Straight answers, from application engineers

A tank of water-based cleaning solution is driven with high-frequency sound. The sound alternately stretches and compresses the liquid; during the stretch phase the liquid is literally torn apart, forming microscopic vacuum bubbles, and in the compression phase those bubbles collapse with enormous local energy. Millions of these implosions per second, against every wetted surface, is what does the cleaning.

Because the work is done by the liquid itself, it cleans wherever the liquid reaches — galleries, threads, blind holes and internal passages that no brush, jet or blast line can touch — without anything contacting the part.

The bubble is the tool. Each cavitation bubble grows during the low-pressure half of the sound cycle, then implodes in the high-pressure half, producing an intense micro-jet against the nearest surface. The energy of one implosion is tiny; the cleaning power comes from the number — the field contains millions of events per second.

Bubble size is set mainly by frequency: lower frequencies give bubbles more time to grow per cycle, so they are larger and implode harder. Higher frequencies produce smaller, gentler, more numerous bubbles. That single relationship drives most machine selection decisions.

Correctly parameterised, no — and the proof is that aerospace and medical industries specify it for components that must not be abraded. But the honest answer has more texture: low frequencies can mark soft substrates like aluminium and brass if exposure is long, because larger bubbles implode harder. The damage risk rises with exposure time, which is why delicate work runs higher frequencies, moderated power, or sequenced frequencies with short exposures at the aggressive end.

This is exactly what the cleaning trial resolves: your material, your finish, the real parameter set — proven on your parts before production ever runs.

Two different jobs, and it does both. Insoluble particles — swarf, dust, polishing media, carbon fines — are physically dislodged by the implosions and carried into suspension. Soluble and reactive soils — oils, greases, fluxes, salts — are dissolved by the chemistry, with cavitation constantly stripping away the saturated boundary layer at the surface so fresh solution keeps reaching the soil.

That second mechanism is why ultrasonics accelerates chemistry so dramatically: without it, dissolution slows to the rate at which soil can diffuse through a stagnant layer. With it, the layer never gets to form.

Each fails a different way. Solvent soaking barely touches coked or polymerised deposits — once contamination has cross-linked under heat, it no longer dissolves in what created it. Pressure washing cannot reach internal geometry and can drive contamination deeper into it. Blasting removes parent material, rounds edges, and embeds media that later migrates out of soft surfaces and circulates through the machine you just rebuilt.

Ultrasonics is the only method that cleans geometry it cannot see while leaving dimensions and surfaces exactly as they were — which is the whole requirement on gauged, sealed or coated components.

Cycle time follows the contaminant, not the clock. Light oils release in minutes because the chemistry and cavitation work on a thin, accessible film. Heavy coked carbon may need staged cycles of 20–45 minutes with inspection between — the deposit comes off in layers, and the bath needs filtration between stages so removed soil does not redeposit.

A machine cannot promise a number; a trial can. We quote cycle times from documented trial results on your parts, which is the only honest way to do it.

Frequency sets the size and energy of the cavitation bubbles. At 20–28 kHz bubbles are large and implode hard — right for gross deposits on robust steel and cast components: engine parts, mining hydraulics, exchangers. At 40 kHz and above the bubbles are smaller, more numerous and gentler — right for precision surfaces, fine detail, and soft or delicate substrates.

Where a load carries both heavy soil and delicate features, sequencing frequencies works better than compromising on one: a short, aggressive low-frequency pass for the bulk, then higher frequency for the finish — the short exposure keeps the aggressive phase below the damage threshold of the substrate.

A watt is a rate, not an amount — one joule per second. Cleaning is done by accumulated energy delivered into the bath over the cycle, so a nameplate figure by itself tells you very little: what matters is how many energetic cavitation events actually occur at your part.

Delivered watts per litre is the useful working measure, and even that does not scale simply — laboratory results at small volume cannot be linearly scaled to a large industrial tank, because bath geometry, transducer placement and load all change how the energy is used. This is precisely why we size industrial systems from trials and experience rather than a formula, and why an underpowered large tank never quite develops a working field.

Sized to the process, not the part: the largest basket load with room for solution to circulate around it, plus the working depth your components need. Oversizing wastes heating and chemistry every single day; undersizing forces part-by-part cleaning that erases the labour saving.

Our standard range runs 3.3 to over 42,000 litres across seven families, so the recommendation can follow the application rather than the inventory — and purpose-built systems cover what the catalogue does not.

