A sliding window does not fail because nobody cared for it. It fails because the caring was done out of order, by someone who was never told what the window is actually doing. Cleaning and maintenance get printed on the same checklist, scheduled on the same afternoon and handed to the same person, but they pull in opposite directions. Cleaning takes material off. Maintenance puts material back, or adjusts it, or leaves it alone on purpose. Only one of the two can permanently undo the other four.
That matters more than it sounds, because of who ends up doing it. The person who touches your windows most often over the life of a building is almost never the person who read the manual. It is a cleaning crew, a painter, a stucco team, a pressure-washing contractor, or a new facilities manager inheriting a building with no handover file. The window was designed by people who assumed somebody would be told three things: that the bottom track is a drain, that the finish has a temperature limit, and that the small holes at the bottom are the most important feature on the product. None of those three are visible in a showroom, and none of them appear on a quotation.
This page is about the order of operations, about the parts of a sliding window that are drains rather than surfaces, and about the small number of actions that quietly cancel everything else. It is not a product page. Where we give a number, we say whose number it is.
Ask ten people what "sliding window maintenance" means and you will get ten versions of the same two-minute job: wipe the track, spray something, done. That description collapses five distinct operations into one, and the collapse is where the damage happens. Each task has a different failure mode, a different skill level and a different consequence when it is skipped.
| Task | What it actually does | Who usually does it | How it fails |
|---|---|---|---|
| Clean | Removes deposits: salt, dust, grit, pollen, mortar, insect debris, soap film | Cleaning crew, occupant | Wrong chemistry or wrong tool damages coating, glass edge or seals. This is the only task that can destroy the others. |
| Inspect | Looks for wear that only shows on a clean surface: flat-spotted rollers, split gaskets, blocked weeps, loose fasteners | Nobody, usually | Skipped entirely. Most "sudden" failures were visible a year earlier on a surface nobody cleaned. |
| Adjust | Sets roller height, lock point compression, anti-lift clearance | Glazier or service technician | Done on a dirty window, which misdiagnoses the fault; or done too far, which bends hardware. |
| Lubricate | Restores a low-friction film at the roller and track interface | Whoever has a spray can | Wrong product turns remaining dust into abrasive paste. Over-application floods the drain path. |
| Record | Creates the only evidence that the other four happened | Almost nobody | Absent. When a claim arrives, there is no way to separate wear from neglect from defect. |
Look at the third column. Two of the five tasks have no natural owner. Inspect and Record are the two that decide whether a building still has working windows in year twenty, and they are the two most often assigned to nobody. If you take one organisational action from this page, make it that one: name a role, not a person, for inspection and for the log.
Almost every maintenance guide online argues about how often. How often matters far less than in what order, because three of the five tasks are ruined if the one before them was skipped. A maintenance programme that does the right five things in the wrong sequence is worse than one that does only cleaning, because it hides the evidence.
| If you do this | Before that | What you have actually created |
|---|---|---|
| Lubricate | Clean | A grinding compound. Grit is now suspended in a sticky film that stays exactly where the roller runs. Several published guides describe this in the same words: an abrasive paste that accelerates wear. |
| Adjust | Clean | A false diagnosis. A sash that drags because the track is full of sand gets its roller height altered instead; when the track is eventually cleaned the window is now out of alignment in a second way. |
| Inspect | Clean | Nothing useful. A split gasket, a hairline crack at a corner and a blocked weep are all invisible under a film of grime. |
| Clean the glass | Clean the frame | Scratched glass. The frame carries the grit. One cloth, moved from frame to glass, drags it across the pane. |
| Record | Any of the above | Nothing, because there is nothing to record. The log has to be filled in during the work, not reconstructed afterwards from memory. |
There is one more ordering rule that sits outside this table and catches people out: never lubricate to diagnose. If a sash is stiff and a spray makes it better for a week, you have not fixed anything. You have silenced the symptom of a roller that has flattened, a height that has drifted or a track that has been gouged. The correct sequence is clean, then operate, then decide.
