Plastic Part Design: 20 Rules That Avoid Tooling Rework

Tooling rework is one of the most expensive surprises in product development. Once steel has been cut for an injection mold, even a straightforward design correction — a boss that's too thick, a wall that won't fill, a snap-fit that can't eject — can mean days of delay and thousands of dollars in modifications. In the worst cases, it means scrapping the tool entirely.

The good news is that the vast majority of tooling rework is avoidable. Every mold modification triggered by a design flaw traces back to a decision made at the CAD stage — a decision that could have been made differently. These 20 plastic part design rules exist precisely to prevent that. They reflect the core principles of Design for Manufacturability (DFM) as they apply to injection-molded plastic parts, covering everything from wall thickness and draft angles to weld line management, snap-fit geometry, and pre-tooling validation. Apply them before a single line of steel is machined, and you protect both your schedule and your budget.

Why Tooling Rework Is So Expensive

The economics of injection mold correction are asymmetric in a way that catches many product teams off guard. Fixing a design flaw in CAD takes minutes or hours. Fixing the same flaw after a mold has been machined involves recutting steel, re-fitting mold components, re-running trial shots, and re-inspecting the parts — a process that typically costs between $1,000 and $10,000 depending on complexity, and adds days or weeks to your launch schedule.

There is also a directional constraint that compounds the problem. In injection molding, removing steel from a mold (which makes a cavity larger and a wall thinner) is relatively straightforward machining work. Adding steel — which means shrinking a cavity to thicken a wall or correct a dimension — requires welding in an insert, which is expensive, time-consuming, and can affect surface quality. This is why engineers refer to designing "steel-safe": erring on the side that makes the easier correction if a change is needed. The 20 rules below are built around this logic. Each one describes a design decision that, made correctly, either eliminates the need for tooling correction or ensures that any minor adjustment is the easier, cheaper kind.

The 20 Plastic Part Design Rules

Rule 1: Maintain Uniform Wall Thickness

Uniform wall thickness is the single most impactful design rule in injection molding. When plastic enters a mold as a hot, pressurized fluid, sections with different thicknesses cool at different rates. Thicker areas solidify more slowly, and as they contract, they pull the surrounding material inward — producing visible sink marks, internal voids, or warped geometry that falls outside dimensional tolerance. The practical solution is to maintain a consistent nominal wall thickness across the part. Where thick sections are functionally necessary, hollow them out using coring: remove material from the interior to bring the effective wall back to the nominal thickness without sacrificing structural performance.

Rule 2: Respect Material-Specific Wall Thickness Ranges

Uniform thickness is the goal, but the target value depends on the resin you've selected. Every plastic material has a recommended wall thickness range that balances fill reliability against cooling time and mechanical performance. Walls below the minimum for a given material risk incomplete fill — a "short shot" where the plastic doesn't reach the far end of the cavity. Walls above the maximum increase cycle time and material consumption, and they amplify the risk of sink and void defects. For most engineering thermoplastics, the recommended nominal wall falls between 1.5 mm and 3.0 mm, but this varies significantly: polycarbonate runs well at 1.0–3.8 mm, while materials like PEEK can be successfully molded as thin as 0.5 mm. Confirm the range for your specific resin before finalizing geometry.

Rule 3: Use Smooth Thickness Transitions

Even when you can't maintain perfectly uniform wall thickness across an entire part, abrupt changes between thick and thin sections cause the same differential cooling problems as consistently non-uniform walls. Wherever the design requires a thickness change, blend the transition using a chamfer or a fillet rather than a sharp step. A gradual ramp stretches the transition zone over a longer distance, giving the plastic time to equalize pressure and temperature. This reduces the residual stress that builds up at abrupt intersections and that can cause parts to crack, distort, or fail in service long after they appear dimensionally correct.

