WO2026019455A1 - Non-circular carbon fiber for cost-specific compressive strength - Google Patents

Abstract

For wind energy and other cost-driven industries, carbon fiber material cost is often prohibitive for its adoption and compressive strength is nearly as critical of a driver as the tensile strength. Optimized non-circular carbon fiber geometries for use in carbon fiber reinforced polymers have been developed to increase the compressive strength per unit cost. Analyses have been performed to identify fiber geometries that have been shown to reduce the material processing costs as well as increase the compressive strength. These studies have identified non-circular carbon fiber geometries with three or more lobes that (1) reduce the diffusion thickness to enable higher throughput and lower costs during production and (2) increase the perimeter and area moment of inertia for a given fiber area. Composites using lobed non-circular carbon fiber geometries with a carbon fiber cross-sectional area greater than 41 micron² offer cost effective alternatives while also increasing compressive strength.

Description

NON-CIRCULAR CARBON FIBER FOR COST-SPECIFIC COMPRESSIVE STRENGTH

REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to patent application U.S. Ser. No. 19/074,986, entitled “NON-CIRCULAR CARBON FIBER FOR COST-SPECIFIC COMPRESSIVE STRENGTH”, filed March 10, 2025, which claims priority to provisional patent application U.S. Ser. No. 63/672,103, entitled "NON-CIRCULAR CARBON FIBER FOR COSTSPECIFIC COMPRESSIVE STRENGTH", filed July 16, 2024, both disclosures of which are incorporated herein by reference in their entireties.

GOVERNMENT INTEREST STATEMENT

[0002] This invention was made with Government support under Contract No. DE- NA0003525 awarded by the United States Department of Energy/National Nuclear Security Administration. The U.S. Government has certain rights in the invention.

FIELD OF THE INVENTION

[0003] The present invention is directed to improved carbon fibers. In particular, the present invention is directed to optimized non-circular carbon fiber geometries and sizes for use in carbon fiber reinforced polymers.

BACKGROUND OF THE INVENTION

[0004] Carbon fiber reinforced polymers are among materials with the highest massspecific strength and stiffness available. When considering its mechanical performance, it is obvious why carbon fiber usage is dominant in many fields. Carbon fiber materials to date have predominantly been developed for optimal tensile performance for use in the aerospace industry. The considerations of mass and cost also drive material selection for wind turbine blades, which are exceeding 100-meters in blade length. Many industries have relied upon the development of improved fiber composite materials. For example, the automotive, aircraft, defense, recreational, container and wind industries have relied upon improvements of strength, flexibility and cost to improve their products. The global commercialization of the wind industry has relied upon, and furthered, the development of composite materials to enable blade mass and cost reductions. [0005] The specific strength and stiffness (strength and stiffness per unit mass) of carbon fiber composites offer superior performance for a variety of industries including those mentioned above. The wind industry has relied upon carbon fiber composites to facilitate longer wind turbine blades. Most carbon fiber development has been focused on performance, not cost, where improved performance is necessary for achieving various system objectives. Additionally, for aerospace applications, the material cost is typically a minor contributor to the final produced part cost which can be orders of magnitude higher than the carbon fiber material cost. While many of the aerospace applications can justify the relatively high cost of carbon fiber based on performance requirements, most other applications are driven by cost-performance tradeoff assessments.

[0006] The successful development of wind energy has resulted in having the first or second lowest levelized cost of energy (LCOE) for new energy generation in the US. Blade mass scales with blade length to a power greater than two, which means that the longer blades that have resulted in LCOE reductions are becoming increasingly more massive. Increases in blade mass additionally drive system costs up, such as drive train and support structures, so controlling blade mass is critical to further LCOE reductions. It may be surprising then that most turbine OEMs do not conventionally design their blades using carbon fiber reinforced polymers. The improved mechanical properties of carbon fiber are often negated by the lower cost of glass fiber reinforced polymers in many wind turbine blade designs.

[0007] For wind turbines, the material modulus and compressive strength typically drive the material requirements in the main structural element, the blade spar cap, and not tensile strength. The spar cap component is essentially an internal beam spanning the blade length that provides the majority of the downstream structural resistance to wind loads. Due to the cantilevered nature of wind blades with distributed aerodynamic loads, the spar cap component acts as a beam in bending with similar magnitude compressive and tensile loads on the opposing blade surfaces. The similar maximum compressive and tensile loads combined with the lower compressive resistance of fiber-reinforced polymers results in the compressive strength being the controlling material strength for spar caps. A metric that is useful for material assessment in wind turbine blade spar caps is the costspecific compressive strength, particularly for materials with similar moduli. Increasing the cost-specific compressive strength of carbon fibers having a similar modulus will increase the value of that material for wind turbine spar caps and reduce the system LCOE.

[0008] Thus, there is a need for tough, damage resistant composite materials that offer increased compressive strength per unit cost of wind turbine blades.

SUMMARY OF THE INVENTION

[0009] The disclosure is directed to optimized fiber geometries for use in carbon fiber reinforced polymers with increased compressive strength per unit cost. The disclosed fiber geometries reduce the material processing costs as well as increase the compressive strength. The disclosed materials include non-circular carbon fiber geometries with lobes. In an embodiment, the fibers may include three or more lobes. In an embodiment, the fibers may include three, four, six or eight lobes. The lobes (1 ) reduce the diffusion thickness to enable higher throughput and lower costs during production and (2) increase the perimeter and area moment of inertia for a given fiber area. The noncircular lobed carbon fibers also have lower energy costs and emissions during carbon fiber conversion when compared to prior art circular fibers. The non-circular fibers may be used in the automotive, aircraft, defense, recreational, container and wind industries to improve strength, durability and cost.

[0010] In one embodiment, a carbon fiber tow includes a plurality of non-circular carbon fibers. A carbon fiber shape of the non-circular carbon fibers includes extensions in an outer surface to create a lobed geometry. The lobed geometry includes at least three lobes extending from the circumferential core of the non-circular carbon fibers and each of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area equal to or greater than 41 micron² (pm²). The new and unexpected discovery that cross-sectional areas equal to or greater than 41 pm² is based on the finding that conversion energy costs and emissions can be reduced by the adoption of non-circular carbon fiber geometries of three, four, six or eight lobes for increased fiber cross-sectional areas equal or to or greater than 41 pm².

[0011] The non-circular carbon fibers have a cross-sectional area of between 41 pm² and 200 pm². In an embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 140 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 100 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 100 m² and 140 pm². These cross-sectional areas provide unexpected cost effective compressive structural strength.

[0012] In another embodiment, a carbon fiber composite includes a matrix material and a plurality of non-circular carbon fibers embedded in the matrix material. The matrix material may be polymers including thermoset, thermoplastic, or vitrimer resin systems. Some common thermoset resins include epoxies, polyolefins, vinylesters, and polyurethanes. In another embodiment, the matrix may be carbon or silicon carbide. For example, the matrix may be formed of graphite or carbon nano-particles.

