Design for Injection Molding: The Complete DFM Guide for Engineers Moving from Prototype to Production

A comprehensive engineer's guide to designing parts for injection molding — covering wall thickness uniformity, draft angles, rib and boss design rules,…

·

You've validated your design with a dozen FDM prototypes, fine-tuned the tolerances with SLS nylon, and confirmed the snap-fit engagement force on an MJF print. Now your customer wants 50,000 units. You reach for the phone to call an injection molder — and they send back a markup of your CAD file with 47 redlines.

Wall thickness varies from 1.2 mm to 6.8 mm in a single part. There are zero draft angles on any vertical face. The boss is 3× taller than the nominal wall and will sink. The undercut that took you 15 minutes to model will require a 12,000 side action. And you spec'd a surface finish that only a 180,000 hardened tool steel mold can achieve.

This is the valley of death between prototyping and production. 3D printing forgives nearly everything. Injection molding forgives nothing.

This guide is the bridge. It covers every DFM rule that separates a prototype from a production-ready injection molded part — with the equations, tables, and cost data you need to get it right on the first tooling iteration.

1. How Injection Molding Actually Works

Before diving into design rules, you need to understand what happens inside the machine. An injection molding machine has three subsystems: the injection unit, the mold (tool), and the clamping unit.

The Machine

Thermoplastic pellets are gravity-fed from a hopper into a heated barrel containing a reciprocating screw. The barrel heats the material to its melt temperature — typically 180°C to 350°C depending on the resin. The screw rotates to convey material forward, and the mechanical shear provides additional frictional heating (\dot{Q}_{\text{shear}} \approx \eta \dot{\gamma}^2 where \eta is melt viscosity and \dot{\gamma} is shear rate). This is why screw design matters: the compression ratio (typically 2:1 to 3:1 for general-purpose screws) controls how much work is imparted to the melt.

Material accumulates in front of the screw as a shot — the volume needed to fill the cavity, compensate for shrinkage, and maintain a small cushion (~5–10 mm of screw travel) to prevent the screw from bottoming out. When the shot is ready, the screw acts as a plunger, injecting the melt at high velocity (typically 50–200 mm/s) through the nozzle, sprue, runner system, and gate into the cavity.

The Injection Cycle

The cycle has four phases, each with its own physics:

1. Filling (velocity-controlled). The screw moves forward at constant velocity. Melt fills approximately 95–98% of the cavity. This phase is shear-rate dominated — the non-Newtonian viscosity of polymer melts means the relationship between pressure drop and flow rate is nonlinear. The governing equation for flow in a thin cavity of thickness h is the Hele-Shaw approximation:

In practice, this means that doubling the wall thickness reduces filling pressure by a factor of 8 (h^3 dependence). This is the single most important number in injection molding DFM.

2. Packing (pressure-controlled). At a transfer position (typically 95–98% fill by volume), the machine switches from velocity control to pressure control. Packing pressure — typically 50–80% of the injection pressure — forces additional melt into the cavity to compensate for volumetric shrinkage as the polymer cools. Packing continues until the gate freezes (solidifies). Insufficient packing → sink marks and voids. Excessive packing → flash, warpage, and molded-in stress.

3. Cooling. Once the gate solidifies, no more material can enter. The part continues cooling inside the mold until it reaches a temperature below the heat deflection temperature (HDT) of the material — typically the ejection temperature. Cooling accounts for 60–80% of the total cycle time. Cooling time scales with the square of wall thickness:

where \alpha is thermal diffusivity (~0.1 mm²/s for most polymers), h is wall thickness, and the temperatures are melt, mold, and ejection. For a 2 mm ABS wall at 230°C melt/50°C mold/90°C ejection: t_{\text{cool}} \approx 8–10 seconds. For a 4 mm wall: 32–40 seconds. This quadratic relationship is why uniform thin walls aren't just about quality — they're about economics.

