helen@hbxinnuorollforming.com
Dovetail closed floor deck cold roll forming production line equipment
36 stations with cam-actuated mandrel rollers for throat lock forming
Geometric concrete interlock — no reliance on surface embossment bond
Twin 15KW drive, 450H frame, 2.5-inch chain, hydraulic floor shear
This dovetail deck roll former produces closed-profile steel floor decking panels where each rib narrows at the opening — creating the distinctive dovetail geometry that permanently locks the cured concrete slab into the steel deck. Unlike open trapezoidal decks that depend on surface dimples for composite bond, the dovetail shape itself forms a mechanical wedge: concrete fills the wide rib base, cures, and cannot be extracted through the narrowed throat. This is structural interlock at the geometric level — immune to embossment wear, unaffected by minor surface corrosion, and proven under cyclic seismic loading where dimple-based bond can progressively degrade. The machine achieves this complex forming geometry through a specialized 36-station progressive roller sequence, with cam-actuated internal mandrel rollers that support the rib cavity during the critical wall-inversion forming phase.
| Principle | Traditional Dovetail Joint (Wood) | Dovetail Deck (Steel-Concrete) | Engineering Significance |
|---|---|---|---|
| Geometry | Trapezoidal pin fits into matching socket. Pin is wider at the end than at the base — cannot be pulled straight out once assembled. | Concrete rib is wider at the base (inside the deck cavity) than at the throat (deck surface). Cured concrete is mechanically trapped — cannot lift out of the deck. | The interlock does not depend on adhesion, friction, or surface texture. It is a geometric constraint — as fundamental as a key in a lock. This is why dovetail deck maintains composite action even under conditions that degrade other bond mechanisms. |
| Load Transfer | Pulling force on the joint is transferred as bearing pressure between the angled pin faces and socket walls. The wood fibers carry compression — the dovetail geometry converts tension into bearing. | Longitudinal shear force between concrete slab and steel deck is resisted by bearing of the concrete wedge against the inclined dovetail rib walls. The concrete is in compression at the interface — its strongest loading mode. | Both systems exploit the same physical principle: convert tension/shear into compression bearing through inclined surfaces. Concrete's compressive strength (30-50 MPa) far exceeds its tensile bond strength (1-2 MPa) — the dovetail geometry ensures the load path uses the strong material property. |
| Failure Mode | Wood joint fails by: (a) pin shearing off at the base (insufficient pin thickness), or (b) socket wall splitting (insufficient material around socket). Both are geometric design failures — not adhesive failures. | Dovetail deck fails by: (a) concrete crushing at the rib wall interface (insufficient rib depth or concrete strength), or (b) deck rib wall buckling outward (insufficient steel gauge). Both are capacity failures — not bond failures. | Failure modes are predictable and calculable using standard material properties. Unlike dimple-based bond (where m+k values are empirical from testing), dovetail interlock capacity can be estimated from first principles — geometry × concrete strength × steel gauge. |
| Cyclic Loading | Furniture drawer opened/closed thousands of times. Dovetail joint stays tight because the geometry does not wear — there is no adhesive to fatigue. Slight wood shrinkage actually tightens the joint. | Building experiences thousands of load cycles (occupancy, wind, seismic). Dovetail interlock does not degrade because there is no bond to break. Concrete creep over time actually increases the bearing preload at the interface. | This is the critical seismic advantage: under earthquake loading, the deck-concrete interface undergoes repeated slip reversals. Dimple embossments elongate plastically with each cycle — progressively losing bond. Dovetail geometry cannot elongate — the concrete wedge simply bears harder against the rib wall. |
| Design Parameter | 820 Dovetail, 0.9mm | 820 Dovetail, 1.2mm | 750 Dovetail, 1.0mm | How Value Is Determined |
|---|---|---|---|---|
| Longitudinal Shear Resistance (τu,Rk) | 0.35-0.45 N/mm² | 0.50-0.60 N/mm² | 0.30-0.40 N/mm² | Full-scale 4-point bending test per EN 1994-1-1 Annex B, minimum 3 specimens per configuration |
