PERBOND
← The Journal
· The PerBond Team

Overlap Length or Joint Width: Which Adds More Strength to a Bonded Joint?

You have a bonded lap joint that needs to carry more. There are two ways to add bonded area: make the overlap longer, or make the joint wider. On paper they are the same, because area is length times width and the usual strength estimate is area times shear strength. In the joint they are not the same at all. This guide puts both options through the Lap Joint Efficiency Calculator and the Lap Joint Designer on one example, so the difference can be seen in numbers, and then sets out how far those numbers can be trusted.

Read this first

This is a comparison of two layouts using simplified models. None of the loads below is a rating for a real joint, and nothing here is a structural design. Joints that are load-bearing, overhead or relied on for anyone's safety need a qualified engineer, the adhesive manufacturer's tested design values, and normally mechanical fasteners too.

The example

Two strips of 2 mm aluminium, 25 mm wide, bonded with a 25 mm overlap, using an adhesive with a data-sheet lap-shear strength of 20 MPa. The efficiency calculator's defaults are used for the adhesive stiffness (1,000 MPa shear modulus) and the bond line (0.2 mm).

Two figures are compared throughout:

  • The straight-line figure: bonded area times shear strength, with no safety factor. This is the "lab-value ceiling" the Bond Strength Calculator shows.
  • The elastic-model figure: the straight-line figure multiplied by the elastic efficiency from the efficiency calculator. It represents the load at which the most highly stressed adhesive, at the ends of the overlap, reaches the data-sheet strength, assuming nothing yields.

For the starting joint the straight-line figure is 12,500 N. The elastic efficiency is 30 percent, with the shear at the ends 3.35 times the average, giving an elastic-model figure of 3,732 N.

Option A: double the overlap

Lengthen the overlap from 25 mm to 50 mm and keep the width at 25 mm. The area doubles and the straight-line figure doubles with it, to 25,000 N. The elastic efficiency, though, falls to 15 percent, and the elastic-model figure comes out at 3,742 N. That is a gain of about 0 percent over the starting joint. The overlap is now 25 times the sheet thickness, and the calculator flags it as past the point where length helps.

Push it further, to a 100 mm overlap, and the straight-line figure reaches 50,000 N while the elastic-model figure stays at 3,742 N, with efficiency down to 7 percent. Four times the adhesive, four times the bonded area, and on this model no more load.

Option B: double the width

Keep the 25 mm overlap and widen the joint from 25 mm to 50 mm. The area doubles, exactly as in Option A, and the straight-line figure is the same 25,000 N. But the stress pattern along the overlap has not changed, so the efficiency stays at 30 percent, and the elastic-model figure doubles to 7,465 N: a gain of 100 percent.

Layout (width × overlap)Straight-line figureElastic efficiencyElastic-model figureGain over start
25 × 25 mm (start)12,500 N30%3,732 N
25 × 50 mm (longer)25,000 N15%3,742 N0%
25 × 100 mm (much longer)50,000 N7%3,742 N0%
50 × 25 mm (wider)25,000 N30%7,465 N100%

Why the two behave so differently

Under load the strips stretch. Each carries the full load where it enters the overlap and none at its free end, so the two strips are stretched at opposite ends, and the adhesive at each end is sheared hard to make up the difference. In the middle the strips have stretched about equally and the adhesive is nearly idle. The stress along the overlap is a valley with a peak at each end.

Lengthening the overlap widens the valley floor. The two peaks stay where they were and stay as high as they were for a given load, so the load at which they reach the adhesive's limit hardly changes. Widening the joint is different in kind: it places more material side by side, each strip of width with its own pair of ends carrying its own share. Nothing about the stress pattern changes; there is simply more joint. That is why the calculator's own guidance is that length has a limit and width does not.

The same lesson when sizing from a load

The Lap Joint Designer works from a required load to an overlap. Ask it for 6,000 N on 2 mm aluminium with the 20 MPa adhesive and its default safety factor of 3, and try two widths. Both need the same bonded area, 900 mm².

