Chemical Crosslinking of Polyethylene and its Effect on

Chemical Crosslinking Of Polyethylene And Its Effect On-Free PDF

  • Date:20 Nov 2020
  • Views:2
  • Downloads:0
  • Pages:7
  • Size:2.87 MB

Share Pdf : Chemical Crosslinking Of Polyethylene And Its Effect On

Download and Preview : Chemical Crosslinking Of Polyethylene And Its Effect On


Report CopyRight/DMCA Form For : Chemical Crosslinking Of Polyethylene And Its Effect On


Transcription:

710 F Ciuprina et al Chemical Crosslinking of Polyethylene and its Effect on Water Tree Initiation and Propagation. Since the most common technique used today for XLPE cable pellets of A were pressed 5 min at 150 C and 40 bar and. insulation manufacturing is not irradiation but chemical afterwards the samples were cooled in air at room temperature. crosslinking it was of interest to perform a similar study using The crosslinkable polyethylene pellets used for obtaining XLA. different chemically crosslinkable compounds designed for cable were pressed 14 min at 195 C and 4 MPa 40 bar and then the. insulation The chemical crosslinking using dicumyl peroxide cooling was made with a rate of 15 C min In order to obtain. has been extensively studied and it has been shown that by identical samples the quantity of pellets used in each hole of the. products originating from decomposition of the peroxide in this pattern was set by taking into account the whole volume and the. case acetophenone influence the propagation of water trees thus density of polyethylene. introducing an additional parameter which has to be taken into. consideration 3 Experiments which were performed about. twenty years ago on samples where the by products have been. removed did not reveal any difference concerning the growth of. water trees between thermoplastic and crosslinked materials 4. However the large dispersion of the water tree lengths observed. by de Bellet and al in 1987 has led us to question if a better. accuracy presently available could reveal a difference between. XLPE and LDPE with respect to water tree growth kinetics. The materials used in this study were three polyethylenes. designed for use as insulation of MV cables One sample was a. base polymer containing only dicumyl peroxide to avoid the. possible influence of any other components on water treeing Figure 1 Pattern used to manufacture the samples. other than that of polymer crosslinking Then two different. compounds based on the same polymer but containing After manufacturing the crosslinked samples were degassed in. additionally a tree retardant additive intended to reduce the vacuum at 90 C for 72 h. number and the size of water trees were included as well. The sample preparation for the thermoplastic samples B and C. was performed in a somewhat different way compared to the. 2 EXPERIMENTAL description above The mould used for pressing was preheated at. 130 C for 10 min Then the pellets were pressed into a plaque of. 2 1 MATERIALS, 0 6 0 65 mm thickness at 130 C 2 min at 2 MPa 20 bar and 5. The water treeing tests were performed on samples provided min at 10 MPa 100 bar followed by cooling to room. by Borealis AB All materials for the samples in thermoplastic temperature at a cooling rate of 15 C min For the preparation of. form labelled A B and C were based on low density the crosslinked samples XLB and XLC the mould was preheated. polyethylene LDPE prepared by the high pressure for 10 min Then the pellets were pressed into a plaque of 0 6. polymerization process usually used to manufacture compounds 0 65 mm thickness using the following conditions 2 min at 130. for power cable insulation and they all had a Melt Flow Rate C and 2 MPa 20 bar followed by 20 min at 200 C and 10 MPa. 2 16 kg MFR2 of 2 g 10 min 100 bar and then cooling to room temperature at a cooling rate. The material A contained no further additives of 15 C min. The materials B and C were based on material A and Five plaques were prepared of each formulation After. contained antioxidant to impart thermo oxidative ageing manufacturing the crosslinked samples were heat treated in an. protection during cable manufacture and service life These oven at 90 C no vacuum for 72 h to remove the peroxide. samples also contain different water tree retardant additive decomposition products The level after heat treatment of. systems one for B and another for C both of polar nature dicumyl peroxide acetophenone cumyl alcohol alpha methyl. styrene and cumene were all found to be below 10 ppm Also the. Many studies have been conducted on the effect of peroxide thermoplastic samples were heat treated in the same way in order. crosslinking on a number of important properties of cable to have the same thermal history in all the samples The purpose. insulations 5 7 In keeping with these studies the samples of the heat treatment was to remove the crosslinking byproducts. labelled XLA XLB and XLC were obtained from the materials that are known to have large influence on the treeing 8 The. A B and C respectively by crosslinking using 2 wt of conditions used in this experiment do not significantly affect the. dicumyl peroxide These levels provide very comparable insulation morphology as they are considerably below those. performance with insulations used for a range of power cables employed in cable manufacture Furthermore any small effect. 