How to Design for Moldability: 9 Geometry Rules

Somewhere between a part's CAD file and its first production shot, every geometry decision comes due. A wall that looked harmless at 3.2 mm on screen becomes a sink mark the size of a coin on a cosmetic surface. A sharp corner that seemed fine becomes the crack that sends a USD 40,000 mold back to the bench. Design for moldability is not a review step bolted onto the end of a project — it is a geometry discipline that decides whether your injection mold design runs clean from shot one or burns budget in rework. This guide walks through the nine geometry rules our toolmakers apply on every mold we build, with the industry thresholds that separate a moldable part from a costly experiment.

Why Moldability Is a Geometry Problem — and Why It Hurts Most in Production

Roughly 70 percent of injection molding defects trace back to geometry decisions that were locked in before the first design review: uneven wall thickness, missing draft, corner radii that were never modeled. By the time a defect shows up on the production floor, the fix is no longer a CAD edit — it is a steel edit. That distinction is what makes moldability the cheapest insurance in plastic injection mold design.

Follow the cost curve. Changing a wall thickness in CAD costs minutes. Changing it after the mold base is assembled costs machining time, weld repairs, and schedule. Changing it after first-article approval costs retooling, production downtime, and customer confidence. On a typical two-cavity production mold, a single geometry-driven rework cycle can add two to three weeks to delivery and thousands of dollars to the tool price — precisely the outcome that a 24-hour DFM pass is designed to prevent.

At DieStrike, every mold we quote passes through a structured design-for-manufacturability review before steel is ordered. Our engineers run the nine geometry rules below against your model, flag conflicts, and return a mold design and DFM report within 24 hours. The rules themselves are not proprietary — they are the accumulated standard practice of the injection molding industry, and they apply whether you are building a mold for automotive, medical, consumer electronics, or packaging. The same disciplines that earned us IATF 16949 certification in automotive mold programs are the ones that protect a first-time part designer from a costly first mold.

The Snapshot: Nine Geometry Rules in Sixty Seconds

If you take nothing else from this article, keep this list. Each rule is stated with the industry rule-of-thumb value our toolmakers treat as a starting point — not a ceiling:

• Uniform wall thickness — nominal 1–3 mm, adjacent sections within 25% of each other.
• Corner radii — R ≥ 0.5 mm everywhere; ideally 0.5–1.5× the nominal wall.
• Draft angle — 1–2° per side on every vertical face; more on textured surfaces.
• Rib design — rib base 0.5–0.6× wall thickness; rib height ≤ 3× base thickness.
• Boss design — outside diameter ≈ 2× inside diameter; surrounding wall 0.6–0.75× nominal.
• Design undercuts — route around side actions and lifters; keep them shallow and few.
• Shrinkage — plan for 0.5–2% (material-dependent); build cavities oversize to compensate.
• Texture depth — deeper texture demands more draft (≈ 1° per 0.025 mm of depth).
• Tolerance allocation — standard ±0.1 mm; critical features ±0.005 mm, best-case ±0.002 mm.

Each rule below explains the mechanism behind the number, the material context, and what happens when you violate it. Together they form the checklist our engineers use for the free 24-hour DFM review included with every DieStrike quotation.

Rule 1: Keep Wall Thickness Uniform — 1–3 mm Nominal

Wall thickness is the single most influential variable in mold design because it drives filling, cooling, and shrinkage simultaneously. The rule of thumb: hold the nominal wall between 1 and 3 mm for typical engineering thermoplastics, and never let adjacent sections differ by more than 25% of the nominal value. A part designed at 2.5 mm with a 4 mm boss base is a part with a built-in sink mark.

Why 25%? When molten polymer flows from a thin section into a thick one, it hesitates, cools unevenly, and packs unevenly. The thick section cools last and contracts more, pulling the thinner skin inward — the classic sink. On the flow side, uneven thickness produces hesitation marks, short shots at the end of fill, and internal voids in the thick sections. On the cooling side, the penalty is geometric: cooling time scales roughly with the square of wall thickness. Doubling the wall from 1.5 mm to 3 mm can nearly quadruple cycle time — a difference that multiplies across millions of shots.

When a functional requirement forces a thickness change — for example, a threaded section or a mounting pad — step the transition gradually and radius every step. Coring is almost always cheaper than thickening: a hollow rib pattern delivers stiffness without the mass. If the part is structural and genuinely needs 3–5 mm walls, keep the thickness constant across the whole part rather than spot-thickening, and talk to a mold engineer before committing — flow-length-to-wall ratios beyond roughly 150:1 for unfilled resins start to need flow analysis to guarantee fill.

