Wide-Belt Sander

identity

one-line: A conveyor-fed machine using a wide endless abrasive belt backed by one or more configured working units (rigid or resilient contact rollers/drums, pads/platens, crossbelts, sectioned pads, or combined heads), with the workpiece carried beneath on a powered conveyor bed, used to calibrate panels to a uniform thickness and/or finish-sand them, at continuous throughput [A1; A4; A6; A9]. The class spans three access bands: compact single-phase trade/educational machines (roughly 18 to 25 inches, sometimes 5 HP, still heavy and demanding of extraction and compressed air), professional/offline machines (commonly 25 to 43 inches, single- or three-phase depending on model, one or more heads), and automated industrial lines (multi-head, crossbelt/segmented/planer options, high throughput, often three-phase and integrated with facility dust and material handling) [manual, relayed]. Its capability and safe operating envelope come from the exact machine, head, abrasive, feed, workpiece and dust-control system, not from the name alone. It does the drum sander's dimensioning job at larger scale, on a belt rather than a drum.

name: Wide-belt sander (the standard name) / calibrating sander or abrasive calibrator (when the job is thicknessing panels to tolerance) / production wide-belt / panel sander; model-shorthand by belt width — "25," "37," "43," "52," "64" inches [A1; A4; A9]. NAMING NOTE: distinct from the handheld Belt Sander (a portable hand tool — same belt mechanism, a thousandth of the scale) and from the small-shop drum sander (which wraps abrasive directly around a rotating drum; the wide-belt separates the belt from the drum). A "calibrating sander" is a wide-belt configured with a rigid roller/drum head for thicknessing; a "finishing sander" is configured with a pad/platen head; combined and multi-head machines can sequence both jobs. Exact capability depends on the configured unit, not the marketing noun alone.

family / category: Abrading tools — a member of the COATED-grit POWERED node the File founded, abrading by a coated abrasive (an endless belt of aluminium-oxide/zirconia grit on cloth/paper/film backing), so it sits on the same coated-grit-vs-toothed-steel family boundary flagged OPEN since Run 155 [A1; A2]. Within the dimensioning branch opened by Run 211, it is the industrial ancestor of the drum sander: same job (bring a board to a calibrated thickness by abrasion), but on a belt instead of a drum, at production scale and throughput. It is also the structural cousin of the Belt Sander (the same endless-loop mechanism) doing the drum sander's dimensioning job — the point where the family's linear-loop mechanism and its dimensioning function converge.

✦ how NOT to use it (the heaviest section — five misuse classes)

not-for (jobs it should never do):

• Bulk hogging like a planer. It abrades; it is engineered to remove small, uniform amounts. Timesavers itself separates "widebelt sanding and calibrating" from "abrasive planers" as product families. Over-deep passes overload the belt, burn the work, and destroy the abrasive fast [A9; M9].
• Flattening warped/cupped/twisted stock — the single most persistent conceptual misuse. It calibrates to a uniform thickness but follows the bed: feed a cupped board, get a thinner cupped board. It has no jointer-like reference table. Correct sequence: joint one face first (or use an appropriate sled/fixture and machine-specific procedure), then plane or calibrate the opposite face parallel, then wide-belt for final calibration and finish [M11; M12].
• Short / narrow / thin pieces below the machine's stated minimums — they can tip into the head or be ejected. Minimum length, width and thickness, permitted number and spacing of workpieces, and feed orientation are model- and configuration-specific: Grizzly G0410 specifies a 14-inch minimum length and 3/16-inch minimum thickness [manual, relayed]; a South Bend 37-inch manual warns not to insert more than one board at a time because stock may be ejected [manual, relayed]; an SCM industrial configuration is designed to process several differently thick workpieces with an anti-kickback pressure system [manual, relayed]. Follow the exact machine manual [M1].
• Metal-contaminated stock (nails, staples, screws) — tramp metal generates friction sparks that enter the dust stream and ignite accumulated dust. Exclude foreign material through purchasing/inspection and appropriate detection or separation systems (a magnet does not detect nonferrous contamination); use the DHA and system designer to determine whether spark detection/extinguishment, isolation, abort/diversion, suppression, venting, or other protection is required [M5].

dangerous (injures the operator, and can destroy the shop) — operator hazards and the facility fire pathway are coequal:

