Bow & Twist Testing
Quality & Inspection Standards
PRODUCTS CENTER
Get Custom Quote
PRODUCTS CENTER
Bow & Twist Testing – PCB Warpage Measurement per IPC‑TM‑650 2.4.22
Bow and twist (warpage) is a critical dimensional attribute of printed circuit boards that affects SMT placement accuracy, solder joint integrity, and assembly yield. Excessive warpage leads to tombstoning, head‑in‑pillow defects, and poor coplanarity for BGA and QFN components. This engineering framework from UltroNiu describes the standardized measurement of bow and twist using IPC‑TM‑650 2.4.22 (Bow and Twist). The method applies to rigid printed boards, flex boards, and laminates. IPC‑TM‑650 provides only the test procedure; acceptance criteria (e.g., maximum warpage percentage for SMT or through‑hole assembly) must be defined by users based on product requirements, typically per IPC‑6012 or customer specifications.
Table of Contents
- Scope – IPC‑TM‑650 2.4.22 Bow & Twist
- Definitions – Bow vs Twist vs Warpage
- Failure Modes Caused by Excessive Bow & Twist
- Test Equipment & Setup per 2.4.22
- Measurement Procedure (IPC‑TM‑650 2.4.22)
- Bow & Twist Calculation Formulas
- Typical Acceptance Criteria (Not in IPC‑TM‑650)
- Sampling Plan per IPC‑TM‑650
- FAQ – Bow & Twist Testing
- Bow & Twist Test Checklist
1. Scope – IPC‑TM‑650 2.4.22 Bow & Twist
IPC‑TM‑650 2.4.22 (Bow and Twist) is the standard test method for measuring the flatness deviation of printed boards, laminates, and finished assemblies. The method applies to:
- Unpopulated rigid printed boards (bare boards)
- Copper‑clad laminates
- Flexible printed circuits (with support)
- Assembled boards (with caution for components)
IPC‑TM‑650 does not specify:
- Maximum allowable bow or twist percentage
- Pass/fail criteria for specific board types or assembly processes
Users must define acceptance criteria based on IPC‑6012 (for bare boards) or customer SMT assembly requirements.
2. Definitions – Bow vs Twist vs Warpage (per IPC‑TM‑650 2.4.22)
- Bow: A cylindrical or spherical curvature of the board. The board curves in a single direction, like a cylinder segment. Measured as the maximum vertical deviation from a flat plane, divided by the diagonal length of the board.
- Twist: A deformation where opposite corners are not in the same plane, resembling a saddle shape. Measured as the difference in height between opposite corners, divided by the diagonal length.
- Warpage (general): Any deviation from flatness; often used interchangeably with bow, but technically bow + twist.
The method distinguishes bow (uniform curvature) from twist (corner lifting). For most PCBs, both are measured, and the larger of the two determines acceptability.
3. Failure Modes Caused by Excessive Bow & Twist
| Assembly Issue | Cause | Consequence |
|---|---|---|
| Tombstoning | Component lifted on one side due to warpage during reflow | Open circuit, assembly rejection |
| Head‑in‑pillow (BGA) | BGA ball does not contact pad due to warpage | Intermittent electrical connection, field failure |
| Solder voiding | Non‑coplanar PCB prevents uniform flux flow | Voids in BGA or QFN joints |
| Poor stencil printing | Board not flat, paste smeared or missing | Insufficient solder, bridging |
| Pick‑and‑place errors | Vacuum nozzle cannot pick due to warpage | Component placement failure |
4. Test Equipment & Setup per IPC‑TM‑650 2.4.22
The following equipment is required to perform the test:
- Flat granite surface plate: Grade A or better, with flatness ≤0.001" per foot.
- Dial indicator or height gauge: With resolution of 0.001" (0.025 mm) or better. Digital height gauge recommended.
- Support pins or blocks: To hold the board without imposing stress. Typically three pins at corners (leaving one corner free) or a fixture that supports the board along edges.
- Indicator stand: To position the dial indicator over measurement points.
The board shall be measured in a stress‑free condition (no external clamps or weights). Measure at controlled temperature (23±2°C) and humidity (typically 50% RH, though humidity effect on warpage is not corrected in this method).
5. Measurement Procedure (IPC‑TM‑650 2.4.22)
- Condition the board: Boards should be stored at room temperature for at least 24 hours before measurement. No baking is required unless specified.
- Mount the board: Place the board on three support points located near the corners. The supports should be at the same height (coplanar). The fourth corner remains unsupported. The board should be positioned such that the measurement points (corners) are accessible.
- Measure corner heights: Using the dial indicator, measure the vertical height of each corner (or point near each corner, approximately 10 mm from the edge). Record the values as A (highest corner), B, C, and D.
- Measure center (optional): For bow measurement, also measure the height at the center of the board.
- Calculate twist: Twist is the difference in height between two opposite corners, minus the other two corners. The method uses a specific formula to isolate twist from bow (see Section 6).