[0013] The carbon fiber shape of the non-circular carbon fibers includes extensions or lobes in an outer surface to create a lobed geometry. The lobed geometry includes at least three lobes extending from a core of the non-circular carbon fibers. In some embodiments, the core may be circumferential. In some embodiments, the number of lobes in the non-circular carbon fibers may be three lobes, four lobes, six lobes, or eight lobes. The lobes allow the fibers to have a smaller diffusion thickness per area, higher perimeter per area, and increase the fiber area moment of inertia compared to circular fibers. The lobes can also enhance the composite compressive strength and enable fiber nesting to arrest crack growth and improve damage tolerance of the composite.

[0014] In another embodiment, a structure is disclosed that includes a plurality of non- circular carbon fibers embedded within the matrix material. The geometry of the non- circular carbon fibers is discussed above. In an embodiment, the structure may be but is not limited to wind turbine blades and wind turbine components, automotive structures, recreational equipment, recreational equipment, and related to defense applications. In an embodiment, the structure may be a wind turbine blade or a portion of a wind turbine blade. In an embodiment, a wind turbine blade is disclosed that includes a carbon fiber composite including a matrix material and a plurality of non-circular carbon fibers as disclosed above embedded within the matrix material.

[0015] According to an embodiment of the disclosure, a carbon fiber tow is disclosed that includes a plurality of non-circular carbon fibers, wherein a carbon fiber shape of the non- circular carbon fibers includes extensions in an outer surface to create a lobed geometry. The lobed geometry is selected from the group consisting of three lobes, four lobes, six lobes and eight lobes extending from a circumferential core of the non-circular carbon fibers and each of the plurality of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area greater than 41 micron².

[0016] According to another embodiment of the disclosure, a carbon fiber composite is disclosed that includes a matrix material and a plurality of non-circular carbon fibers embedded in the matrix material, wherein a carbon fiber shape of the non-circular carbon fibers includes extensions in an outer surface to create a lobed geometry. The lobed geometry is selected from the group consisting of three lobes, four lobes, six lobes and eight lobes extending from a circumferential core of the non-circular carbon fibers and each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than 41 micron².

[0017] According to another embodiment of the disclosure, a wind turbine blade is disclosed that includes a carbon fiber composite comprising a matrix material and a plurality of non-circular carbon fibers embedded within the matrix material. The carbon fiber shape of the non-circular carbon fibers includes extensions in an outer surface to create a lobed geometry and the lobed geometry is selected from the group consisting of three, four, six and eight lobes extending from a circumferential core of the non-circular carbon fibers. Each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than 41 micron².

[0018] Other features and advantages of the present invention will be apparent from the following more detailed description of the preferred embodiment, taken in conjunction with the accompanying drawings which illustrate, by way of example, the principles of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] FIG.1 illustrates a wind turbine according to an embodiment of the invention.

[0020] FIG. 2 shows normalized fitness values as a proxy for cost-specific compressive performance versus the number of lobes according to an embodiment of the disclosure.

[0021] FIG. 3A illustrates the cross-sectional geometry of a three lobe carbon fiber geometry according to an embodiment of the disclosure. [0022] FIG. 3B illustrates the cross-sectional geometry of a four lobe carbon fiber geometry according to an embodiment of the disclosure.

[0023] FIG. 3C illustrates the cross-sectional geometry of a six lobe carbon fiber geometry with alternating lobe length/height according to an embodiment of the disclosure.

[0024] FIG. 3D illustrates the cross-sectional geometry of an eight lobe carbon fiber geometry with alternating lobe length/height according to an embodiment of the disclosure.

[0025] FIG. 4 shows performance parameters of the exemplary geometries from FIGS. 3A-3D relative to a circular fiber having the same cross-sectional area according to an embodiment of the disclosure.

[0026] FIG. 5 illustrates an exemplary fiber packing arrangement of multiple three lobe carbon fiber geometries in a composite.

[0027] FIG. 6 is a plot of compressive strength test averages versus effective fiber diameter for two distinct precursor systems (Dolan and Kaltex) in two composite test forms (unidirectional [0] and cross-ply [0/90]). This figure illustrates a positive correlation between increased fiber size and/or area moment of inertia and improved compressive strength.

[0028] FIG. 7A shows numerical predictions of compressive strength for a fiber diameter of 7.2 microns and a fiber that has twice the diameter of 14.4 microns as a function of initial fiber misalignment angle.

[0029] FIG. 7B is a plot of the compressive strength percent difference of the larger diameter fiber based on predictions in FIG. 7A with improvements of 2.5% to 4.5%. The numerical results for improvement in compressive strength with fiber size underpredict the trends observed from the test results in FIG. 6.

[0030] FIG. 8A is a plot of the fiber area moment of inertia variation with fiber cross- sectional area to compare a circular fiber with the exemplary three lobe and exemplary six lobe fibers from FIG. 3A and 3C, revealing the improvement in this property for the non-circular fibers. [0031] FIG. 8B is a plot of the fiber perimeter divided by its area variation with fiber cross- sectional area to compare a circular fiber with the exemplary three lobe and exemplary six lobe fibers from FIG. 3A and 3C, revealing the improvement in this property for the non-circular fibers. The interfacial area for a fiber is the product of the fiber perimeter and its length. For a composite with a given fiber volume fraction, this figure illustrates that the total interfacial area in the composite is decreased as fibers of any specific geometry are increased in size. The non-circular fibers enable the same or higher total interfacial area in a composite while using fibers with larger cross-sectional areas (horizontal dashed line as a reference).

[0032] FIG. 8C is a plot of the fiber diffusion thickness during oxidation variation with fiber cross-sectional area to compare a circular fiber with the exemplary three lobe and exemplary six lobe fibers from FIG. 3A and 3C, revealing the improvement in this property for the non-circular fibers. Reducing the diffusion thickness reduces the required residence time during the oxidation stage of carbon fiber conversion (FIG. 9). The non- circular fibers enable the same or lower diffusion thickness while using fibers with larger cross-sectional areas (horizontal dashed line as a reference).

[0033] FIG. 8D is a plot of the required residence time during oxidation variation with fiber cross-sectional area to compare a circular fiber with the exemplary three lobe and exemplary six lobe fibers from FIG. 3A and 3C, revealing the improvement in this property for the non-circular fibers. Reducing the required residence time during the oxidation stage of carbon fiber conversion (FIG. 9) increases production throughput and reduce carbon fiber costs.

[0034] FIG. 9 illustrates the major sequential steps in carbon fiber conversion as used in the cost model to predict cost trends with carbon fiber geometry.

