DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Objections
Claims 3-4, 8, 10-11, 14-15, 19 and 21-22 are objected to because of the following informalities:
Claims 3-4 and 14-15 have a line about, “cyclindrical coordinates”, which is misspelled and should be, “cylindrical coordinates”,
Claims 4 and 15 have a line about, “partial directive”, which should be, “partial derivative”,
Claims 8 and 19 have a line, “deformation and roughness patterns are leaned from”, which should be, “deformation and roughness patterns are learned from”,
Claims 11 and 22 have a line, “additive manufacturing machine comprising an arch torch…the arch torch”, which should be, “additive manufacturing machine comprising an arc torch…the arc torch”,
Claim 15 depends upon claim 1, but it appears that the same dependent claim is already dependent on claim 1 in claim 4. The Examiner has assumed that claim 15 should be dependent on claim 12, if this is true then correction should be done on the dependency,
Claims 10 and 21, line 1, “whrein the machine learning”, should be, “wherein the machine learning”.
Due to the amount of misspellings, Examiner requests applicant to review the language of their claims as there might be other overlooked misspellings.
Appropriate correction is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 5-13, and 16-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yang et al. (US 20170368753 A1, hereinafter Yang) in view of Li et al. (CN 113435083 A, hereinafter Li) and Buller et al. (US 20170239892 A1, hereinafter Buller).
Regarding claim 1, Yang discloses a method for forming a target object by additive manufacturing, the method comprising:
a) receiving, by a computing device, an initial three dimensional model of the target object to be fabricated by additive manufacturing (Abstract, “system and methods are provided comprising generating a nominal computer-aided design (CAD) image of a component”);
b) applying a compensation plan to a initial three-dimensional model to form a modified three-dimensional model to compensate for deformation during fabrication of the target object (Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold; modifying the nominal CAD image by the compensation field; and producing a physical representation of the component from the modified nominal CAD image.”),
the compensation plan being determined by the computing device with a computer implement method comprising (Para. 0003, “method that compensates for the distortions/defects to reduce iterations of non-relevant AM produced components.”):
converting point-cloud data to functional representation (Para. 0023, “physical component 102 is measured via a measurement device 108 to capture the geometry of at least one of the external and internal surface of the physical component 102 (FIG. 4B)…In one or more embodiments, the measurement data may be point-cloud coordinate data. In one or more embodiments, the point-cloud coordinate measurements may be transformed (e.g., via postprocessing at the AM device) into triangular-mesh data”);
calculate total shape deviations (Para. 0024, “In one or more embodiments, the deviation 400 may be determined between each geometry in the nominal CAD image 101 and each corresponding measurement 109 of the physical component 102.”);
constructing an engineering-informed tensor-product basis representation of deformation patterns for each object (Para. 0033, “Then the tensor fitting is used to generate a three dimensional (3D) compensation tensor field (e.g., compensation field 508) in S316. The compensation field 508 may then be applied to the nominal CAD image 101 to modify the nominal CAD image as described above with respect to S224 in FIG. 2.”); and
creating an optimal compensation plan from a predicted deformation patterns to be applied to the target object that minimizes its shape deformation (Para. 0037, “For example, the processor 610 may receive nominal CAD image data and then may apply the geometrical compensation module 106 via the instructions of the programs 612, 614 to generate a modified CAD image to generate a component 102.”); and
c) forming target object by additive manufacturing with the modified three- dimensional model (Para. 0030, “Then at S226, a modified physical component 404 (FIG. 4E) is produced from the modified nominal CAD image 402. In other words, the modified shape ( e.g., modified triangular mesh) may replace the nominal triangular mesh as the start point of the next production cycle of the physical component, in one or more embodiments.”).
