Dual cure additive manufacturing

Abstract

A method for manufacturing a composite material includes providing a matrix comprising a fiber material and a thermoset polymer. A length of the fiber tow is impregnated with the thermoset polymer via a Localized-In-Plane Thermal Assist (LITA) impregnation system. A dual curable polymer is provided between a first layer of the matrix and the second layer of the matrix, and the dual curable polymer includes a thermally curable component and a photocurable component curable by exposure to actinic radiation. The dual curable polymer is exposed to actinic radiation to at least partially cure the photocurable component. In this way, a layer of the composite material is formed, and the aforementioned steps are repeated until a desired number of layers have been stacked together to create a structure. The layer or the structure is heated to thermally cure the thermoset polymer and the thermally curable component of the dual curable polymer.

Description

[0001] DUAL CORE ADDITIVE MANUFACTURING

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This application claims priority to United States Provisional Application No. 63 / 452,537, filed March 16, 2023, the content of which is incorporated herein by reference in its entirety for all purposes.

[0004] REFERENCE TO U.S. GOVERNMENT SUPPORT

[0005] This invention was made with government support under Grant No. DE- AR0001576 from the Department of Energy Advanced Research Projects Agency- Energy (ARPA-E) and Grant No. DE-EE0010205 under the Department of Energy Advanced Materials and Manufacturing Technologies Office (AMMTO). The United States has certain rights in the invention.

[0006] FIELD OF THE INVENTION

[0007] Aspects of the present invention relate generally to composite additive manufacturing, and more particularly to a printing method to additively manufacture continuous carbon fiber reinforced thermally curable thermoset composites.

[0008] BACKGROUND OF THE INVENTION

[0009] Composite additive manufacturing (CAM) is a process of manufacturing composite materials layer by layer using additive manufacturing techniques (e.g. 3D printing) in a controlled manner to build up a final part or component. This process has applications in several industries, including but not limited to aerospace, automotive, marine, construction, and sports equipment. CAM enables the production of lightweight and high-strength components that can be tailored to specific applications, leading to improved performance and efficiency. Additionally, CAM can reduce waste, as parts can be produced on demand and with minimal material waste.

[0010] Among various types of composites, continuous fiber reinforced thermoset composite and its additive manufacturing are particularly desirable. First, continuous fiber reinforced composites have higher strength, stiffness, and fatigue resistance compared to short fiber composites, due to continuous fibers being able to provide a continuous load path, resulting in better stress transfer between fibers, and higher interfacial bonding between fiber and matrix. Second, thermoset polymer performs suitably in material quality and processability than thermoplastics, such as higher temperature resistance, better chemical resistance, lower processing temperature window, and lower cost. While thermoplastics and thermosets are present in the composites manufacturing depending on specific applications, thermoset composites are suitable to the areas where high-temperature stability, chemical stability and high structural performances are required. Therefore, composites of continuous fibers and thermosets are suitable for additive manufacturing.

[0011] A problem with thermally curable thermosets is the rapid decrease in viscosity due to the rise in temperature from heating and then the sudden increase due to the curing chemistry. As rapid curable resins (photopolymerization and frontal polymerization) are developed, resin formulation to 3D printing of composites are applied, including for example, ultraviolet (UV)-based snap cure resin, a rapid curable and high glass transition temperature polymer matrix, which is considered an alternative to the thermally curable resin (e.g., epoxy). UV snap cure resins containing mixed photopolymer and epoxy have been demonstrated in fiber tow-based additive manufacturing. However, aUV's limited penetration though the carbon fiber tow, and the disposition of that tow between the resin matrix and the source of photo radiation, may lead to an undesirably low curing degree and low cross-linking density.

[0012] In addition, continuously reinforced thermally curable polymer composite additive manufacturing poses challenges, including slow curing time of thermally curable resins, and the temperature-dependent thermoset viscosity change, which limit the shape design and geometry control of additive manufacturing. In addition, achieving high fiber volume fraction in 3D printed composites can be difficult due to limitations in impregnation and consolidation. Furthermore, interlaminar bonding between printed layers can also be an issue due to the presence of weak boundary layers between the printed layers, leading to reduced mechanical properties.

[0013] Carbon fiber composites offer desirable strength-to-weight and stiffness-to- weight ratios and can provide major weight reductions (up to 70%) and corresponding energy savings (fuel savings during the use of a vehicle) when used to replace traditional structural materials such as steel. Composites manufacturing challenges include low process throughput, complex and multiple processing cycles, intensive energy consumption, low part design flexibility, and poor quality control, all of which can lead to high costs and variable performance. Especially for small-format, complexshaped components, these limitations can limit the use of lightweight advanced composites. Over 70% of the cost of composite products comes from the product fabrication process, and on similar structures, the labor-material cost ratio for composites is much higher than that for metals. State-of-the-art (SOA) technologies for composites manufacturing (e.g., thermoset) include high-cost, complex tooling designs and fabrication along with energy and time consuming, labor-intensive manufacturing steps that include fiber / fabric layup, and resin impregnation / transfer. Although commercial robotic automated fiber / tape placement systems (AFP and ATP) have been used for large-scale manufacturing of aerostructures, these processes typically rely on expensive molds for placement of wet thermoset prepregs combined with long processing times. The geometric complexity of composite structures is often limited by restrictions on the mold design due to the process versatility. Furthermore, additional cost and energy consumption may be experienced for prepregs that must be kept cold prior to manufacturing (so that they don't cure) and for heating prior to using them. All of the foregoing may increase composites manufacturing time, energy consumption, and processing cost in previous composite manufacturing factories. These restrictions constrain the structural and functional capabilities of composites for lightweight component development and applications.

[0014] Conventional apparatuses and systems typically fabricate carbon fiber composites using thermoplastic or a narrow range of rapid curable resins (photopolymerization and frontal polymerization), which may fall short of meeting a wide range of mechanical and durability requirements. For example, the limited penetration of the photo radiation (e.g., ultraviolet, UV) though the carbon fiber tow presents a major challenge to achieving optimal in-situ curing. The snap-cure resin often contains a high volume / weight fraction of UV resin over the thermally curable resin (e.g., epoxy), which compromises epoxy performance, causes limited UV curing kinetics, and induces non-uniform phase-separated domain composition in the matrix. Furthermore, the limited chemical bonding between two highly cured laminates (e.g., via frontal polymerization layer-by-layer 3D printing) may present another major challenge to achieving optimal interlaminar strength.

[0015] Therefore, there remains a desire for improvements with thermoset composite additive manufacturing, and more particularly there remains a desire for improvements in long-term curing chemistry and rapid shape retention, which can be achieved, for example, via 3D printing of composites using continuous reinforcement and thermally curable resins.

[0016] SUMMARY OF THE INVENTION

[0017] Aspects of the invention relate to an apparatus for manufacturing a composite material, such as pursuant to the methods as described herein. The apparatus includes a tow-preg module and a layering module. The tow-preg module includes a source of fiber material, means for causing the fiber material to move along a material path having a direction, and means for impregnating a length of the fiber material with a thermoset polymer to form a matrix. The layering module includes a coater for coating the matrix with a photocurable polymer to form a photocurable- polymer-coated matrix; an interlayer heater positioned to heat the interlayer to a temperature sufficient to cause the thermoset polymer and the photocurable polymer to diffuse together; means for stacking one layer of photocurable-polymer-coated matrix on another layer of photocurable-polymer-coated matrix, forming an interlayer therebetween; a roller positioned to compress the one layer and the other layer; and a source of actinic radiation positioned to expose the photocurable polymer to the actinic radiation in a location along the material path between the interlayer heater and the roller. The apparatus may comprise a robot, wherein the means for stacking the one layer of photocurable-polymer-coated matrix on the other layer of photocurable- polymer-coated matrix is an end effector of the robot, and the robot has a controller programmed with instructions for causing the robot to dispose the layers in a predetermined position relative to one another. The apparatus may further include a source of thermal energy sufficient for thermally curing the thermoset polymer to form the composite. The means for impregnating the length of the fiber tow with the thermoset polymer may be a Localized-In-Plane Thermal Assist (LITA) impregnation system comprising a LITA heater configured to heat the fiber material to generate a thermal gradient in the fiber material trailing the LITA heater relative to the material path direction, and a feeder configured to dispense the thermoset polymer on the heated fiber material.

[0018] Another aspect of the invention relates to a composite material comprising a plurality of cured layers and an interlayer between each adjacent pair of cured layers, each layer comprising a fiber in a matrix of a thermoset polymer, each interlayer comprising interpenetrating network of a dual-cured polymer, the dual-cured polymer comprising a thermoset component and a photocured component. The thermoset component and the thermoset polymer may comprise the same polymer. In some embodiments, the dual-cured polymer may comprise an in-situ diffusion of the photocured component into the thermoset polymer of the matrix. The interlayer may comprise two outer sub-interlayers each comprising the interpenetrating network of dual-cured polymer, and a middle sub-interlayer comprising the photocured component without the thermoset component between the two outer sub-interlayers.

[0019] Still another aspect of the invention relates to a method for manufacturing a composite material. The method includes the steps of providing a matrix comprising a fiber material and a thermoset polymer; providing a dual curable polymer between a first layer of the matrix and the second layer of the matrix, the dual curable polymer comprising a thermally curable component and a photocurable component curable by exposure to actinic radiation; and exposing the dual curable polymer to actinic radiation to at least partially cure the photocurable component. Performing the foregoing steps creates a layer of the composite material, and the method optionally includes repeating the foregoing steps until a desired number of layers have been stacked together to create a structure. Then, the method includes heating the layer or the structure to thermally cure the thermoset polymer and the thermally curable component of the dual curable polymer. The thermoset polymer and the thermally curable component of the dual curable polymer may comprise the same polymer. The dual curable polymer may be formed in situ by coating a surface of the thermoset polymer of the matrix with the photocurable polymer and causing the thermally curable polymer and the photocurable polymer to diffuse together at an interface thereof, which may include heating the thermally curable polymer and the photocurable polymer to cause the diffusion of the photocurable polymer and the thermally curable polymer together to form an interpenetrating network (IPN).

[0020] Another aspect of the invention relates to a method for manufacturing a composite material, the method comprising providing a matrix comprising a fiber material and a first polymer; contacting the matrix with a second polymer; causing the first polymer and the second polymer to diffuse together at an interface between the first polymer and the second polymer; curing the second polymer; and then curing the first polymer. In embodiments, the second polymer has a faster curing rate than a curing rate of the first polymer, uses a different curing initiator than the first polymer, or a combination thereof and / or the second polymer is less viscous than the first polymer. For example, the first polymer may be thermally curable and the second polymer may be photocurable, such as with UV radiation. The fiber material may comprise a carbon fiber, and the matrix may comprise a tow-preg or a pre-preg. The first polymer may comprise a thermoset epoxy. The step of causing the first polymer and the second polymer to diffuse together may be thermally assisted and performed at a temperature above room temperature, such as in a range of 50-60 C, inclusive. Performing the steps (except for curing of the first polymer) may form a layer and may be repeated one or more times to form a multilayered structure prior to curing the first polymer in the multilayered structure as a whole.