4. Ejection. Mold opens, ejector pins push the part off the core. Draft angles are critical here — without them, ejection forces spike, and the part may stick, warp, or drag.

2. Material Selection: The Foundation of Every DFM Decision

You cannot make DFM decisions without first choosing a material. Every rule — wall thickness, draft angle, shrinkage allowance — depends on the resin.

Shrinkage Is Everything

When a thermoplastic cools from melt temperature to room temperature, it shrinks. The volumetric shrinkage is primarily a function of the material's semicrystalline or amorphous nature:

Material · Type · Mold Shrinkage (%) · Typical Wall Range (mm) · Melt Temp (°C) · Mold Temp (°C)

ABS · Amorphous · 0.4–0.7 · 1.0–3.5 · 200–260 · 40–80

Polycarbonate (PC) · Amorphous · 0.5–0.7 · 1.0–3.0 · 280–320 · 80–120

PC/ABS blend · Amorphous · 0.5–0.7 · 1.2–3.5 · 240–280 · 60–100

Nylon 6 (PA6) · Semicrystalline · 1.0–1.5 · 0.8–3.0 · 230–270 · 60–90

Nylon 66 (PA66) · Semicrystalline · 1.5–2.0 · 0.8–3.0 · 260–290 · 60–100

30% GF Nylon 66 · Semicrystalline · 0.3–0.7 · 1.5–4.0 · 270–300 · 80–120

Polypropylene (PP) · Semicrystalline · 1.0–2.5 · 0.8–3.5 · 200–250 · 20–60

20% Talc-filled PP · Semicrystalline · 0.8–1.2 · 1.2–3.5 · 210–260 · 30–60

POM (Acetal/Delrin) · Semicrystalline · 1.8–2.5 · 0.8–3.0 · 180–220 · 60–120

PBT · Semicrystalline · 1.5–2.2 · 1.0–3.0 · 230–260 · 40–80

30% GF PBT · Semicrystalline · 0.3–0.8 · 1.5–4.0 · 240–270 · 60–100

HDPE · Semicrystalline · 1.5–4.0 · 0.8–3.5 · 180–240 · 20–40

PMMA (Acrylic) · Amorphous · 0.2–0.8 · 1.0–3.5 · 220–250 · 50–90

PS (Polystyrene) · Amorphous · 0.4–0.7 · 1.0–3.5 · 180–260 · 20–60

TPE/TPU · Elastomeric · 1.0–3.0 · 0.5–3.0 · 170–230 · 20–50

Key rule: Amorphous materials shrink less (~0.5%) and more isotropically. Semicrystalline materials shrink more (1–4%) and anisotropically — more in the flow direction than transverse. Unfilled semicrystalline materials are difficult to mold to tight tolerances. Glass fiber filling reduces shrinkage dramatically (0.3–0.8%) but introduces warpage anisotropy — fibers align with flow, creating differential shrinkage parallel vs. perpendicular to flow.

Melt Flow Index (MFI)

MFI (g/10 min at specified temperature and load) measures how easily a polymer flows. For injection molding:

For thin-wall parts (<1.5 mm), use high-flow grades (MFI > 15). For structural parts where strength matters more, use medium-flow grades and design walls accordingly.

3. Wall Thickness: The Single Most Important Rule

Uniform wall thickness is the First Commandment of injection molding DFM. Every other rule is downstream of this one.

Why Uniformity Matters

When walls vary in thickness, they cool at different rates. The thick section is still molten and shrinking while the thin section has already solidified. This creates:

The Golden Rule for Ribs and Bosses

Every projecting feature — ribs, bosses, gussets — should be:

For bosses (threaded inserts, screw bosses):

If a boss wall would exceed 60% of nominal wall thickness, you must core it out from the back side. A solid boss on a 2 mm wall will sink like the Mariana Trench.