• ⚑ PRIMARY CONTROL BLOCK (every operator, every shift). Follow the exact machine manual for minimum length, width and thickness, permitted number and spacing of workpieces, feed orientation, and pressure/anti-kickback settings. Do not defeat guards, interlocks, doors, pressure bars, or emergency stops; check the emergency stops, brake, tracking shutdown and extraction before operation, and do not open doors before the machine has stopped. Wear the specified eye, hearing and respiratory protection. Stop and isolate electrical, pneumatic, stored mechanical, belt-tension and conveyor energy before belt changes, jam clearing, cleaning inside the enclosure, adjustment or service; apply the employer's lockout/tagout procedure where applicable. Keep hands, hair, clothing and tools clear of in-running rolls and the point of operation. Use trained operators. [manual, relayed] (from the reviewer's Fix 4 and Safety Notes; OSHA's Wide Belt Sanders page, read this pass, supports the point-of-operation and in-running-roll controls below).
• Point of operation and in-running rolls. Contact with the abrasive surface may occur during operation, adjustment or maintenance, and hands can be pulled into the sanding area while the operator adjusts the machine. Clothing, hair or hands can be caught and drawn into the automatic feed mechanism, or pinched between stock and rolls. OSHA's controls: a semi-cylindrical feed-roll guard adjustable for any stock thickness (29 CFR 1910.213(p)(1)), a fixed guard on power-transmission pulleys (1910.219), barriers at the loading and unloading ends, a wide metal strip or bar guard on the feed rolls (1910.213(n)(3)), and keeping hands away from the abrasive [M1; M14] (read this pass).
• Ejected stock and kickback. Workpiece ejection, entanglement and unexpected startup can cause severe injury; current manuals warn about ejection, opening doors before stop, and hazardous energy during adjustment and service [manual, relayed].
• Belt break. A broken belt releases abrasive at high speed, can eject fragments, and deposits fragments into the dust stream (a fire pathway); causes include mistracking, excessive removal, faulty joints and moisture-warped belts [M9]. Investigate heat, burning odor, mistracking, abnormal load, damaged joints or broken belts immediately under the safe-stop procedure [manual, relayed].
• Noise, eyes, lungs. Hearing, eye and respiratory protection as the manual and the exposure assessment specify; use exposure monitoring and local exhaust rather than PPE as the only dust control [manual, relayed].
• Wood dust — IARC Group 1 (confirmed human) carcinogen (nasal cavity, paranasal sinuses; hardwoods worst). OSHA names wide-belt sanders, shapers, and routers as generating the greatest amounts of wood dust, and notes conventional control is "not very effective" for sanders. Limits as the original run recorded them from OSHA's eTool: OSHA PNOR 15 mg/m3 total / 5 mg/m3 respirable; NIOSH REL 1 mg/m3; ACGIH TLV 0.5 mg/m3 (western red cedar; inhalable fraction). These are different dust fractions and jurisdictions, are not interchangeable, and workplace users must verify the applicable current standard [M1; M2] (the fractions were read on OSHA's wood-dust chemical-data page in the sibling Sanding Block fix pass).
• ⚑ FIRE & DUST EXPLOSION — a facility-scale hazard added to the above. Frictional belt heat + large volumes of fine combustible wood dust + possible tramp-metal sparks make the wide-belt sander one of the most fire-dangerous machines in a woodshop. Wood dust can be explosible, but its deflagration properties are sample-specific. OSHA's Hazard Communication Guidance for Combustible Dusts says the actual St class "is sample specific and will depend on varying characteristics of the material" and lists wood flour as a typical St 2 example (read this pass). Use representative testing and the facility's Dust Hazard Analysis (DHA) to establish Kst, Pmax, MIE, MEC and required controls; do not assign a universal St class to "wood dust." The earlier "ST1, Kst 100-200" figure is withdrawn. Standards framework. In the United States, evaluate the installation against the current adopted fire/building codes and NFPA 660 (2025), Standard for Combustible Dusts and Particulate Solids, especially its wood-processing chapter, plus the machine instructions and the authority having jurisdiction [manual, relayed] (NFPA 660 is the consolidated successor of the earlier NFPA 652, 654 and 664 cited in this entry's original run; the NFPA page was not reachable). NFPA is a consensus standard rather than a freestanding federal OSHA regulation (OSHA's consensus-standards page, read this pass: NFPA standards are not OSHA regulations and may be required by state or local governments or individual companies; under the General Duty Clause they can supply a feasible means of abatement and show what employers and industry know about a hazard). A DHA is a facility/process obligation under the applicable adopted standard; do not infer compliance from machine size alone. The earlier 5,000 CFM figure applies to a restricted enclosureless-collector pathway, not to every indoor collector, and mechanically fed sanders are not permitted on that pathway [manual, relayed]. The ignition chain: a malfunction or belt-break sends sparks or burning belt into the duct, ignition in the duct or collector, a primary explosion that suspends settled dust, and a larger secondary explosion. Exclude foreign material through purchasing/inspection and appropriate detection or separation systems; the DHA and system designer decide whether spark detection/extinguishment, isolation, abort/diversion, suppression, venting or other protection is required [M5; M6]. Keep extraction running at verified performance and use housekeeping suited to combustible dust. Documented incidents (preliminary, secondary-sourced): Holmes County, OH, March 18 2026 — per the fire chief a malfunctioning belt sander fed sparks into the dust system; the explosion blew a wall off its foundation and injured six; OSHA's inspection remains open and no final cause is implied [M7]. Mountainside High School, OR, May 2026 — a sanding belt may have broken and entered the dust system, where the fire originated; the exact cause was still under investigation [M8]. ⚑ Honest gap: NFPA section text is paywalled and not read in this pass; the belt-sander-as-priority-spark-source language is carried via secondary sources.

damaging (ruins the work or the abrasive):

• Belt burn / loading / glazing — the #1 damage — too much per pass, too fine a grit, too slow a feed, or a worn belt; resinous/dense species (maple, pine, cherry) load and burn "almost instantly." Cure: open-coat belts, faster feed, smaller passes, fresh belts [M10; M9].
• Chatter marks — even lines across the work; usually a machine fault (wrong/worn/out-of-round contact roller, worn bearings, bad belt splice, wrong tension). Set belt tension and supply air to the exact machine-and-belt manufacturer's values; do not transfer pressure values between machines. (The generic vendor psi table by belt construction is removed: Timesavers' own guidance, read this pass, says to set tension pressure to the value recommended by the manufacturer, and notes that too little tension can fail to keep the belt aligned while excessive tension can overpower the tracking cylinder.) [M9; M13]
• Snipe — a dip at the leading/trailing edge from loose feed rollers or bowed stock not lying flat; even hold-down pressure and roller stands are the fix [M11].
• Wavy / washboard surface — soft/damaged rollers for the task, asking the platen to remove significant stock (it conforms instead of bridging), too-high platen pressure, or skipping grits [M9].
• Wrong head for the job — using a pad/platen unit at a calibrating depth tends to give waviness; using a steel calibrating drum for a finish pass tends to leave coarse scratch. Rigid roller units are the usual primary calibration choice and softer rollers and pads favor surface conformity and finish; exact capability depends on the configured unit, not a two-word rule [M10].
• Belt mistracking damaging the work — an off-track belt abrades against housings and gouges/ridges the work edge; causes: dirty laser eye, wrong tension, moisture-warped belt (stored flat on concrete), un-parallel rollers [M9; M13].
• Static-driven loading — high static makes dust cling inside the machine (loading, belt scoring, tracking issues); ground the machine in cold/dry conditions [M9].

beginner-trap (looks fine until it isn't):

• The flatness misconception (the biggest trap) — expecting it to flatten warped stock like a jointer. It calibrates to uniform thickness but follows the bed; a cupped board springs back after the rollers release. Mild cup (1–2 mm) may compress flat; 1/4 inch+ will not [M11; M12].
• Skipping grits — 80 → 180 can leave 80-grit valleys the finer belt can't reach; invisible on bare wood, glaring under clear finish. Use the abrasive and machine manufacturer's sequence; avoid grit jumps that leave scratches the next step cannot remove [M9].
• Too heavy a pass — beginners drop the head to speed removal → burn, scorch, loading, belt break; each grit has a designed removal rate [M13].
• Cross-grain / angled feed — a deliberate slight angle is a pro technique to reduce snipe/burn on dense species, but a true cross-grain feed (especially coarse) leaves scratches far harder to remove and telegraphing under finish [M10].

lineage-misuse (treating it like its relatives):

• Like a PLANER — it abrades (small amounts per pass on most configurations) and can't hog like a planer. Neither a conventional thickness planer nor an ordinary conveyor sander reliably establishes the first flat face from unsupported warped stock: joint one face first or use an appropriate sled/fixture and machine-specific procedure, then plane or calibrate the opposite face parallel [M10; M11].
• Like the handheld Belt Sander or the small drum sander — the scale is decisive: production belt widths and speeds, automatic conveyor feed, and dust volumes that overwhelm any collector not engineered for it. Running one without engineered extraction that meets the maker's requirements and the applicable combustible-dust requirements is not a nuisance — it's a documented explosion hazard [M1; M3].
• Like a calibrating roller (treating the pad/platen head as a heavy-removal head) — the pad conforms to the surface and refines; asking it to remove significant stock tends to leave a wavy surface [M10].

