This series has taken apart a car lamp, an espresso machine, a toothbrush and a cordless drill — heat and oil, oil and alkali, water and foam, dust and vibration. This time it is the two dirtiest appliances in the house: a wet-dry floor washer and a vacuum cleaner, where several of those duty conditions stack up at once — detergent, hair, grit, negative pressure, and someone rinsing the whole thing after every use.
Users judge these machines on three things: does it pick up properly, does it leave streaks, and after a year does it smell or leak. None of those is decided by the motor or the battery. They are decided by a handful of unremarkable soft parts.
The counter-intuitive conclusion: in cleaning appliances, soft parts rarely fail from water. They fail because negative pressure deforms them and hair grinds them away. Moving to a more water-resistant material usually changes nothing.
1. Five soft parts in a floor washer
- Brush-chamber squeegee: the long strip that scrapes water off the floor into the intake. It has to flex, resist abrasion and not mark the floor — mostly TPU, with some premium units using a silicone composite.
- Dirty-water tank gasket: a profiled silicone strip sealing lid to body. This one decides whether the machine dumps dirty water on the floor.
- Intake and hose joint: the transition seal between two rigid tubes, usually a silicone or TPE cone.
- Keypad and display seal: the ring between panel and housing, doubling as key return.
- Battery bay and charging contacts: overmoulding where the charging pins pass through, plus the bay door seal — the first place to suffer when the machine goes back on its dock with wet hands.
A vacuum trades the squeegee for two others: the dust-cup gasket (any leak is lost suction) and the HEPA frame seal (a good filter with a leaking frame filters nothing).
2. Three duty conditions stacked
- Detergent is harder to seal than water: surfactants drop surface tension to around 30 mN/m, so a gap that holds clean water will let detergent solution creep through. This is why a unit passes a water test and leaks once cleaning fluid is added.
- Negative pressure deforms the seal: under suction the gasket is pushed toward the gap. Give it too much compression without a back-up limit and the rubber extrudes into the clearance — the "nibbled edge" you see as a cut ring when you open it.
- Hair and grit are abrasives: hair wraps the brush shaft and grinds the lip and squeegee edge with grit in between. Once it thins, contact pressure drops and the leak starts there.
3. Material assignment: more water resistance is not the answer
| Material | Role in a cleaning appliance | Strength | Weakness |
|---|---|---|---|
| Silicone VMQ | Tank gasket, keypad membrane, contact overmould | Low compression set, good rebound, wide temperature range, low odour | Moderate abrasion and tear strength; detergent immersion needs validation |
| Fluorosilicone FVMQ | Hot water and aggressive detergent contact | Oil and solvent resistance with low-temperature flexibility | Higher cost, lower tear strength than TPU |
| EPDM | Tank and pipe seals | Good water and detergent resistance at low cost | Poor oil resistance, limited colour options, slight odour |
| TPU | Squeegee, wear corners | Abrasion and flex resistance, effective water scraping | Creeps under long compression, lower temperature ceiling |
| TPE | Hose joints, grip overmould | Fast moulding, good bond to plastics | Moderate chemical resistance, can bloom |
The decisive line is silicone's weakness: it handles water but not abrasion. So the industry splits the work — rebound and sealing to silicone, scraping and wear to TPU, aggressive detergent to FVMQ or EPDM.
4. Four places that fail first
1. Dirty-water tank gasket
Profiled, many corners, compression set by the lid latch. A slightly loose latch or a stiff strip and the corners lift. The fix is a ribbed profile positioned against the tank edge, 20–30% compression, with locating posts at the corners.
2. Brush-chamber squeegee
It decides whether the floor streaks. Too hard and it does not pick up; too soft and it rolls over; the wrong angle pushes water back. Typical is 70–80 Shore A TPU with a thin lip and a support rib behind it. Hair collects in the gap between strip and housing, so the structure needs a removable opening.
3. Charging-pin overmould
Wet hands on the dock, splashes while washing — this point takes water and insertion wear together. An O-ring alone gets pushed out by the pin; overmoulding bonded to the housing is more stable, with radial compression at the pin hole.
4. Dust-cup gasket
A leak here is lost suction, which the user feels as "it gets weaker over time". The issue is rarely the material but the match between latch travel and gasket thickness — 0.2 mm of thickness is enough to stop the latch closing properly.
5. Six structural rules
- Face the lip toward the pressure difference: vacuum should seat it, not lift it;
- Add an anti-extrusion limit: above 30% compression the groove needs volume for the rubber;
- Squeegee angle and backing: a thin lip with a back rib resists rolling better than a thicker strip;
- No hair traps: every joint between soft and hard parts is either removable or generously radiused;
- Drain slope: no standing water against the sealing face;
- Serviceability first: a gasket you can take out and rinse outlasts a permanently sealed joint.
6. How to test before shipping
- Whole-unit water resistance: IPX4 splash and IPX7 short immersion on the assembled machine, not on a single part;
- Detergent immersion: the customer's actual cleaning fluid at 1:50 for 72 hours, measuring volume change, hardness change and compression set;
- Vacuum cycling: repeated start-stop to see whether the gasket deforms or extrudes;
- Hair-wrap rig: fixed duration with realistic hair load, measuring suction loss and squeegee wear;
- Drop and tip-over: full tank, 0.8–1 m, checking latch and seal integrity;
- Odour and mould: 72 hours idle, then smell — confirm there is no dirt trap.
7. Ten checks for the design team
- Is the tank seal a profiled strip or a flat gasket, and are the corners located?
- Have latch compression and strip thickness been calculated together?
- Have squeegee hardness, lip thickness and angle been compared on a real machine?
- Is the charging pin sealed by an O-ring or by overmoulding?
- Do the dust-cup gasket and the latch travel match?
- Can hair collect in any soft/hard joint?
- Are the actual detergent type and dilution in the validation plan?
- Under vacuum, can the gasket be drawn into the clearance?
- Is the water test on the assembled unit or on a part?
- After 200 removal cycles, how much compression is left in the serviceable gasket?
Conclusion
The job silicone does here is specific: hold a reliable ring of contact pressure in a wet, dirty, pressure-differential environment. It rebounds well, smells of little and takes compression set slowly — but it is weak in abrasion and needs separate validation against detergent and grease. Hence scraping goes to TPU, aggressive chemistry to FVMQ or EPDM, and sealing and keypads stay with silicone.
Next time a machine leaks at six months or loses suction, do not swap the material first: check whether vacuum has deformed the gasket, whether hair has ground a groove in the lip, and whether there is still enough compression at the latch.
The series continues with another industry and another part. If you have soft parts in cleaning appliances, small kitchen machines or outdoor equipment — water plus abrasives — send the structure, duty condition and test requirements, and we will start with a sealing and material assessment.