Fiber vs. CO2 Laser Cutting: How to Choose the Right Machine
Choosing between a fiber laser and a CO2 laser is not simply a matter of which machine is newer, faster,…
Rough striations, bottom dross, incomplete penetration, burned corners, and dimensional errors often prompt operators to change settings immediately: increase the gas pressure, slow the machine, add power, or adjust the focus.
The problem is that the same visible defect can have several opposing causes. Roughness may result from insufficient gas flow, but excessive or turbulent flow can also create a rough edge. Slowing the cut may restore penetration, yet if the speed is already too low, it can produce heavy slag and an oversized heat-affected zone. Soft dross may result from excessive heat, poor melt evacuation, or an incorrect focus position.
The edge confirms that the process is wrong, but it rarely identifies a single setting on its own.
Fast troubleshooting therefore relies on evidence rather than random tuning. Preserve the original condition, identify the defect pattern, inspect the physical system, and change one variable at a time. This approach identifies the cause more quickly and prevents a setting change from masking a hardware, material, gas, or motion problem.
A stable laser cut requires several systems to work together. The beam must deliver the correct amount of energy into the material, while machine motion carries that energy along the programmed path. Assist gas must enter the kerf—the narrow slot created by the laser—and remove molten material. Material condition, focus geometry, nozzle condition, optics, and height control all affect this interaction. For operations looking to coordinate these variables within a CNC-based workflow, ADH Machine Tool’s single-table fiber laser cutting machine provides a relevant next step toward more controlled implementation.
Changing several settings at once destroys the evidence. Additional power may partly compensate for a dirty protective window. Lower speed may force a cut through warped material. Higher pressure may drive more gas through a damaged nozzle while making the jet more turbulent. The next sample may look better, but the original fault remains.
A useful test must either eliminate or support one cause. This requires three things: an intact baseline, a clearly described defect, and one controlled change.
Keep the failed part or a clearly marked section of it. Photograph the top, bottom, and cut edge. Mark the cutting direction, pierce point, machine orientation, and the side that faced the nozzle.
Describe the pattern precisely. Instead of writing “bad edge,” record observations such as “hard dross along the lower third of every straight contour,” “roughness only on the machine-right wall,” or “burning begins when the head decelerates into corners.” These details determine the next test.
Create a baseline record for the defective cut, then compare it with the most similar verified successful job. Record:
For concrete machine specifications to support this comparison, consult ADH Machine Tool’s downloadable technical resources, covering its CNC-based solutions for advanced laser-cutting and sheet-metal applications.
Do not compare unrelated jobs. A successful cut using a different thickness, coating, gas, nozzle, or program is not a reliable baseline. If no verified record exists, label the setup as unverified rather than reconstructing it from memory.
A sudden decline suggests a discrete event, such as a nozzle collision, a contaminated protective window, a change in gas supply, a loose component, a program revision, or an incorrect setup entry. Gradual deterioration points more strongly to contamination, consumable wear, declining gas delivery, or mechanical drift.
If the problem began immediately after maintenance, inspect everything that was disturbed before changing the recipe. Check the nozzle installation, optic seating, focus calibration, gas connections, height-sensing components, and restored parameters.
If the problem began with a new material batch, verify the actual thickness, flatness, coating, oxidation, film, and contamination. If the old material still cuts correctly using the saved baseline, investigate the new sheet before retuning the machine.
The key question is not simply, “When did the cut become bad?” but rather, “What changed immediately before the evidence changed?”
Some quality problems can be addressed through a controlled test sequence. Others require an immediate stop.
Stop cutting after a head crash or if there is a suspected gas leak, a loose or damaged nozzle, a cracked or burned optic, an abnormal sound, a sustained flame, unstable height control, a cooling fault, an open safety interlock, or an inhibited beam. Follow the machine manufacturer’s shutdown and isolation procedures, as well as site safety procedures.
