John Bozzelli is the founder of Injection Molding Solutions (Scientific Molding) in Midland, Mich., a provider of training and consulting services to injection molders, including LIMS and other specialties. Email email@example.com or visit scientificmolding.com.
Screws and Pellets: One Size Does Not Fit All
My last few columns have dealt with various issues pertaining to the nozzle and barrel. I’ve discussed the important roles these components play in providing the mold with material that is uniform in both temperature and consistency. Here I’d like to discuss the role of the screw in the process, starting with the basics of screw design and moving into often overlooked issues such as pellet size and shape and the melting behavior of various materials.
IN THE ZONE
Screws have three zones: one for feeding; a second for transitioning solids into melt, also known as the compression or melting zone; and a third zone for metering.
The function of the feed zone is to convey unmelted plastic to the transition zone. Usually the feed zone accounts for about half the total length of the screw, and the flights in this section are deep. The feed section of the screw is not designed to melt the plastic, but to auger the granules forward and prepare them for melting.
The granules are “gravity fed” through the feed throat, and there is lots of air between the granules in this part of the screw. This air, as well as volatiles such as residual moisture and light fractions of the polymer (and perhaps of the additives), must be vented through the feed throat as the material begins to heat in the rear zone.
So the feed throat is a de facto vent. Don’t look into the hopper or feed throat without eye protection and without giving some serious thought to the condition of the barrel. I have seen many molding plants with patched roofs, the result of hoppers shot through the ceiling. Also, recognize that this is one reason you want the feed throat to be warm; you don’t want to condense those volatiles.
The transition zone is the workhorse of the melting process. The deep flights in the feed zone transition or taper into the shallow flights of the metering zone, which accounts for about 25% of the screw length. This transition zone compresses the granules against the barrel wall, generating roughly 80% of the energy needed to melt the granules. Plastic sticks to the barrel wall, and as the screw rotates the flights wipe off a thin film to form a melt pool between the flights of the screw. This melt pool rolls in a spiral and grows larger as the resin progresses toward the metering zone.
The transition zone is supposed to melt the pellets completely, but this is often not the case. I have a quite a collection of parts containing partially melted pellets, as well as runners with unmelted granules stuck in the gates. Ever wonder why a part in a multi-cavity mold did not fill? Unmelted or partially melted pellets may have been the culprit.
If you are present when a screw is pulled from a barrel, I encourage you to take a close look at the flights in the transition and early stages of the metering zone. Nine times out of 10 you will see a carbon layer built up behind the flights. Often you will see unmelted granules, or even residue of colors that you ran weeks ago. This shows that all the “new” resin arriving in the transition zone is not pushing all of the “old” resin out in front of it. And it means that there are dead spots behind the flights where resin “hangs up.” It sits there partially melted and degrades to carbon.
This carbon on the flight is the root cause for the dreaded black specs or carbon showers that cause you endless hours of purging. And if you decide to bite the bullet and tear it down for cleaning, don’t use wire brushes and the like. For the screw to work properly it must be smooth and highly polished with no scratches or undercuts. The concept is for plastic to stick to the barrel and “slip” off the screw. Those who mold clear parts are especially sensitive to carbon, as it is a significant cause of rejects. Eliminating the dead spots on the screw will significantly reduce carbon formation.
The metering zone of the screw pumps plastic forward to form the shot and overcome the backpressure set by the operator. Normally it is not designed to finish melting or mixing the plastic. If unmelted or partially melted granules make it through the transition zone, the metering zone will not complete the melting process.
A PELLET IS A PELLET?—NO!
Now that we’ve gone over some screw design basics, what about the resin? Can the type of resin or shape or size of the pellets influence the melting process? Amorphous resins melt differently than semi-crystalline resins. Amorphous resins like polystyrene, polycarbonate, and ABS melt like butter—they soften and mush easily as they come up to temperature. Semi-crystalline resins like PE, PP, or nylon melt like ice. They stay hard up until their melting temperature, then melt. Most semi-crystalline resins require nearly twice as much energy to melt as do amorphous resins. Amorphous resins are somewhat forgiving in the melting process, while semi-crystalline resins present a difficult challenge to get them melted uniformly.
The size and shape of pellets, and the means by which they have been pelletized, also influence melting. My experience has been that a general-purpose screw will melt almost anything as long as the granules are the same size, shape, and cut. But how often is that the case? Problems develop whenever you have a mixture of granule sizes or shapes. Non-uniform regrind—especially fines, regrind with virgin, tiny color granules with larger virgin granules, cut strand mixed with underwater die-face cut beads—all wreak havoc with the melting process. They begin to melt at different spots along the screw, creating non-uniform pockets, partially melted granules, degradation, black/white specks etc.
Most molders do not have the luxury of running just one kind of granule, so what can be done? Start by working with competent screw designers—not just the sales person, but a technical person or engineer who can prove the design works and will stand behind the product. Get prints and review them before any steel is cut. A screw designed mainly for melting semi-crystalline resins should be capable of running amorphous resins, too.
Here are some more general rules of thumb: L/D ratios should be at least 20:1 or greater. Demand that flights have a large radius (more like a farmer’s plow) that eliminates the dead spot on the flights. Make sure the metal of construction is appropriate for your resins. Ensure your screws are polished and scratch-free and have reasonably sharp flight edges.
If working with clear materials, eliminate fines and make sure your screw is chemically resistant. In addition, inspect pellets for uniformity from the supplier. Use only 25% to 65% of the barrel capacity. Evaluate grinders for granule uniformity and minimum fines generation. If you do micromolding or run screws smaller than 25 mm diam., the criteria are even more stringent. I’ll get into that in another column.
Mold Filling Simulation: What, When, Why, How
Click Image to Enlarge
Mold-filling simulation and analysis can tell you a lot about how well and how fast a mold will fill under a given set of conditions. But the accuracy of results depends on many factors, including the accuracy of the material data and the experience of the analyst in plastics materials, molding, and tooling. (Photo: Autodesk Moldflow)
If a mold is having a problem making good parts, and a Pressure Loss Experiment shows results like these, flow analysis will have a better shot at providing a solution than you’ll get simply by enlarging the sprue and runner.
Accuracy of flow simulation depends partly on the quality of the material data manipulated by the program. If your material grade is not in the software database, using a supposedly “equivalent” material should be checked by comparing a simulation of that material in a known existing tool. (Photo: Moldex 3D)
Mold-filling analysis is the computer simulation of molten plastic flow into a given mold geometry. This could be a plastic part, hot or cold runner, gate, nozzle tip, sprue, rib, etc. For a mold, the flow is predicted to the point of 99% filled; packing is a separate simulation. With proper application by an experienced user, this flow simulation can predict—before the mold is built—how the flow front progresses through the flow path of nozzle, sprue, runner, gate, and part. This allows for possible prediction of gate location, fill time, flow restrictions, plastic pressure distribution, air entrapment, venting issues, temperatures, weld lines, and other concerns associated with plastic part production.
Possible is emphasized here because analysis of a virtual process in the computer is sometimes accurate and sometimes not. The software continues to improve, and more advanced programs attempt to simulate packing, warpage, and cooling. This article will focus on only the filling or flow simulation aspects. My perspective on flow simulation is that it is a type of computer-aided engineering (CAE), with the emphasis on aided. The computer can’t do it alone. The software helps an experienced user determine the fill pattern and associated phenomena. Computer software does not have the magic capability to provide good results if the person using it has little or no experience with plastics materials, processing, and tooling.
