TL;DR: Overlap adhesive shear strength per [ASTM D1002](https://www.astm.org/d1002-10.html) tells you almost nothing about long-term structural performance — the parameter that predicts field failure is lap shear retention after thermal cycling, which most Chinese supplier COAs omit entirely.
TL;DR: In our supplier qualification program, 4 out of 7 Chinese structural adhesive suppliers we evaluated in 2024 could not provide thermal cycling data across 500 cycles from -40°C to 85°C — a gap that eliminated them from consideration for any metal-composite bonding application.
Selection Criteria That Separate Structural from Semi-Structural Performance #
The line between a structural adhesive and a high-strength assembly adhesive is not marketing — it is a testable threshold. For load-bearing applications, the minimum qualifying bar we apply is 10 MPa lap shear strength per ASTM D1002 on the target substrate combination, combined with less than 20% strength reduction after 1,000 hours of humidity aging at 85°C/85% RH. Below that retention threshold, the adhesive is not structural in any meaningful field sense, regardless of how the datasheet describes it.
For UV-curing systems specifically, the second criterion is full-depth cure at the actual bond line thickness in your assembly. A UV adhesive achieving 20 MPa on a 0.1 mm film may deliver 9 MPa on a 0.5 mm bond line if the photoinitiator package is optimized for thin-film cure. This is not a failure you catch at incoming inspection — it appears after assembly, under load.
The six criteria we use internally across our adhesive evaluation program (what we call the SAQ-6 matrix) are: lap shear strength at temperature, thermal cycling retention, humidity aging retention, cure depth at bond-line thickness, substrate-specific peel adhesion per ASTM D1876, and glass transition temperature (Tg) relative to maximum service temperature. Suppliers who can provide documented data on all six from a single product datasheet represent roughly the top 15% of what is available from Chinese manufacturers in this category.
Supplier Qualification — What to Request and What the Response Tells You #
Ask for lap shear data at three temperatures: 23°C (ambient baseline), -30°C (cold brittleness check), and the maximum rated service temperature. The response time matters. A qualified supplier returns this within two business days because the data exists in their QC records. A supplier who needs to “arrange testing” before responding does not have it — and that means lot-to-lot consistency is also unverifiable.
For structural adhesives going into metal-to-composite joints — increasingly common in EV battery enclosures and industrial robotics — also request failure mode documentation. Cohesive failure at the adhesive layer is acceptable. Adhesive failure at the substrate interface, especially on treated aluminum or CFRP, signals a surface energy mismatch that primer selection alone will not fully resolve.
Request three consecutive production batch COAs. Check whether the viscosity range tightens or widens across batches. A viscosity drift of more than ±15% across three consecutive lots is a formulation control warning — not necessarily a disqualification, but a trigger for incoming spot-testing on every subsequent delivery. One thing we specifically flag in our supplier audit form (Form SAQ-07b) is whether the listed viscosity measurement method specifies spindle number, RPM, and temperature. Without all three, the viscosity value on the COA is not reproducible and not verifiable.
For UV systems, ask for photoinitiator content by class (Type I vs Type II per ISO 11040) and the specific UV absorption peak. Suppliers who provide absorption spectrum data alongside their cure speed claims are operationally more credible than those who only quote mW/cm² exposure values without defining the spectral distribution.
There is a version of this supplier response you see regularly from mid-tier Chinese manufacturers: a well-formatted datasheet with all the right headings, but the test conditions are missing. Shear strength of “18 MPa” with no substrate, no test geometry, no temperature, and no standard cited. That document is not a COA — it is a brochure formatted like a COA. Experienced incoming QC teams catch this immediately. First-time buyers from Chinese suppliers sometimes do not.
Cost-Performance Trade-offs in Structural vs UV-Structural Systems #
Single-component UV-cure structural adhesives carry a raw material cost premium relative to two-part epoxies, primarily because the photoinitiator package and resin purity requirements are more demanding. At volume, the delta between a qualified two-part structural epoxy and a UV-cure structural adhesive from a Chinese tier-1 supplier typically runs 30–60% on a per-kilogram basis — though this varies significantly with order volume and resin type.
The counterargument for UV systems is cycle time, not material cost. If your assembly line is currently limited by epoxy cure schedules (even fast-cure two-part systems require 20–40 minutes for handling strength), switching to a UV-cure structural system with on-demand cure in under 30 seconds changes the economics entirely. The adhesive cost per unit increases; the labor and throughput cost per unit decreases. For high-volume electronic assembly and optical bonding, the calculation almost always favors UV despite the material premium.
For low-volume or prototype applications, the calculus changes. Two-part structural epoxies sourced from China in 5 kg cartridge quantities are significantly more cost-accessible, and the qualification burden is lower because the cure mechanism is less equipment-dependent. You are not buying a lamp, validating intensity, or managing shelf life under controlled light conditions.
One segment where buyers consistently over-invest is UV-cure for bond lines that are partially shadowed. A standard single-cure UV adhesive used in a geometry with any shadow area is a latent failure risk regardless of price point. The right answer in that case is a dual-cure system — UV primary, moisture or heat secondary — which is a different product family with different suppliers and different qualification criteria. Paying more for a premium UV adhesive does not solve a geometry problem. See our coverage of pump valve seals for a parallel case where cure geometry drives material selection in ways the datasheet does not flag.
Thermal Cycling Performance — The Parameter Most Structural Adhesive Evaluations Miss #
Lap shear strength at 23°C is the number every supplier leads with. It is also the least predictive parameter for structural adhesive performance in real assemblies, because real assemblies cycle thermally. Every powered-on, powered-off cycle in an industrial controller, every summer-to-winter transition in a bonded automotive panel, every sterilization cycle in a medical device — these are not static load events. They impose differential thermal expansion across the adhesive bond line, and the adhesive’s ability to absorb that strain without fatigue cracking determines its useful life.