[0035] FIG. 10 is a plot of the carbon fiber material cost estimates as a function of fiber diffusion thickness and cross-sectional area. Non-circular fiber geometries have lower diffusion thickness values and follow distinct lines within the sampled space. The cost curves are shown for the exemplary three lobe and exemplary six lobe fibers from FIG. 3A and 3C relative to a circular fiber (having the highest possible diffusion thickness for a given area representing the upper bound in this plot). [0036] FIG. 11 is a plot of the carbon fiber material cost versus the fiber cross-sectional area for the exemplary profiles from FIG. 3 compared with a circular fiber. The diamond markers reveal the numerical minimum of each shape’s cost curve. It is apparent from this figure that the non-circular geometries reduce the cost relative to circular geometries while also enabling the cost-effectiveness of carbon fibers with large cross-sectional areas.

[0037] FIG. 12 is a plot of the energy costs during carbon fiber conversion versus the fiber cross-sectional area for the exemplary profiles from FIG. 3 compared with a circular fiber. The diamond markers reveal the numerical minimum of each shape’s cost curve. The energy costs have associated emissions, and the figure illustrates how the noncircular carbon fiber geometries can have reduced energy costs and associated emissions beyond 41 pm² carbon fiber cross-sectional areas whereas the quantities increase beyond this value for circular fibers.

[0038] Wherever possible, the same reference numbers will be used throughout the drawings to represent the same parts. In addition, items referred to by reference number plus letters refer to the same reference number part with an additional condition.

DETAILED DESCRIPTION OF THE INVENTION

[0039] The invention includes carbon fiber cross-sectional shapes for carbon fiber reinforced polymers that provide cost-specific compressive strength of a carbon fiber composite structure. In one embodiment, the carbon fiber composite is used in wind turbine blades.

[0040] The disclosed carbon fiber shapes include extensions that extend away from a circumferential core of the carbon fiber to create lobed geometries. In some embodiments, the disclosed carbon fiber shapes include a periodic oscillation in their outer surface to create the lobed geometries. In other embodiments, the extensions that create the lobed geometries include lobes that are all the same size, alternating sized lobes or lobes of multiple varying sizes. These carbon fiber shapes are also described as “lobed noncircular carbon fibers” herein. The carbon fiber composites disclosed herein have lobed non-circular carbon fibers embedded in a polymer matrix. The composite has improved compressive strength compared to prior art circular carbon fibers. [0041] As defined herein, “lobes” (or “lobed”) are extensions that extend away from a circumferential core of the carbon fiber. The “circumferential core”, as defined herein, is the central portion of the carbon fiber. While the circumferential core is shown as circular herein, other shapes, such as squares, triangles, or ovals, are within the scope of the current invention.

[0042] Put another way, lobes are perturbations of material extending away from the circumferential core. The minimum lobe height of each lobe is at least 10% of the diameter of the circumferential core of the noncircular carbon fiber. A lobed geometry includes a plurality of lobes.

[0043] The carbon fiber composite disclosed herein includes large diameter lobed noncircular fibers that increase compressive strength while also being cost effective compared to prior art circular carbon fibers. Lobed non-circular carbon fibers have a higher area moment of inertia at a much lower cost than circular carbon fibers. This also allows the carbon fibers to be straighter, with less misalignment than prior art circular carbon fibers.

[0044] Composites including the lobed non-circular fibers disclosed herein may be interlocking, resulting in better damage tolerance than circular carbon fibers. Water absorption is also slowed down, so degradation decreases with lobed non-circular fibers in certain applications especially where water absorption is prevalent. Lobed non-circular fibers also have lower emissions than circular carbon fibers with the same cross-sectional area.

[0045] In an embodiment, the non-circular carbon fibers with extensions in their outer surface have at least three lobes. Any of the lobes in the non-circular carbon fibers may have the same or different amplitude than other lobes in the same non-circular carbon fiber.

[0046] In an embodiment, the non-circular carbon fibers with extensions in their outer surface have three, four, six or eight lobes. In some embodiments, the lobes in the three- lobe embodiment are all the same and equidistant from each other. In other embodiments, the lobes may not be equidistant from each other. In other embodiments, the three lobes may all have a different amplitude, or one of the three lobes may have a different amplitude than the other two lobes. The lobes in the four-lobe embodiments are equidistant from each other and may either be all the same or alternating in height/length such that there are two sets of two different sized lobes. In some embodiments, the angular orientation of the lobes could be symmetric or variable (for each geometry).

[0047] In some embodiments, the lobes in the six-lobe and eight lobe embodiments are equidistant from each other and preferably each include two lobe types, each of a different height/length, which alternate with each other. In other embodiments, the lobes in the six- lobe or eight-lobe embodiments all have the same height/length. In an embodiment, the alternating lobes in the four, six or eight lobe embodiments have variable amplitude (height/length). In one embodiment, the non-circular carbon fibers are used in turbine blade structures.

[0048] The invention includes carbon fiber composites including the non-circular carbon fibers described herein within a matrix. The composite may be made of unidirectional carbon fibers, multi-axial carbon fibers, or a combination of unidirectional carbon fiber and multi-axial carbon fiber. The unidirectional carbon fibers and multi-axial carbon fibers may be continuous carbon fibers with a long length or discontinuous carbon fibers with a short length. The non-circular carbon fibers may all have the same non-circular shapes, or the non-circular carbon fibers within the matrix (or within a carbon fiber tow) may have more than one of the non-circular carbon fiber shapes disclosed herein.

[0049] In an embodiment, tows formed of the disclosed carbon fibers may include multiple fiber types in one fiber-reinforced polymer (FRP) composite. This could be different reinforcement fiber materials (e.g., carbon + aramid + kevlar, etc.). These may be referred to as “hybrid” composites. Additionally, there could be carbon fiber tows with multiple different fiber shapes in a single tow. For example, a tow that includes multiple non- circular fiber shapes and possibly also including circular fiber shapes.

[0050] In some embodiments, carbon fiber composites include long length, continuous lobe-shaped non-circular carbon fibers aligned in the same direction. In other embodiments, the carbon fiber composites include long length, continuous lobe-shaped non-circular carbon fibers aligned in two or more directions. In other embodiments, the carbon fiber composites include short length, discontinuous lobe-shaped non-circular carbon fibers aligned in one or more directions or aligned randomly. [0051] The lobe-shaped carbon fibers described herein have a combination of a larger area moment of inertia and perimeter at a given cross-sectional area than conventional circular carbon fibers and produce a higher composite compressive strength. Compressive strength of carbon fiber composites can be further increased with carbon fiber size.

[0052] The lobe-shaped carbon fibers described herein also are significantly more cost effective than circular carbon fibers at higher carbon fiber areas (pm²). The three lobe and six lobe carbon fibers described herein have lower carbon fiber conversion costs than conventional circular carbon fibers, and larger three-lobe and six-lobe carbon fibers are not cost-prohibitive, unlike the costs for larger conventional circular carbon fibers. Three- lobe and six-lobe carbon fibers result in manufacturing cost improvements and promising fiber packing.

