Understanding Tolerances Injection Molding After 22 years of building molds and troubleshooting dimensional issues, I can tell you that tolerances are where most projects either succeed or go sideways.
I’ve seen engineers specify ±0.001” on a part that doesn’t need it,and watched procurement scramble when the quotes come back 3x higher than expected. I’ve also seen critical dimensions get overlooked, leading to parts that look perfect but won’t assemble. what actually matters when it comes to tolerances in injection molding.
The Reality of Plastic Tolerances Here’s what most people don’t understand:
plastic isn’t metal. It shrinks, warps, and moves in ways that steel and aluminum simply don’t. A tolerance that’s easy to hold on a machined part might be nearly impossible,or prohibitively expensive,on a molded one. The general rule of thumb I use: standard commercial tolerances for injection molding are about ±0.005” per inch (±0.127mm per 25mm). That’s your baseline. Anything tighter than that starts adding cost and complexity.
ISO Standards You Need to Know The international standard for plastic tolerances is ISO 20457 (which replaced the older DIN 16742).
It defines tolerance grades various TG9 (coarsest), based on:
- Nominal dimension
- Material shrinkage characteristics
- Part geometry complexity Tolerance Grade Application Relative Cost TG1-TG2 Precision parts, tight fits Very High (4-5x)TG3-TG4 Engineering applications High (2-3x)TG5-TG6 Standard commercial Baseline TG7-TG9 Non-critical dimensions Lower In the US, you’ll also encounter SPI (Society of Plastics Industry) guidelines, which break tolerances into:
- Fine: Achievable with extra process control
- Commercial: Standard production capability
- Coarse: Relaxed tolerances for non-critical has
Material-Specific Tolerance Capabilities This is where it gets interesting.
Different materials behave differently, and your tolerance expectations need to match.
Amorphous vs. Semi-Crystalline Materials Material
| Type | Shrinkage Range | Typical Tolerance CapabilityAmorphous (ABS, PC, PMMA) |
|---|---|---|
| 0.4-0.8%±0.002-0.003”/inSemi-Crystalline (PP, PE, Nylon, POM) | 1.5-3.0%±0.004-0.006”/inFilled Materials (Glass-filled) | 0.2-0.5% (flow) / 0.4-1.0% (cross-flow)±0.002-0.004”/in |
Material-Specific Tolerance Chart
| Material | Shrinkage | Commercial Tolerance | Fine Tolerance |
|---|---|---|---|
| NotesABS0.4-0.7%±0.004”/in±0.002”/in | Predictable, good choice for tight tolerances | Polycarbonate | 0.5-0.7%±0.004”/in±0.002”/in |
| Excellent dimensional stability | Nylon 6/ | 61.5-2.2%±0.006”/in±0.004”/in | Moisture absorption affects dimensions |
| Polypropylene | 1.5-2.5%±0.008”/in±0.005”/in | High shrinkage, harder to controlPOM (Acetal) | 2.0-2.5%±0.006”/in±0.003”/in |
| Consistent but high shrinkage | Glass-filled Nylon | 0.3-0.5%±0.003”/in±0.002”/in | Anisotropic,different in flow vs. cross-flow |
The Shrinkage Factor I can’t overstate this:
shrinkage is the enemy of tight tolerances. When plastic cools, it shrinks,and that shrinkage isn’t always uniform. Factors that affect shrinkage consistency:
- Wall thickness variations , Thicker sections shrink more
- Gate location , Parts shrink toward the gate
- Cooling uniformity , Hot spots = inconsistent shrinkage
- Packing pressure , Under-packed parts shrink more
- Material lot variations , Even the same grade can vary
Real-World Example I had a project with a PP housing that needed a ±0.010” tolerance on a 4” dimension.
Sounds reasonable, right? But PP shrinks about 2%, so we’re starting with 0.080” of shrinkage to manage. Factor in wall thickness variations, and we were chasing our tails for weeks. The solution? We switched to a glass-filled PP with 0.5% shrinkage. Problem solved, and the part actually cost less because we could run faster cycles.
