An artificial intelligence-based 3D printing cement-based material design method

Abstract

The application provides a 3D printing cement-based material design method based on artificial intelligence. The intelligent design method comprises the following steps: determining a mix proportion of component raw materials in the 3D printing cement-based material according to a required performance index and a mix proportion constraint range, wherein the component raw materials comprise cementitious materials, sand, water and admixtures, and the mix proportion constraint range is a preset value range of the amount of each raw material in the component raw materials; determining printing parameters, and a target printing parameter is a printing parameter that meets the printability under the constraint range of the printing parameters and rheological parameters, wherein the constraint range of the printing parameters and rheological parameters is a value range of each printing parameter; and printing the material and testing the performance of the material based on the mix proportion and the target printing parameter. The intelligent design method provided by the application can save design time and cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the fields of 3D printing cement-based materials and machine learning technology, and particularly relates to a 3D printing cement-based material design method based on artificial intelligence. BACKGROUND

[0002] 3D printing technology is a disruptive manufacturing technology that can convert three-dimensional digital models of components into physical components during the design phase. It can build complex geometries by layering materials, so it is called additive manufacturing technology. This technology not only greatly improves manufacturing efficiency and reduces manufacturing costs, but also improves the reliability, strength and quality of components. In the development of 3D printing technology, the construction industry has also gradually begun to explore its application, and 3D printing construction has become a field of great concern.

[0003] In the field of construction, traditional construction processes require the production of a large number of auxiliary tools such as templates and scaffolding, which not only takes time and effort, but also increases costs and wastes resources. 3D printing technology can directly convert components into physical objects through digital design, avoiding the need for manufacturing auxiliary tools required by traditional construction, thereby greatly reducing construction costs and labor costs. In addition, 3D printing technology can also achieve the construction of free-form surfaces and fine control of details and structures, making architectural design more personalized and flexible. At the same time, the high efficiency, high precision and repeatability of 3D printing technology also make construction safer, faster and more efficient.

[0004] There have been many successful 3D printing construction cases at home and abroad. For example, an American construction company used 3D printing technology to build a 6-meter-high building, while a Dutch construction company used 3D printing technology to design and build a multi-functional building. Although 3D printing technology has achieved some successful applications in the construction field, research on printing concrete materials is relatively less, and technology needs to be improved and improved.

[0005] At the same time, with the rapid development of artificial intelligence technology, its application in the field of 3D printing is becoming more and more widespread. Artificial intelligence can use a large amount of data for training and learning to obtain more accurate and efficient results, thereby improving the efficiency and accuracy of 3D printing concrete design. At present, in the field of 3D printing concrete, artificial intelligence has been widely used in material performance prediction, printing path optimization, structure design and other aspects, providing strong support for the development of 3D printing concrete.

[0006] In view of the problems existing in the current 3D printing concrete, the present application proposes a 3D printing cement-based material design method based on artificial intelligence. The method first collects a large amount of experimental data, and combines advanced machine learning algorithms to model, obtaining a model with predictive ability. Then, through the model, the design of cement-based materials is carried out, and the mechanical properties of the materials and the printing process parameters are optimized to meet different application requirements.

[0007] Specifically, the method includes the following steps: first, a large amount of concrete experimental data is collected, including material composition, mixing ratio, mechanical properties, etc. Second, use machine learning algorithm to model and train data, establish a model that can predict the performance of concrete. Then, through the model, the design of cement-based materials is carried out, and the composition and proportion of the materials are optimized to achieve the expected mechanical properties. Finally, the designed cement-based materials are printed into shape by 3D printing technology, and the final product is obtained.

[0008] Compared with the traditional trial and error method, the method has high efficiency and accuracy, can greatly shorten the product development cycle, reduce cost and improve product quality. At the same time, the method also has certain universality and operability, and can provide strong support and guarantee for the application and development of 3D printing concrete. SUMMARY

[0009] The present application provides a 3D printing cement-based material design method based on artificial intelligence, aiming to reduce the amount of cement in 3D printing cement-based materials, thereby reducing the preparation cost of 3D printing cement-based materials, and promoting the sustainable development and intelligent process of the construction industry. The method includes intelligent mix design and target printing parameter design, and is based on optimization algorithm to realize intelligent design. The intelligent design method of the present application can effectively reduce the design cost and time, and improve the preparation efficiency and quality of 3D printing cement-based materials.

