Epoxy vs Polyurethane Adhesive: Strength, Flexibility and Use

Epoxy adhesive suits rigid, high-strength, high-temperature metal joints; polyurethane adhesive suits flexible joints that must absorb impact, movement or vibration. Datasheets list 28-34 N/mm² steel lap shear for a toughened epoxy and 12-20 MPa for a two-part polyurethane, both ISO 4587 (Permabond ET5428 TDS, 2026; Permabond PT326 TDS, 2019). Epoxy's heat advantage needs a heat cure (Permabond ET5401 TDS, 2016).

• ISO 4587:2003, edition 3, measures tensile lap-shear strength of rigid-to-rigid bonds and was last reviewed and confirmed in 2021, and as of 2026-10-13 is under systematic review (stage 90.20) (ISO, 2003).

• One two-part epoxy datasheet lists 10-15 N/mm² lap shear on grit-blasted mild steel after 7 days at 25°C, and 20-30 N/mm² after 1 hour at 80°C (Permabond ET5401 TDS, 2016).

• The same epoxy lists a glass transition temperature of 110°C when heat cured and 50°C when cured at room temperature (Permabond ET5401 TDS, 2016).

• In the EU, products with 0.1% or more diisocyanates by weight may be used industrially or professionally after 24 August 2023 only by users who have completed safe-use training (EU Regulation 2020/1149, Annex).

This comparison covers two-part structural epoxies and two-part and one-part polyurethane adhesives on metal and composite joints. It does not cover foams, coatings, wood glues or construction sealants, and datasheet values are typical results, not design allowables.

How do epoxy and polyurethane adhesives differ?

Epoxy and polyurethane adhesives differ most in stiffness: epoxy adhesives cure rigid with high shear strength and low elongation, while polyurethane adhesives range from rigid two-part grades to elastic one-part grades that stretch several times their length (Sika, 2025). Polyurethane is therefore a family, not one property set, and a comparison has to name which kind. Both families are reactive adhesives that cure by chemical reaction, so cure schedule, surface preparation and bond-line thickness shape the final values as much as the chemistry does.

The table below compares datasheet values for a toughened epoxy, a rigid two-part polyurethane and an elastic one-part polyurethane, with the test method in each cell, and decides which family fits a joint's load and movement.

The elongation row decides most choices: the toughened epoxy breaks below 5% elongation, while the elastic one-part polyurethane reaches 400% (Sika, 2025). Tg and service temperature are different measures: compare hot-strength curves or service limits from the same maker, not a Tg against a service range. All of these are typical values; the epoxy datasheet states that it "offers guideline information and does not constitute a specification" (Permabond ET5428 TDS, 2026).

Is epoxy or polyurethane adhesive stronger in shear?

Epoxy adhesive is usually stronger in lap shear than polyurethane adhesive on prepared metal. A toughened two-part epoxy lists 28-34 N/mm² on mild steel and a two-part polyurethane lists 12-20 MPa on steel, both tested to ISO 4587 (Permabond ET5428 TDS, 2026; Permabond PT326 TDS, 2019). One N/mm² equals one MPa, so the epoxy range sits about twice as high. An elastic one-part polyurethane lists 2.5 MPa (Sika, 2025).

Cure conditions move epoxy shear strength as much as the chemistry does. One two-part epoxy lists 10-15 N/mm² on grit-blasted mild steel after 7 days at 25°C and 20-30 N/mm² on mild steel after 1 hour at 80°C (Permabond ET5401 TDS, 2016). A one-part heat-cure epoxy lists 27-41 N/mm² on steel after an oven cure such as 75 minutes at 130°C (Permabond ES558 TDS, 2019).

Lap-shear numbers compare adhesives; they do not size joints. ISO 4587:2003 states that "This test procedure does not provide design information" (ISO, 2003). ASTM D1002-10(2019), the active ASTM method for metal-to-metal single-lap joints, warns that using its results "as design-allowable stress values for structural joints could lead to product failure, property damage, and human injury" (ASTM International, 2019). Compare only values from the same method, substrate and surface preparation.

Which adhesive handles vibration and impact better?

Polyurethane adhesive generally handles vibration, impact and joint movement better than a rigid epoxy, because its higher elongation lets the bond line deform instead of cracking, an inference from the elongation data above; toughened epoxies narrow that gap. An elastic one-part polyurethane datasheet lists the product as "Vibration-damping" and states that it "is suitable for assemblies that are subject to dynamic stresses" (Sika, 2025).