What Affects Achievable Tolerances
Part Design Factors
| Factor | Impact on Tolerance | Recommendation |
|---|---|---|
| Wall thickness uniformity | High | Keep walls within ±10% of nominal |
| Part size | High | Larger = more variation |
| Geometry complexity | Medium | Simplify where possible |
| Gate location | Medium | Gate near critical dimensions |
| Draft angles | Low-Medium | Adequate draft reduces ejection variation |
Process Factors
| Factor | Impact on Tolerance | Control Method | Melt temperature |
|---|---|---|---|
| High±5°F control | Mold temperature | High±3°F control | Injection speed |
| Medium | Consistent fill time | Packing pressure | High |
| Pressure profiling | Cooling time | Medium | Sufficient for full solidification |
Tooling Factors
| Factor | Impact on Tolerance | Specification |
|---|---|---|
| Steel selection | MediumP20 for commercial, H13 for fine | Cooling design |
| High | Uniform cooling within ±5°FVenting | Low-Medium |
| Adequate venting prevents shorts | Tool maintenance | Medium |
Measurement Techniques You can’t control what you can’t measure.
Here’s what I recommend for different tolerance levels:
Measurement Method Selection Tolerance
| Level | Recommended Method | Measurement Uncertainty±0.001” or finerCMM (temperature controlled)±0.0002”±0.002-0.005”CMM or optical comparator±0.0005”±0.005-0.010”Calipers, micrometers, go/no-go gauges±0.001”±0.010” or coarser | Standard inspection tools±0.002” |
|---|
Critical Measurement Considerations
- Temperature matters , Measure at 68°F (20°C) or specify otherwise
- Conditioning time , Parts need 24-48 hours to stabilize
- Moisture content , Nylon parts change dimensions with humidity
- Measurement location , Specify exactly where to measure
- Sample size , Use Cpk studies for critical dimensions (minimum 30 parts)
Tolerance Stack-Up Analysis When parts assemble together, tolerances add up. I use this simple approach:
Worst-case stack-up: Add all tolerances arithmetically
Statistical stack-up (RSS): √(tol₁² + tol₂² + tol₃² + …)
Example: Three-Part Assembly
| Part | Dimension | Tolerance | Part A1.000”±0.003”Part B0.500”±0.004”Part C1.500”±0.005” |
|---|
- Worst case: ±0.012” total variation
- Statistical (RSS): ±0.007” total variation (more realistic)
Cost Impact of Tight Tolerances Here’s the uncomfortable truth about tolerances and cost:
Tolerance Level Cost Multiplier What It Requires Standard (±0.005”/in)1.0x Normal process Tight (±0.003”/in)1.5-2x Process optimization, tighter material specs Very Tight (±0.002”/in)2-3x Scientific molding, SPC, tool modifications Precision (±0.001”/in)3-5x Specialized equipment, controlled environment
Practical Tolerance Specification Checklist Before you finalize your tolerances, run through this:
Design Review: Are tight tolerances only on critical dimensions? Have you considered material shrinkage characteristics? Is wall thickness uniform (±10%)? Are gate locations optimized for critical has?
Material Selection: Is the material appropriate for required tolerances? Have you accounted for moisture absorption (if applicable)? Is filled material needed for dimensional stability? Documentation: Are critical dimensions clearly marked (CTQ)? Is the measurement method specified? Are measurement conditions defined (temperature, humidity)? Is sampling plan established for production? Communication: Has the molder reviewed and agreed to tolerances? Is there a tolerance capability study (Cpk) requirement? Are first article inspection requirements clear?
What I Tell Every New Engineer Start with the loosest tolerance that will work, then tighten only where absolutely necessary.
Every thousandth of an inch you add to a tolerance specification has a cost,in tooling, in process control, in rejection rates, and in your timeline. If you’re not sure whether a tight tolerance is necessary, ask yourself: “What happens if this dimension is at the edge of the tolerance band?” If the answer is “nothing critical,” you probably don’t need that tight tolerance. And when you do need tight tolerances, make sure your molder knows which dimensions are truly critical. I’d rather hit three critical dimensions perfectly than chase five non-critical ones. The best tolerance strategy isn’t about precision everywhere—it’s about precision where it matters.