[0053] The non-circular carbon fibers have a cross-sectional area of between 41 pm² and 200 pm². In an embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 140 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 100 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 60 pm² and 140 pm². These cross-sectional areas provide unexpected cost effective compressive structural strength for three, four, six and eight lobe non-circular cross-sectional carbon fibers.

[0054] There is a reduction in carbon fiber cost with non-circular fiber cross-sections. Lobed non-circular carbon fiber shapes reduce oxidization time during manufacturing which decreases the needed energy input and increases the production rate and thus lowers the manufacturing cost. While accounting for manufacturability, there is a 27% to 32% reduction in the resulting carbon fiber cost based purely on the fiber geometry changes while maintaining the precursor chemistry, fiber cross-sectional area, and tow count. As defined herein, a “carbon fiber tow” is a bundle of carbon fiber filaments. The carbon fiber tow count is the number of filaments in a bundle.

[0055] In some embodiments, the carbon fiber tow count is greater than or equal to 6 k. In other embodiments, the carbon fiber tow counts are greater than or equal to 24 k. In other embodiments, the carbon fiber tow counts are greater than or equal to 50 k. In some embodiments, the carbon fiber tow count may be as between 50 k and 475 k. [0056] While there is an unexpected positive correlation observed for compressive strength with fiber size, increasing the carbon fiber size substantially for circular fibers is not optimal because reduced factory throughput during carbon fiber conversion increases costs.

[0057] Non-circular fibers produce an opportunity to realize the compressive strength improvements of larger carbon fibers without the negative cost consequences of using circular fibers. This is accomplished through reducing the diffusion thickness parameter when using non-circular fibers.

[0058] When estimating the carbon fiber cost for specific fiber geometries, lobed non- circular carbon fibers unexpectedly have cost-optimal performance at larger fiber cross- sectional areas compared to circular fibers. This invention discloses that carbon fiber cost for the non-circular carbon fibers decreases with the use of larger fiber cross-sectional area.

[0059] The decreased costs for the non-circular carbon fibers are partly sourced from reduced energy load, such that the use of the lobed non-circular carbon fibers disclosed herein results in lower emissions than using traditional circular carbon fibers.

[0060] The non-circular carbon fibers with carbon fiber areas between 41 pm² and 200 pm² disclosed herein have decreased energy costs during carbon fiber conversion. There is an unexpected trend where the energy cost decreases as the carbon fiber area increases forthree-lobed, four-lobed, six-lobed and eight-lobed non-circular carbon fibers. Other energy related parameters, for example emissions, are also decreased as the carbon fiber area increases for three-lobed, four-lobed, six-lobed and eight-lobed non- circular carbon fibers.

[0061] FIG. 1 illustrates a wind turbine 100 according to an embodiment of the disclosure. The wind turbine 100 includes three wind turbine blades 101 attached to a nacelle 103 or hub, which is attached to a tower 105. While FIG. 1 shows three blades 101 , any suitable number of blades 101 , such as, but not limited to two, four or five blades 101 may be used. It should be noted that FIG. 1 is not drawn to scale.

[0062] In FIG. 1 , the wind turbine blades 101 include a matrix of lobed non-circular carbon fibers, the cross-sections of various embodiments of which are shown in FIGS. 3A-3D (see discussion of FIGS 3A-3D below). In another embodiment, the matrix includes a plurality of three-lobed non-circular carbon fibers, four-lobed non-circular carbon fibers, six-lobed non-circular carbon fibers or eight-lobed non-circular carbon fibers.

[0063] The non-circular carbon fibers have a cross-sectional area of between 41 pm² and 200 pm². In an embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 140 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 41 pm² and 100 pm². In another embodiment, the non-circular carbon fibers have a cross-sectional area of between 100 pm² and 140 pm². These cross-sectional areas provide unexpected cost effective compressive structural strength for three, four, six and eight lobe non-circular cross-sectional carbon fibers.

[0064] The non-circular cross-sectional fiber shapes disclosed herein have been found to provide the highest fitness performance. If the fiber packing factor constraint is ignored, only two fitness factors were considered (normalized by fiber production costs). Fitness 1 only cares about the area moment of inertia. Fitness Values combines weights the area moment of inertia and the fiber perimeter. The disclosed fibers provide for effective area moment of inertia for fitness function (which does not penalize for fiber packing factor).

[0065] FIG. 2 shows the distribution of fitness results when using the penalized FPF approach with a 68% target. The best performing shapes when penalizing FPF calculations (whether including perimeter effects or not) have three, four, six, or eight lobes. Shapes with five, seven, and nine lobes do not perform as well due to the lower fiber packing factor values that must fit within the circular geometrical element which does not have an idealized perfect packing. FIG. 2 illustrates this trend where five, seven, and nine lobe shapes always have a lower area moment of inertia for a given maximum fiber packing factor.

[0066] Referring to FIG. 2, the normalized performance “fitness” values are highest for shapes with a high cost-specific area moment of inertia. For this fitness calculation, the optimal shapes are those that move the most area furthest from the shape center. Fitness 1 values are therefore highest for shapes with an even number of lobes where the alternating peaks are removed R to 0 as these shapes can have long, narrow lobes. Calculating the fitness objective functions using the minimum area moment of inertia compared to the effective area moment of inertia removes the Fitness 1 peak for two- lobed shapes due to the asymmetric area moment of inertia of these shapes with asymmetric lobe heights. The Fitness 2 values include the effect of shape perimeter in estimating compressive strength of the composite. Based on this approach, there is nearly a direct correlation between number of lobes and shape fitness performance as the higher-lobed shapes have associated higher perimeters.

[0067] FIGS. 3A-3D show dimensions for exemplary 3, 4, 6 and 8 lobe fibers. These exemplary fibers with these numbers of lobes have shown the best performance for increasing fiber area moment of inertia and perimeter at reduced costs relative to circular fibers while enabling efficient fiber packing arrangements to yield high fiber volume fraction composites. When using a constrained FPF to achieve a high fiber volume fraction, the three-lobe (see FIG. 3A) and six-lobe (see FIG. 3C) shapes are the best performers. If there is nesting of the fiber lobes for the fiber packing factor estimate, the two-lobe shape or another shape may be a better performer if they nest properly.

[0068] FIG. 4 shows various performance metrics for non-circular carbon fiber geometries with three, four, six, and eight lobes compared with a circular fiber of the same cross- sectional area. The values are normalized to those for the reference 7.2 micron diameter circle to illustrate why they are optimal over a circular fiber with the same cross-sectional area. Each of the shapes have a higher area moment of inertia and perimeter compared to the circle, and a lower cost (because of the shorter diffusion pathway for oxidation).