| m-value (Mechanical Interlock) | 220-280 N/mm² | 280-340 N/mm² | 200-260 N/mm² | Slope of linear regression from shear bond test data. Higher m = stronger interlock. Dovetail m-values are 30-50% higher than equivalent open deck. |
| k-value (Friction Component) | 0.02-0.05 N/mm² | 0.03-0.06 N/mm² | 0.02-0.04 N/mm² | Intercept of regression line. Dovetail k is lower than open deck because friction is secondary — geometry dominates. |
| Maximum Unpropped Span (130mm slab) | 3.8m single span | 4.5m single span | 3.5m single span | Calculated per code from section properties × construction stage load limits. Verified with deflection measurement during testing. |
| Composite Stage Capacity (3.5m span) | 4.5 kN/m² | 6.0 kN/m² | 4.0 kN/m² | Design value after applying partial safety factors (γVS = 1.25 for shear bond). Sufficient for office (3.0 kN/m²) and retail (5.0 kN/m²) occupancy. |
| Fire Rating Achievable | R90 with 130mm slab | R120 with 130mm slab | R90 with 120mm slab | Fire test per ISO 834. Dovetail geometry restricts hot gas flow in rib cavity — achieves rating with thinner slab than open deck equivalent. |
| Diaphragm Shear Capacity | 12-18 kN/m (side-lap screwed at 450mm centers per SDI DDM). Dovetail deck diaphragm values are comparable to open deck — the difference is bond performance, not in-plane shear. | Panel assembly test per SDI DDM or AISI S310. Dependent on fastener type and spacing, not profile geometry. | ||
| Station Group | Rows | Roller Function | Material State Entering | Material State Exiting |
|---|---|---|---|---|
| Entry & Leveling | Pre-form guide | Side guides center strip; 5-roller leveler removes coil curvature. Strip enters first forming station flat and centered. | Curved coil strip with residual set from coiling | Flat, centered strip, zero camber, zero cross-bow |
| Dimple Embossment | 1 | Clutch-controlled dimple wheel stamps bond pattern into what will become the internal rib surface. Dimples are secondary bond mechanism — backup to geometric interlock. | Flat strip, full width | Strip with embossed pattern at rib locations; remainder of strip flat |
| Initial Rib Forming | 2-8 | Progressive edge-forming rollers begin lifting the rib from the flat strip. Outer edges of each rib are bent upward in small increments — typically 5-8° per station. | Embossed flat strip | Strip with ribs partially formed — rib walls angled outward at approximately 60-70° from horizontal |
| Rib Deepening | 9-16 | Rib walls continue bending upward toward vertical. Rib base is established. Internal rib cavity volume is created. This is the standard trapezoidal forming phase — identical in principle to open deck forming. | Partial ribs at 60-70° wall angle | Fully deepened ribs at 90° wall angle (vertical walls). Rib cavity is at its maximum width — the trapezoidal pre-form is complete |
| Wall Inversion Initiation | 17-20 | Critical phase begins. Upper rollers begin pressing rib walls inward past vertical. Each station tilts the wall 3-5° inward. Internal mandrel rollers deploy to support the rib cavity against collapse. | Vertical rib walls, open cavity | Rib walls leaning inward at 5-15° from vertical. Mandrel rollers are supporting the cavity from inside |
| Throat Narrowing | 21-26 | Wall inversion accelerates. Each station pushes walls further inward — wall angle now 15-40° from vertical. Rib throat opening progressively narrows. Mandrel rollers begin to compress as available internal space reduces. | Walls leaning 5-15° inward | Walls leaning 25-40° inward. Throat opening is 60-75% of the original rib width. Dovetail shape is now visible |
| Throat Lock | 27-30 | Most technically demanding stations. Walls are folded past 45° inward — throat opening approaches final dimension. Over-bend stations push walls slightly past target angle. Mandrel rollers are nearly fully compressed — they must withdraw cleanly as the roller set opens. | Walls leaning 25-40° inward | Walls at target dovetail angle — throat width at specification. Mandrel compression synchronized with roller set opening cycle |
| Sizing & Springback Compensation | 31-34 | Over-bend at stations 31-32: walls are pushed 2-3° past target angle. Release at 33-34: walls spring back to exactly the target geometry. Each of these 4 stations is independently micrometer-adjustable. | Walls at target angle (but will spring back when released) | Walls spring back to final specification. Throat width, rib depth, and wall angle all within tolerance |
