PERBOND
← The Journal
· The PerBond Team

Bonding Wood, Plastic or Painted Metal: When the Surface, Not the Glue, Sets the Strength

A "multi-surface" or "bonds almost anything" adhesive makes one promise: it will stick to many materials. It makes no promise about how strong the result is on each of them, and the difference between materials is far larger than the difference between adhesives. A bonded joint is a stack of layers, and it fails in whichever layer is weakest. Very often that layer is not the glue. This guide uses the Bond Strength Calculator, which has a separate input for the strength of the surface, to show how much that changes the answer.

Read this first

Every load figure below is a simplified estimate for comparison. None of it is a structural rating. The surface strengths are illustrative round numbers chosen to show the effect, not data for your materials. Do not use this guide or the calculator to decide whether a joint is safe for anything overhead, load-bearing or relied on for someone's safety. That needs a qualified engineer and the adhesive manufacturer's tested values for your exact substrate and preparation.

The layers in a "simple" glued joint

Bond a steel bracket to a painted steel panel and count what the load passes through: the bracket, the bracket's surface, the adhesive, the paint's top surface, the body of the paint film, the bond between paint and primer, the primer, the bond between primer and steel, and finally the panel. The adhesive is one layer of nine. Its 20 MPa on the data sheet was measured on clean, abraded metal with none of the other layers present.

The same thinking applies to other materials:

  • Wood fails by a thin layer of fibres tearing out of the surface. On plywood, the face veneer can peel away from the layer beneath.
  • Painted and powder-coated surfaces fail where the coating lets go of what is under it.
  • Galvanised steel can fail within the zinc layer or its surface oxide.
  • Plastics vary enormously. ABS, PVC and acrylic bond well. Polyethylene, polypropylene and PTFE have low surface energy: most adhesives sit on them without properly wetting the surface and peel off cleanly.
  • Concrete and masonry fail in the weak, dusty surface layer unless it is removed.

Worked comparison: one adhesive, five surfaces

Take a bonded patch 40 mm by 50 mm, an area of 2,000 mm², and an adhesive with a data-sheet lap-shear strength of 20 MPa. In the calculator choose "Max load from a bonded area", enter the area and the 20 MPa, leave the safety factor at 3, and put the surface figure in the optional "Substrate strength" box. The surface values here are illustrative only.

Bonded toIllustrative surface strengthGoverned byLab-value ceilingWorking estimate (÷3)
Clean, abraded bare steelnot limitingadhesive (20 MPa)40,000 N13,333.3 N
ABS plastic8 MPasubstrate (8 MPa)16,000 N5,333.3 N
Softwood, face grain6 MPasubstrate (6 MPa)12,000 N4,000 N
Painted steel3 MPasubstrate (3 MPa)6,000 N2,000 N
Untreated polypropylene1 MPasubstrate (1 MPa)2,000 N666.6 N

Same adhesive, same area, same care in application. The estimate on the painted panel is 85 percent lower than on bare steel, and the adhesive is not the reason. In every row but the first, the result box reads "Governed by: substrate".

A stronger adhesive does not help once the surface governs

The instinctive fix for a weak joint is a stronger glue. Try it in the calculator: keep the painted panel at 3 MPa and raise the adhesive from 20 MPa to 30 MPa. The working estimate stays at 2,000 N. The extra strength in the adhesive is never called on, because the paint lets go first.

The calculator handles the opposite case sensibly too. Enter a surface figure higher than the adhesive (40 MPa against the 20 MPa adhesive) and the result goes back to 13,333.3 N, governed by the adhesive. It always uses the weaker of the two numbers.

So the questions to ask about a weak joint, in order, are: which layer failed, can that layer be removed or improved, and only then, is the adhesive right? Looking at a failed joint tells you the first. If there is adhesive on both parts, the adhesive itself broke. If one part is clean, the bond to that surface failed, which points to preparation or to a surface the adhesive cannot grip. If there is paint, fibre or a skin of material stuck to the adhesive, the surface gave way.

