What are the hidden risks of tool transfers? Part 2: Assessing risk in part design

Believe it or not, the design of a part has an important impact on a tool transfer.

It’s important when bringing a new part from the design stage to the first part out of the machine that the right steps are taken. I can remember a few troubled molds that had very small processing windows and could only run in one machine. This happened because the designer missed some of the crucial elements in the process of part design. The one thing you must understand is that the customer doesn’t always know what’s best. In most cases, what is a functioning part is not always one that we as molders can manufacture very easily. 



The first task in assessing risk in your part design is understanding how mold details affect the plastic. Take, for instance, how molecules react around these details. Let’s break it down a little further. In order to get the plastic into the mold, we need it to flow from the injection unit to the cavity. The melt takes whatever route we provide in the mold/part details to get into the cavity. Since plastic is a non-Newtonian fluid, we know that we can change the viscosity 100 times more through flow rate than anything else. Also, since plastic is compressive (and not hydrostatic), it does not transmit energy very well. I like to relate this to a sponge, because if you apply pressure to the center of the sponge, the energy does not transmit equally to the sides. Figure 1 shows how the molecular chain reacts during flow.

Image A represents the polymer chain at its most relaxed state, without flow. During this time, the polymer chain is coiled up onto itself. The viscosity is at it highest in this state. When the viscosity is high, the plastic is more compressible and it transmits energy poorly.



Image B represents the polymer chain during the molecules’ normal flow. The polymers align themselves, allowing better flow and better energy transmission. The viscosity is lower, giving molders better control over the flow portion of the process.

Image C shows how the polymer chains react with a fast injection rate. Here, viscosity is very low and the material flows easily. However, it does pick up some frictional heat. Imagine holding the ends of a rubber band in your hands and place the center of it between your lips. Now, really fast, stretch and contract it. Repeat this several times without stopping. What do you feel? Heat is generated, much the same way plastic reacts when you stretch and contract the chains. The next step will shed some light on the subject.



In determining risk of your parts, you need to think about how you will be treating the plastic in the cavities. Take the following steps in “risking” your part design.

Step 1: Sharp corners
Sharp corners can cause a few issues in the final product. Stress in the corners causes a weakened condition, burns, sinks, and shrinkage, especially in thick-walled parts. One of the jobs I faced was a five-sided box, about 2 inches high, 1 inch wide, and 1.5 inches deep. The customer had moved this job to multiple companies because the molders failed to hold a crazy specification on warp: less than 0.002 inch using PET. The mold was a hot runner, low-voltage, internally heated manifold. We tried everything under the sun, including making cores that were bowed in the opposite direction. The only thing I fell short on was looking at the part’s sharp corners. This mold could only run in “press 6,” since it was the newer machine. We had some production runs that went very smoothly and others that were a nightmare. It wasn’t until I took a class at RJG 10 years ago that the light finally came on and explained the problem: The sharper the corners, the more stress concentration there is. 



Let’s take a closer look at the sharp corners shown in Figure 2. You can see how three different viscosities of material are flowing through the cavity. All three react differently with flow and packing, not to mention the shear heat the polymer picks up around the corner. What we know about this scenario is that the material flowing around the corner is not consistent.

Let’s add another concern to this—cooling. We have a core that shapes the interior of our part and a cavity that shapes the outside. The problem lies in the corner’s cooling capabilities. You must consider where the heat is being generated. The inside corner has only one cooling capability, where the largest content of BTUs is being generated, while the cavity block has four cooling capabilities, where the least amount of BTUs is being generated. We have now encouraged uneven cooling, creating a greater variable, as well as a variable shrink rate. Inside sharp corners take longer to cool, while the outside cools very quickly. This causes a conflict in cooling rates, and either will void in the center wall section or sink on the outside of the part—or both. 



To understand the full impact of this problem, we must understand what most process technicians would do to correct the sink and void problem. Typically, when these defects detected, the process tech raises the pack and hold pressures, maybe even profiling them to iron out the quality defect. The problem is, we have now increased the density of the plastic in that area, as well as in others. When the density goes up, the shrink rate goes down. It shrinks less, so the part is too big. Pressure losses across a part can be great, and therefore higher pack and hold pressure is not the same everywhere in the part. Pressure is typically highest near the gate and lowest farthest from the gate; it is not linear. Many times, the approach to this is dictated by the talents or knowledge of the process tech.

The problems escalate when we try to transfer this mold to another machine. If the waterlines are connected differently, or the water flow varies, you will not make the same part. If your machine is not tuned the same as the first machine, in this case for injection speed, you will change the shear rates and the distribution of the plastic when injected. If you do not posses the same talents that were used in tweaking the process to get the approved part, you may not make the same part.