✦ matching (the workbench's first read — match the OWNED tool to the PROJECT)

does — calibrates glued-up panels to a uniform thickness across full width (abrasive planing); surfaces planed/rough-sawn panels to remove machine marks; dimensions figured/reversing-grain stock while avoiding cutterhead tearout; finish-sands panels to 120–180+ before the finishing room; levels edge-banding flush; calibrates/dimensions veneer and thin stock (segmented machines to ~3 mm); removes veneer tape after pressing (cross-belt head); sealer/lacquer sanding between coats (segmented platen, fine grit); runs continuous-feed at production throughput [A1; A2; A4; A6].

suits — production cabinet panels (plywood/MDF box parts, door panels, face frames after glue-up); hardwood flooring blanks; veneered panels and substrate calibration; wide glued-up slabs (especially figured stock a planer would tear out); production lutherie (guitar tops/backs/sides to spec); mass-timber/CLT surfacing; any panel-intensive work where random-orbital time is the production bottleneck [A1; A4; A6; A9].

kit-fit —

• pairs with (upstream): the jointer + planer (rough-flatten and bring stock to within 1/16–1/8 inch before the wide-belt calibrates to final); the panel press / clamps (the wide-belt is the machine glued panels run through after cure).
• pairs with (downstream): the random-orbital sander (clears the residual cross-grain scratch a drum head leaves); the stroke sander (the traditional high-end step before finishing); the spray/finishing line (the wide-belt is usually the last machine before the booth) [A10].
• substitutes for: the small-shop drum sander at scale (wider, faster, flatter, more precise, finishing-capable — the drum sander is the access-limited home version); a planer on figured stock (abrades without tearing).
• can't replace: a jointer or a flattening method (neither a conventional thickness planer nor an ordinary conveyor sander reliably establishes the first flat face from unsupported warped stock; it calibrates to a uniform thickness around whatever shape it's given); a planer for fast bulk removal on straight-grained lumber (the planer hogs 1/16 inch+; the wide-belt takes thousandths); a random-orbital/stroke sander for fully removing cross-grain scratch before finish [A10; M10; M11].

not-for (project-keyed) — don't reach for this when: you need to flatten a cupped or twisted board (that's a jointer first — the wide-belt follows the bed and a thinner cupped board comes out); you need to take a rough board down fast (that's the planer; hogging here burns the belt and scorches the wood); you want to sand a curved/concave edge (that's the Oscillating Spindle Sander); you have any chance of tramp metal — nails, staples, screws — in the stock (a spark into the dust system is the documented fire/explosion pathway) [M1; M5; M12].

access — three bands, not one. Compact trade/educational (roughly 18 to 25 inches, sometimes 5 HP single-phase; hand-fed/offline machines used in schools and colleges, per the reviewer's reading of Timesavers' 13 Series) and professional/offline machines are within reach of small shops, makerspaces and vocational programs; automated industrial lines are floor-standing (800 lb to multiple tons), often three-phase, and need facility-level extraction and material handling [manual, relayed]. Uncommon for first-time makers to own, especially at industrial scale. Every band needs appropriate extraction, guarding, power, compressed air where specified, training, and floor/loading capacity. Setup, grit selection, feed speed, pad pressure and belt tracking require training; mistakes destroy expensive belts, scorch panels, injure operators, or start fires. Sold through industrial machinery dealers/auctions and direct (Timesavers, Grizzly/Baileigh at the accessible end; SCM, Bütfering, Heesemann, Felder at the production end) [A4; A5; A6].

revival — the machine a serious maker seeks time on, not ownership of: at a makerspace/community production shop, a contract cabinet/CNC shop that sells machine time, or a school/vocational program. The reasons a one-person maker needs it: glued-up panels too wide for a thickness planer; figured wood (walnut crotch, curly maple) that tears on a planer; veneer work needing calibrated substrate flatness; volume panel work where hand/orbital sanding is the bottleneck [A10; H3].

anatomy (named parts + what each does)

• Sanding head(s) — the heart of the machine; one to four per machine:
  • Contact drum (roll) — a rubber- or steel-covered cylinder (4–18 inch dia, 30–95 durometer rubber, or solid steel) that presses the moving belt against the work in line contact, concentrating pressure for aggressive stock removal / calibration. Serrated steel drums + coarse belt (36–80 grit) = the "abrasive planer" calibrating head [A1].
  • Platen (graphite pad / smoothing bar) — a flat, non-rotating backup pad (graphite-impregnated cloth over dense felt) spanning the full belt width; broad, low-unit-pressure contact for finishing — longer dwell, finer scratch from a given grit. A pad/platen still removes material; it is simply not normally the primary rigid calibration reference [A1; A2].
  • Combination head — one belt loop over both a contact drum and a platen in sequence; operator engages drum, platen, or both [A1; A6].
  • Cross-belt head — a narrow belt (6–12 inch) running at 90° to feed, cutting cross-grain (aggressive) — veneer-tape removal, perpendicular-grain panels [A1; A2].
  • Segmented platen — the platen divided into individually addressable pneumatic/electromechanical pads, each firing independently behind a thin steel plate + felt + graphite cloth, to follow variable-thickness contours and prevent veneer sand-through; an infeed sensor bar reads the part profile and fires segments accordingly (Bütfering SWT: 26–35 mm segments; Heesemann CSD) [A3; A7].
• Abrasive belt — an endless coated-abrasive loop (aluminium-oxide/zirconia on paper/cloth/film), 25–64 inch wide in production; grit 24–400 by head and job; run ~1,800–5,000 SFPM (~9–25 m/s) for wood [A2; A4; A9].
• Belt-tracking + oscillation system — keeps the belt centered (an off-track belt destroys itself against the frame in seconds): the belt is deliberately oscillated laterally (±½–1 inch) by a pneumatic cylinder tilting an idler roller (distributes wear, prevents edge loading), and an electronic eye (photocell — "POSI-TRAK" on Timesavers) detects drift and actuates the air piston to re-center. Belt tracking may use pneumatic, optical/photoelectric, air-jet or other systems; "laser eye" is not universal. An inline oiler lubricates the pneumatics; if its oil runs low, tracking fails [A4; A9; M13].
• Conveyor / feed belt + bed — a rubber-coated endless conveyor over a flat bed, driven by a separate motor, carrying the workpiece under the head(s); hold-down pressure rollers / shoes grip the work and prevent lift. Feed speed is set by a VFD (≈10–60 FPM production; two fixed speeds on small machines) [A4; A5].
• Table (bed) elevation — motorized raise/lower of the conveyor bed to set cut depth / final thickness, with digital readout to 0.1 mm on industrial machines; advanced machines (Felder findustry) auto-measure incoming thickness by fibre-optic/laser and auto-set the head [A4; A8].
• Dust hoods + collection — close-fitting hoods over each head's belt entry/exit, ducted to a central collector; Provide engineered local exhaust that meets the machine manufacturer's airflow and static-pressure requirements at every active port and maintains transport velocity appropriate to the conveyed residue. OSHA's woodworking eTool gives a general 2,500 to 4,000 FPM transport-velocity range for different wood residues [manual, relayed], but duct velocity alone does not establish hood capture or collector performance; verify at the machine under operating conditions and do not size the system from a generic FPM number. A wide-belt generates enormous fine-dust volume and a fire risk if dust settles [A4; M1].
• Power — three-phase on many industrial machines (220/440 V; sanding motors from the teens to 75 HP); single-phase models also exist: Grizzly G9983 (15 inch, 5 HP) per the original run, and per the reviewer a 37-inch, 10 HP, 220 V single-phase model (G0571) and an 18-inch, 5 HP model [manual, relayed]. Power class is model-specific, not definitional [A4; A5; A6].

