Epoxy & Anaerobic Adhesives — Application & Performance Guide

What COA Data Actually Predicts — And What It Misses #

Procurement teams sourcing epoxy and anaerobic adhesives from China receive COAs that reliably report three values: viscosity, cure time, and lap shear strength at ambient temperature. These are the three parameters that are easiest to test and least predictive of joint survival in real operating conditions.

The parameter that actually determines whether a bonded joint survives 18 months of service is not on most Chinese supplier COAs. For epoxies, it’s glass transition temperature (Tg) combined with the coefficient of thermal expansion (CTE) mismatch between adhesive and substrate. For anaerobics, it’s fixture strength retention after fluid immersion — a test that takes 168 hours and that most Chinese suppliers will not run unless you contractually require it.

I’d prioritize Tg data above lap shear strength for any epoxy application where the service temperature exceeds 80°C. Lap shear at room temperature tells you about mix ratio compliance. Tg tells you whether the joint will creep at operating temperature.

Three Operating Scenarios — Performance Data and Failure Thresholds #

The three scenarios below represent the conditions where we’ve seen the largest divergence between COA data and field performance when sourcing from China. Each requires a different specification priority.

Thermal Cycling: -40°C to +125°C

A standard two-part epoxy with a Tg of 95°C will begin to exhibit viscoelastic behavior — measurable creep under sustained load — at temperatures above 70°C, roughly 25°C below its Tg. In a thermal cycling application between -40°C and +125°C, the joint is operating above this creep threshold on every high-temperature excursion. After 500 cycles (per IEC 60068-2-14 thermal shock testing), a Tg-95 epoxy typically shows 15–22% reduction in shear strength; a Tg-145 structural epoxy under the same protocol retains greater than 90% of initial strength in our qualification data.

CTE mismatch is the mechanism. Aluminum substrate runs at 23 µm/m·°C. A standard bisphenol-A epoxy runs at 55–65 µm/m·°C. That 30–40 µm/m·°C differential drives peel stress at the bondline with every cycle. Flexible-modified epoxies can reduce adhesive CTE to 40–45 µm/m·°C, which changes the failure trajectory significantly — but the modification shows up nowhere on a standard COA.

Strength retention data from 500-cycle IEC 60068-2-14 protocol, -40°C/+125°C, 15-minute dwell. Anaerobic data from internal qualification program, 12 lots across 4 suppliers.

The table above reflects what we see in qualification testing, not what supplier datasheets claim. Supplier datasheets — Chinese or Western — almost universally report single-point lap shear at 23°C. The 500-cycle column is where the product selection actually matters.

For thermal cycling applications, the selection decision comes down to this: if service temperature peaks above 100°C, specify Tg ≥ 130°C with a written COA requirement. If you specify Tg ≥ 130°C and the supplier returns a COA showing Tg of 118°C, that is not a minor deviation — it is a disqualifying result.

Chemical Exposure: Immersion or Splash Contact

Anaerobic adhesives outperform standard epoxies in most single-fluid exposure scenarios — hydraulic fluid, lubricating oil, diesel — because the cured acrylic polymer matrix is relatively non-polar and resists swell. Where anaerobics fail is in mixed chemical environments: anything involving ketones, chlorinated solvents, or concentrated acids above pH 2.

For epoxies under chemical exposure, the specification that actually predicts performance is water absorption after 24h immersion per ASTM D570, not chemical resistance listed on a datasheet. A standard BPA epoxy absorbs 0.1–0.3% by weight after 24h at 23°C. A poorly formulated epoxy — and we have received these from second-tier Chinese suppliers during Category B qualification reviews — can absorb 0.8–1.2%, which plasticizes the matrix and drops lap shear strength by 20–35% after sustained immersion.

Concentration matters more than chemical class. A 10% sulfuric acid solution may show minimal epoxy degradation over 30 days; 50% sulfuric acid will begin visibly attacking most standard epoxies within 72 hours. Buyers sourcing adhesives for chemical process equipment who state only “acid resistance required” on their RFQ get products optimized for the supplier’s interpretation, not theirs.

Pressure and Load Conditions: Static vs. Dynamic

This is the scenario where anaerobic threadlockers most often get misspecified. Vibration-induced loosening and static axial load are two completely different failure mechanisms, and the grade selection logic differs accordingly.