| Straightening & Final Pass | 35-36 | Straightening roller set corrects any panel camber or twist accumulated through 34 stations of asymmetric forming. Final pass roller verifies profile geometry — effectively a go/no-go gauge roller. | Fully formed dovetail profile | Straight, dimensionally verified dovetail panel, ready for hydraulic cutoff |
| Aspect | Detail |
|---|---|
| Function | The mandrel is a spring-loaded roller assembly that expands to fill the rib cavity during forming and compresses to withdraw as the forming roller set opens. It prevents the rib walls from collapsing inward during the wall inversion phase — the rib cavity must remain open for concrete to flow in, but the forming pressure pushing walls inward would collapse the cavity without internal support. |
| Location | Installed at stations 17-30 (wall inversion + throat narrowing + throat lock). Not needed at stations 1-16 (trapezoidal pre-form — ribs are open, no internal support required). Not needed at stations 31-36 (sizing — walls are at final geometry, forming pressure is light). |
| Actuation | Cam-actuated, mechanically timed to the roller set position. A cam lobe on the upper roller shaft pushes the mandrel assembly to expand as the roller closes. As the roller opens, a return spring compresses the mandrel. This is purely mechanical — no hydraulics, no pneumatics, no electronics. Mechanical timing cannot drift due to sensor failure. |
| Mandrel Roller Material | Cr12MoV tool steel, quenched to HRC 58-60, polished to Ra 0.2µm surface finish. Harder than the forming rollers (HRC 48-52) because the mandrel contacts the inside surface of the rib — any surface defect on the mandrel marks the panel interior. |
| Spring Specification | Die spring (chrome silicon steel), rated for 500,000+ compression cycles. Spring force is calibrated to provide sufficient internal support pressure without excessive force that would mark the panel. Correct spring force = panel gauge × rib wall area × 2-3 MPa. |
| Adjustment Points | Three adjustments per mandrel station: (1) spring pre-load — controls internal support pressure; (2) cam follower gap — controls the timing point where mandrel begins expanding; (3) mandrel roller height — controls the vertical position of the internal support within the rib cavity. |
| Common Failure Modes | (a) Spring fatigue — reduced support pressure, rib walls begin bulging inward. Replace spring set annually. (b) Cam follower wear — changes mandrel timing, causing panel edge marking during roller opening. Inspect follower rollers monthly. (c) Mandrel roller bearing seizure — roller stops rotating, scratches panel interior surface continuously. Replace bearings at first sign of roughness during manual rotation check. |
| Maintenance Interval | Daily: visual inspection of mandrel surfaces for scoring or embedded debris. Weekly: manual rotation check of each mandrel roller bearing (should spin freely, no roughness). Monthly: measure spring free length (replace if shortened >3% from new). Annually: full mandrel assembly removal, disassembly, cleaning, bearing replacement, spring replacement, re-installation with timing verification. |
| Region / Standard | Dovetail Deck Reference | Design Basis | Testing Requirement |
|---|---|---|---|
| Europe — EN 1994-1-1 (Eurocode 4) | Recognized composite deck type. Design by m+k method (Annex B) or partial connection method (§9.7). Dovetail profiles classified as "re-entrant" in code terminology. | Full-scale bending tests to establish m+k values. Minimum 3 tests per deck type, gauge, and concrete grade combination. | EN 1994-1-1 Annex B test protocol. Additional requirements in EN 1990 for statistical determination of characteristic values. |
| North America — ANSI/SDI-C / IBC | Recognized as "composite steel floor deck — dovetail profile" per SDI. Load tables published per SDI-DDM methodology. Accepted under IBC Chapter 22 (Steel). | SDI-DDM calculation methodology. Composite slab design per AISC 360 Chapter I or ACI 318 composite provisions. | Full-scale testing per SDI test standard. Load tables must be verified by independent registered engineer. ICC-ES evaluation report optional but recommended. |
| Australia / New Zealand — AS 2327.1 | Recognized as "re-entrant profile" composite deck. Design per AS 2327.1 Section 6 — Composite Slabs. Specific provisions for dovetail geometry in Clause 6.4. | m+k method identical to Eurocode approach. Australian amendment requires additional cyclic testing for structures in Seismic Category 3-4. | AS 2327.1 test protocol — similar to EN 1994-1-1 Annex B but with Australian loading protocols. NATA-accredited laboratory required. |
