Sample concept study · The deliverable format
Capturing microfibres in washing-machine discharge.
An anchor-free read on which embodiment to commit to tooling — four genuinely different concepts, the honest tradeoffs, and a recommendation you can defend. This is the document a client receives at the end of the five days.
This is a sample. It shows the concept-study format on real engineering analysis. The client framing is fictionalised; the flow physics, the calculations, and the prediction-to-failure account are genuine — drawn from physical prototype testing of the Oyster filter in the portfolio.
The short version
You've decided to capture microfibres at the machine's discharge rather than at the drum, and you want an outside read on which way to build it before you spend on tooling. Good place to pause — the embodiments in front of you look similar on a slide and behave very differently in the pipe.
The headline: a naive in-line filter across the discharge pipe will work on day one and fail you by month two. The flow physics are against it. Once a fibre mat builds on a flat mesh sitting square to the flow, back-pressure rises, discharge velocity collapses, and water starts returning to the drum. That isn't a maintenance nuisance — it's a re-contamination and standing-water problem that generates warranty calls.
The embodiment that survives contact with reality is a rotating cylindrical filter driven by a small dedicated motor, mounted in the existing service compartment. It buys roughly a 10× increase in active filter area for the same envelope, and it actively sheds the fibre mat instead of accumulating it. The cost is a motor, a seal, and a shaft you must not under-build — and that last detail is the one most teams get wrong, which we have direct evidence of.
The recommendation is Concept C, with two non-negotiable design conditions attached. Concepts A and B are cheaper — and I explain precisely where they break, because you'll be tempted by them and you should know the failure mode you're buying.
What's actually being decided
Synthetic garments shed microfibres — typically 10–30 µm in diameter, hundreds of microns to a few millimetres long — every wash. A single domestic load can release on the order of a million fibres. These pass straight through the machine and into the wastewater system because nothing in a conventional machine is sized to stop them.
You want to intercept them inside the appliance, at the discharge side, so that captured fibres are contained for periodic removal rather than released; the wash function and discharge behaviour are unchanged from the user's point of view; and the solution fits a machine you already manufacture, without redesigning the cabinet or the drum.
The decision is not whether to filter — that's settled — but which physical embodiment to take to tooling. That's a design-fixation risk: the obvious answer (a fine mesh across the outlet) is the one most likely to have been prototyped first and championed internally, and it's the one the flow physics quietly kills. The job here is anchor-free eyes on the full embodiment space before the tooling spend locks you in.
What the concepts are scored against
Where a number is assumed rather than given, it's marked — confirm or correct it and the matrix updates.
| # | Criterion | Target / basis |
|---|---|---|
| 1 | Capture rate | ≥ 90% of shed microfibre mass per wash (stretch — see §05) |
| 2 | Maintenance interval | ≤ 12 interactions/year (≈ monthly), tool-free access |
| 3 | Discharge integrity | Clears water to standpipe at existing head; no return-to-drum |
| 4 | Envelope | Fits service compartment, ≤ 512 × 162 × 165 mm (assumed) |
| 5 | Unit cost | ≤ £40 at production volume (assumed) |
| 6 | Added mass | ≤ 3 kg |
| 7 | Service life | Match machine life ≈ 10–11 yr / ≈ 4,000 cycles |
| 8 | Sustainability | Production material in mainstream kerbside recycling |
| 9 | Integration | No change to cabinet, drum, or 40 mm plumbing interface |
The two criteria that do the real discriminating work are #3 (discharge integrity) and #7 (service life). Most of the concept differences collapse to two questions: how do you keep flow up as the filter loads, and what's the one component that fails first.
Four embodiments, one basis
Spanning the genuine design space from "cheapest thing that filters" to "actively self-clearing."
In-line flat mesh across the discharge pipe
A fine mesh disc (≈ 25 µm aperture) seated square across the discharge pipe. Water passes through; fibres are retained on the upstream face. The simplest possible intervention — minimal parts, nothing moving.
Cheapest to make, lowest mass, trivially within envelope, no power. If filtration were a static problem, this would be the answer.
