Pull testing and shear testing load a bond in fundamentally different directions. A pull test applies force upward along the Z-axis, while a shear test applies force laterally across the XY plane. The correct choice depends on the bond type, the sample geometry, and the failure mode you need to evaluate. The sections below answer the most common engineering questions about both methods.
How does a pull test actually load a bond?
A pull test applies an upward force to a wire, ribbon, bump, or component, pulling it away from the substrate along the Z-axis. The force is transferred through the bond interface until the weakest point fails. The result is a measured force value combined with an observed failure mode, and both pieces of information matter.
In standard wire pull testing, a 90-degree hook is positioned beneath the wire loop. The Z-stage moves upward, bending the wire and loading both bond points simultaneously. The hook position along the wire affects how the load is distributed between the two bonds. Precise hook alignment is therefore critical for reproducible measurements.
Tweezer pull testing works differently. The tweezers grip the wire, ball, pin, or component directly and apply force straight above the bond. Because the force is applied directly overhead, the measured force equals the applied force at the bond. This makes tweezer pull particularly useful when you need to isolate and test a single bond point rather than loading two bonds at once.
Cold Bump Pull (CBP) is a variation used for bumped wafers and advanced packaging applications. The tweezers grip a solder bump or copper pillar and pull it upward. CBP is especially relevant for wafer-level testing where standard hook-based wire pull is not applicable.
Pull tests can be performed in destructive or non-destructive (NDT) modes. In destructive testing, the force increases until the bond fails. In NDT, the force is raised to a defined limit and then released, leaving the bond intact. NDT is used when you need to verify that every bond meets a minimum strength requirement without destroying the part.
How does a shear test load a bond differently?
A shear test applies a lateral force to a bond or component, pushing it horizontally across the substrate surface. Instead of pulling upward, the shear tool moves in the XY plane and contacts the side of the ball, die, wedge, or other structure. The bond is loaded in shear rather than tension.
Shear height is one of the most important parameters in shear testing. It defines how high above the substrate surface the shear tool contacts the bond. Setting the correct shear height is critical. A shear height that is too high often produces bulk material failure rather than bond interface failure, which means the test does not reveal the actual bond strength. The optimal shear height is the lowest value that consistently produces the failure modes of interest and the highest test force.
Sample clamping is equally important. If the sample moves during a shear test, the effective shear height changes and results become inconsistent. Different clamping methods are available, including leading edge stops, trailing edge clamps, side clamps, and vacuum. Combinations are common. For example, a leading edge stop combined with vacuum holding provides both lateral restraint and vertical support.
Shear testing covers a wide range of bond types and structures. Gold ball shear, copper ball shear, solder ball shear, die shear, wedge shear, copper pillar shear, and ribbon shear are all lateral force tests, but each involves different tooling geometry, shear height requirements, and failure mode interpretation. The test method must match the specific bond and application.
Which test method should you choose for your application?
The correct test method depends on the bond type, the sample geometry, the failure mode you need to evaluate, and any applicable industry standards. Pull testing is generally used for wire bonds, ribbon bonds, and bumped structures where upward loading reflects the relevant service stress. Shear testing is used for ball bonds, die attachments, wedge bonds, and surface-mounted structures where lateral loading is appropriate.
A useful way to approach method selection:
- Wire bonds (gold, aluminum, copper): Wire pull test using a 90-degree hook, or tweezer pull when isolating a single bond
- Ribbon bonds: Ribbon pull or ribbon peel, depending on the geometry and failure mode of interest
- Solder balls, gold balls, copper balls: Ball shear test with appropriate shear height and tool geometry
- Copper pillars with solder caps: Copper pillar shear, with shear height set to pillar height plus approximately 2 µm
- Die attach bonds: Die shear test, with tooling selected based on die size, thickness, and bond material
- Wedge bonds: Wedge shear test for the bond pad interface
- Bumped wafers and advanced packaging: Cold Bump Pull (CBP) for upward loading of individual bumps
When the service load on a bond is primarily tensile, a pull test is more representative. When the service load is primarily lateral, shear testing is more relevant. In practice, many engineers run both methods to characterize a bond from multiple directions, particularly during process development or failure analysis.
If you are unsure which method applies to your sample, the bond test type overview provides structured guidance across pull, shear, peel, and push methods.
What failure modes does each test reveal?
The failure mode observed after a bond test tells you where the bond broke and what that means about bond quality. The measured force alone is not sufficient. A high force with the wrong failure mode can indicate a poor test setup rather than a strong bond. Always interpret force and failure mode together.