[0069] The six-lobe and eight-lobe shapes both have the highest perimeter (FIG. 4) meaning these shapes would be more optimal if the composite compressive strength is more correlated to the interfacial area than was assumed. The three-lobe and four-lobe shapes have the lowest cost, but also have lower area moments of inertia than the six- lobe (FIG. 4) shape. The eight-lobe (FIG. 4) shape would perform higher overall but does not achieve the fiber packing factor target and is penalized at the idealized prediction.

[0070] Compared to the circular fiber, the three and six lobe fibers have a 50-60% higher area moment of inertia, 40-80% increased perimeter, and a 30% reduction in the fiber cost (see FIG. 4).

[0071] Optimal fiber shapes have been identified as those with the greatest potential to maximize the cost-specific compressive strength of carbon fiber composites based on the analysis approach within the previous section. The selected optimal shapes focus primarily on fitness functions which penalize based on fiber packing factor predictions. These two fitness calculations are considered the most relevant for wind turbine blades due to the importance of the fiber volume fraction at reducing blade weight and cost. Unknowns in the objective formulation prevent a singular optimal shape from being identified by this analysis. These unknowns include the area moment of inertia that drives composite performance, minimum or effective, as well as the quantitative relationship between the fiber-resin interfacial area and resulting compressive strength.

[0072] FIG. 3A shows a three-lobed carbon fiber geometry. The three-lobed fibers include three lobes 310 equidistant from each other and having the same shape and amplitude (height/length).

[0073] FIG. 3B shows a four-lobed geometry, where the lobes 320 are all the same length. While three-lobed shapes and six-lobed shapes perform best for higher fiber volumes in applications like wind turbine blades, in applications where the user wants a lower fiber volume fraction, the four-lobed shapes with all the same size lobes or alternating lobes may perform better.

[0074] FIG. 3C shows an alternating six-lobed carbon fiber geometry. The alternating lobes have variable amplitude (height/length). The alternating six-lobed carbon fiber geometry includes three lobes 330 and three lobes 340 having a different length than the three lobes 330. The lobes alternate around a circumferential core of the six-lobed carbon fiber.

[0075] FIG. 3D shows an alternating eight-lobed carbon fiber geometry. The alternating eight-lobed carbon fiber geometry includes four lobes 350 and four lobes 360 having a different length than the four lobes 350. The lobes alternate around a circumferential core of the eight-lobed carbon fiber.

[0076] The three-lobe shape (FIG. 3A) does not depend on the effective area moment of inertia and performs comparably regardless of including perimeter effects or not in the fitness calculation. This shape also has thicker lobes which benefit the qualitative considerations related to manufacturing, resin wetting, and lobe durability. Additionally, it is considered the most likely to achieve its fiber packing factor estimate due to the I ikelihood of adjacent fibers nesting well between lobes.

[0077] There is a strong inverse relationship between fiber packing factor and area moment of inertia, indicating that the optimal shapes should differ substantially for applications with varying fiber volume fraction targets. Increasing the shape perimeter also decreases the maximum fiber packing estimate, generally. There is a complementary relationship between area moment of inertia and perimeter. Six-lobe shapes can achieve the highest area moment of inertia for a given fiber packing factor, while six and eight lobe shapes achieve the highest perimeter for a given fiber packing factor.

[0078] FIG. 5 illustrates an exemplary fiber packing arrangement considered for the three lobe shape. Oftentimes, the fiber volume fraction controls what the base-shape is. A fiber volume fraction 0.68 was targeted for all shapes and all fiber packing arrangements.

[0079] The selected fiber shapes are obtained with the following parametric equation: r = h [1 + S((cos(/<0/4)²)'⁷ + /?(sin( c^4)²)ⁿ)] where r is radial position of the fiber boundary, 0 is the parameter and varies from 0 to 2TT, / s a positive integer which prescribes the number of lobes, is defined to take on positive values and adjusts the amplitude of the lobes, /Pis a value between 0 and 1 which controls the relative amplitude of alternating lobes; /’ must equal one when k is an odd number for the shape to be continuous; /F allows for non-axisymmetric shapes in some cases, and n adjusts the curvature of the shape, where a large n results in steep lobes that are narrower. The area of the generated shape is normalized to the baseline fiber by the variable h.

[0080] The three-lobe and six-lobe fibers are predicted with compressive strength increases of 6-10% for a reference fiber alignment distribution. The increased area moment of inertia for the non-circular carbon fiber geometries is hypothesized to relate to improved fiber alignment. The sensitivity of these improvements is calculated for a 10% improvement in fiber alignment (as quantified by the standard deviation of the fiber angle distribution) which increases the compressive strength improvement to 10-14%.

[0081] The longitudinal compressive strength of unidirectional composites can be improved using non-circular fibers including, but not limited to, six-lobe and three-lobe shaped carbon fibers. These non-circular fibers also provide cost reductions on the order of 30%.

[0082] In addition to the predicted compressive strength improvements for a given unit cell, there is another potential advantage for these non-circular fibers regarding the fiber alignment distribution in the composite. The three-lobe and six-lobe fibers have a 52% and 62% increase in the area moments of inertia relative to a circular fiber of the same cross-sectional area. This increased area moment of inertia results in a higher bending resistance and may contribute to reduced fiber waviness and improved alignment during composite manufacturing.

[0083] Non-circular fibers can reduce manufacturing costs by up to 30%. The results herein show that six-lobed fibers can improve strength from 10% to 13% and three-lobed fibers can be 6% to 9% stronger than circular fibers. These strength improvements can be further augmented with increased fiber alignment resulting from the lobed non-circular carbon fibers with higher area moments of inertia and bending resistance.

[0084] Carbon Fiber Size and Carbon Fiber Processing Costs

[0085] The current micromechanical theories on compressive strength of carbon fiber consider compressive strength to be independent of fiber size. However, as described herein, alterations of carbon fiber geometry increase compressive strength through alteration of the carbon fiber geometry by increasing the fiber area moment of inertia and/or the fiber perimeter and interfacial area.

[0086] FIG. 6 shows that carbon fiber systems with increasing size illustrate a favorable correlation for compressive performance, which is not predicted from a micromechanical failure model (FIG. 7A and FIG. 7B). The manufacturing and mechanical test results show that alterations to fiber geometry can be used to produce improvements of the compressive strength of carbon fiber reinforced polymers.

[0087] For wind technologies, cost is essential. Wind technologies also care about compressive strength as much as tensile strength due to mostly symmetric bending loading conditions. In the present invention, large diameter lobed non-circular fibers increase cost effective compressive strength. [0088] For wind turbine blades, the desirable mass-specific modulus inherent with carbon fiber systems must also be met with system cost reductions to facilitate widespread implementation. Additionally, the deficit in compressive strength relative to tensile strength for carbon fiber reinforced polymers (CFRP) makes compressive strength a controlling property for wind turbine blades, while improving tensile strength would not reduce the material demand. For cost-driven applications with fairly symmetric loading demands, such as for wind blades, increasing compressive strength and decreasing material cost are the two most meaningful targets for designing future carbon fiber material systems.

