Views: 0 Author: Peter Cui Publish Time: 2026-06-02 Origin: Mitour Silicone
Table of Contents
TL;DR — Silicone mold design is the single largest cost driver in any custom silicone product project. Poor mold design decisions — wrong parting line, inadequate draft angle, insufficient wall thickness — show up as expensive T2 and T3 revision rounds, visible flash, or rejected first articles. At Mitour Silicone, our in-house tooling team has cut over 2,000 custom molds since 2005 across 300+ patented designs. This guide gives brand owners the vocabulary and decision framework they need to walk into a mold discussion and not get surprised later.
Parting line placement is the most consequential design decision — it determines where flash appears and where witness lines are visible on the finished part.
Draft angle: minimum 0.5° on all vertical walls for LSR; 1–3° for HTV. Insufficient draft causes tearing during demolding.
Minimum wall thickness: 0.5 mm for LSR, 1.5 mm for HTV. Thin walls below these thresholds cause incomplete fill and surface defects.
Mold material: P20 steel for prototype/low-volume molds; H13 (1.2344) hardened steel for production molds above 100,000 shots.
Cavity count vs. mold cost: a 4-cavity mold costs roughly 3–3.5× a single-cavity mold and delivers 4× the output per press cycle.
Cold runner vs. hot runner: irrelevant for HTV; critical for LSR. Cold runner LSR molds eliminate sprue waste and reduce cycle time.
Always request a DFM report before mold purchase. DFM identifies design features that will increase cost or reduce quality before you commit tooling budget.
Brand owners think about their product in terms of how it looks, how it functions, and how it is packaged. Factories think about it first in terms of how it can be manufactured — and specifically how it can be released from a mold 50,000 times without destroying itself, producing flash, or creating parting lines in inconvenient places.
The disconnect between these two perspectives costs the industry millions of dollars per year in rework, mold revision, and rejected tooling. It is not the buyer's fault for not knowing tooling principles. But a basic vocabulary of mold design concepts will help you ask the right questions, understand the answers, and avoid the most common expensive surprises.
This article covers the seven most important mold design concepts for silicone products, with specific numbers, trade-offs, and examples drawn from our two decades of production experience.
The parting line is where the two halves of a mold meet. Every silicone product has one, and on every finished part, the parting line leaves either a witness mark (a faint line in the surface) or flash (a thin film of excess material). You cannot eliminate parting lines — you can only choose where to put them.
The rule: place the parting line where it will be least visible, least functional, and easiest to trim if flash occurs.
Good parting line locations:
At a natural geometric edge or corner (the bottom rim of a bowl; the equator of a spherical part; a shelf step in the design)
On a non-functional surface (the back or underside of a product)
Along a decorative groove or recess that disguises the witness line
Bad parting line locations:
Across a flat, high-gloss cosmetic surface (parting witness marks are highly visible on shiny silicone)
Through the center of a sealing surface (flash on a sealing face causes leak failures)
Across text or texture patterns (parting witness marks interrupt pattern continuity)
When we review a buyer's design, the first question is always: "Where do you want the parting line?" Many buyers have not thought about it. Our DFM report includes a parting line recommendation with visual markup showing our proposed location and the reason for it. If the buyer's aesthetic requirements conflict with our tooling recommendation, we discuss trade-offs explicitly.
A real example: In Q3 2023, we revised a baby spoon design where the buyer's concept placed the parting line across the top of the spoon bowl — a high-visibility cosmetic surface. Our DFM proposed moving it to the rim edge of the bowl. The mold modification cost USD 180 versus what would have been a T2 revision at USD 350 plus a delayed sample timeline. We catch these decisions before tooling cuts, not after.
A draft angle is the slight taper applied to vertical walls in a mold to allow the part to release cleanly when the mold opens. Without sufficient draft, the silicone grips the steel mold wall and tears during demolding — or the mold cannot open without the part deforming.
Draft angle requirements by process:
Process | Minimum draft angle | Recommended draft |
LSR injection | 0.5° | 1–2° |
HTV compression | 1° | 2–3° |
Overmolded silicone | 1° on silicone surfaces | 1.5–2° preferred |
The lower the Shore A hardness of the silicone, the more important draft becomes — soft parts at Shore A 20–30 are more prone to tearing during release than stiffer Shore A 60–70 parts.
When buyers push back on draft angles: Brand owners sometimes want perfectly vertical walls because it looks cleaner in their CAD rendering. The visual difference between a 1° draft and a 0° wall is essentially invisible to the human eye on any part shorter than 50 mm. We always add the draft. The draft is not optional on production molds.