Pull test failure modes
In wire pull testing, the failure can occur at several locations. Failure at the first bond (the ball bond) or second bond (the wedge bond) indicates the strength of that specific interface. Failure in the wire mid-span indicates the wire itself is weaker than both bonds, which is generally a sign of good bond strength. Pad cratering, where the pad material separates from the substrate, indicates a substrate or pad adhesion issue rather than a bond process problem.
Shear test failure modes
In gold and copper ball shear testing, failure in the intermetallic compound (IMC) layer reveals the bond strength at that interface. Failure in the bulk material of the ball means the bond was stronger than the ball itself, which indicates good bond strength but does not quantify it precisely. Pad cratering points to a substrate or bonding process issue. Inconsistent shear height combined with bulk failure suggests a test setup or clamping problem rather than a bond quality issue.
In solder ball shear testing, failure in the IMC layer is the most informative result because it reflects the actual bond interface strength. Bulk solder failure means the solder failed before the interface, which does not characterize the bond directly. Pad failure and pad cratering indicate substrate or process issues.
In die shear testing, die bond failure indicates a weak bond. Die fracture means the die itself failed before the bond, which can happen when the die is thin relative to its area and the shear load exceeds the die’s yield strength before the bond fails. Die splintering at the test load application point is a sign of a test limitation rather than a bond quality problem.
What are the key parameters that affect test accuracy?
Test accuracy in both pull and shear testing depends on several interacting parameters. Getting these right is what separates a meaningful result from a misleading one.
For pull testing, the most critical parameters are:
- Hook position: The hook must be placed consistently along the wire loop. Small positional variations change the load distribution between the two bond points.
- Test speed: Pull speed affects the measured force. Higher speeds generally produce higher force readings. Consistent speed is essential for repeatable results.
- Gripping force (tweezer pull): For tweezer-based pull tests, grip force must be controlled precisely. Too little grip and the wire slips; too much and the wire deforms before the bond is loaded. USB tweezers with closed-loop force control provide the most consistent grip across multiple tests and multiple machines.
- Sensor accuracy: The force sensor must have sufficient accuracy and resolution for the load range being tested. The Sigma platform is specified at ±0.075% sensor accuracy.
For shear testing, the most critical parameters are:
- Shear height: As described above, shear height directly determines which failure mode occurs. It must be set and verified carefully for each application.
- Landing force: The force used to touch down on the substrate before stepping back to the programmed shear height. Consistent landing force is necessary for consistent shear height control.
- Tool geometry: The clearance angle, rake angle, and tip sharpness all affect how the tool loads the bond. A clearance angle of approximately 5 degrees is standard. A zero or negative clearance angle corrupts the shear height and must be avoided.
- Sample clamping: If the sample moves laterally or vertically during the test, the shear height changes and results become unreliable.
- Test speed: Shear speed affects the measured force, particularly for solder materials that exhibit rate-dependent behavior.
Can pull testing and shear testing be combined?
Yes. Pull testing and shear testing are complementary methods, and using both together gives a more complete picture of bond quality than either method alone. Many process development and failure analysis workflows use pull and shear tests in combination to characterize the same bond from different loading directions.
In practice, a wire bond process might be evaluated using wire pull testing to assess the bond interface strength in tension, and gold ball shear testing to assess the same bond laterally. If the pull test produces consistent wire-break failures but the shear test reveals IMC failures at lower-than-expected forces, that combination points toward a specific process variable affecting the ball bond interface.
Modular bond testers support this approach by allowing multiple test heads to be mounted and selected automatically during a test sequence. A system configured with both hook-based pull sensors and shear tools can switch between methods without manual intervention, which is particularly useful in automated production testing or high-throughput laboratory workflows.
Peel testing is a third method that is sometimes added to pull and shear workflows, particularly for ribbon bonds and adhesive film applications. A peel test applies force simultaneously in the Z and X or Y directions to maintain a consistent 90-degree peel angle, which produces a different loading profile than either a pure pull or a pure shear test.
What standards apply to pull and shear testing?
Several industry standards define test methods, parameters, and acceptance criteria for pull and shear testing in semiconductor and electronics manufacturing. The applicable standard depends on the specific test type, the bond material, the industry, and the customer or regulatory requirement.
For wire pull testing, MIL-STD-883 Method 2011 covers destructive bond pull strength testing. MIL-STD-883 Method 2023 covers non-destructive bond pull testing. DVS 2811 provides additional guidance for wire pull testing with pull angle modeling, which is relevant for more precise characterization of the load applied to each bond point.
For die shear testing, MIL-STD-883 Method 2019 covers die shear strength testing.