The standard protocol we apply for thermal cycling evaluation is 500 cycles between -40°C and +85°C with 30-minute dwell times at each extreme, per the general methodology in IEC 60068-2-14 (thermal shock, test method Na/Nb). After cycling, we measure residual lap shear per ASTM D1002 and require a minimum 75% retention relative to the pre-cycle baseline. For adhesives bonding dissimilar substrates — aluminum to polycarbonate, steel to CFRP — we tighten this to 80% retention because the coefficient of thermal expansion (CTE) mismatch amplifies interfacial stress at every cycle boundary.
| Adhesive Type | Lap Shear at 23°C | Retention After 500 Thermal Cycles | Tg (Typical) |
|---|---|---|---|
| Two-part structural epoxy | 18–28 MPa | 70–85% | 80–130°C |
| UV-cure acrylate (structural grade) | 12–22 MPa | 60–78% | 50–90°C |
| UV-cure epoxy-acrylate hybrid | 15–25 MPa | 72–87% | 70–110°C |
| Polyurethane structural adhesive | 8–16 MPa | 80–92% | -20–40°C |
Retention data based on our evaluation of 14 Chinese and multinational suppliers, 2023–2024, using aluminum-to-aluminum lap joints per ASTM D1002 geometry.
Two observations from this data are worth highlighting. First, polyurethane structural adhesives consistently outperform epoxies and acrylates on thermal cycling retention, because their lower modulus allows elastic accommodation of CTE mismatch strain rather than storing it as crack-initiation energy. The trade-off is lower peak lap shear and a much lower Tg, which limits them to below 60°C continuous service. Second, UV-cure acrylates show the widest spread — we have seen Chinese-sourced UV structural adhesives come in at 55% retention and at 80% retention from suppliers quoting similar virgin datasheet values. The spread is almost entirely explained by the resin oligomer architecture, which is not disclosed on most Chinese supplier datasheets.
The open question we are still tracking across our supplier panel: whether the 80% retention threshold holds after 1,000 cycles rather than 500. Our dataset only covers 500-cycle testing to date. For applications requiring 10+ year service life with daily thermal excursions, we would recommend specifying 1,000-cycle testing explicitly in the qualification protocol — even if it means a longer supplier lead time before volume commitment.
Practical Guidance for Buyers #
When sourcing structural or UV-cure structural adhesives from China, the first specification to request is not tensile strength or viscosity — it is lap shear retention after thermal cycling on your specific substrate combination. That single datapoint screens out more unsuitable suppliers faster than any other parameter in our evaluation experience.
The specific risk to manage: a supplier who performs well on initial sample approval but delivers production batches with a shifted resin oligomer ratio. This typically shows up as a 15–25% drop in post-cycle retention without any visible change in the ambient lap shear value — meaning standard incoming inspection at room temperature will not catch it. The trigger for this type of substitution is almost always a compounder-level change driven by raw material availability, not intentional non-conformance. The mitigation is a quarterly incoming spot-test protocol: three samples per lot, tested at ambient and after 50 thermal cycles minimum.
Before volume commitment, insist on a 30-piece qualification run with documented cure parameters (for UV systems: irradiance in mW/cm², dose in mJ/cm², lamp spectral output), accompanied by lap shear test reports on the substrate combination in your assembly. Thirty pieces is not a large ask — it is the minimum statistically meaningful sample for a Cpk calculation at the bond strength threshold. Suppliers who push back on this ask are not qualified for structural applications regardless of their pricing.
FAQ
What is the minimum lap shear strength that qualifies as “structural” for industrial bonding applications?
The threshold we apply in our qualification program is 10 MPa on the substrate combination in the actual assembly, measured per ASTM D1002. Below that, the adhesive may handle secondary loads but should not be specified for primary load paths.
Does Tg need to exceed maximum service temperature for structural applications?
Yes, with a margin. We specify Tg at least 20°C above the maximum continuous service temperature — a structural adhesive with Tg of 80°C operating continuously at 75°C is working in its rubbery plateau regime and will exhibit progressive creep under sustained load. For thermal cycling applications, the margin should be wider still because Tg depression occurs under repeated thermal fatigue.
Can a UV adhesive be used for structural bonding on opaque substrates?
It depends on the joint geometry. If even one side of the bond line is transparent and the adhesive film thickness is within the cure depth specification — typically under 3 mm for standard UV structural grades — then yes. For fully opaque or shadowed geometries, you need a dual-cure system. A UV adhesive used on an opaque substrate where only edge irradiation is possible will partially cure at best; specifying it for structural duty in that geometry is a design error.
What should I check on a Chinese supplier COA for UV-cure structural adhesives?
Viscosity with full measurement conditions (spindle, RPM, temperature), photoinitiator content by class, cure speed at a defined irradiance and wavelength, and lap shear with test substrate and standard cited. A COA missing any of these is incomplete. The one field we flag most often in incoming review: viscosity listed without measurement temperature.
Is two-part epoxy or UV-cure acrylate better for aluminum-to-CFRP bonding?
For thermal cycling environments, the data favors epoxy-acrylate hybrids over either pure system — our evaluation data shows 72–87% lap shear retention across 500 cycles versus 60–78% for UV acrylates alone. For applications below 50°C with limited cycling, a polyurethane structural adhesive is worth evaluating for its superior strain accommodation, even though its peak lap shear is lower.
Published by sinoraw.com Technical Team | Request a sourcing consultation
