CARBON FIBER BUNDLE

Abstract

 A carbon fiber bundle including fibers having a surface on which a fibril(s) is/are present along the fiber axis direction, wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 14%, and wherein the number of filaments is 48,000 to 60,000. An object of the present invention is to provide a carbon fiber bundle having excellent strength, and excellent process stability during further processing, while having high total fineness.

Description

TECHNICAL FIELD

[0001] The present invention relates to a carbon fiber bundle having excellent strength, and excellent process stability during further processing, while having high total fineness.

BACKGROUND ART

[0002] Since carbon fiber bundles have high specific strength and high specific elastic modulus, they have been widely used as reinforcing fibers for composite materials in the field of aerospace and the like. In recent years, carbon fiber bundles have been used also for industrial applications such as automotive parts and wind power generation. In particular, since wind power generation requires lightweightness and rigidity, carbon fiber bundles having excellent specific elastic modulus are often used therefor. Thus, in recent years, there has been an increasing demand for carbon fiber bundles for wind power generation.

[0003] In industrial applications, reduction of the cost of the final composite material product is strongly demanded. Therefore, emphasis is placed not only on reduction of the cost of the carbon fiber bundles, but also on processability of further processing in the production of intermediate substrates such as prepregs, towpregs, woven fabrics, and sheet molding compounds (SMCs), and carbon fiber-reinforced composites such as pultrusions. In order to increase the processability of further processing, it is especially important for a carbon fiber bundle to hardly generate fuzz, to have excellent spreadability, and not to cause fracture of the entire carbon fiber bundle or single carbon fibers during the course of unwinding from a bobbin and running through the production process, and to show favorable process stability.

[0004] In industrial applications for which reduction of the cost is strongly demanded, the so-called large-tow carbon fiber bundles, which have a single fiber fineness of not less than 0.6 dtex and whose number of filaments is not less than 40,000, are often used. In an attempt to reduce the cost of large-tow carbon fiber bundles, for example, polyacrylonitrile-based precursor fibers prepared by application of the wet spinning method, which is highly productive, are used to increase the processing unit and the processing density, to thereby increase the productivity, or simple equipment for acrylic fibers for clothing is applied. Compared to regular-tow carbon fiber bundles, whose number of filaments is 12,000 to 24,000, large-tow carbon fiber bundles are advantageous in terms of the cost. However, at present, large-tow carbon fiber bundles are inferior to regular-tow carbon fiber bundles in terms of the strand strength, fuzz number, and processability of further processing of carbon fiber bundles, so that further improvement that can be achieved without deterioration of the productivity has been demanded.

[0005] In order to address such an issue, Patent Document 1 proposes a technique for a large-tow carbon fiber bundle in which a polyacrylonitrile-based precursor fiber bundle having a dynamic viscoelastic property and a silicon content within particular ranges is subjected to heat treatment and drawing under particular conditions, to produce a high-quality large-tow carbon fiber bundle with high productivity.

[0006] Patent Documents 2, 3, and 4 propose techniques in which the composition of an oil agent to be applied to, and the amount of such an oil agent to be attached to, a polyacrylonitrile-based precursor fiber bundle are controlled to improve the processability in the stabilization process, to thereby improve the strand strength and the quality of the carbon fibers obtained.

[0007] Patent Documents 5 and 6 propose techniques in which a particular defect that forms a fracture origin of the resulting carbon fiber bundle is controlled within a certain range, to improve the strand strength and the quality of the resulting carbon fiber bundle.

PRIOR ART DOCUMENTS

[Patent Documents]

[0008] 

[Patent Document 1] JP 2006-299439 A

[Patent Document 2] JP 2016-199824 A

[Patent Document 3] JP 2021-50428 A

[Patent Document 4] JP 2021-123812 A

[Patent Document 5] JP 2019-112730 A

[Patent Document 6] JP 2020-153051 A

SUMMARY OF THE INVENTION

PROBLEMS TO BE SOLVED BY THE INVENTION

[0009] However, the prior art has the following problems.

[0010] Patent Document 1 discloses that improved strand strength is achieved in a large-tow carbon fiber bundle, and that generation of fuzz during its production process can be effectively suppressed. However, there is neither disclosure nor suggestion for improvement of the process stability during further processing of the fiber bundle. Further, although the kinematic viscosity of an oil agent is important for suppressing voids of not less than 100 nm and hence for effectively suppressing fuzz generation during further processing, the kinematic viscosity of the oil agent (which was, for example, 450 mm²/sec in Example 1) is insufficient, leading to insufficiency of the improving effect by the oil agent, which has been problematic.

[0011] Patent Document 2 discloses that improved strand strength is achieved in a regular-tow carbon fiber bundle, and that generation of fuzz during its production process is effectively suppressed. However, there is neither disclosure nor suggestion for improvement of the process stability during further processing of the fiber bundle. Further, although the kinematic viscosity of an oil agent is important for suppressing voids of not less than 100 nm and hence for effectively suppressing fuzz generation during further processing, the kinematic viscosity of the oil agent is insufficient (not more than 5000 mm²/sec), leading to insufficiency of the improving effect by the oil agent, which has been problematic.

[0012] Patent Document 3 discloses that improved strand strength is achieved in a regular-tow carbon fiber bundle, and that generation of fuzz during its production process is effectively suppressed. However, there is neither disclosure nor suggestion for improvement of the process stability during further processing of the fiber bundle. Moreover, although the document discloses that the kinematic viscosity of an oil agent, important for suppressing voids of not less than 100 nm and hence for effectively suppressing fuzz generation during further processing, is 3500 to 20,000 mm²/sec, there is no specific disclosure on the void state of the polyacrylonitrile-based precursor fibers to which the oil agent is applied, and the draw ratio in warm water, which is important for reduction of voids, is insufficient (the ratio was, for example, 3.5 in Example 1), so that the improving effect is insufficient, which has been problematic.

[0013]  Furthermore, this invention is based on the use of a polyacrylonitrile-based precursor fiber bundle containing only a small number of filaments, obtained by the dry-jet wet spinning method. Therefore, in cases where the method is applied to a large-tow carbon fiber bundle in which fibrils are present on the fiber surface, and which has a large processing unit and a high processing density, an excessive silicon content leads to insufficiency of the effect that suppresses voids of not less than 100 nm, so that fuzz generation during further processing cannot be effectively suppressed, which has been problematic.