Yang does not disclose:
where the additive manufacturing method is by wire and arc and fabrication is done by wire and arc additive manufacturing;
where data is used for each training object in a set of training objects;
where deviations are calculated for each training object;
where a tensor basis representation is included for each training object, where roughness can be taken into account, and
learning deformation patterns and roughness patterns that constitute the total shape deviations from the set of training objects
However, Li discloses, in the similar field of additive manufacturing (Abstract, “a prediction method and system for additive manufacturing”), where wire and arc additive manufacturing is used (Page 2, Para. 2, “wherein the laser-arc composite additive manufacturing has high forming efficiency”, and Page 5, last Para., “distance between the laser spot and the welding wire is not more than 2mm.”), where a machine learning model can take in training data that includes deformations and use those deviations caused by the deformations to make correlations on the proper corrections to be made, where these corrections can then be predicted in the future through the machine learning model (Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the point-cloud, tensor representation of deformation, and compensation plan to correct for deformations in Yang to include using those deformation data points as training data for a machine learning model in order to predict the deformation as taught by Li, where the prediction from Li would allow for the correct compensation plan to be selected.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use machine learning in order to predict a deformations with less intensive computing power needed, where predictions would allow a user to have more confidence in the compensation model as there would be training data used to in order to reach the prediction conclusion, as stated by Li, Abstract, “The invention accurately predicts the residual stress and deformation of the additive manufacturing sample under different paths with low calculation cost.”, and Page 2, Para. 8 from end, “thereby solving the problem that the current additive manufacturing residual stress and deformation of the prediction technology is difficult to ensure the prediction precision and reduce the calculation cost of the technical problem.”.
Further, Buller discloses, in the similar field of additive manufacturing (Para. 0002, “Three-dimensional (3D) printing ( e.g., additive manufacturing)”), where roughness patterns can be taken into consideration for the total shape deviation of the object (Para. 0326, “The 3D object can have various surface roughness profiles, which may be suitable for various applications. In some examples, the surface roughness is the deviations in the direction of the normal vector of a real surface ( e.g., average or mean planarity of an exposed surface of the 3D object), from its ideal form.”, and Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the shape deviations in modified Yang to take into consideration surface roughness patterns as taught by Buller.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use surface roughness as another parameter for determining the amount of correction needed for the total shape deviations of an object, as stated by Buller, Para. 0232, “the meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”.
Regarding claim 2, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein the target object has a net shape (Yang, Para. 0030, “In other words, the modified shape ( e.g., modified triangular mesh) may replace the nominal triangular mesh as the start point of the next production cycle of the physical component, in one or more embodiments.”, where the final triangular mesh of the modified shape is a net shape of the entire object).
Regarding claim 5, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein the optimal compensation plan minimizes the absolute volume deviations (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, and Para. 0029, “producing a physical representation of the component from the modified nominal CAD image to provide a path for the laser for each layer in the AM process”, where each layer has the deviations determined, where these deviations in total would be a volume deviation that is minimized during the compensation step).
Regarding claim 6, modified Yang teaches the method according to claim 5, as set forth above, discloses wherein the optimal compensation plan minimizes total absolute area deviation (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, where each layer has the total absolute area deviation minimized to a nominal area).
Regarding claim 7, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein the optimal compensation plan minimizes deviations for predicted deformation by calculating an amount of compensation that is equivalent to an area deviation (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, where the amount of compensation needed is the area deviation, Para. 0026, “If the deviation 400 is not within the pre-defined tolerance threshold, the process 200 continues to S222, and a compensation field 508 (FIG. 5) is determined, as described in more detail with respect to FIG. 3. As shown in FIG. 5, for example, the compensation field 508 may be the area between the nominal CAD model 502 and a perimeter 503 of the compensated geometry, represented by the arrows 505.”).
Regarding claim 8, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein deformation and roughness patterns are leaned from the set of training objects by training a machine learning algorithm (Yang, Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold”, where deformation patterns are taken into consideration; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness patterns are taken into consideration; teaching from Li, Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”, where the deformation and roughness data sets can be inputted as training data for a machine learning model, where the predictions on the optical compensation models would be based on the training data used to train the machine learning model).
