A gearbox housing is one of the awkward parts to put in front of a new cleaning process. It carries the marks of whatever the housing went through before it reached the cleaning bay — cutting fluid, drawing compound, assembly grease, and the fine metal fines that come loose while gears mesh — and it has internal cavities, machined mating faces, sealing grooves and blind holes that a rag and a bucket of solvent will never reach properly. The cleaning trial is where you find out whether the method you have chosen actually suits that geometry. The preparation work done before the first shot of steam decides whether the trial gives you a number you can trust or just a set of photographs.
If the trial is being run with a fixed steam cleaning machine such as the PowerDe HW02, the preparation steps below are the ones we ask customers to complete in advance, because a fixed 220 V unit runs continuously at the bench or in the maintenance bay and the trial usually shares the working day with normal production.
1. Record the part and the soiling before anything is cleaned
Photograph the housings as they arrive, and write down what they are carrying. A gearbox housing from a machining cell behaves very differently from one pulled out of a working drive. Fresh machining residue — cutting oil and coolant film, small chips and swarf packed into threaded holes — responds to heat and pressure fairly quickly. A housing taken out of service carries compacted grease around the bearing seats, hardened sealant at the split line, and abrasive fines that have been ground into the surfaces under load; that is the harder case and needs longer dwell and more passes.
Note the material and the finish, because those two facts control what the process is allowed to do. Cast housings — aluminium or iron — can have porosity that holds oil, and a machined face that has been deburred by hand may have burrs and lap marks that trap soil. Record the part number, the batch, and how many parts are in the trial set. A trial run on one part tells you almost nothing; run the same method on a small, repeatable set so that variation between parts is visible.
Ask the customer or the process owner what the clean part has to be fit for. A housing that goes on to the assembly line has to be clean enough for sealant to adhere and for bearing bores to accept their tolerance; a housing being returned from a service repair mainly needs to be clean enough to inspect for cracks and wear. These two acceptance levels are not the same, and the trial is only meaningful against a target agreed before it starts.
2. Strip the housing down to a cleanable state
A cleaning trial run on an assembled gearbox tests the seal, not the cleaning process. Before the housing goes into the trial, remove everything that is not part of the housing: bearings and retainers, shafts, gears, gaskets and O-rings, nameplates and stickers, and all fasteners that can come out. Grease packed under a retained bearing cannot be reached by any spray from outside, so any result that looks acceptable on an unstripped part is misleading.
Take a decision on plugs and protective fittings. Deep cavities and oil galleries that must stay dry need temporary plugs; open machined bores that must stay free of debris need to be covered during the passes around them. Note where each plug was fitted, because a plugged bore that arrives at inspection with liquid and grit trapped behind it is a preparation failure, not a cleaning failure.
Sealant beads and old gasket material should be removed mechanically before the wet process begins. Steam and pressure can soften the residue, but they will not push a cured bead out of a groove in one pass, and trying to force it wastes trial time and risks damaging the mating face. Scrape and brush the grooves first, then treat the remaining film as part of the trial.
3. Set up the trial area so the parts can be inspected between passes
Clean parts on a stand or rack that holds the housing with the openings facing down or sideways, so that rinse water and loosened soil drain away instead of settling back into a bore. Keep the parts off the floor and away from the machine's own hose and lance. Lay out the personal protective equipment that the operator will wear throughout the trial — high-temperature steam at the lance can scald, and a fixed machine running at high vapour pressure is not something to be handled in ordinary workshop gloves and an apron.
Check the services at the position where the trial will run. A fixed machine needs the correct supply voltage and a water inlet; the water quality in the shop affects what is left on the part after drying, so record what is being fed to the machine. Confirm that the area drains, that the floor is not going to be a slipping hazard for the rest of the shift, and that parts waiting for inspection will not be knocked or re-contaminated by traffic through the bay.
Make the machine itself known to the operator before the trial starts. On the PowerDe HW02 steam cleaner the detail page lists the dry-steam and wet-steam modes alongside the vacuum function, together with leakage, short-circuit, high-temperature, high-voltage and water-shortage protection. Dry steam and wet steam behave very differently on oily surfaces, and that difference is one of the main variables you want to compare during the trial rather than discover halfway through it.
4. Choose what the trial is going to compare, and keep it to a small number of variables
Decide the sequence of passes and stick to it. The usual order is a short pre-rinse or dry pass to take off loose chips and free oil, then the main cleaning pass on the heavy soil at the bearing seats, split line and grooves, then a final flush over the faces that will be assembled. Where a housing has both external grime and internal cavities, treat them as separate stages and log them separately; the external surface is not evidence about the internal one.
Write down the settings for every pass: the mode used, the distance of the lance from the surface, the speed of travel, and how many times the same area was gone over. These are the numbers you will reproduce when the trial becomes a work instruction. Do not change two variables at once, because you will not know which one caused the improvement.
Keep a control part. Set aside an identical soiled housing that will not be cleaned by the trial method. When you compare the treated part against the untreated one, the difference is obvious at a glance, and it prevents people arguing about whether a surface "looks clean enough".
5. Inspect the same way every time and log the result
Inspect immediately after each pass, before the part dries and leaves residues that look like the soil you were trying to remove. Wipe a clean white cloth or filter paper across the machined faces, into the grooves, and along the split line, and judge the cloth rather than the part — a cloth that comes across with dark streaks at the bearing bore says more than a surface that merely looks brighter. Check the bores, threaded holes and blind holes with a light and a probe for residue that has been pushed to the bottom of a hole rather than removed.
Run the trial set through the acceptance checks agreed in step 1, and photograph the same features in the same order after every pass. Where the target is adhesion for sealant, do the adhesion check on the cleaned face rather than assuming it follows from a clean appearance. Where the target is assembly, check that the bores and grooves still hold their condition and that nothing has been scored, corroded or left with trapped liquid.
6. Turn a passing trial into an instruction
When the parts pass, write down what you did while it is still fresh: the stripping steps, the plug map, the pre-treatment of grooves and old sealant, the mode and pass sequence, the number of passes, the drying and draining arrangement, and the inspection method with the acceptance rule. Attach the photographs. Add a line about what failed, because the parts that did not pass are the most useful part of the record.
If the trial fails, say where it failed before changing the machine or buying anything: soil that survived on the split line points to a dwell and pass problem, soil at the bottom of a blind hole points to a drainage and access problem, and a part that passed visually but failed on the cloth usually means the inspection point was in the wrong place.
The preparation described here is generic mechanical-workshop practice, so it holds whatever cleaning method you are comparing. Where the trial is being run against PowerDe equipment, the cleaning equipment range covers fixed, hand-pushed and mobile machines, and the customisation service covers cleaning equipment specified around a customer's parts and shop layout. For a trial on your own housings, send the part details, the soiling you are dealing with and the acceptance target to sales@powerde.net or call +86 13023198376.
