微細發泡射出成型發泡行為探討及模流分析與實際成型驗證

Abstract

近年來世界朝向減塑的方向邁進，而發泡製品為減少塑料使用的極佳方案，其優異的機械性質，以及具有絕緣、絕熱、減震、隔音等特性，可廣泛應用於工業界中;以汽車產業為例，汽車保險桿與車門飾板皆是由微細發泡射出成型加工而成且可有效減低重量與塑料使用，然而發泡製品通常為大型部件，為防範氣泡數量及尺寸不均導致其機械性能下降等成型缺陷，需有良好的數值模型與CAE模流分析軟體於成型前推估泡體結構，事先了解其製品可能存在的缺陷。 發泡行為動態且複雜，過往學者皆是以視覺化設備觀察氣泡的細微變化，以及微細發泡射出成型參數對於泡體結構的影響，本研究擬將過往學者視覺化研究成果比對現今模流分析軟體分析結果，針對高壓射出發泡成型與抽芯技術中，成型參數對於泡體結構之影響，比較模擬分析結果中氣泡成長動態行為是否與視覺化實驗結果相符。發現高壓射出發泡成型模擬分析結果與過往視覺化實驗結果相符，成型過程中，氣泡動態行為也與視覺化實驗觀察結果相近;在抽芯技術中，儘管模擬分析模內壓力與視覺化實驗結果相近，但模擬仍無法反應壓力梯度對於氣泡數量密度的影響，與視覺化實驗結果不符。 後續實際成型驗證實驗主要透過田口實驗搭配L9直交表同時進行模擬分析與實際成型實驗，比對模擬分析與實際成型驗證結果，評估模擬軟體是否能有效判別於高壓射出發泡成型以及抽芯技術中，重要之成型參數以及每一成型參數的貢獻度，是否模擬能助於開發者找到最佳化成型參數。結果顯示在高壓射出發泡成型中，模擬可有效預估成型參數對氣泡數量密度的影響，但未能預估對氣泡尺寸的影響，實際成型驗證，試片氣泡尺寸不均，而模擬則假設氣泡以球對稱的方式於模穴內均勻成長，與實際成型狀況不符;在抽芯技術實驗結果則與視覺化比對結果一致，模擬分析仍無法分析抽芯技術，本研究認為抽芯時模內已有膚層產生，因此導致模內壓力梯度變化與熱傳效應對於氣泡結構的影響更加難以預估。

In recent years, the world is demanding the manufacturer to reduce plastics. As one of the best solution, foamed products reduce use of plastics, while exhibiting excellent mechanical properties, electrical insulation, heat insulation, shock absorption and sound insulation. However, all of the properties mentioned above were related to the structure of bubbles, which requires comprehensive numerical models to estimate the bubble structure and the product characteristic before forming. This study is devoted to the validation of the numerical CAE software of Moldex3D that is mainly divided into two sections : the comparison of the simulation results with the visualization in literature, and with the actual forming experiments. The former study was conducted to investigate the effect of molding parameters on the cell structure in high-pressure injection molding and core-back technology. In the high-pressure injection molding, the simulation results were consistent with the visualization results, and the bubble dynamic behaviors were also in agreement with the visualization results.Moldex3D can effectively identify the influence of molding parameters on the structure of the bubble. But in the core-back injection molding, the CAE software obviously can’t determine the final structure of the bubbles. The reason is that the pressure gradient is non-isotropic and the heat transfer effect in the mold is more complicated because of the single-mold directional movement of the core-back technique and the skin layer generated during the core-back. The latter section was the experimental validation of simulation using Moldex3D. The comparison is evaluated whether the simulation software can effectively identify key molding parameters and the contributions of these parameters in high pressure injection molding and core-back by Taguchi's method with L9 table. In the high pressure injection molding, the simulation can effectively evaluate every molding parameter’s influence to the cell density, but is unable to determine its diameter. The cell size is not uniform in actual injection molding, while the CAE software assumes the bubbles to grow in a spherically symmetric manner. In the core-back experimental validation, the CAE software can’t evaluate the foaming behavior in core-back. The reason is that the heat transfer effect in the mold is more complicated because of the single-mold directional movement of the core-back technique and the skin layer generated during the core-back. It caused numerical model unable to evaluate the foaming behavior in core-back.