Hybrid model and method for determining manufacturing properties of an injection-molded part
View Patent ↗A method of determining properties relating to the manufacture of an injection-molded article is described. The method makes use of a hybrid model which includes at least one neural network and at least one rigorous model. In order to forecast (or predict) properties relating to the manufacture of a plastic molded part, a hybrid model is used which includes: one or more neural networks NN 1 , NN 2 , NN 3 , NN 4 , . . . , NN k ; and one or more rigorous models R 1 , R 2 , R 3 , R 4 , . . . , which are connected to one another. The rigorous models are used to map model elements which can be described in mathematical formulae. The neural model elements are used to map processes whose relationship is present only in the form of data, as it is typically impossible to model such processes rigorously. As a result, a forecast (or prediction) relating to properties including, for example, the mechanical, thermal and rheological processing properties and relating to the cycle time of a plastic molded part can be made.
1. A method of determining properties relating to the manufacture of an injection-molded part comprising:
(a) providing a hybrid model comprising,
(i) a first input ( 1 ) for inputting parameters,
(ii) at least one neural network (NN 1 , . . . NN k ) having an input for inputting at least one of said parameters into the neural network, and having a first output for outputting a forecast value,
(iii) at least one rigorous model (R 1 , R 2 , R 3 ; R 4 , R 5 , . . . ) having a second input for inputting at least one of said parameters and said forecast value, and a second output for outputting a calculated value relating to the properties,
(iv) a means for mapping an input vector of parameters (P 1 , P 2 , P 3 , P 4 , P 5 . . . P 1 ) onto a subvector (PN 1 , PN 2 , PN 3 . . . . PN 1 ), and
(v) a means ( 4 ) for mapping the subvector onto corresponding inputs of the neural networks;
(b) selecting said parameters;
(c) inputting said parameters into said hybrid model; and
(d) obtaining at least one of said forecast value and said calculated value relating to at least one of said properties.
2. The method of claim 1 wherein said hybrid model comprises a multiplicity of neural networks (NN 1 , . . . NN k ) for outputting a corresponding multiplicity of forecast values.
3. The method of claim 1 wherein the parameters input into each neural network are selected independently from at least one of molded-part parameters, recipe parameters, process parameters and machine parameters.
4. The method of claim 1 wherein said forecast value relates to at least one of the following properties: rheological properties, processing properties, mechanical properties, thermal properties, and processing time.
5. The method of claim 1 wherein the output of each rigorous model relates independently to at least one of the following properties: rheological properties, processing properties, mechanical properties, thermal properties, and processing time.
6. The method of claim 1 wherein each neural network independently outputs at least one of the following forecast values: melt viscosity, Carreau parameters, filling pressure, plastifying capacity, modulus of elasticity, yield stress, yield elongation, tearing stress, breaking strain, work to yield stress, work to breakage, nominal breaking strain, static friction, sliding friction, shrinkage, demolding temperature, bending modulus, bending stress, peripheral fiber strain, resistance to breaking, density, sealing temperature, and molded part mass.
7. The method of claim 1 wherein each rigorous model independently outputs at least one calculated value relating to at least one of the following properties: cycle time, mold-filling time, injection time, metering time, cooling time, customer-specific filling pressure, Carreau melt viscosity, effective thermal conductivity, modulus of transverse elasticity, transverse tearing stress, and transverse breaking strain.
8. The method of claim 1 wherein said hybrid model further comprises a rigorous model for calculating a mold-filling time from a volume of the molded part and a speed of injecting thermoplastic material into mold.
9. The method of claim 1 wherein said hybrid model further comprises a rigorous model for calculating the mass of the molded part from a volume of the molded part and a density of the molded part.
10. The method of claim 1 wherein said hybrid model further comprises a rigorous model for calculating an effective thermal conductivity from a thickness of the injection-molded part, and a sealing time, a mass temperature, a mold temperature and a sealing temperature of a molding process by which said injection-molded part is prepared.
11. The method of claim 1 wherein said hybrid model comprises a rigorous model for calculating a cooling time of the manufacturing process by which said molded part is prepared, said cooling time being calculated from at least one process parameter selected from: the effective thermal conductivity and the thickness of said molded part; and a mass temperature, a mold temperature and a demolding temperature of the manufacturing process by which said molded part is prepared.
12. The method of claim 1 wherein said hybrid model comprises a second rigorous model, one input of the second rigorous model being connected to at least one output of another rigorous model in order to input the calculated value.
13. A method of determining properties relating o the manufacture of an injection-molded part comprising:
(a) providing a hybrid model comprising,
(i) a first input ( 1 ) for inputting parameters,
(ii) at least one neural network (NN 1 , . . . NN k ) having an input for inputting at least one of said parameters into the neural network, and having a first output for outputting a forecast value, and
(iii) at least one rigorous model (R 1 , R 2 , R 3 ; R 4 , R 5 , . . . having a second input for inputting at least one of said parameters and said forecast value, and a second output for outputting a calculated value relating to the properties,
(b) selecting said parameters;
(c) inputting said parameters into said hybrid model; and
(d) obtaining at least one of said forecast value and said calculated value relating to at least one of said properties
wherein each rigorous model independently outputs at least one calculated value relating to at least one of the following properties: cycle time, mold-filling time, injection time, metering time, cooling time, customer-specific filling pressure, Carreau melt viscosity, effective thermal conductivity, modulus of transverse elasticity, transverse tearing stress, and transverse breaking strain.
14. The method of claim 13 wherein said hybrid model further comprises a rigorous model for calculating a mold-filling time from a volume of the molded part and a speed of injecting thermoplastic material into a mold.
15. The method of claim 13 wherein said hybrid model further comprises a rigorous model for calculating the mass of the molded part from a volume of the molded part and a density of the molded part.
16. The method of claim 13 wherein said hybrid model further comprises a rigorous model for calculating an effective thermal conductivity from a thickness of the injection-molded part, and a sealing time, a mass temperature, a mold temperature and a sealing temperature of a molding process by which said injection-molded part is prepared.
17. The method of claim 13 wherein said hybrid model comprises a rigorous model for calculating a cooling time of the manufacturing process by which said molded part is prepared, said cooling time being calculated from at least one process parameter selected from: the effective thermal conductivity and the thickness of said molded part; and a mass temperature, a mold temperature and a demolding temperature of the manufacturing process by which said molded part is prepared.
18. The method of claim 13 wherein said hybrid model comprises a second rigorous model, one input of the second rigorous model being connected to at least one output of another rigorous model in order to input the calculated value.