If a laser tube cut starts clean and then becomes rough after about 30 minutes, I check the protective window and focus stability first. Heat, vapor, spatter, or contamination can shift the focus position during a sustained run, so the same program that cut well at 9:00 can leave dross, taper, or rough edges at 9:30. The fastest fix is to stop guessing: inspect the protective glass, nozzle, assist gas, cooling, and focus, then run one controlled test cut and judge the edge by ISO 9013 terms such as Rz5 and perpendicularity u.
What does this problem look like, and what does it cost to ignore?
This problem usually looks like a clean first batch followed by rougher edges, bottom dross, color change, taper, or unstable piercing after the machine runs under load. The cost is not only the bad tube. You also pay for rework, deburring labor, lost schedule, and operator time spent adjusting settings without a clear cause.
On our factory floor in Foshan, I usually ask the operator one question first: “Did the first parts pass, and did the later parts fail with the same program?” If the answer is yes, I do not blame the drawing first. I look for something that changes with time.
A 30-minute pattern is not a standard threshold. It is a symptom pattern. It tells me the process may degrade after heat, vapor, spatter, or gas demand builds up.
Look at the edge, not only the surface appearance. ISO 9013:2017 classifies thermal cuts, including laser cuts from 0.5 mm to 32 mm, by geometrical product specification and quality tolerances. In practical shop language, that means you should separate these problems:
- Roughness: the cut face feels or measures rougher, often tracked by mean height of profile
Rz5. - Taper or angle error: the wall is not square, linked to perpendicularity or angularity tolerance
u. - Bottom dross: re-solidified molten metal sticks to the lower edge.
- Oxidation or color: stainless turns yellow, blue, gray, or rough when shielding or nitrogen cutting is unstable.
- Pierce instability: the machine starts punching through slower or with more spatter after a sustained run.
If you keep running, the machine may still make parts, but your floor pays twice. Your operator cuts once, then another worker grinds, sorts, or rejects. That is exactly the labor cost tube lasers should remove.
What is actually causing the cut to get rough after 30 minutes?
The most likely causes are focus drift, protective-window contamination, nozzle or gas-flow degradation, assist-gas instability, and cooling instability. All five can start small and become visible only after sustained cutting.
| Symptom / observation | Likely cause | How to confirm it |
|---|---|---|
| Cut starts clean, then edge gets rough on the same tube and program | Time-dependent focus drift | Run a focus check before and after the fault appears; compare edge roughness and perpendicularity |
| Roughness appears after smoke, vapor, or spatter builds near the head | Protective-window contamination | Remove and inspect the protective glass; replace it and repeat the same cut |
| Bottom dross grows during the run | Nozzle damage, poor centering, or gas-flow degradation | Inspect nozzle roundness, clean spatter, check nozzle-to-beam alignment, and repeat |
| Stainless edge turns colored, gray, or rough | Nitrogen flow, purity, leakage, or turbulence problem | Check gas supply stability, line condition, nozzle condition, and oxygen exposure at the cut |
| Piercing becomes unstable after continuous work | Cooling instability or head thermal load | Check chiller alarms, temperature trend, laser source status, and head condition |
| Only one batch of tube fails | Material batch, coating, seam, scale, or wall variation | Cut the earlier good tube and the later bad tube with the same program |
| The defect changes with part geometry | Program path, heat concentration, or local dwell | Compare straight cuts, holes, slots, and tight patterns separately |
Protective glass deserves special attention because it can fool a busy floor. A contaminated protection glass can absorb more energy, heat up, change optical behavior, and shift the focus position. A ScienceDirect paper on industrial laser beam processing describes this focus-shift mechanism in contaminated protection glass. Applied Optics also notes that protective glass contamination can reduce transmissivity and affect production quality, and visual inspection may miss it.
Focus shift also has direct support in laser cutting research. A Karlsruhe Institute of Technology publication studied thermally induced focus shift and evaluated its effect on 3 mm stainless steel cut edges using roughness and perpendicularity tolerance according to ISO quality measurement.
I do not start by changing ten parameters. If the same program made a good part earlier, I first ask what changed during the run: optics temperature, protective glass condition, nozzle condition, gas flow, cooling, or material.
How do you fix each cause?
Fix the cheapest and fastest checks first: protective window, nozzle, alignment, gas supply, cooling, then focus and cutting parameters. If you change speed, power, focus, gas pressure, and nozzle all at once, you may hide the real problem and make it return next week.
Stop and save one good part and one bad part. Mark the time, material, tube size, wall thickness, program, assist gas, and operator shift. You need a comparison, not a memory.
Inspect and replace the protective window. If the roughness appears after sustained cutting, I treat the protective glass as a wear part, not as a permanent optic. Contamination can shift focus because it changes heat absorption and optical behavior.
Check the nozzle. Look for spatter, oval wear, collision damage, and off-center flow. A small nozzle defect can disturb the gas jet. Assist gas removes molten material from the kerf and shields the cut zone from unwanted oxidation, so poor flow shows up as dross, roughness, and color change.
Re-center the beam and nozzle. Poor coaxial alignment can make one side of the tube edge look worse than the other. BNL Laser offers a Precision Coaxial alignment tool for this job. It brings the nozzle and beam back to true center, and it works with auto-focus and manual-focus heads on tube and sheet machines.