WHEN TO DO FILLING ANALYSIS
There are two main applications for mold-filling simulation. One is in optimizing a new mold design before steel is cut. The second is solving problems with an existing mold. Both can save lots of time and money.
Designing a new mold: Going from a part drawing to production involves hundreds, perhaps thousands of details. Anything you can do to catch errors in part and tool design upfront will pay huge dividends. Factor in the demand for ever-shorter lead times, ever-increasing part and tooling complexity, and the costs of molds, and it’s obvious that the stakes are high. With the right data (part geometry, material rheology, and processing setup), flow simulation can help avoid costly errors as development proceeds from art to part. The problem is that the simulation has to be done right to provide the cost benefit, and many are not getting the full benefit. My guess is that nearly 80% of all hot runners made by major vendors are designed with flow simulation. (But much of this information never reaches the molder. When buying a hot runner, be sure to ask for the estimated pressure loss, which you’ll need to make sure your machine has enough injection pressure.) My bet would be that about 50% of molds are built with flow analysis. Here are some issues to consider if you have a simulation done on a new mold:
1. Make a list of what is expected of the analysis and decide if the benefits justify the cost. Do you want iterations at different fill times, temperatures, and gate locations? Do you want to know the clamp-pressure requirement? (Note: When a mold has slides, flow simulation generally does not predict the forces projected against the clamp over the heel blocks.) Want to make sure you know weld-line locations, high-stress areas, venting issues, or pressure distribution? Do you need to make sure your equipment has the shot size and injection speed/pressure necessary to mold this part?
2. Make sure there is a commitment within your firm to evaluate and use the report. I have seen too many analyses that are ordered, paid for, and then totally ignored. Worse, it is exceptionally rare that anyone provides or explains the simulation information to the processor before he/she trials the new tool. (Keep in mind that the simulation was performed according to an assumed set of injection conditions. If you run the mold under different conditions, there may be little benefit gained from the simulation.) You should compare the predicted results with actual data once the mold is built. Don’t you want to know if you are getting the benefits of this expenditure or evaluate the capability of the analyst?
3. Be prepared for some hard compromises. The filling analysis may identify an ideal location for the gate from a processing viewpoint, but building a mold with that gate location may be impossible or extremely costly, or the gate location may not be not acceptable from a product performance or aesthetics viewpoint. The simulation analyst should have access to sufficient design information to be aware of such restrictions and should have access to the design team to be able to explore any flexibility allowed in part design and application.
4. Is there good rheology data for this resin? Flow simulation software requires certain material data to predict flow and cooling. If the software materials database does not include your particular formulation (including glass or filler content), using data on an “equivalent” material to perform the analysis may or may not provide accurate results.
5. Can you evaluate the filling analysis on a similar part? The wise user finds an existing mold with a few of the same features of the mold being analyzed (perhaps a similar nominal wall thickness) and does a flow analysis on this mold or section of the mold to show that the software’s predictions correlate with actual molding results on the existing mold. If the software under- or over-estimates fill times, pressures, etc., this data can be used to “correct” the flow analysis for the new mold, providing significantly more confidence in the results.
6. Mold filling analysis alone may not predict cosmetic problems. You may need additional packing, shrinkage, and warpage analysis to predict sinks and distortions.
7. Many simulation programs fall short in predicting the critical balance of filling in multi-cavity tools. It is an unfortunate fact that a “naturally balanced” or “geometrically balanced” runner system will not necessarily produce balanced filling of a multi-cavity tool. If you do not have balanced fill you will not have identical parts. (For details of this discussion see my column in March 2010. Unbalanced filling of supposedly balanced molds has also been explained in numerous articles and papers by Dr. John Beaumont of Beaumont Technology Inc.).
Fixing problems in an existing mold: If an existing mold is not producing good parts or has a high reject rate and an experimental “Pressure Loss” analysis has been done and shows a problem such as that indicated in the accompanying illustration, a flow analysis should be performed. (For a description of Pressure Loss Analysis, see my column in Dec. 2010.) This 20-minute experiment showed that the the sprue and runner were taking up nearly half of the total injection pressure to fill the mold. Obviously something is wrong. While most would simply open up the runner to reduce pressure loss, one actually needs to weigh both sides of the equation. If you open up the runner, pressure will drop to the 4th power relative to the change in diameter. However, the viscosity will increase because the shear rate decreases, and this complicates things significantly. You may win on pressure but pay the price with much stiffer resin. What is the best compromise? Flow analysis is ideal for this situation. First model the flow channel and part, then get the software to match the actual molding conditions, and then do the iterations on channel size to find the best compromise.
WHO SHOULD DO THE FLOW ANALYSIS?
Experience counts. Repeat that 10 times. The ideal analyst has years of experience, not only with the software, but also has shop-floor experience in processing, materials, tooling, and part design. Each of these play a role in flow analysis and affect the quality of the results. Finding someone with this diversity of experience is unfortunately rare, so look for a shop that uses a team of these players to do the flow analysis. Try to find a team that will actually test the results on an existing tool. Request, perhaps demand, that they show you a case history where experiments on the tool matched their predictions.
HOW OT GET GOOD RESULTS:
•Select a qualified team with appropriate experience to do the flow analysis.
•Require at least a 12-layer analysis on complex parts. (Simulation programs “slice” the melt in the flow path into multiple layers and calculate velocities, shear, temperature, etc. separately for each layer.)
•Define and then specify the results you want. No need to waste resources, time, and money on data you won’t use—or on an analysis that doesn’t tell you what you do need to know.
•Make sure you provide the analyst with the correct part/tool dimensions and the current design drawing. And keep the analyst posted on design changes. Also, check to make sure the tool is designed for the correct material shrinkage rate.
•Use a team approach. Involve design, material, tooling, and processing specialists to work with the analysis team.
•If the analyst is using “equivalent” resin data, require him/her to do a check on an existing tool. Volunteer to do this in your shop on a tool you can use as a standard. Actually I would do this for any resin, for what is at stake and the money involved, it is worth the double-check.
•Understand that if you do not request an analysis of the entire flow path, you’ll only get the pressure requirements for what is analyzed. You must add to that the pressure requirements of the nozzle, sprue, runner, etc. Too many processors have come to the unpleasant realization that their machine is pressure-limited. That is, the additional flow paths of sprue and runner maxed out the machine and now the part cannot be filled, or the process runs pressure-limited, providing poor consistency.
•Understand that balance of fill in multicavity molds may not be predicted accurately. Balance of filling multi-cavity molds is critical for identical parts. (See my column in March 2010.)
•Communicate the analysis results to those involved in bringing the tool into production. For example, the flow analysis will usually suggest the fill time for the part. My bet is that this critical number is not passed on to the processor 99% of time. If you do not shoot the tool with the suggested fill time you have essentially wasted your money on the flow analysis. Injection time establishes viscosity; a different fill time will often provide a different part!
In summary, mold-filling analysis is the right direction to go, but if you want to arrive at your destination—good parts and an efficient process—you’ll need to pay attention to all the details and select an experienced professional to do the analysis. Even then, the simulation accuracy may not be on target.
Short Shots Redux
In several columns I have addressed why you’d want to make a short shot on purpose. Doing so is an important aspect of practicing “Scientific Molding.” Among many reasons, deliberately making a short shot permits you to do a scientific molding viscosity curve. It will also prevent you from damaging the mold by overpacking if you set shot size incorrectly.