Regarding claim 9, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein deformation and roughness patterns are leaned from the set of training objects by training a machine learning algorithm such that once trained provides deviations and/or the compensation plan for the target object (Yang, Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold”, where deformation patterns are taken into consideration and a final compensation plan is developed based on the deformations; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness patterns are taken into consideration; teaching from Li, Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”, where the deformation and roughness data sets can be inputted as training data for a machine learning model, where the predictions on the optical compensation models would be based on the training data used to train the machine learning model; where modified Yang would have the machine learning model from the teaching of Li, where the predictions would be aligned compensation models from Yang, where the machine learning model after training would be able to identify the specific deformation and roughness patterns that map to a specific compensation model and then use that specific compensation model).
Regarding claim 10, modified Yang teaches the method according to claim 9, as set forth above.
Modified Yang does not disclose:
wherein the machine learning algorithm is selected from the group consisting of support vector machines, regression tree systems, gradient tree enhancement systems, neural networks, Bayesian neural networks, k nearest neighbor algorithms, random forest algorithms, and combinations thereof.
However, Li discloses where the machine learning algorithm can be a gradient tree enhancement system (Page 5, Para. 5, “The invention claims a machine learning additive manufacturing residual stress and deformation prediction method based on improved GBDT model”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the machine learning model in modified Yang to be within the gradient tree enhancement system category as taught by Li.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to quickly and accurately predict deformations in the additive manufacturing sample, which can save on computing power and time, as stated by Li, Page 5, Para. 5, “The invention claims a machine learning additive manufacturing residual stress and deformation prediction method based on improved GBDT model, which can quickly and accurately predict residual stress and deformation of additive manufacturing sample”.
Regarding claim 11, modified Yang teaches the method according to claim 1, as set forth above, discloses where the laser is in electrical communication with the computing device and receives instructions therefrom for forming the target object (Yang, Para. 0030, “one or more layers are then extracted from the modified nominal CAD image 402 prior to producing a physical representation of the component from the modified nominal CAD image to provide a path for the laser for each layer in the AM process. Then at S226, a modified physical component 404 (FIG. 4E) is produced from the modified nominal CAD image 402.”, where the CAD instructions are sent to the laser path in order to form the object, where the instructions sent electronically through a computing device, Para. 0041, “operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function”).
Modified Yang does not disclose:
wherein step c) is performed by a wire and arc additive manufacturing machine comprising an arch torch, a wire-feed system, and a power source, the arch torch and wire-feed system being in electrical communication with the computing device and receiving instructions therefrom for forming the target object.
However, Li discloses where the manufacturing step includes a wire and arc additive manufacturing machine with an arc torch (Page 5, last Para., “electric arc process is cold metal transition; the laser technique is continuous optical fibre laser; welding gun is vertical; the laser is installed on one side of the arc”), a wire feed system (Page 5, last Para., “the distance between the laser spot and the welding wire is not more than 2mm.”, where wire arc additive manufacturing includes a wire feeder system to feed the wire into the welding torch), and a power source (Page 10, Para. 1, “P is the laser power; a2, b2 is respectively long axis in the x direction of the elliptical laser and the short axis in the y direction; q2 (x, y) is the heat source model of the laser.”, where a power or heat source is included). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the additive manufacturing system with a laser in modified Yang to include an arc torch, wire feeder, and power source as taught by Li; where the combined modified Yang system would have the arc torch and wire-feed system receiving electrical instructions from the computing device in Yang in order to create the product, as the laser from Yang is replaced by the arc torch and wire-feed system from Li.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage of using another additive manufacturing method to create objects, as stated by Li, Page 5, last Para., “The forming technique is laser electric arc composite additive manufacturing technology”.
Regarding the specific additive manufacturing method, it is the Examiner's position that one of ordinary skill in the art would have found it obvious to try as there are a limited amount of additive manufacturing systems that each have their own advantages and disadvantages. However, all the additive manufacturing systems have the same end result of being able to produce an object. As a result, the selection of a specific additive manufacturing system would be a mere matter of user design choice for a user to select the desired advantages needed.
