Steel Tongue Drum Capstone Print Packet

Generated: 2026-05-08

Packet folder: `/mnt/c/Users/Tony/Documents/GitHub/steel-tongue-drum`

File Map

design.md

Project intent, catalog metadata, assumptions, and validation plan.

Steel Tongue Drum Design

Generated: 2026-05-08

Intent

Create a traceable root-mode packet for a tuned steel tongue drum. The packet documents the first-pass geometry from `steel-tongue-drum-design-table.xlsx`, the manufacturing assumptions that can move pitch, and the validation workflow needed to turn estimates into a tuned instrument.

This design is intentionally conservative about claims. The workbook formulas size tongues from a cantilever model, but the final instrument must be tuned by measured cutting and strike tests.

Source Artifacts

Workbook Inputs Used

Governing Model

Use the cantilever beam model only as an initial layout estimate:

f = K_steel * t / L^2
L = sqrt(K_steel * t / f)

Where:

• `f` is target frequency in Hz.
• `K_steel` is the workbook steel constant, currently 32098 for the selected steel row.
• `t` is steel thickness in inches.
• `L` is free tongue length in inches.

The model assumes an ideal fixed-free tongue. A real steel tongue drum departs from that ideal because the tongue remains part of a stressed plate/shell, is cut by a finite kerf, may be heat affected, may retain burrs, and is coupled to the vessel body and support system.

Target Tongue Table

The values below are first-pass estimates from the workbook formula family. Treat every length as a starting cut length with trim allowance, not a final tuned dimension.

Shell / Helmholtz Estimate

The workbook includes a first-pass vessel estimate using a 12 in diameter, 6 in high partial volume assumption and a 2.5 in port. This is useful for checking whether the shell may reinforce or color the tongue field, but it should not be treated as a final resonator design.

For this packet, the shell resonance is documented as a support variable:

• Verify actual internal volume after forming, welding, or purchased shell selection.
• Measure the port diameter after deburring and finish.
• Compare struck tongue spectra mounted and unmounted.
• Record whether final mounting shifts pitch or sustain.

Empirical Tuning Workflow

1. Rough-cut the tongues long or conservative, with rounded stress-relief termini.

2. Deburr both sides without thinning the root unpredictably.

3. Strike each tongue with the same mallet and support condition.

4. Log measured Hz and cents error in `validation.csv`.

5. Remove material in controlled passes, favoring tip/length correction first.

6. Re-measure after each tuning pass and after final mounting.

7. Re-check drift after finish cure and 24-72 hours of rest.

Manufacturing Assumptions

• Steel grade is not locked. Mild low-carbon sheet is the default planning case; stainless is acceptable only after revisiting K, cutting method, corrosion plan, and work hardening behavior.
• Thickness must be measured at the actual blank, not assumed from nominal sheet gauge.
• Laser cutting is fast but risks heat affected zones and local stress changes.
• Waterjet cutting reduces heat risk but can leave taper, abrasive edge texture, and local burrs.
• CNC milling can control slot geometry but needs serious fixturing and coolant/chip planning on thin steel.
• Every inside slot end needs a smooth radius to reduce crack initiation.

Unresolved Tuning Assumptions

• Actual steel grade, modulus, density, hardness, and residual stress.
• Actual kerf after the chosen cutting process.
• Whether the blank is flat sheet, preformed shell, salvaged vessel, or welded assembly.
• Final support/isolation scheme.
• Mallet hardness and strike point standard.
• Finish mass and corrosion-protection method.
• Whether the bass tongue and high tongues remain balanced in volume and sustain after final trimming.

Acceptance Criteria

The packet becomes public-review ready only after:

• Each tongue has target Hz, measured Hz, cents error, mallet condition, support condition, and environment logged.
• Rough-cut, deburr, first strike, tuning pass, final mount, and drift check rows are complete.
• At least one process note records actual kerf, burr condition, and heat/warp observations.
• `validation-report.md` is updated from measured data rather than assumptions.

bom.csv

Starter bill of materials with part categories, quantities, drawing refs, and notes.

sourcing.csv

Supplier/search tracker with specs, price/date fields, lead time, substitutes, and risks.

cut-list.csv

Rough/final stock sizes, material, grain/orientation, operations, yield, and offcuts.

drawing-brief.md

Manufacturing drawing and technical product sketch brief.