Nominal Wall Thickness Guidelines

Material · Min Wall (mm) · Typical (mm) · Max Recommended (mm)

ABS · 0.8 · 1.5–2.5 · 3.5

PC · 0.8 · 1.5–2.5 · 3.5

PC/ABS · 0.8 · 1.5–2.5 · 3.5

Nylon (unfilled) · 0.6 · 1.0–2.5 · 3.5

Nylon (30% GF) · 1.0 · 1.5–3.0 · 4.5

PP (unfilled) · 0.6 · 1.0–2.5 · 4.0

POM · 0.5 · 1.0–2.5 · 3.5

PMMA · 0.8 · 1.5–3.0 · 4.0

HDPE · 0.6 · 1.0–2.5 · 4.0

TPE/TPU · 0.4 · 0.8–2.0 · 3.0

Thin-wall injection molding (wall < 1.0 mm) is a specialized process requiring high-speed machines, hot runner systems, and high injection pressures (up to 200 MPa). It's used for packaging, phone cases, and connector housings. Don't attempt it with a general-purpose molder and a commodity resin.

Flow Length to Thickness Ratio

Every material has a maximum flow length it can achieve at a given wall thickness before the melt front freezes. The flow length ratio (L/t) is:

Typical L/t ratios for various materials:

This means: with 2 mm wall ABS, the gate can be at most ~300–400 mm from the farthest cavity edge. If your part is larger, you need multiple gates or a thicker wall.

4. Draft Angles: The Price of Verticality

Every surface parallel to the mold opening direction must be tapered — drafted — so the part can release from the mold without dragging, scoring, or sticking.

Physics of Ejection

The ejection force required to push a part off the core is:

where \mu is the coefficient of friction between polymer and steel (typically 0.2–0.5 for most thermoplastics on polished tool steel), and P_{\text{shrink}} is the shrinkage-induced contact pressure. Without draft, the part shrinks onto the core like a compression fitting — and every micron of surface roughness becomes a mechanical interlock.

With draft angle \theta, the ejection force required along the mold opening axis drops to:

At \theta = 1^\circ: the reduction is subtle but measurable. At \theta = 3^\circ: ejection force drops by ~20–30% compared to zero draft.

Minimum Draft Angle Guidelines

Surface Type · Minimum Draft · Recommended Draft

Outer walls (cavity side) · 0.5° · 1.0–2.0°

Inner walls (core side) · 1.0° · 2.0–3.0°

Textured surfaces · 1.0° + 1.0° per 0.025 mm texture depth · See mold-tech specs

Deep ribs (h > 5t) · 0.5° per side minimum · 1.0–1.5° per side

Bosses (ID) · 0.25° · 0.5–1.0°

Shut-off surfaces · 3.0–5.0° · 5.0°+

Why inner walls need more draft: The part shrinks onto the core as it cools — the plastic contracts, gripping the core tighter on inner surfaces. Outer surfaces shrink away from the cavity wall, so less draft is needed.

Draft and Surface Texture

Textured surfaces require additional draft because the texture creates micro-undercuts. Mold-Tech (the industry standard for chemical etching textures) provides draft angle recommendations for each texture spec:

Mold-Tech Spec · Texture Depth (mm) · Additional Draft Required

MT-11000 (fine matte) · 0.025 · +1.0°

MT-11010 (light texture) · 0.04 · +1.5°

MT-11020 (medium texture) · 0.08 · +3.0°

MT-11030 (coarse) · 0.12 · +5.0°

MT-11040 (leather grain) · 0.15 · +7.0°

A textured surface on a vertical wall without sufficient draft will result in drag marks — visible scratches running parallel to the ejection direction. These can't be polished out; the mold must be re-cut.

5. Ribs, Bosses, and Gussets: Structural Design in Plastic

Plastic parts get their stiffness from geometry, not material thickness. A flat plate's bending stiffness is proportional to t^3 — doubling the thickness gives 8× the stiffness. But doubling wall thickness also doubles the cooling time, creates sink marks, and wastes material.