Here is the single most consequential fact about a sliding window, and it is the one nobody is told: the bottom track is designed to get wet, and water standing in it is the system working, not the system failing.
The physics is straightforward. Wind-driven rain runs down the outside face of the glass, lands in the outer track, and is meant to leave through small slots or holes on the exterior face of the frame. Published care instructions from window manufacturers say this in almost the same words: it is normal for water to collect on the sill or in the outside track during a rain event, and it is not a leak. One manufacturer's troubleshooting page puts it more bluntly still: if you can see water in the track from inside, you are looking at a weep drainage system doing its job.
Everything that goes wrong with sliding windows downstream of this fact goes wrong because somebody treated a drain as a surface.
Two corollaries follow, and both are counter-intuitive enough that they get argued with on site.
Do not mop it out. One manufacturer's guidance is explicit: do not try to mop standing water out of the track and do not stuff a towel into it, because a towel wicks water onto the sill and into finishes that were never meant to get wet. Wait. The system is draining on its own schedule, not yours.
Never seal a weep hole, and never let a trade do it. Several published care sheets warn that blocking the weep path can void part or all of the product warranty, and one specifically instructs owners not to let a painter or a rendering crew caulk over the drainage openings. This is not a hypothetical. In one repair account from a UK trade supplier, a homeowner with water on an internal sill responded by smearing fresh silicone all along the external perimeter joint, repeatedly, while the actual cause was a compacted plug of dead insects, moss and brick dust sitting in an internal collection channel. The silicone was ugly, cost money, and did nothing, because the water was already inside the frame. The instruction that comes out the other end of that story is worth memorising: when water appears on the inside, the first suspect is not the perimeter seal. It is the drain.
Most maintenance guides stop at "vacuum the track and wipe it." That instruction has a hidden assumption: that the visible channel is the drain. On many sliding systems it is not. The visible channel is a snap-in liner, and the water passes through it into the frame below before it reaches the exterior weep.
At least one manufacturer's care page describes exactly this failure. Debris accumulates under the track liner and clogs the weep from beneath, so cleaning the liner changes nothing. Their published remedy is to lift the liner out: hook an L-shaped pick under the outer edge of the sill track and pull up to release the snap connection, remove it completely, clean underneath, reinstall with an even gap at each end, and then repeat the pour test. The same page notes that if the window still will not drain after the track itself has been cleaned, removing the liner is the next step, not calling a glazier.
This matters for how you write a maintenance specification. "Clean the track" is unenforceable and incomplete. "Remove the sill track liner, clear beneath it, reinstall evenly, and verify drainage with a pour test, twice yearly" is a sentence a contractor can price and a facilities manager can inspect. Whether your particular window has a removable liner is a question for the supplier, and it is a good one to ask before handover rather than after the first overflow.
Lubricant choice is the place where a well-meaning person does the most durable harm, because the damage is invisible for weeks and then permanent. The rule across nearly every published source is the same, and it is worth stating in the negative: do not put anything on a sliding window track that stays sticky.
| Product | What it does in a track | Verdict |
|---|---|---|
| Silicone-based spray | Lays a dry-feeling film, does not hold dust, tolerates temperature and moisture | The most commonly recommended option. Apply thin, cycle the sash, wipe the excess. |
| Dry PTFE spray | Leaves a slippery coating with no liquid residue at all | Preferred by several sources specifically for dusty or sandy environments, where any wet film is a liability. |
| Petroleum oil, machine oil, general grease | Attracts and holds grit; the mixture becomes a lapping compound | Never. Every source we consulted lists these as the thing to avoid, for the same reason. |
| General-purpose penetrating spray | Initially frees movement, then dries tacky and collects dust; can also degrade the grease already inside a sealed mechanism | Not for tracks and not for locks. One source warns specifically against penetrating oils on hardware for this reason. |
| Cooking oil | Goes rancid, sticks, holds dust | It appears on enough "home remedy" pages that one manufacturer's care guide names and rejects it explicitly. |
Now the honest complication, which almost no guide mentions. Published advice genuinely conflicts on whether you should lubricate the rollers at all. One North American manufacturer's care document states that the roller mechanisms in their sliding window are self-lubricating and that additional lubricant is not required, adding only that in areas of heavy pollution a light coat of spray silicone may be applied to the rollers. Other sources recommend lubricating rollers every six months as routine, and coastal guides recommend it quarterly.
