How Do You Detect Dry Porous Parts Without Contact?

Dry porous parts can be rejected reliably after the sensing method is matched to the property that changes most consistently when the part is wet. At a conveyor speed of about 4 ft/s, the practical choices are optical color or reflectance sensing, noncontact temperature sensing, a specialist reflected-light moisture analyzer, or a custom extended-range capacitive measurement. Start with optical testing because it offers the simplest noncontact installation when wetting changes color or surface reflectance.

Detection approaches and decision criteria

Before anything else, confirm which measurable property separates acceptable wet parts from dry parts. Moisture itself is not always the easiest production variable to detect. A repeatable change in color, reflectance, temperature, weight, or dielectric constant can provide a more stable pass/fail signal.

Do not select a sensor from nominal range alone. Test whether the wet-to-dry separation remains larger than the combined variation caused by acceptable parts, position, motion, lighting, temperature, and conveyor material.

Recommended inspection architecture

Use a presence sensor to establish the inspection window and a second measurement to classify the detected part. The existing optical missing-piece check should not double as the moisture decision unless its signal provides repeatable wet/dry separation.

1. Retain or add a presence channel that detects every part independently of moisture condition. Confirm one clean presence event for each part across the full allowed position range.
2. Bench-test color or reflectance first. Confirm that the distributions for wet and dry samples remain separated after changing part position and orientation within production limits.
3. If optical separation is weak, measure surface temperature at the intended inspection location. Confirm that the temperature difference persists until every part passes the sensor.
4. If neither proxy is stable, evaluate a specialist reflected-light moisture analyzer over the required standoff.
5. Reserve extended-range capacitance for an engineered installation with controlled geometry, or where an electrode can be placed beneath the conveyor behind a nonconductive window or opening.

This order minimizes mechanical complexity. It also distinguishes a missing part from a present but dry part, preventing one threshold from being asked to solve two different detection problems.

Commissioning prerequisites

Collect production samples before setting any threshold. A calibration made from one wet part and one dry part does not describe normal process variation.

1. Assemble known-dry parts, known-acceptable wet parts, and borderline parts from the real process. Include the normal range of color, texture, dimensions, and retained moisture.
2. Record the closest and farthest surface positions caused by movement. The stated requirement is at least an inch or two of clearance, preferably more; use the worst-case surface distance when evaluating sensing range.
3. Run samples on the actual conveyor at about 4 ft/s. Confirm that vibration, spacing variation, and orientation do not create extra triggers or missed measurement windows.
4. Measure the sensor's usable detection-zone length in the direction of travel. Calculate available observation time as t = L / 4 ft/s when L is expressed in feet, or t = L / 48 in/s when L is expressed in inches.
5. Check that sensor response, controller filtering, classification, and output transfer all complete inside that observation time. Do not move on until the controller records one classification per presence event.
6. Define the acceptable process state from reference measurements. Calibrate air as a no-part baseline where applicable, then establish separate dry, acceptable-wet, and borderline populations.

Optical color and reflectance procedure

Optical sensing is the preferred first trial when wetting changes color or makes the surface more reflective. A controlled light source and receiver can detect a small difference, but the sensing geometry must remain fixed.

1. Mount the emitter and receiver so that normal part movement stays inside the specified sensing field. For diffuse color change, favor a geometry that reduces sensitivity to small angular changes. For wet-surface shine, test the angle that captures the reflectance difference without saturating the receiver.
2. Shield the measurement from changing ambient illumination. Confirm that switching nearby lights on and off does not move a dry sample across the wet threshold.
3. Trigger acquisition from the independent presence sensor. Sample the same area of each part rather than using an uncontrolled continuous signal.
4. Record readings from the dry, wet, and borderline sample groups. Set the threshold inside the measured gap between unacceptable and acceptable populations, not at the average of two individual parts.
5. Repeat the test at the nearest and farthest part positions. Confirm that distance-induced signal variation remains smaller than the wet/dry separation.
6. Run consecutive wet and dry parts at line speed. Confirm correct classification without requiring extra spacing beyond the normal somewhat equal production spacing.

If surface texture creates bright flashes, use the stable portion of the inspection window or combine multiple samples into one part decision. Reject this method when position or finish changes produce the same signal excursion as moisture.

Infrared temperature procedure

Temperature sensing works only when the process produces a persistent thermal difference between wet and dry blocks. Evaporation can cool a wet surface, while recently applied liquid can also impose the liquid's temperature. The direction of the difference must be measured at the inspection point.

1. Measure known-wet and known-dry parts at the exact delay between wetting and inspection used in production. Confirm a nonoverlapping temperature range across representative samples.
2. Select a sensor field of view that falls within the target surface throughout the allowed standoff. Background conveyor temperature must not dominate when a small or displaced part passes.
3. Keep the viewed surface and angle consistent. A wet, shiny surface can change infrared emissivity and make indicated temperature differ even when actual temperature does not.
4. Set the classification threshold from the worst-case wet and dry populations. Confirm the threshold after the process, liquid, and room temperatures reach their normal operating ranges.
5. Challenge the system after a stop and restart. Do not move on until parts held on the conveyor and newly arriving parts are classified correctly.

Use thermal sensing as a moisture proxy, not an absolute moisture measurement. If wet and dry temperature ranges overlap during warm-up or production drift, move to a direct moisture-sensitive method.

Extended-range capacitive measurement

A custom capacitive system measures the effective relative permittivity of air, the porous product, retained water, the conveyor, and nearby structures between or around conductive electrodes. Water-driven dielectric change can separate wet and dry parts, but the installation measures the whole electric-field path rather than the object alone.

1. Choose a repeatable electrode geometry. Opposed conductive plates can place the moving product within the field, while an electrode below the transport system may work through a nonconductive window or opening.
2. Measure the empty-conveyor value as the air and conveyor baseline. Confirm that belt seams, supports, contamination, and vibration do not resemble a product event.
3. Pass known-dry parts through the complete allowed position envelope. Record the minimum and maximum capacitance response rather than one centered reading.
4. Repeat with acceptable wet and borderline parts. Continue only if the wet range remains distinct from the dry range at the required inch-or-two clearance.
5. Use a sensitive capacitance bridge or equivalent measurement circuit to compare the result with the calibrated acceptable range. Transfer a discrete decision to the controller only after the presence channel validates that a part occupies the field.
6. Test buildup and gradual baseline drift. Confirm that empty-conveyor and dry-part values remain on their assigned sides of the thresholds during an extended run.

A standard capacitive proximity switch is a poor starting point for this geometry because its switching range is commonly intended for close target detection. Increasing sensitivity without controlling surrounding metal, belt motion, product position, and contamination creates unstable switching. The longer standoff may therefore require custom electrodes and measurement electronics.

Decision timing and fault isolation

Track each presence event, classification result, and reject command as separate states. At 4 ft/s, a stable sensor value that arrives after the part leaves the inspection zone is not usable.

1. Run an ordered challenge sequence containing known wet, dry, missing, and borderline positions.
2. Verify one presence record for every physical part and no presence record for the intentional gap.
3. Verify one moisture classification is attached to the correct presence record.
4. Measure the elapsed time from detection to valid classification against the observation time calculated from the sensing-zone length.
5. Confirm the reject output follows the dry part rather than the preceding or following part.
6. Repeat at the minimum and maximum expected spacing, surface position, and production-condition extremes. Do not release the station until all challenge parts produce the expected result.

Frequently asked questions