[0021] BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The foregoing summary and the following description will be better appreciated and understood in conjunction with the non-limiting examples illustrated in the attached drawing figures, of which:

[0023] FIG. 1 depicts an exemplary method of manufacturing a composite material according to an embodiment of the present invention;

[0024] FIG. 2 is a schematic diagram of the method of FIG. 1;

[0025] FIG. 3Ais a schematic diagram of an apparatus for manufacturing a composite material according to an embodiment of the present invention;

[0026] FIG. 3B depicts the apparatus of FIG. 3A including a robot according to an embodiment of the present invention;

[0027] FIGS. 4A-4F are graphs depicting a characterization of material properties of a thermoset polymer and thermally curable component in accordance with testing of the exemplary method of FIG. 1;

[0028] FIG. 5A is a graph showing a viscosity of the thermoset polymer and thermally curable component of FIGS. 4A-4F in accordance with laboratory assessment of the exemplary method of FIG. 1;

[0029] FIG. 5B is an image of the temperature-dependent diffusion of the thermoset polymer and thermally curable component of FIGS. 4A-4F in accordance with laboratory assessment of the exemplary method of FIG. 1;

[0030] FIG. 6A is a graph depicting a characterization of the thermally curable component of FIGS. 4A-4F at different printing speeds of the apparatus of FIGS. 3A-3B and different number of radiation sources;

[0031] FIG. 6B depicts an image of a matrix formed according to conventional methods and another image of a matrix formed according to the method of FIG. 1;

[0032] FIGS. 7A-7B depict images of the apparatus of FIG. 3B for manufacturing composite coupons in accordance with laboratory assessment of the exemplary method of FIG. 1;

[0033] FIG. 7C depict images of the composite material formed by the apparatus of FIGS. 7A-7B; FIG. 7D is a graph depicting a microstructural characterization of the composite material of FIG. 7C;

[0034] FIGS. 7E-7F depict scanning electron microscope images of a cross-sectional section of the composite material of FIG. 7C;

[0035] FIGS. 7G-7H depict nano-computed tomography images of the composite material of FIG. 7C;

[0036] FIG. 71 depicts a pore distribution image of the composite material of FIG. 7C;

[0037] FIG. 8 depicts another exemplary method of manufacturing a composite material according to an embodiment of the present invention;

[0038] FIG. 9 is a schematic diagram of an exemplary apparatus for manufacturing a composite material according to an embodiment of the present invention;

[0039] FIG. 10A is a schematic diagram showing steps of the method of FIG. 8;

[0040] FIG. 10B is an image of a component of the apparatus of FIG. 9;

[0041] FIG. 10C depicts a schematic diagram (top) and an image (bottom) of a portion of a multi-layered structure formed in accordance with FIG. 10A;

[0042] FIG. 10D is a fluorescent confocal image of a portion of the multi-layered structure formed in accordance with laboratory assessment of the method of FIG. 10A;

[0043] FIG. 11A is an image of an exemplary apparatus for manufacturing a composite material according to an embodiment of the present invention;

[0044] FIG. 11B is a schematic diagram showing an exemplary source of actinic radiation of the apparatus of FIG. 11A;

[0045] FIG. 11C depicts images of a multi-layered structure formed in accordance with laboratory assessment of the apparatus of FIG. 11A;

[0046] FIG. 12A-12B depict images of a composite material according to an embodiment of the present invention;

[0047] FIG. 12C depicts confocal microscopy images of a cross-section of the composite material of FIGS. 12A-12B;

[0048] FIG. 12D depicts a 3D depth map of a surface of the composite material of FIGS. 12A-12B;

[0049] FIG. 13A depicts raw nano-computed tomography images showing a reconstruction of an internal structure of a composite material in accordance with laboratory assessment of the method of FIG. 10A;

[0050] FIG. 13B depicts utilization of machine-learning techniques for image segmentation in accordance with laboratory assessment of the method of FIG. 10A; FIG. 13C depicts a segmented 3D image identifying components of the internal structure of the composite material of FIG. 13A;

[0051] FIG. 14A depicts a schematic diagram of an exemplary apparatus for manufacturing a composite material without a mold; and

[0052] FIG. 14B-14D depict images of the composite material of FIG. 14A. DETAILED DESCRIPTION OF THE INVENTION

[0053] Although the invention is illustrated and described herein with reference to specific embodiments, the invention is not intended to be limited to the details shown. Rather, various modifications may be made in the details within the scope and range of equivalents of the claims and without departing from the invention.

[0054] Additionally, various forms and embodiments of the invention are illustrated in the figures. It will be appreciated that the combination and arrangement of some or all features of any of the embodiments with other embodiments is specifically contemplated herein. Accordingly, this detailed disclosure expressly includes the specific embodiments illustrated herein, combinations and sub-combinations of features of the illustrated embodiments, and variations of the illustrated embodiments.

[0055] In the following detailed description, numerous specific details are set forth by way of examples in order to provide a thorough understanding of the relevant design features. However, it should be apparent to those skilled in the art that the present design features may be practiced without such details. In other instances, well known methods, procedures, components, and circuitry have been described at a relatively high-level, without detail, in order to avoid unnecessarily obscuring aspects of the present design.

[0056] In general, according to one aspect of the invention, with reference to FIGS. 1- 7, a Rapid Interlayer Curing Assisted (RICA) 3D printing strategy is provided. This strategy includes in-situ formation of a dual-cure resin system to enable an interpenetrating network (IPN) at the interlayer in between thermally curable resin / fiber laminates, thereby achieving faster solidification and shape retention in 3D printing, and a subsequent formation of a complete polymerization of major matrix resin (e.g., epoxy) across the interlaminar in post-baking. This strategy allows use of any types of commercially available thermally curable resin and fibers to develop 3D printed composites. A non-limiting example of a composite 3D printing technology, such as a Carbon Form Additive Manufacturing or CarbonForm3D™, is provided. The underlying mechanism of the CarbonForm3D™ is based on the formation of an IPN at the interlayer between thermally curable thermosetting polymer / fiber composites to address the challenge of difficulty in maintaining shape during printing and weakness in interlaminar strength.

[0057] Thus, according to one embodiment of this present invention, a method for manufacturing a composite material generally includes a diffusion of thermoset polymer (e.g., epoxy) and the thermally curable component (e.g., UV resin) at the interface forms a dual-cure resin system that enables an interpenetrating polymer network (IPN) 180 at the interlayer during UV curing and sequential thermal curing. For example, FIG. 1 illustrates method 1000, which includes the following steps:

[0058] (a) providing a matrix comprising a fiber material and a thermoset polymer (step 1010);

[0059] (b) providing a dual curable polymer between a first layer of the matrix and a second layer of the matrix (step 1020);

[0060] (c) exposing the dual curable polymer to actinic radiation to at least partially cure the photocurable component (step 1030); and

[0061] (d) heating the layer or the structure to thermally cure the thermoset polymer and a thermally curable component of the dual curable polymer (step 1040).

[0062] Step 1010 includes providing a matrix having a fiber material and a thermoset polymer. In one embodiment, as illustrated in the schematic diagram of FIG. 2, the matrix 110 comprises a carbon fiber structure 112 impregnated with a high-viscous and thermally slow-curable resin 114 (e.g. tow-preg, pre-preg, etc.).

[0063] Step 1020 includes providing a dual curable polymer 130 between a first layer of the matrix 110 and a second layer of the matrix 110. In one embodiment, the dual curable polymer includes a thermally curable component 134 and a photocurable component 136 curable by exposure to actinic radiation. As shown in FIG. 2, the dual curable polymer 130 is formed in situ by coating a surface of the thermoset polymer 114 of the matrix 110 with the photocurable polymer 136 and causing the thermally curable polymer 134 and the photocurable polymer 136 to diffuse together at an interface thereof. In one non-limiting example, the dual curable polymer 130 includes a layer of low-viscous and fast-curable resin, such as a thin layer of ultraviolet (UV) curable resin, coated on the matrix. Additionally or optionally, the thermoset polymer 114 of the matrix 110 and the thermally curable component 134 of the dual curable polymer 130 comprise the same polymer. Step 1030 provides that the dual curable polymer 130 is exposed to actinic radiation (Step 2 in FIG. 2) to at least partially cure the photocurable component 136. With performance of this step, the interlayer between a first layer of photocurable- polymer-coated matrix 102a and another layer of photocurable-polymer-coated matrix 102b, is solidified and its shape is maintained. Further, actinic radiation (e.g. UV radiation) initiates the cross-linking at a nip point of two thermally curable resin / fiber laminates to hold the shape during deposition.

[0064] In this way, performing steps 1010, 1020, and 1030 creates a layer 140 of the composite material. Optionally, method 100 includes repeating steps 1010, 1020, and 1030 until a desired number of layers 140 have been stacked together to create a multi-layered structure, and step 1040 is performed as a subsequent step.

[0065] Step 1040, as shown in Step 3 of FIG. 2, includes heating the layer 140 or the multi-layered structure to thermally cure the thermoset polymer 114 and the thermally curable component 134 of the dual curable polymer 130. In this way, the post thermal curing (Step 3 in FIG. 2) completes the polymerization and cross-linking of the thermally curable polymer of the matrix. Additionally or optionally, the post thermal curing can occur in oven, autoclave, or out-of-oven method such as heating blankets, depending on the resin cure cycle and applications.

[0066] Additionally or optionally, method 1000 further includes a step of heating the thermally curable polymer 134 and the photocurable polymer 136 to cause the diffusion of the photocurable polymer 136 and the thermally curable polymer 134 together to form an interpenetrating network (IPN), as illustrated by Step 1 in FIG. 2. In this way, the application of heat triggers a dynamic polymer diffusion between the two different resins (e.g. the thermoset polymer 114 of the matrix 110 and the thermally curable component 134 of the dual curable polymer 130) at the interface to in situ form a dual-cure system. The application of heat brings about a reduction of polymer viscosity and promotes the diffusion of UV resin 136 and epoxy resin 134.

[0067] As a result of the performance of steps 1010-1040, the shape and interlaminar strength as well as the majority use of high-performance thermally curable resin as matrix become feasible for 3D printed composite materials 190 comprising layers 140 (FIG. 3B).

[0068] With reference to FIGS. 3A-3B, the foregoing method 1000 may be performed by an apparatus 100 for manufacturing a composite material, such as composite material 200. In general, apparatus 100 includes a tow-preg module 150. The tow-preg module 150 includes a source of fiber material 152; means for causing the fiber material to move along a material path having a direction (FIG. 3A); and means 154 for impregnating a length of the fiber material 152 with a thermoset polymer to form a matrix 110.

[0069] The apparatus 100 further includes a layering module 170. The layering module 170 has a coater 172 and an interlayer heater. The coater 172 is configured for coating the matrix with a photocurable polymer 136 to form a photocurable-polymer-coated matrix 102a, 102b. Further, the interlayer heater is positioned to heat the interlayer to a temperature sufficient to cause the thermoset polymer 134 and the photocurable polymer 136 to diffuse together. The layering module 170 also includes a means for stacking one layer of photocurable-polymer-coated matrix 102a on another layer of photocurable-polymer-coated matrix 102b, forming an interlayer therebetween. The layer module 170 also includes a roller 176 and a source 174 of actinic radiation positioned to expose the photocurable polymer 136 to the actinic radiation in a location along the material path between the interlayer heater and the roller 176. In this way, apparatus 100 initiates curing of the photocurable polymer 136 at the interface 180 between the photocurable polymer 136 of the first layer 102a and photocurable polymer 136 of the second layer 102b prior to compacting the first layer 102a and the second layer 102b together (e.g. via roller 176). Thus, in embodiment in which the polymer is photocurable (e.g. UV curable polymer 136), and initiating the curing includes directing actinic radiation (e.g. by shining the source 174 of UV light, such as a plurality of UV LEDs) toward the interface 180.