Both can be right, because they are describing different hardware. A sealed bearing wants no added lubricant and can be damaged by a solvent carrier that strips its factory grease. An open nylon wheel on a plain axle wants a dry film. The instruction you need is therefore not "lubricate quarterly." It is "tell me which roller is in this sash, and give me the hardware maker's care instruction for that roller." If a supplier cannot answer, that is the finding.
Every maintenance schedule you will find is a table of intervals. Almost all of them are honest and almost none of them agree, and the reason is that they were each written for a different atmosphere. Salt, industrial fallout, traffic film and desert dust do not attack the same component and do not arrive at the same rate, so a single interval cannot be correct for all four.
| Exposure | Published interval (varies by source) | What accumulates | What it attacks first |
|---|---|---|---|
| Coastal, close to tidal water (one Australian guide uses roughly 500 metres; another uses within 5 kilometres of the sea) | Full wash every 8 to 12 weeks plus a monthly fresh-water rinse; other sources say every 3 months, or monthly in the worst cases | Salt crystals that hold moisture against the surface | Coating edges, cut ends, fixings, then the substrate. Pitting is the end state. |
| Urban, near traffic or light industry | Every 3 to 4 months, or every 6 months | Traffic film, exhaust particulates, soot | Coating gloss and colour; then the drain path, because film binds with dust into a plug. |
| Arid, sandy or high-dust inland | Track cleaning quarterly at minimum; one source says monthly | Fine abrasive mineral dust | Rollers and the track running surface. This is the exposure where a wet lubricant does the most damage. |
| Sheltered suburban or rural | Two to three full cleans a year; one source says annual where rainfall does the work | Pollen, organic matter, occasional leaf litter | Weep holes and the liner, from leaves and insect debris rather than from film. |
The intervals in that column genuinely conflict. An Australian coastal guide says eight to twelve weeks. A Turkish facade guide working from AAMA and Qualicoat references says every three months and monthly at worst. An American coating care sheet says twice a year minimum and four times a year in coastal or corrosive environments. All three are describing real buildings. There is no universal interval, and pretending otherwise is how a specification becomes unenforceable.
The practical response is not to pick the strictest number. It is to pick one, write it down, and tie it to a named person and a named month. A schedule that exists in writing and gets signed is worth more than an optimal schedule that lives in someone's head.
There is a second driver that has nothing to do with weather: duty cycle. The same logic that sets a coastal interval also applies at the other end of the scale. A commercial swing door in a hotel lobby or a hospital corridor may complete more operating cycles before lunch than a bedroom sliding window completes in a year, and its hardware interval should be measured in cycles rather than in months. When you write a maintenance plan for a mixed building, do not use one interval for every opening. Separate the facade-driven items, which are about atmosphere, from the hardware-driven items, which are about how many times the thing moves.
Cleaning architecturally finished aluminium is not a matter of opinion. There is a published guide for it: AAMA 609 & 610, Cleaning and Maintenance Guide for Architecturally Finished Aluminum, currently at its 2015 revision and first issued in the late 1970s in two separate parts. Its stated scope covers anodised and painted architectural aluminium, rolled or extruded, and it names window and door frames explicitly alongside curtain wall, storefronts, mullions, panels, railings and hardware. It distinguishes between anodic coatings and organic coatings, because they do not tolerate the same treatment. It cautions against strong acids, strong alkalis and abrasives, warns about fire and toxicity when solvents are used and about never mixing cleaning agents, and advises that cleaning be carried out in cool, mild conditions.
We raise it for one reason: if a cleaning contractor tells you they have a method for aluminium, ask which standard or which coating supplier's written instruction it follows. The question is not hostile. It is the fastest way to find out whether they have a method or a habit.
