Adhesive bond failure investigation form for field complaints

What is an adhesive bond failure investigation form for, and who uses it?

An adhesive bond failure investigation form is a field record that fixes what a failed bonded joint looked like, how it was made and what it saw in service, so that the engineer who later decides the cause works from evidence rather than from recollection. Adhesive suppliers' technical service engineers, distributors handling a customer complaint, and the customer's quality or manufacturing engineers use it when a bonded part comes apart in production, in a test or in the field.

The form has seven sections, and each one is a checklist item that must be either filled or marked unknown with a reason:

1. Incident overview and affected scope: what happened, when, where, how many parts, and whether a safety signal is present.

2. Initial response and control measures: parts held, their location and state, and what releases them.

3. Failure surface: the failure pattern named from ISO 10365:2022, percent area per pattern, and photo references.

4. Substrate and preparation: material, coating or treatment, preparation method and time to bonding.

5. Adhesive and cure: product, lot, mix ratio, bond line thickness, cure temperature and time.

6. Service environment and load: temperature, humidity, load type and exposure.

7. Open hypotheses: possible causes, each marked unconfirmed until a test or record settles it.

A Completed by block closes the form with the name and organization, the date (YYYY-MM-DD) and the case or report ID the form feeds. The first two section names follow the incident and containment chapters of a problem resolution or corrective action report, so the form can feed that report without rework; cause, actions and effectiveness checks belong to the report, not to this form. The editable Word form is at the end of this page.

Which details of the failed bond should a field investigation form capture first?

A field investigation form for a failed adhesive bond captures the incident facts, the affected scope and the containment state first, because those details decide who must act today and they degrade fastest. A safety signal comes before everything else: if the failure involves a fall, a leak, a fire risk or exposure of people, the case goes to the site emergency procedure and the responsible safety function first, and the form records only that it was escalated, when and to whom.

The preview below lists the fields of sections 1 and 2, with an example value for each; every example is illustrative and does not describe a real case.

The row that most often goes wrong is the state of each control: a hold that has only been requested is recorded as requested, not as done. Each affected group of parts is recorded as affected, unaffected or unconfirmed, and a group nobody has checked stays unconfirmed. Write observed facts and reported facts on separate lines, with the name of the person who reported them, so the later report can tell what the investigator saw from what was passed on by phone or email. Note the time between the failure and the visit too: parts that sat outdoors or were reworked for days before anyone looked at them can show surfaces that no longer match the moment of failure, and the reviewer needs to know that before reading section 3.

How do you record the failure mode of an adhesive bond using ISO 10365 names?

Record the failure mode of an adhesive bond by naming each failure pattern visible on both mating surfaces with its ISO 10365:2022 designation, estimating the percent of the bonded area for each pattern so the shares total 100, and photographing both surfaces before anyone handles them further. ISO 10365:2022 specifies the designations for the main types of failure pattern of bonded assemblies and illustrates their appearance with diagrams (ISO 10365:2022). The 2022 edition added failure with stress whitening of adhesive (SWCF) and debonding due to bondline corrosion (COR) to the 1992 list (BSI, 2022), so an older internal sheet may lack two patterns that field failures can show.

Copy the designation and its definition from the standard itself into the form legend; the form does not reproduce the standard's text. For plain-language labels in remarks, use the terms of ASTM D907-15(2023), the terminology standard for adhesives (ASTM, 2023). For bonded fiber-reinforced plastic, ASTM D5573-99(2019) names adhesive, cohesive, thin-layer cohesive, fiber-tear, light-fiber-tear and stock-break failure, and shows mixed failures as percentages, such as 32% adhesive failure with 68% fiber-tear failure (ASTM D5573-99(2019), Fig. 10). That practice bases each classification on visual observation without magnification and does not address whether any failure mode is acceptable (ASTM, 2019).

The failure surface fields decide what the photos and estimates must show.

In the illustrative row, a joint with 70% of its area failed at the panel interface points the investigation at the panel surface, but the form only records that pattern; the cause belongs to section 7 as a hypothesis.

Which surface preparation, cure and environment records explain an adhesive bond failure?

The surface preparation, cure and environment records that explain an adhesive bond failure are the ones that show how each joint was actually made and loaded: substrate and preparation method, time from preparation to bonding, adhesive lot and mix ratio, bond line thickness, cure temperature and time, and the service temperature, humidity and load. ISO 17212:2012 describes the usual procedures for preparing metal and plastic surfaces before bonding (ISO 17212:2012), and ASTM D2651-01(2024) covers procedures for preparing wrought metal surfaces, not cast metals (ASTM, 2024), so the form asks which documented procedure was followed and whether it was followed on the failed lot.