Rule 4: Apply Draft Angles to Every Vertical Wall

Draft is a slight taper applied to any surface that runs parallel to the mold's draw direction — the path along which the two mold halves separate during ejection. Without sufficient draft, the part grips the mold steel as it cools and contracts, creating friction that drags, scratches, or distorts the part during ejection. At minimum, apply 1 degree of draft per side on smooth-surfaced walls. Internal features like ribs and boss walls typically need at least 0.5 degrees per side. The mold construction also determines where draft should be directed: features should draft toward the ejector side of the tool so the part stays on the half that carries the ejector pins rather than sticking in the cavity.

Rule 5: Scale Draft with Texture Depth

Textured surfaces require significantly more draft than smooth ones because the texture on the mold steel mechanically interlocks with the solidified plastic. If the draft angle isn't sufficient to break this grip before the texture scratches the part surface, the result is dragging marks, torn texture, or a part that simply won't eject reliably. The industry guideline is to add approximately 1.5 degrees of additional draft for every 0.025 mm (0.001 inch) of texture depth. A light texture at 0.025 mm depth therefore requires around 3 degrees total; a heavy texture at 0.075 mm depth requires 5 degrees or more. Specify your surface finish requirement before finalizing draft — don't adjust it at the end of the design process as an afterthought.

Rule 6: Design Ribs to 50–60% of Nominal Wall Thickness

Ribs are one of the most efficient ways to add stiffness and load-bearing capacity to a thin-walled plastic part without adding significant weight or material cost. But a rib that is too thick becomes its own source of sink marks: the extra volume cools more slowly than the adjacent nominal wall, pulling material inward on the opposite face. Keeping rib thickness between 50% and 60% of the nominal wall (often written as 0.5T to 0.6T) prevents this. Add a fillet at the base of each rib with a radius of approximately 0.25T to 0.5T to improve plastic flow into the rib during filling and to reduce stress concentration at the root. Both measures together significantly reduce the chance of needing tooling corrections after the first trial shots.

Rule 7: Keep Rib Height Under 2.5x Wall Thickness

Tall ribs create molding challenges independent of their thickness. A rib that extends more than 2.5 times the nominal wall thickness is difficult to fill uniformly because the thin cross-section loses heat quickly as plastic travels to its tip. It is also structurally vulnerable to ejector pin loads and more likely to stick in the mold. Where the design requires more stiffness than a single shallow rib can provide, use multiple shorter ribs spaced apart rather than one tall rib. This distributes the structural load, maintains fillability, and simplifies the tooling. Apply at least 0.5 degrees of draft per side to all rib walls.

Rule 8: Size Boss Walls at No More Than 60% of Nominal Thickness

Bosses — the cylindrical features that accept screws, threaded inserts, or locating pins — concentrate material at a point on the part, making them a common source of sink marks and void defects. The outer wall of a boss should not exceed 60% of the surrounding nominal wall thickness. If the boss must be taller than the nominal wall, support it with ribs that connect it to the nearest wall rather than letting it stand unsupported, which would require excessive material. When a boss is designed to accept a self-tapping or machine screw, verify that the inner diameter is appropriate for the fastener: if the hole is too small, the boss will crack during assembly. Design for adequate engagement length from the outset rather than modifying the tool after assembly testing reveals the problem.

Rule 9: Replace Sharp Corners with Radiused Fillets

Sharp internal corners in a plastic part are stress concentrators. During and after molding, the material tries to shrink uniformly, but a sharp corner resists this, building residual stress that can cause the part to crack or deform under load. Sharp corners on the mold tool also require electrical discharge machining (EDM) rather than standard CNC cutting, which adds tooling cost and lead time. The design rule is to make the internal radius at least 50% of the nominal wall thickness and the external radius the sum of the internal radius and the wall thickness. Starting both radii from the same point maintains uniform wall thickness through the corner and keeps cooling rates consistent throughout the transition.

Rule 10: Place Parting Lines Strategically

The parting line is where the two halves of the mold meet, and it leaves a witness mark on the finished part — a faint line or slight step that is impossible to eliminate entirely. Strategic parting line placement means locating it where it is least visible and least functionally problematic. Sharp edges are good candidates because they conceal the parting line mark and reduce tooling complexity. Avoid placing parting lines across radiused or curved surfaces: the tight-tolerance fit required to prevent flash on a curved parting surface increases mold cost and raises the risk of flash defects if the mold wears slightly. When possible, move the parting line to the bottom or back of the part, as a consumer product designer would do, rather than defaulting to a centreline split that bisects the most visible face.