For vibration resistance, the DIN 267-28 Junker test is the correct qualification method. A medium-strength anaerobic (ISO 10123 Class 3, roughly equivalent to Loctite 243 in performance class) will retain 90–95% of breakaway torque after 20,000 Junker cycles on M10 zinc-plated steel. A low-strength grade (Class 1) will begin showing measurable loosening after 8,000–12,000 cycles under the same protocol. Most Chinese suppliers do not publish Junker test data. When we request it during qualification, roughly one-third can produce it for their flagship grade and fewer than one in five can provide data for secondary grades.

For static press-fit retention, what matters is compressive strength of the cured bead and gap-filling capability. Retaining compounds cure effectively up to a 0.25 mm radial gap; beyond that, cure inhibition from oxygen exclusion becomes incomplete and fixture strength is unpredictable.

The Variable That Shifts the Decision: Lot-to-Lot Tg Consistency #

Standard product comparisons focus on peak-rated performance. The variable that actually changes the sourcing decision — particularly when buying from China at volume — is consistency of that performance across production lots.

Across our 22-month evaluation of 8 Chinese epoxy suppliers (logged under our QC-14 adhesive lot tracking protocol), Tg variation within a single supplier’s consecutive lots ranged from ±3°C for the best performer to ±17°C for the worst. The worst performer had passed initial sample qualification on all COA parameters. The Tg drift emerged at month 4 of production supply, triggered by a resin batch change at their raw material compounder that was not disclosed.

This is a specific risk mechanism, not a general quality concern. Chinese epoxy formulators operate at two tiers: large compounders with in-house resin synthesis and smaller formulators who purchase base resins from commodity suppliers. The smaller formulators — which represent the majority of Chinese suppliers in the sub-$50,000 annual order range — have limited control over incoming resin EEW (epoxide equivalent weight). When EEW shifts by 5–8%, stoichiometric balance with the hardener changes, and Tg shifts correspondingly. The COA will still show passing viscosity and gel time. Only differential scanning calorimetry (DSC) will catch the Tg drift.

For applications where thermal cycling or elevated-temperature service is in scope, three consecutive lot COAs showing DSC-confirmed Tg within ±5°C of specification is a reasonable qualification threshold. If a Chinese supplier cannot provide DSC data at all — which is more common than buyers expect — that is a signal about their internal quality capability, not just a documentation gap.

The industry does not agree on how often to requalify adhesive suppliers. Some procurement teams run annual audits regardless of performance history. Others requalify only after a formulation change notification. A third practice, which we apply to high-risk adhesive categories in our AVL gate review process, is quarterly DSC spot-checks on production lots for any supplier whose raw material sourcing is not vertically integrated. For most Chinese epoxy formulators below tier-1, that means quarterly checks.

Implementation Notes — Incoming Inspection and Early-Shipment Red Flags #

After the material selection decision is made and a supplier is conditionally qualified, the incoming inspection protocol determines whether you catch problems before they reach the assembly line.

For epoxy adhesives, the minimum incoming inspection battery we recommend for critical applications is:

• Viscosity check against COA value (±10% tolerance, Brookfield method)
• Gel time at 23°C (±15% of stated value; deviations above this indicate hardener ratio shift)
• Shore D hardness of fully cured specimen at 24h/23°C (±3 Shore D from specification)
• DSC for Tg on every fifth lot, or on any lot where gel time is outside tolerance

For anaerobic adhesives, gel time is the primary incoming check. Breakaway torque testing on a sample M10 bolt per ISO 10123 takes less than 2 hours and will catch the most common failure mode — under-strength cure from contaminated activator or degraded resin from temperature excursion in shipping.

Chemical exposure applications add one step: water absorption per ASTM D570 on a cured specimen, 24h at 23°C. A result above 0.4% on a product specified at less than 0.2% is a disqualifying result and almost always traces to a base resin substitution.