lineage

parent / ancestor (upstream):

• Endless coated-abrasive belt technology — the enabling substrate; the wide-belt sander cannot exist before a durable, wide, endless coated-abrasive loop. Widely attributed to 3M's pioneering of cloth-backed endless abrasive belts for industrial grinding (~1915) — UNVERIFIED to a primary source this run, an honest gap [H13].
• Industrial drum / cylinder "sandpapering" machines (late 19th–early 20th c.) — a rotating abrasive-wrapped cylinder over a conveyor, used in shoe and furniture factories; the functional predecessor the wide belt improved on (interchangeable belts, lateral oscillation, longer abrasive life, flatter calibration) [H1; A11].
• The continuous-feed wide-belt machine itself — Timesavers dates its company history to 1946; a 2026 trade profile attributes its origin to R. F. Pendergast working with 3M on a continuous-duty sanding platform [H1; H2]. Without a primary catalog, patent, or dated machine record, treat this as a company-history claim, not a verified first for the class. Timesavers International's separate Dutch lineage (Linden Machines added wide-belt sanders in 1963) is a different corporate history and is not blended with the North American 1946 claim. [manual, relayed] (Timesavers' own About page supports 1946 but not the Pendergast/3M wording).
• Coated abrasive — the functional substrate; like every powered-tip member it CROSSES the toothed-steel family boundary flagged OPEN [A1].

contemporaries / siblings (same coated-grit powered node, different scale/job):

• The Belt Sander / linisher — the same endless-loop mechanism at a portable, owner-operated scale; the wide-belt is its production-scale, conveyor-fed, calibration-capable big sibling.
• The Disc Sander and the flat-pad finishers (Random-Orbital Sander, Orbital Finishing Sander, Detail Sander) — owner-operated surfacers; the wide-belt's platen head does their finishing job at production scale, and a wide-belt-sanded panel still usually hands off to a random-orbital to clear residual cross-grain scratch [A10].
• The Oscillating Spindle Sander — the other production-adjacent abrader, but a concave-edge surfacer, not a flat-face calibrator.
• The cross-family thickness PLANER — the cutting-mechanism tool that does the same dimensioning job by a knife; the wide-belt avoids cutterhead tearout on figured/reversing grain and usually runs after the planer in sequence; abrasive sanding can still produce fuzz, scratches, heat damage, washboarding or veneer sand-through [M10; M11].

variations / descendants (downstream):

• The small-shop drum sander (Performax/Gary Green, US 4,720,940, 1985/1988) — the democratized descendant, built Run 211; this entry is its larger-scale comparator (attributed interpretation, not a proven single line of descent) [H3; H11].
• Calibrating-type, finishing-type and combined configurations. Rigid/grooved roller units are the usual primary calibration choice; softer rollers and pad/platen units progressively favor surface conformity and finish quality; combined and multi-head machines can sequence those jobs. A pad/platen still removes material; it is simply not normally the primary rigid calibration reference. Calibrating, sanding, superfinishing, crossbelt, planer and brush/planetary units may coexist in modular lines (SCM's current material describes rubber-coated roller units spanning high-removal calibration through finishing of veneered and painted panels, multiple hardnesses and sectioned pads) [manual, relayed] [A1; A6].
• Multi-head machines (2–4 heads — e.g. SCM Sandya 10-RCS-110: calibrating steel-drum head + combination finishing head; Sandya 15S: three heads) — calibrate-then-finish in one pass [A6].
• Segmented-platen machines (Bütfering SWT series; Heesemann CSD electronic segmented pad) — individually addressable pneumatic/electromechanical pads that follow variable-thickness contours and prevent veneer sand-through [A3; A7; H12].
• Cross-belt head (a narrow belt at 90° to feed — veneer-tape removal, cross-grain panels) [A1; A2].
• Top-and-bottom (double-sided) machines (Grizzly G0486/G0581) — sand both faces in one pass [A4].
• Advanced auto-calibrating machines (Felder Format-4 findustry — fibre-optic/laser auto thickness measurement, head auto-adjust to 0.1 mm; Steinemann SATOS mineral-cast frame) [A8; H12].

TREE NOTE. This entry adds the Wide-Belt Sander as a built MEMBER on the COATED node's dimensioning branch — the family's first member with production-scale configurations (a corpus reading about the upper end) and an industrial comparator of the drum sander (attributed interpretation, not a proven line of descent). It is integrated into the _Abrading Family — Lineage Tree this run with a light inline touch: the WIDEBELT ancestor ghost (created Run 211) is promoted to a built node, the drum sander's "democratize" descent now runs from a built ancestor, one node-table row is added, and a dashed same-mechanism link is drawn to the Belt Sander (endless loop) — NOT a refresh.

history & origin

The honest headline: the wide-belt class has a company-history anchor — Timesavers' own 1946 date — and an unusually strong PATENT spine from the 1970s–80s (belt-edge sensing, platen oscillation, feed, dust collection). What it lacks is a primary 1940s catalog, a museum accession, and a verified date for the foundational 3M coated-abrasive belt.

• Timesavers Inc. — a company-history date (1946). Timesavers dates its company history to 1946; a 2026 trade profile says it was founded in Robbinsdale, MN by R. F. Pendergast "in a joint effort with 3M to create the first continuous duty sanding platform." That wording is the trade profile's, not a verified first for the class (see lineage). Now in Maple Grove, MN; ~300 machines/year; a claimed ~25,000 units in the field (manufacturer self-report). Entered its 80th year in 2026 [H1; H2]. The first-machine technical configuration (belt width, head count) is not documented in a primary catalog this run — an honest gap.
• The European makers (deep roots). Heesemann (Karl Heesemann Maschinenfabrik) — founded 1933, Bad Oeynhausen, Germany; the oldest of the major makers by founding date; credited with the first profile sander, the cross-belt unit, and the patented CSD® (Computerized Sanding Dynamics) electronic segmented pad (patent numbers not located — gap); now in the HOMAG Group [H12]. Bütfering (Beckum, Germany; founding year not located) — wide-belt specialist, now HOMAG. SCM Group / DMC (Italy) — the Sandya line. Costa Levigatrici (Schio, Italy, founded 1982). Steinemann (St. Gallen, Switzerland) — the SATOS series, "first wide-belt sander with a mineral-cast frame" for vibration damping [H12].
• The PATENT spine (the strongest proof — all verified on Google Patents this run except where flagged):
  • US 3,789,552, "Tracking mechanism for belt sanders," Singer Co., 1974 — a pivoting-idler linkage that self-corrects belt drift [H6].
  • US 3,832,807 (Kiser & Stinn, 1974), US 3,859,758 (Fair, 1975), US 3,872,627 "Wide Belt Sanding Machine with Improved Dust Collector" (Schuster, 1975) — a Timesavers engineering cluster on feed, head adjustment, and dust collection [H9].
  • US 3,971,166, "Belt position sensor for wide belt sanding machine," Timesavers Inc., 1976 — the pneumatic air-nozzle belt-edge sensor (opposing air streams detect belt position), the ancestor of the "electronic eye" tracking that defines the modern machine [H4].
  • US 4,294,044, "Tracking mechanism for a belt type sander," Emerson Electric, 1981 — spring-biased tilting idler drum [H7].
  • US 4,651,474, "Wide belt sanding machine with platen oscillating means," Eugene C. David / Timesavers, filed 1986, issued 1987 — the oscillating platen head, the finishing-side refinement [H5].
  • US 4,369,601 (Gerber, 1983), "Apparatus for tracking belt for abrasive grinding machine" — optical belt tracking; verifiable per the review (Justia) [manual, relayed]; the earlier "unverified" flag is removed [H10].
  • Context: US 3,971,166's specification calls the abrading head and powered conveyor conventional and cites earlier tracking art, evidence the class predates that 1976 patent [manual, relayed].
  • US 5,184,424 (1993) — a purely-mechanical self-correcting wide-belt tracking apparatus [H8].
• The democratization (the descendant's origin). Gary L. Green / Performax, mid-1980s — the small-shop open-end drum sander (US 4,720,940, filed 1985, issued 1988), built explicitly because a wide-belt sander wouldn't fit a garage shop. The primary trade source for this link: Woodworker's Journal, 2014 [H3; H11].
• The foundational coated-abrasive belt (~1915, 3M). Repeated across secondary sources but not traced to a primary citation this run — an honest gap [H13].