| Middle East — Various (typically Eurocode-based) | Accepted under project-specific performance specification. UAE Fire & Life Safety Code references composite deck fire rating requirements. | Eurocode-based design with local modification factors. Dubai Municipality, Abu Dhabi UPC, and Qatar QCS each have submission requirements. | Fire test to ISO 834 with local authority witness (typically Dubai Civil Defence or equivalent). Material certificate traceability required. |
| Southeast Asia — Various National Codes | Increasing adoption — project-specific acceptance for high-rise commercial (Philippines, Thailand, Vietnam). National codes transitioning from prescriptive to performance-based. | Typically referenced to Eurocode, AISC, or BS 5950 depending on country's historical code lineage. Local engineering consultant familiar with international standards required. | Full-scale testing to recognized international standard. Local university laboratory often used for credibility with building authority. Engineer's certification of compliance. |
| Action | Dovetail 750 to Dovetail 820 | Dovetail 820 to Dovetail 688 | Time Estimate |
|---|---|---|---|
| Roller Change Scope | Complete Phase 2-4 roller set change (stations 17-36). Phase 1 rollers (stations 1-16) may be re-usable if rib base width is similar — verify against profile drawings. | Full roller set change (stations 1-36) — rib configuration changes from 3 ribs to 2 ribs, fundamentally different roller geometry throughout. | 750→820: 3-4 hours. 820→688: 5-6 hours |
| Mandrel Roller Set | Change mandrel rollers at stations 17-30. Dovetail angle and throat width differ — mandrel geometry must match the target profile exactly. | Complete mandrel set change. 688 is 2-rib vs. 820 3-rib — mandrel count and spacing are different. | Mandrel change alone: 1-1.5 hours |
| Shear Blade | Blade width remains same if coil width unchanged (1250mm for both). Verify blade edge condition — changeover is a good time for re-grind if nearing 500K cuts. | Blade width may change if coil width differs (1000mm for 688 vs. 1250mm for 820). Wider blade can cut narrower panel — narrower blade will leave uncut edge on wider panel. | Blade change if needed: 30-45 minutes |
| Entry Guide | Adjust side guide width from 1250mm to 1250mm — no change needed if both use same coil width. | Adjust side guide width from 1250mm to 1000mm. Manual adjustment with scale — 5 minutes. | Guide adjustment: 10-15 minutes |
| PLC Parameters | Update: profile name, target throat width (for gauge roller feedback), encoder calibration (if coil width changed, recalculate pulses per meter). Save as new recipe. | Same process — different target values in recipe. | PLC update: 10-15 minutes |
| Trial & Verification | Mandrel timing verification (slow-speed run through stations 17-30). Full-speed panel sample: measure throat width, rib depth, wall angle at both ends and mid-panel. | Full roller set verification at slow speed, then mandrel timing check, then full-speed trial with dimensional inspection. | Trial & verification: 1-1.5 hours |
| Measurement | Instrument | Tolerance | Sample Frequency | Action If Out of Tolerance |
|---|---|---|---|---|
| Throat Width | Custom go/no-go gauge plate machined to throat width ±0.5mm. Slide gauge through rib throat — must pass without binding, must not have >0.5mm side clearance. | Nominal ±1.0mm | Every 50th panel, both ends | Throat too wide: increase over-bend at stations 31-32. Throat too narrow: reduce over-bend. Re-check next 5 panels. |
| Rib Depth | Digital depth micrometer, 0.01mm resolution. Measure from panel crest (top of rib) to panel soffit (underside between ribs). | Nominal ±1.0mm | Every 50th panel, all ribs | Depth undersize: forming not reaching full depth — check roller gap at deep-form stations (9-16). Depth oversize: over-forming — reduce gap. |
| Dovetail Wall Angle | Digital angle gauge with magnetic base. Place on rib wall surface, measure angle from vertical. For 820 dovetail: target 20-25° from vertical inward. | ±1.5° from specification | Every 50th panel, inner and outer wall of one rib | Angle too shallow: increase over-bend. Angle too steep (wall too vertical): insufficient over-bend — dovetail interlock will be weak. |
