← All work

Product development

Engineering a flush valve from first principles

Barros & Moreira, a Portuguese manufacturer of flushing systems, needed a dual-flush valve that meets the EN 14055 standard for both its full and reduced flush. We took the product through the hard technical stretch of development: a calibrated hydraulic model, rapid 3D-printed prototypes and a designed experiment run on our own test bench.

The challenge

EN 14055 sets acceptance windows for both the full and the reduced flush volume of a cistern. Those volumes depend on hydraulic effects, such as buoyancy, surface tension and the float’s piston effect, that are hard to predict from first principles alone. A mechanism that gets them wrong either fails the standard or delivers the wrong amount of water on every flush, for the life of the product.

What we did

  • Modelling and prototyping: a parametric hydraulic model calibrated on the bench to steer the design, CAD iterations tracked version by version, and 3D-printed prototypes tested in water within days.
  • Root cause, DOE and testing: a root-cause analysis and a 2³ factorial experiment run on our own bench, then a regression model and an optimisation against the standard.

Approach

1. Model the valve, then prototype it

We started from the functional targets, the full and reduced flush volumes the standard requires, and the mechanisms that could reach them. To steer the design rather than guess, we built a parametric hydraulic model of the valve covering volumes, flow rates, buoyancy, weight and timing, and calibrated it against bench measurements. It predicts how each geometric change moves the discharged volume (raise the float, narrow a tube, shift the equilibrium point), so design decisions rested on physics rather than trial and error.

Complete flush: discharged volume vs time02468012345Time (s)Volume (L)
The parametric hydraulic model (line) calibrated against bench measurements (points): discharged volume over a complete flush. Once calibrated, it predicted the effect of each design change before a part was printed. Figures simplified for publication.

We then iterated in CAD and 3D-printed every candidate (tube diameters, float volumes, seal cones, counterweights) to test it in water within days, keeping what worked and setting aside what didn’t. After several prototype generations the full flush was on target and the design was converging on fewer, simpler parts.

Bagged and labelled 3D-printed prototype parts from successive design iterations, next to two assembled valve mechanisms and a laptop.
Successive 3D-printed prototype iterations, each bag holding one configuration under test (pin diameter, hole diameter), labelled and tracked generation by generation.

2. Find what controls the flush

When early prototypes fell short of the target flush, we ran a root-cause analysis of every contributor to the discharged volume, from flow rates, buoyancy and weight to dynamic effects such as surface tension and the float’s piston effect, and tested the suspects one by one.

To pin down what actually controls the flush, we designed a 2³ factorial experiment with three factors (tank water level, full-flush window, reduced-flush window) over eight runs. We ran every test ourselves on a bench rig, timing each flush and measuring the discharged volume and flow.

A flushing valve mechanism mounted inside a clear acrylic enclosure on a wooden test stand, plumbed to a water supply line.
Our bench rig, with a clear enclosure around the valve so every flush could be watched and timed directly, plumbed to a live water supply.
Effect on discharge volume (L)Full flushReduced flushWater level+1.38+1.23Full window+0.53+0.02Reduced window+0.08+0.03
Main effects from the 2³ experiment. Water level dominates both flushes; the full-flush window affects the full flush only. Higher-order interactions were negligible, so the system is essentially linear.

3. Optimise against the standard

Because the effects were essentially linear, we fitted a regression model to the orthogonal design and used it to run thousands of virtual flushes. Scoring each setting by its distance to the standard’s acceptance windows, a Monte Carlo search found the setting closest to the standard.

Normative space: full vs reduced discharge11.82.53.343456786 / 3 L4 / 2 LoptimumoptimumFull discharge (L)Reduced discharge (L)
Each flush plotted as full vs reduced volume against the standard's acceptance windows. From 3,000 virtual settings the model found a single setting that lands inside both, satisfying the full and reduced flush at once.

The result is a product decision: a single factory setting can satisfy both the 6/3 L and 4/2 L variants. That means fewer product codes, simpler assembly and a path to removing the adjustment windows altogether.

Outcome

Barros & Moreira has a prototype tested against EN 14055 and a calibrated model that lets its team explore new settings without going back to the bench for every change. The same approach, model first and then prove it on the bench, now runs through the other products we are developing with them.

Facing a similar challenge?

A first conversation takes about 30 minutes and costs nothing.