[0014] Patent Document 4 discloses that improved strand strength is achieved in a regular-tow carbon fiber bundle, and that generation of fuzz during its production process is effectively suppressed. However, there is neither disclosure nor suggestion for improvement of the process stability during further processing of the fiber bundle. Moreover, the kinematic viscosity of the oil agent, important for suppressing voids of not less than 100 nm, and hence for effectively suppressing fuzz generation during further processing, is insufficient. In addition, there is neither disclosure nor suggestion about the void state of the polyacrylonitrile-based precursor fibers to which the oil agent is applied, and the draw ratio in warm water, which is important for reduction of voids, is insufficient (the ratio was, for example, 3.5 in Example 1), so that the improving effect is insufficient, which has been problematic.

[0015] Furthermore, this invention is based on the use of a polyacrylonitrile-based precursor fiber bundle containing only a small number of filaments, obtained by the dry-jet wet spinning method. Therefore, in cases where the method is applied to a large-tow carbon fiber bundle in which fibrils are present on the fiber surface, and which has a large processing unit and a high processing density, an excessive silicon content leads to insufficiency of the effect that suppresses voids of not less than 100 nm, so that fuzz generation during further processing cannot be effectively suppressed, which has been problematic.

[0016] Patent Documents 5 and 6 disclose that a particular defect that appears on a fracture surface of a carbon fiber bundle resulting from a single fiber tensile test at a gauge length of 10 mm is controlled to improve the strand strength of a regular-tow carbon fiber bundle and to effectively suppress generation of fuzz during its production process. However, there is neither disclosure nor suggestion for improvement of the process stability during further processing of the fiber bundle. Further, although the defect that appears on the fracture surface resulting from the single fiber tensile test of the carbon fiber bundle at a gauge length of 10 mm showed a good correlation with the achievement of the strand strength, the above defect is different from the defect that causes the generation of fuzz during further processing in terms of the type and the existence probability, so that the above defect does not contribute to identification or improvement of the cause of the generation of fuzz, which has been problematic.

[0017] Thus, although the prior art has proposed techniques for improvement of the strand strength and suppression of the generation of fuzz during the carbon fiber bundle production process, there has been neither a disclosure on a technique for improving the generation of fuzz during further processing of a carbon fiber bundle, nor a disclosure on the defect that causes the generation of fuzz during further processing of a carbon fiber bundle or on a technique for identification of such a defect. Therefore, the fundamental reduction of the generation of fuzz during further processing of a carbon fiber bundle has been difficult.

[0018]  Further, regarding suppression of voids and adhesion, which suppression is effective for suppressing the generation of fuzz during further processing, there has been no technique based on a large-tow carbon fiber bundle that comprehensively proposes control of the surface shape and voids of polyacrylonitrile-based precursor fibers, and the composition of an oil agent suitable therefor and control of the amount of such an agent to be attached. As a result, improvement of the generation of fuzz during further processing of a large-tow carbon fiber bundle has been insufficient.

MEANS FOR SOLVING THE PROBLEMS

[0019] In order to solve the above problems, the carbon fiber bundle of the present invention has the following constitution.

(1) A carbon fiber bundle comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,

wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 14%, and

wherein the number of filaments is 48,000 to 60,000.

(2) The carbon fiber bundle according to (1), comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,
wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), when the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is A (%), and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is B (%), A and B satisfy the relationship of the following Formula (1):

(3) The carbon fiber bundle according to (1), comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,
wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 15%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 10%.

(4) The carbon fiber bundle according to any one of (1) to (3), having a fibril width of 100 to 600 nm.

(5) The carbon fiber bundle according to any one of (1) to (4), having a strand strength of 4.5 to 6.0 GPa.

EFFECT OF THE INVENTION

[0020] The present invention enables production of a carbon fiber bundle having excellent strength, and excellent process stability during further processing, while having high total fineness, wherein mechanical properties are likely to be achieved when the carbon fiber bundle is prepared into a carbon fiber-reinforced composite material.

BRIEF DESCRIPTION OF THE DRAWINGS

[0021] 

Fig. 1 is a scanning electron microscopic (SEM) image of a fracture surface of a carbon fiber. Radial streaks that converge to a single point can be seen.

Fig. 2 is a magnified image of an area around the fracture origin of a fracture surface of another carbon fiber. A fibrillar substance with an aspect ratio of 3.0 to 10.0 can be seen.

Fig. 3 is a magnified image of an area around the fracture origin of a fracture surface of another carbon fiber. A void of not less than 100 nm can be seen.

Fig. 4 illustrates the aspect ratio of a fibrillar substance.

Fig. 5 illustrates the aspect ratio of a fibrillar substance.

Fig. 6 illustrates the aspect ratio of a fibrillar substance.

MODE FOR CARRYING OUT THE INVENTION

[0022] The carbon fiber bundle of the present invention comprises fibers having a surface on which a fibril(s) is/are present along the fiber axis direction, wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 14%, and wherein the number of filaments is 48,000 to 60,000.

[0023] In the carbon fiber bundle of the present invention, fibrils need to be present along the fiber axis direction on fiber surfaces. The fibrils have a width of preferably 100 to 600 nm, more preferably 200 to 400 nm. By the presence of the fibrils along the fiber axis direction on the carbon fiber surfaces, the coefficient of friction can be within an appropriate range. As a result, generation of fuzz during further processing can be reduced, and the carbon fiber bundle can have favorable spreadability. Further, by the presence of the fibrils, adhesion of fineness to each other especially in the early stage of stabilization can be prevented, so that the amount of a silicone-containing oil agent attached, which leads to formation of voids, can be reduced to allow reduction of voids. The presence and the width of the fibrils can be confirmed by observation of the fiber surfaces using a scanning electron microscope. The fibril width can be determined by observing 10 fibers at a magnification of ×25,000 to measure the width in the direction perpendicular to the fiber axis at 10 positions per fiber, and then calculating the arithmetic average of the measured values. The presence and the width of the fibrils can be controlled, for example, by employing wet spinning as the spinning method for the polyacrylonitrile-based precursor fiber bundle, by the coagulation conditions, or by the draw ratio in warm water.

[0024] In the carbon fiber bundle of the present invention, in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 14%.

[0025] The ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present is preferably 1 to 15%, more preferably 1 to 13%, still more preferably 1 to 10%. The ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present is preferably 1 to 10%, more preferably 1 to 6%, still more preferably 1 to 4%. For either type of fracture surface, the smaller the ratio of the number of fibers having such a fracture surface, the more easily the effect of the present invention can be obtained. However, in the industrial production scale, a sufficient effect can be obtained by decreasing the ratio to 1% in most cases.