| Rib Spacing (Center-to-Center) | Vernier caliper, 0.02mm resolution. Measure center of one rib throat to center of adjacent rib throat. | Nominal ±1.5mm | Every 100th panel | Spacing error indicates uneven forming across width — check roller alignment at Phase 1 stations (2-8) where rib positions are established. |
| Panel Camber (Straightness) | String line along panel edge, 3m length. Measure maximum deviation from straight line with steel rule. | ≤2mm per 3m length | Every shift start + every coil change | Camber indicates uneven roller pressure left vs. right — re-check roller gap symmetry across machine width. Check frame level. |
| End Flare | Place panel on flat surface plate. Measure height of any corner lift-off with feeler gauge. | ≤2mm at any corner | Every shift start | End flare at shear indicates blade gap too large or blade dull — adjust shear blade gap or re-grind/replace blade set. |
| Panel Surface Condition | Visual inspection under 500 lux lighting. Look for: longitudinal scratches (roller damage), transverse marks (mandrel timing error), zinc coating damage (bare steel visible). | No scratches >0.1mm deep. No bare steel exposure. No mandrel timing marks. | Continuous by operator during production | Scratches: identify damaged roller, polish or replace. Timing marks: check mandrel cam follower and spring. Zinc damage: check roller surface for embedded debris. |
| Inspection Point | Method | Acceptance Criteria | Why It Matters to Your Customer Relationship |
|---|---|---|---|
| Profile Verification | Compare delivered panel sample against approved shop drawing. Measure all key dimensions: throat width, rib depth, rib spacing, dovetail angle. | All dimensions within published tolerances on the data sheet. Any single dimension outside tolerance = batch quarantine pending investigation. | This is the moment of truth. If the first delivery passes dimensional inspection, the engineer trusts your quality. If it fails, every subsequent delivery will be scrutinized — regardless of your machine's capabilities. |
| Gauge Verification | Micrometer measurement at 5 points across panel width. Compare against specified nominal gauge and tolerance. | Measured thickness within ±0.05mm of nominal. No point thinner than nominal minus 0.05mm. | Undergauge steel = reduced structural capacity. If the engineer finds undergauge panels on site, they may require load testing of installed slabs — at your expense. |
| Zinc Coating Verification | Coating thickness gauge (magnetic induction type). Measure at 5 points: panel crest, both rib walls, both inter-rib flats. Compare against mill certificate. | Coating mass ≥ specified minimum (typically Z275 / G90). No single point below 90% of minimum. | Zinc coating is the panel's corrosion protection. Under-spec coating = premature rust = warranty claim. The cost of replacing rusted deck in a completed building dwarfs the panel supply value. |
| Embossment Pattern | Visual comparison against approved reference sample. Check dimple depth with dial indicator at 3 points per rib. | Dimple depth within ±0.2mm of reference sample. Pattern consistent — no missing or partial dimples. | While dovetail interlock is the primary bond mechanism, embossments are the backup. Missing dimples on dovetail deck are less critical than on open deck but still indicate machine maintenance issues that need attention. |
| Panel Straightness | Lay panel on flat surface. Measure gap between panel edge and straight edge reference line. | Camber ≤2mm per 3m. Twist: panel lies flat — no corner lifted more than 2mm from surface. | Bowed or twisted panels don't sit flat on the supporting beams. The installer compensates with extra fasteners, concrete leaks through gaps, and the floor finish may show unevenness. All of these become your problem when the contractor complains. |
| Bundle Condition | Check for transport damage: crushed bundle edges, broken strapping, water ingress through torn wrapping, panel corrosion. | No panels with visible deformation. No wet bundles. No broken strapping that allowed panels to shift during transport. | Transport damage is your problem even if the carrier caused it — you're the supplier. Proper bundling and wrapping is part of your quality system. Damaged panels on site = replacement panels shipped at your cost + delay claims from contractor. |
Hebei Xinnuo is a professional cold roll forming machinery manufacturer with 20+ years of experience. We specialize in C/Z purlin machines, roof panel roll formers, floor deck machines, and custom profile solutions — exported to 50+ countries worldwide.
