Flow Rate Calibration: Finding the Magic Number for Every Filament

When this week’s Monday hub opened the topic of slicer tuning for engineering filaments, flow rate calibration is the unassuming step that decides whether well-tuned temperatures and pressure advance actually deliver the expected result. Flow rate, known in PrusaSlicer as extrusion multiplier and in Bambu Studio as flow ratio, determines how much plastic the nozzle actually extrudes compared to the theoretical volume. A poorly set value shows up in the top layer, in dimensional accuracy and in part strength, and worse, it commonly masks other problems. This guide walks through both the visual and precise calibration methods, shows typical value windows for the most-used filaments, and explains when it makes sense to repeat the calibration.

What flow rate is and why the magic number differs for every filament

Flow rate refers to the amount of filament the extruder pushes through the nozzle, usually expressed as a percentage or multiplier of the default value. In slicer terms, this is the extrusion multiplier (PrusaSlicer) or flow ratio (Bambu Studio, OrcaSlicer), while the printer firmware offers a parallel flow parameter that can be overridden at runtime via M221. According to Prusa Research documentation, the total flow rate equals the firmware flow multiplier times the slicer extrusion multiplier, with a change to one not affecting the other.

Every filament reacts differently because they differ in melt viscosity, thermal conductivity and behavior inside the nozzle chamber. Bambu Lab documentation notes that PLA flows easily, ABS and nylon are denser, PETG is sticky, TPU is elastic and viscous, polycarbonate is tough, and composite filaments are additionally abrasive. The differences apply not only to chemistry, but also to the specific batch, dye and moisture content of the material.

The magic number therefore does not exist as a universal constant, but as a narrow window of values for a given combination of printer, nozzle, filament and profile. The goal of calibration is to find this window and record it in the material profile so you do not have to hunt for it every time.

When calibration makes sense and when it just masks another problem

Bambu Lab notes in its official documentation that on quality printers with genuine filament, mechanical tolerances are minimal and filament standards are high. Many defects that look like bad flow actually come from an untuned dynamic calibration, a bad Z offset or an inappropriate temperature. So before you even start flow rate calibration, you should already have a temperature tower, Z offset and pressure advance dialed in, or flow dynamics calibration where applicable.

The signals that really point to bad flow are specific. Overextrusion shows up as blobs on the surface, lumpy top layers and dimensions above tolerance. Underextrusion produces gaps between perimeters, weak infill and visible cracks in the top layer. Prusa Research also describes a combined case under stress, where overextruded top layers prevent two parts from bonding together due to unevenness.

Calibration, on the other hand, is worth it for fiber-filled composites, foaming materials such as LW-PLA, recycled filaments and cheap no-name brands, where values vary even between spools. Flow rate is also more sensitive at higher print speeds, because any deviation shows up over the shorter contact time with the nozzle.

• Lumpy, glossy top layer with excess material = candidate for reducing flow
• Gaps between perimeters and a weak top wall = candidate for increasing flow
• XY dimensions consistently above nominal on thin walls = likely overextrusion
• Layer delamination without visible gaps = more likely temperature or cooling, not flow

Visual method: a quick path to a usable number

The visual method is the pragmatic approach described by both Prusa Research and Bambu Lab in their manual calibration guides. The principle is to print a series of test cubes or plates, where each sample carries a different flow ratio, usually in steps of 2 or 5 percent in the range of 90 to 110. After printing, you evaluate the top layer with the naked eye and, if needed, with side lighting at a low angle, which highlights any excess or shortage of material.

The ideal sample has a smooth, full top layer without ridges, without glossy bumps and without visible gaps between the individual extrusion lines. According to the Bambu Lab wiki, it is best to choose the sample that looks matte and flat, because the balance between a light overlap layer and a clean surface is what corresponds to optimal flow.

The visual method is supported by Bambu Studio in its Flow Rate Calibration wizard as well as by OrcaSlicer, where manual flow ratio calibration is part of the Calibration section. Practice shows that the visual method is sufficient for the vast majority of users and delivers a value accurate to about 1 to 2 percent, which is fully sufficient for everyday printing.

The precise method using thin-wall part measurement

The precise method described in the Prusa Knowledge Base targets users for whom visual assessment is not enough and who want the flow rate backed by a numerical measurement. The principle is to print a thin-walled sample, typically a calibration square with a single perimeter wall of a nominal width matching the profile (for example 0.45 mm for a 0.4 mm nozzle), and then measure the actual wall thickness with a micrometer or a quality digital caliper.

The ratio of the measured to the nominal value determines the extrusion multiplier correction. If the wall comes out two percent thicker, you reduce the multiplier by two percent, and vice versa. The method requires a calibrated measuring tool and a clean wall free of artifacts from the layer start, otherwise the measurement is meaningless. Prusa Research recommends measuring in three spots and using the average to rule out local variation.

The precise method delivers a result accurate to under one percent, but its added value shows up mainly on parts with tight tolerances, on assemblies with threaded joints where excess material prevents assembly, and on repeatable engineering applications. For decorative prints, the visual method is entirely sufficient.

