Steel Cord Conveyor Belt Splice Failure: An Evidence-Based Inspection Guide

Key findings

• ISO 15236-1:2016 defines construction and performance requirements for general-use steel cord conveyor belts. It is not a prediction of field splice life.
• In a 2010 reviewed conference paper, researchers at Leibniz University Hannover described expanded dynamic test capability for belts with nominal strength above 7,800 N/mm. That test capability is not a field failure rate.
• Researchers at Wrocław University of Science and Technology have published studies on magnetic methods for examining steel cord belts and splice condition. Magnetic signals support follow-up inspection; they do not prove the root cause.
• Public records do not provide a harmonized denominator of splice failures per belt-year, operating hour, tonne conveyed, or splice installed.

Why “splice failure rate” is usually the wrong first question

A splice is affected by more than belt nominal strength. Design, cord alignment, bonding materials, cover preparation, press settings, cure time, pulley geometry, take-up behavior, belt tension, loading, tracking, contamination, and inspection quality can all change the result.

These variables are not recorded consistently in public accident or maintenance datasets. One mine may count a re-splice as a failure. Another may record only an unplanned belt stop. Without a common definition and denominator, a percentage can look precise while measuring different things.

What the main evidence types measure

Common conditions are not one-to-one diagnoses

An observed condition can have several contributing factors. Use the condition to choose the next check, not to declare a root cause before the evidence is complete.

What standards contribute

ISO 15236-1:2016 covers steel cord belts for general use and specifies constructional and performance requirements. ISO 15236-3:2017 adds special requirements for belts used in underground installations.

ISO 15236 defines requirements for belt construction. It does not replace the approved splice drawing, material instructions, vulcanization procedure, or site-specific engineering review. The applicable edition, belt application, local regulation, and project specification must be checked before a requirement is copied into a purchase or repair document.

What dynamic testing tells engineers

Leibniz University Hannover’s Institute of Transport and Automation Technology has studied conveyor-belt splice fatigue using test rigs that apply pulsating loads. In a 2010 reviewed conference paper, the group described test capability for high-strength belts above 7,800 N/mm nominal strength.

This evidence is useful for design review because static belt strength alone is not enough. A splice sees repeated loading. A test result is meaningful only with its belt construction, splice design, load history, pulley arrangement, and test method. It should not be converted into a promised service life for another conveyor.

Finite-element work from the same research group shows why splice analysis can require detailed stress modelling. The model is an engineering tool, not a substitute for checking the actual joint, cure record, and operating conditions.

What magnetic inspection can add

Researchers at Wrocław University of Science and Technology have published studies on magnetic methods for examining steel cord belts and splice condition. A 2020 study describes diagnostics based on magnetic-field measurements and signal analysis. A 2025 study examines magnetic, non-destructive assessment of splice geometry.

For a plant, the value is trend detection. Record the belt location, splice identifier, inspection date, method, signal interpretation, and follow-up action. A change from the belt’s previous baseline deserves review. It does not, by itself, identify whether the cause is a manufacturing defect, installation error, overload, mistracking, impact, or another mechanism.

A practical splice-investigation workflow

1. Make the area safe

Stop the conveyor under the site’s isolation procedure. Control stored tension and movement before removing guards, opening the belt, or approaching a take-up.

2. Preserve evidence

Photograph both belt surfaces, splice edges, cover condition, cord exposure, nearby idlers and pulleys, cleaners, loading zone, and take-up. Mark belt direction and splice position. Do not clean away evidence before it is recorded.

3. Compare the joint with its records

• approved splice drawing and revision;
• belt construction and cord grade;
• bonding-material batch, expiry, and storage records;
• press identification and calibration status;
• required environmental conditions;
• recorded temperature by zone, pressure, heating, hold, and cooling cycle;
• alignment and dimensional measurements;
• installer and supervisor records;
• post-installation acceptance and baseline scan;
• previous repair history and commissioning date.

Do not apply generic vulcanization temperature, pressure, or time values. Use the approved material system and procedure. If records are missing, classify that as a data gap.

4. Check system conditions

Review start-and-stop history, overloads, belt tension, take-up travel, pulley diameters, tracking adjustments, material impact, contamination, temperature exposure, and changes to loading or speed. A splice can be the visible failure location while the initiating cause sits elsewhere.

5. Select testing proportionate to the decision

Use visual and dimensional checks for an initial screen. Add magnetic inspection when belt length, consequence of failure, or suspected hidden damage justifies it. Use laboratory or destructive testing only with a defined question, sample location, and approved method. Acceptance criteria and operating decisions should come from the approved site procedure and qualified technical authority.

6. Classify the mechanism

Separate installation or cure defects, design mismatch, operating overload, tracking or pulley problems, impact damage, environmental degradation, and unknown cause. “Splice failed” is an outcome, not a root-cause classification.

How to calculate internal indicators

• Splice intervention rate = number of unplanned splice interventions ÷ total exposure of the defined splice population, reported as interventions per 100 splice-years or per specified number of splice-operating hours.
• Early-life splice intervention rate = splices requiring unplanned intervention within a defined early-life period ÷ commissioned splices that have completed that observation period or experienced the event.
• Corrective-action recurrence rate = eligible corrective actions followed by recurrence of the same classified mechanism within a defined follow-up period ÷ corrective actions completing that follow-up period.

Data limitations

The sources reviewed do not provide a harmonized public global dataset of steel cord splice failures by belt grade, splice design, operating hours, loading, maintenance quality, and mechanism. Research papers often describe controlled tests or diagnostic methods. Standards define requirements and test methods. None of these sources alone supports a universal failure percentage or fixed service-life promise.

This limitation tells a plant what to improve: consistent splice IDs, complete cure and acceptance records, repeatable inspections, and a failure taxonomy that can be analyzed across the site or fleet.