Mold Building Without Rule Breaking

Push the mold in the beginning of the build process to discover any weaknesses before it gets to the molder.
#fiveaxis #casestudy


Facebook Share Icon LinkedIn Share Icon Twitter Share Icon Share by EMail icon Print Icon

Have you ever been standing by a machine processing on a mold for flash, shorts, dimension, etc? Does your company have technicians who have that sole responsibility? They are the ones you call when no one else can get it running. Have you noticed that that particular mold seems to need your attention every time it runs and no one thing seems to work in every situation? Have you noticed that the parts you make off of that mold seem to always get rejected both internally and externally? Why can’t we seem to ship that part on time? Why does this happen? Why is it so pervasive in the injection molding industry today?

The answer is simple: large viscosity shift, and large static and dynamic pressure losses. It is inherent in polymer processing. Some types of molding are more sensitive to this than others. Injection molding is probably the most sensitive to these variables than any other type of molding process.

The mold designer and builder are responsible for producing a mold that will absorb the variation it will see in production. Simple things like getting the part out the first time you eject are so commonly overlooked that there is a huge amount of money lost in cycle time loss, and mold wear and tear.

Rigorous mold tryout is about building the mold so the molder does not have to break the rules to get it running, and so that any molder can start and run production from the mold.


Rigorous Mold Tryout

1. Flow Rate
First, flow rate must be established. How is this usually done? Is it even studied at all? The answer is “It isn’t studied.” The sad fact is that it is so easy to do. The injection molding machine itself is the most available and easiest rheometer there is to use.

Start with at least 90 percent of the machine’s injection rate. Then decrease the fill speed and document the viscosity change. What you end up with is a shear rate curve (based on shear during fill). At a certain point the viscosity takes a sharp turn upward. This is where you do not want to be. At these flow rates the process becomes super sensitive to machine fill speed changes. When the viscosity changes it is amplified in pressure loss more than if you were on the correct side of the shear rate curve.

2. Flash
Knowing this, what should you do if you are flashing the mold just because you are filling it fast? (Of course, you cannot hit the end of the cavity as that will result in flash). If the mold flashes simply because the viscosity is lower due to the high shear rates, it is a mold problem; not a process problem. You fix the flash. Failure to do so results in process failures later down the line because the flash never gets better. The wizard on the floor then does the best finessing he can do only to come back after the viscosity has changed and find that he has made shorts. Or the flash suddenly got worse. The typical situation is that they think they have made it go away only to find out that a couple of bad parts got to the customer. Wouldn’t it be better to push the mold in the beginning and know its weaknesses? If you correct them up front, they won’t come back.

3. Pressure
What if when you filled the mold your injection pressure reached maximum? This is the worst situation you could be in because you are no longer in control of flow rate. Flow rate happens to be the primary process control variable. If this happens you either find where the loss is maximum inside the mold or you move to a machine with a higher intensification ratio. Slowing down may be an option so long as it doesn’t put you on the wrong side of the shear rate curve. If the viscosity increases by ways other than flow rate, the machine will need to use more pressure so you have to make sure you are not close to the pressure difference between the flow control valve and the pump.

As for pressure, there must be adequate pressure to offset the effects of static pressure loss during pack. The issue with this is if the tool is not instrumented with cavity pressure sensors, it is impossible to truly quantify pressure losses. It is considered negligence to not look inside the cavity for this information. Yes, you can pack the part to a small short and measure the loss from the injection pressure.

For example, if a small short occurs at 1000 psi, it is safe to assume that there is a 1000 psi loss from injection to end of cavity, but if you add another 1000, it does not mean there is 1000 at the end of cavity. Why? Because of static pressure loss. The long polymer molecules do not transmit hydrostatic pressure very well. There is a certain point in pressure loss where things really get critical.

For example, if there is 12000 psi at the injection unit and 4000 at the end of fill, this is an 8000 psi loss. If the viscosity is increased by 10 percent (not uncommon at all) then you could take away 10 percent from the pressure loss. This is 800 psi. Subtract that from the original 4000 and your new end of fill pressure just dropped to 3200 psi.

