Breadcrumb: Home > Applications > Fluorescent Leak Testing
Updated:2026-11-14
Views: Fluorescence leak detection succeeds or fails on a handful of physical quantities: whether the excitation wavelength matches, whether dye concentration sits inside its window, whether irradiance at the read-out distance holds above 1000 µW/cm², and whether enough time has passed. This article works through six layers - spectrum, fluorescence efficiency, sensitivity, tracer chemistry, dosage worked examples, read-out and recheck, cleanliness and safety - and closes with a quick-reference table of 18 field questions that can serve as a parameter sheet before selection and before a job starts. Read-out practice follows ASTM E3022, ISO 3059 and ISO 4406.
Explaining fluorescence leak detection in a complicated way is easy; explaining it clearly has just one route: follow the chain of photon, dye and eye. If any link in that chain fails, the field symptom is the same - no fluorescence, or fluorescence that is not real. This article follows the chain in order, with parameters taken from the LUYOR and Spectroline ranges we supply, which can be checked against the product pages.
Dyes used for fluorescent leak detection have absorption bands concentrated at the long-wave end of UV-A. For the oil-soluble LUYOR-6100, for example, the excitation peak is about 494nm and the emission peak about 535nm; water-based products work the same way, with the carrier as the difference. For efficient excitation the source energy has to land inside the absorption band while not pushing out visible light at the same time - visible light raises background brightness and reduces fluorescence contrast.
| Metric | 365nm professional leak-detection lamp | 395nm violet flashlight |
|---|---|---|
| Spectral position | Long-wave UV-A, main peak 360–370nm | Close to the visible violet edge, significant visible component |
| Filter | UV filter that blocks visible light above 400nm | Mostly unfiltered, or simply tinted |
| Background brightness | Low; strong fluorescence contrast in a dark environment | High; violet light "drowns" weak fluorescence |
| Use | Pinpointing seepage, read-out and acceptance | Coarse screening, non-critical situations |
So "looks more violet and brighter" does not mean stronger. To judge whether a lamp can be used for leak detection, check three things: whether the main peak is at 365nm, whether it carries a filter, and whether irradiance is up to standard. ASTM E3022 gives measurement methods for the emission peak and visible leakage of LED UV-A lamps, and that convention can be used for comparison; a more practical field measure is irradiance - whether the level at the read-out distance stays reliably above 1000 µW/cm².
Dye fluorescence intensity is not a monotonic function of concentration. As concentration rises from zero, fluorescence brightens; past a certain point, intermolecular self-absorption and energy transfer take over, fluorescence dims and the color washes out. That is concentration quenching.
In the field this shows up as two opposite misreadings: too little dye gives no glow or a faint one, read as "no leak"; too much makes the whole area pale and bright, so the boundary of the leak is lost, read as "a bad leak". Either way the conclusion is distorted.
The detection limit of the fluorescence method is set by three quantities together: dye concentration in the seepage, the volume that seeps out, and the thickness of the residue left on the surface. A workable rule of thumb is that once the seepage surface accumulates to the microliter level, it can be read under reasonable irradiance. Converted to specific systems:
| System | Typical detectable magnitude | Notes |
|---|---|---|
| Refrigeration / air conditioning | Slow seepage of a few grams of refrigerant per year | Refrigerant carries PAG/POE oil, and the dye seeps out with the oil and deposits on the leak surface |
| Hydraulics and lubrication | Weeping points that have not yet formed a drop | A thin film on the surface is enough to register, weeks before it becomes visible to the eye |
| Water systems | Slow weeping, damp traces | Water-based tracer seeps out with the water and leaves a fluorescent trace after drying |
What needs emphasis is the role of time: the same 0.01 ml/h leak may show nothing after 10 minutes of illumination but be obvious after 8 hours of accumulation. A leak-detection plan therefore has to state how long passes between dosing and read-out, otherwise "found nothing" is just "looked too early".
Chemically, LUYOR-6100 is a solvent-borne concentrated dye - Solvent Red 175 dissolved in a highly active hydrogenated naphthenic solvent system. That sets three properties: fully miscible with petroleum-based oils, not miscible with water, and able to remain in the system long term without precipitating (flash point above 85℃, not classified as a hazardous or flammable chemical).
| Dimension | Oil-soluble (6100) | Water-based (6200/6300) |
|---|---|---|
| Carrier | Solvent-borne, miscible with mineral oil | Water-soluble, miscible with water in any ratio |
| Media to avoid | Water systems, water-containing hydraulic fluids (HFA/HFB and similar) | Oil systems, drinking water and product-contact systems |
| Fluorescent color | Yellow-white under UV; the oil takes an orange-red tint in daylight | 6200 green / 6300 blue |
| Residue handling | No flushing needed, stays with the oil; surface residue wiped off with a dedicated cleaner | Rinses clean with water |
| Material compatibility | The solvent carrier carries some swelling risk for certain elastomers; evaluate old seals on a small sample beforehand | Friendly to metals and common plastics, though long soaking at high concentration still needs evaluation |
Two further boundaries are easily overlooked: 1. do not mix across media - oil-soluble dye added to a water tank is not "slightly effective", it is completely ineffective and contaminates the system; 2. do not use it in food or pharmaceutical contact chains - even water-based products are for industrial water systems and must not enter drinking water or product contact surfaces.