Proof of existence (summary): a company-history date for a leading maker (Timesavers, 1946; the 3M joint-effort wording is a trade-profile claim) [H1; H2]; a verified cluster of dated, named US patents (Singer 1974; Timesavers 1974–87; Emerson 1981) proving the machine was an established, actively-improved industrial category by the early 1970s [H4; H5; H6; H7; H9]; living makers with deep roots (Heesemann 1933; Costa 1982) [H12]; trade-encyclopedia records (WoodTecPedia/Höchsmann) and current manufacturer catalogs [H12; A9]. No museum accession with a catalog number was located — VintageMachinery.org does not index industrial wide-belts (its records skew to small-shop/consumer machines), which is itself a sourced finding, not just an absence [H-gap].

human side & trades — who runs and relies on it

• Cabinet & furniture shops — the primary market. Calibrating glued-up door and face-frame panels to uniform thickness, removing planer snipe and glue squeeze-out before the finishing room. The machine sits at the end of milling and the start of finishing — a production bottleneck when it goes down [A1; A10].
• Flooring mills — dimensioning hardwood plank to final thickness at high volume [A9].
• Panel & plywood plants — chipboard, MDF, plywood calibration and surface finishing; veneered-panel substrate flatness [A2; A9].
• Production lutherie — guitar/instrument factories thickness tops, backs, and sides to spec; individual luthiers usually can't afford or house one and "befriend cabinet shops" or barter machine time. (One forum builder: finding or building a thickness sander "is one of the rites of passage for acoustic builders") [H-forum].
• Fire culture — specific and serious. "Never leave it running unattended" is a fire rule: a loaded/overheating belt or a jam is the difference between catching a smolder and arriving at a collector fire. Shops with one as a production tool typically run spark detection in the duct, abort gates on return-air systems, scheduled (not run-to-failure) belt changes, end-of-shift vacuuming (never compressed-air blow-down), and a Dust Hazard Analysis where the applicable standard calls for one. A loss-control consultant's read of a shop: "are the people managing the sawdust, or is the sawdust managing the people?" [M6].
• Belt-tracking & grit lore — keep the laser eye scrupulously clean of dust (a common neglect point); verify rollers parallel with a dial indicator (uneven rollers are the top structural cause of chronic mistracking); hang belts (never store flat on concrete — moisture warps them out of round); vary the stock's entry point to spread belt wear; use the abrasive and machine manufacturer's grit sequence; for resinous species, open-coat zirconia + faster feed + scheduled belt changes [M9; M13].

proof of existence

• Company-history anchor (not a verified first for the class): Timesavers' own 1946 date; the Robbinsdale/Pendergast/3M wording is a 2026 trade-profile claim [H1; H2].
• Patents (named, dated, verified except where flagged): US 3,789,552 (Singer, 1974) [H6]; US 3,832,807 / 3,859,758 / 3,872,627 (Timesavers cluster, 1974–75) [H9]; US 3,971,166 "Belt position sensor for wide belt sanding machine" (Timesavers, 1976) [H4]; US 4,294,044 (Emerson, 1981) [H7]; US 4,651,474 "platen oscillating means" (David/Timesavers, 1986/87) [H5]; US 5,184,424 (1993) [H8]; US 4,369,601 (Gerber, 1983, "Apparatus for tracking belt for abrasive grinding machine" — verifiable per the reviewer [manual, relayed]) [H10]; and the democratizing descendant US 4,720,940 (Green/Performax, 1985/88) [H11].
• Living makers with deep roots: Heesemann (1933), Costa Levigatrici (1982), SCM/DMC, Bütfering, Steinemann, Felder [H12].
• Trade-reference & catalog record: WoodTecPedia (Höchsmann); current manufacturer catalogs (Timesavers, SCM, Grizzly, Bütfering, Felder) [H12; A9].
• Gaps: no primary 1940s Timesavers catalog; no museum accession — VintageMachinery.org does not index industrial wide-belts (a sourced absence); Heesemann CSD patent numbers not located; the foundational 3M coated-abrasive-belt date (~1915) unverified.