Red flags in early production shipments are typically not visible in the adhesive itself. Watch for:

• Viscosity creeping upward over consecutive lots (indicates resin batch variability or EEW drift)
• Gel time shortening with no formulation change notice (indicates accelerator ratio shift)
• Color inconsistency between lots on pigmented systems (indicates raw material sourcing instability)
• Cure inhibition spots in anaerobic beads applied to low-activity metals — an issue that rarely appears in qualification samples because those are often made with activated test substrates

Establish a shelf-life tracking system from day one. Most Chinese epoxy suppliers ship product with a 12-month shelf life from manufacture date. If the COA manufacture date is absent or inconsistent with the batch number format, request re-verification before accepting the shipment. We have intercepted product with 3 months remaining shelf life being shipped into a 6-month production cycle — a situation that creates mid-program adhesive failures with no clear root cause if the receiving team does not check dates.

Set a formal milestone at the end of the first three production lots: review all incoming inspection data, compare Tg across lots, confirm gel time stability. If any parameter shows a drift trend rather than random noise, raise it as a corrective action before it becomes a field problem.

Practical Guidance for Buyers #

When sourcing epoxy or anaerobic adhesives from China for applications involving any of the three scenarios above, do not start the RFQ with lap shear strength requirements. Lap shear at ambient temperature is the specification every Chinese supplier can meet — it is the easiest number to optimize during sample preparation and the least predictive of joint performance in service.

Start with service temperature and Tg. If your application sees temperatures above 80°C, state a minimum Tg requirement on the drawing and request DSC confirmation on the COA. If a supplier cannot provide DSC data, your qualification decision should reflect that capability gap directly.

The specific risk scenario to plan for: a supplier who passes initial qualification at 8 lots and then shifts base resin at month 5 of production. This is not hypothetical — our QC-14 tracking data shows it as the most common mechanism for mid-program adhesive failures when sourcing from non-vertically integrated Chinese formulators. The shift does not show in viscosity or gel time. It shows in Tg — typically a 7–12°C drop — and in field performance 6–9 months later.

Before volume commitment, require a three-consecutive-lot DSC dataset showing Tg within ±5°C of specification, plus breakaway torque data (for anaerobics) or water absorption data (for epoxies in chemical exposure service) from the same three lots. Sample size: minimum 5 specimens per test per lot. This is a 2–3 week exercise, not a 6-month program. Any qualified supplier should be able to produce it. If they cannot, that information is worth more than any price negotiation you will have with them.

FAQ

Does Tg on the COA always reflect what you’ll see in the field?
Only if the cure schedule was followed exactly. Tg is cure-schedule-dependent — an epoxy cured at 23°C for 24 hours will have a Tg 15–25°C lower than the same product post-cured at 80°C for 2 hours. If your assembly process does not include a post-cure step, specify Tg at ambient cure conditions, not at full post-cure, and require the COA to match.

Can Chinese anaerobic adhesives reliably replace Western-brand threadlockers in critical assemblies?
It depends on the assembly classification. For standard M6–M16 steel fasteners in non-safety-critical applications, qualified Chinese medium-strength anaerobics perform comparably in Junker testing. For safety-critical fasteners, aerospace, or assemblies requiring traceable qualification documentation, the documentation trail from most Chinese suppliers does not currently support the level of traceability buyers need — and that is a compliance problem, not a chemistry problem.

What is the minimum order quantity to get DSC testing included in the COA?
Ask for it regardless of order size. The cost of a DSC run is roughly $30–60 per sample at a Chinese third-party lab. A supplier refusing to include DSC data on a $5,000 order is making a business decision, not a capability claim — and that tells you something about how they manage quality systematically.

How do you handle shelf life when sourcing epoxy adhesives from China with long lead times?
Specify manufacture date on the PO, not just expiry date. Require COA to include both. For a 12-month shelf-life product with a 10-week lead time and a 6-month production cycle, you need manufacture date to be no older than 8 weeks at time of shipment — state that explicitly. We have seen this missed consistently on first-time orders.

Is water absorption testing really necessary for every chemical exposure application?
No. For splash-contact or incidental exposure to lubricating oils or hydraulic fluids, standard epoxies with less than 0.3% water absorption at 24h/23°C per ASTM D570 are typically sufficient and the test is not worth the delay. For immersion service or contact with solvents, acids, or bases, it is non-negotiable — the 48-hour test will predict 6-month field behavior more accurately than any chemical resistance table from a supplier datasheet.

For related sourcing guidance, see epoxy and structural adhesive bond performance and pump and valve sealing material qualification.

Published by sinoraw.com Technical Team | Request a sourcing consultation