Toughened epoxies are not brittle by default. One toughened two-part epoxy is described as having "excellent resistance to impact and vibration" and lists an impact strength of 30-40 kJ/m² to ASTM D950 (Permabond ET5428 TDS, 2026). The two-part polyurethane datasheet claims "Good resistance to impact and vibration" but gives no impact value (Permabond PT326 TDS, 2019), so a numeric impact comparison between the two families is not possible from these datasheets.

Thermal movement is a separate check. The two-part polyurethane lists a coefficient of thermal expansion of 85 × 10⁻⁶ per K to ASTM D696 (Permabond PT326 TDS, 2019), against 45 × 10⁻⁶ per °C below Tg for a one-part epoxy (Permabond ES558 TDS, 2019), so the polyurethane itself expands more. On joints between dissimilar materials, an elastic polyurethane bond line likely absorbs differential expansion better because it stretches, an inference from elongation, not a measured result. Where a joint sees peel, ask for peel data that matches the adherends: floating-roller peel (ISO 4578, as the epoxy datasheet uses) when one adherend is rigid, and T-peel when both are flexible, since ASTM D1876-08(2023) measures "the relative peel resistance of adhesive bonds between flexible adherends" (ASTM International, 2023).

Which adhesive tolerates moisture during application?

Neither standard epoxy nor standard polyurethane adhesive tolerates wet surfaces: both makers' datasheets ask for clean, dry surfaces, and on metal a two-part polyurethane carries a specific risk because its isocyanate reacts with surface water. The epoxy and two-part polyurethane datasheets both state "Surfaces should be clean, dry and grease-free before applying the adhesive" (Permabond ET5428 TDS, 2026; Permabond PT326 TDS, 2019).

The polyurethane risk is chemical. A polyurethane adhesive patent explains that "Any water, even as moisture absorbed or adsorbed on a substrate surface, presents a problem in bond strength" because water reacting with isocyanate forms a thin, weak cellular layer at the interface (US Patent 3,886,122, 1975). The same patent states that with impermeable substrates such as metals and plastics "this problem is critical" (US Patent 3,886,122, 1975). A one-part elastic polyurethane does use water, but from the air: it "cures by reaction with atmospheric moisture" and still requires surfaces that are "clean, dry and free from grease, oil, dust and contaminants" (Sika, 2025). Its open time is "significantly shorter in hot and humid climate" (Sika, 2025).

Epoxies can be formulated for damp work. One moisture-tolerant structural epoxy for concrete states "Substrate must be dry or mat damp (no standing water)" (Sika, 2026). On damp concrete, a moisture-tolerant epoxy grade is the documented option; on metal, both datasheets cited here ask for a dry surface whichever adhesive is used.

How do you switch a joint from epoxy to polyurethane adhesive?

Switching a joint from epoxy to polyurethane adhesive takes a redesign check, not a product swap, because the polyurethane carries less shear stress per unit of bond area. Work through six steps.

1. Recheck the load path. With datasheet lap shear of 12-20 MPa for a two-part polyurethane against 28-34 N/mm² for a toughened epoxy on steel (Permabond PT326 TDS, 2019; Permabond ET5428 TDS, 2026), the overlap or bond area likely has to grow; confirm by test.

2. Set the bond-line thickness. The two-part polyurethane fills gaps up to 5 mm (Permabond PT326 TDS, 2019); an elastic grade is applied as a triangular bead for a uniform bond line (Sika, 2025).

3. Prepare surfaces to the maker's instructions, including any primer or abrasion.

4. Plan the cure. The two-part polyurethane mixes 1:1 by volume and reaches full cure in 4-5 days at 23°C or 30 minutes at 90°C (Permabond PT326 TDS, 2019).

5. Read the SDS of the isocyanate component before use. In the EU, products with 0.1% or more diisocyanates by weight require completed safe-use training for industrial and professional users after 24 August 2023 (EU Regulation 2020/1149, Annex). Training must be renewed at least every five years, and handling warm, incompletely cured parts after a heat cure needs the advanced training elements (EU Regulation 2020/1149, Annex, entry 74, paragraphs 4 and 8).

6. Test the new joint on the actual substrates to ISO 4587 or ASTM D1002, including the service environment.

Lubinpla can work a switch like this from your own TDS, SDS and test records and return one cited document for your review.

What drives epoxy and polyurethane adhesive cost over service life?

The cost of an epoxy or polyurethane adhesive joint over its service life is driven more by cure time, process energy, rework and failure risk than by the adhesive itself.