The correct solution: thin walls + ribs.

Rib Design Rules

A rib's bending stiffness contribution is proportional to its height cubed (h^3). A rib that's 3× the wall thickness adds approximately 27× the bending stiffness of the wall alone — with a fraction of the material.

Correct rib geometry:

    ← t_wall →
    ┌─────────┐
    │         │  ↑
    │         │  h_rib ≤ 3 × t_wall
    │    ↑    │  ↓
    │  t_rib  │
    │    ↓    │
    └───╮ ╭───┘
        ╰─╯ R_base ≥ 0.25 × t_wall
    ← t_rib ≤ 0.6 × t_wall →

The five rules of rib design:

  1. t_{\text{rib}} \leq 0.6 \times t_{\text{wall}} — thicker ribs cause sink on the opposite face
  2. h_{\text{rib}} \leq 3 \times t_{\text{wall}} — taller ribs are difficult to fill and eject
  3. R_{\text{base}} \geq 0.25 \times t_{\text{wall}} — radius the rib-to-wall junction to reduce stress concentration and improve flow
  4. Draft both sides — minimum 0.5° per side, 1.0° preferred
  5. Spacing: rib spacing should be at least 2 \times t_{\text{wall}} to avoid creating a locally thickened section

Boss Design for Self-Tapping Screws

Bosses are the most common defect source in injection molded parts. A badly designed boss will sink, short-shot, or crack during screw insertion.

For a standard M3 self-tapping screw into ABS:

If the boss is taller than 3× the wall, add gusset ribs (3 or 4, equally spaced) connecting the boss to the base wall. Gusset rib thickness ≤ 0.4 × t_wall.

For threaded inserts (heat-set or ultrasonic):

Gussets

Gussets support tall ribs or free-standing walls. They're triangular reinforcements that prevent buckling during ejection:

6. Undercuts and Side Actions: The $10,000 Feature

An undercut is any feature that prevents the part from being ejected straight out of a simple two-plate mold — a hole perpendicular to the mold opening axis, a snap-fit tab, a side window, an external thread.

Internal vs. External Undercuts

External undercuts (on the outside of the part) can sometimes be resolved with side actions — sliding cores that retract before the mold opens. Side actions are actuated by angle pins (cams), hydraulic cylinders, or springs.

Internal undercuts (inside the part, like an internal snap-fit latch) require lifters — angled ejector pins that move laterally as they push the part off the core.

Cost Hierarchy

Solution · Typical Added Cost/Tool · Cycle Time Impact

Parting line reorientation · $0 (design change) · None

Pass-through core (shut-off) · $0–500 · None

Bump-off (flexible material) · $0 · None (requires TPU/TPE)

Lifter (internal undercut) · $1,000–3,000 · +0.5–1.0s

Side action — angle pin · $2,000–5,000 · +1–2s

Side action — hydraulic · $5,000–12,000 · +2–4s

Collapsible core · $8,000–20,000 · +1–2s

Unscrewing mechanism (threads) · $10,000–25,000 · +2–5s

Before you add a side action to your design, ask: Can I reorient this feature so it's in the line of draw? Can I use a pass-through (hole through both walls) instead of a blind pocket? Can the feature be a separate part that snaps or welds on?

The 5 mm Rule for Lifter Travel

A lifter's lateral travel is limited by the ejector stroke. For a typical mold with 50 mm ejector stroke and a lifter angled at 10°:

Subtract clearance and you typically get 5–7 mm of usable lateral travel. If your undercut is deeper than that, you need a side action.

7. Gate Types and Placement: Where the Plastic Enters

The gate is the small orifice through which melt enters the cavity. Its type, size, and location control filling pattern, knit line formation, gas entrapment, and the cosmetic appearance of the final part.