Because aluminium windows are finished rather than bare, most of what ruins them is chemical or thermal rather than mechanical. The following items appear in published coating care sheets and manufacturer guides, and each one is specific enough to be written into a cleaning contract.
Dark finishes deserve a line of their own. The temperature rule is not academic on a dark elevation, because a dark coating absorbs more solar radiation than a light one and sits above the safe-to-clean threshold for much longer. A black entry door or a near-black frame on a west-facing wall can easily be too hot to wash at eleven in the morning and perfectly washable at seven in the evening. If your building has dark frames, the cleaning schedule should specify which elevations are cleaned at which hour, and the specification should say so rather than leaving a contractor to discover it.
If a window could nominate the single worst afternoon of its life, it would be the first clean after the trades leave. More than one industry guide makes this point explicitly: the greatest damage to architectural aluminium occurs during post-construction cleaning, not during decades of service.
The mechanism is chemistry. Cement, plaster, lime, mortar splatter and grout residue are strongly alkaline. Left in contact with aluminium, and especially with an anodic or organic coating, they attack it quickly. The published instruction is consistent: wash such contamination off immediately with copious clean water, and do not attempt to scrape it off once it has dried, because scraping scratches the finish far worse than the stain. Keeping protective film or tape on the profiles until the dirty trades are finished is described as the safest approach, and we would add the obvious corollary: the film itself has a removal deadline, and leaving it on too long under strong sun is its own problem. Ask your supplier for both dates and put them in the programme.
This is also the point in a project where the weep holes get sealed, because the same trades who splash mortar on the frame are the ones who run a bead of sealant along the bottom. Two of the most common lifetime defects on a sliding window, coating burn and a blocked drain, are both created in the same week, by the same people, before the building is even occupied.
Glass cleaning is the part everybody thinks they already know, which is why it produces the most expensive mistakes on the list. Four rules cover nearly all of them.
One more distinction is worth having in the handover document, because it prevents a great deal of pointless argument: condensation you can touch is not the same as fogging you cannot. Moisture on a surface is a humidity and ventilation question. Moisture or fog between the panes means the sealed unit has lost its edge seal, and no amount of cleaning, drying or adjusting will change it. That is a glass unit replacement, not a maintenance item, and the usual advice is that the sash and frame stay in place.
This is the table to pin in the plant room. Each row is a real, frequently performed action that is done in good faith, costs money, and either does nothing or makes things worse.
| Symptom reported | What people do | Why it fails | Do this instead |
|---|---|---|---|
| Water on the internal sill | Smear silicone along the external perimeter joint | The water is already inside the frame. You have sealed the outside of a problem that starts on the inside. | Run the pour test. Clear the internal channel and the weep path first, then re-test, and only then consider the perimeter. |
| Sash is stiff to slide | Spray more lubricant, repeatedly | Stiffness after a proper clean usually means a flat-spotted roller, a drifted height or a scored track. Lubricant masks it for a week while the roller keeps cutting the track. | Clean, then test dry. One source is blunt on this: if it still grinds after a thorough clean, the rollers are at end of life. Continuing to run them gouges the aluminium track, which is not repairable. |
| Draught at the meeting stile | Force the handle harder, or adjust the lock keep | On a sliding unit, the pile or brush seal is a contact seal by construction. Forcing the mechanism transfers load into the hardware, not into the gap. | Inspect the pile for flattening and the gasket for compression set. If it has taken a permanent set, replacement is the only fix. |
| Frame looks chalky or dull | Wash it harder, more often | Chalking is the coating binder breaking down under UV, releasing pigment. It is degradation, not dirt. | One guide gives the test: if the chalky residue returns within weeks of a proper wash, the problem is past the point of cleaning and needs assessment. |
| Grey or green film in a corner | Treat with a biocide and repaint over weep holes later | Persistent mould usually signals standing moisture from a blocked drain or failed seal. Killing the mould does not remove the moisture. | Treat the growth, then find and clear the moisture source, and never let anyone paint or caulk over the drainage openings while doing it. |
Every maintenance schedule quietly assumes you can reach the window. On the ground floor that assumption holds. Above the second floor it is the single largest cost in the whole programme, and it is decided at design stage, long before anyone buys a squeegee.