Regarding claim 12, Yang discloses a method for determining a compensation plan to be applied to an initial three- dimensional model to form a modified three-dimensional model to compensate for deformation during fabrication of a target object (Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold; modifying the nominal CAD image by the compensation field; and producing a physical representation of the component from the modified nominal CAD image.”), the method comprising:
converting point-cloud to functional data (Para. 0023, “physical component 102 is measured via a measurement device 108 to capture the geometry of at least one of the external and internal surface of the physical component 102 (FIG. 4B)…In one or more embodiments, the measurement data may be point-cloud coordinate data. In one or more embodiments, the point-cloud coordinate measurements may be transformed (e.g., via postprocessing at the AM device) into triangular-mesh data”);
calculate total shape deviations (Para. 0024, “In one or more embodiments, the deviation 400 may be determined between each geometry in the nominal CAD image 101 and each corresponding measurement 109 of the physical component 102.”);
constructing an engineering-informed tensor-product basis representation of deformation patterns for each object (Para. 0033, “Then the tensor fitting is used to generate a three dimensional (3D) compensation tensor field (e.g., compensation field 508) in S316. The compensation field 508 may then be applied to the nominal CAD image 101 to modify the nominal CAD image as described above with respect to S224 in FIG. 2.”); and
creating an optimal compensation plan from a predicted deformation patterns to be applied to the target object that minimizes its shape deformation (Para. 0037, “For example, the processor 610 may receive nominal CAD image data and then may apply the geometrical compensation module 106 via the instructions of the programs 612, 614 to generate a modified CAD image to generate a component 102.”).
Yang does not disclose:
where data is used for each training object in a set of training objects;
where deviations are calculated for each training object;
where a tensor basis representation is included for each training object, where roughness can be taken into account, and
learning deformation patterns and roughness patterns that constitute the total shape deviations from the set of training objects
However, Li discloses, in the similar field of additive manufacturing (Abstract, “a prediction method and system for additive manufacturing”), where a machine learning model can take in training data that includes deformations and use those deviations caused by the deformations to make correlations on the proper corrections to be made, where these corrections can then be predicted in the future through the machine learning model (Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the point-cloud, tensor representation of deformation, and compensation plan to correct for deformations in Yang to include using those deformation data points as training data for a machine learning model in order to predict the deformation as taught by Li, where the prediction from Li would allow for the correct compensation plan to be selected.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use machine learning in order to predict a deformations with less intensive computing power needed, where predictions would allow a user to have more confidence in the compensation model as there would be training data used to in order to reach the prediction conclusion, as stated by Li, Abstract, “The invention accurately predicts the residual stress and deformation of the additive manufacturing sample under different paths with low calculation cost.”, and Page 2, Para. 8 from end, “thereby solving the problem that the current additive manufacturing residual stress and deformation of the prediction technology is difficult to ensure the prediction precision and reduce the calculation cost of the technical problem.”.
Further, Buller discloses, in the similar field of additive manufacturing (Para. 0002, “Three-dimensional (3D) printing ( e.g., additive manufacturing)”), where roughness patterns can be taken into consideration for the total shape deviation of the object (Para. 0326, “The 3D object can have various surface roughness profiles, which may be suitable for various applications. In some examples, the surface roughness is the deviations in the direction of the normal vector of a real surface ( e.g., average or mean planarity of an exposed surface of the 3D object), from its ideal form.”, and Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the shape deviations in modified Yang to take into consideration surface roughness patterns as taught by Buller.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use surface roughness as another parameter for determining the amount of correction needed for the total shape deviations of an object, as stated by Buller, Para. 0232, “the meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”.
Regarding claim 13, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein the target object has a net shape (Yang, Para. 0030, “In other words, the modified shape ( e.g., modified triangular mesh) may replace the nominal triangular mesh as the start point of the next production cycle of the physical component, in one or more embodiments.”, where the final triangular mesh of the modified shape is a net shape of the entire object).