[0070] In one embodiment, the method 1000 is performed by the apparatus 100 comprising a robot 132. The robot 132 has an end effector 160 for stacking the one layer of photocurable-polymer-coated matrix 102a on the other layer of photocurable- polymer-coated matrix 102b in a predetermined position. With reference to FIGS. 3A- 3B, the apparatus 100 including a robot 132, such as Carbon Form Additive Manufacturing (CF-AM™), includes the end effector 160 mounted on a robotic arm that controls the material path direction and a consolidation force to deposit, laminate, and cure tow-pregs 110 into composites 200.

[0071] In one embodiment, as shown in the schematic diagram illustrated in FIG. 3B, the tow-preg module 150 of the apparatus 100 includes the source of fiber material 152, such as dry carbon fiber tow, which may be fitted to a dry carbon fiber creel. The tow-preg 150 module includes the means 154 for impregnating the length of the fiber material with the thermoset polymer, which in one non-limiting example, includes a Localized-In-Plane Thermal Assist (LITA) impregnation system. The step 1010 comprises impregnating the fiber 152 and step 1010 thus includes performing the LITA technique as described in U.S. Published Patent Application No. US20220048246A1, titled METHOD AND SYSTEM FOR DYNAMIC CAPILLARY-DRIVEN ADDITIVE MANUFACTURING OF CONTINUOUS FIBER COMPOSITE, filed August 17, 2021, incorporated herein by references in its entirety for all purposes. The LITA system includes a LITA heater configured to heat the fiber material 152 to generate a thermal gradient in the fiber material trailing the LITA heater relative to the material path direction, and a feeder configured to dispense the thermoset polymer on the heated fiber material. Additionally or optionally, the apparatus 100 includes a compaction roller 176 to laminate and consolidate the deposited tow-prepregs 110.

[0072] As stated above, the layering module 170 of the apparatus 100 has a coater and an interlayer heater. In one embodiment, the coater 172 includes a UV-curable resin coating system and the interlayer heater includes a heating gun. Additionally, the source 174 of actinic radiation includes a UV LED. Thus, in one non-limiting example of an operation of apparatus 100, the matrix 110 comprising the fiber material 152 and thermoset polymer 114 includes tow-pregs 110 containing thermally curable epoxy 114 and carbon fiber 152 having high fiber volume fraction (60 voL%) and B-stage epoxy state. The B-stage tow-preg is then transferred to the UV-curable resin coating system 170 for coating a thin UV-curable resin 136 (10~15 pm) onto the both sides of tow- preg 110. The source of thermal energy from the heating gun and the source 174 of actinic radiation from the UV LED are applied. The heat or thermal energy is to initiate a dynamic resin diffusion of the two resins (134, 136) to form a dual-cure resin system at the interface 180. The UV light from the UV LED 174 is to initiate the curing of the UV-curable resin 136 portion of the resin. Heating and UV lighting are designed at the nip point when the compaction roller 176 laminate and consolidate the deposited towprepregs 110.

[0073] According to another aspect of the invention, with reference to FIGS. 8-14, to address the challenges in thermally curable polymer composite additive manufacturing, a method 2000 for manufacturing a composite material 290 (FIG. 9) includes a rapid composite preformation strategy, such as a Tow-Preg Cladding™ (TPC). This TPC approach utilizes a thin and rapidly curable dual-cure resin 230, comprising both a thermally curable resin 234 (such as epoxy) and a rapid photo-curable resin 236 to coat over a thermally curable resin tow-preg 210. The objective is to create a multilayered structure 240 comprising a core-shell tow-preg structure with a rigid cladding as the shell that self-supports the tow-pregs and allows them to be preformed into desired geometries and patterns. The dual-cure shell structure 240 (FIG. 9) also contains fibers 252, which allows the printed composites to possess more uniform reinforcement. The tow-preg 210 undergoes a two-step curing mechanism, rapid UV curing and post thermal curing, into an interpenetrating polymer network (IPN) 280 at the interlayer between laminates once the thermally curable resin 234 is cured, resulting in high interlaminar bonding in the composites.

[0074] For example, FIG. 8 illustrates method 2000, which includes the following steps:

[0075] (a) providing a matrix comprising a fiber material and a first polymer (step 2010);

[0076] (b) contacting the matrix with a second polymer (step 2020);

[0077] (c) causing the first polymer and the second polymer to diffuse together at an interface between the first polymer and the second polymer (step 2030);

[0078] (d) curing the second polymer (step 2040); and

[0079] (e) curing the first polymer (step 2050).

[0080] Step 2010 includes providing a matrix comprising a fiber material and a first polymer. In one non-limiting example, as described in the embodiment above with respect to the method of FIG. 1, the matrix 210 is similar to matrix 110, in that matrix 210 comprises a carbon fiber structure 212 impregnated with first polymer 214. The first polymer 214 includes a thermoset epoxy.

[0081] Step 2020 includes contacting the matrix with a second polymer. In one nonlimiting example, as described in the embodiment above with respect to the method of FIG. 1, the second polymer 236 includes a photocurable component (e.g. curable by exposure to UV radiation). Additionally or optionally, the second polymer 236 has a faster curing rate than a curing rate of the first polymer 214, uses a different curing initiator than the first polymer 214, or a combination thereof. Still further, the second polymer 236 is less viscous than the first polymer 214. Still further, in one embodiment, wherein 2010 and 2020 include sequential steps of step 2010 including impregnating the fiber 212 with the first polymer 214 to form the matrix 210, then step 2020 including coating the second polymer 236 over the matrix 210. Additionally or optionally, step 2020 includes performing the localized in-plane thermal assist (LITA) technique. In one embodiment, at least two layers of second-polymer coated matrix are formed and the the at least two layers are compacted together. In this way, an interface 280 is formed between the second polymer 236 of the first layer and the second polymer 236 of the second layer, and curing of the second polymer 236 is initiated at the interface 280 prior to compacting the first layer 202a and the second layer 202b together. When the second polymer 236 is photocurable, initiating the curing comprises directing a source 274 of actinic radiation toward the interface 280. Additionally or optionally, the second polymer 236 is UV curable, and directing the source 274 of actinic radiation comprises shining a source of UV light at the interface 280. For example, the source 274 of UV light comprises a plurality of UV LEDs. In an exemplary embodiment, as shown in FIG. 11B, the plurality of UV LEDs 274 comprises a first LED 274a for selective curing of the second polymer 236 on a first surface of the first layer 202a, a second LED 274b for selective curing of the second polymer 236 at the interface 280, and a third LED 274c for selective curing of the second polymer 236 on an opposite surface of the first surface of the first layer 202a.

[0082] Step 2030 includes causing the first polymer 214 and the second polymer 236 to diffuse together at an interface, such as interface 280 (similar to interface 180 described above) between the first polymer 214 and the second polymer 236. In one embodiment, step 2030 is thermally assisted and performed at a temperature above room temperature. For example, step 2030 is performed at a range of temperature between is 50°C to 60 °C.

[0083] In one embodiment, steps 2010 to step 2040 form a layer, such as layer 240, and repeating steps 2010 to step 2040 one or more times forms a multilayered structure prior to performing step 2050 on the multilayered structure.

[0084] Additionally or optionally, as shown in FIGS. 10A-10D, method 2000 further includes partially removing the first polymer 214 from an outer surface of the matrix 210. Additionally or optionally, partially removing the first polymer 210 from the matrix 210 is performed prior to contacting the matrix with a second polymer 236 (step 2020). This may be achieved using a scraper 300 (FIG. 10B) configured to partially remove the first polymer 214 from the outer surface of the matrix 210. The scraper 300 includes teeth 302 extending for a predetermined length and spaced apart from each other for a predetermined distance, thereby creating a plurality of gaps in the first polymer 214 (FIG. 10C) to receive the second polymer 236 (as shown in FIG. 10D). In an exemplary embodiment, the predetermined length is in a range between 0.1 and 0.2 mm. Additionally or optionally, the predetermined distance is in a range between 1 mm and 5 mm. In this configuration, the teeth are configured to penetrate into matrix 210 and remove excessive polymer 214, thereby ensuring a sufficiently high fiber content (e.g. relative to the polymer content) in the matrix 210 for a better quality.

[0085] EXAMPLES

[0086] The co-inventors assessed the exemplary methods and apparatuses as disclosed herein in a laboratory setting, to validate feasibility and functionality of the components of the subject systems, as well as verified any updates or improvements made.

[0087] Example 1: ADDITIVE MANUFACTURING OF CONTINUOUSLY REINFORCED THERMALLY CURABLE THERMOSET COMPOSITES WITH RAPID INTERLAYER CURING Materials and Methods

[0088] A commercial diglycidyl ether of bisphenol A (DGEBA)-based epoxy resin, mixed with a methyl tetrahydrophthalic anhydride (MTHPA)-based hardener in the weight ratio of 5 : 1 were used as the thermally curable thermoset matrix. The curing temperature for the epoxy-hardener resin mixture is 154 °C. To match the epoxy's past-bake at 154 °C, an acrylate-based UV-curable resin was developed for the high temperature use.

[0089] Tensile, flexural, and interlaminar shear tests of the printed composite specimens were all conducted in Instron 5567 (Instron, Norwood, MA) with a 250-kN load cell. The test rate for uniaxial tensile testing was 0.05 in / min, following ASTM D3039. The test rate for flexural testing was 0.50 in / min, with span-to-thickness ratio as 25 : 1, following ASTM D790. The test rate for interlaminar shear testing was 0.50 in / min, with span-to-thickness ratio as 10 : 1, adjusted according to actual thickness of specimens, and following ASTM D2344. Tensile and flexural tests of the fully cured epoxy and UV resin thin films (thickness : 0.50 mm) were also conducted in Instron 5567. The test rate for tensile testing of cured resin thin films was 12.5 mm / min, following ASTM D882. The flexural test rate was 0.50 in / min, with span-to-thickness ratio as 25 : 1.

[0090] Thermogravimetric analysis (TGA) of the cured epoxy and UV resin was evaluated using Discovery TGA (TA Instruments, New Castle, DE) equipment in air, with heat ramp rate as 10 °C / min from room temperature to 650 °C. Differential scanning calorimetry of the cured resins was tested using Discovery DSC (TA Instruments, New Castle, DE) in air, with heat ramp rate as 10 °C / min from room temperature to 300 °C. Dynamic mechanical analysis (DMA) was conducted in tension mode in Q800 (TA Instruments, New Castle, DE), with frequency as 1 Hz, and strain as 0.02%. Fourier-transform infrared spectroscopy (FTIR) of the UV resin was measured on an ATR crystal using Thermo Nicolet 6700 FT-IR spectrometer (Thermo Fisher Scientific, Waltham, MA). Nano-CT was conducted in Rigaku nano3DX (Rigaku, Tokyo, Japan) for 800 projection images, and the obtained images were processed using Dragonfly software (Object Research Systems, Montreal, Canada).