This is a 20 percent change in pressure from a 10 percent change in viscosity. What if your processing window was between 4000 and 3200 psi? You have a problem. So, during sampling, you want to know those pressure losses and try to minimize them by finding where the most loss is occurring. The critical point is about 50 percent of the packing pressure at the end of fill location. For example, if our end of fill pressure was increased by reducing pressure loss to 6000 then the change would be 10 percent at the end of fill.

4. Temperature, Cooling Rate and Time
How do you optimize plastic temperature, cooling rate and time? You start with the manufacturer’s minimum and maximum limits and set them right in the middle. From there you start the process and try not to have to manipulate this variable. If it becomes necessary to change them, stay within the limits of the supplier’s recommendations. If you have to vary outside that range, identify what the limiting factor is and attempt to correct it. Because temperature changes affect cooling rate and the time needed to cool, try valiantly to not change this variable.

At this point the cooling time is set at whatever it takes to make the part rigid enough to be ejected; however, oftentimes there are other issues that drive the cycle up. It has been found that to meet quoted cycles breaking the rules is often required. It is a very last resort to be out of the specifications; however, it does occasionally happen. The focus here is to reduce normal variation in all phases of the cycle.

You may ask the question, “Does it really matter if I am out of the supplier’s specifications as long as I am making an acceptable part?” Yes and no. If you have a problem later down the line and it is big enough to get the customer concerned, the first thing that happens is the material supplier shows up and points out what you are doing wrong. Even if it is explained why you had to go around the problem, it won’t be good enough in the customer’s eyes. Surprisingly, it is probably something they felt they couldn’t live with that caused it to happen that way in the first place. Hence, it is so important to be able to explain the potential problems with any adverse action to the process and get an upfront agreement (in writing of course) that it could be an issue later down the line. Wouldn’t it be better to not let it get to that point?


Compromised Process Control

When any of the four variables is compromised the tool should be flagged as incapable not due to statistical measures, but due to the fact that process control has been compromised. In other words, if you have to pack the part with just enough pressure to keep from over-packing it (a tooling or design issue) and those two pressures are within a narrow range from each other, it is certain that you will make a short and once it is made it is almost impossible to keep it from getting out the door.

If you have to fill the mold at an extremely slow rate, it is certain that you will make vastly different parts if not shorts and flash along with it. This depends on the pressure loss, but if your pressure loss is large and you are on the wrong side of the shear rate curve, you will get a short or flash. However, the short is more likely and flash would depend on the process window.

Now it is time to explain to the customer that it is highly likely that a reject part will be made and shipped. It isn’t that you want to or don’t care; it’s just inevitable if the process is weak. Of course, they don’t like to hear that and they hate it even more when it happens, but when you didn’t build the mold and weren’t part of the qualifying process, it is up to you to explain that the product they thought they were getting isn’t going to pass the test.


Forces of Change

Lately, the trend has been to go overseas. Get the molds built faster and cheaper and that is exactly what is happening. Molds are put through rigorous mold tryout and then it’s discovered that you have weeks and sometimes months of work to do to make the tool capable of withstanding normal variation.

In the case of transfer tooling, there has been success with avoiding rejects because incapability is documented upfront. When the reject happens they can’t reject you because you restated the terms based on the mold’s capability. Of course alternatives are offered to get out of the situation, but it doesn’t go over well when the cost to get a tool into shape is $5 to $10,000. So instead of fixing the tool they would rather keep getting shorts.

This type of situation is so pervasive because that’s the way it’s always been done. Why should we change? What is forcing the changes now are customers’ new demands for “0” rejects. No more ppm acceptance criteria. Just don’t ship one ever. If you do they will rip into you and put you on a ridiculous sorting program. Is it possible? It is if you don’t break the rules and have a robust tool.

So, the industry is changing and it is forcing the molder to be extra careful when bringing in new tooling. Unfortunately, the method of making a few parts and throwing the mold over the wall to the molder is still very common. It isn’t working and the mold builders need to get on board with the concept.