A cleaning process that only works when a particular supervisor is in the building is not a process. It is a favour. For a machine shop that wants to clean machine-tool exteriors on a schedule rather than on request, the difference between those two states is written down long before the lance moves: which surfaces are in scope, which fluids are on them, what must be shielded, how the pass is made, and what the crew writes back after it is finished. The title question is about repeatability, so this article answers it as a build: define the surface, match the method to the deposit, protect the openings, set the consumable limits, trial the sequence, and train from the record the trial leaves behind. For a business, that build is also the difference between a clean shop for one audit week and a clean shop every week.
Why machine-tool exteriors are the hard part for a business
Inside a machine, the process is the manufacturer's problem: ways, guides, seals, ball screws, coolant paths, and the tolerance chain that depends on all of them. Outside the machine, the surfaces belong to the shop, and the shop decides what happens to them. That boundary should be stated in one sentence in the written procedure, and it should be unambiguous enough that a new operator never guesses whether a cover plate is in scope.
Exteriors also fail differently from floors and walls. A machine has painted panels, machined castings, plastic windows, cable chains, labels and joints where two materials meet. Each reacts differently to heat and water. A process that is safe on a flat painted door can be wrong on a gasket, or on the edge of a decal where the film has begun to lift.
Exteriors are also the part of the machine the customer sees. When a buyer walks a shop to judge a supplier, they do not open the enclosure; they look at the outside of the equipment and at the floor around it. A dull, oil-filmed machine communicates a maintenance story that no marketing page can argue with. Cleaning the exterior is therefore a commercial activity dressed as a housekeeping one.
That is exactly why the process has to be repeatable rather than heroic. A one-off deep clean before an audit changes nothing about the machine tomorrow. A written sequence that two shifts can run to the same result does.
Start from a surface list, not from a machine list
Most shops already have a machine list. It does not help here, because cleaning decisions are made surface by surface. The first deliverable in the process is a per-machine exterior inventory written by someone who has actually walked the machine with a flashlight.
For each machine, record the painted sheet panels, the bare or machined castings, the chip and coolant covers, the enclosure windows, the control cabinet and its door seals, the cable and hose carriers, the nameplate and safety labels, the base and the plinth, and the surfaces that face the operator versus the surfaces against the wall. Note what each one is made of and what it is coated with, because the material and the coating, not the word "machine", determine what the cleaning step may do to it.
Then add a second column that most shops forget: reachability. Which surfaces can be reached standing on the floor, which need a step, which need the door swung open, and which cannot be reached without removing a guard. A process that assumes access nobody has will be quietly abandoned by the second week.
A third column carries a small symbol: is the machine on a live production sequence, and can it be stopped for the time the procedure needs? If it cannot, that machine belongs to the weekend list, and the procedure has to say so rather than leaving the crew to negotiate it at the machine.
Name the deposits before choosing the method
Machine-tool exteriors carry a small number of recurring deposits, and each behaves differently under heat and water. Cutting oil spread over the front of a machine by operator hands and by splatter. Way oil that has crept out past a wiper or a seal and baked onto a warm surface. Coolant mist that dries into a pale film across panels and windows. Hydraulic bleed from a weeping fitting that darkens the area directly below it. Swarf stuck to anything still oily, which turns a film into a crust that will not be cleaned at all until the crust is lifted first.
The practical point is that the deposit, not the calendar, drives the method. A fresh mist film is a short contact-time job. A baked-on oil patch at a way cover needs the heat held on the spot and the pass repeated. A swarf-bound crust needs a mechanical lift beforehand, and any method that skips that step will just move the crust around.
Write the deposit type onto the surface sheet beside each surface. That column lets the crew pick the setting from a rule instead of from memory, and lets a new operator reach the same decision as an experienced one.
It also protects the machine. A film that has sat on a panel for months has already survived normal shop cleaning; what changes its behaviour is heat. Where a label, a decal or a gasket sits under a dried oil film, the film is hiding the condition of the material beneath it, and that is the spot to inspect before anything warm touches it.
Dry steam, wet steam and hot water: decide by surface, not by habit
The cleaning range on the industrial steam cleaning equipment pages describes mobile units offering dry steam, wet steam and hot water modes, and a fixed unit such as the Fixed Steam Cleaning Machine PowerDe HW02 publishes dry and wet steam switching among its specification rows. Which mode is right is a shop decision per surface, and the written procedure should record the decision rather than leave it to whoever happens to be holding the wand.
A lower-moisture pass is the default anywhere water must not travel: around cabinet doors and their seals, past label edges, near cable carriers, at shut lines where a panel meets a casting, and on any surface above an electrical entry. The trade-off is time on the spot: with less water carrying the soil away, contact time and repeat passes do the work.
A higher-moisture pass suits open, robust areas where the run-off can be controlled: painted flat panels away from seams, the outside of chip covers, the base and the plinth, and floor-adjacent surfaces where the soil is thick. The trade-off moves to the floor. Water that runs off a machine into a gutter is fine; water that runs off a machine into a joint, under a control cabinet, or toward the aisle where someone walks is a different problem, and the procedure should name the run-off route for each wet task the way a maintenance document names a lock-out point.
Hot water carries more soil per pass, which is useful on encrusted areas, and it leaves more water behind, which is exactly what the near-electrical and near-seam surfaces cannot accept. Many shops split the job accordingly: heavy areas wet, sensitive areas dry, and everything left dry at the end of the shift so that no machine is handed back wet.
Whatever the split, write down how the pass is actually made. Stand-off distance, the speed of the lance, how many passes over one spot, and what the crew does when a spot does not lift on the second pass. Those small details are where two crews currently diverge, and they are the reason a repeatable process needs a paper version of the technique.