For very deep draws — a 50 mm deep silicone cup, for example — we recommend 2–3° draft and a separate ejector design using a stripper plate. Without the stripper plate, a deep-draw cup in Shore A 40 silicone will invert itself during release and require manual reversal, adding labor cost to every shot.
Wall thickness is where most first-time silicone product designers make their most expensive mistake. Silicone is not plastic. Thin walls in silicone do not just look fragile — they cause processing failures: incomplete fill, air entrapment, surface dimpling, and inconsistent curing.
Minimum wall thickness by process:
Process | Absolute minimum | Recommended minimum |
LSR injection | 0.5 mm | 0.8 mm |
HTV compression | 1.5 mm | 2.0 mm |
Below these minimums, fill pressure cannot overcome the viscosity of the material and force it to the extremities of the cavity before gelation begins. The result is short-shot — an incompletely formed part.
Maximum wall thickness: both processes have upper limits too, though the failure mode is different. For LSR, walls thicker than 6 mm create cure uniformity issues — the center of the wall may cure more slowly than the surface, leading to internal stress. For HTV, very thick sections extend cure time proportionally (doubling wall thickness roughly doubles cure time) and may require stepped mold temperature profiles.
Wall thickness uniformity: more important than absolute thickness in many designs. A part with consistent 2 mm walls cures uniformly. A part with sections alternating between 1 mm and 5 mm creates differential cure rates, residual stress, and dimensional variability. Good silicone product design maintains wall thickness variation below 2:1 ratio wherever possible.
Ribs and texture: Ribs (structural reinforcing features on silicone) should be 0.6–0.8× the adjacent wall thickness. Texture patterns (cross-hatch, dimples, leather grain) should not reduce local wall thickness below the minimum for the process.
An undercut is any feature on a part that prevents the mold from opening in a straight pull. Examples: a groove around the outside of a cap that allows it to snap over a bottle lip; a hook or clip feature on the side of a part; an internal channel.
Why undercuts matter: they require additional mold mechanisms — side actions, sliders, or lifters — that increase mold complexity and cost.
The good news for silicone: because silicone is highly elastic (elongation at break of 300–600%), many features that would require side actions in plastic molding can be demolded by manual or robotic distortion of the flexible silicone part. A groove that is 10% undercut can often be stripped from a straight-pull mold if the part is designed with sufficient flexibility and draft on the non-undercut surfaces.
When undercuts genuinely require mold actions: deep grooves (depth-to-width ratio above 0.5); rigid-substrate overmolds where the substrate cannot flex; internal channels with no draft; multi-direction features on a single part.
Side actions add USD 300–1,500 to mold cost per action, plus maintenance complexity. In our DFM, we specifically flag undercuts that can be stripped (no added cost) versus those that require actions (quantified cost increase). Often a 1–2 mm geometric modification to the design eliminates the undercut entirely.
Not all steel is equal for silicone tooling. The choice of mold material has a direct impact on dimensional stability, surface finish quality, flash control, and mold life.
P20 pre-hardened steel (HRC 28–32): the standard prototype and medium-volume mold material. Machines well, polishes acceptably (up to SPI B-1 surface finish), and handles 50,000–150,000 shots comfortably. Good choice for ODM evaluation molds, short-run OEM projects, and designs likely to change before high-volume production.
H13 (DIN 1.2344) hardened steel (HRC 48–52): the production standard for high-volume LSR molds. Hard steel holds shutoff tolerances (<0.005 mm) that prevent flash in platinum-cure LSR. Polishes to SPI A-2 (near-mirror) for optical parts. Expected life: 500,000–1,000,000+ shots with proper maintenance.
Aluminum: occasionally used for silicone prototype molds where speed and cost are paramount. Machines quickly; poor shutoff durability for LSR; acceptable for HTV prototyping. Life typically under 10,000 shots.
Our standard: all production OEM molds at Mitour Silicone are cut in H13 hardened steel. We do not offer aluminum production molds — the false economy of lower upfront cost is eliminated by their short life and inconsistent shutoff performance.
Single-cavity molds are fine for prototypes and low-volume production. For volume above 5,000 units/year, multi-cavity tooling changes the economics entirely.