  • 25 mm wide: overlap 36 mm, which is 18 times the thickness. Elastic efficiency 21 percent.
  • 75 mm wide: overlap 12 mm, 6 times the thickness. Elastic efficiency 58 percent.

Neither is flagged by the designer, since both are under 20 times the thickness (40 mm for 2 mm sheet). But the wide, short joint uses its adhesive nearly three times as effectively on the elastic model. Given the choice, it is the better layout. A flag-free result from the designer means the overlap is not absurd. It does not mean the layout is the best available.

Other things that move the efficiency

Using the 50 mm overlap from Option A (15 percent on 2 mm aluminium) as the reference:

  • Thicker parts. In 4 mm aluminium the same overlap shows 21 percent. Thicker strips stretch less.
  • Stiffer parts. In 2 mm steel it shows 26 percent, for the same reason.
  • A more compliant adhesive. With the adhesive shear modulus lowered from 1,000 to 200 MPa it shows 33 percent. A less rigid adhesive spreads the load further along the joint. Flexible adhesives have lower strength figures, so this is a trade, not a free gain.

How far to trust the elastic-model figures

As a ranking of layouts: a good deal. As loads: not at all. Three reasons.

  1. Real adhesives yield. The model treats the adhesive as perfectly elastic to failure. Tough structural adhesives yield at the ends and shed load inward, so a real joint goes on gaining strength with length for longer than the model shows, and breaks well above the elastic figure. The calculator describes its own output as the pessimistic end. With a brittle adhesive the model is closer to the truth.
  2. The data-sheet strength is not a peak-stress limit. It is the average stress at which a short standard test joint failed (12.7 mm overlap on 1.62 mm aluminium, itself at 51 percent by this model). Multiplying it by an efficiency for a different geometry, as the elastic-model figure does, penalises the joint twice. The straight-line figure at 50,000 N for the 100 mm overlap is far too high; the elastic-model figure of 3,742 N is too low. The real answer is between them and only a test finds it.
  3. Shear is not the whole story. A single lap bends under load and peels at the ends. Peel usually starts the failure in thin material, and neither tool models it.

What survives all three caveats is the comparison. Longer overlaps give diminishing returns, wider joints do not, and both effects are large.

What the tools do not cover

  • Peel, bending, fatigue, creep, impact, temperature and moisture.
  • Adhesive or adherend yielding.
  • Parts of different material or thickness on the two sides.
  • Joint types other than a single lap. Double-lap, strap, scarf and stepped joints balance the load path and behave better; they need their own analysis.
  • The quality of your surface preparation and cure.

Design rules to take away

  • Size the overlap to a modest multiple of the thickness, then get the rest of the area from width.
  • If the designer returns an overlap beyond 20 times the thickness, do not build it. Widen the joint, thicken the parts or change the joint type.
  • Bond the full width available. Unbonded width is capacity thrown away.
  • Leave the squeeze-out fillets at the ends, and consider a tougher adhesive before a longer overlap.
  • Add fasteners at the ends where peel is possible.
  • Make and break test joints before relying on any figure.

For the same effect in thin sheet steel see why a longer overlap stops adding strength in sheet metal, and for a step-by-step sizing example see how long a lap joint overlap should be.

Safety and limits

Structural adhesives are hazardous until cured. Epoxy resins and hardeners are skin sensitisers; two-part acrylics give off flammable, irritating vapour; polyurethanes contain isocyanates. Wear nitrile gloves and eye protection, work with good ventilation, wash with soap and water and never solvent, and read the product's Safety Data Sheet (SDS). Follow the manufacturer's mix ratio and cure schedule, and do not load a joint before full cure.

If a joint failing could hurt someone or cause serious damage, the design belongs with a qualified engineer working from the adhesive manufacturer's tested values, not with a calculator.

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