[0091] Characterization of EDOXV and UV Resin

[0092] FIG. 4A shows a thermogravimetric analysis (TGA) of cured epoxy and UV resin, tested from room temperature to 650 C° with heat ramp of 10 C° / min in air. FIG. 4B is a graph showing a differential scanning calorimetry (DSC) curve of epoxy and UV resin, tested from room temperature to 300 C° with heat ramp of 10 C° / min in air. FIG. 4C is a graph showing storage modulus, loss modulus and tan(delta) obtained via dynamic mechanical analysis (DMA) testing in tension mode (frequency: 1 Hz, strain: 0.02 %). FIG. 4D is a graph showing temperature-dependent variation of XY area shrinkage and volume shrinkage. FIG. 4E is a graph showing tensile strength and tensile modulus of fully cured epoxy and UV resin thin films, tested using the ASTM D882 Standard. The test rate was 12.5 mm / min. FIG. 4F is a graph showing flexural strength and modulus of fully cured epoxy and UV resin thin films, tested using ASTM D790 Standard. The test rate was 0.50 in / min, and the span-to-thickness ratio was 25:1.

[0093] The TGA profile (heat ramp rate: 10 C° / min in air) of the cured UV-curable resin is shown in FIG. 4A. Thermally cured epoxy resin is the control sample. The result shows that the weight loss of fully cured UV resin was only 0.33 % at the composites curing temperature (154 °C), and it is slightly smaller than weight loss of epoxy (0.41 %) at the same temperature. Both epoxy and UV resin exhibit similar thermal stability with weight loss less than 1 % by 250 °C.

[0094] The TGA also shows the weight loss of the decomposition process of epoxy and UV resin in cured state, as shown in FIG. 4A. The first stage (the "higher knee") featured drying and physical escape of CO2 from the resin. For epoxy, this stage ranged from room temperature to 262 °C, with weight loss of 1.10%. For UV resin, this stage ranged from room temperature to 271 °C, with weight loss of 0.88%. The higher knee is at the transition between the first stage and the second stage. The second stage (the "middle knee") featured pyrolysis of the resin. For the epoxy, this stage ranged from 262 to 506 °C, with the weight loss as 76.66%. The maximum weight loss ratio of 1.93% / °C occurred at 404 °C. Bisphenol-A (BPA) was released by cleavage of the ether bond. Then BPA was decomposed into phenol. The cleavage of C-C bond neighboring to the benzene ring also resulted in formation of phenol. For UV resin, this stage ranged from 271 to 502 °C, with the weight loss as 84.81%. The maximum weight loss ratio of 1.96% / °C occurred at 443 °C. The lower knee is at the beginning of the third stage. The third stage (the "lower knee") featured complete degradation of the resin. For the epoxy, the remaining BPA further degraded. It reached a weight loss of 99.70% by 650 °C, with the maximum weight loss ratio of 0.27% / °C at 586°C. For UV resin, it reached a weight loss of 99.57% by 650 °C, with the maximum weight loss ratio of 0.20% / °C at 559 °C.

[0095] A differential scanning calorimetry (DSC) (heat ramp rate of 10 °C / min in air) was applied to the two resins as shown in FIG. 4B. Result shows that there is no exothermic curing peak for epoxy, indicating it is fully cured. The glass transition of epoxy is observed at 134.8 °C, and the glass transition of UV resin is observed at 36.4 °C and 193.8 °C. In dynamic mechanical analysis (DMA) (as shown in FIG. 4C), the storage modulus of UV resin is almost identical to epoxy (~ 2500 MPa) from room temperature to the Tg of epoxy, and becomes one magnitude higher over the Tg of epoxy. The loss modulus and the tan(5) curve of epoxy are observed with a peak at 150 and 155 °C, respectively, which marks the glass transition temperature range. As the ratio of loss modulus to storage modulus, tan(6) of epoxy is mostly lower than UV resin, meaning that cured epoxy is more elastic than UV resin. For UV resin, no significant changes in values of storage, loss moduli and tan(s) curve are observed, indicating that no glass transition occurs within the measuring temperature range of 190 °C. The DMA results indicate that the UV resin can serve the temperature of the epoxy post curing (154 °C) and could be even stable at higher service temperature than epoxy.

[0096] To study the UV resin dimensional thermal stability, UV resin cubic samples (lcmxicmxicm) were prepared and the sample s dimensional shrinkage by X-Y area and volume (X-Y-Z) were measured from 80 to 160°C in air. As statistically summarized in Table 1 below and FIG. 4D, the X-Y area and volume of samples had an expansion from 80 to 120 °C, then the samples had shrinkage as temperature increases. The dimensional change is within 1%, which is comparable with epoxy resin thermal shrinkage. The tensile and flexural properties of the cured epoxy and UV resin are shown in FIGS. 4E and 4F. Following the standard established by ASTM D882, the thin film specimens (60 mm x 10 mm x 0.50 mm) were tested for tensile strength and modulus properties at a rate of 12.5 mm / min. For epoxy, the tensile strength was 21.14± 2.07 MPa, and the tensile modulus was 2.78 ± 0.23 GPa. For UV resin, the tensile strength was 24.44 ± 4.32 MPa, and the tensile modulus was 2.86 ± 0.18 GPa. The flexural strength and modulus of epoxy and UV resin thin film specimens were tested following the standard established by ASTM D790 at a test rate of 0.50 in / min, and the span-to-thickness ratio was 25: 1. For epoxy, the flexural strength was 112.36 ± 0.51 MPa, and the flexural modulus was 3.43 ± 0.34 GPa. For UV resin, the flexural strength was 104.03 ± 4.02 MPa, and the flexural modulus was 2.75 ± 0.08 GPa.

[0097] Table 1 : Dimensional change of UV resins at different temperatures (sample size: lcmx lcmx lcm)

[0098] Demonstration of a Temperature-Dependent Diffusion of Epoxy and UV Resin at Elevated Temperature

[0099] FIG. 5A shows a graph showing a viscosity of epoxy and UV resin vs. temperature (shear rate: 1 s-1). FIG. 5B is a photo image of the diffusion of epoxy and UV resin between 25°C to 60 °C, with one enlarged photo image of the diffusion of epoxy and UV resin at 50°C.

[0100] The temperature-dependent dynamic polymer diffusion of the epoxy and UV resin makes it possible to form a dual-cure resin system that can be later used to create an interpenetrating polymer network at the interlayer between laminated epoxy / fiber tow-pregs, followed by exposing UV light and heat. This dynamic polymer diffusion phenomenon is illustrated in FIGS. 5A-5B by visually observing the diffusion which occurred at the interface between epoxy and UV resin as the temperature increased. The viscosity vs. temperature of epoxy and UV resin is measured, respectively to find out the onset temperature for the lowest viscosity of the two resins. The initial viscosity at room temperature for epoxy and UV resin is 4.76 Pa*s, and 0.57 Pa«s. It was found that at 60 °C, the epoxy exhibited the lowest viscosity of 0.42 Pa«s. Beyond this temperature, the cross-linking starts and viscosity increases. The UV resin also exhibited the lowest viscosity of 0.15 Pa«s at 60 °C (FIG. 5A). Based on this, BOBO °C is determined as the rapid diffusion temperature range. The epoxy and UV resins were placed in a liquid state on a glass slide and a hot plate underneath the glass slide was used to control the temperature. The interface of two polymers were recorded by a camera. As the hot plate temperature increased, polymer diffusion begins to merge with each other. When zoomed in, the magnified view shown in the photo of the interface of two polymers heated at 50°C that the low viscous UV resin diffused faster and merged into the epoxy (FIG. 5B). This diffusion is a kinetically controlled process, and it provides evidence that by simply heating up a UV resin coated epoxy / fiber tow- preg, a dual-cure resin system could be formed on tow-preg surface, which could be further used to create a sequential IPN bonding at the interlayer between printed tow- pregs.

[0101] Characterization of UV Resin Cured at Different Printing Speeds and UV Powers To determine the optimal printing speed and the power of UV LEDs for the best 3D printing performance at apparatus 100, including the the CF-AM end effector 160, the solidification of the UV resin was tested at different moving speed of end effector (50 mm / s, 75 mm / s, 100 mm / s) and different number of UV LEDs (1 to 3, 2.9 W / cm2for each UV LED). The reduction of C=C peak are used to determine the solidification of acrylate-based UV resin coating. The solidification can be determined by using resin curing degree, which is the percentage of acrylate monomers that transform into crosslinked networks. The curing degree can be calculated by the change of the characteristic acrylate peak intensity obtained from Fourier Transform Infrared Spectroscopy (FTIR). According to the FTIR spectra showing the UV resin cured at one, two, and three UV LEDs and three printing speeds, the characteristic peak at 1720 cm1represents the C=O bond of the acrylate monomer, and is used as the reference peak for the intensity calculation. The characteristic peak at 810 cm1represents the C=C bond of the acrylate monomer. The UV curing is a double bond cleavage process that opens the C=C bond and form a cross-linked C-C structure. The reduction of C=C bond refers to the cross-linking and curing of acrylate UV reins. As UV light power increased and printing speed reduced, the peak at 810 cm1decreased and it indicates a higher degree of curing. A fully cured UV resin sample prepared by digital light processing (DLP) 3D UV curing and flooded UV curing was used as the control. The areas of the peak at 810 cm1and the reference peak at 1720 cm1are used to calculate the degree of curing, following the equation (1). The results of the degree of curing (DOC), which is calculated according to Equation 1 below, are shown in FIG. 6A. The resin cured at 50 mm / s with 3 UV LEDs (8.7 W / cm2) showed the highest degree of curing at 64.1 %, which is close to that of fully cured control sample (73.3 %). Equation 1

[0102] The TGA of the sample cured at 50 mm / s and with 3 UV LEDs was measured. The weight loss of the cured UV resin cured is only 1.17 % at 154°C (the epoxy curing temperature), so it can meet the temperature requirement for post-bake of the epoxy resin. FIG. 6B is a photo image of an epoxy / fiber tow-preg drooped from the clamp, while the epoxy / fiber tow-preg with cured UV resin maintained the shape. This photo images confirm that the thin UV resin (10-15 pm) could effectively and mechanically support the shape retention of the epoxy / fiber tow-preg.

[0103] Microstructural Characterization of Printed Carbon Fiber Composites

[0104] FIG. 7A is a photo image of a 3D printer of a composite material, such as composite coupons. FIG. 7B is a photo image of epoxy / fiber tow-pregs being deposited on a print platform with UV curing at the interlayer. The width of the tow-preg was 8 mm. FIG. 7C is a photo image of a 3D printed composite coupon with UV-assisted interlayers. FIG. 7D is a graph showing a thermogravimetric analysis (TGA) of all the components in the composites (heat ramp rate: 20 C / min) in nitrogen. FIG. 7E is a scanning electron microscope (SEM) image of the cross-sectional printed composites. FIG. 7F is an enlarged SEM image showing the interlayer between epoxy / fiber laminates. FIGS. 7G-7H show nano-computed tomography (nano-CT) images of the printed composites. FIG. 71 shows a pore distribution image of the printed composites. The pore volume was calculated to be 0.71%.