Rule 11: Choose Gate Location Before Finalizing Geometry

Gate location is one of the highest-leverage design decisions in injection molding, yet it is frequently treated as a mold-maker's problem rather than a part designer's responsibility. Where plastic enters the cavity determines the fill direction, the location of weld lines (where two flow fronts meet), the distribution of residual stress, and where any gate vestige will appear on the finished part. Making the gate location decision after part geometry is locked in often forces compromises: a gate placed for cosmetic reasons may create a weld line across a structural feature, or vice versa. Determine the gate location early — ideally during the initial design phase — and adjust surrounding geometry to accommodate it rather than retrofitting the gate to a finished design.

Rule 12: Locate Ejector Pins on Non-Cosmetic Surfaces

Ejector pins push the part out of the mold after cooling, and they always leave a small circular mark or slight witness on the part surface. On a non-cosmetic surface — the inside of a housing, a mating face, or a structural rib — this is inconsequential. On a visible external surface, it is a cosmetic defect. Specify ejector pin locations on surfaces that won't be visible in the finished product. Distribute the ejector pin pattern so that the ejection force is spread evenly, applying the force to the part's strongest areas rather than to thin walls or sloped surfaces. Uneven ejection force is a common cause of part distortion and gate drag that isn't immediately obvious from the CAD model but becomes apparent on the first trial shots.

Rule 13: Minimize and Manage Undercuts

An undercut is any feature that prevents the part from being ejected in the straight pull direction of the mold — a side hole, an external hook, an internal groove, or a recessed feature that faces the wrong direction. Undercuts require additional mold components: side actions, lifters, or collapsible cores that move laterally as the mold opens. Each of these adds cost, adds potential points of wear and misalignment, and makes the mold more complex to maintain. Review every feature in your design and ask whether it can be reoriented to align with the mold pull direction. A through-hole perpendicular to the pull direction is an undercut; the same hole oriented parallel to pull is not. Where undercuts are genuinely necessary for function, flag them explicitly during DFM review so the tooling solution can be planned from the start.

Rule 14: Design Snap-Fits for Clean Mold Release

Snap-fit features are assembly-friendly and can eliminate fasteners entirely, but they are a frequent source of tooling complications when not properly designed. External snap-fits are typically undercuts that require side actions to release from the mold. Where possible, design snap-fits to flex clear of the mold on ejection without side actions — this means orienting the deflection direction so that the snap arm bends away from the core during pull rather than being trapped. Also consider that weld lines and snap-fits are a dangerous combination: the tensile strength at a weld line is typically 15–25% lower than the surrounding material, and a snap-fit that cycles repeatedly at a weld line will fail prematurely. Use mold flow simulation to verify that your snap-fit geometry won't sit on a weld line before committing to the tooling design.

Rule 15: Control Weld Line Placement

Weld lines form wherever two separate flow fronts meet inside the mold cavity and fuse together. They occur naturally wherever there is a hole in the part (the flow splits around the core pin and reunites on the downstream side), wherever there are multiple gates, or wherever complex geometry creates separated flow paths. Weld lines are weaker than the surrounding material and can be visible as a faint line on the surface. The designer's job is not to eliminate them — in most complex parts, that's impossible — but to control where they form so they land in low-stress, non-cosmetic areas. Gate position is the primary lever: relocating a gate can move a weld line away from a structural feature or a visible surface without any change to the part itself.

Rule 16: Account for Material Shrinkage Early

All plastics shrink as they cool from the melt temperature to room temperature, and the mold must be cut larger than the desired finished part dimensions to compensate. The amount of shrinkage varies significantly by material: semi-crystalline polymers like nylon and polypropylene shrink more (typically 1.5–2.5%) than amorphous materials like ABS or polycarbonate (typically 0.4–0.8%), and glass-filled grades shrink less and more directionally than unfilled versions. If shrinkage is not accounted for at the design stage, the finished parts will be dimensionally incorrect and the mold will require expensive recutting. This is especially critical for parts with tight tolerances on mating features. Lock in your material choice early and provide the confirmed shrinkage rate to the toolmaker before steel is ordered.