For horizontal sliding windows there is a published procedure worth knowing. One North American manufacturer's care document describes removing the operating sash for cleaning: disengage the lock, slide the sash to the centre so it clears the anti-lift stop located in the head of the frame, grip it at the centre on both sides, lift it up into the frame, swing the bottom out towards the room, and lift it away. It adds two warnings that matter. On large units the sash can be heavy, so have a surface ready to lay it on and get a second person. And it carries a caveat that is more honest than most: if the window has only one operating panel, that manufacturer does not recommend cleaning the exterior glass from inside the home at all, and advises a hose and a telescoping squeegee instead for anything above ground floor.
That caveat is the useful part. A sliding window with one fixed and one moving panel always leaves one external face that cannot be reached from the room. Inward-opening and tilt-and-turn formats solve the same problem differently, by bringing the sash into the room so the outer face is accessible from inside; UK suppliers of easy-clean sash and vertical slider systems describe exactly this, with sashes tilting inward to give safe access to the outer pane without ladders or scaffolding, which is why those formats are specified for upper-floor apartments and dense urban sites.
So the question to answer before ordering is not "how do we clean it" but "from where, and how often, and by whom, and what does that cost per visit." Glass-heavy projects make this sharper rather than easier, because the cleaning burden scales with glass area and with the length of joint and frame, not with the number of openings. A china 100 aluminum sun room is mostly glass by area and mostly frame by perimeter, and both numbers grow faster than a room's footprint does. Decide the access strategy while the scaffold is still priced as a construction item, not after handover.
A maintenance programme that pretends everything is maintainable produces false comfort. These are the conditions where the correct response is a part, not a task.
A maintenance document is only as good as the data behind it, so here is the honest split for our own range. Some of it is directly useful to whoever writes your cleaning contract, and some of it is not about cleaning at all, and pretending otherwise would waste your time.
| Published | What it does and does not tell you |
|---|---|
| 6060-T66 alloy, 1.8 to 2.5 millimetres depending on series | Useful for the substrate, not for the finish. The coating is what you clean, and coating care depends on the coating system, not on the alloy. Ask for the coating specification and its written care instruction. |
| PA66GF25 multi-chamber insulation strips | A thermal-break material, and one that solvents should not meet. It is another reason the solvent rule above is not optional. |
| Track geometry: 53 and 31 millimetre track heights on the TS130 side-press window; 43 millimetre high track and 8 millimetre low track on the SL126/182 series; 45 millimetre high-ｄｒｏｐ bottom track with an optional 85 millimetre water barrier on the SL112/167; passive check-valve drainage on the SL155 | This is the data your cleaning method has to respect. A deep high-ｄｒｏｐ track and a near-flush low track do not collect the same debris, do not drain at the same rate and cannot be cleaned with the same tool. Get the detail drawing and mark the drainage path on it before writing the method. |
| 240-hour salt spray test and 5,000 newton shear strength, AW90 series | These are test results on a specific series, they are not air, water or wind performance classes, and they do not transfer to other series. More importantly for this page, they are not a cleaning rating. A coating test tells you how the finish was qualified, not which detergent it will tolerate. |
| CMECH hardware tested to 30,000 cycles | A bench test figure, not a service life, and a handle cycle is not the same test as a loaded sash cycling on its rollers. It is useful context for duty-cycle planning and it is not a maintenance interval. |
| Pulley load ratings of 400, 220 and 200 kilograms, quoted per sash | Per sash, and dominated by glass weight. Relevant here because it explains why a worn roller is a structural matter and not a comfort one, and why running a flat-spotted roller until it scores the track is a serious error. |
| No published air permeability, water tightness or wind resistance class for any series | We say this plainly because it changes what a maintenance report can mean. A defect is a departure from a declared threshold. With no declared class, "the window lets water in" is not a measurement, it is an argument. Ask for the class and the standard behind it if your project needs one. |
What you should ask for in writing, and what no catalogue will volunteer: the care instruction for your specific coating system; the care instruction for the specific hardware in your sash, including whether the rollers are sealed; the sash removal procedure and the sash weight, in kilograms, for the largest panel on the project; a labelled drainage drawing showing where water enters, where it travels and where it leaves; and a spare parts schedule covering pile, gasket, rollers, handles and touch-up material, ordered with the original shipment so that a replacement never becomes a second import.