• Print a single-wall sample with top and bottom infill disabled
• Measure wall thickness at 3 to 5 spots with a micrometer
• New multiplier = original multiplier times (nominal width / measured width)
• Verify by iteration, since too large a jump can overshoot

Typical value windows for the most-used filaments

Specific values always tie to the combination of printer, nozzle and brand, but community data and slicer vendor documentation give a band that realistic results do not stray from. According to Prusa Research, the usual range sits between 0.9 and 1.1, with a typical starting point of 1.00 in PrusaSlicer and around 0.98 in Bambu Studio for PLA, which reflects the different profile calibration.

PETG and PETG-CF often land in the 0.95 to 0.98 range due to stickiness and a tendency to overextrude at the edge of the top layer. ABS and ASA usually sit around 0.96 to 1.00, with a thermally stable enclosure significantly boosting repeatability. TPU and other flexible filaments are the most sensitive, and their flow ratio should be tuned in finer steps, typically 1 percent, because elastic deformation at the extruder source introduces another variable.

For composites with carbon or glass fibers, Bambu Lab documentation and community guides on the OrcaSlicer wiki point out that the abrasive fibers gradually wear down even a hardened nozzle and shift the actual flow. For long-term printing with composites, periodic recalibration pays off, for example after every 500 hours of intensive printing, so that changes in nozzle geometry do not go unnoticed.

Automatic flow rate calibration in Bambu Studio and OrcaSlicer

Bambu Studio offers an automatic flow rate calibration mode that, on X1-series printers, uses the lidar sensor and an internal algorithm to evaluate a reference sample without user intervention. Bambu Lab documentation states that the automatic mode is reliable for most common filaments, but for special materials, for example translucent, foaming or heavily filled composites, it is advisable to verify the result with the manual method.

OrcaSlicer, forked from Bambu Studio and extended by the SoftFever community, contains a dedicated Calibration section where flow ratio calibration is a two-stage process. The first stage roughly sweeps the range in steps of 5 percent, and the second stage refines the result in steps of 1 percent around the winner of the first pass. This two-pass strategy matches what the Creality wiki describes as step 1 and step 2 calibration, that is, first find the window and then narrow it down.

The difference between the automatic and manual methods in practice comes down mostly to time savings and the removal of subjective judgment. The cost is less insight into why a particular value came out. For engineering applications, it is worth backing up an automatic calibration result with a thin-wall measurement, as described in the precise method above.

Common mistakes that ruin a calibration

Flow rate calibration only makes sense when the previous steps are stable. The most common mistake is calibrating on material that has not been dried. Wet PETG, nylon or TPU behaves chaotically, its flow changes between individual layers and the calibration result is unusable. Before the test, drying is recommended per the datasheet, easily 8 hours at 70 degrees for nylon.

The second common mistake is calibrating at a different speed or temperature than the target application. Higher print speeds change pressure advance and indirectly the optimal flow. If you calibrate PETG-CF at 60 mm/s and then print an engineering part at 200 mm/s, the resulting number does not hold. Calibration should run on the profile that is used for the target print.

The third mistake concerns nozzles. A worn brass nozzle after several hundred hours of printing with composite has a widened bore, so the same extrusion multiplier will overflow. Before a major calibration, it is a good idea to replace the nozzle, or at least check it, otherwise the result becomes a moving target.

• Calibrate dried filament, otherwise you are measuring moisture, not flow
• Calibrate at the target speed and temperature, not on the default slow profile
• Check nozzle condition, wear shifts the actual flow
• Save the result to the filament profile, not to the global printer settings

When to repeat calibration and how to organize the results

Flow rate is not a value you set once and forget. Bambu Lab and Prusa Research both recommend repeating calibration whenever you change the filament brand, change the nozzle diameter, or make a major change to the print profile, for example moving from 0.2 mm layer height to 0.08 mm. For composites, it makes sense to recalibrate after a nozzle change, because the bore diameter shifts over time.

A sensible approach is to keep an overview of calibration results in a spreadsheet or directly in the profile names. Both Bambu Studio and OrcaSlicer let you save custom filament profiles with their own flow ratio values, which simplifies switching between projects later on. For engineering prints, where the same materials keep coming back, investing half an hour into naming and organizing profiles pays off.

Experience shows that once a profile is calibrated for a specific combination of filament and nozzle, it holds up for many print hours without noticeable drift. The signal to recalibrate is a drop in top layer quality, a change in dimensional accuracy, or the simple fact that a new spool from a different batch has arrived. A compact record of values also helps when tuning other materials in the same family, since the shift tends to be consistent.

Flow rate calibration is the step that turns a well-tuned printer into a predictable tool. The visual method delivers a usable result in an afternoon, the precise method with a micrometer takes values below one percent, and the automatic modes in Bambu Studio and OrcaSlicer shorten the whole process even further. The key is to calibrate dried material, on the target profile, and with a known nozzle state. A recorded and organized filament profile is then a small investment that pays back every time a part leaves the printer with tight tolerance and a clean top layer.