Slice‑Based Mold Automation in Architectural Digital Fabrication

This research documents a workflow automation developed in response to a recurring challenge in architectural digital fabrication: how to efficiently translate large volumes of custom geometry into production‑ready molds without design effort scaling out of control.

Architectural Context

Contemporary architecture increasingly demands complex, highly differentiated elements—domes, coves, cornices, columns, vaulted ceilings, clerestory assemblies, and single‑ or double‑curvature ceiling systems. These components appear most often in large civic and institutional buildings such as airports, universities, museums, libraries, and resorts.

Despite advances in additive manufacturing, mold‑based fabrication remains the dominant and most reliable method for producing these elements at architectural scale. Material cost, surface quality, repeatability, and schedule certainty continue to favor CNC‑fabricated molds assembled from layered sheet stock.

How Slice‑Based Molds Are Made

A typical mold begins as a three‑dimensional design model. To make this geometry fabricatable, thickness is introduced—both in the cutting direction and in the backing direction—to define a mold volume.

This volume is then sliced into constant‑thickness sections based on the available sheet material, commonly MDF. Each slice represents a physical layer to be CNC‑cut, nested within sheet boundaries, and later assembled into a hollow, structurally supported mold using gables and ribs.

The slicing strategy is well‑established. The challenge arises in what comes next.

Problem Statement

Each slice must be spatially offset to prepare it for nesting and cutting. In standard workflows, these offsets are applied one layer at a time. What appears trivial at the scale of a single mold becomes a major bottleneck when multiplied across dozens or hundreds of unique molds.

In large architectural projects, inefficiencies compound not per mold, but across hundreds of unique mold instances.

As projects grow, design time scales with the number of slices rather than the number of design decisions. This mismatch directly impacts production schedules and limits how much geometric complexity a project can realistically afford.

Research Objective

The objective of this research was to restructure slice‑based mold design so that offset configuration occurs once per mold rather than once per layer—allowing slice count to increase without penalizing design time.

The focus was not on inventing a new fabrication method, but on removing friction from an existing, proven one.

Method: Workflow Automation in Practice

Mastercam’s design environment supports extensibility through .NET NetHooks—compiled modules that can operate directly on design data. A custom NetHook was developed to automate vertical offset distribution across an entire slice stack.

Instead of treating each slice as an independent object, the NetHook treats the mold as a single structured system. Offset parameters are defined once, and the required transformations are applied programmatically across all layers in a controlled, repeatable manner, within seconds.

The implementation exists as a compiled library and is supported by auxiliary tooling hosted in a public GitHub repository, allowing the method to be reused, validated, and extended.

Results in Fabrication Context

• Higher slice resolutions introduced without increasing setup effort
• Rapid turnaround across large collections of unique molds
• Improved consistency and reduced manual error
• Greater confidence in meeting aggressive production schedules

The method has been applied across large sets of unique architectural molds in active production, demonstrating reliability under real delivery constraints.

Project‑Scale Insight

At architectural scale, efficiency is measured per project—not per part.

By shifting effort from slice‑level repetition to mold‑level computation, this research demonstrates how established fabrication techniques can remain viable—even as architectural ambition and geometric complexity continue to grow.

Positioning

This work represents independent applied research in architectural fabrication computation—bridging design modeling, production tooling, and project‑scale delivery through computational abstraction.

Original publication: Medium article