[0089] There are some additional factors that need to be considered when increasing the fiber diameter of circular fibers. For a given geometry, increasing the fiber size increases the fiber area moment of inertia (FIG. 8A) but reduces the perimeter to area ratio of each fiber (FIG. 8B), resulting in less total fiber-resin interfacial area within the composite. Additionally, larger fibers have higher diffusion thicknesses (FIG. 8C) which require more residence time during the oxidation stage of carbon fiber conversion (FIG. 8D). The oxidative stabilization process rate is controlled by diffusion of oxygen into the full fiber thickness, with diffusion time related to the diffusion thickness traditionally assumed to be a power of 2, but this relationship might be closer to a power of 1.5. Nevertheless, the relationship favors smaller diameter fibers for increased line speed and maximum tensile performance.

[0090] One opportunity to enhance the benefits of larger fibers while controlling costs is to utilize non-circular carbon fiber geometries, such as multi-lobed geometries or asymmetric fibers. Non-circular carbon fibers enable increases in the fiber area moment of inertia while maintaining diffusion thicknesses in ranges like those for commercial fibers, resulting in similar production line speeds. The non-circular geometry also increases the wetted perimeter of the fiber over more circular fibers having the same total cross-sectional area, thereby potentially providing opportunity to increase interfacial adhesion.

[0091] Non-circular carbon fiber materials for cost-driven industries permit the opportunity to reduce the processing costs and increase compressive strength. Examples of non- circular carbon fiber geometries are discussed herein and shown in FIGS. 3A-3D. These geometries relate to composite compressive strength and decreasing the carbon conversion costs. More specifically, these geometries offer enhancements in cost- effective compressive strength of carbon fiber composites.

[0092] This invention determines and evaluates the relationship between carbon fiber size or, more broadly, the shape area moment of inertia to compressive strength and material cost to identify optimal carbon fiber sizes and geometries for fiber reinforced polymer composites in wind turbine blades and other applications.

[0093] To investigate the effects of increasing carbon fiber size, textile acrylic fibers produced by partner Dolan GmbH in three different effective diameters of about 14.4, 17.7, and 23.9 microns (typical carbon fiber precursor diameters range from 10-13 microns) were used. The resulting carbon fibers have effective diameters of about 5.8, 7.0 and 9.5 microns and tensile results are shown in FIG. 6.

[0094] One textile acrylic fiber is a fiber product manufactured by Kaltex which is produced in very large tows of 457,000 filaments. The wet spinning process employed by Kaltex naturally yields a fiber that is highly non-circular and frequently characterized as a “kidney bean” shape. The Kaltex fibers were also included in the fiber size comparison in FIG. 6.

[0095] Carbon fiber geometries with higher area moments of inertia can increase compressive performance. These materials are all based on polyacrylonitrile (PAN) precursors. The commercial baseline carbon fiber systems were tested to provide a set of known references for the experimental carbon fiber systems. The set of textile acrylic fiber systems from suppliers Dolan and Kaltex were used to assess the impact of varying the size of the carbon fiber within consistent precursor systems.

[0096] The compressive test data are presented in FIG. 6 for the unidirectional [0] coupon tests as well as the cross-ply [0/90] coupon tests using the backout factor from Equation 11 . In addition to the actual test results, the unidirectional compressive strength data are normalized to a 60% fiber volume fraction for a more direct comparison according to Equation 12. While the compressive strength is not perfectly linear with fiber volume fraction, this estimation is appropriate for the minor variations in fiber volume fraction associated with the test data.

[0097] The carbon fiber materials developed to assess the impact of size on carbon fiber compressive performance are compared in FIG. 6. The comparison is made across two precursor systems, including the more circular Dolan fibers and the kidney bean shape Kaltex fibers. Unfortunately, the shape of the Dolan fibers does become less circular with increasing size. To supplement this fiber system, carbon fiber produced using the Kaltex precursor was also tested for two systems with distinct fiber weights. A correlation of increasing compressive strength with fiber size is apparent from these data for both precursor material systems. This positive correlation is consistent for both composite compressive test approaches as well and illustrates the opportunity to increase compressive strength by increasing carbon fiber size and/or area moment of inertia.

[0098] In prior art theories, compressive strength of carbon fiber at the micromechanical level is considered independent of fiber cross-sectional size due to the vast differences in length scale between the length of the fiber column and cross-section of the fiber. Composite test results in FIG. 6 show a correlation of improved compressive strength with fiber size, and area moment of inertia, which are not predicted from numerical modeling (FIGS. 7A-7B). This positive correlation provides further reason and opportunity for noncircular carbon fiber geometries which enable the higher area moment of inertia without negative cost impacts that would be realized for circular fibers. In contrast, increasing the carbon fiber size substantially for circular fibers is not optimal because reduced factory throughput during carbon fiber conversion increases costs. The relationship between fiber size and increased compressive strength is not expected.

[0099] To supplement the test results, a prediction of compressive performance for circular fibers with increasing fiber diameter was made using a finite element micromechanical failure model (FIG. 7A). A reference case simulating the Zoltek PX-35 material was used (D=7.2 microns) in addition to a scenario to predict the impact of doubling the fiber diameter (D=14.4 microns). These numerical predictions were made using a hexangular packing with 62% fiber volume fraction and periodic boundary conditions on the lateral faces. Some of the model parameters, as well as a replication of the repeating unit cell and half-cosine perturbation of the mesh, were taken from Camarena et al., Development of a compressive failure model for carbon fiber composites and associated uncertainties, 2021 , herein incorporated by reference in its entirety.

[00100] The compressive strength prediction as a function of initial fiber misalignment for the reference and increased diameter fibers are shown in FIG. 7A. The larger diameter fiber is predicted to outperform the baseline, however, with only a 2.5% to 4.5% increase in compressive strength for initial misalignments up to 3 degrees for the larger fiber as seen in FIG. 7B.

[00101] The experimental and numerical results both illustrate an increase in compressive strength with increasing carbon fiber size. However, the increase predicted from the numerical simulations are meaningfully lower than what was observed from the textile-grade carbon fiber systems. One theory for this distinction is that the larger fibers with increased area moment of inertia may maintain improved fiber alignment during composite manufacturing. This secondary benefit of improved alignment could be more significant for increasing composite compressive performance than the resistance to micro-buckling, due to the high sensitivity of compressive strength on fiber alignment (as seen in FIG. 7A).

[00102] The textile-grade carbon fiber systems show that there is a correlation between compressive strength and fiber size. This was observed across two precursor systems with somewhat circular and kidney-bean shape fibers. It is considered this relationship is primarily driven by increases in fiber area moment of inertia. Increasing carbon fiber size for circular carbon fibers is cost-prohibitive. However, since lobed fiber shapes can increase area moment of inertia without increases in cross-sectional area, these shapes are beneficial for cost-effective non-circular carbon fibers.