Based on the feedback I have received, it seems I have not communicated the procedure with enough clarity. I have suggested you make a shot size that is 95% to 99.9% of a full part. What exactly does this mean—95% full by weight, volume, size? The answer is volume—the part should be visibly short. You do not want the flow front to hit the end of the cavity. Why not? There are four reasons.
First, most of the time, and especially when you are working the viscosity-curve and mold-flow analysis, we are using the Hagen Poiseuille flow equation. For a round tube, the equation is:
Q = πr4∆P
Q is the volumetric flow rate.
π is the math constant 3.146.
r is the radius of the tube, gate, runner, or part.
μ is the dynamic viscosity.
L is the length of the tube, gate, runner, or part.
∆P is the pressure loss or drop for the plastic to flow through the flow path.
The key data needed to work the equation is the pressure drop—more specifically, the pressure loss from the beginning of flow (the nozzle) to the end of flow or the last area to fill. Without cavity-pressure sensing at the exact end of flow, which most molders don’t have, the only way we can find the pressure loss from the nozzle to the end of flow is to run a short shot.
With a short shot we know the pressure at the end of flow is zero. No plastic = no pressure. If the flow front hits the parting line, there is some pressure, but no way to tell how much, unless you are fortunate enough to have a cavity-pressure sensor at the short location.
We find the pressure drop for short shots by taking the pressure at transfer, as the machine transfers from first to second stage. This pressure is readily provided by most machines. This is the pressure in the nozzle at the moment the screw reaches the set transfer position. In a hydraulic machine, the hydraulic pressure at transfer must be converted to plastic pressure by multiplying the hydraulic pressure by the machine and barrel intensification ratio.
For example, if the pressure at transfer is 1675 psi hydraulic pressure and the barrel provides a 11.6:1 intensification ratio (it’s rarely 10:1 anymore), the plastic pressure in the nozzle is 19,430 psi. In an electric machine, the pressure at transfer would read 19,430 psi directly. For a short shot, this is the pressure drop or pressure loss to fill the part 98% full by volume. For this mold, we now know that the pressure drop to use in the Hagen Poiseuille equation is 19,430 psi, and this allows us to calculate the plastic’s viscosity. If you are surprised that we lost every single psi of the 19,430 psi injection pressure, it is common to see pressure drops in thin-wall applications over 30,000 psi.
The second reason for short shots is to check the results of a mold-flow analysis. While few molders check flow analyses, those that do should note that the part has to be short to compare what is predicted vs. what actually happens. For filling a mold, the flow analysis is done only to a short shot (then it is followed by packing analysis).
Third, you don’t want to smack the parting line with the plastic flow front during your tests, or in production. This way, we do not have a sudden pressure spike on the parting line that may cause flash and wear the parting line.
Fourth, if you are practicing scientific molding you are processing under velocity control conditions. That is, you have adequate pressure available on first-stage injection to ensure velocity control. If you have this extra pressure available and you begin to pack the mold under first-stage pressure, the machine may overpack a portion of the cavity.
Processing Tip of the Month: This short-shot pressure drop is also of interest in evaluating the difficulty of obtaining high Cpk or consistent parts. The higher this pressure drop, the greater the amplification of all the variables involved in molding. This pressure loss perhaps should be called a “Murphy Multiplier,” referring to “Murphy’s Law,” meaning that anything that can go wrong will go wrong. Anything you can do to minimize the pressure drop will provide for a more robust process. Dissecting how much pressure loss occurs in each of the components of the plastic’s flow path can help troubleshoot problematic molds.
Calculate Shot Size Vs. Barrel Capacity
Calculating shot volume to make sure your barrel has enough capacity may seem like a dull topic, but it will overcome the emotional experience that follows when you put a new mold into a machine and you find out there is not enough barrel capacity to make a full shot.
Now I’m sure this has never happened to you, but a few of us have been in this spot. So as not to emulate our mistakes, you need to do your homework before the machine is ordered. Not having a properly sized barrel for your shot is a show (and a part) stopper.
How do you figure out what barrel size you’ll need? If it is a new part and your mold hasn’t been built yet, running a mold-filling analysis will give you the volume of the part and runner. If your mold is in the building stage, your moldmaker may have calculated the volume of your shot—and make sure to also include the runner volume if it’s a cold-runner mold. If you already have the mold and you are purchasing a new machine, you have two sets of data: the part and runner weights, plus the shot size of the existing machine. With either part and runner weight, or the volume of the total shot, you have your starting point.
Let’s start with the scenario in which you know the part weight. I like to work in grams but most machine specifications are in ounces. If the part, runner (cold), and sprue together weigh 164 g, the equivalent is 182.2 cc of PP (at 0.90 g/cc density), or 6.16 oz (1 cc = 0.0338 oz). Assuming you have a barrel capacity of 8 oz, you might be tempted to conclude that you have plenty of shot volume. Unfortunately this is where things start to unravel.
First, what if the machine-capacity specification in ounces is for one material, often polystyrene, and you are running polypropylene? (Note: Many machine specs today quote barrel capacity both for PS and HDPE.) There is an important density difference between these resins. At room temperature the density of PS is 1.04 g/cm3 and PP has a room-temperature density 0.90 g/cm3. You might then think, “Okay, for my 6.16 oz of PP you can take the ratio of densities 1.04/0.90 or 1.16 and arrive at a required barrel capacity of 7.14 oz.” With an 8-oz barrel you’re still okay, right?
Unfortunately, this is when Mr. Murphy usually knocks with his lesson that what can go wrong usually does. He’ll remind you that these are room-temperature densities, and in molding we deal in melt-temperature conditions. Sorry, there’s another step of complexity: dealing with melt densities.
When melted, the polymer molecules are farther apart and density decreases. The problem is that melt-density data is harder to come by. It is not on many material specification sheets and you have to do a bit of research to find it. (Plastics Technology’s PLASPEC Global Material Datacenter is one place to look.)
For PS the melt density is 0.945 g/cc and for PP it is about 0.74 g/cc. This changes the ratio for our calculation; it is now 0.945/0.74 or 1.28. Now we need a minimum of 6.16 x 1.28 or 7.88 oz of shot capacity. We are still (just barely) within our 8-oz barrel capacity, but unfortunately Mr. Murphy is still lurking.
Consider our check valve: It leaks! Is it practical to try to use 98% of the barrel capacity (7.88/8.0 oz)? Not in my book, and you need some cushion and room for decompression—especially for hot runners. I recommend shot-sizes of 25% to 70% of barrel capacity. If we add all this up, plus a 10% safety factor for check-valve leakage, cushion, and decompression and to ensure uniformly melted plastic, we need a barrel with at least 11-12 oz capacity.
But we’re not done yet. We have to think about plasticating capacity. PP is a semi-crystalline resin and it melts differently, perhaps more stubbornly, than amorphous resins. Generally, semi-crystalline resins stay hard up to their melting point and often require twice as many BTUs to melt as amorphous resins. This double whammy makes semi-crystalline resins significantly harder to melt uniformly than amorphous ones. You don’t want to use too much of your barrel capacity, or you’d be running the risk of having unmelted or partially melted solids in the melt stream. Warpage, shrinkage, and physical properties would be significantly affected.
Once again, you do not want to underestimate this detail of properly calculating shot size vs. barrel capacity. Too much is at stake. Send me your email address and I’ll send you a spreadsheet to do this calculation. You plug in ounces or grams and resin melt density, then it calculates the suggested barrel capacity in ounces.
The next task will be to determine if your screw can melt this shot within the quoted cycle time. That’s for the next column.