Regarding claim 16, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein the optimal compensation plan minimizes the absolute volume deviations (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, and Para. 0029, “producing a physical representation of the component from the modified nominal CAD image to provide a path for the laser for each layer in the AM process”, where each layer has the deviations determined, where these deviations in total would be a volume deviation that is minimized during the compensation step).
Regarding claim 17, modified Yang teaches the method according to claim 16, as set forth above, discloses wherein the optimal compensation plan minimizes total absolute area deviation (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, where each layer has the total absolute area deviation minimized to a nominal area).
Regarding claim 18, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein the optimal compensation plan minimizes deviations for predicted deformation by calculating an amount of compensation that is equivalent to an area deviation (Yang, Para. 0025, “the deviation map 500 includes a model 502 of the nominal CAD image 101, a model 504 of the physical component 102, and a deviation or displacement field 506 (shown as the area between the physical component model 504 and a perimeter of the nominal CAD model 502, represented by the arrows 507)”, where the amount of compensation needed is the area deviation, Para. 0026, “If the deviation 400 is not within the pre-defined tolerance threshold, the process 200 continues to S222, and a compensation field 508 (FIG. 5) is determined, as described in more detail with respect to FIG. 3. As shown in FIG. 5, for example, the compensation field 508 may be the area between the nominal CAD model 502 and a perimeter 503 of the compensated geometry, represented by the arrows 505.”).
Regarding claim 19, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein deformation and roughness patterns are leaned from the set of training objects by training a machine learning algorithm (Yang, Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold”, where deformation patterns are taken into consideration; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness patterns are taken into consideration; teaching from Li, Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”, where the deformation and roughness data sets can be inputted as training data for a machine learning model, where the predictions on the optical compensation models would be based on the training data used to train the machine learning model).
Regarding claim 20, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein deformation and roughness patterns are leaned from the set of training objects by training a machine learning algorithm such that once trained provides deviations and/or the compensation plan for the target object (Yang, Abstract, “determining a deviation between geometry associated with the nominal CAD image and the obtained measurement data; determining a compensation field for the deviation, if the deviation is outside of a tolerance threshold”, where deformation patterns are taken into consideration and a final compensation plan is developed based on the deformations; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness patterns are taken into consideration; teaching from Li, Abstract, “performing the finite element dimensional model of the additive manufacturing sample under different manufacturing path to obtain a plurality of residual stress and deformation data; taking the three-dimensional model information containing additive manufacturing sample and the three-dimensional matrix of the manufacturing path as input data; taking the three-dimensional model information containing additive manufacturing sample and three-dimensional matrix of residual stress and deformation data as output data; combining the input data and the output data as the data set; training the classification regression tree to obtain the prediction model. using the prediction model to predict the residual stress and deformation of additive manufacturing sample under new manufacturing path.”, where the deformation and roughness data sets can be inputted as training data for a machine learning model, where the predictions on the optical compensation models would be based on the training data used to train the machine learning model; where modified Yang would have the machine learning model from the teaching of Li, where the predictions would be aligned compensation models from Yang, where the machine learning model after training would be able to identify the specific deformation and roughness patterns that map to a specific compensation model and then use that specific compensation model).
Regarding claim 21, modified Yang teaches the method according to claim 20, as set forth above.
Modified Yang does not disclose:
wherein the machine learning algorithm is selected from the group consisting of support vector machines, regression tree systems, gradient tree enhancement systems, neural networks, Bayesian neural networks, k nearest neighbor algorithms, random forest algorithms, and combinations thereof.
However, Li discloses where the machine learning algorithm can be a gradient tree enhancement system (Page 5, Para. 5, “The invention claims a machine learning additive manufacturing residual stress and deformation prediction method based on improved GBDT model”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the machine learning model in modified Yang to be within the gradient tree enhancement system category as taught by Li.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to quickly and accurately predict deformations in the additive manufacturing sample, which can save on computing power and time, as stated by Li, Page 5, Para. 5, “The invention claims a machine learning additive manufacturing residual stress and deformation prediction method based on improved GBDT model, which can quickly and accurately predict residual stress and deformation of additive manufacturing sample”.
