The pycnometric density of the pyrocarbon matrix ρpyc is determined experimentally by two methods. In the first pyrocarbon deposits are removed from the surface of disks and their density is determined pycnometrically after grinding particles to a size of less than 1mm. In the second method small specimens of 1D-structure reinforcement made from silica threads are placed over the radius in a preform of carbon fibers which were densified by pyrocarbon according to the standard procedure. After densification of silica specimens are extracted from the carbon-carbon billet and turned over the surface in order to obtain cylinders of regular shape. Thenthe silica fibers are etched out in fluoric acid to a constant mass and pycnometric density of the specimens containing only pyrocarbon is determined Monitoring for the pressure within them of silica fiber residues is carried out by burning and determining the ash content which was 0.15-0.2 mass% as in the majority of carbon-carbon materials. With both methods a value of 2.1-2.15 g/cm3 was obtained for the pycnometric density of pyrocarbon. In subsequent calculations we adopt ρpyc =2.1g/cm3.
The ultimate possible level of filling cavities with pyrocarbon η is determined in accordance with the cluster theory of permeability threshold. According to this theory in porous material pores cease to communicate with each other when the overall content of pyrocarbon becomes less than 15-17%, i.e., the cavity filing factor with pyrocarbon cannot exceed 0.85-0.83. In future calculations we adopt and average valueη=0.84.
As already mentioned, for brake disk production material is required with a density of not less than 1.7 g/cm3. With a pycnometric density of the original filler of not less than 1.8g/cm3 in order to provide a higher final material density we are interested in reducing the specific content of filler in a preform. This unavoidably leads to a reduction in mechanical strength of the material since it is mainly provided by the strength of the carbon fiber.
With a pycnometric density exceeding or equal to 1.8g/cm3 it is desirable to increase the content of filler in a preform. Apart from the increase in mechanical strength of the composite material with use of high-density carbon fiber as a filler there is an increase in the final density of the composite material.
In order to check theoretical results we carried out a number of experiments. We dwell in detail on some of them.
We prepared two series of full-scale brake disks based on two sorts of carbon fabric: URAL TM-4 and “Etan”. Ring segments of two standard sizes were cut from a sheet of carbon fabric 500-540 mm wide using templates: with diameters 245*450 mm and 275*480mm.
The choice of the dimensions and shape of segments is due to the requirement of optimum pattern cutting of the carbon fabric. Blanks of model disks were formed by laying up segments in a mold. Segments were compressed in the molds to the required density by means of special disks having perforations for sewing the fabric layers with thread. The lobes compressed in the mold were sewn by hand with carbon thread URAL-NSh-24 with a pitch of -20*20mm. The volume content of carbon fabric was varied by the number of laid-up lobes and degree of their compression. For fabric URAL TM-4 the number of layers was 21-25, and for “Etan” fabric it was within the limits 50-63.
In order to determine the pycnometric density of the carbon fiber, fabric specimens were selected from each batch. The fiber pycnometric density was determined by the standard procedure of vacuum impregnation of carbon fibers with kerosene, and then the values were averaged.
The pycnometric density of fibers for URAL TM-4 fabric was 1.36 g/cm3 and its specific content in preforms was within the limits 0.66-0.75 g/cm3. The pycnometric density of fibers for “Etan” fabric is 1.26 g/cm3 and its specific content in performs was within the limits 0.50-0.53 g/cm3.
The preforms were densified simultaneously in a GF-2 device by a radially moving pyrolysis zone. The rate of pyrolysis zone movement was 0.25mm/h which, as experience shows, provides the maximum densification of porous preforms with any filler.
We also checked the conformity of calculated and actual values of density in real brake disks of different reinforcement structures whose densification we have been concerned with for many years. In all cases the deviation did not exceed 10-12%.