[0026]  In cases where the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present is not more than 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present is not more than 14%, generation of fuzz during further processing can be suppressed, so that favorable process stability can be obtained.

[0027] In the carbon fiber bundle of the present invention, when the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present is A (%), and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present is B (%), A and B preferably satisfy the relationship of the following Formula (1).

[0028] A and B more preferably satisfy the following Formula (2), still more preferably satisfy the following Formula (3).

[0029] In cases where A and B satisfy the preferred relationship described above, both the ratio of the number of fibers having a fracture surface where the fibrillar substance(s) is/are present and the ratio of the number of fibers having a fracture surface where the void(s) is/are present are low. Therefore, the number of defects in the carbon fibers is small, and hence generation of fuzz during further processing can be suppressed so that favorable process stability can be obtained.

[0030] The strength of single fibers of the carbon fibers is dependent on the sizes, the types, and the existence probabilities of defects. A change in the gauge length results in changes in the sizes and the types of the defects included along the gauge length, so that the strength changes. The strand strength is commonly used as an index of the strength of a carbon fiber bundle, and shows a good correlation with the single-fiber strength at a gauge length of about 10 mm. On the other hand, in a study by the present inventors, the process stability during further processing was found to be correlated with the ratio of a particular defect at a gauge length of 50 mm. The reason why the process stability during further processing is correlated with a longer gauge length than that in the case of the strand strength is not necessarily unclear. However, this could be due to the fact that a serious defect with a relatively low existence probability causes fiber fracture upon application of tension or abrasion during the further processing.

[0031] The fibrillar substance with an aspect ratio of 3.0 to 10.0 is a defect that is thought to be generated in the course of production of the carbon fiber bundle, due to adhesion of single fibers to each other followed by their peeling. A polyacrylonitrile-based precursor fiber is an aggregate of fibrils along the fiber axis direction, and the above-described adhesion and peeling tend to cause destruction in the unit of fibrils. The result of a study by the present inventors indicates that fibrillar substances representing defects that are thought to be derived by the destruction in such a unit often have an aspect ratio of 3.0 to 10.0. The ratio of the number of fibers having a fracture surface where the fibrillar substance(s) is/are present can be calculated according to the method described later.

[0032] For controlling the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), it is important to suppress adhesion of single fibers to each other during the process of production of the carbon fiber bundle. In a generally known method for suppressing adhesion of single fibers to each other, a silicone-based oil agent is applied to polyacrylonitrile-based precursor fibers. However, a study by the present inventors showed that the simple use of a silicone-based oil agent is insufficient. A preferred control method is described later as a preferred production method for the carbon fiber bundle.

[0033] The void of not less than 100 nm is a defect that is thought to be formed in a case where a void present in a polyacrylonitrile-based precursor fiber remains without disappearance even after the subsequent stabilization process, the pre-carbonization process, and the carbonization process. Examples of the cause why the void present in the polyacrylonitrile-based precursor fiber does not disappear include the fact that the size of the void present in the polyacrylonitrile-based precursor fiber before the application of the oil agent is large, and the fact that the oil agent infiltrates into the void upon the application of the oil agent, to inhibit densification. Based on the result of a study by the present inventors, the size of the void generated by the above mechanism is often not less than 100 nm. The ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present can be calculated according to the method described later.

[0034] For controlling the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), it is important, for example, to reduce the size of voids present in the polyacrylonitrile-based precursor fibers, and to suppress infiltration of the oil agent into the voids. A preferred control method is described later as a preferred production method for the carbon fiber bundle.

[0035] The number of filaments in the carbon fiber bundle of the present invention is 48,000 to 60,000, preferably 50,000 to 55,000. The number of filaments is the number of single fibers constituting the carbon fiber bundle. The larger the number, the higher the productivity of the carbon fiber-reinforced composite material. However, in cases where the number of filaments is too large, spreadability of the carbon fiber bundle may decrease, and mechanical properties of the obtained carbon fiber-reinforced composite material may decrease from the viewpoint of the resin impregnating property. In cases where the number of filaments is 48,000 to 60,000, excellent productivity can be achieved in the molding of the composite material, so that the carbon fiber bundle can be favorably used in industrial applications. The number filaments can be controlled based on the number of holes of the spinneret in the fiber production process of the polyacrylonitrile-based precursor fiber bundle, and dividing and combining of fibers.

[0036] The strand strength of the carbon fiber bundle of the present invention is preferably 4.5 to 6.0 GPa, more preferably 4.6 to 6.0 GPa, still more preferably 4.8 to 6.0 GPa. The strand strength can be measured by the method described later and, in cases where the strand strength is 4.5 to 6.0 GPa, the carbon fiber bundle can be favorably used for industrial applications such as blade materials for windmills, reinforcing materials for pressure vessels, and structural parts for automobiles.

[0037] A method of producing a carbon fiber bundle preferred for obtaining the carbon fiber bundle of the present invention is described below.

[0038] The carbon fiber bundle of the present invention is preferably produced by

providing a polyacrylonitrile-based precursor fiber bundle comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction, the fiber bundle having a number of filaments of 48,000 to 60,000, wherein the fiber bundle is obtained by a method of producing a polyacrylonitrile-based precursor fiber bundle, the method comprising the steps of: subjecting a polyacrylonitrile-based polymer to wet spinning; drawing the resulting fibers at a ratio of 5.0 to 8.0 in warm water at 30 to 99°C; and applying an oil agent containing a silicone having a kinematic viscosity at 25°C of 6000 to 20,000 mm²/sec thereto;

subjecting the polyacrylonitrile-based precursor fiber bundle to stabilization in an oxidizing atmosphere at 200 to 300°C while the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ is controlled at 0.06 to 0.09% by mass;

performing pre-carbonization in an inert atmosphere at 500 to 1200°C; and

performing carbonization in an inert atmosphere at 900 to 2000°C.

[0039] In the method of producing the polyacrylonitrile-based precursor fiber bundle described above, the polyacrylonitrile-based polymer means a polymer containing at least acrylonitrile as a major component of the polymer unit, wherein the major component means a component that accounts for 90 to 100% by mass of the polymer unit.