Plasticating Rates: Your Profits Are at Stake
You know the pressure you’re under to provide a low part cost. To make matters worse, you often do not have enough time or data to do it right, rendering most quotes mere “guesstimates.”
I would love to see how your estimated or quoted cycle times compare with actual results. Every molder ought to have a spreadsheet ranking their jobs on quoted cycle time vs. actual (available free on my website, scientificmolding.com). It will tell you where you are losing money. If, for example, you are quoting based on a cycle time of 17.5 sec but are actually running parts at 19.3 sec, you have lost 41% of your profit margin (see illustration). One second can mean a gain or loss of $50,000/yr for some plants. Think about it.
To get the best cycle time you need the right machine, and I am appalled at how few molders take the time to specify this critical component of the process. In our last column we dealt with calculating shot size and shot capacity of an injection press. The calculation was not that simple, as machines are typically specified for the melt densities of polystyrene and you might be running something else. Still, you must do this calculation to find out if the barrel can handle the shot required.
But this calculation is only part of the story. Remember Murphy’s Law: Whatever can go wrong usually does. Well Mr. Murphy has more bullets for you to dodge. One of them is plasticating rate, or screw recovery time. Can the screw melt the shot in the time needed to obtain the quoted cycle time? If not, it’s your money out the window.
Machine suppliers usually provide data on plasticating or recovery rates. The specifications can be gram/sec, oz/sec, or lb/hr, but read the fine print. Because, once again, this data is usually presented for melting PS. And as we discussed last month, you have to do the conversion for your resin.
For the purposes of this column, let’s say our shot of parts, runner, and sprue (16-cavity, cold runner), is 224 g of PP. Our project engineer quoted a cycle of 20 sec, but marketing forced it down to 17.5. This is a cap mold, so fast cycles are demanded. The 17.5-sec. cycle time is the sum of:
1. Fill or injection to fill the part
2. Hold/pack time, packing the part to dimensions
3. Cooling time
4. Mold opening
6. Mold closing and lock-up.
So how much time is there for recovery or melting? Assigning some rough numbers for these functions might look something like the table above. It appears that we have about 5.2 sec to melt the 224-g shot. Not quite that much time, though, because within this cooling time, your machine has to be doing some other things:
•Screw delay or suckback before screw rotation. This is recommended to minimize wear on the screw tip, check ring, and screw drive motor. What happens if you do not have this delay? Ever wonder why non-return valves wear so fast or screw tips break? All you need is 0.1 to 0.3 sec of screw delay to allow the melt pressure to decay to a reasonable level.
•Plasticate or melt your next shot.
•Screw decompression or suckback after screw rotation to reduce drool.
•Idle time. Why? Because you cannot set plasticating time. You set screw rpm and backpressure, and plasticating time is a result. During production, screw rotate time will vary relative to the amount of colorant, size/type of granules, backpressure, and other process variables. So, you have to leave at least 0.2 to 2.0 sec of idle time so your cycle time is not dependent on plasticating time.
It’s apparent that you don’t really have 5.2 sec for plastication in our sample shot. It’s more like 4.9 sec. Now check the machine specs for “recovery rate” or plasticating capacity. Our specs say 428 lb/hr. It’s more relevant to view this as 1.9 oz/sec or 54 g/sec. So if we need 224 g for the shot, that’s 224 g ÷ 54 g/sec = 4.2 sec. No problem!
Again, we have to remember that the specification is for PS and we are molding PP, and the densities and specific heats or heat capacities are different. In this case, heat capacity rules the calculations. PS has a heat capacity of 1.2 kJ/kg-°K, and for PP it’s 1.8. We actor in the difference as a ratio: 1.2/1.8 or 0.67. So we now have a plasticating rate of 54 g/sec x 0.67 = 36 g/sec. And 224 g ÷ 36 g/sec now gives us a platicating time of 6.2 sec. Oops, there goes cycle time and lost profits.
Yet there’s more to consider: The specific heats are per °K, the same as a °C, so we have to factor in the melt temperature of each resin—PS at 425 F and PP at 400 F. Converting to °F, we get 218/204 = 1.07. Our plasticating rate is thus 36 g/sec x 1.07 or 38.5 g/sec and a predicted plasticating time of 224 g ÷ 38.5 g/sec = 5.8 sec. The equation to use (as per John Klees) is:
Plasticating Rate (Pr) for PP =
Pr PS x (Sp Heat PS/Sp Heat PP) x
(Melt Temp PS °C/Melt Temp PP °C)
There’s still more to consider. The machine spec is for the maximum plasticating capacity. Do you want to bet your profit margin on a maximum spec, implying maximum screw rpm and minimum backpressure? Considering check-valve leakage, recovery-time variances, screw wear, higher backpressure for mixing colors, and other factors, a prudent safety factor would be 10% to 15% additional capacity. Now we are up to 246 g and a possible plasticating time of 6.4 sec or longer.
Let’s settle on a required 7.0 sec for cooling—that is, screw rotate time plus screw delay of 0.2 sec, plus 0.2 sec for decompression after screw rotate, plus at least 0.2 sec for idle time.
What happened to our profit margin? See the illustration again. The extra 1.8 sec on the cycle ate 41% of your quoted profit.
A little trick here for jobs that are tight on plasticating time is to use a nozzle shutoff valve. This option allows the machine to utilize the time during mold opening and closing for plasticating. A significant time saver, but nozzle shutoffs are usually high-maintenance.
What Percentage of Barrel Capacity Should Your Shot Size Be?
The last two columns covered calculating shot size and pasticating capacity and also dealt with using a given percentage of barrel capacity per shot. These articles generated a few emails with questions concerning how much of the barrel or shot-size capacity should be used. Some felt that this is determined by experience—with little science involved—and most seemed to think it is guesswork. Actually, it’s a combination of both experience and science.
My recommendation is that your process should take somewhere between 25% and 75% of the barrel or shot capacity. So what is the rationale and science (if any) behind this advice?
If you’re using less than 25% of the barrel capacity, the material’s residence time will probably be too long and you’ll likely see resin or color degradation during production. This begs the question: What is the residence time for any percent of the barrel used? My procedure is this: As you are starting up and have set up the rough shot size, put a few granules of different color into the feed throat—best if it’s the same, or at least a compatible, resin type. Then count how many shots it takes until you see the different color come out. Take the number of shots and multiply by the cycle time and you have an idea of the residence time. Some people try to determine residence time purley by calculating with certain equations, but I have not seen good results with that approach.
In addition, if the different color is the same as or compatible with the resin, look at the parts made with that color. If the different color is streaky, the screw is not providing uniformly melted polymer. You can expect lower physical properties, poor weldlines, warpage, inconsistent dimensions, color variations, and a host of other profit-killing issues. On the other hand, if the color comes out as a pastel or mostly blended with the resin, you have evidence the screw is providing good mixing and you have a uniformly melted shot. But I digress…
Another reason you want to use more than 25% of the barrel is that the longer stroke allows for better velocity control. One of the premises of “Scientific Molding” is that you must keep fill time the same; it helps stabilize viscosity and allows production of more consistent parts. It is easier for the machine to control velocity over a longer stroke than a shorter stoke.
However, you don’t want to use more than 75% of the barrel capacity either, or you’ll most likely have problems with melt uniformity. Remember that as the screw turns and backs up to make the shot, the feed section of the screw actually becomes shorter from the point of view of the pellets. Typically the feed section of a screw is half the screw flight length. With large shots a good portion of the feed section winds up behind the feed throat as the pellets fill the flights. That leaves only a few flights to prepare the resin for the transition or compression zone.