These conditions make process tuning invalid. A collision can disrupt head geometry or beam-to-nozzle alignment. A loose nozzle changes the standoff and gas direction. A damaged optic can distort the beam and absorb excess heat. A leak means the displayed pressure may not reflect the gas reaching the kerf. Unstable height control continuously changes focus and gas coupling during the cut.
Do not compensate for suspected damage by increasing power or reducing speed. Additional energy may temporarily force the cut through while accelerating optic damage, widening the heat-affected zone, or concealing a delivery problem. Resume test cutting only after a qualified inspection confirms that the head, optics, nozzle, gas path, cooling system, interlocks, and height control are sound.
Once urgent faults have been ruled out, organize the possible causes into five groups: material, beam delivery, focus geometry, assist gas, and process parameters. Inspect them in that order so that one variable does not mask another.
Confirm that the sheet matches the program and recipe. Check the material grade, measured thickness, coating, rust, scale, oil, protective film, contamination, and batch identity.
Flatness is especially important. A bowed or shifting sheet changes the nozzle standoff as the head moves. This affects both the focal relationship and how the gas enters the kerf. A defect that follows a raised area of the sheet indicates something different from one that repeats identically on every part.
Where possible, cut the same controlled feature on a verified flat sample or on material from a previously successful batch. If the defect follows the material rather than the machine position or geometry, material condition moves higher on the list of suspected causes.
Inspect the nozzle using the approved procedure. It should be clean, round, secure, correctly sized, and centered on the beam. Spatter, impact damage, a partially blocked opening, an incorrect diameter, or improper seating can distort the gas jet even when the pressure display appears normal.
Inspect operator-accessible protective optics using approved lighting and cleaning methods. Haze, dirt, spatter, or damage can reduce the delivered energy or distort its distribution. Do not touch optical surfaces with bare hands, use improvised cleaning materials, open a sealed beam path, or perform unauthorized alignment.
A defect that consistently favors one kerf wall or varies with travel direction is valuable evidence of asymmetry. It may indicate nozzle damage, poor centering, uneven gas flow, beam delivery issues, head geometry, or directional motion rather than a general power problem.
Focus position describes where the narrowest part of the beam is located relative to the material. Nozzle standoff is the physical gap between the nozzle and the sheet. They are separate variables, but both determine how energy and gas act through the material thickness.
An incorrect focus can leave too little useful energy at the bottom of the cut. An incorrect or unstable gap can weaken gas entry, disrupt the jet, and change the effective geometry. A commanded value on the control does not prove that the physical condition is correct.
Confirm proper nozzle installation, sheet flatness, sensor cleanliness, stable height tracking, and calibration. Then use the approved focus-test routine rather than judging from the production edge. Keep the material, gas, nozzle, power, speed, standoff, and test geometry fixed while changing only the focus within the documented range.
Evaluate penetration, taper, top and bottom edge quality, and dross—not just the top kerf width. If the best result lies far outside the established range, investigate calibration, optics, material, or height control rather than adopting an extreme focus value as the new recipe.
Assist gas is more than a pressure reading. Gas type affects the cutting reaction, purity affects consistency, and effective flow removes molten material through the kerf.
A normal upstream gauge reading cannot rule out a leak, restriction, weak regulator, delayed valve, incorrect nozzle, blocked opening, poor centering, or excessive gap. Static pressure when no gas is flowing is especially weak evidence. Observe pressure under demand and use approved diagnostics where available.
Confirm that the gas type, purity information, nozzle diameter, nozzle type, standoff, line condition, and pressure form a matched system. If increasing the pressure does not improve the edge or makes it rougher, return to the baseline. Inspect for restrictions, leaks, turbulence, nozzle damage, misalignment, and regulation issues rather than continuing to increase pressure.
Process tuning should become the primary approach only after the material, physical condition, focus geometry, and gas delivery have been verified as credible.
Power and speed together determine energy per unit length, but they do not offer a universal correction. Corners, small contours, acceleration changes, lead-ins, and pierces affect local heat input. A recipe that works correctly on long straight sections may fail only when the machine slows down or changes operating mode.