[0040] The polyacrylonitrile-based polymer preferably contains a copolymer component such as itaconic acid, acrylamide, or methacrylic acid, for example, from the viewpoint of improvement of the fiber production efficiency, and from the viewpoint of efficiently carrying out the stabilization. In the production of the polyacrylonitrile-based precursor fiber bundle, the method of producing the polyacrylonitrile-based polymer can be selected from known polymerization methods such as solution polymerization and aqueous suspension polymerization. The polyacrylonitrile-based polymer is provided as a spinning dope solution in which the polymer is dissolved in a solvent, for the production of the polyacrylonitrile-based precursor fibers. The solvent used for the spinning dope solution can be selected from known solvents in which polyacrylonitrile is soluble, such as dimethyl sulfoxide, dimethylformamide, and dimethylacetamide, aqueous nitric acid solutions, aqueous zinc chloride solutions, and aqueous sodium rhodanide solutions.

[0041] The method of producing a polyacrylonitrile-based precursor fiber bundle described above comprises a step of subjecting a polyacrylonitrile-based polymer to wet spinning. The wet spinning herein means a spinning method in which the polyacrylonitrile-based polymer is directly discharged into a coagulation bath through a spinneret. By the application of the wet spinning, a fiber surface shape having fibrils suitable for the production of the carbon fiber bundle of the present invention can be obtained.

[0042] In the method of producing the polyacrylonitrile-based precursor fiber bundle described above, the number of holes of the spinneret is preferably 3000 to 200,000 from the viewpoint of achievement of the above-described number of filaments of the carbon fiber bundle. By dividing and combining of the fibers, a polyacrylonitrile-based precursor fiber bundle with a predetermined number of filaments can be obtained.

[0043] In the method of producing the polyacrylonitrile-based precursor fiber bundle described above, the composition of the coagulation bath preferably contains a solvent used as the solvent of the spinning dope solution, such as dimethyl sulfoxide, dimethylformamide, or dimethylacetamide, and the so-called coagulation-promoting component. The solvent is more preferably dimethyl sulfoxide or dimethylformamide from the viewpoint of allowing the formation of appropriate fibrils on the surface of the polyacrylonitrile-based precursor fibers without deteriorating the productivity. As the coagulation-promoting component, a component in which the polyacrylonitrile-based polymer is insoluble, and which is compatible with the solvent used for the spinning dope solution, can be used. Water is preferably used.

[0044] The method of producing a polyacrylonitrile-based precursor fiber bundle described above comprises a step of drawing the fibers at a ratio of 5.0 to 8.0 in warm water at 30 to 99°C. The fibers obtained by the wet spinning of the polyacrylonitrile-based polymer are drawn in warm water while the solvent is washed away therein. The washing and the drawing may be carried out either at the same time or separately as long as the fibers are drawn at a ratio of 5.0 to 8.0 in warm water at 30 to 99°C. In cases where the fibers are drawn in warm water, the drawing is preferably carried out stepwise in a plurality of warm water baths. The temperature of the warm water is preferably 50 to 99°C, more preferably 70 to 99°C. The higher the temperature of the warm water, the more easily the fibers can be drawn, but the more likely the fibers are to adhere to each other. Therefore, it is preferred to use a plurality of warm water baths to increase the temperature of the warm water in a stepwise manner. The draw ratio in the warm water is preferably 5.5 to 8.0, more preferably 6.0 to 8.0. The higher the draw ratio, the smaller the number of voids present in the obtained polyacrylonitrile-based precursor fiber bundle, which is preferred for the production of the carbon fiber bundle of the present invention. In cases where the draw ratio is not more than 8.0, breakage of fibers due to the drawing can be suppressed to enable stable production of a polyacrylonitrile-based precursor fiber bundle with high quality.

[0045] The method of producing a polyacrylonitrile-based precursor fiber bundle described above comprises a step of applying an oil agent containing a silicone having a kinematic viscosity at 25°C of 6000 to 20,000 mm²/sec. The kinematic viscosity of the silicone at 25°C is preferably 10,000 to 20,000 mm²/sec, more preferably 15,000 to 18,000 mm²/sec. In cases where the kinematic viscosity of the silicone at 25°C is not less than 6000 mm²/sec, when the oil agent is applied to a fiber bundle in which the number of voids is small, drawn at a ratio of 5.0 to 8.0 in warm water at 30 to 99°C, adhesion of fibers to each other can be suppressed, and moreover, infiltration of the oil agent into the voids can be effectively suppressed. In cases where the kinematic viscosity of the silicone at 25°C is not more than 20,000 mm²/sec, uneven attachment can be suppressed, so that a stable strand strength can be achieved in the obtained carbon fiber bundle. The kinematic viscosity at 25°C can be measured according to JIS-Z-8803 (2011) or ASTM D 445-46T using, for example, an Ubbelohde viscometer.

[0046] The silicone used in the method of producing a polyacrylonitrile-based precursor fiber bundle described above is preferably an amino-modified silicone from the viewpoint of its uniform attachment. The amino-modified silicone is a silicone containing polydimethylsiloxane as a basic structure wherein side-chain methyl groups are partially modified with amino groups. The amino-modified silicone used may contain other modifying groups added thereto in addition to the amino groups. Although the amino groups as modifying groups may be either of a monoamine type or a polyamine type, polyamine-type amino groups are preferred from the viewpoint of promoting cross-linking. Diamine-type amino groups are more preferably used.

[0047]  The amino equivalent, which is an index of the amount of amino groups (NH₂), in the amino-modified silicone is preferably 1000 to 14,000 g/mol, more preferably 1500 to 6000 g/mol, still more preferably 2000 to 4000 g/mol. In cases where the amino equivalent is not less than 1000 g/mol, uneven attachment due to excessive progress of cross-linking can be suppressed, and a stable strand strength can be achieved in the obtained carbon fiber bundle as a result. In cases where the amino equivalent is not more than 14,000 g/mol, the silicone can be sufficiently cross-linked, and a stable strand strength can be achieved in the obtained carbon fiber bundle as a result. The amino equivalent can be measured by a known method such as neutralization titration. The amino equivalent can be controlled, for example, by the amount of amine added during the polymerization of the amino-modified silicone.

[0048] The oil agent used in the method of producing a polyacrylonitrile-based precursor fiber bundle described above may contain a surfactant, an antioxidant, an antistatic agent, a lubricant, or the like in addition to the silicone having a kinematic viscosity at 25°C of 6000 to 20,000 mm²/sec.

[0049] In the method of producing a polyacrylonitrile-based precursor fiber bundle described above, dry-heat treatment is preferably carried out by a known method following the wet spinning, the drawing in warm water, and the application of the oil agent. By carrying out the dry-heat treatment, densification of the voids can be promoted, which is preferred. The dry-heat treatment temperature is preferably 120 to 180°C.

