2.1. Material

The material used was PGLA multifilament. PGLA is a copolymer of polyglycol acid (GA) and polylactic acid (LA). Besides good biological compatibility, PGLA also presents a reasonable degradable rate for peripheral nerve regeneration. Previous work has shown that PGLA (90:10) material (90% GA, 10% LA ) degrades completely after 90 days of an in vivo degradation test. This is compatible with the period of the peripheral nerve regeneration (about 3 months) [5, 6]. Some properties of the 5.8tex PGLA (90:10) multifilament used in this work are shown in Table 1.

Table 1. Some properties of PGLA filament (90:10)

2.2. Braiding and post-treatment

Braiding is a very suitable textile technology for the production of peripheral nerve regeneration conduits with very small diameters. In order to fabricate such conduits on a braiding machine with a given number of yarns to obtain the desired structural effect, it is essential to choose a proper yarn linear density at the first step. In this work, 40.6tex and 23.2tex PGLA yarns were used as braiding yarns and axial yarns respectively. However, the PGLA multifilament commercially sold was 5.8tex. For this reason, 4 strands and 7 strands of 5.8tex multifilament were respectively ply-twisted together to obtain the desired yarn linear densities.

The braiding machine used was a conventional tubular braiding machine equipped with 16 spindles. In order to prevent the conduits from being flattened due to the taking-up operation, and to obtain the desired inner diameter, a cylindrical metal core mandrel was adopted during braiding.

Three kinds of typical tubular braided structures were used; diamond, regular and triaxial, as shown in Figure 2. The diamond and regular structures have 1/1 and 2/2 intersecting repeat patterns respectively (Figure 2(a) and (b)). The triaxial structure is a kind of braided structure in which the axial yarns are introduced through stationary guide eyes. Figure 2(c) is a triaxial structure in which the axial yarns are introduced into a 2/2 repeating pattern.

(a)

(b) (c) Figure 2. Structures of the braided conduits

(a)

Diamond structure (b) Regular structure (c) Triaxial structure

Besides the braided structure, the braiding angle can also influence the radial compressive properties of the conduits. As shown in Figure 2 (a), the braiding angle α is defined as the angle formed between the

braiding yarns and the axial direction of the tubular fabric. According to the adjustment limits of the braiding machine, 4 braiding angles of 45, 50, 55 and 60 degrees were selected for each braided structure respectively.

The metal mandrel in conduits must be drawn out after braiding to form a tubular hollow structure. However, braided fabrics without any treatment can easily change their shape after the mandrel is drawn out. Because of this, a post-treatment is necessary to help them keep their form. In this work, the braided conduits were firstly soaked in chitosan solution for half an hour, and then set at 70°C for 15 minutes. Chitosan is a kind of biodegradable material compatible with human tissue, which has a favourable biodegradable rate.

Combining 3 braided structures and 4 braiding angles, 12 kinds of conduit samples were produced. The structural parameters of these conduits measured after the post-treatment are listed in Table 2.

Table 2. Structural parameters of the braided conduits after the post-treatment

2.3. Radial Compressive Testing

Due to the very small diameters of the nerve conduits, conventional testing devices for fabrics are not suitable for testing their compressive properties in the radial direction. For this reason, a custom-built compression device was constructed by Laizhou Electron Instrument Co. Ltd. in China to perform this kind of test. Figure 3 schematically shows its working principle, in which the upper fixed plate is set up with a sensor and the testing conduit is placed on the mobile lower support rising to compress the conduit along its radial direction. Because there are no Testing Standards for testing the radial compressive properties of such conduits, the following testing conditions were chosen, based on the previous works and the biomechanics of peripheral nerve under radial compression action [7]:

Diameter of the upper fixed plate: 5mm,

Length of specimen: 7mm,

Compression-decompression cycles: 5,

Maximal displacement: 50% of the initial diameter,

Displacement rate: 2 mm/min.

(a) (b) Figure 3. Schematic diagram of the working principle for the radial compressive testing device

(a) Radial direction (b) Longitudinal direction

The tests were undertaken under the standard air conditions (20±2°C, RH 65±2%). Five specimens were tested for each testing case, and all the results were statistically analysed after testing.

3.1. General Load/Strain Behaviours

The typical radial compressive load/strain curves after the first compression cycle for the different braided structures and the different braiding angles are illustrated in Figure 4 and Figure 5 respectively. In Figure 4, the braiding angle is kept constant (60 degrees) for three curves, while in Figure 5 the braided structures are kept the same (triaxial structure) for four curves. It can be seen from Figures 4 and 5 that all the curves have a similar form. This means that they have similar radial compressive characteristics. In order to better understand these characteristics, each curve has been divided into 4 parts for analysis. The first is the initial part (around 1%) of the strain in which the compressive loads are very small or even nearly zero. This is due to incomplete contact between the braided conduits and the plates of the testing device. Because the interlaced points are not even, only a few points of the conduits contact the plates at the beginning of testing. After the initial part, the conduits begin to undertake the compressive loads. Thus, the second part is characterised by a near-linear increase of the compressive loads with compressive strain from about 2% to an inflexion region between 15-20% depending on the braided structures. The slopes of the load/strain curves in this part are relatively bigger than other parts. After the inflexion region, the increasing rate of the compressive load with the strain starts to decrease, and a platform around 15 to 45% appears. This is the third part. Here, the shapes of the conduits are noticeably deformed in the radial direction. However, the compressive loads do not increase significantly. It can also be seen from Figure 4 that the platform of the curve for the triaxial structure is more obvious than the other structures. In the following part, the compressive loads increase significantly again with the compressive strain until 50% of compression.

