Seam welding wheels are consumables, and consumables punish two opposite mistakes equally: running out, and hoarding. A line down for want of a wheel costs production; a shelf of ten spare wheels costs cash and hides the real wear data. The way out is to understand what actually wears on a wheel, and to pair that with a supplier who can deliver replacements fast enough that you do not need to insure against them.
How a wheel actually wears
Nothing eliminates wheel wear — the goal is wear that is slow, even, and therefore predictable. On a properly specified tungsten-copper wheel, the wear modes are:
- Face growth. The contact face widens as it wears, which gradually changes current density and seam width. Dressing restores it; the interval between dresses is your primary wear metric.
- Diameter loss. Every dress and every running hour removes material from the OD. On most machines the wheel wears toward the position the arm was set up for (our wheels are cut 0/+0.5 mm on OD for exactly this reason) — but past a point, diameter loss changes wheel speed and seam geometry enough to trigger replacement.
- Profile rounding. Bevel edges round, R profiles grow, knurls dull. When redressing can no longer restore the profile, the wheel is done regardless of remaining diameter.
The failure mode matters as much as the rate. A W/Cu face that dresses evenly can be tracked on a simple diameter or dress-count log. A softer alloy face that mushrooms or glazes — the classic behavior on galvanized strip — cannot, because its degradation is non-linear. Predictability, more than raw life, is the argument for the harder W/Cu 75/25 grade on long, high-force campaigns; W/Cu 70/30 suits lines that dress on a fixed schedule anyway and need the extra conductivity (≈ 42% IACS).
The reorder trigger: diameter, not fear
Once wear is even, the reorder trigger becomes arithmetic. Most operations land on one of these:
- Dress count. Replace after N dresses, where N comes from your own log of when profile restoration stopped holding.
- Diameter window. Replace when the OD wears to a limit your process defines — seam width drift is usually the first indicator.
- Calendar. Coarse, but workable for stable, single-product lines with steady duty.
Whatever the trigger, the correct response to it is a standing order, not a bigger shelf. That only works when lead time is short enough to ride the consumption rhythm — which is why we hold standard sizes at a 10–14 day lead time with no minimum order. When replacement arrives in under two weeks, the reorder trigger is diameter wear, not fear of the supply chain.
One spare per station, not ten
The stock question follows directly. With wheels machined to the same drawing every time and delivery inside two weeks, the rational spare policy for most lines is:
- One spare wheel per station, sized to the next replacement on your wear log — enough to cover changeover plus one normal lead time.
- A full set only when lead time dominates. If your line cannot tolerate even a two-week exposure, the honest fix is a standing schedule, not capital sitting in boxes.
This is a real difference from the copper-alloy world, where unpredictable wheel-to-wheel variation often forces shops to buffer with quantity. Consistent machining to a held drawing revision is what lets the buffer shrink.
Standing schedules for weekly consumers
Lines that consume wheels weekly — can and drum plants, high-speed tube mills — usually graduate from reorder triggers to a standing schedule: agreed dimensions, agreed profile, shipments paced to your consumption rhythm, each batch carrying its inspection record. Nothing about the product changes between shipments; only the delivery dates move. For buyers with incoming inspection requirements, dimension records per wheel — bore H7, OD, runout at 0.02 mm, thickness — ship against the drawing revision you hold.
Reading your own wear data
Two habits make any of the above work better. First, log dress interval per wheel by station — the spread between stations tells you whether the problem is the wheel or the machine (cooling flow, arm alignment, force calibration). Second, keep failed wheels until the replacement is validated: a worn face photographed and sent to us tells us more about your next wheel than a purchase order does.
Get a quote
Send us your wheel drawing — or a worn wheel to measure — plus your consumption rate, and we will quote wheels to the same revision every time. Request a quote: 10–14 days standard, 20–25 days for fully custom geometry, no minimum order.