7 that may be present due to the heat treatment is excluded from. the experiment by applying the same conditions to all of the. 2 2 SAMPLE PREPARATION tested samples Initially there were some concerns that the. Disks of 0 5 mm thickness and 50 mm diameter were press temperatures and times employed for the degassing of the. moulded from pellets of polyethylene in a CARVER press samples might induce some morphological change that would. model 2696 using a mould with 12 holes of 50 mm diameter affect the electrical performance of the insulations. and 0 5 mm thickness Figure 1 The thermoplastic polyethylene Confirmation of the fact that this is not true comes from the. Authorized licensed use limited to Florin Ciuprina Downloaded on June 24 2010 at 20 03 29 UTC from IEEE Xplore Restrictions apply. IEEE Transactions on Dielectrics and Electrical Insulation Vol 17 No 3 June 2010 711. inspection of data obtained from experiments where selected of the ionic solution have to be well defined and consistent As. crystallization and annealing conditions were applied to a defects acting as initiation sites for water trees small needle like. similar type of LDPE to that discussed in this paper 9 These imprints were created on one side of the compression moulded. experiments show that the conditions used here for degassing sample by pressing a sheet of abrasive paper P240 type with an. only effect the breakdown strength by 3 and these changes average particle size smaller than 25 m for 2 min at 50 MPa. are not statistically significant due to the inherent scatter in 500 bar at room temperature. breakdown tests It is important to note that the results The images in Figure 2 show the typical distribution of. reported in the present paper display large and more defects achieved for each of the tested samples Inspection of. significant changes than can be attributed to any the images shows that the initiation sites are evenly in. morphological changes separation and size on the surface of the samples Figure 3. The crosslinking degree was assessed by gel fraction gives the concentrations of the initiating defects in a Box and. measurements in accordance to the ASTM D2765 procedure by Whisker format segregated for the three material types both. refluxing xylene close to its boiling point and the extraction was thermoplastic and crosslinked versions The median defect. carried out until the insoluble gel reached a constant weight The contribution for this approach is expected to lie within the. extraction time was of at least 96 h range set by the lower and upper quartiles 100 to 140 defects. per mm2 Analysis of the data Kruskal Wallis test shows that. 2 3 WATER TREES there is no significant difference between the material types. To compare the water tree resistance of polymeric materials it is Figure 3a However the same analysis shows that the. critical to grow water trees under well defined conditions which thermoplastic versions have 25 more initiation sites than the. include a a defect at the water polymer interface which crosslinked versions Figure 3b and it also confirms that this. amplifies the electric field b an ac electric field c an ionic difference is statistically significant Consequently one would. solution For consistency of the results the shape of the defect expect the thermoplastic samples to contain 25 more. the frequency and amplitude of the voltage and the concentration initiated water trees. Figure 2 Defects acting as initiation sites for water trees created on samples by pressing a sheet of abrasive paper before water tree tests. Authorized licensed use limited to Florin Ciuprina Downloaded on June 24 2010 at 20 03 29 UTC from IEEE Xplore Restrictions apply. 