A batch of turbine blades goes through the same ultrasonic cleaning machine, the same cycle time, the same chemistry. Sample A comes out clean. Sample B still carries carbon residue in a cooling passage. Sample C shows an oil film at a thread root that should have been removed in the first minute of cavitation. The batch fails inspection—not all of it, but enough of it to trigger 100% re-inspection, selective rework, or scrapping of parts that are otherwise sound. The ultrasonic cleaning machine was running. The result was not consistent.
This pattern is familiar in aerospace component manufacturing, and it is widely misdiagnosed. When cleanliness varies within a batch, the instinct is to blame the cleaning chemistry, the operator, or the contamination level of the incoming parts. Those factors matter. But in a properly configured system, they are not the source of the variance. The variance comes from the cleaning system itself—from where cavitation is strong, where it is weak, and whether every part in the basket experiences the same acoustic conditions.
The most common architecture for ultrasonic cleaning is also the least suited to batch consistency: a single tank that performs degreasing, cleaning, and rinsing in the same volume of liquid.
The first problem is cross-contamination. As blades release cutting oils, carbon particles, and metal fines into the bath, the liquid becomes progressively dirtier. Parts cleaned early in the cycle emerge from relatively clean solution; parts cleaned later are immersed in a bath that has already accumulated the contamination removed from their predecessors. Contaminants can re-deposit onto blade surfaces before the cycle ends, producing a cleanliness gradient that follows the order of loading rather than the cleaning capability of the machine.
The second problem is acoustic non-uniformity. Even in a tank that “has cavitation," the acoustic field is rarely uniform. Standing waves create alternating zones of high and low cavitation intensity—sometimes called peak-and-valley zones. A blade positioned in a high-energy zone receives aggressive cleaning; a blade a few centimeters away in a low-energy zone receives far less. The machine is performing to specification in the aggregate, while individual parts within the batch are not. For a turbine blade with film-cooling holes measured in fractions of a millimeter, that difference is the difference between a clean passage and a blocked one.
Batch-to-batch and within-batch consistency is not a feature that can be added to a tank after the fact. It is an outcome of three engineering decisions: how the acoustic field is distributed, how the process is staged, and how the part is held.
Acoustic uniformity has to be designed, not assumed. Transducer placement determines whether cavitation reaches every position in the basket or concentrates in convenient geometry. Whale Cleen’s engineering team uses finite element simulation to predict and optimize ultrasonic energy distribution before a tank is built, positioning transducers on side walls as well as the bottom to eliminate dead zones. The objective is straightforward: every blade—regardless of its position in the fixture—should receive comparable cavitation intensity.
Process separation prevents the batch from contaminating itself. Instead of a single tank performing every function, Whale Cleen systems decompose cleaning into discrete, independently controlled stages: ultrasonic cleaning, filtration, rinsing, and drying. Each stage has a defined task, and contamination removed in one stage does not follow the part into the next. This is what allows every blade in a batch to follow the same cleaning path under the same conditions, rather than experiencing a bath that degrades from the first part to the last.
Filtration is a batch-consistency component, not a maintenance item. Continuous multi-stage filtration removes suspended oil and particles from the cleaning bath, keeping the liquid condition stable across the full production run. Without it, the bath itself becomes a variable—clean at the start of the shift, loaded with contaminants by the end—and batch uniformity becomes impossible to hold.
Frequency and power must match the contamination, not a single setpoint. Turbine blade contamination is layered: baked-on carbon, oxide scale, residual oils, and particulate debris respond to different cavitation energies. Whale Cleen systems incorporate multi-frequency capability with adaptive power regulation, allowing the process to use lower frequencies for bulk carbon removal and higher frequencies for fine particulate dislodgement without eroding blade surfaces. The point is not to run every blade through the same intensity, but to match intensity to the specific cleaning task at each stage.
Even a well-designed acoustic field and staged process can drift if the operating window is not locked. Temperature fluctuations shift cavitation intensity; inconsistent chemical concentration changes the wetting and displacement behavior of the cleaning liquid; manual adjustments introduce operator-to-operator variation. Whale Cleen addresses this with closed-loop process control: real-time temperature compensation, bath management, and PLC automation with recipe storage, so that every batch runs under the same parameters regardless of who is on shift. The machine is documented with “one machine, one file" traceability records, logging component batch numbers through final test results—a discipline that matters when cleanliness must be verified, not just inspected.
For turbine blades specifically, the fixture is part of the process window. A blade held in a basket that traps air in a cooling passage blocks cavitation from entering that passage, no matter how uniform the tank field is. Whale Cleen engineers fixturing around the geometry of the part, positioning blades so the cleaning liquid wets the features cavitation needs to reach and trapped air can escape. In a documented aerospace application involving blades with cooling holes smaller than half a millimeter, a custom Whale Cleen system achieved full contaminant removal with zero blade damage in validation testing.
Rework cost in batch blade cleaning is usually tracked as a labor and chemical expense. The larger cost is the inspection burden it creates: when a batch is not confidently clean, every part must be checked, and the parts that fail must be re-cleaned, re-inspected, or scrapped. Repeated cleaning cycles can alter surface conditions and introduce new risks of their own.
The useful question is not “how do we clean the failed parts more aggressively," but “why did the first pass produce a mixed result." In a single-tank system with uneven acoustics and a degrading bath, the answer is structural. Whale Cleen’s approach is to treat batch uniformity as an engineering specification: simulate the acoustic field, separate the stages, filter the bath, lock the parameters, and fixture the part so cavitation reaches the features that determine cleanliness. When every blade in a batch has to meet the same standard, the cleaning system cannot afford to treat any position in the basket as a lower priority than the others.
