Key takeaways

• The biggest cost and lead-time drivers are usually master pattern effort, silicone mold complexity, and finishing + inspection, not “resin price.”

• You reduce cost fastest by (1) simplifying demolding and (2) tightening tolerances only where they change function.

• You reduce lead time fastest by sending a minimum viable spec that prevents clarification loops.

• Plan your project as three layers—pattern, mold, parts—so you can control what you’re paying for.

A typical vacuum casting workflow looks like this:

1. Master pattern: made from CNC or (very commonly) SLA/SLS 3D printing.

2. Silicone mold: silicone is poured around the master, cured, then cut open to form the cavity.

3. Casting: two-part polyurethane resin is mixed, degassed, poured into the mold under vacuum (and/or cured under vacuum), then trimmed and finished.

Where vacuum casting wins

Vacuum casting tends to be a strong fit when you need:

• 10–50 plastic parts with repeatable geometry

• better surface finish than many “as-printed” parts

• engineering resin options (rigid, flexible, clear, flame-retardant)

• parts fast enough to keep an EVT/DVT/PVT schedule moving

Where it tends to lose

Vacuum casting is usually the wrong tool when:

• you need hundreds to thousands of parts (tooling amortization flips)

• you need tight tolerances on many dimensions (inspection and post-machining dominate)

• the design includes lots of fragile, demolding-unfriendly micro-features

A quote is easier to control when you understand its structure. Most vacuum casting projects can be modeled as:

Total cost = Master pattern + Mold(s) + (Per‑part cost × Quantity) + Secondary ops

If you know which layer you’re inflating, you can reduce cost without compromising the functional requirements.

Layer A — Master pattern cost (the “source of truth”)

The silicone mold will replicate whatever the master pattern is—good or bad. Pattern effort increases when you require:

• high cosmetics on many faces

• delicate features that break or warp during printing/handling

• part splits or design changes late in the process

Cost/lead-time lever: separate cosmetic and non-cosmetic surfaces. If only one face is customer-visible, don’t pay to polish everything.

Layer B — Silicone mold cost (complexity charges you rent)

Silicone molds are inexpensive compared to steel tooling, but geometry still affects mold cost. Mold effort increases with:

• complex parting lines

• deep undercuts that require careful cutting

• large parts (more silicone, more handling)

• designs that tear silicone during demolding

Many service references cite silicone mold life in the ~25–50 shot range in common setups before dimensional drift becomes noticeable

What that means for your quote: your design can indirectly set how many molds you’ll need, which sets your effective per-part tooling cost.

Layer C — Per-part cost (labor often dominates)

Per-part cost grows when you add:

• demolding difficulty

• trimming and hand finishing

• painting/texture matching

• inserts and secondary machining

• expanded inspection + documentation

If your part requires a lot of manual work, vacuum casting stops being a “cheap” process. It’s still valuable—but you should treat it as a controlled manufacturing workflow.

A vacuum casting project is a chain. If one link is vague, everything waits.

The schedule, simplified

1. DFM and requirement clarification

2. Master pattern build + finish

3. Silicone mold build + cure

4. Casting cycles + curing

5. Finishing + inspection + packing

The three most common lead-time killers

1) Missing requirements

Missing tolerances, finish requirements, color targets, inserts, or acceptance criteria cause clarification loops. Each loop is usually a day lost.

2) Pattern revision

If you discover late that an A-surface can’t show a parting line—or a snap is too thin—the master pattern may need changes.

3) Unplanned finishing and QC docs

Finishes and documentation are work packages. If you don’t specify them early, you either (a) get a quote change later or (b) get parts that don’t match expectations.

Pro Tip: If you only have time to do one pre-RFQ task, create a one-page “critical requirements list”: critical dimensions + finish + quantity + ship date + acceptance criteria. It’s the highest ROI document in the project.

If you remember only five things from this vacuum casting design guide, make them these.

Lever 1 — Wall thickness (thick sections are expensive twice)

Thick sections increase resin usage and cure risk.

• more resin volume

• slower cure and higher heat buildup

• higher risk of sink, distortion, or cosmetic defects

Cost/lead-time move: keep walls as uniform as function allows. For stiffness, prefer ribs rather than “make everything thicker.”

Lever 2 — Undercuts and demolding difficulty

Silicone flexes, so undercuts are often feasible—but they can still cost you via:

• mold cutting complexity

• slower demolding

• increased mold wear/tearing

Decision rule:

• If an undercut doesn’t change function, remove it.

• If it’s functional (snap fit), minimize depth and put it on non-cosmetic faces.

• If it’s unavoidable, ask how it affects mold life and whether a multi-part mold is required.

Lever 3 — Tolerances (tight everywhere is the fastest way to inflate cost)

You don’t pay for the number. You pay for the control and verification behind the number.