Automation is a specification decision, not a quality grade. Varied, low-volume work is usually best served manual — flexible, inexpensive, little to go wrong. Semi-automation (basket lift, agitation platforms, timed transfer) removes the heavy and repetitive parts of the job while keeping the operator's judgement in the loop. Full PLC automation earns its cost where cleanliness is audited: every part demonstrably sees an identical cycle and the log proves it.

Decide before manufacture. Automation is engineered into the machine — lifts, transfer, control architecture — not bolted on afterwards.

Yes — heat exchanger bundles, large fabrications, unusual geometries and duty cycles are engineered to requirement: tank, transducer arrays, heating, filtration and handling designed around the site and the components. Large engineered systems are commissioned against a documented test program, with full FAT documentation for involved cleaning solutions.

Where the process is not already proven, we scope a trial: parts carrying the real contamination come to our laboratory, and we clean them and document everything — cycle time, chemistry, concentration, temperature, frequency, sequence — and you see the result and the data before committing capital. The machine is then specified around a verified process rather than a hope.

If ultrasonics is the wrong answer for your parts, the trial shows that too, and we will tell you. A supplier who will not run your parts first is asking you to carry the process risk for them.

A transducer array is a resonant system, and its resonant frequency moves constantly — with solution level, temperature, basket load, and the slow ageing of the piezo ceramics. The mechanical resonance is extremely sharp: drift even slightly off it and the energy delivered into the bath collapses, while the reflected energy heats the transducers instead.

The generator senses the array continuously and retunes in real time, holding the drive on resonance for the entire cycle — so the last basket of the shift cleans like the first, and the transducer array is protected from the off-resonance heating that is the classic cause of early failure.

Instead of holding one fixed frequency, the generator deliberately modulates around the centre frequency across a defined bandwidth, at a defined sweep rate. A fixed frequency in a tank of fixed geometry sets up standing waves — stationary zones of high and low intensity — so parts clean in stripes, and bubbles get trapped at pressure nodes instead of doing work.

Sweep keeps the field moving, evens the intensity across the load, and avoids exciting resonances in the parts themselves — which matters for thin sections and delicate assemblies. It also continuously degasses the bath as a side effect.

Fresh solution is saturated with dissolved air, and it matters more than most people expect: gas-filled bubbles are cushioned — they absorb ultrasonic energy and collapse softly, doing almost no work. Until the bath is degassed you are paying for cavitation you are not getting.

Heat helps enormously — hot water holds far less dissolved gas than cold, which is one more reason a properly heated bath outperforms. On current platforms, continuous frequency modulation degasses the solution as it runs, so no separate degas program is needed — but the principle still applies after every fresh fill or top-up: give the bath time at temperature before judging the clean.

The generator delivers the same energy into the tank regardless of changes in liquid level, load or temperature — instead of letting output wander as the impedance of the system shifts. Results stay consistent through the shift, and the electrical stresses that variable loading would otherwise create in both generator and transducers are avoided.

Over-temperature, over-current, short-circuit and no-load protection shut the drive down safely under abnormal conditions. No-load protection is the one most often missing on cheap generators and the most expensive to lack — drive a generator into a failed or disconnected transducer and you lose the generator too.

Each generator monitors its own temperature and shuts down safely to cool if it exceeds limits. On multi-generator systems you can isolate a single generator from the panel and keep running at reduced power until a replacement arrives — a fault becomes reduced throughput, not a stopped line.

Honestly — yes, eventually. The same implosions that clean your parts also act on the radiating surface itself, and over years of service cavitation erosion of transducer faces is a physical reality no manufacturer escapes. What you can control is how fast: correct power density, proper bath maintenance, and not running aggressive chemistry hotter than the process needs all extend life substantially.

What matters commercially is what happens then: on our platforms transducers and resonator assemblies are individually replaceable service items with spares held in Australia — not a reason to replace the machine.

Connected machines support remote diagnostics, programming, control and software updates. Most faults can be identified — and many resolved — without a site visit, while the machine is still in front of the operator. Parameters like frequency and power are monitored live through the HMI, which also makes drift visible before it becomes a failure.

Across Australia, New Zealand, PNG and Indonesia, this is the single biggest difference in what ownership actually costs: a fault is a remote session, not flights and accommodation.

Most general detergents foam — and foam is fatal to ultrasonics. Entrained air and surface foam absorb and scatter the sound field, so cavitation intensity collapses exactly where you need it. Ultrasonic chemistry is formulated low-foam, with the wetting and soil-suspension behaviour that cavitation depends on.

It is the factor most often blamed on the machine: a good tank with the wrong chemistry looks like a bad tank.

