Join Suhas Kulkarni of Fimmtech, as he introduces the concept of Aesthetic and Dimensional Process Windows.
Results from the Design of Experiments provide process parameters that will yield an acceptable process to meet the quality requirements. However, this process that yields dimensionally acceptable parts may not be robust, since it may be in a cosmetically unacceptable area with possibilities of defects such as short shots or flash. The talk focused on this aspect of the process qualification process.
Q: Does your software provide finding the right material for a particular application?
A: No. Use the Prospector plastics search engine from IDES for this.
Q: Why should viscosity curve study be not performed for a hot runner mold?
A: You can perform a viscosity curve for a hot runner mold. There are problems associated with performing a viscosity curve study on insert molded parts (possible displacement of inserts)and cosmetic parts (splay, gate blush).
Q: How do we verify if the effect of a factor on the output is statistically significant?
A: Run at least 30 parts at one process and measure the statistical distribution. Change the factor to the next desired level and mold another 30 parts and measure the statistical distribution. Perform an ANOVA.
Q: The question if the method "robust processing" will work for loaded resins, the reason for the question is your viscosity changes by adding colorant, silicone antioxidants etc... you really do not have truly recommended temperature sheet...
A: That is correct. There is a definite effect of additives such as colorants on the processing. Unfortunately, the material suppliers do not do a good job giving you this information. However, in their defense, colorants can be obtained from any source and added to their resins. This would be difficult to control and characterize. Unfortunately, you must modify the settings using past experience.
Q: Does the method work for loaded resins for example Delrin, Colorant and silicon, antioxidant etc?
A: If you are referring to the mold qualification method. Yes.
Q: Is DOE result is machine specific? Or part specific?
A: DOE is part and mold specific. If you are not using cavity pressure sensors, then it is also machine specific.
Q: What kind of problem would trigger you to try profiled injection speed.
A: A part that is not designed for even flow from the gate to the end of fill. For example, Gate-Thin section – Thick section – Thin section.
Q: What is the maximum mold imbalance allowed on small parts?
A: There is no number that one can place of cavity imbalance. If there is an imbalance, record it and then fill the mold with pack and hold pressure. Check the process window. If the process window is wide enough, then measure the parts. If the parts are in spec at the center of the process window, you do not need any further adjustment.
However, if the process window is small and you can unable to make good parts in spec, the mold must be fixed. An example of a small process window is having flash on some parts and having shorts on others. On wide tolerance parts and molds with a large process window, you can have large cavity imbalances and the opposite is true in case of tight tolerance parts and molds with a narrow process window.
Q: Has this software been validated?
A: Not yet. This will be done some time in 2009 to early 2010
Q: What does decoupled molding eliminate as far as the DOE/Scientific Molding?
A: Nothing. Decoupled Molding only helps in telling you where to transfer into hold. The point of transfer is extremely important and using this technique is therefore a must.
Q: Can you use the machine outputs i.e. peak pressure to predict a dimension?
A: There is a theoretical relationship between machine outputs and dimensions. However, the resolution of measurements of the machine outputs is not enough to perform such a co-relation. Cushion value is probably the parameter that best correlates to the part dimension. But the resolution of cushion values on the machines are not enough. A change in 50 psi could affect the part dimension, however this may not show up on as a change in the cushion value and/or may be within the natural variation of the machine. Peak pressure is not a good indicator. The best output would be with the help of a cavity pressure sensor.
Q: Do you suggest drying of a glass filled polypropylene resin?
A: Surface drying only – 100 deg F for short times is enough.
Q: What are the exceptions with insert molded parts and viscosity curve. Faster speeds refers to injection speeds?
A: Faster speeds refers to injection speeds. Refer to Answer above.
Q: The limits on the process should depend on the part and mold design, so that we do not overpack the mold and/or break slender cores. Who determines the limits and based on what ?
A: The process engineer will have the best knowledge on this. However, this must be considered during mold and part design. The part designer is responsible to make sure that the design lends itself to the molding process. There are some rules of thumb and experienced mold makers and process engineers will be able to guide this process.
Q: Based on the DOE, am I going to get parts on the low side of the nominal? What about Cpk values? What should I do: change the nominal or cut steel?
A: For low Cpk values, you should change the nominal and/or open up the tolerances. If this is not possible, the mold steel must be modified.
Q: Have you ever used doe to set up the correct back pressure and decompression to give you shot to shot consistency
A: Back Pressure – Yes, but no experience with decompression. The best back pressure is when you have a consistent screw recovery time. Use only as much back pressure as you need. You also need good back pressure to take care of cosmetic issues such as splay.
Q: If we are monitoring cavity pressure do we still need to put alarm limits on other machine variables as well.
Monitoring of the cavity pressure is the best way to monitor part quality. If you set alarm limits on cavity pressure, then you do not need to set limits on the holding pressure on the machine. However, you will need to set limits on the other variables that are not related to the pressure. E.g. Temperature.
Share This Page
•
• Free Access to 80,000 Datasheets