If you are making a noncritical dimensional part, this may not be a big deal to you. However, in my case, it was huge. From a scale of 1-5, 1 being the best situation and 5 being the worst, I would rate this corner a 5 (see Figure 3).



Step 2: Draft
I cannot tell you how many times I have run across transfer tools, or even new molds we built, where the customer asked for no draft or even a reversed draft. You and I know that plastic needs some draft to come off the cores. I had a mold making cylinders that were 8.00 inches long and 3.00 inches in diameter, with no draft. The material was a wide-spec polypropylene. The designer went with the customer’s request to have no draft on the part. No matter how fast or how slow we ran the mold, the part would just fold inside-out. The fix? We had to make new cores, at close to $5000 each, with draft on them. We added 2° draft and explained to the customer that this was needed in order to manufacture the part.

Another thing people seem to overlook on the jobs I have consulted with is the draft for the details or for additions and subtractions to the walls. The risk for this job was easily a 5. However, after we added the draft, the risk went down to a 3. There were still risks of sticking during startup, due to only having 2° of draft. But since we had to compromise because of the functioning aspect of the part, that was all we could do (see Figure 4).



Step 3: Nonuniform walls
It is important to verify that you have nominal wall thicknesses across the whole part. If you have a solid model program, it would be easy to draw a circle within the wall sections where the circle would make a two-point contact and drag it around the part geometry to make sure the walls are consistent. In Figure 5, you can see how, in the thicker sections of the walls, you would have to draw a larger circle in order to make the same contact as the thinner walls. In these sections, you may have issues packing out sinks. You may get voids, shrinkage, and many other problems. 



The example in Figure 5 is a five-sided box with all of the problems you may face. This box is a housing for a signal relay that goes into a truck.

Problems:
1. Sharp corners

2. Thick-to-thin sections
3. Thick-walled additions



If I rated this part as it stands, I would have to rate it a 5.



Step 4: Additions and subtractions
Additions and subtractions should be considered throughout the part design phase, as there are rules that need to be followed. In Figure 6 we will talk about three major rules that should be considered during the additions phase.



Figure 6 shows how the rib thickness should be no thicker than half the wall thickness (6A). I once ran a battery housing made of a fire-retardant Cycoloy material. With ribs that were more than half the thickness of the part’s main wall, we were unable to pack out the sink. I remember trying every trick in the book to get rid of these sinks. Finally, the customer did allow the sink and we were able to run the job. However, every once in a while the sink would get worse due to viscosity shifts in the material.



The diagram also shows that ribs should be placed at a distance that’s a minimum of two times the part’s main wall thickness (6B). If you bring the ribs closer than that, there is a chance of getting defects like sinks and voids. I have even run across issues of sticking on the core, since it is harder to add sufficient draft on these ribs when they are too close.  



Any time I get a call from a client about short ribs, the first questions I ask are, what is the melt index of the material? What is the thickness of the part’s main feeding wall (6C)? And how high is the rib? Almost every time, the rib is much more than three times the thickness of the part wall. Remember, when we close the mold, air is occupying the space between the cavity and the core. Air is also occupying the space in the rib detail as well. We can add venting on the parting lines and in knockouts, and possibly in mold details, but it is very hard to vent ribs. Also, plastic will always take the path of least resistance.

Now, let’s look at the five-sided part in a solid model for the truck turn signal relay (Figure 7). My risk rating of this part design before changes were made is shown in Figure 8.



Let’s sum up the scores of our examples. Out of a total of 35 points for a total high-risk situation, we received a score of 27 points for the part design. To me, this job would not be accepted through the design phase as the score is pretty high. Typically, anything higher than 15 can be in question; however, I do have customers who are even more critical than I am. After all, we don’t get paid to make constant tool adjustments.



Sharp corners    5

Draft    5

Nonuniform walls    5

Additions and subtractions    12

Total    27 points



After modifying the design of the five-sided box, we can create a successful part design (Figure 8, after) that will allow the processor to have a greater processing window.



So, how does this help us? We all have been in situations of tool modifications because critical elements were overlooked up front. Some modifications may have been small while others may have been very costly. If we can think smarter up front in the part design phase, and think about it in a sense of the plastic’s point of view, we can eliminate those costly mistakes. Let’s put this a different way: How costly is it to change a drawing vs. cutting or replacing steel?

Missed Part 1 of this series? Find it here.

And there’s more: Part 2 of “Coordinating an acceptable transfer” covers the toolroom manager’s responsibilities.

Author Dan Clark (dan.clark@rjginc.com) is a Consultant/Trainer with Scientific Molding Implementation Specialists RJG Inc.

More on this topic:
What are the hidden risks of tool transfers? Part 1
What are the hidden risks of tool transfers? Part 3: Noninstrumented tool transfers