Rule 17: Design Tolerances Steel-Safe

When a part dimension is at or near the boundary of what is achievable, it is worth deliberately biasing the initial tool dimensions toward the easier correction. In injection molding, "steel-safe" means designing the initial mold so that the part comes out slightly undersized on a critical dimension, because making it larger (by removing steel from the mold) is straightforward. Making the part smaller — which requires adding steel back via welding — is expensive and risky. On critical mating features, specify your nominal target with the understanding that the first shots may run slightly small, and plan an intentional polish-and-verify cycle before signing off on final dimensions. This is a standard practice among experienced molders and it consistently reduces the risk of costly rework.

Rule 18: Avoid Unnecessarily Tight Tolerances

Tight tolerances cost money at every stage of the manufacturing process: they require more precise mold cutting, longer cycle times for process stabilization, more rigorous inspection, and higher scrap rates. More importantly, they increase the probability of tooling rework if the first samples fall outside the specified window. The solution is not to loosen tolerances on features where precision matters — it's to be selective. Audit your drawing and challenge every tight tolerance: does this dimension actually drive assembly fit or function, or was it specified as a conservative default? Apply fine tolerances only where they are genuinely required and use commercial tolerances everywhere else. This focused approach reduces tooling risk and keeps manufacturing costs proportionate to functional requirements.

Rule 19: Run Mold Flow Simulation Before Cutting Steel

Mold flow simulation software predicts how molten plastic will fill, pack, and cool inside the mold cavity before any tooling exists. It identifies potential short shots and incomplete fill, predicts where sink marks, weld lines, and air traps will form, evaluates the effect of gate position on fill balance and residual stress, and models warpage after cooling. Running a simulation before committing to tooling is one of the highest-leverage investments available in the design process — catching a fill problem or a weld line conflict in simulation typically costs a fraction of a percent of what it costs to correct after steel is cut. Simulation results can also drive gate repositioning, wall thickness adjustments, or cooling channel layout changes that improve both part quality and cycle time.

Rule 20: Validate with a Prototype Before Committing to Tooling

No simulation replaces physical validation, and no tooling investment should be made without at least one round of prototype testing. 3D printing can produce functional prototype parts in days, allowing fit checks, assembly trials, and basic load testing before the mold design is finalized. CNC machining can produce high-accuracy parts in the actual target material for more demanding mechanical validation. For parts that need to closely replicate the properties of injection-molded components — including overmolded assemblies and multi-part housings — vacuum casting delivers production-quality surface finish and material properties from a low-cost silicone tool. Each of these prototyping methods provides an opportunity to discover design problems while the cost of correction is still low. The cost of a prototype is measured in hundreds of dollars; the cost of a tooling revision is measured in thousands.

Get a DFM Review Before You Cut Steel

These 20 rules cover the design decisions that most commonly drive tooling rework — and applying them consistently is the most reliable way to arrive at a first tool that produces good parts. But no checklist replaces an experienced set of eyes reviewing your specific geometry against the realities of the tooling process. The interaction between wall thickness, gate location, material shrinkage, and parting line placement is always part-specific, and the right answer for one design may be wrong for another.

At NICE Rapid, our engineering team reviews every part for DFM risk before tooling begins. Our plastic injection molding service includes a design review that examines draft, wall uniformity, gate position, undercut risk, and dimensional feasibility — identifying issues while they're still cheap to fix. We also support the full product journey: from early-stage 3D printing and vacuum casting for prototype validation, through rapid tooling for first articles, to low, mid, and high volume manufacturing as your program scales. Whether your next part is a consumer electronics housing, a medical device component, or an automotive assembly, the right design decisions made at the right time protect your schedule, your budget, and your launch.