Cavity count economics for LSR on a 4-cavity mold:
Cycle time: 30 seconds per shot
Output per 24-hour cell: 4 cavities × (3,600 ÷ 30) shots/hour × 24 hours = 11,520 parts
Mold cost for 4-cavity vs single-cavity: approximately 3–3.5× (not 4×, because base structure, runner, and ejector are shared)
Break-even versus single-cavity: typically at 10,000–15,000 units, depending on part cost
The constraint: cavity count is limited by press platen size and machine tonnage. A four-cavity baby nipple mold at 8 g per cavity occupies roughly 300 × 250 mm platen space and runs on an 80-ton press. Scaling to 8 cavities at 600 × 250 mm requires a larger press. We size the mold to the most appropriate machine in our fleet, not the largest available.
Buyers often ask: "Can we start with a 2-cavity mold and add cavities later?" Yes, for some mold architectures — specifically, modular mold designs where cavities are interchangeable inserts in a common frame (often called a "family mold" or "insert mold" system). This is exactly the approach we use for our standard ODM mold base system. For buyers who want to validate design before committing to a full multi-cavity tool, we recommend a 1- or 2-cavity T1 mold followed by a multi-cavity production mold once the design is locked. The T1 mold cost is credited against the production mold in our standard quoting.
For HTV compression, the runner system is simple — material is placed directly in the cavity as a preform. For LSR injection, the runner system is a significant engineering decision.
Direct gate (hot sprue): the simplest configuration. Material enters the mold through a single sprue, travels through a runner channel (which is at mold temperature = hot = cured), and into the cavities. Waste: the cured sprue must be removed by the operator or robot after each shot. Sprue waste typically runs 8–15% of shot weight. Acceptable for low-volume or single-cavity molds.
Cold runner system: the runner manifold is cooled to 5–10°C, keeping material in it liquid and uncured between shots. No sprue waste — the material in the runner is re-injected on the next shot. Cold runner systems save 8–15% material cost, eliminate a secondary operation (sprue removal), and reduce cycle time by 2–5 seconds (no cooling of a thick sprue in the hot mold zone).
Cold runner cost: USD 1,500–4,000 added to mold cost for a typical 4-cavity mold. Payback calculation at USD 8/kg compound, 15% sprue fraction, 30,000 units/year at 10 g/part: saves approximately USD 360/year in material. So for high-volume molds, cold runner pays back in 4–11 years on material alone — the real payback driver is operator labor elimination and cycle time improvement.
Our default: all multi-cavity production LSR molds at Mitour Silicone specify cold runner systems unless the buyer explicitly requests direct gate for cost reasons. For baby-contact products, cold runner is mandatory — operator contact with hot sprue material in proximity to baby-contact parts is a contamination risk.
Every OEM project starts with a DFM (Design for Manufacturability) review. Here is exactly what that looks like.
Step 1 — Brief receipt (Day 0)
You send us your concept: CAD file (STEP, IGES, SolidWorks, or any standard format), hand sketch with dimensions, or a physical sample with annotated reference dimensions. Tell us: intended application, target material hardness, any regulatory requirements, target unit cost range, and expected annual volume.
Step 2 — Engineering review (Days 1–2)
Our tooling engineer and process engineer jointly review the design. They check: parting line feasibility, draft angles on all surfaces, wall thickness uniformity, undercut identification, expected mold type (LSR/HTV) and cavity count recommendation, and any geometric features that require special mold mechanisms.
Step 3 — DFM report delivery (Day 2)
We deliver a marked-up PDF showing:
Recommended parting line (green) with alternatives (yellow)
Surfaces requiring draft modification (red, with specific angle recommendations)
Thin-wall warnings (orange, with recommended modifications)
Undercut analysis and mechanism recommendations
Mold concept sketch (single-cavity schematic showing core/cavity split)
Preliminary mold cost range and lead time estimate
Step 4 — Design revision (Days 3–7, buyer side)
You or your designer incorporate DFM feedback. Most designs require 1–2 rounds of iteration. We review revised files at no additional charge within the same project.
Step 5 — Mold purchase commitment
Once DFM is approved, you commit to mold tooling. Lead time for mold cutting begins from this date.
Specifying a high-gloss surface finish on a surface that will have a parting witness line. SPI A-2 (mirror polish) costs extra to achieve and maintain, and witness lines are most visible on high-gloss surfaces. Match your surface finish specification to the parting line reality.
Designing logos into the mold without specifying minimum feature size. Silicone fills detail well, but raised text below 0.5 mm height tears easily during repeated use. Debossed logos (recessed into the surface) are more durable than raised logos in silicone.
Ignoring the ejector system. Brand owners rarely think about how the part exits the mold — but ejector pin marks leave witness circles on the part surface. Specify where ejector marks are acceptable before the mold is designed; do not discover them on your T1 sample.