Gate Types (Ranked by Cost and Complexity)

Gate Type · Vestige · Pros · Cons · Cost

Edge gate · Small nub at parting line · Simple, works for most parts · Manual trimming or auto-degating, visible mark · $0

Tab gate · Tab extending from part edge · Redirects jetting, better for thick parts · Requires secondary trimming · $0–500

Fan gate · Wide, flat vestige · Excellent for flat parts, reduced warpage · Larger vestige, trimming required · $0–500

Pin/point gate (3-plate mold) · Small dot on surface · Auto-degated, small vestige · Requires 3-plate mold (+5,000–10,000) · +5,000–10,000

Submarine (tunnel) gate · Small nub on side wall · Auto-degated, can gate below parting line · Limited to certain geometries · $500–1,000

Hot tip gate (hot runner) · Tiny vestige, no runner scrap · Ideal for high cavitation, no runner waste · 2,000–5,000 per drop, requires temperature control · +2,000–5,000/drop

Valve gate (hot runner) · Minimal, flush · Sequential filling, no drool, premium cosmetics · 5,000–10,000 per drop, complex controls · +5,000–10,000/drop

Diaphragm/disk gate · Ring on ID or OD · Excellent for cylindrical parts (gears, bushings) · Trimming required, concentricity critical · $500–2,000

Gate Placement Rules

  1. Gate into the thickest section. This ensures the gate remains molten long enough to pack the cavity. Gating into a thin section that freezes before the thick section means zero packing pressure reaches the thick area → sink marks.
  1. Avoid gating near stress concentrations. The gate leaves a weld line and a material discontinuity. Don't put it at the root of a snap-fit beam or near a hole that sees hoop stress.
  1. Consider knit lines. When two flow fronts meet, they form a knit (weld) line — a mechanically weak zone where polymer chains haven't fully entangled across the interface. Knit line strength is typically 60–80% of bulk material strength. Place the gate so that knit lines form in low-stress areas, not at the base of a snap-fit or a load-bearing boss.
  1. Cosmetics matter. The gate vestige is permanent. On a consumer product's A-surface, a gate mark is unacceptable. Use a tunnel gate (auto-degated, hidden below the visible surface) or valve gate (near-invisible vestige).
  1. Flow balance. In multi-cavity molds, all cavities must fill simultaneously. Unbalanced runner systems cause some cavities to pack while others are still filling → inconsistent part quality. Naturally balanced runner layouts (H-pattern, radial) are preferred; artificially balanced runners (restrictor plugs) are a compromise.

8. Ejector Pin Marks: Hidden in Plain Sight

Ejector pins push the part off the core after cooling. They leave witness marks — circular depressions or protrusions typically 0.05–0.15 mm raised (if worn) or recessed.

Design Rules

Alternatives to Ejector Pins

9. Tolerances: What You Can Actually Hold

Injection molding can't hold machining tolerances. The process involves molten plastic shrinking at different rates in different directions on a tool that thermally expands during every cycle. Anyone who tells you ±0.05 mm on a 100 mm dimension in unfilled PP hasn't run production.

DIN 16901 / ISO 20457 Tolerance Standards

The industry-standard tolerance table gives achievable tolerances as a function of material type and dimension. Here are commercial-grade tolerances (not precision):

Nominal Size (mm) · ABS/PC/PS (Amorphous) · PA (Unfilled, Semicrystalline) · PA GF30 (Filled) · PP (Unfilled)

1–3 · ±0.07 · ±0.10 · ±0.07 · ±0.12

3–6 · ±0.10 · ±0.14 · ±0.10 · ±0.16

6–10 · ±0.12 · ±0.16 · ±0.12 · ±0.20

10–18 · ±0.15 · ±0.20 · ±0.15 · ±0.25

18–30 · ±0.18 · ±0.24 · ±0.18 · ±0.30

30–50 · ±0.22 · ±0.30 · ±0.22 · ±0.38

50–80 · ±0.28 · ±0.38 · ±0.28 · ±0.48

80–120 · ±0.35 · ±0.48 · ±0.35 · ±0.60

120–180 · ±0.45 · ±0.60 · ±0.45 · ±0.78

180–250 · ±0.55 · ±0.75 · ±0.55 · ±0.96

Key insights from this table:

