Meltblown Nonwovens: Fibre Formation and Filtration Trade-Of

Start with the web's job

A meltblown line can make a fine web that still fails its intended use. The roll may meet its average weight yet have thin lanes. A filter sample may capture more particles while making airflow too hard. A charged web may pass when fresh and lose performance after handling or storage.

Good process control follows a chain. It links the polymer, forming conditions and web structure to a defined product test. Fibre diameter is one part of that chain. Weight, pore structure, charge, strength and defects also matter. A useful trial must show which condition changed and which result followed.

This guide follows a hypothetical filtration web from melt delivery to roll release. Its examples use invented production and test data. They show how to calculate and make decisions. They do not give a recipe for a particular resin or machine. The proposed acceptance limits are engineering examples. They do not establish a medical product claim or respirator approval.

How the meltblown process forms a web

The process starts with a polymer suited to the line and product. An extruder melts it, while the delivery system controls the flow toward the die. Fine polymer streams emerge from small openings. Fast air draws those streams into much thinner fibres. They then reach the collector.

Reicofil's meltblown line description shows a single-row spinneret. Hot air acts on both sides. That is one equipment arrangement; other die designs exist. A collector gathers the fibres into a web. Downstream equipment supports, winds or converts it as required.

The fibres form and the web builds in a linked process. A change near the die can alter what reaches the belt. A change in belt speed can alter mass per area even when melt delivery stays fixed. Airflow at the collector can also affect how the fibres settle.

EDANA's nonwoven manufacturing overview places meltblown within the spunmelt routes. It describes low-viscosity polymer streams that are still partly molten. Hot air acts on those streams. Thin meltblown fibres may need support for later use. Converting can include cutting, folding, joining or sealing, according to the product.

That distinction helps prevent a common purchasing error. The meltblown layer may provide a function within a larger structure. The finished article needs strength and resistance to handling. It may also need to form a seal. The meltblown layer need not provide all those features. Qualify the layer and its final assembly separately.

Define the output before setting the line

Write the product requirements before running a process trial. State the intended use, finished width and target mass per unit area. Add the required mechanical properties, visible defect rules and functional tests. Identify which requirements apply to the web and which apply after conversion.

For filtration work, “high efficiency” is incomplete. Specify the particle measure, test conditions and pressure-drop limit. Define the result. Does it describe a chosen particle-size band or a broader aerosol response? Include conditioning, loading or ageing tests when the product brief calls for them.

Identify the resin grade and lot in the process record. Add the additives and any reprocessed content. A melt-flow value needs its stated test conditions. A bare number cannot safely establish that two materials will run alike. Use supplier and machine guidance to set the safe trial range.

Record melt delivery, relevant temperatures and air settings. Add collection distance, belt speed and suction settings. Keep their actual units. Record measured values as well as setpoints where sensors permit. A displayed temperature does not describe every part of the melt's thermal history. A pressure setting is also not the same thing as a measured air mass flow.

Calculate mass flow before changing speed

Suppose a line forms a dry, uniform web across 3.0 metres. Its belt travels at 40 metres per minute. The target basis weight is 20 grams per square metre, often written as 20 gsm.

The formed area per hour is:

3.0 m × 40 m/min × 60 min/h = 7,200 m²/h.

The gross dry web output is:

7,200 m²/h × 20 g/m² ÷ 1,000 g/kg = 144 kg/h.

This balance assumes steady operation and the stated average weight. It describes material in the formed web. It excludes start-up purge, unstable running and rejected rolls. It also excludes material retained elsewhere in the system.

Now trim the accepted roll to 2.8 metres without changing speed or basis weight. Accepted area becomes 6,720 m²/h. Accepted web mass is 134.4 kg/h, while edge trim is 9.6 kg/h. The trim share of gross web mass is about 6.67%.

Do not call 144 kg/h the saleable rate under those conditions. Keep gross formed mass, usable width and accepted output distinct. Any recycled trim needs its own material and quality controls. This calculation does not assume it can return to the same product.