This is a problem because in the feed section granules start to compress, and thermal heating is provided via the rear zone. You may get by when running amorphous resins, as they melt gradually, like butter. But semi-crystalline resins melt differently, more like ice: They stay hard until they reach their melting temperature and then melt.
With only a few flights for preparation, semi-crystalline resins present severe problems to the transition zone. You can get screw slippage in this barrel section when the screw does not back up as it rotates. You’ll also foul up the friction of the granules on the barrel wall and lose a major mechanism for driving energy into the pellets. In addition, you have to factor in that melting most semi-crystalline resins requires nearly twice the BTUs (energy) of melting amorphous materials.
For example, amorphous ABS takes about 150 BTU/lb to bring it from room temperature to melt temperature. The same amount of polyethylene will take about 325 BTU/lb, more than double the energy required for melting. This explains why some PE molders see flights eroded or even completely missing when they pull screws. In addition you now can also explain why you have unmelt in your parts: The transition or melting zone of the screw just could not generate the energy required to get the job done. The metering zone does not finish melting; it is merely a pump.
The process has to factor in screw slippage, room for the non-return valve to seat, and some decompression after screw rotate. These factors are hard to predict or calculate, so if you go over 75% shot size you can expect trouble. Also, large shots require longer screw rotate times that eat into profit margins.
So pay attention to the percentage of your barrel capacity your jobs are using. It may be one of those molding details that gets overlooked but it could be stealing a chunk of your profits.
Why Multi-Cavity Molds Fill Unevenly
In my March 2010 column I wrote about the importance of uniform filling or balanced flow for each cavity of a multi-cavity mold. Balanced filling is critical for making identical parts, achieving high CPKs, holding tight tolerances, and getting “good” data from design of experiments (DOEs).
Balanced flow is critical for both filling and packing (filling influences packing). What I did not cover in the March article were the reasons for uneven or imbalanced filling. Here are the “classic” reasons for non-uniform filling of a multi-cavity mold (let me know if I missed any). The real problem is determining which one or combination of these is affecting your parts:
1. Non-uniformly melted plastic. Unmelted or partially melted plastic can disrupt flow in the runner, gate, or part. This may seem like a minor issue, yet my experience tells me it is significant. For example, temperature control of the nozzle tip and body is notoriously problematic, especially with cold runners. Poor screw design may also result in non-uniform melt. Then there is temperature control of the manifolds and tips in hot-runner systems. How many of you are setting some “unusual” temperatures to get the hot runner to function?
2. Differential venting among the cavities or flow path. Yes cold runners should be vented. Air, water vapor, off-gases, or volatiles from the polymer can build pressure in cavities and restrict filling. I demonstrate this at every one of my seminars. When troubleshooting venting issues for fill balance, one trick is to make short shots at 90+% full and 65-80% full to see if the balance changes.
3. Non-uniform cooling of the mold, hot-runner system, or hot tips. Plugged or partially plugged water lines; water lines too close to the parts or too far away; coolant taking the path of least resistance; laminar flow; air trapped in a coolant channel; or other equipment coming on and off line can all wreak havoc with proper cooling. While rarely done, a case can be made for regulating water coolant flow separately through each channel of your molds.
4. Part design, particularly non-uniform wall thickness. If you have thick and thin sections in the part, the thick section will fill relatively easily compared with the thin section. If gated into the thick section, flow may “hesitate” at the thick-thin junction, causing a seesaw filling pattern. This factor must be considered in “living hinge” applications.
5. Lack of proper velocity control during filling. Too many profiled velocity changes; running the process pressure-limited; or setting up with different fill times—these are big culprits for filling imbalances. Fill time basically establishes shear rate, which in turn establishes viscosity. Varying fill time (shear rate) from shot to shot, run to run, or machine to machine provides a “different” process due to huge changes in viscosity. Change the viscosity and you can the change fill pattern.
6. Gates not all the same size. No problem, as this is easy to check with a pin gauge, right? Wrong. Relatively few pin gauges measure gate land length. It is difficult and time-consuming to measure land length, but it’s essential because gate land length establishes pressure drop. To achieve balanced filling you need each gate to provide identical pressure drop. There is a reason most mold builders have a rule for maximum land length.
7. Unbalanced flow path. If flow distance and/or path geometry are not identical, non-uniform filling is inevitable.
8. Laminar flow. This is perhaps most sinister reason on my list, because you can have everything else down perfectly and still have poor balance in filling. Polymer melt flows in layers (laminar flow) of different velocity, temperature, shear rate, and viscosity. There are concentric layers within any given flow path—a nozzle, runner, manifold, gate or part. Shear is the prime determinant for non-uniform viscosity. The melt layer near the wall of the flow path will see the highest shear, providing the lowest viscosity (easiest flow). The center path has the lowest (or zero) shear, so its viscosity is the highest (stiffest flow). In the sprue, the concentric layers are symmetrical, but as soon as you split or branch the flow path, this critical symmetry goes out the window and you have an imbalance in shear rate across the layers resulting in unbalanced flow of the layers.
Even in a geometrically balanced (or “naturally balanced”) runner system, runner branching upsets the flow symmetry and all cavities do not fill evenly. It is common to see the inside cavities fill earlier than outside cavities. To achieve identical parts you must maintain the laminar flow symmetry in order to control the rheology (viscosity) of the polymer flow to fill all cavities evenly.
Dr. John Beaumont of Beaumont Technologies Inc., Erie, Pa. has developed a five-step procedure that determines if the problem of flow balance is due to steel problems (like gate-size variations) or because of this inherent laminar flow phenomenon. (Check out ptonline.com for articles on this subject by Dr. Beaumont in Feb. 2007 and Nov. 2008).
What to Calibrate on Your Press
In injection molding, the task is to make all parts “identical” from run to run. What constitutes identical depends on the application—for example, parts for aviation, medical, and the like would have more stringent requirements than would flower pots or drink coasters.
There are different approaches to meeting this goal. At the heart of all them is the process or the machine setup sheet. This sheet documents the process, and attempts to specify the machine, mold, and resin process variables so that “identical” parts are produced, shot to shot, run to run, and machine to machine. Of the thousands of molders in North America, each has its own setup sheet. My bet would be you cannot find two that are identical.
To be fair, this is a difficult task, as the processor must deal with hundreds of variables, from the moisture content of the material to the temperature of the melt. If we want to make “identical” parts we need to ensure these processing variables are controlled and held within a reasonable tolerance. That mandates that when you set up a process, you input the correct process setpoints and actually achieve those process conditions, whether they be temperatures, pressures, times, injection rates, cooling rates, or whatever you have determined is important.
Does the industry agree on what process variables are necessary to specify and control? No. So what’s a molder to do? In my view, you should have only one setup sheet per mold, not a setup sheet per machine. The point is that whatever you input into the machine controller—pressure, time, etc.—the the machine should provide that accurately within a reasonable tolerance. This means the machine must be calibrated so it can produce the required conditions to make identical parts. Before we get to the list of machine calibrations, we need to discuss two terms: repeatability and accuracy.