[0105] The microstructural characterization of a 3D printed composite coupon is shown in FIGS. 7A-7I. The carbon fiber composite was fabricated on the exemplary end effector of apparatus 100 as shown in FIG. 7A. The carbon fibers used were T300 12k continuous fibers. The end effector motion was controlled by UFactory xArm 6 with xArm Studio as the software. Three UV LEDs were installed on the end effector to cure the photocurable resin sheath of epoxy-fiber tow-preg. The UV LED has wavelength of 405 nm, and maximum intensity of 2.9 W / cm2. The printing speed was set at 50 mm / s (Table 2). Table 2: Printing parameters of CarbonForm 3D technique

[0106] To print the composite, a dry carbon fiber tow (12K) was first preimpregnated with epoxy resin in the epoxy reservoir to form an epoxy tow-preg (width: 8 mm). The fiber volume fraction of the tow-preg was a controlled by the LITA epoxy impregnation technique. Then the tow-preg was coated with a thin layer of UV-curable resin to form a core-sheath structure, where the epoxy tow-preg is the core and the UV-curable resin is the sheath. Heat gun and UV LEDs were applied to initiate the polymer interdiffusion and solidify the UV-curable resin. The printing direction is parallel to the fiber direction of the tow-preg. Multilayer tow-pregs were placed by repeating the continuous process including polymer impregnation, coating, and UV curing to form a thick composite, as shown in FIG. 7B. Post thermal baking at 154 °C in oven was applied to fully cure the epoxy matrix of printed composite. A 5-layer unidirectional (UD) composite coupons were printed by the apparatus 100, such as UV-assisted CarbonForm3D™ end effector and the sample photo image is shown in FIG. 7C.

[0107] The fiber volume fraction of the printed composites was measured by thermogravimetric analysis (TGA) method, shown in FIG. 7D. Samples were treated in nitrogen from room temperature to 1100 °C with a heat ramp rate of 20 °C / min. The weight fraction of the fiber and resin matrices was calculated from the weight losses of all the components in the specimen by the end of the test, according to ASTM D3171. Based on the TGA results in FI.G 7D, the calculated fiber weight fraction was 56.06 wt %, the epoxy weight fraction was 39.19 wt%, and the UV resin coating was 4.75 wt%. The fiber volume fraction was then calculated from equation (2), according to ASTM D3171. Pspecimen

[0108] Vf = X WfEquation

[0109] P fiber

[0110] 2

[0111] SEM image in FIG. 7B shows that a UV resin interlayer existed between two deposited carbon fiber layers, and the enlarged SEM image (FIG. 7F) shows an interlayer with a thickness of 20-30 pm sandwiched between two epoxy tow-pregs. No voids were detected at the interlayer. X-ray computed tomography (CT) was adopted for non-destructive inspection of the fiber architecture and identification of fabrication- induced defects, if any, in the internal structure of the composite (FIGS. 7G and 7H). Carbon fibers are highly aligned and densely packed within the epoxy matrix. A UV resin interlayer could be observed between two dense epoxy / fiber layers, due to a relatively low carbon content at the interlayer. The entire scanned model shows that the printed composite sample has a dense structure, as shown in FIG. 7F.

[0112] Comparison of Mechanical Performance from Various Additive Manufacturing

[0113] Mechanical testing coupons for uniaxial tensile and flexural properties, and interlaminar shear strength were prepared and characterized according to the ASTM standards. To compare the testing results, we prepared control samples (60 mm x 8 mm x 0.60 mm, the same dimensions of composites samples with UV resin coating) for mechanical testing. The control sample was made by the same epoxy / carbon fiber tow- pregs, which were prepared using the same process of epoxy impregnation but without UV-curable resin coating. Samples were placed by the end effector and post cured in oven with the same heating setup. The printed sample exhibited a longitudinal tensile strength of 1075.27 ± 81.71 MPa, and a longitudinal elastic modulus of 45.24 ± 5.03 GPa. The control sample exhibited a longitudinal tensile strength of 964.49 ± 105.98 MPa, and a longitudinal elastic modulus of 57.53 ± 14.87 GPa. The uniaxial tensile testing was performed on four specimens roughly (60mmx8mmx0.60mm) following ASTM D3039, with a test rate of 0.05 in / min. The flexural strength was 1209.50 ± 128.06 MPa for the 3D printed composites, and 822.79 ± 97.06 MPa for the control sample. The flexural modulus was 68.01 ± 12.79 GPa for the 3D printed composites, 55.75 ± 9.57 GPa for the control sample. Those flexural properties were obtained following ASTM D790 for four specimens roughly (60mm x 8mm x 0.60 mm), with a test rate of 0.50 in / min and a span-to-thickness ratio of 25: 1. The interlaminar shear strength was 37.17 ± 4.54 MPa for the 3D printed composites, and 50.45 ± 1.48 MPa for the control sample, following ASTM D2344, with a test rate of 0.50 in / min and span-to-thickness ratio of 10:1 (Table 3). The interlaminar shear strength of 3D printed composites is lower than the control sample, possibly due to the rich resin interlayer observed in the cross-section SEM images of printed composites.

[0114] Table 3 : Mechanical Properties of 3D-Printed Composites (with UV resin coating) and Composites Control Sample

[0115] The mechanical performance of the prototype 3D printed composite is the result of the printed sample using fiber material including a T300 carbon fiber (tensile strength: 3530 MPa, elastic modulus: 230 GPa) as the reinforcement, epoxy resin (tensile strength: 21.14 MPa, elastic modulus: 2.78 GPa) as the matrix, and UV resin (tensile strength: 24.44 MPa, elastic modulus: 2.86 GPa) as the interlayer. If intermediate modulus carbon fibers (e.g., Toray T1100) is implemented in connected with apparatus 100 and method 1000, it is expected that the projected longitudinal tensile strength of the printed composites would be further improved to reach 2302.14 MPa, and the longitudinal projected elastic modulus would be 86.63 GPa. Likewise, the flexural strength is projected as 2046.18 MPa, and the flexural modulus would be 144.46 GPa if T1100 is used. The projected performance is calculated based on the rule of mixture for composite.

[0116] For the projected tensile strength:

[0117] Based on existing fiber volume fraction and resin system, the Rule of Mixture was utilized, as well as theoretical tensile strength of T300 carbon fiber and experimental tensile strength of epoxy resin and UV-curable resin, and experimental tensile strength of composites to obtain a coefficient acf(axis the modified coefficient for the carbon fiber portion of Equation 3 below). quation

[0118] The projected tensile strength of composites when using T1100 fibers is calculated with the same a1-.

[0119] The tensile modulus is calculated using the same method (as shown in Equation 4): Equation

[0120] 4

[0121] 165GPa X 44.2% X a2+ 2 7QGPa X 46.0% + 2.86 GPa X 5.9% = 45.25 GPa a2= 0.600

[0122] Thus, the projected performance using T1100 as fibers:

[0123] 320GPa X 44.2% X a2+ 2.78GPa X 46.0% + 2.86 GPa X 5.9% = £composttes86.31 GPa

[0124] Alternative strategy to the 3D printing of composites with continuous reinforcement and thermally curable resins is to use UV or dual-cure resin as the matrix. The limited penetration of the UV radiation though the carbon fiber tow presents a challenge to achieving optimal in-situ curing. Furthermore, for dual-cure resin (also known as snap-cure resin), a high volume / weight fraction of UV resin over thermally curable resin (e.g., epoxy) has to be adopted as the matrix material to ensure the rapid UV-induced shape retention. This will compromise the epoxy mechanical performance and cause limited UV curing kinetics as well as non-uniform phase-separated domain composition in the matrix. Thus, the UV-assisted interlayer strategy as described herein addresses those challenges. The main polymer matrix is still thermally curable epoxy and can be extended to many other commercially available resin systems. The thin UV curable resin coating on epoxy / fiber tow-prepreg or prepreg could be thermally triggered to promote polymer diffusion to form a thin dual-cure system formed at the interface, and could be further formed into an interpenetrating polymer network at the interlayer between epoxy / fiber laminates and cross-link together with the entire epoxy matrix in the sequential thermal heating.

[0125] Thus, a dynamic polymer diffusion approach during 3D printing forms an interpenetrating network at the interlayer in between thermally curable resin / fiber laminates, thereby enabling faster solidification and improved shape retention, and when combined with a complete post thermal curing, a unanimous cross-linking of the entire composite from the interlayer to the bulk matrix is enabled . The steps of method 1000 may be performed by the apparatus 100 which includes a robot via an end effector mounted on an automated robot arm, thereby yielding a composite 3D printer that enables printing of continuously reinforced thermally curable thermoset composites. The printed composite achieved fiber volume fraction at ~44.2 vol%, tensile strength at 1075.27 MPa and modulus at 45.24 GPa, flexural strength at 1209.50 MPa and modulus at 68.01 GPa, and interlaminar strength at 37.17 MPa.

[0126] Example 2: CORE-SHELL STRUCTURED TOW-PREGS ENABLED ADDITIVE MANUFACTURING OF CONTINUOUSLY REINFORCED THERMOSET COMPOSITES ("UV- ASSISTED 3D PRINTING OF CONTINUOUSLY REINFORCED THERMALLY CURABLE THERMOSET COMPOSITES WITH DUAL-CURED INTERPENETRATING POLYMER NETWORKS (IPN)")

[0127] Materials and Methods

[0128] The carbon fibers are 12 K T300 continuous fibers. The epoxy is the diglycidyl ether of bisphenol A (DGEBA)-based liquid epoxy resin, as manufactured by from Fibre Glast Developments Corp, of Brookville, Ohio. The epoxy is combined with a compatible hardener, methyl tetrahydrophthalic anhydride (MTHPA)-based hardener, in a weight ratio of 5 parts epoxy to 1 part hardener. The UV resin is the acrylate-based photocurable resin.

[0129] As for composite preparation, the dual cure resin-coated composite samples were 3D-printed using the robotic-based end effector installed on UFactory xArm 6. As a comparison, control samples were prepared using the same printing setup, but without dual cure resin coating . To solidify the dual cure resin coating, UV LEDs (CLIO, Dongguan UVET Co., Ltd .) were employed with specific parameters. The UV LEDs emitted light at a wavelength of 405 nm, with a spot size (or UV laser beam diameter) of 10 mm, power intensity of 2.9 W / cm2, and focus length of 22 mm. This application of UV light took place at the compaction stage. Both the printed samples with dual cure resin coating and the control samples were subjected to compaction. A compaction roller was used in this process to ensure uniformity. Subsequently, all samples, both coated and uncoated, were fully cured in an oven at a temperature of 191 °C for a duration of 1 h.

[0130] Regarding material characterizations, thermogravimetric analysis (TGA) was conducted on the fully cured epoxy, UV curable resin, and their mixtures with varying ratios using the Discovery TGA 5500 instrument from TA Instruments. The measurement involved heating 20 mg samples in alumina pans at a rate of 10 °C / min, starting from room temperature and reaching 650 °C, under ambient air conditions. Differential scanning calorimetry (DSC) was performed on the fully cured resins and the resins containing solely UV-responsive components. The Discovery DSC from TA Instruments was utilized for this test. Each sample, weighing 10 mg, was encapsulated in Tzero aluminum pans and sealed using a Tzero press. The measurement involved a heat ramp rate of 10 °C / min, starting from room temperature and reaching 300 °C, in nitrogen. For dynamic mechanical analysis (DMA), samples were examined in tension mode using the TA Q800 instrument from TA Instruments. The measurement was conducted with a frequency of 1 Hz and a strain of 0.02%. Fourier-Transform Infrared spectroscopy (FTIR) was employed to investigate the resins containing UV-responsive components. The Thermo Nicolet 6700 FT-IR spectrometer from Thermo Fisher Scientific was used for this test.