In low-volume workflows, vacuum casting is often used for “good functional tolerance” rather than precision machining everywhere. One practical baseline comes from a Kaierwo technical comparison of vacuum casting vs. CNC machining, which uses ±0.2 mm as a typical vacuum casting tolerance reference point for many applications and notes CNC machining can achieve tighter tolerances in appropriate cases (Kaierwo’s vacuum casting vs. CNC machining analysis).

Cost/lead-time move: tolerance only what’s functionally critical. Everything else should be a general tolerance.

Lever 4 — Cosmetic finish and color matching

Cosmetics can dominate schedule because they’re manual, subjective, and iterative.

Cost/lead-time move:

• define A-surfaces

• define acceptable defects on hidden faces

• if you need a specific color, specify the standard (Pantone/RAL sample)

Lever 5 — Inserts, threads, and secondary ops

Threads, inserts, tight bores, and press-fits can be done—but they convert a “casting job” into a hybrid (cast + machine + assemble).

Cost/lead-time move: if it needs post-machining, specify which features, how they’re referenced (datums), and what the acceptance criteria is.

Engineers usually underestimate schedule risk because the per-part casting time feels short. In reality, iteration and ambiguity dominate.

A simple estimator can help you choose what to simplify.

Step 1: Split your part into “cost drivers”

Ask these questions and mark each as Low / Medium / High.

A) Pattern effort

• Low: mostly smooth surfaces, modest feature density, no demanding cosmetics

• Medium: some cosmetics, some delicate features

• High: many cosmetic faces, tight visual requirements, fragile snaps/edges

B) Mold complexity

• Low: easy parting line, minimal undercuts, demolding path is obvious

• Medium: some undercuts, tall walls, delicate features near the parting line

• High: deep undercuts, multi-part mold likely, high risk of tearing

C) Secondary ops

• Low: trim only, as-cast surface acceptable

• Medium: a few inserts, some light sanding/paint

• High: extensive machining, paint + texture matching, tight fit interfaces

Step 2: Translate it into a planning decision

• If A is High, expect extra days for pattern finishing and verification. Consider relaxing cosmetic scope.

• If B is High, expect mold iteration risk. Consider redesigning undercuts or relocating critical faces.

• If C is High, treat it as a hybrid manufacturing job. Consider whether CNC or injection molding is the right long-term process.

Worked example (typical)

Imagine a handheld enclosure:

• Cosmetic outer shell, matte paint → Pattern effort: High

• Several snap fits and internal ribs → Mold complexity: Medium–High

• Heat-set inserts and tight PCB mounting bosses → Secondary ops: Medium

The fastest schedule move is not “faster casting.” It’s:

1. define A-surfaces and allow witness marks inside,

2. soften/relocate the worst snap undercuts,

3. call out only the 3–6 critical interface dimensions.

That combination reduces clarification loops, reduces mold damage risk, and reduces post-machining.

This isn’t a full plastic design textbook. It’s a set of DFM choices that reduce scrap, rework, and mold damage.

1.Use ribs for stiffness, not thick walls

A common guideline is to keep ribs thinner than the adjoining wall; Formlabs notes ribs can be around ~60% of wall thickness in many designs.

Why it matters for cost/lead time: thinner ribs fill more reliably and cure with less distortion.

2.Add radii and fillets to protect the mold

Sharp corners are common mold-tear initiation points.

• add fillets on internal corners

• prefer generous radii where possible

Why it matters: fewer torn molds, easier demolding, more consistent parts over the mold’s life.

3.Add draft when you can

Even with flexible silicone, draft reduces demolding time and mold damage—especially on tall vertical walls.

Why it matters: faster demolding and fewer stuck parts.

4.Treat micro-features as “risk flags”

Tiny embossed logos, micro text, razor edges, and ultra-thin snaps fail in one of two ways:

• they don’t fill cleanly

• they fill, but break during demolding/handling

RFQ move: flag micro-features as critical and ask the supplier to confirm a minimum size based on the selected resin.

“Can you hold ±0.1 mm?” isn’t the right question.

The right question is: which dimensions need to be that tight, and how will we verify them?

1.Why tight tolerances are expensive in vacuum casting

• resins and silicones shrink and behave differently

• geometry sensitivity (large flat panels warp more easily)

• inspection burden (fixturing, measurement time, documentation)

2.A tolerance spec that suppliers can quote quickly

Use a two-zone approach:

• Zone 1 — critical dimensions: explicitly tolerance only what affects fit/function

• Zone 2 — general tolerance: a general tolerance for everything else

Then add:

• datums (so measurement is unambiguous)

• a quick note on verification (gauge, mating part, go/no-go)

Key Takeaway: If you don’t identify critical dimensions, the supplier has to guess—and the safe guess is expensive.

Rework is the hidden tax on both cost and lead time. Most rework comes from a small set of failure modes.