Drawing Brief

Required Drawing Set

1. Top layout with all tongue slots, centerline, port, rim, and orientation mark.

2. Section view through shell showing top thickness, shell height, port, support feet, and rim/edge treatment.

3. Detail view of a tongue root with slot-end radius, kerf, burr limit, and stress-relief note.

4. Fixture view showing temporary tuning support and final isolation feet.

5. Process coupon drawing for kerf and HAZ validation.

Critical Dimensions

• 12.0 in body diameter assumption.
• 6.0 in shell height assumption.
• 0.079 in steel thickness assumption.
• 0.040 in kerf assumption.
• 1.0 in tongue width assumption.
• 2.5 in port diameter assumption.
• Tongue lengths from `design.md` are derived estimates and must be marked "rough-cut; tune by measurement."

Drawing Rules

• Units: inches with metric equivalents where useful.
• Include title block, revision, date, material, finish, and "not final tuned dimensions" note.
• Show datum A as top center, datum B as rim plane, datum C as orientation line.
• Call out all inside slot ends as smooth radiused transitions.
• Do not publish as final manufacturing drawing until validation data exists.

assembly-manual.md

Shop-facing sequence, tools, fixtures, safety, tuning, finishing, and maintenance notes.

Assembly Manual

Safety

Steel cutting, deburring, and tuning can create sharp edges, hot work, abrasive dust, and springy tongues. Wear eye protection, gloves when handling raw cuts, and hearing protection during cutting or repeated striking. Do not strike an unfinished blank with sharp burrs in a way that can throw chips.

Tools

• Calipers and steel rule.
• Layout template or CAD print.
• Cutting process: waterjet, laser, CNC mill, or carefully controlled manual process.
• Deburring tools, small files, abrasive pads, and inspection light.
• Tuner or spectrum analyzer with repeatable microphone placement.
• Medium-soft mallets reserved for validation.
• Temporary support cradle and final isolation feet.

Build Sequence

1. Confirm the actual steel grade, thickness, shell diameter, shell height, and port diameter.

2. Update the workbook or `design.md` assumptions if measured thickness differs from 0.079 in.

3. Mark the top centerline, tongue layout, port, and any logo or orientation mark.

4. Cut the port first if it helps workholding; deburr and record actual diameter.

5. Rough-cut tongue slots using rounded slot ends and the documented kerf.

6. Inspect for heat discoloration, slot taper, burrs, and oil-can distortion.

7. Deburr both sides, preserving root radius and avoiding uncontrolled thinning.

8. Place the drum on a temporary isolation support and record first-strike measurements.

9. Tune iteratively, one tongue at a time, recording measured Hz and cents error.

10. Install final feet, edge trim, and finish only after pre-finish tuning is documented.

11. Re-check pitch after final mounting, after finish cure, and after a 24-72 hour drift rest.

Tuning Notes

For this first prototype, do not chase a perfect tuner reading from one strike. Use the same mallet, strike point, support, and microphone distance. Record at least three stable readings per tongue where possible.

If a tongue is sharp after rough cutting, lower pitch by increasing effective length or removing material in a way that reduces stiffness. If a tongue is flat, options are limited; shorten the effective tongue or accept a revised target. This is why rough-cut length should stay conservative.

Photo Checkpoints

• Raw shell and thickness measurement.
• Layout before cutting.
• Slot kerf sample coupon.
• Rough-cut top before deburr.
• Deburred tongue roots.
• First-strike measurement setup.
• Tuning pass with tool marks visible.
• Final mount and underside/support detail.

validation.csv

Target/measured values, tolerance, environment, result, and tuning/build action log.

supplier-rfq.md

Supplier email/request-for-quote starter.

Supplier RFQ Draft

Subject: RFQ - steel tongue drum blank / cutting service

Hello,

I am preparing a prototype steel tongue drum and need either a steel shell/blank, cutting service, or both. The project is a struck musical instrument, so material consistency, low warp, clean edges, and documented kerf matter more than cosmetic speed.

Requested Quote

• Quantity: 1 prototype, with option for 2-3 follow-up blanks.
• Material: low-carbon mild steel preferred for the first prototype.
• Thickness target: near 0.079 in; please quote actual available thickness and tolerance.
• Body target: 12 in diameter, 6 in high shell/vessel or compatible top/shell blank.
• Cutting: tongue slots and 2.5 in port from supplied DXF/SVG when finalized.
• Edge condition: quote as-cut condition, deburring options, and expected kerf/taper.
• Process options: waterjet, fiber laser, CO2 laser, CNC mill, or recommended alternative.

Please Include

• Material grade and data sheet.
• Actual thickness tolerance.
• Cutting kerf estimate.
• Heat affected zone or warp expectations.
• Deburring/edge-finishing options.
• Unit price and lead time.
• Whether nested test coupons can be included for kerf and tuning trials.

