What kind of gear drive you are testing changes the questions
A works carries more than one kind of gear drive, and they do not ask the same thing of a cleaning method. A close-coupled gearbox on a conveyor or a mixer has a cast cover, a nameplate, one or two breathers, a sight glass, a filling plug and a drain plug, and it sits in a corner where a lance has to reach over a guard. A gear motor adds a fan cowl and terminal box. An open girth gear on a mill, or an exposed rack-and-pinion drive, has teeth, tooth flanks and a splash guard, and the deposit there is usually grit bound with grease rather than a smooth oil film. An enclosed machine-tool headstock has machined mating faces, seals and instrument lines around it.
Write the drive types down separately before the trial and give each one its own set of test panels. A result achieved on a plain painted cover tells you nothing about a toothed open drive, and a method that fails on the open drive does not disqualify the plain covers. Grouping them into one line item on a work order is how a business ends up applying a decision to a machine it never actually tested.
Deposit, surface and boundary: the three columns to fill in first
For every unit in the trial list, record what is on the housing and what it is made of. A thin gear-oil film behaves differently from a thick graphite-like grease holding abrasive dust, and a dried coolant or sugar-rich mist film behaves differently from both. A bare casting with a rough skin holds deposit in pores that a painted or powder-coated cover does not. A polished machined face shows scratches that a painted cover hides.
Then draw the boundary of the job on a sketch of the machine, not in prose. Inside the boundary: cover outer faces, the top and sides of the housing, flange outer edges, guard outer faces, mounting feet, the outside of the motor frame and its cooling fins. Outside the boundary, and written as excluded with an alternative named: shaft seals and their lips, breather and its vent path, sight glass and its seat, drain and fill plugs and their gaskets, encoder and sensor housings, cable glands and junction boxes, terminal-box entries, and any label whose legibility matters. An exclusion with no fallback is an unfinished instruction; the fallback is usually a hand wipe with a damp cloth and a dry follow-up, done by hand with no lance.
Why heat and dryness matter more than pressure on a housing
On a housing exterior, heat does the loosening and the small amount of water does the lifting. High pressure on a housing is mostly a way to drive moisture into places you have just spent the previous section excluding. A heated, low-water method gives the dwell time that oil and grease need, and it leaves the machine drier on the far side of the job, which is the property you want around breathers, seals and electrical entries.
A machine that switches between dry and wet steam lets the operator pick the drier setting for everything near a seal or an electrical entry and the wetter setting for a plain cover face away from them. If a supplier's machine is described as offering several modes rather than one fixed output, ask which mode is intended for the seal side of the boundary and which for the open cover face, and have the answer on the trial sheet rather than left to the operator's judgement on the day.
A steam method is not a replacement for a pressure washer wherever volume is what clears the deposit. If a housing is carrying mud, heavy grit or loose swarf, that material has to be removed dry or with a low-pressure rinse first, otherwise you are converting loose contamination into a paste and pushing it into seams.
What must stay dry, and what a fallback looks like
Make a shielding list and treat it as part of the machine preparation, not as something the operator remembers. For a gear unit the usual list is the breather, the sight glass and its frame, the shaft ends and any lip seal, the encoder or resolver housing, cable glands, the terminal box and its entries, the nameplate if it is not engraved, and any gap at a cover joint that you can see light through. Shielding is a cover, tape and a drip tray, and the time it takes belongs in the cycle time you will report at the end of the trial.
Where a job has to be closed out after a wet pass, write the closing condition as a check the operator performs, not as an assumption. Run a dry pass or a dry wipe over everything that has been wetted, let the housing come back to a temperature you can put a hand on, and only then remove the shielding and sign the machine back over to operations. A housing handed back damp, with a still-warm cover and a wet breather area, is how a cleaning job turns into a bearing or an electrical fault the following week.
Isolation and the oil that is still inside
Nothing on the outside of a gearbox tells you the oil inside is safe from the job you are doing. Before the trial, agree in writing who holds the isolation for the drive, who locks it out, and what condition the unit must be in to be returned to service. Confirm that the lubricant level and the vent path are correct before the cleaning starts, because a hot housing plus a blocked breather is a pressure difference, and a lance aimed at a joint is the easiest way to find out whether the breather was clear.
Keep the trial away from a unit that is already showing weeping at a seal, a loose or missing plug, a cracked sight glass or a cover joint that has been re-sealed badly. Those units go on a repair list and are cleaned after the repair, not used as test pieces. Record them on the trial sheet as excluded machines with the reason, so the trial result is not later read as covering them.
Choosing the trial area, then the pass sequence
Pick two or three units per drive type, and on each unit mark off a test patch with tape: one patch of plain cover face, one patch that includes a corner or a flange seam, and one patch that deliberately sits close to an exclusion item so you can see how the method behaves near it. Photograph each patch before anything touches it, with the same lighting and the same distance each time, and put the tape line in the frame. That photograph is the baseline; without it the trial result becomes an opinion.