Rule 2: Radius Every Corner — R ≥ 0.5 mm Minimum

Sharp corners are stress concentrators, and in an injection mold they are also flow obstacles. The rule of thumb: no internal or external corner below R 0.5 mm, and ideally radius at 0.5–1.5× the nominal wall thickness. For internal corners, a minimum radius of about 25% of the wall thickness is the floor that prevents crack initiation.

Three mechanisms punish sharp corners. First, mechanical stress: a sharp internal corner concentrates ejection and service loads into a single line, which is where molded parts crack — typically within the first thousand cycles, not at first shot. Second, flow: polymer does not like turning 90-degree corners; sharp transitions create dead zones, hesitation, and knit lines exactly where the corner is weakest. Third, tooling: sharp corners in the cavity mean sharp steel edges that are fragile to machine, hard to polish, and the first place a mold wears — even at HRC 62 tool steel hardness, a razor edge is a liability.

The practical fix is simple: apply a constant fillet radius across every inside corner, from ribs to bosses to wall transitions. External corners should be radiused too — a sharp outside edge is a warp initiator and a handling hazard in assembly. When you hand a model to a mold shop, a model with clean, consistent radii reads as a design that has been thought through, and it shortens the DFM conversation considerably.

Rule 3: Draft Every Vertical Face — 1–2° Per Side

Draft angle is the angle a vertical wall leans back from the ejection direction, and it is the cheapest insurance you can add to a part. The rule of thumb: 1° per side as the absolute minimum, 1.5–2° per side as the working default, and 2–3° per side on deep cores and ribs. Textured surfaces need more — roughly 1° of additional draft per 0.025 mm of texture depth.

Without draft, the part grips the steel of the core as it cools and shrinks onto it. Ejection then has to rip the part free, which produces scuffed surfaces, white stress marks, and parts that stick — and stuck parts are how ejector pins bend and molds get damaged. With 1° of draft per side on a 20 mm deep wall, the part releases cleanly, the cycle stays short, and the ejection system lives longer.

Draft direction matters as much as magnitude. Draft the outside walls so the part releases from the cavity side, and the inside walls so it releases from the core. Holes and bosses need draft on their internal surfaces, which makes the hole slightly smaller at the top — account for that when you specify the critical diameter. The good news: for most parts, adding draft is a five-minute CAD operation at the design stage and a five-figure rework at the steel stage. Our mold design engineers at DieStrike check draft on every surface during the 24-hour DFM review and will flag zero-draft faces before they ever reach the tool room.

Rule 4: Design Ribs at 0.5–0.6× Wall Thickness

Ribs are the standard answer to the eternal conflict between stiffness and wall thickness: add structure without adding mass. The rule of thumb: the rib base should be 0.5–0.6× the nominal wall thickness, the rib height no more than 3× its base thickness, and the radius at the rib base 0.25–0.4× the rib thickness.

The 0.5–0.6× ratio exists for one reason: sink marks. A rib that is as thick as the wall it supports forms a local mass concentration that cools last and sinks the opposite cosmetic face. Hold the rib base under 60% of the nominal wall and the sink stays within acceptable limits; push it to 1:1 and you get a visible depression on the show surface and a likely scrap rate at first articles. When stiffness demands more, add a second rib or increase the height within the 3:1 limit — but remember that a taller rib needs more draft (2–3° per side) or it will not release from the core.

Every rib junction is a potential thick spot. Where ribs cross, core out the intersection or reduce the base at the crossing. Radius the base generously and taper the tip. From a tooling perspective, ribs that follow the direction of draw are free to machine; ribs that run across the draw direction may need side actions or lifters, which is exactly the kind of complication Rule 6 exists to avoid. If you are unsure whether a rib layout will hold sink within specification, send the model for a DFM pass — our engineers model the resin flow and shrinkage before a single cavity is cut.

Rule 5: Boss Design — Outside Diameter at 2× Inside Diameter

Bosses are the cylindrical standoffs that carry screws, inserts, and alignment features, and they are the most commonly overbuilt feature in plastic part design. The rule of thumb: the boss outside diameter should be about 2× the inside diameter, and the wall thickness around the boss 0.6–0.75× the nominal wall. A boss designed for an M3 self-tapping screw (2.4 mm pilot) should therefore carry an outside diameter near 4.8–5 mm — not the 8 mm that looks "stronger" in CAD.

The classic boss failure is the opposite of the rib failure: designers over-thicken the boss for strength and create a sink mark on the opposite face, or under-support it and watch it crack when the screw seats. The 2:1 ratio balances strength against mass. For tall bosses — height above about 2× the diameter — add gussets at the base rather than thickening the wall, and radius every junction. For bosses that carry threaded inserts, specify the wall so the insert has material to grip on both sides; too thin a wall bulges when the insert is pressed in.