Always verify the exact method number, revision, and applicability to your specific bond type and industry before citing a standard as the basis for your test configuration. Standards are periodically revised, and the applicable version may differ depending on your customer requirements or regulatory context. Never assume that a standard applies to your application without confirming the scope and method details.
How xyztec supports pull testing and shear testing
With more than 25 years of exclusive focus on bond testing, xyztec provides the technology and expertise to implement, optimize, and automate both pull and shear testing across a wide range of applications. The Sigma platform is designed to support the full range of mechanical bond-strength tests within a single modular architecture.
Key capabilities relevant to pull and shear testing include:
- Sensor accuracy of ±0.075% across the measurement range
- Controlled shear height with precise landing force and step-back positioning
- Smart vision and automatic wire detection for consistent hook placement in wire pull testing
- USB tweezers with closed-loop grip force control for repeatable tweezer pull and Cold Bump Pull testing
- Patented self-aligning die shear tooling for large and thin die applications
- Modular test head configuration supporting multiple test methods in a single automated sequence
- Failure mode analysis including manual grading, Automated Optical Inspection (AOI), and deep learning-based automatic grading
- Custom tools and work holders for applications where standard solutions are not sufficient
If you are evaluating which test method applies to your bond type or want to understand how to configure a test for your specific application, contact xyztec to discuss your requirements with our bond-testing engineers.
Frequently Asked Questions
How do I know if my shear height is set correctly before running a full test batch?
Run a small set of trial tests at incrementally lower shear heights and observe the failure modes. The optimal shear height is the lowest value that consistently produces interface-related failures (such as IMC failure) rather than bulk material failure. If you are seeing bulk failures at your current setting, reduce the shear height incrementally until the failure mode shifts to the bond interface, then verify reproducibility across several samples before committing to that setting for production or characterization testing.
What is the most common mistake engineers make when interpreting bond test results?
The most common mistake is evaluating the measured force value without considering the failure mode. A high force reading paired with a bulk material failure or mid-span wire break does not confirm a strong bond interface — it may simply mean the test loaded the wrong part of the structure. Always record and interpret force and failure mode together, and flag any result where the failure mode is inconsistent with what your test setup is designed to reveal.
Can I use non-destructive pull testing (NDT) in a production environment, and what are its limitations?
Yes, NDT pull testing is well-suited for 100% production screening, where every bond must meet a minimum force threshold without being destroyed. The key limitation is that NDT does not tell you how much margin exists above the threshold — a bond that barely passes and one that would have failed at twice the limit look identical in the data. For this reason, NDT is most effective when combined with periodic destructive testing to monitor the actual strength distribution of your process.
How does test speed affect my results, and should I match the speed specified in an industry standard?
Test speed directly influences the measured force, particularly for rate-sensitive materials like solder. Higher speeds generally produce higher force readings, which means results from different speeds are not directly comparable. When a standard such as MIL-STD-883 specifies a test speed, use that value if your goal is compliance or cross-lab comparability. For internal process development, the most important thing is to keep speed consistent across all tests in a dataset so that speed-related variation does not obscure real process differences.
When should I consider running both pull and shear tests on the same bond, and how do I handle the sequencing?
Running both methods is most valuable during process development, qualification, or failure analysis, where you need to characterize a bond from multiple loading directions. Since both tests are destructive, they must be performed on separate samples — you cannot pull and then shear the same bond. A practical approach is to split your sample set and run pull tests on one subset and shear tests on the other, ensuring both subsets come from the same lot or wafer to keep process variables constant.
Are there bond types where neither a standard pull test nor a standard shear test is the right approach?
Yes. Ribbon bonds and adhesive film applications are common examples where peel testing is more appropriate than a pure pull or pure shear test, because the geometry and the relevant service stress involve a combined Z and XY loading component. Similarly, some advanced packaging structures with very low-profile or recessed bonds may require custom tooling or adapted test configurations that go beyond standard hook-based or shear tool setups. If your bond type does not map cleanly to a standard method, consult with a bond-testing specialist to identify the most representative loading approach.
What should I do if my test results show high variability between measurements on what appears to be a consistent process?
High variability in an apparently consistent process usually points to a test setup issue rather than a real process problem. The most common culprits are inconsistent hook placement in wire pull testing, variable shear height caused by sample movement or inconsistent landing force in shear testing, and insufficient clamping that allows the sample to shift between tests. Systematically check each parameter — hook position, shear height, landing force, clamping method, and test speed — and verify that your force sensor is operating within its rated accuracy range before concluding that the variability is process-related.