If you input a hold pressure of 800 psi hydraulic pressure, and the gauge or controller screen reads 800 ±10 psi on every shot, the machine is repeatable. But that does not necessarily mean the true hydraulic pressure is close to 800 psi. The transducer could be faulty; the amplifier card might be out of adjustment; or perhaps the bourdon tube is fatigued. This is why all machines must be calibrated to read the true pressure. Anybody want to share a story about how a pump was changed or a maintenance technician fixed a machine, and when the next shift started up the machine things were “different”? The bottom line is that to replicate identical parts we need to replicate plastic conditions, not machine conditions.
Here is my list of items to calibrate for molding:
1. Pressures, hydraulic or plastic, depending on machine type.
2. Stroke or screw position, including zero stroke.
4. Oil temperature (if a hydraulic machine).
5. Mold temperature controller temperatures.
6. Mold temperature controller gallons per minute (GPM), if you do not regulate the GPM to coolant channels.
Here’s why each of those factors is important:
1. First-stage injection pressure (peak and at stroke transfer), pack and hold (second-stage) pressures, and backpressure are critical for most of us, thus they must be accurate from machine to machine. These values should be the same in actual plastic pressure as you go from one machine to another, electric or hydraulic. Desired accuracy is ±1.5% for hydraulic, ±2.0% for plastic pressures.
2. Screw position establishes shot size, transfer position, cushion, decompression, and cushion. These establish the volume of the shot, which has to be the same from machine to machine. That is volume, not stroke travel. Accuracy should be ± 0.02 in. (0.5 mm).
3. Actual melt temperature would be nice; however the problem is measuring it. I can provide no easy answer to that one. We can put a man on the moon, we have ISO 9000, Six Sigma, etc., but the fact is we cannot accurately measure melt temperature from shot to shot or run to run. This leaves us calibrating thermocouples as an alternative. Granted there are ways to take single shot temperatures by interrupting the cycle, but repeatability is a problem.
4. Hydraulic machines can behave differently according to the temperature of the oil. So it is important that the displayed temperature be accurate to ±2° F, or ±1° C.
5. Mold temperature influences cooling rate, part dimensions, and some properties, so this has to be repeatable. Note that we usually measure only the coolant temperature; the mold may be running at a much different temperature and takes significant time to equilibrate. Accuracy ±3° F or 1.5° C.
6. Calibrate the mold temperature controller to produce the rated GPM.
7. Fill time, hold time, screw recovery time, cycle time, etc., all need to be accurate for shot-to-shot and run-to-run repeatability. It is critical that timers be accurate to ±0.01 sec. Note: velocity setpoints do not need to be calibrated if you use fill time as measure for injection rate.
Improve Profits by Graphing Injection Pressure
Click Image to Enlarge
This is a typical graph of injection pressure vs. time, with the machine sequences noted. All curves will not look like this, but this represents a typical process. Note that the vertical axis is plastic pressure; a curve of hydraulic pressure will look the same but with a different scale.
Most newer injection molding machines with faster microprocessors now have some graphing capabilities. Usually injection pressure, injection velocity, screw position, and sometimes even backpressure can be plotted vs. time or screw position. This graphing capability offers real help in understanding the process and machine.
So take some time and see if you can figure out how to turn it on and set up the scales to your liking. This will not be as capable as a full process-monitoring system, but it’s a good start. Graphing is a powerful aid in optimizing the process and troubleshooting both process and machine issues.
First, make sure you are dealing with good data. My criteria:
1. Speed of data logging. My preference is data logged at the rate of at least 100 points/sec—1000 points/sec is even better during fast machine events like injection. Logging data at only 10 points/sec can miss some important events, yet it is fine for slow events like second-stage injection pack and hold.
2. Ability to properly scale both the vertical and horizontal axes.
3. Ability to pick the variable of interest, for example injection pressure or screw position.
4. Ability to pick time as the horizontal axis. This is perhaps the most frustrating issue on many machines. Often you will find first-stage injection plotted vs. screw position as the horizontal axis and then changing to time as the horizontal axis at switch-over from the first to second stage. Worse, the two plots are not connected, and how this switchover point happens is critical for good processing. Processors and maintenance need to see it clearly as one continuous curve, so use time as the horizontal axis.
5. Ability to overlay graphs of different variables for one cycle or for several cycles.
6. Ability to plot first-stage injection, second-stage pack/hold, and backpressure on one graph. Not seeing all together seriously limits processing and maintenance folks’ ability to use the data.
7. Ability to select the color of the graph.
All this may sound simple, but remember that the first priority of the controller is to control the machine cycle. Data gathering and manipulation are secondary and have to take a back seat to critical machine functions. So it may not get the computer power or time to do high-speed data gathering or storage. You may need dedicated process-monitoring equipment to get the information you want.
Now that you are collecting good data, what can you learn from these plots? Here’s a partial list:
•Is the machine repeatable?
•Do you have a cold slug problem in the nozzle tip?
•Are you pressure limited on first stage?
•Is second stage stable or trending?
•Is machine response appropriate at switchover from first to second stage?
•Are you starting to turn the screw with too much pressure on it?
•Is backpressure stable?
•Is there a difference between peak injection pressure and pressure at transfer during injection?
•If there is a glitch in machine function can you readily see if it is in first, second, or backpressure stages?
To illustrate the benefits of doing this, refer to Figs. 1 and 2. Begin by looking at only one curve at a time. Two or more can be confusing. Understand what graphing one process variable tells you before going on to another. Keep the same color for each process variable; for example, make injection pressure red for all machines. This saves time in interpretation. Again, try to keep the horizontal axis for the time scale.
Figure 1 is a typical graph plotting injection pressure vs. time, with the machine sequences noted. If you could overlay several shots you could easily see if the machine was repeatable; and if there was a machine glitch, you could easily tell in what stage it was occurring on and if it was on every shot or “random.”
Figure 2 shows what a cold slug in the nozzle looks like. Again, this is injection pressure vs. time. In this case, having the ability to overlay the graph for every cycle would enable the processor to watch the spike grow with each new shot. The cold slug spike would start out small and grow higher with every cycle. The longer the nozzle is in contact with the cold-runner sprue-bushing the more heat loss and the higher the cold slug spike. Interrupt the cycle and pull the injection unit back and the spike would be lower again upon startup.
Note that this cold slug will not always be caught in the cold-slug well. It can easily wind up blocking or restricting a gate. But now you have documentary evidence of the cold slug—it is not just a matter of opinion, so there’s an end to time-wasting arguments. To remedy, make sure the controlling thermocouple is back one-third the length of the nozzle; try to have a free-flow nozzle tip; reduce the tip land; and reduce the tip’s contact area to minimize heat loss.
Understanding the intricacies of these graphs will provide better part consistency, process understanding, and faster troubleshooting. Profits will improve.
Getting Good Data from DOE
The injection molding process encompasses hundreds of variables. There are so many that some practitioners let science fall by the wayside and take the attitude that molding is an art.
Honestly, sometimes I too get a bit frustrated when I cannot explain what is happening. But there is an answer—we just have not found it yet. Often the question is: Which process variable or variables is/are controlling some critical aspect of the part? This aspect could be a dimension, warp, gloss, or whatever.
Dealing with so many variables can get confusing. One way to get a handle on them is to organize or categorize the processing variables into groups. To do this I use Don Paulson’s four plastic variables (Don founded Paulson Training Programs Inc.), but I change the word “variables” to “categories.” This helps separate plastic variables from machine variables. So I review the process problem relative to four categories of variables—from the point of view of the plastic: flow rate (encompassing shear rate and injection velocity); temperature; pressure; cooling rate and time.