Bosses also interact with draft and tolerances. The internal hole needs draft, which changes the pilot diameter along its length — the top of the hole is the size you should dimension. If a boss must sit on a cosmetic surface, keep its wall under 75% of nominal or prepare for sink. And if the boss is on the parting line or near an ejector location, expect witness marks — our engineers routinely relocate ejector pins and gates during the DFM review to keep boss features clean.

Rule 6: Design Undercuts Around Side Actions and Lifters

An undercut is any feature that prevents the part from pulling straight off the mold — a snap hook, a side hole, an internal lip. The rule of thumb: design undercuts deliberately, keep them shallow and few, place them close to the parting line, and plan for them to be formed by side actions (external undercuts) or lifters (internal undercuts) using standard HASCO, DME, or MISUMI components.

Why the caution? An undercut converts a simple two-plate mold into a mold with moving steel. Side actions add slides, wear plates, and hydraulic or mechanical actuation; lifters add angled moving cores that must clear the part before ejection. Every moving element adds machining time, assembly time, and a maintenance point — and it lengthens the cycle, because the mold must open the slide or lift the lifter before ejection can happen. A modest undercut can add 10–20% to the mold cost and push delivery toward the upper end of the 2–4 week band we quote at DieStrike.

Two practical checks. First, ask whether the undercut is necessary at all: can the snap hook face the draw direction? Can the side hole become a slot that draws vertically? Second, for small undercuts on flexible resins such as PP or PE, shallow features under roughly 0.25–0.5 mm can sometimes be force-ejected by the ejector system — but treat that as an exception to be confirmed with the mold maker, never as a default. Where an undercut is unavoidable, our designers position the side action to minimize its stroke, keep the undercut depth as small as function allows, and use standard components so replacement parts are available in days, not months.

Rule 7: Account for Shrinkage — 0.5–2% Rule of Thumb

Every thermoplastic contracts as it cools from melt temperature to room temperature, and the mold must be built to the part's expanded, pre-shrink dimensions. The rule of thumb: plan for mold shrinkage between 0.5% and 2% for most engineering thermoplastics, with amorphous resins like ABS and PC at the low end (0.4–0.7%) and semi-crystalline resins like PP, PA, and POM toward the top (1.5–2.5%). The cavity is cut oversize by the shrinkage factor: a 50 mm nominal dimension on a part in a 1.8% shrink material is cut at roughly 50.9 mm in the steel.

Two mistakes dominate. The first is ignoring shrinkage entirely and relying on the mold trial to "fix" the size — which is expensive because steel can be cut away but not added back. The second is treating shrinkage as a single number. Real shrinkage is anisotropic: it is larger in the flow direction than across it, it varies with wall thickness, packing pressure, and gate location, and it is the root cause of warpage in long, thin parts. Glass-filled grades shrink less but warp more directionally. That is why a competent injection mold design does not stop at multiplying by a data-sheet factor — it positions gates, balances flow, and sizes cooling channels so the part shrinks predictably.

At DieStrike, shrinkage is handled in the DFM stage: we confirm the material and its shrink range with you, build the cavities to the compensated dimensions, and validate with mold trials before production. Because we cut steel in-house across 120+ machines, correcting a shrinkage assumption during trial costs days, not weeks — and our customers get parts that hit dimension the first time they run.

Rule 8: Match Texture Depth to Draft — ≈ 1° per 0.025 mm

Texture is a finish decision made in the design office and paid for in the mold — because a textured cavity surface grips the part like sandpaper grips a hand. The rule of thumb: add roughly 1° of draft per 0.025 mm of texture depth, on top of the base 1–2° draft. A VDI 24–36 class texture, common for consumer and industrial housings, typically needs 3–5° total draft to eject without scuffing.

The failure mode is ejection tearing. As the part shrinks onto a textured core, every microscopic peak of the texture bites into the polymer. Without enough draft, the ejector system drags the part off, leaving witness scuffs, stretched texture, or outright tear-out on the textured face — and a textured cavity is far harder to repair than a polished one, because the texture pattern must be matched. Deeper textures also hide flow defects like sink marks and weld lines, which is why designers love them — but the draft bill comes due at ejection.

Two notes for the design file. First, texture depth specification belongs on the drawing: name the standard (SPI, VDI, Mold-Tech) and the grade, so the mold maker applies the right etching depth and can calculate the draft you actually need. Second, decide texture direction relative to draw: a texture that runs with the draw direction ejects more easily than one that runs across it. When in doubt, our DFM engineers will compute the required draft from the specified texture grade and flag conflicts before etching begins.

