Telling PET and CoPET fibres apart in a CT scan of a nonwoven

This spunbond is made of two polyesters: PET fibres that carry the sheet and CoPET fibres that melt in the calender and bond it. In a CT scan the two look alike, so a segmentation by grey value sees one material. Here they are told apart in a µCT scan, which gives how much binder the sheet holds and where it sits.

The same strip of the nonwoven twice in the same oblique view: on the left all fibres grey as scanned, on the right PET fibres grey, CoPET fibres orange and molten CoPET at the bonds red.
One strip of the sheet as scanned, and with the fibres assigned to their polymer.

The sample

The sample is a commercial PET/CoPET bicomponent spunbond filter medium, thermally bonded in a calender. RJL Micro & Analytic scanned a piece of a few millimetres in a µCT. Two strips away from the calender lines were analysed. The case study on preparing a CT scan uses a cut-out of the same scan at a coarser resolution.

Top view of the µCT scan of the nonwoven with a scale bar, the two calender lines marked in orange and the two analysed strips outlined in blue.
Top view of the scan with the calender lines and the two strips.

Binder fibres and bonds

With the PET removed, the binder stands alone: CoPET fibres that kept their shape, and CoPET that melted and now sits at the contacts between fibres. Every analysed PET fibre in the two strips is bonded.

On the left the binder network of the strip alone, CoPET fibres in orange and molten CoPET in red. On the right a section through bonds in the strip, with fibre cross-sections and molten CoPET between them.
The binder network without the PET, and a section through bonds.

Binder share

The CoPET share is given as a share of the solid volume, for the fibres and the molten material separately. It is a lower bound, because CoPET that has melted into films thinner than the scan resolves is not counted. The two strips lie on either side of a calender line and give different shares; part of the difference lies within the uncertainty of the evaluation.

Stacked bars of the CoPET share of the solid volume for the two strips, split into CoPET fibres and molten CoPET, labelled as a lower bound.
CoPET share of the solid volume in the two strips.

Binder across the thickness

Across the thickness the CoPET share is higher towards both faces of the sheet than in its middle.

For each strip a stacked profile of PET, CoPET fibres and molten CoPET over the height, and below it the CoPET share of the solid per layer, higher at both faces than in the middle.
PET, CoPET fibres and molten CoPET over the thickness of the two strips.

The structure of the sheet

The same strips give pore sizes and porosity.

Pore size distribution of the two strips with the median pore diameter and the porosity of both strips.
Pore size distribution and porosity of the two strips.

Limits

The scan covers a few square millimetres of one sheet and does not describe a roll. The CoPET share is a lower bound. The method has been applied to this one material so far. The assignment of the fibres has not been checked against a chemical analysis of single fibres. Filtration has not been simulated on this structure.


Fibre analysis of nonwovens

cvt engineering offers this analysis for CT scans of other thermally bonded nonwovens with carrier and binder fibres.