DOING DOE RIGHT
To answer the question of which processing variable is influencing part dimensions, warp, gloss, ovality, etc., many like to perform a Design of Experiment (DOE). In a DOE one selects one or more “factors” (process variables like injection velocity or hold pressure), and then assigns levels to test for this factor. Levels are the high and low for a particular variable. For example, choosing the factor of injection velocity, a low level might be 0.50 in./sec, and the high level might be 2 in./sec.
Care has to be taken in choosing factors and levels. Keeping both to reasonable numbers can save huge amounts of time and money. Processing factors or variables that may be important should be identified during Operational Qualification (OQ) for medical molders and Production Part Approval Process (PPAP) for automotive processors. Levels should also be established in this initial work-up of the mold, resin, and process. If we pick injection velocity as a factor, the levels can be found by looking at the viscosity curve and parts at each tested velocity.
Whatever factor you pick, first you have to ascertain if it’s a machine or plastic variable. If the former, you must evaluate it with respect to the four categories listed above to help you organize the plastic variables.
So let’s suppose we’ve chosen injection velocity as a machine variable, and we have a low and high level for velocity from the viscosity curve. Now subject the factor of fill time to the four questions. With injection velocity/fill time we change the flow or shear rate. So what are the ramifications on the plastic? Changes in shear rate will change viscosity and perhaps fill pattern. See the influence? In this case velocity is the machine variable and fill time is the true plastic variable.
Restate the two levels as fill time by finding what the fill times were for the two injection velocities. Fill times can be reproduced from one machine to another, but duplicating velocities will not duplicate plastic conditions. Let’s say the high/low fill times were 3 sec and 0.75 sec. A molder might think all he needs to do is set injection rates of 0.50 in. and 2.0 in/sec in order to achieve those fill times, but the experiment will provide incorrect results.
Why? If the molder changes fill time only by changing the injection velocity, without touching the shot size or position transfer set points, he actually obtains two different shot sizes. If he sets the same stroke transfer position for both levels, watch what happens—again, from the plastic’s point of view. Using the slow velocity first to provide a 3-sec fill time, the molder will find a 0.60 in. transfer position provides a 98% full part by volume. He sets up this transfer to second stage to pack out the part.
Now, he increases the injection velocity to whatever machine setpoint is needed to obtain the 0.75-sec fill time. What happens if he uses the same cutoff position of 0.60 in.? Well, I would not want to be around for this shot if it is done in a tool with a slide in it.
All machines are governed by the laws of physics, and even though our hypothetical state-of-the-art press still transfers at 0.60 in., the ram/screw now has significantly more momentum, kinetic energy, and will coast beyond the 0.60 in. setpoint farther than it does at the slow fill time. This might flash that slide/mold. Even if it doesn’t, the first stage does not make the same 98%-full part, and the same pack and hold pressure will provide significantly different cavity pressures. Yes, you’ll have parts, but the data will not be apples-to-apples for appropriate comparison.
To get good data the processor has to re-establish the position transfer for fast fill to achieve the same 98% full part as with the 3-sec fill time. Note that we have only touched on one of the four questions; do any of the others demand a change in molding procedure? Do changes in fill time influence plastic temperature, pressure, and cooling rate or time? How do you perform the DOE to get good data?
NO GUESSING ALLOWED
Another common error in setting up DOE is to set levels arbitrarily. For example, suppose a molder wanted to test hold pressure as a factor and picked low and high levels of 400 and 800 psi hydraulic hold pressure, respectively. There are two significant problems with these levels. First, they are stated in machine pressures when they should be stated in plastic pressures. (My bet is that 80% of DOE variables are stated as machine variables instead of plastic variables. Always work with plastic variables.)
Second, can the mold be packed properly with the low level? And would trying the high level damage the mold by flashing a slide? Anybody want to fess up and admit a mold was damaged attempting an inappropriate level? I estimate that more than 50% of the DOEs performed have levels incorrectly chosen. (Ever wonder why DOE data don’t always make sense? This is one possibility.)
For hold pressure, it is an easy and fast experiment to find the low and high limits possible for the resin, mold, and process. After determining gate-seal time, and you have established whether the part needs gate seal or unseal, you set the hold time. Then start with 500 to 1000 psi of hold pressure and slowly raise the hold pressure until parts look like they might pass QC. Then continue to raise the hold pressure, slowly, until you get early evidence of flash, sticking, pin push, stress whitening, or any mold-opening or part-ejection problem.
You cannot run near this high pressure, so back off a few hundred to 1000 psi plastic pressure for your high limit. You now have established the low and high possibilities for this mold, resin, and process. And you haven’t busted a mold because somebody plugged in a number dreamed up in an office.
All About ‘Cushion’
Many believe injection molding is more art than science. I don’t have all the answers but have spent many years working to replace as much of the art as possible by using scientific principles.
Perhaps one of the reasons the term “art” pops up is the confusion in our industry about terminology. To date, there is no accepted standard for the terms we use to describe processing and part problems. For example, what is the difference between a weld line, flow line, knit line, meld line, and … you get the picture. How can we get to the science when many of us are using different names for the same issue, and sometimes the same name for different issues?
My goal for today is to pick one term: Cushion. I’ll tell you what it means and why it’s important.
There are two definitions (see what I mean?) for cushion in molding: 1) the position of the screw when the hold or second-stage timer ends; 2) the most forward position the screw reaches during injection (first or second stage). Sometimes this is labeled “cushion minimum.”
While these definitions are not the same, the values they produce may or may not be, depending on the mold, machine, or process. Confused? To duplicate and document a process correctly, the molder needs to know the difference and understand what the machine is reporting.
Let’s use an example to explain. Suppose that during first-stage injection, the screw reaches 6 mm from the forward zero position—the transfer point from first to second stage. Thus, 6 mm is the stroke transfer position (velocity to pressure control, or V to P). And supposed that there is enough hold pressure so there is no bounceback of the screw and it continues forward to pack the part. At the end of the second stage (pack and hold), the screw is at 3 mm, which makes this both the position of the screw at the end of second stage and the screw’s most forward position. In this scenario both definitions of cushion would result in the same 3-mm value.
In another scenario, let’s say the screw reaches the 6-mm stroke transfer position during the first stage, yet bounces back slightly due to a required low hold pressure to minimize overpacking at the gate. (For some parts screw bounceback is acceptable as long as the flow front does not hesitate in filling the cavity. Plastic is compressible.) Here, the position of the screw at the end of hold is at 8.5 mm. Some machines will report a cushion of 6 mm, and the screw position at the end of hold as 8.5 mm. Another machine would simply report 8.5 mm as the cushion, and yet another might report cushion as 6 mm. My preference is to see both reported. This is not a case where one machine is better than another, just that there are two definitions being used and you need to know what the numbers mean. Which definition is right or best? That’s not my call. Better check out what your machines are reporting to you. Is your documentation correct?
Now that the definitions are established, what should the cushion be and what is an acceptable range? My bet is that if we sampled 50 molders we would have lots of different responses. To start answering these questions, let’s state the purpose of a cushion: to provide a pressure pathway to enable packing out the part. To properly pack out a part, there has to be plastic in front of the screw/non-return valve (check ring) to provide a means to transmit plastic pressure through the nozzle, sprue, runner, and gate(s) to the part. There are processes such as gas assist that do not require the injection unit to pressurize (pack out) the cavity, but for our discussion, we will stay with conventional injection molding.