[00103] The cost of carbon fiber has a strong dependency on the conversion processing speed which is limited by the required oxidation time for the precursor, related to the radius (diffusion thickness of the precursor fibers) to a power between 1.5 to 2. However, when only using circular fibers, there is a fixed relationship between quantities of interest (area moment of inertia, fiber perimeter, and diffusion thickness) and the fiber diameter. Alternative shapes enable variation from these fixed relationships.

[00104] Non-circular Carbon Fiber Size and Cost

[00105] FIG. 8A shows a graph of fiber geometry and processing trends for circular 1 , three-lobe 10 and six-lobe 14 carbon fibers, the shapes of which are shown in FIGS. 3A and 3C, respectively. As shown in FIG. 8B, the perimeter per area is lower for the circular fiber than both the three-lobe (FIG. 3A) and six-lobe (FIG. 3C) fibers. FIG. 8D shows that the oxidation residence time for the circular fibers increases significantly as carbon fiber area increase for the circular fiber. While the residence time also increases for the three- lobe and six-lobe fibers, it increases at a much slower rate for both of these non-circular fiber shapes.

[00106] FIG. 8A shows the area moment of inertia (microns⁴) versus carbon fiber area for circular carbon fibers, three-lobed carbon fibers (FIG. 3A) and six-lobed carbon fibers (FIG. 3C). The area moment of inertia is higher for both the three-lobe and six-lobe fibers compared to the circular fiber. As the figure shows, the larger cross-sectional area produces an even larger area moment of inertia for the three-lobe and six-lobe non- circular carbon fiber shapes, producing compounding benefits for improved fiber alignment in composites and increased compressive strength.

[00107] FIG. 8C shows diffusion thickness (microns) versus carbon fiber area. The diffusion thickness is the geometry characteristic that controls the required oxidation residence time. The lower diffusion thickness of the three-lobe and six-lobe shapes enables higher throughput and resulting lower cost.

[00108] Cost modeling was performed for circular fibers, three-lobed non-circular fibers (FIG. 3A) and six-lobed non-circular fibers (FIG. 3C). The cost modeling considered raw materials, precursor manufacturing, and carbon fiber conversion.

[00109] The cost of converting the PAN-based precursor to carbon fiber is estimated using a carbon fiber conversion cost model. The cost model accounts for the capital, labor, energy, and material costs according to the major process steps depicted in FIG. 9. The process steps include precursor, pre-treatment, oxidation, low temperature carbonization furnace (LT), high temperature carbonization furnace (HT), surface treatment, sizing, and post-processing steps (winding, inspection, and shipping). Effluent occurs during the oxidation, LT, and HT steps. The effluent is dealt with during abatement.

[00110] The baseline conversion cost model assumptions represent a $58MM installed capital costs with a 15-year life expectancy, a 3 m roller width, 810 m of heated oxidation furnace length, continuous operation with a plant availability of 82%, and a mechanical yield of 95% and chemical yield of 48%. Additional assumptions and commercial reference cost estimation values used for calibration in this analysis are provided in Ennis et al., Optimized carbon fiber composites in wind turbine blade design. Technical Report (SAND2019-14173), Sandia National Laboratories, November 2019, herein incorporated by reference in its entirety.

[00111] The carbon fiber cost model is used to estimate the conversion costs for variable carbon fiber geometry and tow specifications. Carbon fiber throughput is controlled by the required residence time during oxidation. For the baseline commercialgrade carbon fiber, the oxidation time is estimated at 90 min compared of 1 .5 min for the low temperature (LT) and high temperature (HT) carbonization furnaces. Based on this convention, the time during oxidation is treated as the rate controlling stage for carbon fiber production in the model assuming a fixed oxidation furnace length of 810 m. The required oxidation residence time is related the fiber diffusion thickness according to Equation 16. The line speed during conversion is then calculated as the heated length in the oxidation furnace divided by required residence time. The length of the LT and HT furnaces are then modified for this line speed assuming a fixed requirement of 90 sec residence time, independent of the variable carbon fiber specifications. 6.23 X t diffusion (16)

[00112] The diffusion thickness that controls the residence time during oxidation is based on the fiber geometry during oxidation, not the final carbon fiber geometry. It is therefore necessary to relate these intermediate parameters to the carbon fiber geometry for the carbon fiber optimization studies. During the various stages of conversion there are density changes to the fiber along with stretching as the precursor is being converted to carbon fiber. The relationship between fiber area during oxidation and the final carbon fiber is approximated in Equation 17. A simplifying assumption is made that the area translates proportionally so the diffusion thickness during oxidation can be estimated from the final carbon fiber geometry with Equation 18, using common values from conversion studies.

^diffusion ~ 1-648 X t_carb_on fiber (18)

[00113] The resulting carbon fiber cost is the summation of conversion costs and precursor costs as shown in Equation 22. Precursor cost contributions were estimated at 45% of carbon fiber costs for a commercial-grade, PAN-based carbon fiber where processing yields require 2.2 kg of precursor for every 1 kg of carbon fiber.

[00114] FIG. 10 shows total carbon fiber costs [USD/kg], including precursor costs, for a 50 k tow count. The circular reference fiber 1 is approximately 40 pm² and is shown by the O in the figure. The six-lobe carbon fiber (FIG. 3C) is shown a short-long dashed line, while the three- lobe carbon fiber (FIG. 3A) is shown as a constant dashed line. As shown in FIG. 10, both the three lobed and six lobed carbon fiber shapes have a significantly lower cost than the circular fibers.

[00115] FIG. 11 shows a summary of the carbon fiber costs for 50 k tow counts, including precursor costs, for the circular prior art fibers, the three-lobed fibers (FIG. 3A) and the six-lobed fibers (FIG. 3C), as well as the carbon fiber costs at 50k tow count for the four-lobed (FIG. 3B) and eight-lobed (FIG. 3D) carbon fibers. While 40 pm² is used as the reference circular fiber area (Zoltek PX-35, 50K tow), this is considered large in the industry.

[00116] The cost of circular carbon fibers with high carbon fiber areas (for example, areas greater than or equal to 40 pm²) suffer increased costs. While the costs for all four non-circular fiber shapes decrease with increased fiber area up to 200 pm².

[00117] When estimating the carbon fiber cost for specific fiber geometries (FIG. 11 ), it is apparent that the non-circular carbon fibers have cost-optimal performance at larger fiber cross-sectional areas (black lines) compared to circular fibers (grey line), with the minimum costs shown by the diamond markers. This is unexpected as the one of the main drivers of energy consumption in the carbon fiber conversion process is the maximum diffusion thickness of the fiber. This becomes larger as the fiber area increases meaning the expected outcome should be the smallest area fiber should have the lowest energy cost/mass if that was the only factor considered.