Base the first adjustment on the defect evidence. State the prediction before cutting. For example: “If insufficient energy is the cause, a small reduction in speed should improve bottom penetration without producing a wider, overheated upper kerf.” Change one variable, label the sample, compare it with the baseline, and restore rejected settings to their original values.
The term “rough” has little diagnostic value. Roughness along the bottom of every contour indicates a depth or melt-evacuation problem. A scarred machine-right wall with a clean opposite wall indicates an asymmetry in the nozzle, gas jet, beam, head, or motion system.
A useful observation has three coordinates: its location on the sheet, its location through the cut edge, and its location in the toolpath.
A defect that recurs around nearly every contour points to a condition shared across the job: settings, focus, gas delivery, optics, material specification, or nozzle condition. Inspect the common physical path before modifying the recipe.
A position-specific defect points to factors that vary with machine location. Check sheet bow, local support, dirty or damaged slats, debris, reflected heat, extraction, gantry behavior, and height sensing.
Cut the same small geometry in both an affected and an unaffected area without changing the recipe. If the defect remains tied to the sheet location, investigate local flatness and support. If it remains at the same machine coordinates after relocating the material or geometry, investigate the bed, height control, alignment, or motion system.
Top-edge defects often indicate issues with piercing, focus, nozzle height, contamination, or excessive heat. A crater confined to the contour start isolates the pierce-to-cut transition. A rounded or melted upper edge around the entire contour suggests focus, standoff, surface contamination, or excessive energy. If the defect worsens in corners, check for feature-specific heat accumulation.
Bottom-edge dross or roughness indicates that the process lost control near the exit. Either insufficient useful energy reached the lower region, or molten material formed but was not expelled. Check whether the cut fully penetrated, then inspect nozzle condition, gas delivery, focus, and standoff before adjusting speed or power.
One-sided roughness, taper, or dross indicates asymmetric delivery. Inspect the nozzle for damage, check beam-to-nozzle centering, head verticality, and gas-flow alignment. Compare equivalent walls while keeping the part in machine orientation. A global focus or power adjustment may change the amount of residue, but it cannot correct the asymmetry that caused it.
Corners and small features require the machine to decelerate. The beam then spends more time over a short distance, increasing local heat input. If straight sections are clean but corners burn or melt, confirm that the defect corresponds with deceleration and use approved corner or small-feature controls. Do not retune the entire cut for a localized motion event.
A defect at the contour start suggests piercing completion, gas transition, delay, lead-in length, or start-point placement. A mark at the end suggests lead-out, overlap, pause, or beam-off timing. Moving the start point on an approved test contour can reveal whether the defect follows the toolpath event.
Direction-dependent defects help distinguish motion issues from stationary delivery asymmetry. Cut mirrored or reversed paths without changing the recipe. If the problem reverses with the direction of travel, investigate backlash, vibration, acceleration behavior, and directional gas interaction. If it remains on the same machine-facing side, prioritize the nozzle, beam centering, head geometry, and gas symmetry.
Metal dross is primarily a melt-ejection problem, not simply “burning.” Determine whether sufficient energy reached the full thickness and whether the gas jet could remove the molten material. Adding heat may improve incomplete penetration but worsen a cut that is already molten and poorly evacuated.
Wood, acrylic, and other nonmetals require a different model. Charring, soot, residue, and flame point to dwell time, energy, airflow, extraction, contamination, and combustion. Corner charring suggests deceleration and heat accumulation. Darkening across the entire job suggests focus, optics, surface protection, airflow, or extraction. Any sustained flame requires an immediate safe stop.
First, check machine status, alarms, cooling, interlocks, and beam availability. Then classify the failure.
Failure across the entire cut suggests incorrect material thickness, contaminated optics, improper focus, insufficient delivered power, restricted gas flow, or an unsuitable recipe. A single uncut section suggests warped material, local support interference, an unstable gap, variations in coating or thickness, or a toolpath event. A cut that deteriorates as the machine warms suggests changes in cooling, gas supply, optic contamination, or output stability.