Filtration here is not static. Each aperture behaves as a short pipe in laminar, pressure-driven (Poiseuille) flow, where throughput scales with the fourth power of effective pore radius. As fibres bridge and occlude pores, the radius falls and flow collapses non-linearly. Practical failure — noticeable flow decline — arrives at 50–75% coverage, long before total blockage. Discharge velocity then drops below what's needed to lift water to the standpipe, and water returns to the drum. Fails #3, then #2.
High — it's trivial. But it solves the wrong problem.
Flow-driven rotating cylindrical filter
Replace the flat disc with a cylinder of mesh and spin it. Filter area becomes the cylinder's circumferential surface (≈ 2πrL) instead of the pipe cross-section (πr²), and rotation throws the fibre mat outward to keep the mesh clearing itself. Here the spin is driven by the discharge flow via integral vanes — no motor.
Keeps the area and self-clearing benefits with zero added power and no electrical integration. If it worked, it would be the elegant answer.
It doesn't spin. The discharge flow carries very little usable mechanical power — a Betz-limited lift power on the order of 0.04 W at the available head and flow. Against the torque needed to rotate a wetted, fibre-loaded filter through bearing and seal drag, that leaves angular velocities far too low to clear the mat. The passive-spin idea is appealing on a whiteboard and is killed by the actual energy in the stream — not a tuning problem, an energy-availability problem.
Low, and physics caps it. Once it stops clearing, it loads up like Concept A.
Motor-driven rotating cylindrical filter Recommended
The Concept B geometry — cylindrical mesh, ≈10× area, self-clearing rotation — driven by a small dedicated motor (a 12 V DC-class motor suffices on the torque numbers). It runs only during the discharge phase. Centrifugal action forces water radially through the mesh and continuously sheds the fibre mat to hold throughput as load builds.
The only embodiment that keeps discharge velocity up across the maintenance interval, because it doesn't rely on the mat staying thin — it manages it. Mounts in the existing service compartment, integrates the emergency-drain function, fixes to the compartment floor (no CG shift, no added vibration), stays within envelope. Total annual motor run-time is only a few hours — single-digit pence per household per year, no impact on the appliance's energy rating.
- The drive shaft is the life-limiting component, and it's easy to under-build. Direct evidence: FEA flagged the shaft/crossing region as peak-stress, and on repeat physical testing the shaft snapped at exactly that predicted point — driven by a high inertial load at motor start (no ramp) and a prototype shaft weak in the relevant shear plane. Condition 1: solid moulded or metal shaft + motor soft-start. Not a printed or abrupt-start part.
- The dynamic shaft seal is the second risk. In testing, water tracked along the shaft into the motor. Condition 2: a proper waterproof bearing/seal at the penetration, validated against discharge-phase pressure — not just static.
Medium — proven to filter (see §05), and the two failure modes are known and addressed. A stronger position than an untested clean sheet.
Static cylindrical filter, no rotation
Concept C's cylindrical geometry for the ≈10× area benefit, but delete the motor, shaft, and seal. The enlarged area alone extends time-to-clog relative to the flat disc, even without active clearing.
Removes C's two hardest components and all electrical integration while keeping most of the area advantage. Cheaper, lighter, simpler, more reliable per part — a legitimate de-risking fallback if the motor/seal integration proves too costly at your volumes.
No rotation means no active mat-clearing, so it still loads up — just far more slowly than A because the area is ~10× larger. It pushes the maintenance interval toward acceptable but won't self-recover; eventually back-pressure rises and a service action is needed. Sits between A and C: materially better than A on discharge integrity and life, short of C on both.
Medium-high — fewer novel parts than C. Carry it as the explicit fallback if a review of C's seal cost comes back unfavourable.
