I explain step by step how to set molding conditions with quality engineering(QE). The story this time is the procedure before the experiment using the orthogonal table. In QE, the ideas and policies up to this point are more important than the orthogonal table experiment. If you do not have the ideal image and ideal, nothing will work well. Then, please see the following materials.
Click here → QE7
p.1: When setting molding conditions, engineers tend to list only molding process parameters and perform experiments. Are the characteristics of the raw materials used for molding constant? If there is a process for jointing parts and a heat treatment process after molding, is that process not affected? Are there any problems with the environment for transportation and storage to users and the usage environment for users? In QE, we list up the items that we can control (control factors) and those that we cannot control (noise factors). The control factor applies not only to the molding process but also to the parameters we can control. In QE, defining what we can control is called "system selection".
p.2: You can organize system by creating a "process function expansion table" in Excel. The leftmost column is the product manufacturing flow. After the raw materials are obtained, they are molded, jointed and heat treated into products. After that, the flow is to transport and ship to the customer. The second column specifies whether the system is divided and applied. The third right column is the item. What do we select the basic function and input / output during evaluation? What do we set the control and noise factors? Enter the optimum condition SN ratio, worst condition SN ratio, its estimated gain, and gain after confirmation experiment in the cell below. "L18-1" is used as the result of the first orthogonal table experiment. This table is written only for molding process, so it is not a good example. Actually, it is said that it is preferable to pick up the parameters little by little from each process of molding, welding and heat treatment.
p.3: Select the “control factor” that we can control. Only the molding conditions are picked up this time, but when actually performing it, please add the control factors for the front and rear processes as described above. This time, since it is an orthogonal table experiment of L18 (performing 18 combinations of experiments), factor A can be set to 2 levels and factors B to H can be set to 3 levels. Please set the level as wide as possible to the extent that the molding machine does not break. Previously, a molding engineer set the level, but since he empirically know the range in which molding is possible, the range was set narrow for safety. Please avoid this because you cannot evaluated the difference in parameters in this case.
p.4: Setting the noise factors is more important than setting the control factors. Again, please consider the noise factors (a worst case conditions ) that more than half of the 18 lines will be defective. For example, during transportation in midsummer, being placed in a car that exceeds 60 ° C, negative pressure during transportation by airplane, or the case raw materials manufacturers mix with different degree of polymerization. You may give the sample collectively the factors that increase the output more than ideal and the factors that decrease the output. Additioning up several errors is called "compounded noise factor". This time, the error factor is set to the raw material with extremely different resin flowability (MFR:Melt Mass-Flow Rate).