Substitutions are acceptable only if the grade, thickness, cutting process, and expected edge condition are clearly identified.

visual-bom-brief.md

Art direction for an image-forward visual BOM.

Visual BOM Brief

Goal

Create a one-page visual BOM that lets a fabricator or reviewer understand the steel tongue drum parts without reading every CSV.

Layout

• Header: Steel Tongue Drum, STD-001, quote date, estimated total.
• Main image: top layout drawing from `drawings/steel-tongue-drum-layout.svg`.
• Row images: shell blank, tongue field, port, rubber feet, mallets, finish, deburring tools.
• Columns: item id, part, visual, quantity, required spec, cost basis, tuning risk.

Required Labels

• "Derived estimate" for any tongue geometry.
• "Verify before purchase" for price and supplier information.
• "Measured data required" near tuning-critical parts.

Shot Priority

Until real photos exist, use clean line drawings and explicit placeholders. Do not use atmospheric concept images as proof of manufactured geometry.

steel-tongue-drum-starter.wl

Wolfram starter for physics, optimization, visualization, and validation.

(* Steel Tongue Drum Wolfram starter *)

ClearAll["Global`*"];

(* Inputs from steel-tongue-drum-design-table.xlsx *)
bodyDiameterIn = 12.0;
shellHeightIn = 6.0;
steelThicknessIn = 0.079;
kerfIn = 0.040;
tongueWidthIn = 1.0;
portDiameterIn = 2.5;
a4Hz = 440.0;
kSteel = 32098.0;

targetMidis = <|
  "Bass" -> 50,
  "T1" -> 57,
  "T2" -> 58,
  "T3" -> 60,
  "T4" -> 62,
  "T5" -> 64,
  "T6" -> 65,
  "T7" -> 67,
  "T8" -> 69,
  "T9" -> 72,
  "T10" -> 74
|>;

midiToHz[m_] := a4Hz*2^((m - 69)/12);
tongueLengthIn[f_] := Sqrt[kSteel*steelThicknessIn/f];
centsError[measured_, target_] := 1200*Log2[measured/target];

targetTable = KeyValueMap[
  {#1, #2, midiToHz[#2], tongueLengthIn[midiToHz[#2]], 25.4*tongueLengthIn[midiToHz[#2]]} &,
  targetMidis
];

Grid[
  Prepend[targetTable, {"Tongue", "MIDI", "Target Hz", "Length in", "Length mm"}],
  Frame -> All
]

(* Validation example: replace with measured data from validation.csv. *)
measuredExampleHz = <|"T8" -> 440.0|>;
KeyValueMap[{#1, centsError[#2, midiToHz[targetMidis[#1]]]} &, measuredExampleHz]

(* Simple sensitivity plot: pitch versus tongue length for the selected steel thickness. *)
Plot[kSteel*steelThicknessIn/l^2, {l, 2.0, 4.5},
  AxesLabel -> {"Tongue length (in)", "Frequency (Hz)"},
  PlotLabel -> "First-pass cantilever estimate only"
]

README.md

Project artifact.

Steel Tongue Drum

Root-mode engineering packet for a tuned steel tongue drum: a slit-tongue steel idiophone whose final tuning must be measured and corrected on the actual prototype.

This repo is part of the [tonykoop/instrument-maker](https://github.com/tonykoop/instrument-maker) catalogue and uses the `instrument-maker-v4` root packet shape.

What this is

This packet turns the existing `steel-tongue-drum-design-table.xlsx` into a build-ready documentation set for a circular steel tongue drum. The workbook provides first-pass geometry, material constants, target MIDI notes, a rough Helmholtz body estimate, and tongue length formulas.

The important caveat: steel tongue drums are empirical instruments. The cantilever equation is useful for rough layout, but final pitch depends on steel batch, residual stress, slot kerf, heat affected zone, burrs, mounting isolation, mallet hardness, finish mass, and the way each tongue is trimmed. This repo is therefore not "tuned" until measured prototype data exists in `validation.csv`.