[0010] Traditional cement-based material design methods are usually based on experience and trial-and-error, which are inefficient, costly, and difficult to optimize. The 3D printing cement-based material design method based on artificial intelligence provided by the present application can effectively solve these problems. First, the method determines the mix proportion of the constituent raw materials in the 3D printing cement-based material according to the user's performance indicators and the mix proportion constraint range. These constituent raw materials include cementitious materials, sand, water, and admixtures. The mix proportion constraint range is the preset value range of the amount of each raw material, which can ensure that the 3D printing cement-based material meets the performance indicators while minimizing costs. Second, the method determines the target printing parameters, including the optimal printing parameters under the constraints of printing parameters and rheological parameters, to ensure that the 3D printing cement-based material can be smoothly printed on the 3D printer and the material quality is stable and reliable. Finally, the method prints and tests the performance of the material based on the mix proportion and target printing parameters. Through intelligent design, the present application can realize the intelligent, optimized, and efficient preparation process of 3D printing cement-based materials, further promoting the sustainable development and intelligent process of the construction industry.

[0011] In certain implementations, the method provided by the present application first establishes a 3D printing cement-based material printability model, the input parameters of which include 3D printing cement-based material rheological parameters and 3D printer printing parameters. Rheological parameters include static yield stress, dynamic yield stress, and plastic viscosity, etc.; printing parameters include printing speed, printing height, printer speed, and printing port diameter, etc. The output parameter is printability, including layer height error and layer width error. Through an optimization algorithm, the 3D printing cement-based material rheological parameters and 3D printer printing parameters are intelligently designed according to the user-specified printability indicators. During the optimization process, the differential evolution algorithm can be used to minimize the printability error to obtain the best printing parameters. Specifically, the differential evolution algorithm searches for the optimal solution in the parameter space by performing operations such as "crossover" and "mutation" on individuals in the population. After optimization is complete, the system outputs the optimal printing parameters, including printing speed, printing height, printer speed, and printing port diameter.

[0012] Secondly, based on the optimal printing parameters, a 3D printing cement-based material rheological model is established. In this rheological model, the input parameters are the mix proportion parameters of the 3D printing cement-based material, including cement, fly ash, limestone powder, water, sand, thickening agent, and water reducing agent. The output parameters are the rheological parameters of the 3D printing cement-based material, including static yield stress, dynamic yield stress, and plastic viscosity. In this rheological model, we use the Bingham plastic rheological model commonly used in the concrete community. This model assumes that a certain yield stress needs to be reached for the initial deformation of the material, after which the material can begin to flow. After establishing the rheological model, we can predict the rheological parameters under different mix proportions, thereby guiding actual printing production.

[0013] Further, the evaluation method for evaluating the established machine learning model is to calculate the average relative error, the average absolute error, the root mean square error, the mean square error and the correlation coefficient of the test target value and the predicted target value of the test set.

[0014] The calculation method of the average relative error MRE is:

[0015]

[0016] The calculation method of the average absolute error MAE is:

[0017]

[0018] The calculation method of the root mean square error RMSE is:

[0019]

[0020] The calculation method of the correlation coefficient R 2 is:

[0021]

[0022] The calculation method of the mean square error MSE is:

[0023]

[0024] Wherein, y i is the test target value of the test set, y′ i is the predicted target value of the test set, n is the sample number, is the average value of the true value, i=1,2,…,n.

[0025] The beneficial effects of the present application are that the present application establishes the nano-enhanced cement-based material database, and establishes the performance optimization model of the nano-enhanced cement-based material based on the prediction algorithm and the optimization algorithm; the present application provides an intelligent design method for the nano-enhanced cement-based material, which can recommend the production mixing ratio of the nano-enhanced cement-based material meeting the performance requirements to the user in real time and efficiently, and saves the economic, time and environmental costs.

[0026] In the intelligent design method provided by the present application, we realize the intelligent design of parameters by establishing the printability model and rheological model. This method has advantages in many aspects. First, this method combines traditional experimental design and numerical simulation, effectively shortening the design cycle and reducing the design cost. Second, this method can intelligently design 3D printing cement-based materials according to user-specified performance indicators, making the design more flexible and personalized. Finally, this method improves the production efficiency and quality stability of 3D printing cement-based materials, which is conducive to the sustainable development and intelligent development of the construction industry.