[0050] In the method of producing a polyacrylonitrile-based precursor fiber bundle described above, the fibers subjected to the dry-heat treatment may be further drawn in pressurized steam or under dry heat. However, the total draw ratio is preferably 5.0 to 8.0. The total draw ratio herein is the ratio calculated by multiplying the draw ratio in warm water by the draw ratio after the dry-heat treatment. In cases where the draw ratio in warm water is not less than 5.0, a polyacrylonitrile-based precursor fiber bundle in which the number of voids is small, which is suitable for the production of the carbon fiber bundle of the present invention, can be obtained. In cases where the total draw ratio, taking the draw ratio after the dry-heat treatment into account, is not more than 8.0, breakage of fibers due to the drawing can be suppressed to enable stable production of a polyacrylonitrile-based precursor fiber bundle with high quality. The draw ratio in warm water is more preferably the same as the total draw ratio. In other words, it is more preferred not to carry out drawing after the dry-heat treatment. In cases where a fiber bundle with a number of filaments of not less than 48,000 is drawn in pressurized steam or under dry heat, the temperature tends be uneven in the bundle. Therefore, the drawing is preferably carried out only in warm water.

[0051] The single fiber fineness of the polyacrylonitrile-based precursor fiber bundle in the production process of the carbon fiber bundle is preferably 1.10 to 2.40 dtex, more preferably 1.20 to 2.20 dtex. The single fiber fineness is the mass per unit length of a single fiber. In cases where the single fiber fineness is not less than 1.10 dtex, the carbon fiber bundle can be obtained with sufficient productivity, while in cases where the single fiber fineness is not more than 2.40 dtex, unevenness of treatment in the heat treatment after the stabilization process can be reduced, resulting in a carbon fiber bundle having high mechanical properties. The single fiber fineness can be evaluated by measuring the mass per unit length. The single fiber fineness can be controlled by the discharge rate and the draw ratio in the fiber production process.

[0052] In the production process of the carbon fiber bundle, the circularity of single-fiber cross-section of the polyacrylonitrile-based precursor fiber bundle is preferably 0.86 to 0.98, more preferably 0.87 to 0.96, still more preferably 0.87 to 0.93. The circularity of single-fiber cross-section is defined as follows based on the perimeter L and the area A of the single-fiber cross-section:

[0053] In cases where the circularity of single-fiber cross-section is 0.86 to 0.98, the obtained carbon fibers can have both an excellent bundling ability and high abrasion resistance, which is preferred from the viewpoint of improving the process stability during further processing of the obtained carbon fiber bundle. The circularity of single-fiber cross-section of the polyacrylonitrile-based precursor fiber bundle can be evaluated from an image of a cross-section prepared by perpendicularly cutting a single fiber by the following method. The circularity of single-fiber cross-section of the polyacrylonitrile-based precursor fiber bundle can be controlled by the shape of the discharge holes of the spinneret in the fiber production process, conditions in the coagulation process, and the like.

[0054] In the production process of the carbon fiber bundle, the obtained polyacrylonitrile-based precursor fiber bundle is preferably subjected to stabilization in an oxidizing atmosphere at 200 to 300°C while the silicon content is controlled at 0.06 to 0.09% by mass until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³.

[0055] In the production method of the carbon fiber bundle described above, in cases where the polyacrylonitrile-based precursor fiber bundle is subjected to the heat treatment in an oxidizing atmosphere at a temperature of not less than 200°C, a sufficiently stabilized fiber bundle can be produced. Therefore, generation of fuzz due to insufficient stabilization can be suppressed, and hence the obtained carbon fiber bundle shows excellent process stability during further processing. In cases where the temperature at which the stabilization is carried out is not more than 300°C, the heat generation rate is not too high, so that temperature spots in the stabilized fiber bundle can be reduced. Therefore, a carbon fiber bundle having excellent mechanical properties can be obtained.

[0056] The temperature for the stabilization may be measured by inserting a thermometer such as a thermocouple thermometer into the oxidation oven to measure the furnace temperature. In cases where a temperature spot or temperature distribution is found after measurement of the furnace temperature at several positions, a simple average temperature is calculated. The stabilization temperature can be controlled by the output of heating in a heating method used for a known oxidation oven. For example, in cases of an oxidation oven with internal air circulation, the output of the heater used for heating the oxidizing atmosphere may be changed.

[0057] In the production method of the carbon fiber bundle described above, the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ is more preferably 0.07 to 0.08% by mass. The density of the polyacrylonitrile-based precursor fiber bundle is generally 1.14 to 1.18 g/cm³, and the density generally exceeds 1.30 g/cm³ after carrying out the stabilization. The density specified in the present invention, 1.21 to 1.23 g/cm³, means the range in the early stage of the stabilization. In the early stage of the stabilization, the polyacrylonitrile-based precursor fiber bundle undergoes the treatment at a temperature of as high as not less than 200°C in a state where the stabilization of the structure is incomplete. Therefore, adhesion of single fibers to each other tends to occur, and in addition, densification of the remaining voids tends to occur to cause their disappearance. In polyacrylonitrile-based precursor fibers in which fibrils are present on the fiber surface, in cases where the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ is not less than 0.06% by mass, adhesion of single fibers to each other in the early stage of the stabilization can be suppressed. Therefore, fibrillar substances due to adhesion and peeling of the obtained carbon fibers can be suppressed. In cases where the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ is not more than 0.09% by mass, densification of voids in the early stage of the stabilization is less likely to be inhibited, so that voids in the obtained carbon fibers can be suppressed.

[0058] The silicon content until the time when the density of the fiber bundle becomes 1.21 to 1.23 g/cm³ can be measured as follows. From the oxidation oven in which the polyacrylonitrile-based precursor fibers are continuously subjected to the stabilization, a stabilized fiber bundle in the early stage of the stabilization is sampled. The stabilized fiber bundle is cut at 1-m intervals as measured from the inlet of the oxidation oven, and the density and the silicon content are measured by the later-mentioned method. A portion where the density of the stabilized fiber bundle is 1.22 ± 0.01 g/cm³ is identified, and the silicon content in this portion is defined as the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³.