Current design baseline

Packet map

• `design.md` - design intent, assumptions, formulas, tuning workflow.
• `family-spec.csv` - root-mode family row for the current steel tongue drum.
• `bom.csv`, `sourcing.csv`, `cut-list.csv` - purchasing and shop prep.
• `assembly-manual.md` - rough build, cutting, tuning, finish, and mount steps.
• `validation.csv` - tuning and process validation rows, including rough-cut, deburr, first strike, tuning pass, final mount, and drift check.
• `drawing-brief.md`, `drawings/` - drawing requirements and starter layout SVG.
• `cad/`, `cnc/`, `jigs/` - SolidWorks/CAM/fixture briefs.
• `risks.md` - acoustic, structural, ergonomic, supply, and finish risks with tests.
• `capstone-deck.md`, `print-packet.md`, `print-packet.html`, `site/index.html` - review and presentation layers.

Readiness

This packet is private-prototype ready. It should stay private for public technical review until a rough-cut prototype is struck, tuned, mounted, drift-checked, and logged with measured Hz and cents error.

License

[CC BY 4.0](LICENSE) - see LICENSE for details.

family-spec.csv

Project artifact.

jig-decision.md

Project artifact.

Jig Decision

Decision

Use a two-stage jig approach:

1. A non-marring temporary tuning cradle that supports the shell consistently while leaving the tongue field free.

2. A final isolation-foot layout that is re-measured after installation.

Why

The support condition is part of the instrument. A rigid clamp, bench pad, hand hold, rubber foot, and display stand can all change damping and perceived pitch. Separating the temporary tuning cradle from the final mount makes that shift visible in `validation.csv`.

Requirements

• Cradle must locate the shell repeatably without contacting vibrating tongues.
• Contact points should be soft enough to prevent rattles but stable enough for repeatable measurements.
• Fixture should expose the port and underside for inspection.
• Final feet must be installed before the final-mount validation phase.

Rejected Options

• Freehand holding during tuning: too variable.
• Clamping the shell rigidly to a bench: likely to damp shell modes.
• Tuning only after final finish: hides process causes and makes correction harder.

photo-shotlist.md

Project artifact.

Photo Shotlist

• Raw steel shell or blank with label/spec visible.
• Caliper reading of actual steel thickness.
• Layout on top surface before cutting.
• Kerf coupon or test cut close-up.
• Rough-cut slots before deburring.
• Deburred slot root close-up with light showing radius.
• Temporary tuning support setup.
• Tuner/microphone placement during first strike.
• Tuning pass in progress with tool and tongue identified.
• Final feet or isolation mount underside.
• Finished top view with all tongues visible.
• Detail of port, rim, and edge trim.

resources.md

Project artifact.

Resources

Local Packet Sources

• `steel-tongue-drum-design-table.xlsx` - workbook source for dimensions, K constants, target notes, and formulas.
• `design.md` - narrative design basis and assumptions.
• `validation.csv` - required measurement log.
• `drawings/steel-tongue-drum-layout.svg` - starter layout drawing.

Instrument Maker References

• `instrument-maker-v4/references/new-instruments-v4.md` - steel tongue drum entry.
• `instrument-maker-v4/references/acoustic-models.md` - cantilever beam formulas and guard rules.
• `instrument-maker-v4/references/manufacturing-and-cnc.md` - machining and fixture guidance.
• `instrument-maker-v4/references/sourcing-and-production.md` - sourcing, RFQ, cut list, validation guidance.

Sister Repos

• `tongue-drum` - wooden slit tongue drum done-bar reference and DoE pattern.
• `glockenspiel` - metal bar idiophone sibling concept.
• `handpan` - steel shell idiophone sibling, but different governing model.

Measurement Tools To Use

• Chromatic tuner or spectral analysis app.
• Calipers for thickness, kerf, port, and slot checks.
• Consistent microphone placement.
• Temperature/humidity record for each tuning session.

risks.md

Project artifact.

Risks

Acoustic

Structural

Ergonomic

Supply

Fit / Finish

validation-report.md

Project artifact.

Validation Report

Current Status

No measured prototype data is present yet. The packet is documentation-complete enough to guide a prototype, but it is not public-review ready as a tuned instrument.

What Has Been Validated

• A4 reference and equal-temperament target method are documented.
• Workbook assumptions have been surfaced in `design.md`.
• The required validation workflow exists in `validation.csv`.
• Manufacturing risks for laser, waterjet, CNC, deburring, support, finish, and drift are documented.

What Remains Open

• Actual steel grade and thickness.
• Actual cutting kerf and edge condition.
• First-strike Hz for every tongue.
• Cents error before and after tuning.
• Pitch shift after final mounting.
• Pitch drift after finish cure and rest.

Readiness Call

Private-prototype ready: yes.

Public-review ready: no. Keep the repo private or clearly marked as unvalidated until measured rough-cut, deburr, first-strike, tuning-pass, final-mount, and drift-check data exists.