  1. Amorphous materials (ABS, PC) hold tighter tolerances than unfilled semicrystalline materials (PP, PA unfilled). This is the most important tolerance rule. If you need precision, use an amorphous resin or a glass-filled semicrystalline resin.
  1. Glass fiber filling (GF30) brings semicrystalline tolerance bands down to amorphous levels. The fibers constrain shrinkage.
  1. Tolerances do NOT scale linearly with dimension. The tolerance band grows roughly as \sqrt{\text{dimension}} rather than linearly — larger parts are proportionally more precise than small parts.
  1. Mold tolerances are separate. The mold itself is machined to ±0.01–0.02 mm (CNC) or ±0.005 mm (EDM/grinding). But the plastic shrinks unpredictably on top of that.

Precision vs. Commercial Molding

Grade · Cost Premium · Achievable Tolerance · Requirements

Commercial · 1× (baseline) · DIN 16901 "normal" · Standard mold, standard process control

Technical · 1.3–1.5× · DIN 16901 "close" (~50% tighter) · Tighter mold tolerances, tighter process control

Precision · 2–3× · DIN 16901 "precision" (~70% tighter than normal) · Hardened tool steel, climate-controlled molding room, SPC monitoring, stricter resin lot control

Critical-to-function (CTF) dimensions: Mark only the dimensions that actually matter. If you dimension every feature with ±0.05 mm, the molder will either quote you 3× the price or reject the job entirely. Typically, only 5–20% of dimensions on a part should be CTF.

10. Mold Tooling Economics

The mold (tool) is the biggest upfront investment in injection molding. Understanding what drives tool cost helps you design cost-effective parts.

Mold Base and Cavity Count

A mold consists of:

Cost by Complexity (Single Cavity, Production Mold)

Complexity · Typical Cost (India, 2026) · Typical Cost (China, 2026) · Typical Cost (US/Europe, 2026) · Lead Time

Simple (2-plate, no side actions, no texture) · ₹1.5–4L (1,800–4,800) · 2,000–5,000 · 5,000–15,000 · 4–6 weeks

Moderate (1–2 side actions, basic texture) · ₹4–8L (4,800–9,600) · 5,000–12,000 · 15,000–35,000 · 6–10 weeks

Complex (3+ side actions, hot runner, precision) · ₹8–20L (9,600–24,000) · 12,000–30,000 · 35,000–80,000 · 10–16 weeks

High-cavitation (8–16 cavity, hot runner, hardened) · ₹15–40L (18,000–48,000) · 20,000–60,000 · 60,000–150,000 · 12–20 weeks

Multi-cavity molds reduce per-part cost but increase upfront investment. The breakeven between a 1-cavity and a 4-cavity mold depends on volume:

If a single cavity produces 1 part per 30s cycle, that's 120 parts/hour, ~86,400 parts/month (24/7). A 4-cavity mold produces 4× that. If the 1-cavity tool costs ₹3L (3,600) and the 4-cavity costs ₹8L (9,600), the incremental ₹5L ($6,000) must be amortized over the marginal 260,000 parts/month capacity. At a processing cost savings of ₹2/part, breakeven is 3,000 parts — almost always worth it for any volume above ~5,000 parts.

Tool Life by Steel Type

Steel Type · Hardness (HRC) · Typical Life (shots) · Cost Factor

Aluminum (7075) · ~60 HB (soft) · 1,000–10,000 · 0.3–0.5×

P-20 (pre-hardened) · 28–32 · 100,000–500,000 · 1.0× (baseline)

H-13 (hot work, hardened) · 48–52 · 500,000–1,000,000+ · 1.5–2.0×

S-7 · 54–56 · 1,000,000+ · 2.0–2.5×

420 SS · 48–52 · 500,000–1,000,000+ · 2.0–3.0×

Aluminum molds are appropriate for:

Hardened steel (H-13/S-7) is required for:

11. Surface Finishes: SPI and VDI Standards

The mold surface finish directly transfers to the part (inverse for cavity surfaces). Finishes are classified by the SPI (Society of the Plastics Industry) standard in the US and VDI 3400 in Europe.

