METHOD FOR REAL-TIME CONTROLLING RESIN TRANSFER MOLDING PROCESS

摘要

The present invention is a method for real-time controlling a resin transfer molding process, which is used to control a filling pressure of a resin in a resin transfer molding (RTM) apparatus. In a pre-control RTM process, the current filling pressure, the current permeability and the wave front position at the current time point is input into a prediction-control model to acquire a predicted filling pressure at the next time point. The predicted filling pressure is used as the filling pressure to make the resin flow to the expected position of the wave front at the next time point, whereby to achieve stable quality of RTM products.

主权项

1. A method for real-time controlling a resin transfer molding process, which is used to control a filling pressure of a resin in a resin transfer molding (RTM) apparatus, wherein the RTM apparatus comprises a resin supply unit and a molding unit connected with the resin supply unit, and wherein the molding unit includes a mold cavity accommodating a pre-woven fiber object (a preform) and a plane inside the mold cavity, and wherein the method comprises:
Step 1: establishing a plurality of groups of process training conditions, wherein the process training conditions include an initial filling pressure; Step 2, performing a plurality of training processes according to the process training conditions to acquire a permeability of the resin, a wave front position of the resin at a current time point and a wave front position of the resin at a next time point for a plurality of time points of each training process, wherein the permeability, the wave front position at the current time point and the wave front position at the next time point are acquired with Steps 2A-2F: Step 2A: defining on the plane a plurality of detection positions ym,n whose number amounts to m×n; Step 2B: providing a detection module including a pressure transducer unit, at least one image capture device and a processing unit, wherein the pressure transducer unit is arranged in the detection positions ym,n, and wherein the image capture device is arranged on one side of the plane, and wherein the processing unit is electrically connected with the pressure transducer unit and the image capture device, and wherein the pressure transducer unit includes m×n pieces of pressure transducers; Step 2C: filling the resin into the mold cavity with the initial filling pressure and letting the resin flow on the plane along a direction; Step 2D: using the image capture device to obtain positions of the resin on the plane at a time point ti so as to define on the plane a plurality of measurement positions whose number amounts to i×j, wherein the time point ti and a time point ti−1, or a time point ti+1 and the time point ti, are separated by a sampling interval, and wherein the measurement position xi,j is a position corresponding to a position of a wave front of the resin at the time point ti, and wherein i denotes the ith sampling time point, and j is an integer related to II; Step 2E: setting i and j to be preset values a and r respectively, wherein r is an integer greater than 1 and a is an integer greater than or equal to 1; using the image capture device to obtain the measurement positions xr+1, xr,a and xr−1,a of the resin respectively at the time points tr+1, tr and tr−1; and using the pressure transducer, which is nearest to the measurement position xr,a and the resin has reached, to obtain the pressure Ps,a of the resin at the detection position ys,a, wherein xr+1,a is the wave front position at the current time point and xr+1,a is the wave front position at the next time point in the training process; Step 2F: using the processing unit to obtain the permeability Kr,a of the measurement position xr,a with Equation (1):Kr,a=μφPs,aΔT(xr,a-xr-1,a)(xr,a-ys,a)(1) wherein ψ is a porosity of the preform, μ a fluid viscosity of the resin, ΔT=tr−tr−1, whereby is acquired the permeability of the resin corresponding to the time point ti on the plane in the training process; Step 2G: repeating Steps 2A-2F to acquire the permeability, the wave front position at the current time point and the wave front position at the next time point of the resin for a plurality of time points in the training process; Step 3: defining the initial filling pressure, the permeability and the wave front position at the current time point as an input, and defining the wave front position at the next time point as an output; using a data mining technology to establish a prediction-control model involving the input and the output, which is expressed as
xi+1,j=f(xi,j,Po,Ki,j)wherein Po is the initial filling pressure, xi,j the wave front position at the current time point, xi+1,j the wave front position at the next time point, and the permeability; Step 4: performing a process to be controlled, which uses Steps 2A-2F to acquire the permeability at the current time point and the wave front position at the current time point, using an optimization algorithm to select at least one candidate filling pressure, and substituting the candidate filling pressure, the permeability at the current time point and the wave front position at the current time point into the prediction-control model, which is established in Step 3, to acquire at least one candidate wave front position at the next time point corresponding to the candidate filling pressure; Step 5: comparing the candidate wave front positions at the next time point, which are acquired in Step 4, with the expected wave front position at the next time point to find out the candidate wave front position at the next time point, which is nearest to the expected wave front position at the next time point, and using the result to trace back the corresponding candidate filling pressure; Step 6: transmitting the candidate filling pressure to the resin supply unit, and using the candidate filling pressure as the filling pressure of the next time point to make the resin flow to the expected position of the wave front of the next time point.