

Flower pattern and roll pass design is the engineering foundation of every cold roll forming operation. The flower pattern — a sequential set of cross-sectional diagrams showing the strip shape at each forming station — serves as the master blueprint from which all roller contours are derived. Roll pass design encompasses the complete process of determining station count, bend angle progression, roller geometry, and springback compensation that transforms flat coil into a finished profile.
A well-executed flower pattern directly determines dimensional accuracy, surface quality, and production stability. Conversely, errors in roll pass design lead to edge waves, longitudinal bow, twist, and excessive springback — defects that no machine adjustment can fully correct. For this reason, roll pass design is considered the most critical engineering phase in any roll forming project, preceding machine manufacturing and tooling fabrication.
The flower pattern is a graphical representation of the strip's cross-section at each forming station, from the flat strip at station 0 to the final profile at the last station. The name derives from the way the strip progressively opens and bends across successive stations, visually resembling petals radiating outward from a center. Each layer in the flower diagram corresponds to one forming stand and shows the bend angle, bend radius, and strip position at that stage.
The flower pattern serves three purposes: it defines the bend sequence from flat to final shape, it determines the total number of forming stations required, and it provides the geometric basis from which individual roller contours are machined. A typical roll forming line requires 12 to 30 stations depending on profile complexity, material thickness, and tolerance requirements.
For symmetric profiles such as C-channels or U-channels with equal flanges, the flower pattern is relatively straightforward: bends are applied symmetrically and simultaneously on both sides. For asymmetric profiles such as Z-purlins or hat channels with unequal legs, the design complexity increases substantially because bends on each side must be sequenced independently to prevent twist and lateral drift.
The roll pass design process follows a structured sequence of engineering steps, from profile definition through to final tool validation:
The process begins with a complete drawing of the finished profile, specifying overall width and height, flange lengths, inside bend radii (typically 1.5 to 3 times material thickness), material thickness, and dimensional tolerances. For asymmetric profiles, left and right flange dimensions are specified separately.
The flat strip width is calculated by summing all straight sections and the arc lengths of bends. The formula uses the neutral axis factor k (typically 0.33 to 0.50 for cold forming), which accounts for the shift of the neutral plane during bending:
Developed width = Σ(straight lengths) + π × bend angle (rad) × (inside radius + k × thickness)
The Oehler method provides the most precise calculation, accounting for the relationship between sheet thickness, bending radius, and angle. For radii of 1.5 times thickness, a simplified approximation of 0.5 × thickness per 90° bend is commonly used.
Station count is determined by the total bending angle and the maximum permissible bend per station. A common rule allocates one station per 10 to 15 degrees of total bend per bend line. For a profile with four 90° bends (total 360°), approximately 18 to 24 stations are required if each station contributes 15 to 20 degrees per bend line. Asymmetric profiles generally require additional stations because bends cannot be applied simultaneously on both sides.
With station count and bend angles established, the flower pattern is drawn. This step defines the exact bend angle at each station, the bend radius progression, and the strip centerline position. The designer selects a bending method (constant length or constant radius) and verifies that longitudinal edge strain remains below the elastic limit of the material at every intermediate stage.
Roller contours are derived directly from the profile cross-section at each station. Each station receives a pair of upper and lower rollers whose mating surfaces match the strip geometry at that stage. Side rolls may be added between stations to guide vertical flanges and reduce surface scuffing. Rollers are typically machined from D2, A2, or Cr12MoV tool steel, hardened to HRC 58–62, and ground to tolerances of ±0.01 mm.
Before manufacturing, the complete roll design is validated through Finite Element Analysis. FEA simulation predicts longitudinal strain distribution, springback behavior, and potential defect zones, allowing the designer to optimize bend angles and station count before physical tooling is cut.
Two principal methods govern how bend angles are distributed across forming stations. The choice between them affects springback behavior, longitudinal strain, and final profile quality.
| Method | Principle | Springback Behavior | Best Suited For |
|---|---|---|---|
| Constant Developed Length (Constant Arc Length) | The arc length at each bend remains constant across stations; the bend radius decreases progressively as the angle increases | Springback does not change significantly with increasing pass number, but overall amplitude is greater | Profiles requiring tight radius control; high-strength steel where radius consistency matters |
| Constant Radius | The bend radius remains fixed across stations; the arc length increases progressively as the angle increases | Springback increases with pass number; overall amplitude is lower than constant length method | Profiles where minimizing longitudinal bow is a priority; mild steel applications |
Research published in the Chinese Journal of Mechanical Engineering compared the two methods using experimental testing and finite element simulation. The study found that the difference in fillet radius and arc length between the two design approaches is the primary factor affecting stress-strain distribution and springback. In both methods, springback increases with larger fillet radii and decreases with greater sheet thickness.