What you can do about the surface

  • Remove the weak layer. Sand or strip paint back to sound metal under the bond area, take concrete back to hard aggregate, abrade galvanising lightly. This turns a 3 MPa problem back into an adhesive-limited joint.
  • Degrease, abrade, degrease. Clean first so you do not grind contamination in, abrade to a fresh surface, then clean off the dust. Use the cleaner the adhesive maker recommends.
  • Prime or treat low-energy plastics. Polyethylene and polypropylene need a primer, flame or plasma treatment, or an adhesive formulated for them. Check the data sheet lists your plastic by name.
  • Add area. If the surface cannot be improved, spread the load over more of it.
  • Add a mechanical fixing. A screw, rivet or bolt through the joint takes peel loads and provides a back-up.

How much more area a weak surface needs

Switch the calculator to "Min area for a target load" and ask for a working load of 1,000 N at the default safety factor. On bare steel it wants 150 mm². On ABS, 375 mm². On softwood, 500 mm². On the painted panel, 1,000 mm². On untreated polypropylene, 3,000 mm², which is 20 times the bare-steel figure. Area is cheap on a large panel and impossible on a small part, which is why surface treatment, not area, is the usual answer for difficult plastics.

Where the surface figure comes from

This is the hard part, and the reason the numbers above are labelled illustrative. Sources, best first:

  1. Your own test. Bond a few offcuts of the real materials, prepared the real way, cure them fully, and break them. The failure pattern alone tells you which layer governs.
  2. The adhesive's technical data sheet. Good ones list lap-shear strength on several named substrates (steel, aluminium, ABS, PVC, GRP, wood) and note "substrate failure" where the material broke first. Use the figure for your substrate as the adhesive strength, and you do not need the substrate box at all.
  3. The coating or material supplier, for adhesion or shear values.

A number taken from a general table or a forum is not a basis for anything that matters.

Dissimilar materials bring problems area cannot fix

The calculator is a strength-times-area model. Joining two different materials adds effects it does not see:

  • Different thermal expansion. Aluminium expands about twice as much as steel for the same temperature rise, and most plastics several times more. A rigid adhesive between two materials that move differently is stressed every time the temperature changes, with no external load at all. Long joints and outdoor joints suffer most. A flexible adhesive in a thicker bond line absorbs the movement, at the cost of a lower strength figure.
  • Different stiffness. When one part stretches more than the other under load, the stress concentrates at one end of the overlap. See why a longer overlap stops adding strength.
  • Moisture. Wood swells and shrinks across the grain with humidity; metal does not.
  • Corrosion. Two different metals in contact in a damp place can corrode each other. A continuous adhesive layer helps keep them apart; a patchy one does not.

What the calculator does not cover

  • It does not know any material's strength. The substrate box is empty until you fill it, and the answer is only as good as that number.
  • It models shear only. Coatings and laminated surfaces are weakest in peel, where they fail at far lower loads.
  • It ignores thermal expansion, moisture movement, creep and ageing.
  • It assumes stress is even across the bond area, which it is not.
  • It cannot tell you whether an adhesive will wet a given plastic at all.

Safety and limits

Surface preparation has its own hazards. Sanding old paint can release lead dust in older buildings and vehicles: test before you sand, and use dust extraction and a suitable mask. Solvent cleaners and primers for plastics are flammable and harmful to breathe, so ventilate and keep ignition sources away. Flame treatment of plastics is a fire risk and not a job to improvise.

For the adhesives: epoxy resins and hardeners are skin sensitisers, polyurethanes contain isocyanates, and cyanoacrylate bonds skin instantly and reacts hot with cotton. Wear nitrile gloves and eye protection, wash with soap and water and never solvent, and read the Safety Data Sheet (SDS) for every product you use, including cleaners and primers. Follow the manufacturer's preparation, mix ratio and cure schedule over anything here. For any joint that is load-bearing, overhead or safety-related, get a qualified engineer and tested design values.

For what the strength figure itself means and how the safety factor works, see how much weight a glued joint can hold.

Get new guides by emailOccasional emails. No spam. Unsubscribe anytime.