references

• [H1] Miller Wood Trade Publications (National Hardwood Magazine), "Timesavers LLC: Eight Decades of Innovation in Wide Belt Sanding and American Manufacturing," Mar 2026 (founded 1946 by R.F. Pendergast w/ 3M, Robbinsdale MN). millerwoodtradepub.com — kind: blog/trade
• [H2] Capital Woods Machinery, "Timesavers Inc. History," Jul 2019. capitalwoodsmachinery.com — kind: blog/trade
• [H3] Woodworker's Journal, C. Marshall, "SuperMax Tools: Singular Focus on Woodworking's Most Tedious Task," 2014-12-23 (the wide-belt → drum-sander democratization, Gary Green's garage-shop motive). — kind: blog/trade
• [H4] US Patent 3,971,166, "Belt position sensor for wide belt sanding machine," Timesavers Inc., 1976. Google Patents. — kind: patent
• [H5] US Patent 4,651,474, "Wide belt sanding machine with platen oscillating means," E. C. David / Timesavers, filed 1986, issued 1987. Google Patents / FreePatentsOnline. — kind: patent
• [H6] US Patent 3,789,552, "Tracking mechanism for belt sanders," Singer Co., 1974. Google Patents. — kind: patent
• [H7] US Patent 4,294,044, "Tracking mechanism for a belt type sander," Emerson Electric, 1981. Google Patents. — kind: patent
• [H8] US Patent 5,184,424, "Self correcting belt tracking apparatus for widebelt abrasive grinding machine," 1993. Google Patents. — kind: patent
• [H9] US Patents 3,832,807 (Kiser & Stinn, 1974), 3,859,758 (Fair, 1975), 3,872,627 "Improved Dust Collector" (Schuster / Timesavers, 1975). Google Patents. — kind: patent
• [H10] US Patent 4,369,601 (Gerber, 1983), "Apparatus for tracking belt for abrasive grinding machine" — verifiable at https://patents.justia.com/patent/4369601 per the review [manual, relayed]; not fetched in this pass. — kind: patent
• [H11] US Patent 4,720,940, "Rotary Drum Sander," G. L. Green / Performax, filed 1985, issued 1988 (the democratized descendant). Google Patents. — kind: patent
• [H12] Stiles Machinery (Heesemann brand, founded 1933; CSD® segmented pad); HOMAG (Bütfering); WoodTecPedia/Höchsmann (Costa Levigatrici 1982; wide-belt lexicon); Steinemann SATOS; SCM Group wide-belt sanders. — kind: blog/trade + museum/archive (trade encyclopedia)
• [H13] 3M endless coated-abrasive belt origin (~1915) — repeated across secondary sources, NOT traced to a primary citation this run. — kind: blog/trade (UNVERIFIED)
• [A1] A. J. Hamler, "Sanding head variations have their roles," Woodshop News, Nov 2012 (contact drum vs platen vs combination vs cross-belt; calibrating vs finishing). — kind: blog/trade
• [A2] Fintech Abrasives, "Wide Belt Sanding Head Variations & Their Benefits," Oct 2023. — kind: blog/vendor
• [A3] Fintech Abrasives, "Segmented Platen: About & Machine Setup," Oct 2023 (segment widths, sensor bar, belt speeds). — kind: blog/vendor
• [A4] Grizzly G0410 product page (43 inch, 25 HP, 3-phase, VFD 15–60 FPM, 0.1 mm DRO); Grizzly G0486/G0581 (double-sided). grizzly.com. — kind: manufacturer product page
• [A5] Grizzly G9983 spec sheet (15 inch, 5 HP single-phase, 2-speed feed). — kind: manufacturer spec sheet
• [A6] SCMI Sandya 10-RCS-110 (2-head, 30 HP + 18 HP, 43 inch) + Sandya 15S (3-head) dealer listings. — kind: dealer listing
• [A7] Bütfering SWT 134/RRC (35 mm segments, 18 kW) + SWT 225 CH OPTIMAT (26 mm segments) dealer listings. — kind: manual/vendor
• [A8] Felder Format-4 findustry product page (fibre-optic/laser auto thickness measure, 0.1 mm, 2.4–19 m/s). felder-group.com — kind: manufacturer product page
• [A9] Timesavers product pages (13/23/33/53 series; up to 64 inch). timesaversinc.com — kind: manufacturer product pages
• [A10] Woodweb Cabinetmaking KB, "Should a Widebelt Sander Have a Platen?", 2014-06-16 (platen = what separates a drum from a wide belt; wide-belt → stroke sander → finishing). — kind: forum
• [A11] Wurth Baer Machinery, "Drum Sander vs. Wide Belt Sander"; Benchmark Abrasives, "Difference Between Wide Belt Sander and Drum Sander." — kind: blog/vendor
• [M1] OSHA eTool, Woodworking — Wide Belt Sanders (names wide-belt among greatest dust generators; point-of-operation + in-running-roll hazards). osha.gov — kind: regulator (point-of-operation page read 2026-10-16)
• [M2] OSHA, Wood Dust Hazard Recognition + eTool Wood Dust Carcinogens (IARC Group 1; PEL 15/5; NIOSH REL 1; ACGIH TLV 0.5 WRC). osha.gov — kind: regulatory (peer-reviewed basis)
• [M3] NFPA 664, Standard for the Prevention of Fires and Explosions in Wood Processing and Woodworking Facilities — now consolidated into NFPA 660 (2025) [manual, relayed]. The earlier claims (enclosureless-collector prohibition; 5,000 CFM) came via Stonehouse Process Safety / SonicAire / metalmanufacturing.net summaries (full text paywalled) and are restated in this copy only in the reviewer's narrower form. — kind: consensus standard (secondary summaries)
• [M4] NFPA 652, Standard on the Fundamentals of Combustible Dust — predecessor of NFPA 660 (2025). The earlier "wood dust ST1, Kst 100-200" claim is withdrawn (explosibility is sample-specific); the Sept 7 2020 DHA date is an old deadline under a consensus standard, not universal law. Via Woodworking Network / Höcker North America, Jul 2020. — kind: consensus standard (secondary summary, superseded)
• [M5] Hastings Insurance, "Woodworking Operations — Fire and Dust Explosion Hazards," Mar 2025 (tramp-metal control, spark detection/suppression, housekeeping ≤1/8 inch). hastingsinsurance.com — kind: blog/trade (loss-prevention)
• [M6] Woodworking Network, "Tips for Woodworking Fire Safety" (Pennsylvania Lumberman's Mutual / M. Culbreth), Apr 2023. — kind: blog/trade
• [M7] Woodworking Network, "Malfunctioning Belt Sander, Sawdust Causes Explosion at Ohio Moulding Facility," 2026-03-20 + Cleveland 19 News, 2026-03-18 (Holmes County, OH; 6 injured). — kind: news (secondary, dated)
• [M8] Right Now Oregon / KGW, Mountainside High School woodshop fire, May 2026 (broken belt into dust duct). — kind: news (secondary, dated)
• [M9] Uneeda Enterprizes, "Top 8 Wide Belt Sanding Problems," #1-4 and #5-8 (chatter, snipe, washboard, mistracking, static, tension psi, belt break). sandpaper.com — kind: blog/vendor
• [M10] Woodweb KB, "Wide-Belt Sanding Maple: Burning Problem" (resinous-wood burn/load; platen-vs-drum head; angled feed). woodweb.com — kind: forum
• [M11] Woodweb KB, "Snipe with a Wide-Belt Sander" + "Flattening Oversize Table Tops" (snipe causes; follows-the-bed flatness). woodweb.com — kind: forum
• [M12] Sawmill Creek, "Flattening door with wide belt sander" (the cup-springs-back flatness trap). sawmillcreek.org — kind: forum
• [M13] Timesavers, "Common Wide Belt Sander Problems" (tracking, electronic eye, oiler, too-heavy passes; tension to the manufacturer's value). timesaversinc.com — kind: vendor guidance (read 2026-10-16)
• [M14] OSHA 29 CFR 1910.213, Woodworking Machinery Requirements (incl. 1910.213(p)(1) nip-point guards; (n)(3) in-running-roll guard). osha.gov — kind: regulator
• [N1] OSHA, Hazard Communication Guidance for Combustible Dusts (https://www.osha.gov/publications/3371combustible-dust) — St class is sample-specific; wood flour listed as a typical St 2 example. — kind: regulator (read 2026-10-16)
• [N2] OSHA, Combustible Dust Consensus Standards (https://www.osha.gov/combustible-dust/consensus-standards) — NFPA standards are not OSHA regulations; General Duty Clause use is limited. — kind: regulator (read 2026-10-16)
• [N3] Named by the reviewer and carried [manual, relayed], not re-read: NFPA 660 (2025) listings; ISO 19085-8:2024; Grizzly G0410 and G0571 product pages; South Bend SB1096 owner's manual; SCM dmc eurosystem brochure; Timesavers 13 Series and About Us pages; Google Patents US3971166A; Justia US4369601; OSHA 1991 sander/ventilation explosion accident report; OSHA inspection record for Holmes Custom Moulding (March 2026, open); Cleveland 19 and KPTV reports (preliminary accounts). — kind: mixed (manufacturer product pages, owner/operator manual, brochure, patents, regulator, consensus standard, news)