Several calculation methods exist for determining the number of forming stations. The most widely used approaches are summarized below:
| Method | Formula / Rule | Key Parameters |
|---|---|---|
| Forming Length Rule | F = 40H (standard); F = 70H to 100H (high-tensile material) | F = forming length (mm); H = flange height (mm). Number of stations = F ÷ station spacing + 1 final fixing station |
| Bend Angle Rule | Stations = total bend angle ÷ max bend per station | Max bend per station: 20–30° for mild steel; 10–15° for high-strength steel |
| Halmos Equation | n = 2a ÷ (2d × tan α1) | a = bending edge length; d = roll distance; α1 = estimated angle (1°–1.5°) |
| Shape Factor Method | φ = F × n1 ÷ t | F = sum of vertical edge lengths; n1 = total bending angle; t = thickness. Pass number derived from data charts |
The forming length rule, documented by the Australian Steel Institute, provides a practical starting point. For a simple channel with a 50 mm flange formed through stations at 400 mm intervals: F = 40 × 50 = 2,000 mm, yielding 5 station spacings (6 stations) plus one final fixing station, for a total of 7 stations. For higher-tensile materials used in roofing and walling products, designers modify the factor to 70H or 100H to accommodate reduced ductility.
Typical station counts by profile type:
| Profile Type | Stations | Complexity |
|---|---|---|
| Simple C-channel or U-channel | 12–16 | Symmetric, single bend line per side |
| Standard sigma channel | 18–22 | Symmetric, multiple bend lines |
| Complex asymmetric profile with lips | 24–30 | Asymmetric, multiple bend lines, return flanges |
When metal is bent, the outer fibers stretch and inner fibers compress. After the bending force is removed, elastic recovery causes the material to partially return toward its original shape — a phenomenon known as springback. Roll pass design must compensate for springback by over-bending each station slightly beyond the theoretical angle.
| Parameter | Typical Value | Influencing Factors |
|---|---|---|
| Compensation angle per bend | 0.5°–2° (mild steel) | Increases with material yield strength and bend radius |
| Compensation angle (high-strength steel) | 2°–5° | HSLA and advanced high-strength steels exhibit greater elastic recovery |
| Neutral axis factor (k) | 0.33–0.50 | Lower k for tighter radii; higher k for larger radii relative to thickness |
| Inside bend radius | 1.5× to 3× thickness | Larger radii increase springback; smaller radii risk edge cracking |
Springback calculation depends on material properties (yield strength, elastic modulus), sheet thickness, and the bending radius and angle. Software tools such as UBECO PROFIL compute springback automatically based on the material database, allowing the designer to apply compensation angles directly to each station's roller geometry. The constant radius method produces springback that increases with pass number, while the constant arc length method yields springback that remains relatively stable across stations but with a larger overall magnitude.
Modern roll pass design relies on specialized software that integrates profile definition, flower pattern generation, roll design, and FEA validation into a single workflow. Two industry-leading platforms dominate the market:
| Software | Developer | Core Modules | Key Capabilities |
|---|---|---|---|
| UBECO PROFIL | UBECO GmbH (Germany) | Profile Design → Flower Pattern → Roll Design → PSA → FEA (VRM) | Oehler developed length calculation; constant length / constant radius / track holding methods; band edge stress monitoring; NC code output (DIN 66025); CAD interfaces (DXF, IGES, ActiveX for AutoCAD/SolidWorks); 600+ installations in 50+ countries since 1986 |
| COPRA RF | data M Sheet Metal Solutions (Germany) | Sections → Flower Technology → DTM → Roll Design (SmartRolls) → FEA RF | Parametric flower design via SpreadSheet; DTM longitudinal strain analysis using thin shell theory; SmartRolls automated roll contour generation; COPRA FEA RF nonlinear elastoplastic simulation; RollScanner and ProfileScan for quality inspection |
Both platforms follow a three-step quality management concept: geometric stress calculation (band edge stress or DTM), profile stress analysis (PSA), and full FEA simulation. This progressive validation approach allows designers to identify and resolve potential issues at each stage, from initial flower pattern through to final roll geometry, before any physical tooling is manufactured.
Defects in roll formed profiles often originate from suboptimal flower pattern design. The most common defects, their causes, and prevention strategies are summarized below:
| Defect | Design-Related Cause | Prevention Strategy |
|---|---|---|
| Edge wave (wavy edges) | Excessive edge strain from too few stations or aggressive bend angles at edge bend lines | Increase station count for edge bends; reduce bend angle per station; verify edge strain < elastic limit using band edge stress calculation |
| Longitudinal bow (center bow) | Uneven longitudinal strain distribution between top and bottom surfaces | Select appropriate bending method (constant radius reduces bow); add straightening device after final station; adjust forming line height |
| Twist | Asymmetric bend sequence causing unbalanced lateral forces | Sequence bends to balance forces; add side rolls between stations; verify with FEA simulation |
| Excessive springback | Insufficient compensation angle; large bend radius relative to thickness | Apply material-specific overbend compensation (0.5°–5°); reduce inside radius; use constant length method for stable springback |
| Surface marking (scuffing) | Speed differential between roll surfaces and strip; driven rolls on vertical flanges | Idle flange rolls for legs >70°; use side rolls on vertical axis; apply hard chrome plating (0.05 mm) to roll surfaces |
| End flare | Residual stress release at cut ends; insufficient final station fixing | Add dedicated finishing station; apply coining or bead pass at final stage; use straightener block (aluminum bronze or nylon) close to last pass |
A straightening device is recommended after the final forming station to remove residual twist, camber, and bow. Block-type straighteners made from aluminum bronze or nylon provide approximately 25 to 38 mm of working contact area with 0.38 mm per side clearance around the profile. The straightener should be positioned between the last two passes for pre-cut applications, or immediately after the final station for post-cut lines.