Scored on one basis
Scored 1 (poor) – 5 (strong) against the §02 criteria. Weights reflect that discharge integrity and service life decide field survival; adjust the weights to your priorities and the ranking recalculates.
| Criterion (weight) | A · Flat | B · Flow-spun | C · Motor-spun | D · Static |
|---|---|---|---|---|
| Discharge integrity over interval (×3) | 1 | 2 | 5 | 3 |
| Service life / failure margin (×3) | 2 | 2 | 4 | 4 |
| Capture rate (×2) | 3 | 3 | 4 | 3 |
| Maintenance interval (×2) | 1 | 2 | 5 | 3 |
| Unit cost (×2) | 5 | 4 | 2 | 4 |
| Integration & envelope (×1) | 5 | 3 | 3 | 4 |
| Added mass (×1) | 5 | 4 | 3 | 4 |
| Sustainability — material (×1) | 4 | 4 | 4 | 4 |
| Weighted total | 35 | 37 | 57 | 49 |
The numbers say what the physics says: C wins on the criteria that decide whether the product survives in the field, and pays for it in cost and complexity. D is the rational fallback. A scores well only on the criteria that don't determine field survival — which is exactly how design fixation traps a team.
What I'd commit to
Commit to Concept C — the motor-driven rotating cylindrical filter — with the two production conditions treated as gating requirements, not options.
The reasoning in one line: it is the only embodiment that holds discharge velocity up as the filter loads, and discharge integrity is the criterion that separates a product from a warranty liability.
What the evidence supports
- It filters. In physical prototype testing this embodiment captured ≈ 76% of microfibre mass in a single uncontrolled pass — short of the 90% stretch target, but the shortfall traced to specific, fixable bypass paths (fibres routed around the filter at the inlet; residue left in the rig), not to the filtration principle. Set 90% as the validated production target, but plan around ~75–80% as the honest interim figure until bypass is sealed.
- Its failure modes are already located. The shaft and the seal are the two risks; both have identified fixes (solid/metal shaft + soft-start; waterproof bearing). Most concepts hide their first failure until tooling is cut — this one showed where it breaks and why, including the FEA-predicted shaft failure that then occurred exactly as predicted.
- Material: move off the PLA prototype material. PLA isn't in mainstream kerbside recycling and isn't the right mechanical choice. HDPE via injection moulding fits the duty, suits the simple part geometry, and sits in mainstream recycling — satisfying #8 and the volume economics.
- Fallback: if C's seal/motor integration cost is unacceptable at volume, fall back to Concept D — not A. D keeps the area benefit that makes the difference on discharge integrity; A does not.
A path to a tooling-ready decision
- Confirm the four assumed constraints (envelope, £40 unit cost, capture-rate basis, maintenance interval). Two matrix scores move if these change.
- Seal the bypass paths identified in testing and re-run the capture test under controlled inlet conditions — the fastest, cheapest route from ~76% toward the 90% target.
- Validate C's two failure modes as a gate: solid/metal shaft + motor soft-start against the inertial start load; waterproof bearing under discharge-phase pressure. Pass both before any tooling spend.
- Cost C vs. D at production volume. This single comparison decides whether the motor/seal complexity is justified or whether the static fallback is the commercially correct call.
- (If useful) take the chosen concept into a detailed design review — tolerances, material/seal spec, DFM for the moulded parts.
The honest small print
- This study compares mechanical embodiments for microfibre capture at discharge. It is a design-direction review — not a detailed design, FMEA, or production-readiness sign-off.
- Flow, clogging, energy-budget and torque figures are first-order engineering estimates sufficient to discriminate between concepts, and deliberately conservative (minimum-case clogging, Betz-limit lift power) so the ranking is robust rather than optimistic.
- Capture rate, the prediction-to-failure shaft account, and the casing-life result are drawn from real physical prototype testing of the recommended embodiment.
- Four constraints are assumed and flagged; the matrix recalculates transparently when they're corrected.
- Independence: produced with no commercial stake in any concept and deliberately outside the reviewer's employer's technical domain. The purpose is anchor-free judgement — naming the failure mode you're buying, not selling the exciting option.
- Ownership: all concepts, analysis and recommendations are the client's to use, develop and commercialise on delivery. VG Engineering retains no rights over the outcomes and asserts no claim on resulting IP. This sample is anonymised and carries no real client information.
This is what your problem gets — in five working days.
Concepts sketched, tradeoffs named, a recommendation you can defend — in the format you've just read. And if a concept earns it, the road runs further: developed design, 3D-printed functional models, a full working prototype.