[0131] As for morphology characterizations, morphology of a single tow-preg was characterized using a confocal microscope (Zeiss LSM 510). The overall structure was observed at x 60 magnification. To distinguish between the core and shell layers of the core-shell structured tow-preg, fluorescent-sensitive Congo red dye powders (C6767, Sigma-Aldrich) were added into the dual-cure resin during sample preparation (Concentration: 1 g Congo red / 10 g dual-cure resin). The morphology of multi-layer composite was characterized using a 3D laser scanning confocal microscope (Keyence VK-X3000). The overall structure was observed at x 10 magnification, while finer details were examined at x 20 magnification. To visualize the internal structure of the composite, high-resolution computed tomography (CT) scanning was performed using a Nano-CT system with a Cu X-ray source (Rigaku nano3DX). For high-resolution images, the lens and binning were set as L0270 and 1, respectively. The obtained CT images were processed using Dragonfly software. A single slice in the X-Y plane was selected for deep learning training and color-coded, with carbon fibers (CF) shown in red and the matrix material in blue. The training process was conducted 100 times to ensure robustness and accuracy. Subsequently, the trained model was applied to analyze a specific volume of reconstructed images, measuring 455 x 455 x 325 pm.

[0132] Regarding mechanical characterizations, tensile, flexural, and interlaminar shear tests were conducted in Instron 5567 with a 30-kN load cell. For the tensile tests, ASTM D3039 testing standard was adhered to, with a test rate of 0.05 in / min. Flexural tests was performed in accordance with ASTM D790, employing a test rate of 0.50 in / min and a span-to-thickness ratio of 25: 1. Interlaminar shear tests were conducted following ASTM D2344 with a test rate of 0.05 in / min and a span-to-thickness ratio of 6:1.

[0133] Results

[0134] Dual-cure resin chemistry with varying weight ratios of UV-curable resin and epoxy resin were assessed to select the most suitable resin for composite additive manufacturing. To assess the thermal stability of the dual-cure resin system, TGA was performed. Prior to TGA measurements, all resins were fully cured based on the vendor's recommended thermal curing conditions for epoxy: a heating rate of 3 °C / min, and a curing temperature of 191 °C for 5 h. The TGA results indicated that the epoxy resin exhibited a weight loss of 0.04%, while the UV resin exhibited a weight loss of 0.23% before reaching 191 °C.

[0135] Among the various tested dual-cure resins, the dual-cure resin with an epoxy / UV resin ratio of 1 / 2 (by weight) presented the smallest weight loss of only 0.02% before 191 °C. In comparison, the dual-cure resins with ratios of 1 / 1 and 1 / 0.5 exhibited weight losses larger than epoxy but smaller than UV resin. The superior thermal stability observed in the epoxy / UV resin ratio of 1 / 2 can be attributed to the synergistic effects of the two resins, with the UV resin contributing to the formation of a more crosslinked network.

[0136] Furthermore, Differential Scanning Calorimetry (DSC) was utilized to analyze the exothermic reaction peak temperature and the reaction enthalpy of the intermediate dual-cure resin material after UV exposure. The DSC profile provided insights into the thermal transition of the dual-cure cladding on printed composites. Table 4 below presents the reaction peak temperature Tpeak, and the enthalpy values obtained from DSC. A higher Tpeak suggests a higher activation energy required for curing the resin system, which may contribute to a denser network of crosslinked polymer chains. The enthalpy is the integration area of exothermic peak in DSC profile, which can be determined by using Equation 5. where AH is the enthalpy between To and Ti, CPis the heat capacity, which is defined by Equation 6 below. Equation

[0137] A lower enthalpy indicates a lower concentration of reactive components, primarily the epoxy resin. Overall, the DSC results showed that the epoxy / UV ratio of 1 / 2 exhibited the lowest enthalpy and the highest glass transition temperature among the tested ratios. This indicates that the epoxy / UV ratio of 1 / 2 is the most optimized combination, displaying enhanced thermal stability and crosslinking behavior, making it the preferred choice for use in composite additive manufacturing applications.

[0138] Table 4: Reaction peak temperature and enthalpy of intermediate dual-cure resin with uncured epoxy and cured UV resin

[0139] Dynamic Mechanical Analysis (DMA) was employed to study the mechanical properties of the dual-cure resins by evaluating the storage modulus and the loss factor tan (delta) across a temperature range from room temperature to 200 °C. The storage modulus in DMA represents a material's ability to store energy when deformed elastically, and a higher storage modulus value indicates greater stiffness and resistance to deformation, correlating to a higher elastic modulus. The dual-cure resin with an epoxy / UV resin ratio of1 / 2, which comprises an interpenetrating polymer network (IPN), exhibited a higher storage modulus value in the lower temperature region compared to the other ratios. This observation indicates that the material with this specific ratio and IPN structure is stronger and more rigid, attributing its enhanced mechanical performance to the balanced combination of epoxy and UV resin networks. Furthermore, during DMA analysis, relaxation of polymer chains was observed through a sharp reduction in storage modulus at the onset temperature of glass transition, where the resin enters the rubbery state. To estimate the crosslinking density (ve) of the resin, the storage modulus of a cross-linked resin in the rubbery plateau region can be utilized. This is because the resistance level to deformation of a rubbery material reflects the restriction level on the mobility of polymer chains, aligning with the principles of rubber elasticity theory. The crosslinking density of a dual-cure resin refers to the degree of IPN network formation between polymer chains, which significantly impacts the mechanical, thermal, and chemical properties of the cured resin. Generally, a higher crosslinking density results in a stiffer and stronger cured resin material with improved performance characteristics. The IPN structure in the dual-cure resin contributes to the enhanced viscoelastic and crosslinking behaviors, leading to improved mechanical properties compared to the pristine resins. The crosslinking density (ve) of the cured resin can be calculated using Equation 7.

[0140] E = 3veRT Equation

[0141] 7 where E represents the storage modulus at the temperature of Tg+ 30 °C. Tgis the glass transition temperature measured through DMA by observing the tan (delta) peak. Tg+ 30 °C ensures that the material is fully in the rubbery state, and E values measured at temperatures beyond Tg+ 30 °C tend to stabilize, providing a reliable baseline for studying the mechanical behaviors of the material in the rubbery state. Also, veis the crosslinking density (mol / m3), R is the gas constant (8.314 J / (mol»K)), and T is the absolute temperature. This calculation allows for a quantitative assessment of the crosslinking density, providing valuable insights into the resin's properties and suitability for specific applications.

[0142] The calculated crosslinking densities are presented in Table 5 below. It is evident that epoxy / UV = 1 / 0.5 and epoxy / UV = 1 / 1 exhibit similar crosslinking densities, around 1640 mol / cm3. In contrast, the epoxy / UV resin with a ratio of 1 / 2 demonstrates a significantly higher crosslinking density, reaching 2665.43 mol / cm3, comparable to that of epoxy.

[0143] Table 5 : Crosslinking density of cured resin (except UV resin) The tan (delta) measurements obtained from DMA offer valuable insights into the curing kinetics of the material. Specifically, the tan (delta) peak corresponds to the glass transition temperature (Tg) of the material, providing critical information to determine the optimal curing conditions for the resin. For the UV resin, calculating the exact crosslinking density is challenging. Its storage modulus remains at high values over a higher temperature range, and the tan (delta) displays a linear-like response without a distinct Tgpeak. These observations indicate that the UV resin possesses a highly crosslinked and rigid polymer network. The rigid network limits the mobility of the polymer chains, restricting energy dissipation and resulting in a low damping response. The addition of the UV resin lowers the glass transition temperature of the epoxy resin. Notably, Tgof the epoxy / UV resin with a 1 / 2 ratio increased to 146 °C compared to the lower Tgof 130 °C for the epoxy / UV resin with a 1 / 1 ratio. This change in Tgis confirmed by the increased crosslinking density observed in the dualcure resin with a 1 / 2 ratio. The possibility of a smaller and compact UV resin crosslinking network constraining the mobility of the epoxy network provides a plausible explanation for the observed increase in Tg. To assess the mechanical properties of the dual-cure resin, we conducted tensile tests following the ASTM D3039 standard. All the resins were tested at a constant tensile rate of 0.05 in / min. Among the various dual-cure resins, the cured epoxy / UV resin with a ratio of 1 / 2 exhibited the highest tensile strength, measuring 52.28 MPa. This value is only 1.07% lower than that of the pristine epoxy resin, while surpassing the tensile strength of the UV resin by 1.26%. The tensile modulus of the cured epoxy / UV resin with a ratio of 1 / 2 was determined to be 2.70 GPa. While this value is 1.30% lower than the cured epoxy / UV resin with a ratio of 1 / 1, it still outperforms the cured epoxy / UV resin with a ratio of 1 / 0.5 by 1.48%.

[0144] Overall, the samples made from the dual-cure resin demonstrated comparable tensile properties to those made from pristine epoxy and UV resin. The mechanical performance of the dual-cure resin was promising, showcasing the benefits of the tailored combination of epoxy and UV resin components in achieving enhanced properties. To compare the minimum resin solidification time needed for the dual-cure resin, UV light with a 10 mm spot size (or UV laser beam diameter), 2.9 W / cm2 power density, and 405 nm wavelength to irradiate the resins (thickness: ~30 pm) was utilized. The resin surface hardening time was qualitatively measured to determine the minimum solidification time of the dual-cure resin. The pristine UV resin solidified in just 4 seconds upon UV exposure on the sample surface. On the other hand, the dualcure resin with an epoxy / UV resin ratio of 1 / 2 required approximately 15 seconds to solidify on the surface. The solidification time increased as the epoxy content in the dual-cure resin increased. Based on the analysis of thermal stability, crosslinking density, cure kinetics, mechanical properties, and minimum resin hardening time, we selected the dual-cure resin with an epoxy / UV resin ratio of 1 / 2 to be further studied and tested in our composite additive manufacturing process. In the subsequent paragraphs in this example, unless specifically mentioned, the term "dual-cure resin" exclusively refers to the resin with an epoxy / UV ratio of 1 / 2 by weight.

[0145] In-Situ Formation Process and Morphology Characterization of Core-Shell Structured Tow-Preg.

[0146] FIG. 10A depicts a schematic of an in-situ formation process of core-shell structured tow-preg during printing. The core is formed through epoxy impregnation and partial removal in the epoxy reservoir, while the shell is formed through dual-cure resin coating in the dual-cure resin reservoir. FIG. 10B depicts a schematic and digital photo of resin removal to form the core of tow-preg. FIG. 10C depicts a digital photo of a resin scrapper design for core formation. FIG. 10D depicts a fluorescent confocal image of the printed core-shell structured tow-preg, with the dual-cure shell layers exhibiting fluorescence due to the addition of fluorescent-sensitive Congo red dye.