Regarding claim 22, modified Yang teaches the method according to claim 12, as set forth above, discloses where the laser is in electrical communication with the computing device and receives instructions therefrom for forming the target object (Yang, Para. 0030, “one or more layers are then extracted from the modified nominal CAD image 402 prior to producing a physical representation of the component from the modified nominal CAD image to provide a path for the laser for each layer in the AM process. Then at S226, a modified physical component 404 (FIG. 4E) is produced from the modified nominal CAD image 402.”, where the CAD instructions are sent to the laser path in order to form the object, where the instructions sent electronically through a computing device, Para. 0041, “operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function”).
Modified Yang does not disclose:
wherein step c) is performed by a wire and arc additive manufacturing machine comprising an arch torch, a wire-feed system, and a power source, the arch torch and wire-feed system being in electrical communication with the computing device and receiving instructions therefrom for forming the target object.
However, Li discloses where the manufacturing step includes a wire and arc additive manufacturing machine with an arc torch (Page 5, last Para., “electric arc process is cold metal transition; the laser technique is continuous optical fibre laser; welding gun is vertical; the laser is installed on one side of the arc”), a wire feed system (Page 5, last Para., “the distance between the laser spot and the welding wire is not more than 2mm.”, where wire arc additive manufacturing includes a wire feeder system to feed the wire into the welding torch), and a power source (Page 10, Para. 1, “P is the laser power; a2, b2 is respectively long axis in the x direction of the elliptical laser and the short axis in the y direction; q2 (x, y) is the heat source model of the laser.”, where a power or heat source is included). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the additive manufacturing system with a laser in modified Yang to include an arc torch, wire feeder, and power source as taught by Li; where the combined modified Yang system would have the arc torch and wire-feed system receiving electrical instructions from the computing device in Yang in order to create the product, as the laser from Yang is replaced by the arc torch and wire-feed system from Li.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage of using another additive manufacturing method to create objects, as stated by Li, Page 5, last Para., “The forming technique is laser electric arc composite additive manufacturing technology”.
Regarding the specific additive manufacturing method, it is the Examiner's position that one of ordinary skill in the art would have found it obvious to try as there are a limited amount of additive manufacturing systems that each have their own advantages and disadvantages. However, all the additive manufacturing systems have the same end result of being able to produce an object. As a result, the selection of a specific additive manufacturing system would be a mere matter of user design choice for a user to select the desired advantages needed.
Claims 3 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yang et al. (US 20170368753 A1, hereinafter Yang) in view of Li et al. (CN 113435083 A, hereinafter Li) and Buller et al. (US 20170239892 A1, hereinafter Buller) in further view of Fonte et al. (CN 105637512 A, hereinafter Fonte).
Regarding claim 3, modified Yang teaches the method according to claim 1, as set forth above, discloses wherein when the target object has a circular cross-section, the shape deviations are given by:
PNG
media_image1.png
138
436
media_image1.png
Greyscale
ε ~ N (0, σ-2) (Yang, Para. 0024, “nominal or desired geometry N (x, y, z) and the displacement vectors N for the difference between the nominal CAD image 101 and the physical component 102.
PNG
media_image2.png
170
570
media_image2.png
Greyscale
”, where the total amount of deformation are taken into consideration; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness can be taken into consideration for the compensation model)
wherein: yk(z,t,Ro) is the deviations;
ro is a target radius at position z;
θ , z, ro are cylindrical coordinates:
ψ and ϕ are basis expansion functions;
k1 and k2 are the number of basis functions ψ and ϕ, respectively;
β and yk are coefficient vectors for the basis expansion functions;
α (-) is a function that captures size effects over a roughness pattern;
σ-2 is a variance of measurement error; and
N is the normal distribution.
Modified Yang does not disclose:
where measurement error is taken into consideration.