[0059] The silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ can be controlled by the amount of the oil agent attached to the polyacrylonitrile-based precursor fiber bundle, and the treatment temperature in the early stage of the stabilization. The higher the treatment temperature in the early stage of the stabilization, the lower the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³ can be. For controlling, within the appropriate range, the silicon content until the time when the density of the stabilized fiber bundle becomes 1.21 to 1.23 g/cm³, it is important to apply a sufficient amount of the oil agent to the polyacrylonitrile-based precursor fiber bundle such that the oil agent can be uniformly attached thereto, and then to appropriately control the temperature in the early stage of the stabilization.

[0060] After the production process and the stabilization process of the polyacrylonitrile-based precursor fiber bundle, pre-carbonization is carried out. In the pre-carbonization process, the obtained stabilized fiber bundle is subjected to heat treatment in an inert atmosphere at a maximum temperature of 500 to 1200°C, preferably until the density becomes 1.5 to 1.8 g/cm³. The draw ratio in the pre-carbonization process is preferably 1.00 to 1.30, more preferably 1.10 to 1.25.

[0061] The pre-carbonization is followed by carbonization. In the carbonization process, the pre-carbonized fiber bundle is subjected to carbonization at a maximum temperature of 900 to 2000°C in an inert atmosphere. The draw ratio in the carbonization process is preferably 0.94 to 1.05, more preferably 0.96 to 1.02.

[0062] The thus obtained carbon fiber bundle is preferably subjected to oxidation treatment to introduce oxygen-containing functional groups in order to improve the adhesion to the matrix resin. As a method of the oxidation treatment, gas-phase oxidation, liquid-phase oxidation, or liquid-phase electrochemical oxidation is carried out. Liquid-phase electrochemical oxidation is preferably employed from the viewpoint of the fact that it is highly productive and capable of uniform treatment. The method of the liquid-phase electrochemical oxidation is not limited, and it may be carried out by a known method.

[0063] After the electrochemical treatment, sizing treatment may be carried out in order to impart a bundling ability to the obtained carbon fiber bundle. As the sizing agent, a sizing agent having good compatibility with the matrix resin can be appropriately selected depending on the type of the matrix resin used for the composite material.

<Resin-Impregnated Strand Tensile Test of Carbon Fiber Bundle>

[0064] The tensile strength (strand strength) and the stress-strain curve of a resin-impregnated strand of the carbon fiber bundle are determined according to the "resin-impregnated strand test method" of JIS R7608 (2008). A test piece is prepared by impregnating the carbon fiber bundle with the following resin composition, and performing heat treatment under curing conditions at a temperature of 130°C for 35 minutes.

[Resin Composition]

[0065] 

· 3,4-Epoxycyclohexylmethyl-3,4-epoxy-cyclohexane-carboxylate (100 parts by mass)

· Boron trifluoride monoethylamine (3 parts by mass)

· Acetone (4 parts by mass)

[0066] Six strands are measured, and the arithmetic average of the measurement results is defined as the strand strength of the carbon fibers.

<Measurement of Circularity of Single-Fiber Cross-Section of Polyacrylonitrile-

Based Precursor Fiber Bundle>

[0067] Using a single-edged razor blade, a polyacrylonitrile-based precursor fiber bundle is cut in the direction perpendicular to the fiber axis direction. Using a scanning electron microscope (SEM) "S-4800", manufactured by Hitachi High-Tech Corporation, the resulting cross-section is observed in the direction perpendicular to the fiber cross-section. In the image obtained, the circumference of the fiber cross-section is selected using image analysis software "ImageJ", and the circularity is calculated according to the following definition based on the perimeter and the area of the calculated fiber cross-section. After measurement of the circularity for five single fibers randomly selected in each field of view, measured circularities of a total of 25 single fibers are averaged to determine the circularity of single fiber cross-section of the polyacrylonitrile-based precursor fiber bundle. The circularity of a single fiber cross-section is defined based on the perimeter L and the area A of the single-fiber cross-section as follows:

<Ratio of Number of Fibers Having Fracture Surface Where Fibrillar Substance(s) with Aspect Ratio of 3.0 to 10.0 Is/Are Present, and Ratio of Number of Fibers Having Fracture Surface Where Void(s) of Not Less Than 100 nm Is/Are Present>

[0068] A tensile test of single carbon fibers is carried out according to JIS R7606 (2000). With a gauge length of 50 mm, a carbon fiber is fixed to a test piece mount using a commercially available cyanoacrylate adhesive. The tensile test is carried out using a tensile tester (in Examples of the present invention, "RTC-1210A", manufactured by A&D) together with a test tool that is designed such that the test can be carried out in water. From the fiber bundle, 150 single fibers to be subjected to the test are randomly extracted. For each of the 150 single fibers extracted, a tensile test is carried out under conditions at a strain rate of 0.4 mm/minute, and both of the single fibers resulting by fracture are collected.

[0069] The fracture surfaces of the single fibers collected are observed using a field emission scanning electron microscope (in the Examples of the present invention, "S-4800", manufactured by Hitachi High-Tech Corporation). For highly accurate observation of small defects, vapor deposition treatment for imparting conductivity is not carried out since the treatment may cause surface unevenness. The observation is carried out at an accelerating voltage of 1 keV and a magnification of ×25,000 to ×50,000. Further, in order to allow easier identification of the presence or absence of small defects, the stage is rotated such that the fracture origin faces the near side, and the stage is tilted by 30° so that the fracture surface can be observed obliquely from above. For example, the observation direction is as in Fig. 2 and Fig. 3. The observation is carried out for all fibers that could be collected.

[0070] On a primary fracture surface formed by tensile fracture of a carbon fiber, traces of the progress of the fracture from the fracture origin remain as radial streaks. In view of this, as illustrated in Fig. 1, the streaks present in the SEM observation image are traced to identify a single point at which the streaks converge. This point is defined as the fracture origin. In cases where the streaks cannot be recognized, or where the streaks can be recognized but the fracture origin cannot be observed due to attachment of dirt near the fracture origin in at least one of the pair of fracture surfaces, the pair of the fracture surfaces are excluded from the evaluation. The number of pairs of fracture surfaces that could be finally observed is defined as the total number of fracture surfaces. In cases where the total number of fracture surfaces does not exceed 100 pairs, 150 single fibers are randomly extracted again from the fiber bundle, and the tensile test and the observation of the fracture surfaces are carried out again. After a total of more than 100 pairs of fracture surfaces are secured, whether or not a fibrillar substance(s) or void(s) is/are present at the fracture origin is judged.