Rule 9: Allocate Tolerances Feature by Feature — ±0.005 mm Where It Matters

Tolerances are the budget of moldability: every micron you demand has a price, and the price compounds on every feature. The rule of thumb: hold general tolerances at ±0.1 mm, tighten to ±0.05 mm where assembly requires it, and reserve ±0.005 mm — DieStrike's production capability, with ±0.002 mm on select features — for the two or three dimensions that genuinely determine function.

What drives achievable tolerance is geometry, not machinery. A dimension parallel to the draw direction sits on the core and is controlled by machining and shrink compensation. A dimension across the parting line depends on how the mold closes and is inherently looser. Features near gates and ejector pins carry witness and packing variation. So the smart allocation is feature-by-feature: tight on bore diameters, shaft fits, and mating faces; loose on external profiles, wall positions, and anything the customer will never measure. This is the difference between a moldable part and a part that needs five sampling rounds.

DieStrike's precision capability comes from equipment discipline: 120+ machining and inspection units, wire and sinker EDM for fine details, hardened tool steel to HRC 62, and CMM inspection on every mold before shipment. But the highest-leverage tolerance tool is the DFM review — catching an impossible tolerance before steel is cut. If you are unsure what your part can hold, send the drawing: our engineers will return a tolerance allocation recommendation within the 24-hour DFM window, and we build and inspect every mold to the agreed numbers.

Geometry Rules at a Glance: The Zebra Reference Table

Print this table, pin it above the CAD station, and run your model against it before you send it out for quotation.

FAQ: Mold Design Questions Engineers Ask

What does "moldability" mean in injection molding?

Moldability is the degree to which a part's geometry can be produced reliably and repeatably in an injection mold — filling completely, cooling without distortion, ejecting without damage, and holding dimension. A moldable part is one whose walls, radii, draft, ribs, bosses, and tolerances sit within the ranges the process can support. The nine rules above are the operational definition of moldability.

What is the minimum draft angle for injection molding?

1° per side on smooth, polished steel is the accepted minimum, with 1.5–2° per side recommended as the working default. Deep cores, ribs, and textured surfaces need more: 2–3° per side, plus roughly 1° per 0.025 mm of texture depth. Zero-draft parts are possible with special mold features, but they cost more and run slower.

How thick should a rib be relative to the wall?

The rule of thumb is 0.5–0.6× the nominal wall thickness, with the rib height no more than 3× its base thickness and a base radius of 0.25–0.4× the rib thickness. Going thicker than 60% of the wall invites sink marks on the opposite surface; going taller than 3:1 invites fill and ejection problems.

Can undercuts be avoided entirely?

Many can. Rotating the feature to draw direction, converting side holes to slots, or splitting a housing into two parts that slide together eliminates the undercut without losing function. Where an undercut is unavoidable, keep it shallow, place it near the parting line, and form it with a standard side action or lifter. Every undercut adds moving steel, cost, and cycle time, so treat it as a design decision rather than an accident.

How long does a DFM review take?

At DieStrike, a DFM review of your model and drawing is returned within 24 hours of receiving your inquiry — including geometry checks against the nine rules above, material and shrinkage confirmation, tolerance recommendations, and a quotation. The review is free and carries no obligation; you can see exactly how your part scores on moldability before you commit a single dollar.

What tolerances can a precision mold actually hold?

Standard injection molding holds about ±0.1 mm. With precision mold design, tight features can be held to ±0.05 mm, and DieStrike production molds hold critical dimensions to ±0.005 mm, with ±0.002 mm achievable on select features. Tolerance should always be allocated feature by feature — tight where function demands it, loose everywhere else — because every tightened dimension shows up in the tool price and the sampling schedule.

The Bottom Line: Moldability Is a Design Decision, Made Early

Every one of the nine rules is cheap at the CAD stage and expensive at the steel stage. Uniform walls, generous radii, honest draft, disciplined ribs and bosses, deliberate undercuts, compensated shrinkage, texture-aware draft, and feature-by-feature tolerances — together they are the difference between a mold that runs clean from shot one and a mold that consumes its profit in rework. The good news is that you do not have to hold all of them in your head: a proper DFM pass catches the conflicts before steel is cut.

That is the model we run at DieStrike. We are an IATF 16949-certified precision mold manufacturer with 120+ machines, in-house machining from hardened tool steel at HRC 62, and a full standard parts line — ejector pins, core pins, sprue bushings, hot runner systems, mold bases, springs, leader pins, and punches in HASCO, DME, and MISUMI standards. We quote MOQ 1, deliver production molds in 2–4 weeks, and return your DFM review within 24 hours. Send us your model, and we will tell you — in writing, with numbers — exactly how moldable your part is, and how to make it more so. If you are earlier in the journey, our DFM checklist and guide to buying injection molds cover the procurement side, and for inserts that are already in production, our custom mold inserts page shows how fast a precision component can be delivered.