Expecting silicone to match a Pantone swatch perfectly on the first try. Silicone colorants behave differently from textile or plastic dyes. We match to Pantone within Delta E 2.0 as a standard and Delta E 1.0 for premium cosmetic products — but "matching a Pantone chip" requires a color-matching trial before production. Budget one color iteration cycle.
Confusing nominal dimension with tolerance. Silicone has a natural thermal expansion and dimensional variability. For tight-fit assembly (a silicone cap that must seal a glass bottle), we need the mating dimensions of the bottle, not just the silicone cap nominal dimension. We will calculate the silicone nominal and tolerance to achieve your assembly function.
Mold type | Approximate cost range | Typical life | Best for |
HTV prototype, 1-cavity | USD 400–800 | 5,000–10,000 shots | Design validation, under 500 units |
HTV production, 1-cavity | USD 800–2,000 | 200,000+ shots | 1,000–10,000 units/year |
HTV production, 4-cavity | USD 2,000–5,000 | 200,000+ shots | 5,000–30,000 units/year |
LSR production, 1-cavity | USD 1,800–4,000 | 500,000+ shots | 2,000–10,000 units/year |
LSR production, 4-cavity + cold runner | USD 5,000–12,000 | 500,000+ shots | 10,000+ units/year |
LSR production, 8-cavity + cold runner | USD 10,000–20,000 | 500,000+ shots | 30,000+ units/year |
Overmold tool (insert molding) | USD 2,000–8,000 | 200,000+ shots | Silicone-over-plastic products |
Note: all costs are for tooling manufactured at Mitour Silicone's in-house tooling facility. Third-party tooling purchase from Chinese mold makers (without production commitment) carries similar ranges.
DFM stands for Design for Manufacturability — a written engineering analysis of your product design covering parting line placement, draft angles, wall thickness, undercut features, and process recommendations. You absolutely need one before committing mold budget. At Mitour Silicone, DFM review is included at no charge for projects above our minimum order threshold.
Standard single-cavity production molds take 10–14 days for steel cutting, EDM, and polishing. Multi-cavity molds and complex geometries take 14–21 days. We quote lead times before you commit. Rush mold services (7 days) are available for simple geometries at a 30% tooling premium.
Technically yes — the mold can be shipped to another factory. Practically, there is a re-qualification cost at the receiving factory (typically 1–2 days of press time and a T1 sample run) because machine parameters, material viscosity, and injection conditions differ between factories. At Mitour Silicone, we retain molds in active storage at no charge for 18 months between orders.
All production molds are cut in H13 (DIN 1.2344) hardened steel, heat-treated to HRC 48–52. This steel grade holds the shutoff tolerances below 0.005 mm required for near-zero flash in LSR injection. Prototype molds use P20 pre-hardened steel (HRC 28–32) for faster machining.
H13 production molds, properly maintained with scheduled polishing every 30,000 shots, routinely last 500,000–1,000,000 shots. P20 molds last 50,000–150,000 shots. The limiting factor is parting surface wear, which is why parting surface polishing is part of our preventive maintenance protocol.
Minimum wall thickness is 0.5 mm for LSR injection (recommended 0.8 mm) and 1.5 mm for HTV compression (recommended 2.0 mm). Below these thresholds, fill is incomplete, leading to short shots and surface defects. We flag thin-wall areas in our DFM report before the mold is cut.
A cold runner is a cooled manifold system that keeps LSR material liquid in the runner between shots, eliminating cured sprue waste. Cold runners add USD 1,500–4,000 to mold cost but eliminate material waste (8–15% of shot weight) and labor for sprue removal. At Mitour Silicone, cold runners are standard on all multi-cavity LSR production molds.
You do not need to be an engineer. Send us your product reference (sketch, photo, competitive sample, or CAD file), tell us what the product does and who uses it, and our engineering team handles the DFM. We translate your functional requirements into mold design specs. Email yfsalee@mymitour.com with "DFM Request" in the subject line.
Mold design decisions made before tooling commitment determine the quality and cost of everything that follows. The right time to invest in a thorough DFM is before you spend a dollar on steel, not after your first article comes back with visible flash on a cosmetic surface.
Review our OEM/ODM capabilities
See our manufacturing process
Explore our material options
Learn about our facility and tooling shop
Email: yfsalee@mymitour.com (subject: "DFM Request — [your product description]")
Phone: +86 199 2529 4106
Peter Cui | 21 years of silicone manufacturing experience | 4,500 m² Shenzhen facility | Walmart-, Target-, and Disney-approved supplier | Contact: yfsalee@mymitour.com