[00118] As shown in FIG. 12, the energy costs during carbon conversion for three- lobed, six-lobed, eight-lobed and four-lobed carbon fibers are significantly lower than the carbon fiber costs for circular carbon fibers for all carbon fiber areas at a 50 k tow count. Minimum emission levels occur at higher cross-sectional areas for the multi-lobed carbon fibers, with numerical minima shown by the black markers for each shape. Consequently, there is a lower energy demand during carbon fiber production for the lobed geometries. This reduces the cost and associated production emissions.

[00119] Three-lobed carbon fibers have a carbon fiber area greater than 40 pm² in one embodiment of the invention. In other embodiments of the invention, three-lobed carbon fibers have a carbon fiber area greater than or equal to 41 pm² and equal to or less than 200 pm². In another embodiment, three-lobed carbon fibers have a carbon fiber area of at least 50 pm². Three-lobed carbon fibers have a carbon fiber area of at least 60 pm² in one embodiment of this invention. Three-lobed carbon fibers have a carbon fiber area of at least 80 pm² in another embodiment of this invention. Three-lobed carbon fibers have a carbon fiber area of at least 100 pm² in another embodiment of this invention. Three-lobed carbon fibers have a carbon fiber area of at least 140 pm² in another embodiment of this invention. These sizes provide cost effective compressive strength and reduced energy costs and emissions for the non-circular carbon fibers.

[00120] Six-lobed carbon fibers have a carbon fiber area greater than 40 pm² in one embodiment of the invention. In other embodiments of the invention, six-lobed carbon fibers have a carbon fiber area equal to 41 pm² and equal to or less than 200 pm². In another embodiment, six-lobed carbon fibers have a carbon fiber area of at least 50 pm². Six-lobed carbon fibers have a carbon fiber area of at least 60 pm² in one embodiment of this invention. Six-lobed carbon fibers have a carbon fiber area of at least 80 pm² in another embodiment of this invention. Six-lobed carbon fibers have a carbon fiber area of at least 100 pm² in another embodiment of this invention. Six-lobed carbon fibers have a carbon fiber area of at least 140 pm² in another embodiment of this invention. These sizes provide cost effective compressive strength and reduced energy costs and emissions for the non-circular carbon fibers.

[00121] While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes may be made, and equivalents may be substituted for elements thereof without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention is not limited to the particular embodiment disclosed as the best mode contemplated for carrying out this invention, but that the invention will include all embodiments falling within the scope of the appended claims.

Claims

CLAIMS What is claimed is:

1 . A carbon fiber tow, comprising: a plurality of non-circular carbon fibers, wherein a carbon fiber shape of the noncircular carbon fibers includes extensions in an outer surface to create a lobed geometry; wherein the lobed geometry is selected from the group consisting of three lobes, four lobes, six lobes and eight lobes extending from a circumferential core of the noncircular carbon fibers; and wherein each of the plurality of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area greater than 41 micron².

2. The carbon fiber tow of claim 1 , wherein the carbon fiber shape comprises three lobes of equal length from the circumferential core.

3. The carbon fiber tow of claim 1 , wherein the lobed geometry is six or eight lobes comprising carbon fiber shapes comprising lobes of different lobe lengths from the circumferential core.

4. The carbon fiber tow of claim 1 , wherein the non-circular carbon fibers comprise six alternating lobes, wherein a first set of three lobes has a first length and a second set of three lobes has a second length different from the first length, and wherein the first set of three lobes alternate with the second set of three lobes around the circumferential core of each of the non-circular carbon fibers.

5. The carbon fiber tow of claim 1 , wherein each of the plurality of non-circular carbon fibers in the carbon fiber matrix has a carbon fiber cross-sectional area greater than or equal to 45 micron².

6. The carbon fiber tow of claim 1 , wherein each of the plurality of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area greater than or equal to 60 micron².

7. The carbon fiber tow of claim 1 , wherein each of the plurality of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area greater than or

- l- equal to 100 micron².

8. A carbon fiber composite comprising: a matrix material; and a plurality of non-circular carbon fibers embedded in the matrix material, wherein a carbon fiber shape of the non-circular carbon fibers includes extensions in an outer surface to create a lobed geometry; wherein the lobed geometry is selected from the group consisting of three lobes, four lobes, six lobes and eight lobes extending from a circumferential core of the non- circular carbon fibers; and wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than 41 micron².

9. The carbon fiber composite of claim 8, wherein the carbon fiber shape comprises three or four lobes of equal length from the circumferential core.

10. The carbon fiber composite of claim 8, wherein the lobed geometry is six or eight lobes comprising carbon fiber shapes comprising lobes of different lobe lengths from the circumferential core.

11. The carbon fiber composite of claim 8, wherein the non-circular carbon fibers comprise six alternating lobes, wherein a first set of three lobes has a first length and a second set of three lobes has a second length different from the first length, and wherein the first set of three lobes alternate with the second set of three lobes around the circumferential core of each of the non-circular carbon fibers.

12. The carbon fiber composite of claim 8, wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than or equal to 45 micron².

13. The carbon fiber composite of claim 8, wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than or equal to 60 micron².

14. The carbon fiber composite of claim 8, wherein each of the plurality of non-circular carbon fibers in the carbon fiber tow has a carbon fiber cross-sectional area greater than or equal to 100 micron².

15. A wind turbine blade comprising: a carbon fiber composite comprising a matrix material and a plurality of non-circular carbon fibers embedded within the matrix material; wherein a carbon fiber shape of the non-circular carbon fibers includes extensions in an outer surface to create a lobed geometry; wherein the lobed geometry is selected from the group consisting of three, four, six and eight lobes extending from a circumferential core of the non-circular carbon fibers; and wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than 41 micron².

16. The wind turbine blade of claim 15, wherein the carbon fiber shape comprises three or four lobes of equal length from the circumferential core.

17. The wind turbine blade of claim 15, wherein the lobed geometry is six or eight lobes comprising carbon fiber shapes comprising lobes of different lobe lengths from the circumferential core.

18. The wind turbine blade of claim 15, wherein the carbon fiber shape comprises three or four lobes of equal length from the circumferential core.

19. The carbon fiber tow of claim 1 , wherein the lobed geometry is six or eight lobes comprising carbon fiber shapes comprising lobes of different lobe lengths from the circumferential core.

20. The wind turbine blade of claim 15, wherein the non-circular carbon fibers comprise six alternating lobes, wherein a first set of three lobes has a first length and a second set of three lobes has a second length different from the first length, and wherein the first set of three lobes alternate with the second set of three lobes around the circumferential core of each of the non-circular carbon fibers.

21. The wind turbine blade of claim 15, wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than or equal to 45 micron².

22. The wind turbine blade of claim 15, wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than or equal to 60 micron².

23. The wind turbine blade of claim 15, wherein each of the plurality of non-circular carbon fibers in the carbon fiber composite has a carbon fiber cross-sectional area greater than or equal to 100 micron².