[0027] The present application provides an artificial intelligence-based 3D printing cement-based material design method. This method can realize the intelligent design of parameters by establishing a printability model and a rheological model, effectively shortening the design cycle and reducing the design cost. In addition, this method not only saves cost, but also greatly reduces the amount of cement used in 3D printing cement-based materials, thereby reducing the preparation cost and promoting the sustainable development and intelligent development of the construction industry. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required to be used in the prior art description will be briefly introduced as follows:

[0029] Figure 1 is a schematic flow chart of the artificial intelligence-based 3D printing cement-based material design method provided by the present application. DETAILED DESCRIPTION

[0030] The present application provides an artificial intelligence-based 3D printing cement-based material design method, which aims to reduce the amount of cement used in 3D printing cement-based materials, thereby reducing the preparation cost of 3D printing cement-based materials and promoting the sustainable development and intelligent process of the construction industry. The implementation of the method mainly includes the following steps:

[0031] Step one: determine the performance indicators and mix proportion constraint range

[0032] In this method, the performance indicators and mix proportion constraint range of 3D printing cement-based materials need to be determined first. Performance indicators include strength, density, processability, etc. The mix proportion constraint range includes the usage limit of cement, sand, water, and other raw materials.

[0033] Different performance indicators and mix proportion constraint ranges can be established for different application scenarios and usage requirements. For example, in the construction industry, 3D printed cement-based materials need to have certain strength and stability, while also having good processability to achieve more complex building structures. In road engineering, 3D printed cement-based materials need to have good durability and compression resistance to adapt to different road conditions and loads.

[0034] When determining the mix proportion of the constituent raw materials in 3D printed cement-based materials, multiple factors need to be considered. For example, cementitious materials are an important component of cement-based materials, while sand can adjust the strength and density of the material. Water can promote the flowability and processability of the material, while additives can improve the performance of the material. Therefore, when determining the mix proportion, these factors need to be considered comprehensively, and the cost needs to be reduced as much as possible while ensuring the strength and stability of the material. At the same time, attention also needs to be paid to the range of the mix proportion to avoid waste of raw materials or instability of material performance.

[0035] Example: Suppose a user needs to prepare a 3D printed cement-based material with a strength of 30 MPa, and the mix proportion constraint range is cement dosage of 350-500 kg / m 3 , sand dosage of 700-900 kg / m 3 , water dosage of 100-200 kg / m 3 , and additive dosage of 0-10 kg / m 3 . According to these requirements, an artificial intelligence algorithm can be used to intelligently design the mix proportion of each constituent raw material.

[0036] Table 1 Mix Proportion Constraint Range

[0037]

[0038] Step Two: Determine Target Printing Parameters

[0039] In the manufacturing process of 3D printed cement-based materials, determining the target printing parameters is a very critical step. The target printing parameters refer to the optimal printing parameters that can be smoothly printed within a certain range of rheological parameter constraints, which directly affects the quality, efficiency and cost of 3D printed cement-based materials. Therefore, the process of determining the target printing parameters needs to consider multiple factors, including the characteristics of the printer, the characteristics and rheological behavior of the material, and the requirements and requirements of printing.

[0040] Firstly, the characteristics and limitations of the 3D printer need to be considered. Different models of 3D printers have different printing speeds, printing heights, printing pressures, printing temperatures, nozzle diameters, and other aspects. When determining the target printing parameters, the optimal parameter range of the specific 3D printer model needs to be determined. For example, a certain model of 3D printer has a maximum printing speed of 40 mm / s, a maximum printing height of 0.3 mm, and a maximum printing temperature of 100℃, which need to be considered in the target printing parameter range.

[0041] Secondly, the characteristics and rheological behavior of the material need to be considered. Different cement-based materials have different physical, chemical, and rheological characteristics, such as viscosity, fluidity, plasticity, etc. These characteristics directly affect the printing ability of the material and the quality of the finished product. When determining the target printing parameters, the rheological behavior and characteristics of the material need to be considered to ensure that the material can be normally printed on the 3D printer and obtain the required performance and quality.

[0042] Finally, the requirements and demands of printing need to be considered. Different printing tasks have different requirements, such as printing speed, precision, pressure, density, hardness, etc. When determining the target printing parameters, the requirements and demands of printing need to be considered to meet different printing tasks.