[0147] FIG. 10A schematically depicts a process of in-situ formation of a core-shell structured tow-preg, which includes three steps: epoxy impregnation, partial removal of epoxy from an outer surface of the tow-preg, and a dual-cure resin coating on the tow-preg forming core-shell structure. A comb-like resin scraper is developed to remove the a desired portion of the resin from the outer surface of the tow-preg (FIG. 10B). The teeth length and the spacing between teeth are controlled to ensure the desired epoxy removal from the outer surface. The desired epoxy removal can be determined by one or more variables related to polymer curing kinetics and rigidity, for example. As shown in Fig. 10C, resin scraper teeth remove a portion of the epoxy from the outer fibers of the tow-preg (e.g. provides a desired degree of surface roughening by creating grooves in the epoxy), creating space to coat dual-cure resin to form a core-shell structure. The core-shell structure formation of the 3D-printed tow-preg is validated with confocal microscopy by adding the fluorescent-sensitive Congo red dye into the dual-cure resin. The fluorescent confocal microscope image confirms the creation of the core-shell structure (FIG. 10D), where the epoxy in the core and the red-colored dual-cure resin on the outer shell of tow-preg are present. To understand the impact of UV radiation power and time on the hardening effect of the dual-cure resin, Fourier Transform Infrared Spectroscopy (FTIR) was employed to measure the relative peak intensity changes of characteristic functional groups (C=O and C=C). These data were then utilized to calculate the curing degree of the dual-cure resin under different UV radiation power and time conditions. Initially, we prepared a thin polymer membrane (length: 100 mm, width: 10 mm, thickness: ~30 pm) using a micrometer adjustable film applicator. Subsequently, we utilized three UV LED lights (power intensity: 2.9 W / cm2, spot size or UV laser beam diameter: 10 mm), preinstalled on the end effector, to scan over the surface of the dual-cure polymer membrane at a focal length of 22 mm. The UV power (1, 2, 3 LED light; 2.9 W / cm2 each) and scanning speed (1, 2, 5, 10 mm / s) were controlled during the process.

[0148] In the FTIR spectra of the irradiated resin membrane, the peak was observed at 1720 cm1, arising from the C=O of the acrylate monomer, remained unchanged during the polymerization process. UV curing involves a double bond cleavage process that opens the C=C bond, leading to the formation of a cross-linked C-C structure. The reduction of the C=C bond signifies the cross-linking and curing of the acrylate UV resins. Additionally, the peak at 810 cm1decreased with an increased scanning speed. This particular peak represents the acrylate C=C bond, which undergoes breaking and forms polymeric chains during photopolymerization. The extent of peak reduction corresponds to the degree of conversion from double bonds degree of curing. To calculate the degree of curing achieved by UV scanning at a specific speed, we used the peak areas at 810 cm1and 1720 cm1, applying Equation 8. This analysis provides valuable insights into the curing efficiency and enables optimization of the UV radiation parameters for the dual-cure resin. Equation s

[0149] The results obtained from Equation 8 indicated that the membrane irradiated by more UV LEDs and at a lower scanning speed exhibits a higher degree of curing. For instance, the membrane irradiated at 1 mm / s with 3 UV LEDs achieves a degree of curing of 70.4%. In comparison, the degree of curing for the fully cured control sample is 93.1%. The impact of different UV radiation parameters on the curing efficiency of the dual-cure resin includes that as the number of UV LEDs increases and the scanning speed decreases, the degree of curing improves, approaching closer to the fully cured control sample. A prototype of an apparatus according to the present invention includes a 3D printer having 3 UV LEDs for selective curing of the resin coating on tow-pregs to achieve optimal curing and solidification quality (FIG. 11A). The schematic in FIG. 11B shows that the first UV LED is utilized to solidify the resin coating on one side of the tow-pregs. The second UV LED, which is irradiated at the nip point between the compaction roller and the tow-pregs, is employed to establish interlayer bonding of the deposited fiber laminates. The third UV LED is to improve the degree of curing of tow- pregs on the other side. The setup parameters for additive manufacturing (AM) are listed in Table 6.

[0150] Table 6: AM Setup Parameters

[0151] For characterization purposes, FTIR was employed to measure the intensity change of characteristic functional groups in the resins, allowing the calculation of the degree of curing on the tow-pregs. In comparison to the epoxy-impregnated tow-pregs, the dual-cure resin coating introduced the C=O functional group to the tow-pregs. The C=O peak at 1720 cm1remained relatively stable after UV radiation on the dual-cure resin-coated tow-pregs, while the C=C peak at 810 cm1decreased as the UV resin monomer component polymerized. The calculated results, using Equation 2 above, demonstrated a degree of curing of 33.8% at the 1st UV spot, 36.5% at the 2nd UV spot, and 48.3% at the 3rd UV spot. These values are slightly lower than those obtained for the dual-cure resin membranes, which were scanned by 1 UV LED and 2 UV LEDs, respectively, at a speed of 1 mm / s. The discrepancy may be attributed to some dual-cure resin being trapped in the porous structure of carbon fibers, resulting in insufficient UV exposure. To achieve full solidification, post UV curing and subsequent thermal curing (191 °C for 1 h) were applied to the samples. The fiber volume fraction of the fully cured AM composite was measured through TGA using the burn-off method in nitrogen, following ASTM D3171, with a heat rate of 20 °C / min. The calculated weight fractions were: 39.6 wt% for carbon fibers, 43.8 wt% for epoxy, and 16.5 wt% for UV resin. Using Equation 9 below, the fiber volume fraction was determined to be 29.7 vol%, epoxy volume fraction at 48.9 vol%, UV resin volume fraction at 19.4 vol%, and the void volume at 2.0 vol%.

[0152] Vr PspeclmenA rf — - X Wf Equation 9

[0153] P fiber where the density of the specimen was measured as 1.35 g / cm3, the fiber density was 1.80 g / cm3, the epoxy density was 1.21 g / cm3, and the UV resin density was 1.15 g / cm3.

[0154] FIG. 11C presents a comparison between a self-lifted tow-preg (thickness: ~0.12 mm) coated with dual-cure resin (thickness: 30-60 pm) and UV-cured at 1 mm / s, and a fully cured structure achieved after thermal curing at 191 °C for 1 h. Each photo shows the corresponding structure in a horizontal position, gripped at one end by a pair of needle-nose pliers (of which only the tip is shown), with the opposite end unsupported. This comparison illustrates that the dual-cure resin coating plays a vital role in providing a rigid self-supporting structure during the shape formation of the tow-pregs.

[0155] Large-Area 3D Printing Demonstration and Confocal Microscopy Images of Cross-Sectional of Composites with Interlayers Exhibited

[0156] FIG. 12A depicts a photo of large-area 3D printing demonstration of the unidirectional carbon fiber composites panel. The length and the width of the panel are 100 mm x 60 mm, deposited 5 layers with 10 tow-pregs on each layer. The hatch spacing was 6 mm, and the print speed was 1 mm / s. FIG. 12B depicts a photo of the composites panel held on hand, after it was thermally cured at 191 °C for 1 h. FIG. 12C depicts a confocal microscopy images of the cross-section of the composite with 5 layers of tow-pregs. FIG. 12D depicts a 3D depth mapping shows the voids and cracks created during the printing and polishing process. A magnified microscopic image of the cross-section is shown on the bottom.

[0157] FIGS. 12A-12D illustrates the automated lamination process of the composite and presents cross-section characterizations. For the lamination, a composite panel was created with a hatch spacing of 6 mm, print speed of 1 mm / s, and using 3 UV LEDs, each with an intensity of 2.9 W / cm2 and a wavelength of 405 nm. After printing, the composite underwent thermal curing at 191 °C for 1 h. To examine the crosssection morphology, confocal microscopy was employed. The resulting observations revealed that the thickness of a carbon fiber (CF) layer is approximately 90 pm. Between the CF layers, a matrix-rich region spanning from 30 to 60 pm was observed. It is important to note that some macro-voids were noticed, which may have been caused by the polishing process. For a more comprehensive analysis, a 3D image of the composite surface was generated by stacking surface slices through its depth, allowing for the identification of defects such as deep voids and shallow cracks, as presented in FIG. 12D.

[0158] Nano-CT Image Segmentation Using Deep Learning Training

[0159] To visualize the internal structure of the carbon fiber composite in three dimensions, we reconstructed a randomly selected composite sample's internal structure using X-ray nano-CT images, as depicted in FIGS. 13A-13C. FIG. 13A depicts raw CT images of one slice of the X-Y plane and a 3D view. FIG. 13B depicts CT images after 1st training and 100th training. FIG. 13C depicts the X-Y, X-Z planes and the 3D view of the CT images after 100th training. The image is segmented with a distinction between fiber (in one color or shading) and matrix (in another color or shading).

[0160] The raw CT images reveal visible fibers and matrix, distinguished by different color intensities (FIG. 13A). However, using threshold and manual segmentation to differentiate between the two phases has limitations. Some parts of the matrix display similar color intensities as carbon fibers, making it challenging to precisely distinguish them. Additionally, manually coloring thousands of image slices can be time consuming. To overcome these challenges, we employed deep learning, a machine learning technique that utilizes neural networks to efficiently segment images. Neural networks are model-fitting processes that can identify components based not only on color intensities but also on various information, such as shape and edge details. Initially, the image resembles a segmented image using a threshold, where parts of the matrix with similar intensities are displayed in red. However, after hundreds of training iterations, only fibers are displayed in red (FIG. 13B). The trained information was then applied to the raw CT images, resulting in a segmented 3D image that clearly identifies fibers and matrix. While there are gaps between the prepreg layers in the X-Y plane, the fibers show high alignment in the X-Z plane, producing a morphology that closely resembles images taken with confocal microscopy (FIG. 13C). This deep learning approach provides a more efficient and accurate means of analyzing the CT images and gaining valuable insights into the composite's internal structure.

[0161] Mechanical testing coupons were prepared and characterized for uniaxial tensile, flexural properties, and interlaminar shear strength following ASTM standards. To compare the results, we prepared 3D printed coupons with dual-cure resin coating (60 mm x 8 mm x 0.60 mm), using the same dimensions as the composite control samples and 3D printed coupons with UV resin coating from our previous work. All three types of samples were made using the same epoxy / carbon fiber tow-pregs and prepared with the same impregnation process. The samples were deposited using the robotic end effector and post-cured in an oven with the same heating procedure. The 3D-printed dual-cure resin-coated samples exhibited a longitudinal elastic modulus of 74.2 ± 10.8 GPa and a longitudinal tensile strength of 1058.2 ± 77.4 MPa. In comparison, the 3D printed UV resin-coated samples showed a longitudinal elastic modulus of 45.2 ± 5.0 GPa and a longitudinal tensile strength of 1075.3 ± 81.7 MPa. The control samples exhibited a longitudinal elastic modulus of 57.5 ± 14.9 GPa and a longitudinal tensile strength of 964.5 ± 106.0 MPa. Uniaxial tensile testing was performed on four specimens following ASTM D3039 with a test rate of 0.05 in / min. For flexural properties, the 3D printed dual-cure resin-coated samples exhibited a flexural modulus of 41.7 ± 7.8 GPa and a flexural strength of 1080.4 ± 223.3 MPa. The 3D printed UV resin-coated samples showed a flexural modulus of 68.0 ± 12.8 GPa and a flexural strength of 1209.5 ± 128.1 MPa. The control samples had a flexural modulus of

[0162] 55.8 ± 9.6 GPa and a flexural strength of 822.8 ± 97.1 MPa. Flexural testing was performed on four specimens following ASTM D790 with a test rate of 0.50 in / min and a span-to-thickness ratio of 20:1. The interlaminar shear strength (ILSS) was tested as

[0163] 48.8 ± 3.4 MPa for the 3D-printed dual-cure resin-coated samples. In comparison, the ILSS was tested as 37.2 ± 4.5 MPa for 3D-printed UV resin-coated samples and 50.5 ± 1.5 MPa for control samples. Interlaminar shear strength testing was performed on four specimens following ASTM D2344 with a test rate of 0.05 in / min and a span-to- thickness ratio of 10: 1 (Table 7).