✦ the genuinely-new structural reading (the entry's headline — held as interpretation)

Three braided readings, each an interpretation of the sourced facts. The honest part up front: reading 1 is the genuinely new one (the abrading family's first member with PRODUCTION-SCALE configurations — it introduces a WORKSHOP↔PRODUCTION axis, and with it a facility-scale combustible-dust pathway added to the ordinary machine hazards); reading 2 is the democratization inverted in time (the corpus's first ancestor built after its cheaper descendant — a form-follows-failure not of function but of ACCESS); reading 3 is the convergence + matching reading (it holds both ends of Run 211's SURFACE↔DIMENSION axis in one machine, and it's the Belt Sander's mechanism doing the drum sander's job — and uncommon for first-time makers to own, especially at industrial scale).

1 — The abrading family's first member with PRODUCTION-SCALE configurations: prior abraders are owner-operated; the upper end of this class is a production machine — and it adds a facility-scale combustible-dust pathway to the operator hazards, it does not replace them. Walk the abrading family the corpus has built and every prior member is an owner-operated tool — a person, a workpiece, a tool held or fed by hand:

• the hand tools (Rasp, File, Riffler, scrapers, the Japanese saw-rasps) — one hand, one stroke;
• the benchtop machines (Bench Grinder, Disc Sander, Oscillating Spindle Sander) — bolted to a bench, fed by hand;
• the handheld powered abraders (Random-Orbital Sander, Orbital Finishing Sander, Detail Sander, Belt Sander, Angle Grinder, Die Grinder) — carried in the hand;
• the drum sander (Run 211) — a floor machine, owner-operated, one board at a time.

The wide-belt class reaches further: floor-standing (hundreds of pounds to several tons), a powered conveyor feeding stock at continuous throughput, and at the industrial end sanding motors in the tens of horsepower and three-phase power. It always needs engineered extraction, and an installation must be evaluated against the combustible-dust requirements that apply to it (see dangerous). This draws a WORKSHOP-scale ↔ PRODUCTION-scale axis through the family. The consequence is an added hazard class, not a transferred one: frictional belt heat + large volumes of fine combustible dust + the chance of a tramp-metal spark or a broken belt = a documented dust-fire/explosion pathway, where a spark or broken belt enters the duct, ignites in the collector, and a primary explosion dislodges settled dust into a larger secondary one [M1; M3; M5; M7]. OSHA names the wide-belt sander by name as among the woodworking machines generating the greatest amount of dust [M1]; in March 2026 a malfunctioning belt sander feeding sparks into a dust system was reported to have blown a wall off an Ohio moulding plant and injured six (preliminary, fire-chief-attributed account; OSHA's inspection remains open, so no final cause is implied) [M7]. Wide-belt sanding adds a facility-scale combustible-dust pathway to the ordinary machine hazards; it does not transfer danger away from the operator. Point-of-operation contact, in-running rolls, ejected stock, belt failure, noise, dust exposure, electrical energy and unexpected startup remain coequal controls (see dangerous).

Honesty flag. That every prior abrading member is owner-operated and that the wide-belt class includes production-scale, infrastructure-dependent machines are sourced (the prior entries' own theses; the wide-belt's specs and dust infrastructure across [A4; A6; M1; M3]). That this constitutes "the family's first member with production-scale configurations / a WORKSHOP↔PRODUCTION axis" is my synthesis, labelled, and describes the upper end of the class only: compact single-phase and hand-fed machines also exist. The earlier "danger moves from the hand to the building" reading is withdrawn per the reviewer's Fix 3.

2 — The democratization INVERTED in time: the corpus's first ancestor built AFTER its cheaper descendant — a form-follows-failure not of FUNCTION but of ACCESS. The corpus's usual move is to build a descendant that fixes a functional failure of its ancestor: the Random-Orbital Sander frees the orbital sander's constrained bearing and kills the swirl; the slow-speed cool grinder engineers the temper-drawing out of the Bench Grinder. Here the relation runs the other way. The wide-belt sander is the capable, expensive ancestor; the drum sander (Run 211, built first) is the descendant that gave up capability to gain access. The trade is documented in the primary source: Gary Green built the small drum sander in the mid-1980s precisely because "if you didn't like to sand by hand… the only other option was to invest in a wide belt sander — and Green didn't have room for one of those in his garage shop" [H3]. So the "failure" the drum sander fixes is not a failure of function — the wide-belt works beautifully — it is a failure of ACCESS: too wide, too costly ($5,000–$15,000 then), too power-hungry (industrial machines of that era were three-phase), too big for a one-car garage. The drum sander pays for that access by sacrificing exactly what makes the wide-belt a production tool: belt-vs-drum (no interchangeable-belt economy, the abrasive wraps the drum), the platen head (no true finishing head — the drum sander only dimensions, then hands off to a random-orbital), belt speed and width, multi-head one-pass capability, and, typically, a smaller motor and single-phase power. This is Petroski's form-follows-failure read along an unfamiliar axis — democratization is the act of fixing an access failure by sacrificing function — and the entry is built ancestor-after-descendant, the established corpus pattern (Grindstone built after the Bench Grinder; the orbital finishing sander's FIXORB built after its Random-Orbital Sander descendant).

Honesty flag. The wide-belt → drum-sander democratization, with the cost/space/power motive, is directly and primarily sourced (the 2014 Woodworker's Journal SuperMax history quoting the Performax origin) [H3]. The list of what the drum sander sacrifices (platen, belt economy, multi-head, three-phase) is sourced to the two machines' specs [A1; A6; Run 211]. The "democratization = fixing an access failure by sacrificing function" framing is my interpretation, labelled.