If the belt rises to 50 m/min while gross mass stays at 144 kg/h, basis weight falls to 16 gsm. Holding 20 gsm at that speed and full width would require 180 kg/h in the web. Raising melt flow to reach that figure can change fibre formation. The line must prove the product again under the new conditions.

Separate average weight from uniformity

Average gsm can hide a weak lane. Consider five equal-width zones across a trial web. Their measured basis weights are 18, 19, 20, 21 and 22 gsm. Their equally weighted mean is exactly 20 gsm. The range is 4 gsm, and the outer values sit 10% below and above the mean.

Suppose the trial specification requires every sampled zone to lie from 19 to 21 gsm. The mean meets target, but two zones fail. Five points here illustrate the logic. They are not a universal sampling plan or proof that the rest of the roll conforms.

Choose locations across the width and along the roll. Retain each location with its result. Record the sample area, conditioning and weighing method. Map repeat thin or heavy lanes against die positions and collection conditions. Do this before assigning a cause.

ISO 9073-1:2023 covers determining mass per unit area of nonwovens. Its public scope identifies the method's purpose. Use the full applicable method where conformity is required. This guide does not reproduce its procedure.

ApparelDiary's fabric GSM testing and roll-weight guide explains related area and mass calculations. Apply those balances to clearly defined web samples. Do not let a correct roll-average calculation replace a local uniformity check.

Treat fibre diameter as a distribution

A single average fibre diameter is a weak description of a web. It can hide a broad spread, coarse strands and local defects. Keep the sampling location and imaging method with the measurements. Report enough of the distribution to show both its centre and spread.

Avoid selecting only the cleanest images. Define how fields and fibres will be chosen before measuring them. Separate fibre measurements from fused patches or polymer lumps. A microscope view can reveal structure. It does not directly replace a filtration or strength test.

A primary study of annular melt blowing combined modelling with experiments using polycaprolactone. The team studied changes in polymer throughput and air conditions. These affected fibre size and motion within the studied range. Those findings support controlled investigation of the air–polymer relationship. The study used a small-scale setup and a specific die design. It does not supply a universal industrial polypropylene recipe.

Change settings within a justified range and record the result. “More air makes finer fibres” is too broad to serve as a complete operating rule. Test a defined adjustment under the actual line conditions. Does it improve the required web?

Link filtration efficiency to pressure drop

A filter web must capture particles while allowing the required airflow. Particle penetration is the fraction that passes through under the test conditions. Efficiency is one minus that fraction. Use comparable upstream and downstream concentrations, including any required dilution or background correction.

For the worked example, define:

Penetration, P = downstream concentration ÷ upstream concentration.

Efficiency, η = 1 − P.

The published study by Bement and colleagues gives an equation for filtration quality factor. It uses the natural logarithm:

Quality factor, Q = −ln(P) ÷ Δp = −ln(1 − η) ÷ Δp.

Here Δp is pressure drop. If pressure is in pascals, Q is in Pa⁻¹. Efficiency enters as a fraction, so 99% means 0.99. This expression uses the natural logarithm, not a base-ten logarithm.

Quality factor combines capture and airflow resistance for a defined comparison. It does not replace separate product limits. Nor does it make results comparable when test conditions differ. Retain those conditions with the calculation, and interpret the result within their scope.

Work through a controlled filter comparison

Assume a clean-media laboratory test exposes 100 cm² of each web to 60 litres per minute of airflow. The area is 0.01 m². The flow is 0.001 m³/s, so face velocity is 0.10 m/s, or 10 cm/s.

All three invented samples have a nominal basis weight of 20 gsm. The aerosol is dry sodium chloride. The result below covers one optical-equivalent size channel, 0.45–0.55 micrometres. Aerosol charge condition, humidity and temperature stay the same. So do fixture sealing and the test procedure. These details define the example; they are not a claimed standard test.

M1 is an uncharged base web. M2 is a matched web after an electret treatment. M3 has a different fibre structure and also receives a charge treatment. The corrected upstream concentration is 1,000 particles/cm³ in each comparison.

