Laser not cutting through: the order I check things in

Not cutting through is the most reported symptom in fiber laser cutting and the least useful one on its own. It tells me what the operator sees. It does not tell me what is wrong. Two machines with the same complaint routinely need opposite fixes: one is starving for gas, the other is drowning in it.

So I stopped asking what the machine is doing and started asking three things instead: what shape the failure has, whether it ever worked, and what changed. Everything below follows from those answers.

Did it ever cut this material?

Ask this before anything else. It splits the diagnosis in half.

• It cut this exact material and thickness last week, and now it does not. Something changed. Consumables, gas supply, optics, alignment, or a mechanical fault. The recipe is probably still correct, so do not touch it yet.
• It has never cut this material cleanly. Then it is not a fault at all, it is a capability or recipe problem: power against thickness, nozzle choice, gas type, focus position. The machine is doing what it was told.

I have watched people spend a full shift rebuilding a parameter table on a machine whose protective window was burnt. The window was a twenty dollar part. The shift was not.

What shape does the failure have?

Four shapes, four different places to look.

• Marks the top surface, no penetration at all. The beam is arriving weak or entirely defocused: gas, optics, focus, or the source.
• Cuts most of the way, small bridges left behind. Usually speed against power, or the gas cannot clear the kerf fast enough at that speed.
• Fine in the middle of the sheet, fails on corners or at the start. That is motion, not process: acceleration, backlash, or height sensing.
• Cuts thin sheet perfectly, cannot manage the same material thicker. It is a capability question. Check the power against thickness honestly before blaming the machine.

Gas, because it is the cheapest thing to check

Pressure at the regulator means almost nothing. What matters is the pressure and flow at the nozzle, at the moment of cutting.

• Oxygen on mild steel is doing the cutting, not just the blowing. Low pressure there produces the exact symptom of too much speed.
• Nitrogen on stainless needs purity as much as pressure. A 99.5% supply will cut, badly, and the edge will be yellow. If the edge is straw coloured on stainless, look at the gas before the optics.
• Shop air is the most under-checked supply on the floor. If the dryer has failed, water goes into the beam path and absorbs energy before it reaches the metal.
• A partially blocked line or a leaking joint downstream of the gauge gives you a good reading and a bad cut.

The nozzle tells you more than the screen does

• A chipped, ovalised, or spattered bore changes the gas flow far more than the gauge suggests. Hold it up to the light. It should be a clean circle.
• Diameter against thickness matters. Too small and the kerf cannot clear; too large and the gas loses velocity before it reaches the cut.
• Centring: tape shot at the nozzle, look at where the mark lands. Off centre means the beam is clipping the nozzle on one side, which costs you power exactly where you need it most.
• Standoff and height following: if the head is riding too high, the beam is out of focus at the workpiece no matter what the focus setting says.

Optics: the protective window is a consumable

Treat it like one.

• A burnt, misted, or coated window scatters energy. The source reports full power, the workpiece receives a fraction of it, and nothing alarms.
• Focus drift is the quiet one. If the machine needs a different focus value every week, that is a symptom, not a setting.
• The collimator and the head interior windows are behind the protective window and get checked last, usually after two lenses have been replaced for nothing.

Corners only? Stop looking at the process

If the failure is geometric rather than uniform, the process is innocent.

• Acceleration set too high for the moving mass: the axis is not where the controller thinks it is, so the effective speed varies along the cut.
• Rack and pinion backlash shows up as an unstable cut at a constant commanded speed.
• On thin sheet, height sensing can lose the sheet during a fast direction change, which looks exactly like a gas problem at the corner.

Only then the source

• Red beam test first. Round and concentric, the delivery path is probably intact. A crescent means the beam is clipped somewhere. A soft blob means it is not being focused at all.
• Compare set power against delivered power. A burn test at a fixed distance, or a power meter if you have one, answers in two minutes what an hour of parameter editing will not.
• Condensation is seasonal and easy to miss. On a hot, humid day, water pipes sweating at the source inlet is a warning. If moisture reaches the modules, output falls and the machine stays alarm-free.
• Back reflection on highly reflective material can degrade a source slowly, over months, with no single event to point at.

What I do not do

I do not change three parameters in one test. If the result is good, you have learnt nothing about which change did it, and next month you cannot reproduce it.

I do not raise power to compensate for a dirty lens. It works for a week and then the lens fails properly, usually mid-job.

And I do not start with the parameter table. It is the easiest thing to edit and the last thing that is usually wrong.

If it still will not cut

Then the next step is not another parameter. Send a video of the head while it cuts, a close-up of the cut edge, and the screen showing the program. Loss of power with a clean alarm log is a different article, and most of the time it is downstream of the source, not inside it.

FAQ

Why is my laser not cutting all the way through?

The most common reasons, in the order I check them: a contaminated protective window or focusing lens, a drifted focus position, assist gas pressure too low for the thickness, or cutting speed set above what the power can support. The full order is in the sections above.

Can bad gas cause the laser not to cut?

Yes, in two ways. Pressure that sags under flow starves the melt of momentum, and purity below spec adds oxides that resist cutting — especially on stainless with nitrogen. A gauge reading at standby does not prove either: check pressure during the cut and confirm the purity certificate.