1.Bubbles, voids, and “cosmetic pitting”

Common causes:

• air traps due to poor venting or geometry that creates dead-end pockets

• resin mixing/degassing issues

• surface defects on the master pattern replicated into every part

What to do (as a buyer/specifier):

• flag bubble-sensitive areas (e.g., sealing faces)

• avoid dead-end thin cavities where possible

• define whether tiny surface defects are acceptable on non-critical faces

2.Warpage on large flat panels

Common causes:

• big flat spans with uneven wall thickness

• asymmetric ribbing

• tight tolerances on features that are naturally prone to distortion

What to do:

• add ribs strategically (not just thicker walls)

• allow looser tolerances on non-interface faces

• define flatness requirements only where they matter

3.Dimensional drift across a batch

Silicone molds can wear and deform; many references cite mold lives in the 25–50 shot range for typical molds.

What to do:

• specify which dimensions must remain stable across the entire batch

• ask whether multiple molds will be used for your quantity

• for tight interfaces, consider post-machining those specific features

4.Cosmetic mismatch (color, gloss, texture)

Cosmetics are subjective. They also require iteration.

What to do:

• provide a color standard (Pantone/RAL) if it matters

• define A-surfaces and acceptable defects

• ask for a first-article sample approval step for cosmetic parts

Warning: If you request “tight tolerances” and “high cosmetics” but don’t define acceptance criteria, you’ll often get either (1) a conservative, expensive quote, or (2) a fast quote that later changes. Ambiguity is expensive.

If your project is schedule-sensitive, finishing must be treated as a separate work package.

1.“As-cast” vs cosmetic parts

“As-cast” can be fast. Cosmetic parts often require:

• sanding/polishing

• primer + paint cycles

• masking

• rework

Decision rule: if the part is for internal engineering validation, don’t pay for production cosmetics unless you truly need them.

2.Make acceptance criteria explicit

Cosmetics are subjective. The easiest way to trigger delays is to specify “high quality finish” without defining what defects are acceptable.

Include at least one of:

• reference photo standard

• allowed scratch/pin-hole area

• whether parting line witness is acceptable

If your project is schedule-sensitive, finishing must be treated as a separate work package.

Quantity planning and silicone mold life: pick a batch strategy

When you plan quantity, you’re really planning tooling amortization and drift risk.

1.A practical batch model

• ~1–10 parts: optimize for speed; keep tooling simple

• ~10–50 parts: vacuum casting is often strongest; plan for consistent acceptance criteria

• >50 parts: ask whether multiple molds are needed, and whether another process is a better fit

Service references commonly cite mold lives in the 25–50 shot band for typical silicone molds (for example, the earlier 3ERP reference). Use that as a planning prompt—not as a guarantee.

2.How design affects mold life (and your per-part cost)

Mold life tends to decrease with:

• deep undercuts

• sharp corners and thin edges that tear silicone

• aggressive demolding paths

• high-temperature resins or conditions that stress silicone

Cost implication: if the mold fails early, you pay for additional molds—and your per-part cost rises.

Below is a checklist you can paste into an email or RFQ form.

A) CAD package

• STEP/IGES

• drawing PDF if you have it

• mating-part context if fit matters (CAD or key interface dims)

• A-surface definition (which faces are cosmetic)

B) Quantity + schedule

• quantity for this run

• expected reorder quantity (if known)

• must-hit delivery date and shipping destination

C) Material and performance

• material family target (ABS-like / PP-like / PC-like / elastomer)

• temperature exposure

• transparency (clear/tinted/opaque)

• chemical/UV exposure if relevant

D) Tolerances

• list critical dimensions with tolerances

• define a general tolerance for everything else

• define datums if inspection is important

E) Finish and color

• as-cast vs painted vs textured

• gloss level target

• color standard if required (Pantone/RAL)

F) Secondary ops

• inserts and thread requirements

• drilling/reaming/machining requirements

• assembly/packaging requirements

G) Quality + documentation

• First Article Inspection required? If yes, which features are critical?

• material certifications required?

• labeling/traceability requirements?

A “quote-ready” spec template (copy/paste)

Copy/paste and fill in what you know:

• Part name / revision:

• Quantity:

• Target ship date:

• Material target: (e.g., ABS-like rigid; flame-retardant; clear)

• Color: (e.g., black, Pantone ___; or “as-cast”)

• Finish: (as-cast / matte / gloss / painted; define A-surfaces)

• Critical dimensions: (list 3–10 dims that define fit/function)

• General tolerance: (state your default)

• Inserts/threads: (type, count, placement notes)

• Inspection: (FAI required yes/no; measurement method preferences)

• Packaging: (individual bagging, labeling, ESD, etc.)

This template is intentionally minimal. The goal is to remove ambiguity without building a 20-page