[0043] Example: Suppose the printing speed range of the 3D printer is 20-40 mm / s, the printing height range is 0.1-0.3 mm, the printer speed range is 500-1000 rpm, and the printing port diameter range is 0.3-0.5 mm. According to these requirements, an artificial intelligence algorithm can be used to intelligently design the target printing parameters.

[0044] Table 2 Target printing parameter range

[0045]

[0046] Step three: Intelligent mix design and target printing parameter design

[0047] After determining the performance indicators and mix constraint ranges and target printing parameters, this method enters the intelligent mix design and target printing parameter design stage. This method uses intelligent mix design and target printing parameter design to achieve the best balance between performance indicators and cost-effectiveness. The core of this stage is to use artificial intelligence algorithms to intelligently design and optimize the mix and target printing parameters. These algorithms can automatically identify and optimize the best mix and printing parameters by learning historical data and experimental results, thereby improving the preparation effect and reducing the preparation cost.

[0048] Intelligent mix design is based on artificial intelligence algorithms to intelligently design and optimize the mix. Genetic algorithms, simulated annealing algorithms, neural network algorithms, etc. can be used to optimize the mix. These algorithms can automatically identify the best mix through learning from historical data and experimental results. Under the premise of mix constraint range and performance indicators, intelligent mix design can automatically achieve the best balance between performance indicators and cost-effectiveness.

[0049] Target printing parameter design is based on artificial intelligence algorithms to intelligently design and optimize the printing parameters. For example, deep learning algorithms can be used to model and optimize the 3D printer printing process, predict the rheological properties of the material, and automatically optimize the optimal printing parameters to ensure smooth printing of 3D printed cement-based materials with stable and reliable quality. Through intelligent design and optimization, target printing parameters can achieve the best effect, improving preparation efficiency and preparation quality.

[0050] The comprehensive use of artificial intelligence algorithms for intelligent design and optimization of mix and target printing parameters can achieve the optimal balance of 3D printed cement-based material preparation effect and cost-effectiveness. For example, when users need to prepare high-strength 3D printed cement-based materials, they can use artificial intelligence-based mix design methods to obtain intelligently designed mix, thus meeting the performance indicators and cost-effectiveness requirements of users. This not only improves preparation efficiency and preparation quality, but also promotes the sustainable development and intelligent process of the construction industry.

[0051] Example: Suppose a user needs to prepare high-strength 3D printed cement-based materials, with performance indicators including compressive strength greater than 50MPa and flexural strength greater than 5MPa, and mix constraint range of cement dosage 350-500kg / m 3 , sand dosage 700-900kg / m 3 , water dosage 100-200kg / m 3 , and admixture dosage 0-10kg / m 3 . According to these requirements, an artificial intelligence-based mix design method can be used to obtain an intelligently designed mix of cement dosage 460kg / m 3 , sand dosage 720kg / m 3 , water dosage 120kg / m 3 , and admixture dosage 8kg / m 3 .

[0052] Table 3 Intelligent mix design

[0053]

[0054] Step four: experimental verification and performance evaluation

[0055] The final step is experimental verification and performance evaluation. In the laboratory, the 3D printed cement-based material designed intelligently is prepared, and a series of performance tests and evaluations, such as compressive strength, tensile strength, and durability tests, are conducted to evaluate the performance of the material. According to the experimental results, the intelligent design method can be verified and optimized.

[0056] In the experimental verification and performance evaluation stage, the 3D printed cement-based material designed intelligently needs to be prepared, and various performance tests and evaluations need to be conducted. These performance tests and evaluations can help us evaluate the mechanical properties, physical properties, and durability of the material. For example, the compressive strength test can test the load-carrying capacity of the material under compressive stress; the tensile strength test can test the load-carrying capacity of the material under tensile stress; and the durability test can evaluate the performance of the material under long-term use and various external environments.