Example one: a hydraulic station. Tank volume 200L, at 0.5–2.5 ml/L the dose is 100–500ml; 250ml in the middle of that range will do. Run the system for 10 minutes after dosing so the dye circulates fully, then illuminate point by point. For a repeat check at fixed points, re-illuminate the same system without topping up.
Example two: a cooling water system. A 3000L body of water at 1000:1 needs about 3L of concentrate - beyond a single bottle, so in practice make up 500ml as a stock solution and inject it zone by zone, or use the economical 6300 route (500ml covers about 3000L) and test in sections. One more rule for water systems: a static system diffuses slowly, so concentration is high near the injection point and low at the far end; allow for the gradient when reading out and sample at several points if needed.
| System type | Dosing | Circulation / penetration time | When to read out |
|---|---|---|---|
| Running oil system | 0.5–2.5 ml/L | 5–10 minutes | Can be checked the same shift |
| Static oil system | As above | 2–3 hours | An overnight recheck is more reliable |
| Circulating water system | From 1000:1 | Circulate 30 minutes or more | Read out once circulation is stable |
| Static water body | As above, injected by zone | Hours to a day | Allow for the concentration gradient |
| Container pressure test | Can be diluted to 5000:1 | Hold pressure 30 minutes or more | Scan along welds while under pressure |
Read-out conditions can borrow established NDT practice: UV-A irradiance on the read-out surface not below 1000 µW/cm², ambient white light held below 20 lx (ISO 3059 / GB/T 5097). In engineering terms that means three actions: switch off workstation lights, hang a light-blocking screen, and give the eyes 5–10 minutes to dark-adapt.
False positives are the biggest enemy at the read-out stage. Common sources and how to rule them out:
| Source | Identifying feature | Ruling out |
|---|---|---|
| Residue from an earlier check | Fixed location, sharp boundary, matches the previous photograph | Wipe clean and re-illuminate; see whether it reappears in the same place |
| Assembly or rust-preventive oil | Broad patches, blurred boundary, unrelated to seam direction | Clean, run under pressure, and watch whether the seam shows again |
| Autofluorescence of rubber and plastic | The whole part glows, outline follows the part | Change the angle; autofluorescence does not extend along the leak path |
| Cleaner residue | Blue-white haze, widely distributed | Wipe a second time and look again |
| Ambient reflection | Moves when the lamp moves | Move the lamp; true fluorescence stays where it is |
The most effective read-out sequence is three words: wipe, illuminate, look again. A leak point is accepted when seepage reappears at the same spot after wiping and follows the direction of media flow.
Whether a recheck conclusion can be trusted depends on whether old traces were cleaned off properly. Acceptance, service and warranty work should have cleaners on hand: for oil-phase systems a dedicated fluorescent cleaner wipes off surface residue; water-based products rinse clean with water.
Another point often missed is system cleanliness. Leak detection itself can carry external dust into the system (especially with parts opened or fill ports uncovered), and particulate contamination of the oil directly affects wear rate. Comparing cleanliness codes to ISO 4406 before and after shows whether the job introduced new contamination. For hydraulic and servo systems with tight cleanliness requirements, fill ports and cleaning actions should be handled to cleanroom standards.
| Misconception | Why it is wrong | Correct practice |
|---|---|---|
| Using a banknote marker or a pet violet flashlight instead of a leak-detection lamp | No UV filter, visible light lifts the background and weak fluorescence is swamped | A 365nm lamp with a filter and adequate irradiance on the read-out surface |
| Adding more dye when fluorescence is faint | Once concentration enters the quenching zone, more dye means dimmer, paler results | Oil 0.5–2.5 ml/L; water from 1000:1 |
| Concluding right after a quick sweep | Slow seepage accumulates over time; nothing visible in ten minutes does not mean no leak | Run 5–10 minutes for the initial check, look again the next day for slow seepage |
| Reading out under bright workstation lighting | Ambient white light directly flattens fluorescence contrast | Lights off, light blocked, 5–10 minutes of dark adaptation |
| Putting oil-soluble dye into a water tank | Immiscible, completely ineffective and it contaminates the system | Match the medium, then pick the color |
| Judging leak rate by fluorescence brightness | Brightness is governed by concentration, residue thickness and irradiance together | Use weighing, flow or pressure-decay methods for quantitative work |
| Re-illuminating after a repair without wiping old traces | Old residue is read as "not fixed yet" or "already fixed" | Wipe clean, then circulate under pressure, then re-illuminate |
| Using fluorescence to check compressed air | A gas has no liquid medium to carry the dye | Use ultrasound for gas, with flow-rate accounting |
One more variable easily overlooked is temperature: at low temperature oil viscosity rises, dye circulates more slowly, and the read-out window has to be stretched; at high temperature the solvent evaporates faster, residue on the surface is thinner and fluorescence fainter, so a shorter read-out distance and higher irradiance are needed. The same instrument working in summer and in winter does not have the same read-out conditions, and that is normal.