However, Fonte discloses, in the similar field of additive manufacturing (Claim 10, “computer system configuration is as additive manufacturing equipment generating the customized product of 2D or 3D expression”), where measurement error is taken into consideration for data (Page 20, Para. 3, “measuring a plurality of predetermined size is size of anatomical model scale factor, and h) computer system may, in addition to the reference target in each frame in the size average value or weighted in order to reduce any single size measurement error.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the shape deviation calculation in modified Yang to include measurement error consideration as taught by Fonte.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use an average or weighted average for data to reduce measurement error significance, which can assist with making the data set more accuracy, as stated by Fonte, Page 20, Para. 3, “measuring a plurality of predetermined size is size of anatomical model scale factor, and h) computer system may, in addition to the reference target in each frame in the size average value or weighted in order to reduce any single size measurement error.”.
Regarding the specific equations for deformation, surface roughness, and measurement error, it is the Examiner's position that one of ordinary skill in the art would have found it obvious to try as there are a limited amount of mathematical representations of each of the shape deviation features. Yang, Buller, and Fonte all discuss methods in which to calculate the specific shape deviation features and have associated equations. As a result, the selection of a specific coordinate system and equations that represent deformation, surface roughness, and measurement error would be mere matters of user design choice.
Regarding claim 14, modified Yang teaches the method according to claim 12, as set forth above, discloses wherein when the target object has a circular cross-section, the shape deviations are given by:
PNG
media_image1.png
138
436
media_image1.png
Greyscale
ε ~ N (0, σ-2) (Yang, Para. 0024, “nominal or desired geometry N (x, y, z) and the displacement vectors N for the difference between the nominal CAD image 101 and the physical component 102.
PNG
media_image2.png
170
570
media_image2.png
Greyscale
”, where the total amount of deformation are taken into consideration; teaching from Buller, Para. 0232, “meteorological measurement(s) may facilitate determination and/or subsequent compensation for a roughness and/or inclination of the exposed surface of the material bed with respect to the platform and/or horizon.”, where surface roughness can be taken into consideration for the compensation model)
wherein: yk(z,t,Ro) is the deviations;
ro is a target radius at position z;
θ , z, ro are cylindrical coordinates:
ψ and ϕ are basis expansion functions;
k1 and k2 are the number of basis functions ψ and ϕ, respectively;
β and yk are coefficient vectors for the basis expansion functions;
α (-) is a function that captures size effects over a roughness pattern;
σ-2 is a variance of measurement error; and
N is the normal distribution.
Modified Yang does not disclose:
where measurement error is taken into consideration.
However, Fonte discloses, in the similar field of additive manufacturing (Claim 10, “computer system configuration is as additive manufacturing equipment generating the customized product of 2D or 3D expression”), where measurement error is taken into consideration for data (Page 20, Para. 3, “measuring a plurality of predetermined size is size of anatomical model scale factor, and h) computer system may, in addition to the reference target in each frame in the size average value or weighted in order to reduce any single size measurement error.”). It would have been obvious for one of ordinary skill in the art before the effective filling date of the claimed invention to have modified the shape deviation calculation in modified Yang to include measurement error consideration as taught by Fonte.
One of ordinary skill in the art would have been motivated to make this modification in order to gain the advantage being able to use an average or weighted average for data to reduce measurement error significance, which can assist with making the data set more accuracy, as stated by Fonte, Page 20, Para. 3, “measuring a plurality of predetermined size is size of anatomical model scale factor, and h) computer system may, in addition to the reference target in each frame in the size average value or weighted in order to reduce any single size measurement error.”.
Regarding the specific equations for deformation, surface roughness, and measurement error, it is the Examiner's position that one of ordinary skill in the art would have found it obvious to try as there are a limited amount of mathematical representations of each of the shape deviation features. Yang, Buller, and Fonte all discuss methods in which to calculate the specific shape deviation features and have associated equations. As a result, the selection of a specific coordinate system and equations that represent def