Run the sequence the same way on every patch, and write the sequence down before the first pass: a dry removal of loose material, dwell so the heat works on the deposit, a pass with the setting you have chosen, an immediate wipe with a clean cloth, then a dry pass and a visual check under a lamp. Change one variable at a time between patches, for example the setting or the number of passes, never both. Compare the after photograph with the before photograph against the tape line, and note whether the deposit has been lifted or merely moved and redeposited at the edge of the patch, which is a common false pass on a rough casting.
Stop conditions to write down before the trigger is pulled
Write the stop conditions on the sheet in advance so nobody is improvising when one appears. A lifting or blistering coating, a softened label, a nameplate whose characters start to disappear, moisture appearing under a sight glass or in a seal area, water standing in a joint or a drain path, an unexpected change in how a running unit sounds or feels afterwards, and any sign of a weeping seal that was not there before the job. Each of those ends the test on that patch, and the patch is recorded as a failure with what happened, not dropped from the record.
Add one more line to the sheet for the machine itself. Ask the supplier in writing which protective functions on the unit interrupt a job and on what condition they trip, and record the answer for the exact machine quoted, since protection arrangements are written per model and are not transferable between them.
What to record for each patch
One row per patch, on one page, with these entries: unit tag and drive type; deposit type as you typed it; surface and coating; patch location and its distance from the nearest exclusion item; the setting used and the number of passes; dwell time and total time for the patch including preparation, shielding and the closing dry check; before and after photograph references; outcome, stated as passed, passed with a condition, or failed with the reason; and the signature of the person who inspected it, who should not be the person who cleaned it.
Add a second short block on the same page for the machine you used, filled in from the supplier's written answer rather than from the brochure: working voltage and where your nearest supply point is, how the machine is fed with water and how far the nearest tap actually is, what limits a session and what it takes to resume it, how the machine moves around the plant and whether it will fit the space the gear unit stands in, and how the figures on the page were obtained. A tank capacity is not a working time and a stated inlet thread is not a hose plan; both belong on this page as questions answered for the exact unit on your floor.
The run-off question the site has to sign
Cleaning a housing produces dirty water mixed with oil and grease, and where it goes is not a detail. Before the trial, agree with the site which floor drains are usable, which are not, and what container or sump collects the run-off from each patch. Put drip trays under everything you shield, and write the collection route on the sheet next to the task it belongs to. A trial that produces an unagreed discharge is not a completed trial, and the cost and time of the collection route should appear in the trial's reported time rather than being discovered during the first full round.
How to read the catalogue before you buy the answer
Read what the cleaning pages actually publish. industrial steam cleaning equipment is listed there with mobile units positioned around indoor garage bays and tight outdoor parking areas, and a fixed unit positioned around vehicle exterior work at fuelling and parking sites, interior work and engine compartments. Nothing on those pages names a gear-drive duty among the applications, and no line-cleaning rate for gear drives is published anywhere on them. Treat that as a scope question to ask in writing rather than a gap to fill in with an assumption, and do not let a supplier's product name stand in for your own trial result. The catalogue entry worth opening when you want a fixed unit as an engineering starting point is the Fixed Steam Cleaning Machine PowerDe HW02 page. Any figure on that page belongs to that machine at that supply arrangement, and the honest way to use it is to ask the supplier how it was obtained and what the same figures are for the unit they are quoting you.
Limitations worth keeping in the report
A heated low-water method is the wrong tool in a set of cases the trial will expose, and the report is more useful for naming them. It is not the answer where volume is needed to move heavy loose contamination. It will not remove a polymerised coating or a baked-on varnish without damaging what is underneath, and it will not restore a surface that is corroded. On open toothed drives, nothing that adds moisture to tooth flanks and a splash guard should be approved without the drive maker's own maintenance rules being read first, and gear oil and lubrication are outside the scope of any exterior cleaning method.
Where a plant's requirements go beyond what a catalogue machine is documented to do, the sensible next step is a conversation about a configuration rather than an attempt to force an existing datasheet to fit. PowerDe presents customised cleaning configurations as a service, so customized steam cleaning equipment is the place to put your measured trial sheet in front of an engineer, with your stop conditions and your run-off route attached to it.
Turning the trial into a standard the plant keeps
When the trial is done, hold one review with maintenance, production and cleaning together, and take the patch sheet to it. Decide three things in that meeting: which drive types the method is approved for, which are excluded with a named fallback, and which are parked until a repair is complete. Anything approved goes onto a one-page standard per drive family, with the shielding list, the sequence, the closing dry check and the stop conditions on it, and with the before and after photographs reproduced so the acceptance standard is visible rather than described.
Train operators against that page at the machine, not in a room, and give them the authority to stop and to mark a unit as excluded when they find a weeping seal or a missing plug. Put a review date on the page. If a new gear drive type enters the plant, it does not inherit an existing approval; it gets its own patch, its own photographs and its own row on the sheet, and only then joins the standard.