[0164] Table 7 : Mechanical Properties of Composite Samples

[0165]

[0166] *Example 1 above

[0167] It is noteworthy that the tensile properties of 3D printed composites with dualcure resin coating are improved, due to the formation of an Interpenetrating Polymer Network (IPN) network. However, the flexural properties and interlaminar shear strength are reduced, likely due to the presence of resin-rich regions observed in the confocal microscopy images of the printed composites. While the mechanical performances of the prototype 3D-printed dual-cure resin-coated composite samples may appear relatively lower compared to conventional epoxy composites produced through thermal cure methods, the prototype samples exhibit competitive performances in the AM composite systems in terms of mechanical testing performances (e.g. the tensile and flexural properties of 3D printed dual-cure resin- coated carbon fiber composite coupons with results from prior art literature and data of conventional composites). The dual-cure cladding strategy as described herein provides an advantage by allowing the fabrication of uncured epoxy composite structures without the need for a mold. The process and structure are demonstrated in FIGS. 14A-14D. The schematic illustrates the process of spanning tow-pregs while curing the dual-cure resin shell around them (FIG. 14A). Notably, a clear span without the need for support columns was successfully achieved. During the part fabrication, in situ UV curing enables the soft, flexible tow-pregs to self-lift, supported by the stiffness of the solidified tow- pregs. A digital photo of the printing process of the overhang structure is shown in FIG. 14B. The printing parameters used were a print speed of 1 mm / s, tow-preg width of 6 mm, and the use of 3 UV LEDs with a wavelength of 405 nm and an intensity of 2.9 W / cm2 for each LED. A thermally cured composite structure (191 °C for 1 h) is presented in FIGS. 14C and 14D. The span structure is well self-supported, and no sagging was observed, demonstrating the effectiveness and stability of the dual cure cladding approach in creating complex overhang structures without the need for additional support.

[0168] Thus, the disclosed tow-preg cladding technology demonstrates that the preformation of epoxy / continuous fiber composite structures, and the incorporation of a rigid dual-cure cladding enables mold-free preformation of thermoset composites during additive manufacturing. The tow-preg cladding process facilitates the formation of a sequential interpenetrating polymer network (IPN) at the interlayer between laminates, enhancing the overall structural integrity. The printed composites exhibited excellent mechanical properties, including a longitudinal tensile strength of 1058.2 MPa, a longitudinal tensile modulus of 74.2 GPa, a flexural strength of 1080.4 MPa, and a flexural modulus of 41.7 GPa. This rigid dual-cure tow-preg cladding strategy has effectively addressed the tackiness issue associated with tow-pregs and has provided a viable means for the preformation of thermoset composites during additive manufacturing without the need for molds. Overall, this innovative approach facilitates the fabrication of complex composite structures with enhanced mechanical performance, making it a promising technique for advancing the field of additive manufacturing for thermoset composites.

[0169] According to another aspect of the invention, a composite material is provided. The composite material includes a plurality of cured layers and an interlayer between each adjacent pair of cured layers. In an embodiment, each layer has a fiber (such as fiber 112 or 212 above) in a matrix (such as matrix 110 or 210 above) of a thermoset polymer (such as polymer 114 or 214). Each interlayer, such as interlayer 180 or 280 above, has an interpenetrating network (IPN) of a dual-cured polymer (such as dual- cured polymer 130, 230). Additionally or optionally, the dual-cured polymer 230 has a thermoset component (such as polymer 134) and a photocured component (such as polymer 136, 236). In one non-limiting example, the thermoset component 134 and the thermoset polymer 114 comprise a same polymer. The dual-cured polymer 130 includes an in-situ diffusion of the photocured component 136 into the thermoset polymer 114 of the matrix 110, as discussed above with reference to at least FIGS. 1-7 and method 1000. Additionally or optionally, each interlayer 180, 280 comprises two outer sub-interlayers each comprising the interpenetrating network of dual-cured polymer, and a middle sub-interlayer comprising the photocured component without the thermoset component between the two outer sub-interlayers (as shown in FIG. 2). Although the invention is illustrated and described herein with reference to specific embodiments, the invention is not intended to be limited to the details shown. Rather, various modifications may be made in the details within the scope and range of equivalents of the claims and without departing from the invention.

Claims

What is claimed :

1. An apparatus for manufacturing a composite material, the apparatus comprising : a tow-preg module, comprising a source of fiber material; means for causing the fiber material to move along a material path having a direction; means for impregnating a length of the fiber material with a thermoset polymer to form a matrix; a layering module, the layering module comprising : a coater for coating the matrix with a photocurable polymer to form a photocurable-polymer-coated matrix; an interlayer heater positioned to heat the interlayer to a temperature sufficient to cause the thermoset polymer and the photocurable polymer to diffuse together; means for stacking one layer of photocurable-polymer-coated matrix on an other layer of photocurable-polymer-coated matrix, forming an interlayer therebetween; a roller positioned to compress the one layer and the other layer; and a source of actinic radiation positioned to expose the photocurable polymer to the actinic radiation in a location along the material path between the interlayer heater and the roller.

2. The apparatus of claim 2, wherein the apparatus comprises a robot having an end effector for stacking the one layer of photocurable-polymer-coated matrix on the other layer of photocurable-polymer-coated matrix in a predetermined position.

3. The apparatus of claim 2, further comprising a source of thermal energy sufficient for thermally curing the thermoset polymer to form the composite.

4. The apparatus of claim 2, wherein the means for impregnating the length of the fiber tow with the thermoset polymer comprises a Localized-In-Plane Thermal Assist (LITA) impregnation system comprising a LITA heater configured to heat the fiber material to generate a thermal gradient in the fiber material trailing the LITA heaterrelative to the material path direction, and a feeder configured to dispense the thermoset polymer on the heated fiber material.

5. A composite material comprising a plurality of cured layers and an interlayer between each adjacent pair of cured layers, each layer comprising a fiber in a matrix of a thermoset polymer, each interlayer comprising interpenetrating network of a dualcured polymer, the dual-cured polymer comprising a thermoset component and a photocured component.

6. The composite material of claim 5, wherein the thermoset component and the thermoset polymer comprise a same polymer.

7. The composite material of claim 6, wherein the dual-cured polymer comprises an in-situ diffusion of the photocured component into the thermoset polymer of the matrix.

8. The composite material of claim 6, wherein each interlayer comprises two outer sub-interlayers each comprising the interpenetrating network of dual-cured polymer, and a middle sub-interlayer comprising the photocured component without the thermoset component between the two outer sub-interlayers.

9. A method for manufacturing a composite material, the method comprising:(a) providing a matrix comprising a fiber material and a thermoset polymer;(b) providing a dual curable polymer between a first layer of the matrix and the second layer of the matrix, the dual curable polymer comprising a thermally curable component and a photocurable component curable by exposure to actinic radiation;(c) exposing the dual curable polymer to actinic radiation to at least partially cure the photocurable component; wherein performing steps (a)-(c) creates a layer of the composite material, the method optionally comprising repeating steps (a)-(c) until a desired number of layers have been stacked together to create a structure, and then(d) heating the layer or the structure to thermally cure the thermoset polymer and the thermally curable component of the dual curable polymer.

10. The method of claim 9, wherein the thermoset polymer and the thermally curable component of the dual curable polymer comprise the same polymer.

11. The method of claim 10, comprising forming the dual curable polymer in situ by coating a surface of the thermoset polymer of the matrix with the photocurablepolymer and causing the thermally curable polymer and the photocurable polymer to diffuse together at an interface thereof.

12. The method of claim 11, comprising heating the thermally curable polymer and the photocurable polymer to cause the diffusion of the photocurable polymer and the thermally curable polymer together to form an interpenetrating network (IPN).

13. A method for manufacturing a composite material, the method comprising:(a) providing a matrix comprising a fiber material and a first polymer;(b) contacting the matrix with a second polymer;(c) causing the first polymer and the second polymer to diffuse together at an interface between the first polymer and the second polymer;(d) curing the second polymer; and(e) curing the first polymer.

14. The method of claim 13, wherein the second polymer has a faster curing rate than a curing rate of the first polymer, uses a different curing initiator than the first polymer, or a combination thereof.

15. The method of claim 13, wherein the second polymer is less viscous than the first polymer.

16. The method of claim 13, wherein the first polymer is thermally curable and the second polymer is photocurable.

17. The method of claim 16, wherein the second polymer is curable with UV radiation.

18. The method of any of the foregoing claims, wherein the fiber material comprises a carbon fiber.

19. The method of claim 13, wherein the matrix comprises a tow-preg or a prepreg.

20. The method of claim 13, wherein the first polymer comprises a thermoset epoxy.

21. The method of claim 13, wherein step (c) is thermally assisted and performed at a temperature above room temperature.

22. The method of claim 21, wherein the temperature is 50-60 C.

23. The method of claim 13, wherein performing steps (a)-(d) forms a layer, the method further comprising repeating steps (a)-(d) one or more times to form a multilayered structure prior to performing step (e) on the multilayered structure.

24. The method of claim 13, wherein steps (a) and (b) comprise the sequential steps of (a) impregnating the fiber with the first polymer to form the matrix, then (b) coating the second polymer over the matrix.

25. The method of claim 24, wherein step (a) comprises performing a localized inplane thermal assist (LITA) technique.

26. The method of claim 24, further comprising providing at least two layers of second-polymer coated matrix and compacting the at least two layers together.

27. The method of claim 26, comprising forming an interface between the second polymer of the first layer and the second polymer of the second layer, and initiating curing of the second polymer at the interface prior to compacting the first layer and the second layer together.

28. The method of claim 27, wherein the second polymer is photocurable, and initiating the curing comprises directing actinic radiation toward the interface.

29. The method of claim 28, wherein the second polymer is UV curable, and directing the actinic radiation comprises shining a source of UV light at the interface.

30. The method of claim 29, wherein the source of UV light comprises a plurality of UV LED sources.

31. The method of claim 30, wherein the plurality of UV LED sources comprises a first LED source for selective curing of the second polymer on a first surface of the first layer, a second LED source for selective curing of the second polymer at the interface, and a third LED source for selective curing of the second polymer on an opposite surface of the first surface of the first layer.

32. The method of claim 23, comprising performing steps (a)-(d) with a robot.

33. The method of claim 19, further comprising removing a removed portion of the first polymer from an outer surface of the matrix.

34. The method of claim 33, wherein removing the removed portion of the first polymer from the matrix is performed prior to contacting the matrix with a second polymer.

35. The method of claim 33, wherein a scraper is configured to remove the removed portion of the first polymer from the outer surface of the matrix, the scraper having teeth extending for a predetermined length and spaced apart from each other for a predetermined distance, thereby creating a plurality of gaps in a remaining portion of the first polymer from which the removed portion of the first polymer was removed to receive the second polymer.

Images