[0071] A fibrillar substance is found as a long and narrow attached substance as shown in Fig. 2. The aspect ratio of the attached substance is calculated as follows. As illustrated in Figs. 4 to 6, the circumscribed circle and the inscribed circle having the maximum diameter are determined for the attached substance. The diameters of the circles are defined as D_A and D_B, respectively. The aspect ratio can be calculated by dividing D_A by D_B. In cases where the calculated aspect ratio is 3.0 to 10.0, the attached substance is judged as a fibrillar substance with an aspect ratio of 3.0 to 10.0. In cases where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present on one or both of a pair of fracture surfaces, the fracture surfaces are regarded as fracture surfaces where a fibrillar substance(s) is/are present. The number of such fracture surfaces is divided by the total number of fracture surfaces, to determine the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present.

[0072] A void is found as a hole as shown in Fig. 3. The diameter of the circumscribed circle of the void is defined as the size of the void. In cases where the size is not less than 100 nm, the void is judged as a void of not less than 100 nm. In cases where the above-described void(s) is/are present on one or both of a pair of fracture surfaces, the fracture surfaces are regarded as fracture surfaces where a void(s) is/are present. The number of such fracture surfaces is divided by the total number of fracture surfaces, to determine the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present.

[0073] In the quantification of the aspect ratio of a fibrillar substance or the size of a void, an image is read by image analysis software, and a ruler tool is used.

<Density of Stabilized Fiber Bundle>

[0074] From the central portion of each 1-m piece of a stabilized fiber bundle in the early stage of stabilization collected from the oxidation oven, 1 to 3 g of a sample is taken, and the sample is absolutely dried at 120°C for 2 hours. Subsequently, the absolute dry mass D (g) is measured, and then the sample is impregnated with ethanol while allowing sufficient defoaming, followed by measuring the mass of the fiber bundle in the ethanol solvent bath E (g), and calculating the density as follows:

ρ is the ethanol density at the measurement temperature.

<Silicon Content of Stabilized Fiber Bundle>

[0075] From the central portion of each 1-m piece of a stabilized fiber bundle in the early stage of stabilization collected from the oxidation oven, a sample is taken. The sample is wound around a "Teflon" (registered trademark) plate at 0.03 to 0.07 g per 1 cm², and the silicon content is quantified using a fluorescent X-ray analyzer. Based on a calibration curve prepared using a reference substance with a known silicon content, the measured value of fluorescent X-ray is converted to the silicon content, rather than directly using the measured value of fluorescent X-ray.

<Evaluation of Processability of Further Processing>

[0076] A bobbin of a carbon fiber bundle is placed in a creel, and the fiber bundle is drawn at a tension of 1.6 mN/dtex. After allowing the fiber bundle to pass through 10 free rollers, the fiber bundle is subjected to abrasion through five fixed guides. The fiber bundle is then taken up by a drive roller at a speed of 10 m/minute to be wound on a winder. The fuzz generated in this process is counted for 10 minutes immediately before the drive roller, and evaluation is carried out according to the following index.

A: Less than 10 pieces/m

A: Not less than 10 pieces/m, and less than 50 pieces/m

A: Not less than 50 pieces/m

EXAMPLES

[0077] The present invention is described below more concretely by way of Examples. However, the present invention is not limited thereto.

(Example 1)

[0078] A polyacrylonitrile-based copolymer composed of acrylonitrile, itaconic acid, and methyl acrylate was polymerized by a solution polymerization method using dimethyl sulfoxide as a solvent, to obtain a spinning dope solution. The obtained spinning dope solution was subjected to a wet spinning method in which the solution was introduced from a fiber-forming spinneret with a number of holes of 50,000 into a coagulation bath composed of an aqueous dimethyl sulfoxide solution, to prepare a coagulated fiber bundle. The fiber bundle was then introduced into a plurality of warm water baths at 70 to 99°C, to wash away the solvent and to perform drawing at a ratio of 7.0. Subsequently, an amino-modified silicone oil agent having a kinematic viscosity at 25°C of 15,000 mm²/sec was applied to the fiber bundle drawn in the warm water, and dry-heat treatment was carried out using a heating roller at 130°C, to obtain a polyacrylonitrile-based precursor fiber bundle with a number of filaments of 50,000, having a single fiber fineness of 1.40 dtex. The obtained polyacrylonitrile-based precursor fiber bundle was subjected to stabilization at 220 to 250°C while the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was controlled at 0.075% by mass. Pre-carbonization was carried out under conditions at a maximum temperature of 800°C, and then carbonization was carried out under conditions at a maximum temperature of 1400°C, to obtain a carbon fiber bundle. Properties of the obtained carbon fiber bundle are shown in Table 1.

(Example 2)

[0079] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.088% by mass. Properties of the obtained carbon fiber bundle are shown in Table 1.

(Example 3)

[0080] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the dry-heat treatment was further followed by drawing at a ratio of 1.2 using a heating roller at 180°C to achieve a total draw ratio of 8.4.

(Example 4)

[0081] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the fiber bundle was drawn at a ratio of 6.0 in a plurality of warm water baths at 70 to 99°C in the fiber production process, and that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.085% by mass in the stabilization process. Properties of the obtained carbon fiber bundle are shown in Table 1.

(Example 5)

[0082] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the fiber bundle was drawn at a ratio of 5.0 in a plurality of warm water baths at 70 to 99°C in the fiber production process, and that drawing at a ratio of 1.6 was carried out using a heating roller at 180°C to achieve a total draw ratio of 8.0. Properties of the obtained carbon fiber bundle are shown in Table 1.

(Example 6)

[0083] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the fiber bundle was drawn at a ratio of 6.5 in a plurality of warm water baths at 70 to 99°C in the fiber production process, and that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.085% by mass in the stabilization process. Properties of the obtained carbon fiber bundle are shown in Table 1.

(Comparative Example 1)

[0084] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the draw ratio in warm water was 2.1. Properties of the obtained carbon fiber bundle are shown in Table 1. The ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm was/were present was high, indicating poor processability of further processing.

(Comparative Example 2)

[0085] The same process as in Example 1 was carried out except that the draw ratio in warm water was 9.0. As a result, yarn wrapping and yarn break frequently occurred in the warm-water drawing process, and therefore a polyacrylonitrile-based precursor fiber could not be obtained.

(Comparative Example 3)

[0086] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the kinematic viscosity of the amino-modified silicone at 25°C was 1500 mm²/sec, and that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.160% by mass. Properties of the obtained carbon fiber bundle are shown in Table 1. The ratio of fracture surfaces where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 was/were present, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm was/were present were high, indicating poor processability of further processing.