712 F Ciuprina et al Chemical Crosslinking of Polyethylene and its Effect on Water Tree Initiation and Propagation. Figure 3 Concentration of defect sites estimated from images shown in Figure 2 represented in a box and whisker plot the boxes enclose 50 of the data the. whiskers 100 the central lines represent the medians a materials A B and C both forms b crosslinked and thermoplastic all materials. In our experiments water trees were grown in cells Figure 4. arranged by fixing the sample on a polyethylene tube using. LOCTITE 401 in combination with an adequate surface. treatment The electrolyte was a NaCl solution of concentration. c 0 1 mol l Groups of five cells were fixed in a cell holder and. water trees were grown by applying a voltage of 2 kV at a. frequency of 5 kHz for 25 h at room temperature, After ageing the samples were detached and dyed in a Figure 5 a Slices for measuring water trees dimensions b Water tree. rhodamine solution at 60 C for 3 days to facilitate the length Lk as measured on a slice. measurements of the water tree lengths Three slices of 200. m thickness were microtomed from each sample Figure 5a. and the lengths of all water trees from each slice were. measured using the experimental setup shown in Figure 6 The. average length La for each sample was determined as the. average of the water trees lengths Lk measured on the three. slices of the sample Figure 5b The number of trees in each. slice was counted to determine the water tree density on the. sample surface, Ten samples of each type were tested according to the. procedure described above Figure 6 Setup used to measure water tree lengths and water tree. 3 RESULTS AND DISCUSSION 3 1 RESULT ACCURACY, To evaluate the accuracy of the measurements both the. Table 1 gives the average water tree lengths Lai obtained. scattering between samples and the uncertainty on the. from each of the ten samples tested for each material the. measurements of the water tree lengths need to be, average length of the ten average tree lengths and the.
considered In fact the scattering in the Lai values is very. measured crosslinking degree, small in the order of 3 Based on this finding it was. decided to consider the measurement error on the length. values as well Among the different sources of error the. depth and the shape of the water tree initiation sites as well. as the determination of the water tree front could be. mentioned The last one is by far the most important and. the scatter in the length measurements can be estimated to. be in the range between 10 and 15 depending on the type. of material used It has been observed that different types of. materials give rise to trees that are more or less marked. Finally it was concluded that each material can be. characterized by the average length L affected by an overall. uncertainty of less than 15 which is a lower value than. the previous work on this subject 4 Figure 7 gives the. distribution of the individual average water tree lengths for. Figure 4 Cell used to produce water trees different material types and crosslinking conditions. Authorized licensed use limited to Florin Ciuprina Downloaded on June 24 2010 at 20 03 29 UTC from IEEE Xplore Restrictions apply. IEEE Transactions on Dielectrics and Electrical Insulation Vol 17 No 3 June 2010 713. Table 1 Water tree lengths and crosslinking degree for the tested materials. Samples Cross Water tree lengths of the ten tested samples Average. linking m length m,La1 La2 La3 La4 La5 La6 La7 La8 La9 La10 L. A 0 281 275 282 283 255 273 259 262 262 260 269, XLA 81 281 293 275 279 309 282 283 285 289 283 286. B 0 67 59 68 73 68 69 81 68 67 78 70,XLB 84 100 94 81 81 93 91 83 78 82 89 87. C 0 150 129 149 161 157 133 146 140 163 155 148, XLC 81 5 121 144 139 156 135 147 127 151 129 144 139.
During peroxide crosslinking of polyethylene the radicals. 3 2 INFLUENCE OF CROSSLINKING ON WATER, generated from the decomposed peroxide abstract hydrogen. TREE GROWTH, atoms from the polyethylene chain forming macroradicals. A detailed discussion of the statistical tests is outside the scope When two macroradicals combine a crosslink is formed. of this paper and will be dealt with elsewhere Nevertheless However several authors have suggested that during. inspection of Table 1 and Figure 7 shows three very interesting crosslinking the entanglements naturally present in a. features Firstly these data confirm the results published in earlier polymer melt become trapped to different degrees so called. studies by de Bellet et al 4 generated under other conditions physical crosslinks by the chemical crosslinks thereby. and with a lower accuracy and they are also in agreement with contributing to the network density It has been reported that. the findings in our previous study where water trees were grown in the network formed in crosslinked low density. in irradiation crosslinked polyethylene 2 polyethylene around 1 3 of the crosslinks originate from. Chemical Crosslinking of Polyethylene and its Effect on Water Tree Initiation and Propagation Florin Ciuprina ELMAT Faculty of Electrical Engineering University Politehnica of Bucharest Splaiul Independentei 313 060042 Bucharest Romania Gis le Teiss dre Jean C sar Filippini G2E Laboratory CNRS INPG UJF

Related Books