SPI Finish Standards

SPI Grade · Finish · Ra (μm) · Method · Typical Application

A-1 · #3 Diamond buff · 0.012–0.025 · Diamond paste, rotary buff · Optical lenses, mirrors, transparent parts

A-2 · #6 Diamond buff · 0.025–0.05 · Diamond paste, rotary buff · High-gloss cosmetic surfaces

A-3 · #15 Diamond buff · 0.05–0.10 · Diamond paste, rotary buff · Glossy consumer products

B-1 · 600 grit paper · 0.05–0.10 · Wet sanding · Semi-gloss surfaces

B-2 · 400 grit paper · 0.10–0.15 · Wet sanding · Medium finish

B-3 · 320 grit paper · 0.15–0.30 · Wet sanding · Matte-satin finish

C-1 · 600 stone · 0.30–0.50 · Stoning · Low-gloss, provides tooth for paint

C-2 · 400 stone · 0.50–0.80 · Stoning · Functional surfaces

C-3 · 320 stone · 0.80–1.20 · Stoning · Hidden/internal surfaces

D-1 · Glass bead blast · 1.0–3.0 · Pressure blasting · Textured matte

D-2 · #240 oxide blast · 2.0–5.0 · Pressure blasting · Heavy texture

D-3 · #80 oxide blast · 5.0–10.0 · Pressure blasting · Very rough texture

Cost Impact

Design note: High-gloss SPI A-1/A-2 finishes are merciless. They show every sink mark, flow line, and knit line. If you spec A-1, your wall thickness control and rib design must be flawless.

12. Common Injection Molding Defects: Diagnosis and Fixes

When the first samples come back from the molder looking wrong, here's how to diagnose what went wrong:

Defect · Appearance · Root Cause · Design Fix · Process Fix

Sink marks · Localized depressions on surface · Thick section shrinks more than thin section during cooling · Reduce rib/boss thickness to ≤60% of nominal wall; core out thick sections · Increase packing pressure/time; reduce melt temperature

Voids · Internal bubbles, visible on x-ray or cross-section · Volumetric shrinkage in thick section with rigid skin · Same as sink marks — thin out thick sections · Increase packing pressure; gate into thick section

Warpage · Part twists or bends after ejection · Differential shrinkage due to uneven cooling, fiber orientation, or wall thickness variation · Uniform wall thickness; add ribs for flatness; symmetric gating · Adjust cooling line placement; reduce mold temperature differential

Flash · Thin film of plastic at parting line · Insufficient clamp force; worn mold; overpacking · Increase draft on shut-off surfaces; clean parting line geometry · Reduce injection/packing pressure; increase clamp force; check mold alignment

Short shot · Incomplete part — plastic didn't reach end of cavity · Flow front froze before filling cavity; entrapped air; insufficient venting · Increase wall thickness; add flow leaders; move gate closer to thin areas · Increase melt/mold temperature; increase injection speed; add vents

Knit/weld lines · Visible line where two flow fronts met · Flow fronts cool before merging; air trapped at knit · Move gate to redirect knit line to non-structural area; increase wall thickness at knit zone · Increase melt temperature; increase injection speed; add vent at knit location

Burn marks · Brown/black discoloration, often at end of fill · Diesel effect: trapped air compresses and ignites; resin degradation · Add vents at last-to-fill locations; reduce injection speed in problem zones · Reduce injection speed; add/clean vents; reduce melt temperature

