Bend Allowances

This program calculates sheet-metal bend allowances, helping fabricators determine the total flat-blank length required to produce a part with multiple bends. For each bend, it computes the setback (W) using the tangent of half the bend angle, and the bend allowance (C) using the empirical formula ((0.01743 × R) + (0.0078 × T)) × A, which accounts for both bend radius and material thickness. Up to 20 bends are supported via parallel DIM arrays for sides, angles, radians, setbacks, bend allowances, flat lengths, radii, and a spare K array. Results are tabulated showing each leg’s flat length, setback, bend allowance, angle, and radius, with all values rounded to three decimal places using the INT(x×1000+0.5)/1000 idiom.

---

Program Structure

The program is organized into three logical phases: data entry, calculation, and tabular output. Lines 40–120 initialize nine parallel arrays (each dimensioned to 20 elements) and collect global parameters (number of bends B and material thickness T). Lines 130–400 handle the per-bend input loop, with the first bend processed outside the loop (lines 130–230) and bends 2 through B handled in a FOR loop (lines 240–360). A final leg entry at lines 370–400 closes the geometry. Lines 410–500 produce the summary and tabular output.

Array Usage

Nine arrays are declared at line 40, all of size 20:

• S() — input leg lengths
• A() — bend angles in degrees
• X() — bend angles converted to radians
• W() — setback values
• C() — bend allowances
• I() — declared but unused (shadowed by the loop variable I)
• L() — computed flat leg lengths
• R() — bend radii
• K() — declared but never used

Note that I is both a numeric scalar loop variable and a declared array I(); on this platform these are distinct namespaces, so no collision occurs, but the array I() is entirely superfluous.

Engineering Formulas

Two standard sheet-metal formulas drive the calculations:

• Setback W(i): (R(I) + T) * TAN(X(I) / 2) — the tangent-based setback from the outside mold line to the bend tangent point.
• Bend allowance C(i): ((0.01743 * R(I)) + (7.8E-03 * T)) * A(I) — an empirical linear approximation. The coefficient 0.01743 is approximately π/180, scaling radius to arc length per degree; 7.8E-3 (0.0078) is a material-dependent neutral-axis factor applied to thickness.

The flat leg length L(I) is the nominal leg length minus the setbacks contributed by adjacent bends: for the first leg S(1) - W(1), for intermediate legs S(I) - W(I) - W(I-1), and for the last leg S(B+1) - W(B). The running total M accumulates all flat lengths and bend allowances to yield the required blank length.

Notable Techniques

• 180-degree guard: Line 150 clamps a 180° angle to 179° and line 280 to 179.9°, preventing a division-by-zero in TAN(X/2) when the half-angle reaches 90°. The slightly different clamp values (179 vs. 179.9) between the first-bend and loop paths is an inconsistency.
• Rounding idiom: INT(x * 1000 + 0.5) / 1000 at lines 430, 460–480 rounds results to three decimal places, a standard BASIC technique in the absence of a built-in rounding function.
• Degree-to-radian conversion: Performed explicitly via multiplication by 3.1415926 / 180 and stored in array X(), since no built-in degree-mode trigonometry is available.
• Single-bend shortcut: Line 230 branches directly to line 370 when B=1, skipping the FOR loop entirely.
• Literal π approximation: 3.1415926 is used rather than any built-in constant, which is typical for this platform.

Output Format

The results table at lines 440–500 uses comma-separated PRINT items to align output into the system’s fixed tab columns. Column headers are “LEG”, “LENGTH”, “S/B” (setback), “B/A” (bend allowance), “ANGLE”, and “RADIUS”. The loop runs from 1 to B+1, covering all legs including the final one; however, C(B+1) and A(B+1) and R(B+1) are never assigned for the last leg and will print as zero, which is geometrically correct since there is no bend after the last leg.

Bugs and Anomalies

• The unused arrays I(20) and K(20) consume memory without contributing to the program’s function.
• The 180° clamp differs between the first bend (clamped to 179) and subsequent bends (clamped to 179.9), producing slightly different numerical results for the edge case of a near-flat bend.
• Variable M is used as a scalar accumulator but is never initialized to zero explicitly; it relies on the interpreter’s default zero-initialization of numeric variables.
• Line 510 contains a REM SAVE "BEND ALLOW" LINE 1 comment, suggesting an intended auto-run save command that was commented out rather than executed.