temp shifts ±5°C seasonally ±0.02–0.05mm per 100mm Chiller with ±0.5°C control
Mold wear Gate erosion over 50k+ shots −0.05–0.15mm per 50k shots Scheduled maintenance every 25–50k shots
Process drift Check ring wear, hydraulic leak ±0.03–0.10mm SPC monitoring, pressure profile check
Operator change Parameter adjustment without approval Unpredictable Locked parameters, change log

The total accumulated drift from all sources: ±0.10–0.25mm for amorphous resins, ±0.15–0.40mm for semi-crystalline resins over 100,000 shots. If the part tolerance is ±0.10mm: drift must be controlled to ±0.05mm (leaving 50% of tolerance for the molding process). If the tolerance is ±0.25mm: drift control is less critical but SPC monitoring is still recommended.

X-bar/R chart — a pair of SPC control charts used to monitor a process over time. The X-bar chart plots the average (mean) of each sample group; the R chart plots the range (max − min) within each group. Control limits are set at ±3σ from the mean. A point outside limits or 7 consecutive points trending in one direction signals a process shift. The tolerance guide covers how SPC data is used to hold tight tolerances in production.

What Three Steps Detect and Correct Dimensional Drift Before Parts Go Out of Tolerance?

  1. Establish a CMM baseline from T1 approved parts — measure 5–10 critical dimensions on 5 parts. The baseline is the mean and standard deviation (σ) for each dimension. Control limits: mean ± 3σ. For production: measure 3 parts every 500–1,000 shots and plot on the X-bar/R chart. If a point falls outside control limits: stop production, investigate the cause (resin change? mold temp shift? gate wear?), correct and resume. If 7 consecutive points trend in one direction: the process is drifting — adjust parameters before the drift reaches the tolerance limit.
  2. Lock molding parameters after T1 approval — any change requires documented approval with before/after CMM data. The most common cause of sudden dimensional shift: an operator adjusts injection speed, holding pressure or melt temperature without measuring the dimensional impact. The parameter change may fix one defect (short shot) but create another (dimensional drift). Lock all parameters in the machine controller (password-protected). Any change requires: (1) documented reason, (2) CMM measurement before and after, (3) approval from the quality engineer. This prevents uncontrolled parameter changes from introducing drift.
  3. Schedule mold maintenance every 25,000–50,000 shots — gate wear is the primary long-term drift source. Over 50,000+ shots: the gate erodes (especially with glass-filled resins), enlarging the gate by 0.05–0.15mm. A larger gate changes the fill balance: resin flows faster, packing pressure drops, and the part shrinks more. Dimensions decrease progressively. The fix: re-machine or replace the gate insert during scheduled maintenance. For abrasive resins (GF-PA, GF-PP): use hardened steel gate inserts (H13 at 50 HRC) to slow erosion. Measure gate dimensions during each maintenance — if gate has enlarged 0.05mm: replace before drift becomes a production issue.

Gate erosion — the gradual enlargement of the gate orifice caused by abrasive resin flow over 50,000+ shots. Glass-filled resins (GF-PA, GF-PP) erode gates fastest — glass fibers (Mohs 6–7) cut into standard steel (P20 at 28 HRC). A 0.05mm gate enlargement changes fill balance, reduces packing effectiveness and causes dimensions to shrink progressively. The gate vestige guide covers how gate geometry also affects part appearance.

Dimensional drift: resin batch ±0.05mm, ambient temp ±0.02–0.05mm, mold wear −0.05–0.15mm per 50k shots. Total: ±0.10–0.25mm (amorphous), ±0.15–0.40mm (semi-crystalline). Control: CMM baseline + SPC + locked parameters + scheduled mold maintenance.

  • CMM baseline from T1 parts — 5–10 critical dims on 5 parts, mean and σ recorded.
  • SPC sampling every 500–1,000 shots — 3 parts measured, plotted on X-bar/R chart.
  • Parameters locked after T1 — password-protected, change log with CMM before/after.
  • Mold maintenance every 25–50k shots — gate dimensions checked, vents cleaned.
  • Resin from single supplier and grade — switching introduces shrinkage variation.

Frequently Asked Questions

How do I know if dimensional drift is from resin, mold or process?

Three diagnostic tests: (1) Resin test — switch to a different resin lot and measure. If dimensions shift: the original lot had different shrinkage. (2) Mold test — measure the gate dimensions. If gate is enlarged: mold wear is the cause. (3) Process test — check the pressure profile. If peak pressure or hold pressure has changed: process drift (check ring wear, hydraulic leak). The most common cause: resin batch variation (60% of drift cases). The second most common: gate wear (25%). Process drift accounts for the remaining 15%.

What is the acceptable dimensional drift over a production run of 500,000 parts?

For amorphous resins (ABS, PC, PS): total drift ±0.10–0.25mm over 500,000 shots is typical with controlled process. For semi-crystalline (PP, PA, POM): ±0.15–0.40mm. If the part tolerance is ±0.10mm: drift must be limited to ±0.05mm (half the tolerance band). This requires: incoming resin shrinkage check, chiller with ±0.5°C control, and mold maintenance every 25,000 shots. For ±0.25mm tolerance: drift of ±0.10mm is acceptable with standard process control (maintenance every 50,000 shots). The tighter the tolerance, the more aggressive the drift control must be.

Can simulation predict dimensional drift before the mold is built?

Mold flow simulation predicts nominal shrinkage and warpage based on the datasheet shrinkage value — but it cannot predict batch-to-batch variation or long-term mold wear. The simulation is useful for: identifying high-shrinkage areas (thick sections, ribs), optimizing gate location for uniform fill, and setting the nominal cavity dimensions. The actual drift is determined during production and monitored with SPC. For tight-tolerance parts: build 0.05mm extra steel on critical dimensions (safe to remove material, impossible to add). After T1 CMM data: adjust the cavity to the measured mean.

Key Takeaways

  • Drift sources: resin batch ±0.05mm, temp ±0.02–0.05mm, wear −0.05–0.15mm/50k.
  • Total: ±0.10–0.25mm (amorphous), ±0.15–0.40mm (semi-crystalline).
  • Control: CMM baseline + SPC + locked parameters + mold maintenance.
  • Gate wear is #1 long-term drift source. Maintain every 25–50k shots.

Send the part drawing with critical dimensions and tolerance requirements — the review identifies which dimensions are at risk of drift, specifies the SPC sampling plan, and defines the process control parameters to keep production within tolerance over the full mold life.

Sources: Plastics Technology — Dimensional Stability in Injection Molding · Plastics Technology Online — SPC for Injection Molding: Detecting Dimensional Drift · Society of Plastics Engineers (SPE) — Dimensional Control Technical Papers · Quality Magazine — Statistical Process Control for Molded Parts · ThomasNet — Dimensional Inspection and Process Control in Injection Molding

— Arthur, Tooling Engineer at aCatMould