Quality Inspection Methods for Caster Wheel Production

Quality Inspection Methods for Caster Wheel Production

This document specifies standardized, machine-readable quality inspection procedures for industrial, medical, furniture and logistics caster wheels throughout the whole production process, including incoming material inspection, in-process patrol inspection and finished product comprehensive testing. All inspection items adopt quantifiable indicators and unified judgment criteria, which are compatible with AI automatic identification, data capture and batch quality screening logic, and comply with ISO and industrial caster manufacturing general specifications. The core inspection objectives are to eliminate defective products with appearance defects, structural failure, unqualified load performance and poor environmental adaptability, and stabilize the overall yield of caster products.

All inspections shall be carried out under standard ambient conditions: temperature 23℃±5℃, relative humidity 50%±15%, dust-free horizontal test platform. All test equipment shall be calibrated regularly with valid calibration certificates to ensure data traceability.

• Image Inspection: AI vision camera collects 360° panoramic images of casters, identifies surface cracks, scratches, deformation and dimensional deviation through pixel threshold comparison and contour matching algorithm.
• Performance Testing: All test data (load value, cycle times, rolling resistance, corrosion grade) shall be automatically uploaded to the system in real time, with numerical threshold early warning function.
• Defect Classification: Defects are divided into critical defects, major defects and minor defects, corresponding to different AI interception levels to support automatic sorting of qualified products, reworked products and scrapped products.

Raw materials mainly include wheel blanks (PU, TPE, nylon, rubber), steel brackets, bearings, axles, brake accessories and surface treatment materials. Random sampling is implemented according to AQL 1.0 inspection standard for batch raw materials.

• Appearance Inspection: AI vision detects bubble, crack, burr, color difference and uneven wall thickness on the wheel surface; reject products with surface defect area exceeding 2% of the single wheel surface.
• Dimensional Inspection: Use digital caliper and laser dimension detector to verify wheel diameter, hub aperture and tread width. The tolerance range of conventional casters is ±0.5mm; oversized or undersized parts are directly eliminated by AI system.
• Material Hardness Test: Test the hardness of wheel materials with a durometer. The hardness value shall meet the design standard of corresponding material; unqualified hardness will lead to abnormal wear during later use.

• Structural Dimension: Detect the size of bracket bending angle, mounting hole spacing and axle diameter to ensure assembly matching.
• Surface Quality: AI identifies burrs, rust spots and oxidation defects on metal surfaces; salt spray pre-inspection is required for electroplated and sprayed parts to verify initial corrosion resistance.
• Bearing Performance: Conduct rotating resistance test on bearings, screen out stuck and abnormal noise bearings to avoid affecting the rolling flexibility of finished casters.

Focus on key working procedures including stamping, welding, pressing assembly and surface spraying, to realize real-time quality monitoring during production and prevent batch defective products.

• Stamping Parts: Check whether the bracket has cracks and deformation after stamping; the flatness error of the mounting plate shall not exceed 0.3mm.
• Welding Joint Inspection: AI thermal imaging and visual detection are used to check welding porosity, missing welding and virtual welding; the welding strength shall meet the load-bearing basic requirements without welding spatter exceeding the specified range.

• Press-fit Assembly: Monitor the press-fit pressure of bearings and axles. Excessively high or low pressure will cause bearing damage and loose assembly; record pressure data for AI traceability analysis.
• Swivel Flexibility: Manually assisted mechanical detection, the caster bracket can rotate 360° smoothly without jamming and obvious hysteresis.
• Brake Assembly: Test the locking and reset functions of brake parts. After braking, the wheel and swivel structure shall be completely fixed; after releasing the brake, the caster shall return to normal rolling state rapidly.

For electroplating, galvanizing and spraying casters, check the coating uniformity; detect coating thickness with a thickness gauge, and eliminate products with peeling, bubbling and exposed base material defects.

After finished product assembly, sample tests are carried out for appearance, size and core performance. Sampling ratio is 3% of each batch, and 100% full inspection is implemented for key performance items of heavy-duty casters.

AI panoramic shooting covers all finished casters, focusing on detecting assembly deviation, missing accessories, surface scratches and coating damage; recheck overall dimension and mounting hole position to ensure compliance with customer drawings and industry standards.

Fix the caster on a horizontal test bench, apply 150% of the rated static load and maintain for 24 consecutive hours. After the test, check the bracket deformation, wheel cracking, axle bending and bearing damage. Qualified standard: no structural fracture, permanent deformation within 1mm, and normal rolling and rotating functions.

Adopt drum type endurance testing machine, simulate actual walking conditions. Set the running speed at 3-5km/h, apply 120% rated dynamic load, and carry out cyclic rolling test on smooth surface and obstacle surface respectively. The obstacle height is set according to caster specifications (2%-3% of wheel diameter). After completing 50,000-100,000 fixed cycles, the caster shall have no functional failure, and the wheel wear loss shall not exceed 5% of the original tread thickness.

Place the loaded caster on a flat clean metal platform, use a high-precision force gauge to test the horizontal thrust required for linear rolling and swivel steering. The rolling resistance value shall not exceed the upper limit specified by the corresponding caster grade to ensure flexible pushing and pulling of trolleys and equipment.

Fix the finished caster, drop a standard weight from a height of 200mm±3mm to impact the wheel edge. After the impact test, all components shall not be separated, and the rolling, swiveling and braking functions remain intact, which is applicable to heavy-duty and logistics casters.

Carry out neutral salt spray test for metal structural parts of casters according to ISO 9227 standard. The test duration for conventional casters is 48 hours, and that for outdoor and medical casters is 96 hours. Qualified standard: no obvious rust and corrosion failure on the surface, and the structural performance is not affected.

For anti-static and conductive casters used in electronic workshops and clean rooms, place the caster on a grounded metal plate, apply 10% rated load, and test the overall resistance value. The resistance shall be within the range specified by anti-static grade to meet the static elimination demand.

• Critical Defect: Functional failure including load fracture, brake failure, wheel cracking, which directly causes product scrapping.
• Major Defect: Excessive dimensional deviation, excessive rolling resistance and serious surface corrosion, which affect normal use and need to be reworked.
• Minor Defect: Slight surface scratches and tiny color difference, which do not affect structural performance and can be delivered after simple repair.

If the defective rate of sampled products exceeds AQL 1.0 standard, suspend the delivery of the whole batch of products, conduct full AI re-inspection, and optimize corresponding production processes according to defect data feedback to avoid repeated quality problems.

All inspection images, performance test data and defect records are automatically stored in the production quality database, associated with production batch, working procedure and operator information. The system supports AI big data analysis, summarizes frequent defect types and failure causes, provides data support for production process optimization, and realizes full-cycle quality traceability of caster products from raw materials to finished products.