Sample	Downstream, particles/cm³	Efficiency	Pressure drop
M1	50	95.0%	100 Pa
M2	10	99.0%	100 Pa
M3	5	99.5%	160 Pa

For M1, penetration is 50 ÷ 1,000 = 0.05. Its quality factor is −ln(0.05) ÷ 100, or about 0.0300 Pa⁻¹. M2 has penetration 0.01 and a quality factor of about 0.0461 Pa⁻¹. M3 has penetration 0.005 and a quality factor of about 0.0331 Pa⁻¹.

M3 captures the largest share in this channel, but M2 has the largest quality factor. That comparison does not establish the best finished filter. It shows how the extra pressure drop changes this particular trade-off.

Suppose the trial requires at least 98.5% capture in this channel and no more than 120 Pa. M1 fails capture, M3 fails pressure drop and M2 meets both trial limits. The team still needs the other specified tests before accepting the product. A single channel cannot describe all particle sizes or loading conditions. It cannot establish the performance of the finished assembly.

M2 improves capture by four percentage points over M1. Its penetration falls from 0.05 to 0.01, an 80% reduction. These describe different changes. Calling the capture gain “80 percentage points” would be wrong.

Keep test conditions with every result

Changing exposed area changes face velocity when volume flow stays fixed. At 60 L/min through 50 cm², face velocity becomes 20 cm/s. A result from that setup is not directly interchangeable with the earlier 10 cm/s result.

TSI's filter-testing overview distinguishes tools that assess penetration, pressure drop and fractional efficiency. It also shows why the instrument and measurement basis matter. An integrated aerosol response differs from a size-resolved particle count. They do not describe exactly the same measurement.

Keep the test aerosol, size measure and charge condition with the result. Add flow, exposed area and conditioning. Record the instrument, method, sampling locations and replicate results. Define the reported pressure drop. Is it across the sample alone, or does it include fixture losses?

Check for edge leakage and damaged seals before blaming the web. Keep the blank and instrument checks required by the method. The detector may find no downstream particles. In that case, report the measurement limit and the appropriate bound. A zero count does not prove that the material captures every possible particle.

Understand what electret charging changes

Electret treatment can add an electrostatic contribution to particle capture. It need not produce the same change in airflow resistance as making a denser structure. The effect depends on the material, treatment, exposure and test conditions. Treat charge as a controlled product feature with its own records.

A 2026 primary study of charged PLA meltblown webs found higher filtration after electret treatment. Pressure drop remained similar in the tested samples. The authors also studied retention during storage. This supports checking charge-related performance over time. It does not establish one lifetime for every web. Polymer, treatment and storage conditions still matter.

Extend the fictional M2 example to illustrate the point. Apply a separately defined conditioning exposure. Suppose the downstream concentration then rises to 20 particles/cm³. With the same upstream value, efficiency becomes 98.0%. Assume its pressure drop stays at 100 Pa.

The sample now fails the earlier 98.5% trial limit even though pressure drop is unchanged. Its quality factor becomes about 0.0391 Pa⁻¹. A pressure-only release check would miss that loss. These assumed data demonstrate the decision. They do not predict any actual web's ageing rate.

Define storage, packaging and conditioning tests from the intended product use. Record time from charging to testing. Evaluate any handling, cleaning or conversion process that could affect the qualified material. Do not infer safe reuse or cleaning instructions from a web's fresh filtration result.

Plan trials that can reveal causes

Begin with a stable reference condition and a clear question. Suppose the team plans a melt-flow change to raise output. Can it keep uniformity and filter performance within limits? Select a limited set of responses before the run. Keep production losses visible as well as accepted output.

Change one factor at a time when that answers the question. Use a designed experiment when interactions matter and the team can support it. Either approach needs a stable process, traceable samples and repeat measurements. A complex design cannot rescue poorly identified rolls.

Collect samples after the agreed stabilisation point. Mark which material belongs to a transition and which belongs to a steady condition. Trials may span different resin lots or shifts. Keep those differences in the analysis. Returning to a reference condition can help expose drift during the trial.