Inspect accessible optics, nozzle condition, focus, standoff, material, and gas delivery before adjusting energy. After these checks pass, test either a small reduction in speed or a power increase within approved limits—but not both. Improvement should provide full penetration without unacceptable kerf widening, heat damage, or dross.
Hard, sharp, tenacious residue often indicates insufficient energy at depth, excessive speed, incorrect focus, or weak evacuation. Verify focus and gas delivery first. If both are correct, a small reduction in speed is a meaningful test.
Soft, rounded beads indicate that a substantial amount of material became fluid. Possible causes include excessive heat, slow travel, or melt that is not being expelled. A wide kerf, rounded top edge, discoloration, and buildup concentrated at corners support excessive heat. A narrow upper kerf with persistent bottom beads supports poor evacuation.
In the first case, test a modest increase in speed or reduction in power. In the second, leave the heat settings unchanged and inspect gas flow, nozzle geometry, centering, and standoff. With one-sided dross, prioritize alignment and gas symmetry over another speed adjustment.
Uniform, repeated striations suggest a consistent mismatch in focus, process settings, or gas flow. Compare their depth and angle from the top to the bottom. A controlled focus series is usually more informative than several arbitrary speed adjustments.
Irregular waves suggest instability, such as vibration, sheet movement, unstable height control, nozzle contamination, turbulent gas, or degraded beam delivery. Where permitted, perform a safe beam-off motion observation, inspect the nozzle and sheet, and watch for abrupt height corrections.
Striations that worsen toward the bottom indicate declining beam effectiveness or melt evacuation through the material thickness. Check focus, standoff, the nozzle, and gas delivery before reducing speed. If a slower test widens and overheats the kerf without improving the lower wall, return to the baseline and investigate evacuation or focus again.
Burning along the full path suggests excessive energy per unit length, unsuitable gas, contamination, or poor extraction. Confirm the actual material and gas, remove any unexpected film or residue according to procedure, and verify airflow. Then test either an increase in speed or a reduction in power.
Burned corners and small details indicate insufficient corner compensation or excessive dwell. Use the controls intended for those features rather than changing a recipe that produces clean straight cuts.
Localized marks require a localized comparison. Check the protective film, dirty support slats, reflected heat, smoke buildup, sheet contamination, pauses, overlaps, and pierce locations. If the mark moves with the sheet, investigate the material. If it remains in the same position relative to the bed, investigate the supports and extraction. If it follows the toolpath start, investigate timing.
Use a controlled focus test to distinguish a focal-position error from general overburning. A focus error changes the top-to-bottom balance and taper. General overburning causes broad melting, rounded edges, discoloration, and excessive heat.
Directional bevel suggests variations in standoff, sheet angle, nozzle damage, poor centering, head alignment, or asymmetric gas flow. Measure corresponding edges in opposite orientations before changing the global energy settings.
When the edge is clean and square but the part dimensions are incorrect, inspect kerf compensation, toolpath offset, axis calibration, backlash, sheet movement, and thermal distortion. Do not compromise an acceptable edge to solve a coordinate or programming problem.
Freeze and save the original recipe. Then follow the same sequence every time.
Treat a brief inspection as triage, not as permission to rush. Confirm that the nozzle is clean, round, secure, correctly sized, and centered. Inspect accessible optics, the extraction path, material identity, flatness, coating, and the condition of the supports.
If a common physical fault is corrected, restore the saved recipe and repeat the original cut before making any other changes. A clean result after replacing a damaged nozzle provides useful evidence. A clean result after replacing the nozzle, cleaning the optic, shifting focus, and reducing speed is inconclusive.
Use the manufacturer-approved baseline focal range for the machine, optic, material, thickness, gas, and cutting mode. Run a labeled focus series while keeping all other conditions fixed.