The above analysis has shown that the braided conduits have their particular compressive behaviours which differ from these of conventional planar textile structures, which compress along with the thickness direction. The particular compressive behaviours of the braided conduits include: 1. a very low compressive modulus at the beginning of compression; 2. high radial deformation due to the hollow structure. It is easy to compress the conduits to 50% of the deformation related to their initial diameters;

3. with an increase in the compressive deformation (more than 15%), the buckling effect can appear. This results in a reduction of the compressive modulus in the third part of the curves. 4. After 40% of the deformation, the compressive modulus increases again due to the quasi-contact of the inner layers.

900
750 )	Triaxial Structure
450 600 Compressive Load (cN		Regular Structure
300
150		Diamond Structure
0

Radial Compressive Strain (%) Figure 4. Typical radial compressive load/strain curves after first compression cycle for the different braided structures

0 10 20 30 40 50

Figure 5. Typical radial compressive load/strain curves after first compression cycle for the different braiding angles

3.2. Influence of the braided structures on the compressive behaviours

It can be seen from Figure 4 that the influence of the braided structures on the compressive behaviours is evident. Among the three braided structures analysed, the triaxial structure has a greater ability to resist the mechanical compression than the regular and diamond structures when the braiding angle is kept constant. The diamond structure has the least ability to resist the mechanical compression. This phenomenon can be explained by the facts that there are less bending of the braiding yarns in the regular structure than in the diamond structure, and that the axial yarns introduced into the triaxial structure can restrain the shift of the braiding yarns and improve the stability of the conduits.

45 50 55 60 Braiding Angle (degree)

Figure 6. Percent retention of the compressive load vs. braiding angle

3.3. Influence of the braiding angles on the compressive behaviours

Taking the conduits produced with triaxial structure as an example, the radial compressive loads gradually increase with the increase of the braiding angles, as shown in Figure 5. This is normal, because the thickness and the yarn density in the axial direction increase with the increase of the braiding angle. Consequently, the ability to resist the mechanical compression increases. However, increasing the braiding angle to enhance the capacity of the conduits to undertake higher compressive loads is limited due to the braiding process. Besides, a high braiding angle will result in the reduction of the tensile properties of the braided conduits in the axial direction. In practical application, the choice of the braiding angle must be equilibrated between the radial compressive properties and the axial tensile properties.

3.3. Load Retention after 5 compression-decompression cycles

In order to verify the ability of the conduits to resist mechanical compression during use, the radial compressive loads at 50% of the deformation after the first compression-decompression cycle were compared to those after 5 compression-decompression cycles. The results of the compressive load retention expressed as percentages are shown in Figure 6. It can be seen from Figure 6 that after 5 compression-decompression cycles the load retention is still over 80%. However, the influence of the braided structure and braiding angle on the load retention is relatively small. The difference does not exceed 5% for the same structure when the braiding angle changes, or for the same braiding angle when the braided structure changes. The results have also shown that the conduits braided with the triaxial structure and with a high braiding angle give better results. This means that the triaxial structure braided with 60 degrees is the first choice for the conduits from the point of view of load retention.

Figure 7. Elastic recovery rate vs. braiding angle of the fifth compression-decompression cycle

3.4. Elastic recovery rates of the fifth compression-decompression cycle

In addition to the capacity to undertake the compressive loads, elastic recovery rate is also an important parameter during the utilisation of the conduits. The elastic recovery rate refers to the percentage of the elastic deformation to the total deformation of the conduits. The results of the elastic recovery of the fifth compression-decompression cycle are shown in Figure 7. It can be seen from Figure 7 that both the braided structure and braiding angle influence the elastic recovery of the conduits. However, this influence is relatively small, less than 5%. It is also necessary to point out that the influencing trends are opposite to those for the radial compressive loads. This phenomenon can be explained by the fact that when the braided structures are changed from diamond and regular to triaxial, and when the braiding angle increases from 45 to 60 degrees, the structures of the conduits become more compact, and the contacting areas between the yarns increase. For this reason, the conduits after compression have to overcome more frictional forces between the yarns to recover their initial state. Contrarily, the conduit with a more compact structure can better bear the compressive loads. There will an equilibrated choice between the load retention and elastic recovery. An experimental test for the degradation of the conduits has shown that the conduits with a more compact structure give better results for the clinical application.