Cure records need temperatures, not only times. A 3M white paper states that adhesives that cure by chemical reaction are sensitive to temperature, and gives as an approximate estimation that a 10 °C warmer cure environment halves the open time while a 10 °C cooler one doubles it (3M, 2012). Mix ratio needs the ratio as dispensed, because epoxy components react stoichiometrically (Henkel, 2008).

The preview below shows the fields of sections 4 to 6.

The illustrative cure row, 12 °C for 4 hours before load, is the kind of record that later lets an engineer compare the actual cure against the adhesive's data sheet conditions. Where a value comes from the process specification rather than a log, the form marks it as specified, so nobody later reads an intended cure temperature as a measured one. Bond line thickness is best measured on the failed part itself, at several points across the joint, because a range says more than a single nominal value.

How do you fill in an adhesive bond failure investigation form on site?

Fill in an adhesive bond failure investigation form on site in the order the evidence disappears: safety and containment first, then the failure surfaces, then the process records, then hypotheses.

1. Check for a safety signal and record the escalation if there is one. Handle any cleaning agent, primer or adhesive on site only under the site's rules and the product safety data sheet (SDS).

2. Record the incident facts and the counts of failed, inspected and uninspected parts, and mark each group of parts affected, unaffected or unconfirmed.

3. Record each hold or quarantine with its state and evidence, and the condition for release.

4. Photograph both failure surfaces of each failed joint with a scale and label before bagging, then name each pattern and its percent area in section 3.

5. Bag and label failed parts and, if available, an unused part from the same lot and retained adhesive from the same lot, and write who holds each sample.

6. Copy the preparation, adhesive and cure records for the failed lot from the line or job records, not from the process specification, and mark any value taken from the specification as specified, not measured.

7. Record service conditions and the load at failure if known.

8. List every hypothesis across substrate, preparation, adhesive, mixing, cure, design and service, each marked unconfirmed with the check that would settle it.

Forms most often fail at steps 4 and 6: surfaces photographed after handling, and nominal values copied as if they were measured. A blank with a reason is more useful to the reviewer than a value nobody measured.

What does an illustrative filled bond failure investigation form look like?

An illustrative filled bond failure investigation form for a bracket that detached from a coated steel panel shows every section completed or marked unknown, with hypotheses left open. The example is illustrative: the part, quantities, readings and hypotheses are invented to show the format and are not a real case or test data.

The illustrative incident section reads: 2026-04-08, line inspector found 6 brackets detached of 240 inspected on line 2; no safety signal. Containment reads: 6 failed parts bagged and labelled; 180 assemblies in stock on hold, hold executed with tag photo; release when the cause section of the corrective action report is agreed. The failure surface reads: adhesive failure at the panel coating interface over about 70% of the area, cohesive failure in the adhesive over about 30%, both surfaces photographed, no magnification.

The illustrative hypotheses table keeps every possible cause open with the check that would settle it.

The row that matters most in this illustrative form is the first: a preparation record gap that matches the interface failure pattern, still marked unconfirmed until coupons are retested. The other three rows stay in the table even though their records point the other way, because a hypothesis is closed by a check, not by a preference. The illustrative process section completes the picture: powder-coated steel panel, wipe and abrade before bonding, two-part epoxy dispensed at 2:1 by the line dispenser, bond line 0.2 to 0.6 mm measured on two failed parts, and cure at about 12 °C for 4 hours before the brackets were loaded; all of these values are illustrative.

When does a field bond failure investigation form not fit?

A field bond failure investigation form does not fit when the question is not one failed joint population in production or service but a design qualification, a laboratory test result, or a safety-critical structural failure under formal investigation. Laboratory tests carry their own report fields in the test method; ISO 10365:2022 still applies to name the failure patterns, because it covers all mechanical tests on a bonded assembly (ISO 10365:2022), but specimen geometry, test speed and conditioning belong to the test method's report, not to this form.

A structural failure that injured people or is under a regulator's or insurer's investigation needs that investigation's chain of custody and evidence rules; the field form can be attached but does not replace them. Fiber-reinforced plastic joints need the ASTM D5573-99(2019) mode names, which were developed on glass fiber in thermoset polyester sheet-molding compound (ASTM, 2019), so a carbon fiber or thermoplastic joint may show surfaces that the photographs in that practice do not match. Bonds made with tapes, sealants or hot melts can still use the form, but the cure section changes to the product's own setting or dwell conditions.

The form also stops short of the cause. Cause, corrective action and effectiveness checks belong to the corrective action report that the first two sections feed. In Lubinpla, the completed form and its photos can be attached to a work log, and a shared hashtag on each log keeps one customer's bond failure history in one place.