- Measure bow (if needed): For boards with primarily cylindrical curvature, place the board concave side down on the flat surface. Measure the maximum gap between the board and the surface at the center.
The board is not to be clamped or held down; it must rest freely on the supports. The test is non‑destructive.
6. Bow & Twist Calculation Formulas
From the four corner height measurements (A, B, C, D in inches or mm), the twist (T) is calculated as:
T = |(A + C) − (B + D)| / (2 × diagonal length)Where diagonal length is the distance between opposite corners of the board (usually measured as the hypotenuse of length and width).
Bow (B) is calculated as:
B = (Maximum gap between board and flat surface when board is placed concave side down) / (Diagonal length)Both bow and twist are expressed as a percentage (e.g., 0.75%). For many specifications, the larger of the two is used for acceptance.
Example: A board with diagonal 200 mm, twist measurement difference (A+C)‑(B+D) = 0.8 mm. Then twist = 0.8/(2×200) = 0.002 = 0.2%.
7. Typical Acceptance Criteria (Not Specified by IPC‑TM‑650)
IPC‑TM‑650 does not define pass/fail limits. The following are common limits from IPC‑6012 (bare board qualification) or SMT assembly guidelines. Users must adopt limits appropriate for their product.
| Board Type | Maximum Bow & Twist | Reference (Not TM‑650) |
|---|---|---|
| Surface mount boards (SMT) | ≤0.75% of diagonal | IPC‑6012 Class 2/3 |
| Through‑hole only boards | ≤1.5% of diagonal | IPC‑6012 Class 2 |
| Fine‑pitch SMT (≤0.5 mm pitch) | ≤0.5% of diagonal | Industry guideline |
| Large BGAs (>35 mm) | ≤0.3% of diagonal | Component manufacturer spec |
Note: These limits are examples. Users must set acceptance criteria based on their SMT process capability and product reliability requirements.
8. Sampling Plan per IPC‑TM‑650
While IPC‑TM‑650 does not mandate sample size, common practice for incoming inspection or qualification is:
- Incoming bare board inspection (Class 2/3): Measure 5 panels per lot (or 1 panel per 50 boards, whichever is larger). Record bow and twist values.
- Process qualification (laminate or PCB supplier): Measure 10 boards per panel (or 5 panels) to establish baseline.
- Post‑reflow warpage verification: Measure 5 boards after simulated reflow (or after actual reflow) to determine reflow‑induced warpage change.
- For high‑reliability (automotive/aerospace): 100% measurement may be required for critical boards (e.g., BGA substrates).
9. FAQ – Bow & Twist Testing per IPC‑TM‑650 2.4.22
Q: Does IPC‑TM‑650 2.4.22 require a specific support method?
A: Yes. The board must rest on three points at the corners (typically near, but not at, the corners). The three supports must be coplanar. The fourth corner is free. No clamping or weights are allowed. The method also allows a flat plate with a central cutout for large boards.
Q: What is the formula for bow percentage in IPC‑TM‑650 2.4.22?
A: For bow, place the board concave side down on a flat surface. Measure the maximum gap (H) between board and surface. Bow = H / Diagonal length × 100%. For twist, use the corner height difference formula shown above.
Q: Can IPC‑TM‑650 2.4.22 be used for flexible circuits?
A: Yes, but the flexible circuit must be placed on a flat surface without tension. The same measurement method applies, but the board should be allowed to lie naturally. Bow and twist limits for flex circuits are typically looser.
Q: How does bow and twist affect BGA assembly reliability?
A: BGA balls have limited compliance (typically 0.1‑0.2 mm). If PCB warpage exceeds the ball compliance, some balls may not contact the board, causing opens or head‑in‑pillow defects. For large BGAs, limit bow to <0.3% of diagonal.
Bow & Twist Test Checklist (IPC‑TM‑650 2.4.22)
- 1. Board dimensions (length, width, diagonal) measured.
- 2. Granite surface plate flatness verified.
- 3. Dial indicator calibrated and resolution 0.001" or better.
- 4. Board conditioned at room temperature (no forced warpage).
- 5. Three support points placed near corners, coplanar.
- 6. Board placed on supports with fourth corner free.
- 7. Height measured at all four corners (A, B, C, D).
- 8. For bow measurement, board placed concave side down; maximum gap measured.
- 9. Twist calculated per formula: |(A+C)-(B+D)|/(2×diagonal).
- 10. Bow calculated as gap/diagonal.
- 11. Values compared to internal acceptance criteria (e.g., ≤0.75% for SMT).
- 12. Results recorded in inspection log.
Need to measure bow and twist for your PCBs? Our lab performs warpage measurement per IPC‑TM‑650 2.4.22 for incoming inspection, qualification, and process validation.
Request Bow & Twist Testing →IPC‑TM‑650 Test Methods Referenced: 2.4.22 (Bow and Twist).
Get Custom Quote
PRODUCTS CENTER