Jetting · Snake-like wavy pattern from gate · Melt shoots across cavity instead of filling progressively · Gate into an impingement wall (redirect flow); use tab/fan gate · Reduce injection speed; increase gate size

Splay (silver streaks) · Silver/white streaks on surface · Moisture in resin; thermal degradation; shear heating · — (material issue) · Dry resin per manufacturer spec; reduce melt temperature; reduce screw RPM

Delamination · Surface layers separate/flake · Incompatible material contamination; excessive mold release · — (process/contamination) · Clean hopper/thorough purge; reduce mold release spray

Blush/gate bloom · Hazy/white patch near gate · Material shears too fast through small gate; melt fracture · Increase gate size; smooth gate land · Reduce injection speed; increase melt temperature

The Most Common Mistake

The #1 call molders get from designers holding rejected samples: "Can you just adjust the process to fix the sink marks?"

Answer: Not if the root cause is a 4 mm thick boss on a 2 mm wall. You can increase packing pressure, but if the gate freezes before the thick section is fully packed, no amount of pressure does anything. Design the part correctly; the process can only compensate for small deviations.

13. When NOT to Use Injection Molding

Injection molding is incredible — for the right application. It's also the wrong process for many situations where alternatives are faster, cheaper, or better:

Scenario · Better Alternative · Why

Quantity < 500 parts · 3D printing (MJF, SLS), CNC machining, urethane casting · Tooling cost per part too high. Breakeven for injection molding is usually 500–2,000 parts

Quantity 500–5,000 parts · Urethane vacuum casting (silicone mold) · Tooling cost ($500–5,000), 1–4 week lead time, material properties close to injection

Design still evolving · 3D printing for functional prototypes, then bridge tooling (aluminum mold) when nearly final · Don't cut steel until the design is frozen

Very large parts (>1 m) · Thermoforming, rotomolding, structural foam · Clamp tonnage requirements become astronomical. Injection molding machines top out at ~9,000 tons

Thick, chunky parts (>6 mm wall) · CNC machining from billet plastic, compression molding, casting · Cooling time becomes uneconomical; sink and void problems are severe

Low annual volume, high variety · 3D printing (MJF for nylon, SLS for TPU), CNC · No tooling investment per SKU

Solid, prismatic parts · CNC machining · If a part has no thin walls, no complex internal geometry, and doesn't need the specific surface of a molded part, machining from extruded stock is often cheaper

The Urethane Casting Bridge

For quantities between prototyping and production (50–5,000 parts), urethane vacuum casting is the unsung hero of product development:

  1. A master pattern is 3D printed (SLA or MJF)
  2. A silicone mold is poured around the master
  3. The silicone mold is used to cast 20–50 urethane parts before degrading
  4. Multiple silicone molds can be made from the same master

Cost: $500–5,000 per silicone mold. Lead time: 1–2 weeks (including pattern printing). Material options: rigid urethanes from Shore 60D to 90D (similar to ABS/PC/PP), flexible from Shore 20A to 80A, clear, colored, and glass-filled.

This is how most hardware startups bridge from prototype to first production run. It's also how you validate that your DFM-optimized design actually works before spending ₹5L on a steel mold.

14. The DFM Checklist: Before You Send to the Molder

Run through this checklist before emailing your STEP file:

The Physics You Can't Ignore

Injection molding operates at the intersection of non-Newtonian fluid mechanics, heat transfer, polymer physics, and production economics. The rules in this guide exist because the physics is unforgiving:

Understand these physics and the rules become obvious. Design for them from the start, and you'll have a mold that runs thousands of cycles without intervention, producing parts that meet spec every single shot.

Ignore them, and you'll have a $12,000 paperweight full of burned ABS and broken ejector pins.

Ready to get your injection molded parts made? FabFlow connects you with verified Indian manufacturers who specialize in injection molding — from single-cavity prototype tools to high-cavitation production molds. Post your job on FabFlow and get quotes from manufacturers who understand DFM.

More FabFlow blog posts