Check assist gas stability. Nitrogen is commonly used for stainless steel or aluminum when you need a bright, oxide-free edge. Oxygen can help on some mild steel cuts, but it leaves an oxidized edge. Do not copy a pressure from another shop. Confirm the gas grade, pressure, nozzle, focus, and speed on your tube during a timed sample cut.
Check compressed air quality if you use air assist or shop air near the head. ISO 8573-1:2010 defines compressed-air purity classes for particles, water, and oil. Dirty or wet air can create defects and damage components around the cutting head.
Check cooling and thermal stability. Do not guess at chiller setpoints without the laser source and head manual for that configuration. But do check alarms, temperature trend, water condition, and any change that appears only after continuous work.
Run a controlled sample cut. Cut the same geometry after each correction. Judge the edge by roughness, perpendicularity, bottom dross, color, and pierce stability. If your customer drawing names ISO 9013, EN ISO 9013, JB/T 10045, or another requirement, measure against that requirement.
Only then adjust process parameters. Dross formation depends on feed rate, gas pressure, focus position, and laser power. Change one item at a time. Record the result.
This is how we handle RFQ samples too. You send tube data, and our engineer replies within one business day. Then we cut a free real sample on your own material with a cycle-time estimate before final configuration and quote.
How do you stop the rough cut from coming back?
You stop repeat rough cutting by turning the checks into a short routine: optics, nozzle, gas, cooling, material, and edge measurement. A laser tube cutter should not depend on one experienced operator remembering every setting by feel.
Use this prevention checklist on your floor:
- Keep a protective-window change record. Record date, material, gas, failure symptom, and operator.
- Inspect the nozzle before long runs. Check spatter and roundness before the first batch, not after scrap appears.
- Verify beam-to-nozzle centering. Do this after nozzle replacement, head collision, or unexplained one-sided dross.
- Track assist gas supply. Watch pressure stability, line leaks, regulator behavior, and gas changeover.
- Control compressed air quality. If air assist or pneumatic systems affect the head area, specify air quality under ISO 8573-1:2010.
- Watch cooling trend during sustained cutting. A problem that appears after heat builds needs time-based data.
- Separate material batches. ASTM A500/A500M structural tubing, ASTM A554 stainless mechanical tubing, and EN 10219 hollow sections can come with different drawing requirements and surface conditions.
- Measure the defect. Use ISO 9013 terms where the drawing requires them, especially
Rz5, perpendicularity/angularityu, and dross observations.
Our Easy-Cut control helps here because the operator can program the part on-screen at the machine with no CAD step and no separate programmer. In many shops, that cuts operator training from weeks to about a single day. But simple control does not remove the need for process discipline. A clean lens, centered nozzle, stable gas, and correct cooling still matter.
I built BNL Laser around this idea: make complex technology simple to run, but do not pretend the process has no rules.
Is it worth repairing, or time to upgrade or replace?
Repair the machine if the fault traces to consumables, alignment, gas, cooling, or maintenance; consider upgrading if the machine cannot hold focus, repeatability, support, or the tube range your orders now require. A rough cut after 30 minutes often comes from fixable process drift, so do not buy a new machine before you test the basics.
Repair or tune the current machine if:
- The first parts are good and later parts degrade.
- A new protective window restores quality.
- Nozzle replacement or beam centering fixes one-sided dross.
- Gas supply changes remove color or roughness.
- Cooling alarms or temperature trend explain the timing.
- The machine still fits your tube diameter, wall thickness, part length, and tolerance needs.
Start looking at replacement if:
- The machine needs constant manual correction to keep acceptable edges.
- Your operator needs CAD support for simple tube jobs and this slows every shift.
- Your tube range has moved beyond the machine’s chuck capacity or cutting envelope.
- You need bevel weld prep, automatic loading and unloading, or structural profile cutting.
- Service support cannot trace the problem back to the head, laser source, controls, gas, or cooling system.
For reference, BNL Laser builds fiber laser tube cutting machines, not CO2 machines, and tube lasers are our whole factory. We design, assemble, and service them in-house in Foshan. Our Q Series covers general tube cutting with 1500–6000 W laser power, repeatability of ±0.05 mm, and reference ranges up to φ12–φ120 mm or φ8–φ165 mm depending on Q12 or Q16 configuration. Our M Series handles small tube from Φ8–Φ85 mm with wall thickness from 0.2–3 mm. Our P-Pro cuts 5-axis bevels from 0–60° on Φ12–Φ300 mm tube with wall thickness from 1–20 mm.
For higher-volume production, our L-PurePro line uses 1500–12000 W with automatic loading and unloading. For structural profiles and constrained layouts, our B-NexBeam side-mounted tube laser cuts square, round, rectangular, oval, I-beam, channel, and angle steel. Standard build is three-phase 380 V 50 Hz, and we confirm country-specific voltage and power in writing during the RFQ.
If you ask us for a quote, I will not start with a catalog model. I will ask for your tube data first. Then my engineer can tell you whether repair, process tuning, or a different machine is the honest answer.
Conclusion
A laser cut that starts clean and gets rough after about 30 minutes usually points to a time-dependent process problem, especially protective-window contamination or focus drift. Check the protective glass, nozzle, coaxial alignment, assist gas, cooling, and material batch before you blame the whole machine.
If you want a grounded answer, send us your tube size, wall thickness, material grade, drawing requirement, and a photo of the good edge and bad edge. We can review it and, when needed, cut a real sample on your material with a cycle-time estimate.
— Eric Liu, General Manager, BNL Laser