Protecting what must stay dry is part of the cycle time
On a machine tool, the shielding list is short and very specific. Control cabinets and their door gaps. Operator panels and their fan vents. Terminal boxes and their cable glands. Nameplates and safety labels. Any decal or film with a lifting edge. Way covers and the seams where they meet the casting. Spindle area openings. The gap between a panel and a hinge. Anything that is an opening rather than a surface.
Shielding is not free, and the process should cost it honestly. Someone places the covers, someone else is holding a shield while the lance works a tight spot, and both are minutes per machine. If those minutes are not in the plan, they come out of quality: crews stop shielding, and the shop discovers the cost later, when a cabinet dries out and something still does not work.
A useful habit is to fix the shielding as a standing condition rather than a per-job task. Cabinets that get washed near regularly are better served by a fitted cover kept on a hook beside the machine than by a fresh sheet of plastic hunted for each week. The first attempt is improvised; after that it becomes hardware on the wall, and the process stops depending on improvisation.
Write the exclusions in the same column as the surfaces, with a fallback. If a surface is excluded, say what happens to it instead: wiped, degreased by hand, left alone, or referred to the machine builder. An exclusion without a fallback is how a shop ends up with surfaces nobody owns.
Utilities and consumables set the real length of the job
Before the sequence can be repeated at the same speed, its inputs have to be known and stated. Power: what supply the quoted unit needs and whether that supply exists where the machine actually stands. Water: the inlet connection at the point of use, the quality of the feed, and whether the machine is near a tap or the crew is carrying jerry cans across the shop.
The published rows for a fixed unit give a concrete example of the kind of record to ask for: the HW02 page lists 220 V working voltage, 6 kW heating power, an 80 L SUS304 tank, and a tap-water inlet with a 4-point internal-tooth joint, half inch. Take the equivalent rows for the exact machine being quoted and check them against the supply and the tap at the position where the work will actually be done, because those rows belong to that machine and are not portable to another model or another installation.
Two caveats belong next to those numbers. A tank volume is a quantity carried, not working time; the working time depends on how much the wand is actually open, and it is measured, not calculated. And the distance from the machine to the water point is part of the cycle time, since refilling and repositioning happen between passes, not during them.
Water quality is a process detail with a long tail. Feed water that leaves deposits will show up as a pale haze on panels the crew just cleaned, and the crew will read that as a failed clean and repeat a pass that was already successful. If the shop's water is hard, say so in the procedure and decide what is done about it, rather than letting each crew rediscover it.
Run one bounded trial and let it write the standard
Do not roll a process across the shop on the strength of a single good afternoon. Choose one machine that is representative of the family, that can be stopped for the time needed, and that carries the deposit type the shop is most troubled by. Run the written sequence on that machine only.
Decide the acceptance standard before the trial starts, not after it. A useful form is specific and visible: no film remaining on the painted front panel under the shop's normal lighting; no residue along the edge of the way cover; no water standing in any joint; labels intact and legible; the cabinet interior dry. Write the stop condition too: what result ends the trial, and what result sends the sequence back to the drawing board.
During the trial, collect the record rather than trusting memory. Time per surface group, from the shielding placement to the last dry wipe. Which mode was used on which surface, and how many passes it took. What lifted easily and what did not. Where water went that nobody wanted it. What the crew had to invent, because everything they had to invent is a hole in the procedure.
Close the trial with a condition check on the machine itself before anything is unlocked or handed back: cabinet interiors dry, no water at joints, labels and decals undamaged, no lifting at the edges of film, guards back in place. If the machine was fine before and is not fine after, that is a failed trial even if the surfaces look clean, and the procedure has to change before the second machine is touched.
Training and records turn a clean machine into a process
The output of the trial is a short document. One page per machine family: the surface list with its deposit column and its exclusion column, the mode per surface group, the stand-off and pass rule, the shielding list, the run-off route, the dry condition that closes the job, and the acceptance standard. Anyone who can read can run the job from it.
Train against that document, in the shop, at a machine, with the person who wrote it watching. Ask the trainee to explain what they are about to shield and why before starting; the answer shows whether the document is doing its work or whether the procedure still lives in one experienced head.
Keep the record per machine and per date: who cleaned it, which surfaces were done, which were excluded, what the acceptance result was. This is what lets the shop answer an awkward question later. When a machine starts failing after cleaning has begun, the record shows which machines were cleaned, how, and what was found at hand-back. Without it, the answer is a guess.
Review the document on a fixed date, not when someone complains. The deposits change with a new coolant or a new tooling supplier; the surfaces change when a machine is repainted or a label is replaced. A process that has not been touched in two years is a form being signed, not a process being run.
Where the sequence needs to be built around the shop's real machines, the customized steam cleaning equipment route on the customer-facing side is where surface lists, deposits, supplies and stop-windows get described and answered in writing. Ask, in one exchange, for the mode options, the protection items fitted on the quoted unit and what they do not cover, the inlet and supply rows for that exact machine, and the accessories for seams and joints. A written question list asked once is worth more than three partial conversations.
The five things that make the process repeatable
First, a per-machine exterior surface list with a deposit column, an exclusion column with a fallback, and a reachability note. Second, a mode rule per surface group with the pass technique written out, so two crews do the same job. Third, a shielding list and a named run-off route for every wet task, costed into the time rather than smuggled in. Fourth, a bounded trial with the acceptance standard and the stop condition set beforehand, closed by a machine condition check before hand-back. Fifth, a one-page document per machine family, trained in place, reviewed on a date.
None of those five depends on a particular person being on shift. That is the test. If the cleaning result survives a change of shift, a change of supervisor, and a new starter, the shop has a repeatable process for machine-tool exteriors rather than a habit, and the exterior of the equipment will say so every time a customer walks the line.