Cure time ties up fixtures and floor space. The toughened epoxy reaches handling strength in 30-45 minutes and full cure in 24-48 hours at 23°C (Permabond ET5428 TDS, 2026), while the two-part polyurethane needs 60-90 minutes to handling strength and 4-5 days to full cure at 23°C (Permabond PT326 TDS, 2019). Heat shortens both: an increase of 8°C in temperature will halve the epoxy's cure time (Permabond ET5428 TDS, 2026), and the two-part polyurethane reaches full cure in 30 minutes at 90°C instead of 4-5 days at 23°C (Permabond PT326 TDS, 2019).

Process energy applies to heat-cure epoxies, which need an oven or induction step such as 75 minutes at 130°C (Permabond ES558 TDS, 2019). Polyurethane adds a training cost in the EU, where products with 0.1% or more diisocyanates need trained users (EU Regulation 2020/1149, Annex).

Shelf life and rework count as well. An elastic polyurethane in cartridges keeps 9 months when stored below 25°C (Sika, 2025), and once cured it can only be removed mechanically (Sika, 2025). The largest cost is a wrong choice: a rigid epoxy in a joint that flexes, or a low-shear polyurethane in a joint sized for epoxy, likely fails early; this is an inference, not measured data.

FAQ

Can epoxy and polyurethane adhesives bond plastics?

Yes, with checks. The two-part polyurethane datasheet warns that "This product may affect some thermoplastics and users must check compatibility of the product with such substrates" (Permabond PT326 TDS, 2019). The elastic polyurethane maker advises seeking its advice before use on plastics prone to stress cracking (Sika, 2025). The toughened epoxy lists only 4-5 N/mm² on PEEK (Permabond ET5428 TDS, 2026).

Is ASTM D1002 the same test as ISO 4587?

No. Both are single-lap tension shear tests, but ASTM D1002-10(2019) covers adhesives for bonding metals (ASTM International, 2019), while ISO 4587:2003 covers rigid-to-rigid bonded assemblies (ISO, 2003). Compare values only within one method.

Does a polyurethane adhesive need a primer?

It depends on the substrate. One elastic polyurethane datasheet states that "The surface treatment depends on the specific nature of the substrates and is crucial for a long lasting bond" and points to the maker's pre-treatment chart (Sika, 2025).

Working a case like this? Lubinpla works it from your own documents and site history, step by step with your approval. Free to start, no card.

Lubinpla does not sell chemical products and names no brands in recommendations. Our reviewer also leads a specialty chemicals distribution company.

Sources checked 2026-10-13.

Sources

1. ASTM International. ASTM D1002-10(2019) Standard Test Method for Apparent Shear Strength of Single-Lap-Joint Adhesively Bonded Metal Specimens by Tension Loading (Metal-to-Metal). 2019. https://www.astm.org/d1002-10r19.html

2. ASTM International. ASTM D1876-08(2023) Standard Test Method for Peel Resistance of Adhesives (T-Peel Test). 2023. https://www.astm.org/d1876-08r23.html

3. ISO. ISO 4587:2003 Adhesives, Determination of tensile lap-shear strength of rigid-to-rigid bonded assemblies. Edition 3, 2003, confirmed 2021. https://www.iso.org/standard/34852.html

4. Permabond. ET5428 Two-Part Epoxy, Technical Datasheet. Revision 9, 2026. https://permabond.com/tds/et5428_tds-en/?pdf=1

5. Permabond. ET5401 Two-Part Epoxy, Technical Datasheet. Revision 11, 2016. https://permabond.com/tds/et5401_tds-en/?pdf=1

6. Permabond. ES558 Single-part, heat-cure Epoxy, Technical Datasheet. Revision 10, 2019. https://permabond.com/tds/es558_tds-en/?pdf=1

7. Permabond. PT326 Polyurethane Adhesive, Technical Datasheet. Revision 9, 2019. https://permabond.com/tds/pt326_tds-en/?pdf=1

8. Sika. Sikaflex-252 Product Data Sheet. Version 04.01, 2025. https://industry.sika.com/content/dam/dms/us01/o/sikaflex-252.pdf

9. Sika. Sikadur-31 CF Normal Product Data Sheet. Version 01.03, 2026. https://sgp.sika.com/content/dam/dms/sg01/v/sikadur_-31_cf_normal.pdf

10. US Patent 3,886,122. Polyurethane adhesive composition with minimal moisture sensitivity. 1975. https://patents.google.com/patent/US3886122A/en

11. European Commission. Commission Regulation (EU) 2020/1149, Annex (diisocyanates, REACH Annex XVII entry 74). 2020. https://eur-lex.europa.eu/eli/reg/2020/1149/oj

Written by the Lubinpla team from the sources listed and checked against them before publication.