[0057] In addition to performance tests and evaluations, the experimental verification stage can also verify and optimize the intelligent design method. Through the analysis and summary of experimental results, the advantages and disadvantages of material performance can be determined, potential problems can be found, and the intelligent design method can be optimized accordingly. This repeated experimental verification and optimization process can help us continuously improve the intelligent design method, improve material performance and preparation efficiency, and ultimately achieve the goal of sustainable development and intelligent construction industry

[0058] Example: Suppose the user needs to prepare a 3D printed cement-based material with reasonable microstructure and stable and reliable performance, and the target printing parameter constraint range is that the printing speed is between 25-35mm / s, the printing height is 0.1-0.2mm, the printer speed is between 600-800rpm, and the printing port diameter is 0.4-0.5mm. According to these requirements, the target printing parameter design method based on artificial intelligence can be used to obtain the intelligent design of the target printing parameter: the printing speed is 30mm / s, the printing height is 0.15mm, the printer speed is 700rpm, and the printing port diameter is 0.45mm.

[0059] This method provides a 3D printed cement-based material design method based on artificial intelligence, aiming to reduce the amount of cement in 3D printed cement-based materials, thereby reducing the preparation cost of 3D printed cement-based materials, promoting the sustainable development and intelligent process of the construction industry. This method not only improves the performance and quality stability of 3D printed cement-based materials, but also reduces resource waste and environmental pollution, and has broad application prospects.

Claims

1. An artificial intelligence-based 3D printing cementitious material design method, characterized by, The method comprises the following steps: determining target printing parameters and rheological parameters, the target printing parameters being printing parameters meeting the printability requirements within a printing parameter constraint range, and the target rheological parameters being rheological parameters meeting the printability requirements within a rheological parameter constraint range, wherein the constraint ranges are the value ranges of the respective printing parameters and rheological parameters; determining the mix proportion of the constituent raw materials in the 3D printing cement-based material according to the required rheological performance indicators and the mix proportion constraint range, wherein the constituent raw materials include cementitious materials, sand, water, and additives, and the mix proportion constraint range is the value range of the respective raw material quantities in the 3D printing cement-based material; printing the material based on the mix proportion and the target printing parameters and testing the performance of the 3D printing cement-based material; The determination of the target printing parameters comprises: performing global optimization of the printing parameters by an optimization algorithm based on a printability objective function and the printing parameter and rheological parameter constraint ranges, to obtain P printing and rheological parameters and P printability indicators, the P printing and rheological parameters and the P printability indicators corresponding one-to-one, and determining the printing and rheological parameters corresponding to the optimal printability among the P printability indicators as the target printing and rheological parameters, wherein the printability objective function is a mapping function of the printing and rheological parameters and the printability indicators, P is an integer greater than or equal to 1, and the printability indicators include layer height error and layer width error; The determination of the mix proportion of the constituent raw materials in the 3D printing cement-based material according to the required rheological performance indicators and the mix proportion constraint range comprises: performing global optimization of the mix proportion by an optimization algorithm based on a rheological objective function and the mix proportion constraint range, to obtain N mix proportions and N rheological performance parameters, the N mix proportions and the N rheological performance parameters corresponding one-to-one, and determining the mix proportion corresponding to the rheological performance parameter closest to the rheological performance indicators among the N rheological performance parameters as the mix proportion, wherein the rheological objective function is a mapping function of the mix proportion of the constituent raw materials and the rheological performance parameters, and N is an integer greater than or equal to 1.

2. The method of designing a 3D printed cementitious material according to claim 1, wherein, The optimization algorithm is a differential evolution algorithm.

3. The method of designing a 3D printed cementitious material according to claim 1, wherein, The constituent raw materials include cement, fly ash, limestone powder, water, sand, thickening agent, and water reducing agent.

4. The method of designing a 3D printed cementitious material according to claim 1, wherein, The rheological performance parameters include the following parameters: static yield stress, dynamic yield stress, and plastic viscosity.

5. The method of designing a 3D printed cementitious material according to claim 1, wherein, The printing parameters include printing speed, printing height, printer speed, and printing port diameter, and the printability indicators corresponding to each group of printing parameters, including layer height error and layer width error.

6. The 3D printed cementitious material design method according to any one of claims 1-5, wherein, Further comprising the step of real-time monitoring and feedback adjustment of the 3D printing cement-based material to ensure the quality and stability of the printing process.

7. The 3D printed cementitious material design method according to any one of claims 1-5, wherein, Further comprising the step of analyzing and modeling the experimental data and printing parameters by a machine learning algorithm to optimize the performance and printing effect of the 3D printing cement-based material.

8. The 3D printed cementitious material design method according to any one of claims 1-5, wherein, Further comprising the step of post-processing the 3D printing cement-based material, including the curing step, to further improve its strength and durability.

Images