Most maintenance problems are specification problems that were never visible at quotation stage. These ten lines, filled in before the order, change what you can hold anyone to afterwards. The third column is the one to read.
| Line to fill in | What changes once you fill it | Sounds reasonable, tells you nothing |
|---|---|---|
| Coating system and its written care instruction | You now know which cleaners are permitted and which end the warranty | "Just use soap and water" |
| Roller type: sealed bearing or open wheel | Determines whether lubrication is required at all, and what kind | "Lubricate it regularly" |
| Drainage path, drawn and labelled | Turns "clear the weeps" into a specific, inspectable location | "It has hidden drainage" |
| Whether the sill track liner is removable | Decides whether a blocked weep is a five-minute job or a service call | "The track is easy to clean" |
| Largest sash weight, in kilograms | Sets how many people and what equipment sash removal needs | "Two people can handle it" |
| Which panels are fixed and which move, by opening | Determines which external glass faces are reachable from inside at all | "Standard configuration" |
| Exposure zone of the site | Sets the interval, and sets it differently for facade and hardware | "Normal residential conditions" |
| Spare parts schedule shipped with the order | A worn pile becomes a same-week fix instead of a second import | "Spare parts are always available" |
| Protective film type and removal deadline | Prevents both mortar burn and the film baking on under sun | "It comes off easily" |
| Named owner for inspection and for the log | The only line that makes the other nine enforceable | "Site staff will take care of it" |
Usually no, and the presence of water in the outer track during or just after rain is what the drainage system is for. Do not mop it and do not put a towel in it. The thing that turns it into a problem is when it fails to leave: run the pour test on a dry day, and if the water does not exit through the exterior openings, the path is blocked and needs clearing.
Published answers range from monthly in severe coastal or dusty exposure to two or three times a year in a sheltered inland setting, and they are all describing real buildings. Pick an interval based on your exposure zone, write it down, assign it to a named role, and verify it with the pour test. An interval nobody owns is not an interval.
The sources differ in degree but not in direction. One coating care sheet prohibits high-pressure nozzles and power washers outright because they can damage the finish, the glass seal and the weatherstripping. A UK fabricator warns they can force water past the seals. An Australian guide permits them only at a distance of at least 300 millimetres and on the lowest setting. Adopt the strictest of the three and use a hose with a fan nozzle instead.
Because lubrication was the answer to a question nobody asked. Once the track is genuinely clean and dry, persistent drag or grinding points to a roller with a flat spot, a roller height that has drifted, or a track surface that has already been scored. One hardware guide is direct about the consequence of waiting: continuing to run a failed roller gouges the aluminium track, and that damage is not repairable.
It depends on the configuration, and the honest answer is often no. Some manufacturers publish a sash removal procedure for horizontal sliders, but at least one states plainly that if the window has only one operating panel they do not recommend cleaning the exterior glass from inside, and advise a hose and telescoping squeegee for upper floors. Inward-opening and tilt-and-turn formats are specified precisely because they solve this. Decide it at design stage, not at handover.
More than any single task on the list. Without a dated record of what was cleaned, inspected, adjusted and replaced, there is no way to distinguish fair wear from neglect from a manufacturing defect if a claim is ever made. The log does not need to be elaborate. Date, opening reference, what was done, what was found, who signed. It is the cheapest insurance in the building.
If you are writing a maintenance specification for a project and want the drainage path, the track geometry, the coating care instruction and the spare parts schedule for the specific series you are buying, ask for them in writing before the order rather than after handover. We can supply the drawings, the care documents and a parts list that ships with the goods. What we cannot do is tell you which cleaner is approved for a coating we did not apply, or take responsibility for a cleaning method we did not see.