Observe the head as it moves across the test area. Warped stock, dirty sensing components, damaged insulation, loose parts, or unstable capacitive control can change the gap during travel even if a static check passes. Correct these conditions before accepting the focus-test results.
Confirm the gas type, purity information, pressure under flow, nozzle type and diameter, nozzle gap, hoses, filters, regulators, valves, and accessible connections. Use only approved leak-checking methods and safety procedures.
Displayed pressure is not equivalent to stable flow at the kerf. A pressure drop during piercing or cutting may explain a defect associated with that event. Higher pressure will not correct a damaged, blocked, oversized, or misaligned nozzle and may increase turbulence.
After the preceding checks pass, select the variable that best matches the evidence:
Do not change paired variables at the same time. Record the original value, test value, prediction, penetration, top- and bottom-edge quality, dross, taper, discoloration, dimensions, and keep-or-reject decision. Return each rejected variable to its baseline before the next test.
One clean coupon proves only that the process worked once under a single narrow condition. A production correction must withstand the conditions that exposed the original failure.
Define acceptance criteria before validation. Repeat the test at the original failure location and at representative positions on the bed. Rotate or reverse the toolpath where practical. Recheck long straight sections, corners, small details, pierces, and the feature that originally failed. Then run a controlled first-off part using the actual program or nest.
For teams translating validated first-off results into a production-ready CNC workflow, ADH Machine Tool’s double-table fiber laser cutting machine provides a practical next step within its laser cutting and sheet metal automation portfolio.
A correction that requires unusually low speed, excessive power, or another extreme value should be treated as compensation rather than restored process capability. It lacks margin and may fail again with normal variations in flatness, temperature, surface condition, or gas delivery.
Record the successful material specification, batch, nozzle, gas, pressure under flow, focus, standoff, power, speed, feature controls, program revision, bed positions, maintenance condition, photographs, and acceptance results. Retain the previously approved recipe for traceability.
Escalate when basic checks pass, standard settings cannot produce stable results within approved limits, and the defect continues to track a machine coordinate, maintenance event, or unstable input.
A defect that remains at the same machine coordinate after parts are relocated or rotated suggests problems with bed support, height control, motion, alignment, or beam delivery. Do not create location-specific speed or focus offsets unless an approved manufacturer process explicitly requires them.
A sudden decline after a crash, nozzle change, optic replacement, or maintenance task suggests installation damage, misalignment, contamination, or an incorrect component. Inspect only within the operator’s authority, and preserve removed components when practical.
Unstable gas pressure under demand, recurring optic contamination, inconsistent height readings, deteriorating beam quality, or unexplained degradation during warm-up requires qualified maintenance or service after accessible checks have been completed.
If the next step is to speak with the team directly, contact us fits naturally here.
Also recognize process limits. If the system is stable and repeatable but cannot meet the required targets for roughness, dross, taper, heat effect, dimensions, and production rate within approved settings, the requirement may exceed the capabilities of the material, thickness, laser source, or available gas. The practical solution may be secondary finishing, a different gas, a different material condition, another machine, or a different cutting process.
A useful service report defines the fault boundary rather than merely describing the symptom. “Lower-edge dross recurs at the same machine coordinates in both toolpath orientations while gas pressure remains stable” is far more actionable than “the cut is rough.”
Provide labeled photographs and, where practical, retain both failed and acceptable samples. Include the material identity, program revision, feature type, direction, machine coordinates, nozzle, gas, focus, standoff, power, speed, test results, exact alarms, and any recent maintenance or collisions.
State which inspections were completed and what each inspection found. List every controlled adjustment, including its original value, test value, and result. Identify anything that was not inspected because it fell outside the operator’s authority.
The central rule is simple: a valid recipe follows the part through representative geometry, directions, and bed positions. A fault remains tied to a location, event, asymmetry, or unstable input. Once the evidence establishes that boundary, another random setting change is no longer troubleshooting—it is tuning around the problem.