By soil and substrate together. The formulation must attack the contaminant — alkaline for oils, greases and carbon; acidic-side descalers for mineral scale; specialised systems for inks, resins and fluxes — while the material underneath comes out untouched. Inhibitor packages are what make a strongly alkaline product safe on aluminium; without them the same pH etches it.

We toll-manufacture application-specific formulations and supply Brulin's internationally approved range; the trial pairs the right product with your parts, and the delivered process names product, concentration and temperature.

No — and running rich is usually a symptom, not a solution. Above the working concentration you gain little speed, spend more per litre, drag more chemistry out on parts, need more rinsing, and pay more to dispose of both bath and rinse water. If a chemistry only works above its recommended concentration, that is evidence it is the wrong chemistry for the soil — not evidence you need more of it.

Typical working range is 5–30% depending on soil and duty; the trial establishes the honest figure.

Temperature is the biggest single lever in the tank, and it works three ways at once: it activates the chemistry, it drives dissolved gas out of the bath (hot water holds far less than cold, so cavitation intensifies), and it shifts the liquid's physical properties — surface tension, viscosity, vapour pressure — toward better cavitation, up to an optimum that depends on the solution.

Many industrial chemistries do their best work at 60–70 °C, and heavy carbon wants the upper end. A machine with real heating capacity, properly insulated, is process equipment — not a comfort feature.

Three things exhaust a bath: oil loading, suspended solids, and chemistry depletion. The machine defends against all three — weir and interceptor systems continuously separate floating oil, filtration removes solids between and during cycles, and dosing keeps concentration in range. A defended bath is reclaimed many times before replacement; an undefended one dies in days and takes cleaning quality down quietly on the way.

Bath life is an operating cost. It is designed into the machine, not left to housekeeping.

Inhibited formulations are — that is precisely what the inhibitor package exists for. The same caution applies mechanically: soft substrates prefer higher frequencies and moderated exposure at low frequency, because larger bubbles implode harder against soft surfaces.

Both halves — chemical and mechanical compatibility — are confirmed together in the trial, on your parts, before production ever runs.

Yes. Chemical products are covered by their own product warranty, alongside the equipment warranty on the machine — and because we specify chemistry and machine together from the trial, responsibility for the process result does not fall into a gap between two suppliers. One supplier, one process, one accountable party.

Every machine runs a load and cavitation test before it ships — a minimum of four hours for standard machines, with ultrasonics, heating, controls and safety interlocks verified under real operating load rather than switched on briefly and boxed. The benchmark report travels with the machine and becomes your commissioning baseline.

Larger industrial systems are tested for days — sometimes weeks — as each process stage is run, checked and verified in turn. For large and involved cleaning solutions, full FAT (Factory Acceptance Test) documentation is provided, so the system is proven against its specification before it leaves the floor, witnessed if you wish to attend.

Standard equipment warranty is 12 months on electrical and mechanical components, against defects in materials and workmanship. Fabrication — the tank and stainless structure — is warranted for 24 months, which reflects what heavy 316 construction can honestly carry. Chemistry is covered separately under its own product warranty.

Behind the terms sits what makes a warranty usable: spares held in Australia, every wearing component individually replaceable, and over-the-air diagnostics so a claim starts with a remote session, not a site-visit queue.

Less than most industrial plant, and most of it is bath discipline: keep filtration and oil separation working, watch concentration, and drain settled solids from the sump rather than letting them blanket the tank floor — a layer of settled contaminant on a radiating surface insulates it and makes the transducers work harder for less cleaning.

The machine itself asks for little: periodic inspection of heaters, pumps and seals, and attention to the benchmark figures from the despatch test at each service so drift is caught early.

A service call and a part number — not a capital decision. Transducers, resonators, generators, heaters, pumps and sensors are all individually replaceable, spares are held in Australia, and connected machines are diagnosed over the air first. On multi-generator systems, a failed generator can be isolated from the panel and the machine kept running at reduced power until the replacement arrives.

PLC-controlled machines store recipes per part number and log every cycle — time, temperature, chemistry, program, faults. The evidence your auditor wants is produced as a by-product of running the machine, and for validated environments the delivered process documentation from the trial closes the loop from specification to daily record.

Yes — RS485 MODBUS RTU, PLC/HMI and I/O interfaces let the ultrasonics run inside an automated, logged sequence. Parameters, status and fault states are readable without opening a cabinet, generators can be switched from the HMI or hardwired I/O, and frequency and power can be monitored live — the basis of line automation and validated multi-stage processing.