(Comparative Example 4)

[0087] A carbon fiber bundle was obtained in the same manner as in Comparative Example 3 except that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.081% by mass. Properties of the obtained carbon fiber bundle are shown in Table 1. The ratio of fracture surfaces where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 was/were present, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm was/were present were high, indicating poor processability of further processing.

(Comparative Example 5)

[0088] The same process as in Example 1 was carried out except that the kinematic viscosity of the amino-modified silicone at 25°C was 22,000 mm²/sec. As a result, yarn wrapping and yarn break frequently occurred in the dry-heat treatment process, and therefore a polyacrylonitrile-based precursor fiber could not be obtained.

(Comparative Example 6)

[0089] A carbon fiber bundle was obtained in the same manner as in Example 1 except that the stabilization temperature was changed to 230 to 250°C such that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.053% by mass. Properties of the obtained carbon fiber bundle are shown in Table 1. The ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 was/were present was high, indicating poor processability of further processing.

(Comparative Example 7)

[0090] The same process as in Example 1 was carried out except that the amount of the oil agent attached to the polyacrylonitrile-based precursor fiber bundle was changed such that the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was 0.120% by mass. As a result, yarn wrapping and yarn break frequently occurred in the stabilization process, and therefore a carbon fiber bundle could not be obtained.

(Comparative Example 8)

[0091] A polyacrylonitrile-based copolymer composed of acrylonitrile, itaconic acid, and methyl acrylate was polymerized by a solution polymerization method using dimethyl sulfoxide as a solvent, to obtain a spinning dope solution. The obtained spinning dope solution was subjected to a dry-jet wet spinning method in which the solution was discharged from a fiber-forming spinneret with a number of holes of 3000, and once allowed to pass through air, followed by introduction into a coagulation bath composed of an aqueous dimethyl sulfoxide solution, to prepare a coagulated fiber bundle. The fiber bundle was then introduced into a plurality of warm water baths at 40 to 70°C, to wash away the solvent and to perform drawing at a ratio of 3.5. Subsequently, an amino-modified silicone oil agent having a kinematic viscosity at 25°C of 15,000 mm²/sec was applied to the fiber bundle drawn in the warm water, and dry-heat treatment was carried out using a heating roller at 150°C, followed by drawing at a ratio of 3.7 in steam, to obtain a polyacrylonitrile-based precursor fiber bundle with a filament number of 3000, having a single fiber fineness of 1.40 dtex. Four polyacrylonitrile-based precursor fiber bundles obtained were combined to prepare a fiber bundle with a number of single fibers of 12,000. The resulting fiber bundle was subjected to stabilization under conditions at 220 to 250°C while the silicon content until the time when the density of the stabilized fiber bundle became 1.21 to 1.23 g/cm³ was controlled at 0.075% by mass. Pre-carbonization was carried out under conditions at a maximum temperature of 800°C, and then carbonization was carried out under conditions at a maximum temperature of 1400°C, to obtain a carbon fiber bundle. Properties of the obtained carbon fiber bundle are shown in Table 1. The ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 was/were present was high, indicating poor processability of further processing.

[Table 1]

	Draw Ratio in Warm Water	Total Draw Ratio	Kinematic Viscosity of silicone at 25°C	Circularity of Single Fiber Cross-section of Polyacrylonitrile-based Precursor Fiber Bundle	Silicon Content until the Time When the Density of Stabilized Fiber Bundle Becomes Specific gravity of 1.21 to 1.23	Ratio of the Number of Fibers having a Fracture Surface where a Fibrillar Substance(s) with an Aspect Ratio of 3.0 to 10.0: A	Ratio of the Number of Fibers having a Fracture Surface where Void(s) of not less than 100nm: B	Relationship of the Formula (1) Note)	Fibril Width on the Surface of the Carbon Fiber Bundle	Strand Strength	Processability of Further Processing
	(fold)	(fold)	(mm2/sec)	-	(mass%)	(%)	(%)	-	(nm)	(GPa)
Example 1	7.0	7.0	15,000	0.90	0.075	13	5	S	320	4.6	A
Example 2	7.0	7.0	15,000	-	0.088	10	7	S	310	4.5	A
Example 3	7.0	8.4	15,000	-	0.075	15	3	S	200	4.3	B
Example 4	6.0	6.0	15,000	0.90	0.085	4	11	S	290	4.6	B
Example 5	5.0	8.0	15,000	-	0.075	3	4	S	160	5.2	A
Example 6	6.5	6.5	15,000	-	0.085	8	8	S	140	4.6	A
Comparative Example 1	2.1	2.1	15,000	-	0.075	15	50	U	500	4.0	C
Comparative Example 2	9.0	9.0	15,000	-	-	-	-	-	-	-	C
Comparative Example 3	7.0	7.0	1,500	-	0.160	38	15	U	300	4.2	C
Comparative Example 4	7.0	7.0	1,500	-	0.081	50	13	U	300	4.0	C
Comparative Example 5	7.0	7.0	22,000	-	-	-	-	-	-	-	C
Comparative Example 6	7.0	7.0	15,000	-	0.053	30	5	U	300	4.1	C
Comparative Example 7	7.0	7.0	15,000	-	0.120	-	-	-	-	-	C
Comparative Example 8	3.5	13.0	15,000	1.00	0.075	40	9	U	Absence of Fibril	4.4	C

Note) S : Satisfy U : Unsatisfy

Claims

1. A carbon fiber bundle comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,

wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 20%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 14%, and

wherein the number of filaments is 48,000 to 60,000.

2. The carbon fiber bundle according to claim 1, comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,
wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), when the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is A (%), and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is B (%), A and B satisfy the relationship of the following Formula (1):

3. The carbon fiber bundle according to claim 1, comprising fibers having a surface on which a fibril(s) is/are present along the fiber axis direction,
wherein in a single fiber tensile test at a gauge length of 50 mm in accordance with JIS R7606 (2000), the ratio of the number of fibers having a fracture surface where a fibrillar substance(s) with an aspect ratio of 3.0 to 10.0 is/are present in the fracture origin is 1 to 15%, and the ratio of the number of fibers having a fracture surface where a void(s) of not less than 100 nm is/are present in the fracture origin is 1 to 10%.

4. The carbon fiber bundle according to any one of claims 1 to 3, having a fibril width of 100 to 600 nm.

5. The carbon fiber bundle according to any one of claims 1 to 4, having a strand strength of 4.5 to 6.0 GPa.

Drawing