| # | Question | Key point |
|---|---|---|
| 1 | How to choose between 365 and 395 | 365 for pinpointing small leaks; 395 is suited to coarse screening |
| 2 | Can an ordinary violet flashlight detect leaks | No: no filter, inadequate irradiance, stray light swamps fluorescence |
| 3 | Must dye and lamp be the same brand | Not necessarily, but the absorption band has to match 365nm and irradiance must be adequate |
| 4 | What is the principle | Tracer seeps out with the medium and emits visible fluorescence under 365nm excitation |
| 5 | How small a leak can it find | Seepage accumulating to the microliter level on the surface; refrigerant down to a few grams a year |
| 6 | How soon after dosing can I look | Run 5–10 minutes; static 2–3 hours; slow seepage overnight |
| 7 | Can it be done in daylight | Yes, but shade is preferable; blocking light works better than a brighter lamp |
| 8 | Does any fluorescence mean a new leak | Not necessarily; wipe, then illuminate, and see whether it reappears in place |
| 9 | How is a repair accepted | Wipe off old traces, circulate under pressure, re-illuminate with no new fluorescence |
| 10 | Can the dye stay in the system long term | Industrial-grade dye can, without changing medium performance |
| 11 | Does the UV lamp harm the eyes | Long direct viewing of UV-A carries risk; wear protective glasses throughout |
| 12 | Is the leak-detection dye toxic | Safe when used as specified; avoid ingestion and prolonged skin contact |
| 13 | Dosed, but no fluorescence appears | Work through medium mismatch, insufficient wait, ambient light, lamp, concentration, obstruction |
| 14 | What if fluorescence is faint | Raise irradiance, darken the background, check concentration and shelf life |
| 15 | How does it divide work with bubble testing, sniffing and helium mass spectrometry | Fluorescence locates, sniffing confirms, helium does the limit test |
| 16 | Which applications suit it | Any system with a closed circulating liquid; not suitable for gases |
| 17 | How is the lamp maintained | Keep the filter free of oil and scratches; avoid deep discharge of the lithium battery |
| 18 | Is cleaner mandatory | Yes for acceptance and recheck; on demand for routine self-checks |
Standards and methods (check the original text by number; versions follow the issuing body's current edition)
Check four conditions: read-out distance and irradiance, ambient white light, dark-adaptation time, and whether the dye has actually circulated to that spot. If any one of the four is a grade off, weak fluorescence disappears. Then check whether the lamp-head filter is oily.
Industrial-grade water-based tracer dosed at the recommended ratio is for industrial circulating and cooling water, not for drinking water or product contact surfaces. After the job, drain it on the system's normal water-change cycle; for systems with strict water-quality requirements, confirm acceptability beforehand.
Purely semi-quantitative. Brightness is affected by concentration, residue thickness and irradiance together and cannot be converted directly to leak rate. For quantitative work, pair it with weighing, flow or pressure-decay methods.
It depends on circulation and operating conditions. With the compressor running and the system circulating under pressure, fluorescence is visible within 30 minutes to a few hours; very slow seepage may need 1–2 days of running. Intermittent leaks have to be retested under the fault condition.
The dye changes oil color, and spectroscopic tests such as infrared or colorimetric methods can be affected after dosing. If oil analysis is due soon, dose after sampling, or record the dosing time for the laboratory's reference.
Further reading: How to make early leak detection effective: cost accounting, read-out thresholds and inspection routines, How to trace industrial seepage, Leak troubleshooting in paper mill equipment, Preventive maintenance and inspection plans for hydraulic systems.
This article belongs to the fluorescence leak detection techniques section.
Other sections on the same product line include fluorescent leak detection (add a tracer, then find the leak point with a UV lamp - locate before you dismantle) and industrial test instruments (the entry point for UV inspection, speed measurement, stroboscopic observation and surface inspection).
Models mentioned in this article: LUYOR-6100, LUYOR-6200, LUYOR-6802. Ranges, accuracy and accessory lists for each model are given on the product pages.
Email: info@xinysun.com
Tel: 131-2093-6518
Address: 1525 Minqiang Road, Bldg 26, Songjiang, Shanghai
Copyright 2026 Xinysun Instrument (Shanghai) Co., Ltd. All rights reserved.
ICP:ICP-2026046447
