Research On Flow Behaviors And Mechanisms Of PLLA And PCL Nano-Composite Melts

Rheological behaviors in polymer processing have important effects on processingprocedures and mechanical properties of the final products. The thorough knowledge of avariety of rheological prope rties of polymer materials can effectively optimize productionprocess and improve prod ucts quality, there is important theoretical significance andapp lication value.With excellent biocompatibility and biodegradability, Poly（lactic acid）（PLA） graduallybecome the research hotspo t in the field of biode gradable materials. The sources of PLAdistributed widely, which can gain from grain. However fragility and thermal stability of PLAis very poor, and these unfavorable factors severely restrict the application of PLA. Therefore,in order to obtain the PLA composite which have a good performance, in this paper, thePLLA/na no-CaCO₃,PCL/nano-CaCO₃and PLLA/PCL/nano-CaCO₃composites wasprepared, and the shear flow characteristics parameters of the composites were measuredthrough a capillary rheometer. We analyze melt shear flow behavior and mechanism of thecomposites.The results are following.The entrance of the pressure loss （ΔPen） of PLLA/nano-CaCO₃and PCL/nano-CaCO₃composite melts increase nonlinearly with the increase of shear stress at the channel wall （τw）.The ΔPen of the former vary irregularly with the weight fraction of nano-CaCO₃（Φf）, andΔPen reach the minimum values at Φf=2%; The ΔPen of the later decrease with the increaseof Φf. Melt shear flow curve of PLLA/nano-CaCO₃, PCL/nano-CaCO₃andPLLA/PCL/nano-CaCO₃composites are basically linear, and the relationship between shearstress at the channel wall and the apparent shear rates can be represented by the power law.Melt apparent shear viscosity of the three compo sites decrease with the increase oftemperature and shear rate, and the relationship between melt shear viscosity and temperaturecan expressed by Arrhenius equation. Melt apparent shear viscosity of PLLA/nano-CaCO₃composite is lower than pure PLLA resin, and reach the minimum values at Φf=1%; Meltapparent shear viscosity of PCL/nano-CaCO₃composite increase slightly than pure PCL resin.The relationship of melt apparent shear viscosity of PLLA/PCL/nano-CaCO₃composite andthe weight fraction of PCL is nonlinear, melt apparent shear viscosity increase with the increase of weight fraction of PCL.The BP and PSO-BP neural network model between the melt apparent shear viscosity ofPLLA/PCL/nano-CaCO₃composite and melt temperature, shear rate and the weight fractionof the PCL were established based o n the experimental study. The result showed that accuracyof three layer standard BP neural network model trained by ordinary gradient descent methodwas not very good. The BP neural network model trained by particle swarm optimizationalgorithm had high accuracy, a nd relative error between the fitted values, predicted values andthe experimental values all ranged from0%to4%. It indicated that the PSO-BP neuralnetwork model was logical and feasible.The experiment research results and PSO-BP neural network model established in thispaper better elaborated the general flow law of PLLA composite melts. It provided a usefulbasis for forming of PLLA products, and was helpful to enrich the theory of rheology ofpolymer materials.