3 — The CONVERGENCE node: it holds BOTH ends of Run 211's SURFACE↔DIMENSION axis in one machine, and it is the Belt Sander's mechanism doing the drum sander's job — and uncommon for first-time makers to own, especially at industrial scale. Run 211 drew a new SURFACE-quality ↔ DIMENSION axis through the finishing branch: the drum sander dimensions (calibrates to thickness), the flat-pad finishers surface (improve quality without targeting a dimension). The wide-belt sander is the first member to hold both ends of that axis in a single machine: a contact-drum head (coarse belt, line contact) dimensions — abrasive planing to a calibrated thickness — and a platen head (fine belt, broad low-pressure contact) surfaces — finish-sanding to a finish-ready surface; a combination or multi-head machine does both in one pass (calibrate at head 1, finish at head 2) [A1; A2; A6]. It is also the convergence of two mechanism-and-function strands the corpus tracked separately: the endless-loop mechanism of the Belt Sander (a coated-abrasive loop over rollers, linear contact) carrying the dimensioning function of the drum sander — the belt-sander's mechanism doing the drum-sander's job, at production scale. The matching consequence is corrective: ownership is uncommon for first-time makers, especially at industrial scale (compact single-phase and school machines exist). For many makers the honest match-answer is aspirational — your drum sander is the home version of this; if you need calibrated wide panels, figured stock without tearout at volume, or true finish-sanding before the spray booth, this is the production machine you'll meet at a cabinet shop, a flooring mill, or a makerspace, not own.

Honesty flag. The contact-drum-dimensions / platen-surfaces / multi-head-does-both facts are firmly manufacturer-sourced [A1; A2; A6]. The "holds both ends of the SURFACE↔DIMENSION axis in one machine" and "Belt Sander mechanism doing the drum sander's job" framings are my synthesis on top of those facts, labelled. The ownership point is sourced to the larger machines' access profile (floor-standing, industrial-dealer-sold) [A4; A6]; compact single-phase and school machines are a counterexample, so the claim is qualified.

mechanism & specs

• How a wide belt calibrates/finishes vs a drum sander: the wide-belt separates the abrasive (the belt) from the backup (drum or platen) — so one machine runs interchangeable belts across grit ranges, carries multiple head types, sands far wider (25–64 inch), and runs higher belt speeds than a drum sander (which wraps abrasive directly on the drum) [A1; A11]. Like the drum sander and unlike a planer, it abrades (grit scratches, doesn't lift fiber) — so it avoids cutterhead tearout on figured/reversing grain, though abrasive sanding can still produce fuzz, scratches, heat damage, washboarding or veneer sand-through [M10; M11].
• Calibrating (contact-drum) role: line contact concentrates pressure; coarse belt (36–80) + serrated steel drum is the heavy-removal configuration (1/64 to 1/8 inch per pass on some machines; model-specific) to bring panels to a uniform, predictable thickness (±0.1 mm on servo-controlled machines) [A1].
• Finishing (platen) role: broad low-pressure contact, finer grit (100–220), longer scratch pattern that sands out quickly with a random-orbital — surface quality, not tolerance [A1; A2].
• Belt widths: 25 / 37 / 43 / 52 / 64 inch typical; Grizzly G0410 = 43 inch; Timesavers 33-series up to 64 inch [A4; A9].
• Motor / power: industrial machines run from the teens to 75 HP, often three-phase; compact models run 5 HP single-phase (Grizzly G9983, 15 inch); SCM Sandya 10-RCS-110 = 30 HP (head 1) + 18 HP (head 2) [A4; A5; A6].
• Feed speed: ~10–60 FPM (VFD on production machines; two fixed speeds, ~13/16 FPM, on small single-phase machines) [A4; A5].
• Belt speed: ~1,800–5,000 SFPM for wood; segmented-platen veneer work 12–18 m/s; sealer sanding slower (5–9 m/s) [A2; A3].
• Grit by job: 24–50 abrasive planing; 50–80 heavy removal; 100–150 surfacing/calibrating; 180–400 finishing/sealer (use the abrasive and machine manufacturer's sequence; avoid grit jumps that leave scratches the next step cannot remove) [A1; A9].
• Min stock thickness: ~3 mm (0.12 inch) on segmented machines (Bütfering SWT) [A7]. (Minimum and maximum stock size, removal per pass, belt speed, air pressure and simultaneous-workpiece rules are machine/configuration values; several per-model specs here are MEDIUM/UNVERIFIED. Do not infer that every model permits 1/8 inch per pass; an "abrasive planer" is a distinct heavy-removal configuration.)

The reviewer's corrections (13)

• BLOCKER — 1. Replace the obsolete and legally overbroad combustible-dust framework — Exact claim: “the governing standards: NFPA 664 … NFPA 652 … and NFPA 654”; “since September 2020 a formal Dust Hazard Analysis on file”; “indoor collectors capped at 5,000 CFM.”
• BLOCKER — 2. Remove the fixed St 1 / Kst classification — Exact claim: “wood dust is ST1, Kst 100–200.”
• BLOCKER — 3. Restore the operator-injury hazards to equal prominence — Exact claim: “its signature hazard is not what it does to the operator's hand — it is what it does to the building”; “the danger moved from the HAND to the BUILDING”; “its worst failure isn't a wound.”
• BLOCKER — 4. Add model-specific operating limits, PPE, and hazardous-energy controls — Exact claim: “Short / narrow / thin pieces below the infeed–outfeed roller spacing … can tip into the head”; the safety section supplies no exact model limits or shutdown/lockout rule.
• BLOCKER — 5. Make belt tension and pneumatic settings model-specific — Exact claim: “Belt tension standards: paper 45–55 psi, cloth 55–65 psi, polyester 65–85 psi.”
• SHOULD-FIX — 6. Correct the thickness-planer comparison — Exact claim: “the planer's infeed roller progressively reduces warp over passes; the wide-belt's lighter hold-down cannot.”
• SHOULD-FIX — 7. Stop defining the whole category as three-phase production equipment — Exact claim: “a floor-standing, three-phase, conveyor-fed production machine”; “15–75 HP”; “cannot be run safely without shop infrastructure”; “NOT a beginner-owned or hobbyist tool.”
• SHOULD-FIX — 8. Do not reduce every machine to “drum dimensions; platen finishes” — Exact claim: “Platen … Does not dimension”; “Drum dimensions; platen finishes”; a calibrating sander is “drum-head-only,” a finishing sander “platen-head-only,” and “most production machines are combination.”
• SHOULD-FIX — 9. Correct the dust-collection sizing language — Exact claim: “OSHA-class duct velocity (≈3,500–4,000 FPM) and NFPA-sized collection are required.”
• SHOULD-FIX — 10. Qualify tramp-metal and spark-control prescriptions — Exact claim: “metal detectors/magnets upstream are the control”; equipment with a spark history “should carry spark detection + suppression … per NFPA guidance.”
• SHOULD-FIX — 11. Narrow the “founding of the class” claim — Exact claim: “Timesavers Inc., founded 1946 … with 3M to create the first continuous duty sanding platform — the clearest dated anchor of the class”; “Founding of the class: Timesavers Inc., 1946.”
• MINOR — 12. Resolve the patent and source-status cleanup — Exact claim: US 4,369,601 is “referenced, unverified”; manufacturer pages are repeatedly labeled “manual.”
• MINOR — 13. Qualify remaining absolutes — Exact claim: “figured/reversing-grain stock with no tearout”; “almost no first-time maker owns one”; “never skip more than one grit.”