Observation	What to inspect next	Evidence needed before a change
Repeated thin lane	Local delivery and collection conditions	Width map linked to stable run data
Coarse strands or lumps	Melt state, die condition and air balance	Representative images and maintenance history
Higher pressure drop	Weight, structure and test setup	Comparable flow, area and sample records
Capture falls with similar pressure	Charge history, conditioning and leakage	Matched functional tests and seal checks

These are investigation paths, not one-cause diagnoses. Several mechanisms can produce a similar visible fault. Make safe changes through the site's operating and maintenance procedures. Apply the assigned controls for hot melt, hot air and moving rolls. Electrical charging equipment needs its controls too.

First establish a stable method. Then use ApparelDiary's practical statistical process control guide to plan ongoing monitoring. Use charts for a defined measure and sampling scheme. Product limits and process control limits serve different purposes. A steady process can still make unsuitable material.

Make the sampling record useful to the next shift

A sample label should let someone find the material again. Include the parent roll and run condition. Record the time or length position and the location across the width. Keep a map when the web is slit into narrower rolls. A label that says only “trial B” loses value once several operators handle the material.

Separate repeated tests on one piece from tests on separate web locations. The first can help assess the measurement method. The second can reveal variation in the material. Neither alone shows how the process behaves across different runs. Plan those levels around the decision the trial must support.

Where testing destroys the sample, use nearby companion pieces for related tests. Record their positions and avoid claiming that separate pieces are identical. For a charge comparison, plan matched material before treatment. A hidden change in web structure could otherwise affect the comparison. Keep untreated controls under a defined handling condition too.

Agree on what will happen if a test is missing or a specimen is damaged. Keep that entry visible with its reason. Do not replace every awkward result with a better-looking sample. That can conceal the variation the trial was meant to assess. A valid retest needs a documented reason and the agreed method.

At the shift handover, pass on the current settings and stable running window. Identify accepted rolls and unresolved faults. Mark trial material that still awaits results. An operator should not have to infer release status from the place where a roll has been stored.

Finally, write the decision against the original question. A speed increase may raise gross output yet reduce accepted width or cause more rejects. Show both effects. If a web passes fresh tests but its conditioning result is pending, keep that limit in the conclusion. This makes the trial useful for production, purchasing and quality teams. Each team can see what the evidence supports.

Check strength, conversion and roll history

A successful filter test does not prove that the web will survive winding or assembly. Agree on the mechanical tests and directions that matter. Keep machine-direction and cross-direction results distinct. Match specimen preparation, conditioning and test settings before comparing suppliers or trials.

ISO 9073-3:2023 concerns strip tests for nonwovens. It covers breaking force and elongation. Use it with its published correction. Its public scope includes defined specimen-width and condition options. Choose the applicable method consistently. It does not supply a universal strength limit for all meltblown products.

Inspect material after the real converting route. Excess tension or handling damage can defeat an otherwise suitable layer. Poor joining or a local tear can do the same. Use roll identities to link the resin lot and run conditions. Keep charge treatment and test results with those records. Preserve the links through slitting and assembly.

At release, review the specified functional, physical and visual results together. Resolve missing results and nonconforming zones under the agreed disposition process. Retain samples where the quality plan requires them. A certificate should describe the tested material and method. Keep its claim within the evidence.

Questions that often cause confusion

Does finer fibre always mean a better filter?

No single fibre-size result settles the choice. Evaluate capture, pressure drop, uniformity, durability and the required mechanical properties together. A finer web that fails the pressure or handling limit does not meet that product brief.

Can a higher gsm rescue a poor process?

More mass can change performance. It does not establish uniformity or remove defects. Check the source of thin lanes and unstable formation. Test the revised web under the same stated conditions. Check its function before accepting the added material cost.

Does good meltblown media prove respirator protection?

Media filtration is only part of that question. NIOSH's filtration and fit explanation explains two parts of protection. The filter must work, and leaks around the face seal also matter. A roll-goods result cannot establish whole-product approval or wearer protection. Finished protective products need their applicable evaluation.

What should a useful trial report conclude?

State which defined condition met which limits. Give the accepted output and any remaining tests. Preserve the evidence behind that decision. The goal is a repeatable process that produces a qualified web. Keep a traceable link between settings, structure and actual performance.

Meltblown Nonwovens: Fibre Formation and Filtration Trade-Offs