On a hydraulic machine you can program hold pressure: For example, if you set 850 psi (59 bar), you will get 850 psi hold pressure for the time you set on the second-stage or hold timer. However, if the screw reaches the zero position or “bottoms out” at any time during the hold stage, there is a good chance you’ll make a bad part. Without that “cushion” of material in front the screw, the hydraulic pressure is not developing any plastic pressure in the nozzle, and most likely the part won’t be properly packed.
On an electric machine, if the screw reaches the zero position, the results will be the same. However, the controller is often programed to reduce the hold pressure to some minimum value to avoid unnecessary load on the expensive electric servo motor. The controller knows the screw cannot move forward any further and subsequently reduces the forward pressure.
While I prefer units of volume for shot size and cutoff position, cushion is best stated as a distance. Picking a particular volume as a rule of thumb would be a very small distance on larger screw diameters. My target is about 6 mm, or 0.25 in. This assumes that the numbers on the controller screen are accurate.
If the position sensor is not calibrated correctly, you could be reading a 3-mm cushion on the controller when in fact the screw is at the bottom or end of the stroke. Safety tip: it is critical that the screw tip never touch the barrel end cap, because it has the power (pressure) to blow the end cap off. There is always a minimum clearance; my understanding it is about 1.5 mm (0.060 in.). The hydraulic cylinder has stops in it to prevent going beyond this, providing the screw is the correct length. (I am not sure how electric machines handle this.)
Finally, how much shot-to-shot variability is acceptable? In an ideal world the answer is zero. Unfortunately, variability is a fact of life due to check-ring design, wear factors, contamination, lack of maintenance, and the fact that most nonreturn valves have a propensity to leak differently from shot to shot.
If you would like to get an idea of how much they leak, do a simple test on any given part. Note the actual screw start position before injection and the cushion using definition number 1 above. You can now calculate the volume of the stroke using the equation for the volume of a cylinder:
Volume=π × r² × stroke length.
Now catch all the plastic parts and runners, etc. that drop out of the mold. Weigh them all. Find the melt density for the plastic you are using, and calculate the volume of molten plastic needed to make the parts you weighed. Let me know if the volume calculations match. I’ll bet that the calculated shot volume is 1.3 to 2 times the volume calculated from the plastic weight. Because of these issues with non-return valves, my cushion window is ±1.0 mm. (It should be less for a plunger.)
Understanding Polymer Flow: Interpreting the Viscosity Curve
Let’s take a step back in the development of Scientific Molding with a review of one of the benchmarks of the technology: The on-machineviscosity curve. When it was first presented in the late 1980s it was not well accepted, but some 20 years later it has gained a substantial foothold, perhaps to the point of being a bit overrated.
For those new to Scientific Molding, the viscosity curve is a plot of the viscosity of a polymer vs. shear rate (see graph). Data here is plotted on linear scales rather than the academic (and more correct) log-log method. I like the linear plot because it shows the magnitude of viscosity variation as you change shear rate by changing injection velocity or fill time. Data is obtained by doing this experiment on an injection machine with the mold and resin you plan to run.
Start the experiment by taking hold pressure off. Lower the hold pressure to near zero; then, starting with a high injection velocity, make a short shot, 99% full by volume. Take data as you gradually lower the injection velocity—don’t adjust anything but injection velocity. From the table below, you can use the injection pressure at transfer and fill time to develop the viscosity curve. (For those who want to know the science behind the terms in the chart, the parameters in this experiment are related by a fluid dynamics expression known as the Hagen-Poiseuille equation. You can find the equation on Google.com.)
After interpreting this data in relation to the machine, resin, and part, the processor is guided to a fill time for this particular mold and resin. This fill time is now set for the life of the project. Once you become adept at the procedure and log the data into a spreadsheet, the experiment should take about 30 minutes. That’s time well spent to get the data for determining fill time and the start of a production-capable process.
But how do you interpret the curve? The concept is that if you understand how plastic flows and fills a mold, you can use this information to make robust processes that produce consistent parts. Starting with the basics: The graph shows a typical curve. In fact, once you do a few of these you’ll recognize that if you have good data that spans a decent range of fill times, the curves all have the same shape. Only the ranges of the scales change. That is the way it should be—plastics are non-Newtonian, meaning they all change viscosity with changes in injection rate or fill time. In fact, viscosity is influenced more by shear rate—injection speed—than temperature or even resin lot variations.
Check out the numbers in the table accompanying the graph. If you round up a bit, the viscosity at the fastest speed, a fill time of 0.39 sec, is about 2000 psi-sec. If you double the viscosity it would be 4000, five times the viscosity would be 10,000, and 10 times stiffer would be 20,000. So in this case, just by changing injection rate you can change the viscosity 10-fold or an order of magnitude. Temperature and lot variations cannot come close.
If the biggest single factor in the viscosity of most plastics is injection speed or fill time, what does that mean for processing? First, you must keep the fill time the same to have a stable process that can make the same part on any machine. That is, shot to shot, summer to winter, machine to machine. Our range in this case is a tight ±0.04 sec. It does not matter what injection velocity you set on any given machine—the critical parameter for the plastic is keeping the fill time the same. This keeps the shear rate the same. (A little caveat here: If it is a multi-cavity tool, fill balance is important.)
The next step in interpreting the data may be asking this question: Where on the curve do want the process to run? You have to be somewhere on the curve, because your experiments covered the entire range of the machine: You went as fast as the machine could go to as slow as it could go. Looking at the data you can see a large change in viscosity on the left side of the curve. That does not bode well for a stable process. For process stability, the plastic likes the flat section of the curve, as shown in the center to right side of graph. Viscosity is not changing much in this section; it is a flat line, something like the way water flows. In other words, the plastic likes to go fast; it has lower viscosity, which does not change much with small speed variations.
Because of this, many processors pick a point near the shortest fill time. Unfortunately, things aren’t that simple, as the part, mold, machines (other machines this mold may be scheduled for), or the application often require some tough compromises. For example, if you are molding optical lenses, you will be forced to be on the steep section of the curve, as there is too much birefringence at the fast injection speeds. Plating or painting also require slow injection speeds and longer fill times. Your machine may be pressure-limited, or perhaps cannot control the ram speed. The data is just telling you what the plastic wants to do; all the other components of making a good part have to be considered by testing. This is what I meant above when I used the term “overrated.”
What about profiling injection? Most machines today can profile five to 10 velocity steps for part filling. Does the data tell you when to change speeds and what speeds to use? Some molders waste hours on this, to the point of changing the profile daily. However, the test parts from the velocity curve experiment provide the data you need.
Line up all the shots in order, fast to slow or slow to fast. Inspect each to see if there are any changes in appearance as you changed the fill time. If all parts are short and sinky with no blush, flash, streaks, burns, or other defect, then there is no reason to profile injection speed. If you do see flash or burns, those are mold issues that should be fixed, not worked around. You will have to add time to the cycle and that will kill your profit margin, so fix the tool. Keep it simple and use one injection speed. I would estimate that 80% of all molding jobs should not need profiling.
In short, take the time to do the viscosity curve on new molds. You will learn more in that hour than many learn in years about the process for this tool. But remember that there are other components of making a good part that have to be factored into this complex equation.
Should You Profile Injection Velocity?
In my last column I reviewed the viscosity curve and detailed some of the information that can be gleaned from the data. The objective is to come up with a practical fill time—as opposed to guessing—based on data similar to mold-filling analysis. One of the side benefits of doing this is that one obtains the necessary data to determine if part-filling injection velocity should be profiled. Most machines today have between 5 and 10 stages of injection, but don’t use them unless necessary. “To profile or not to profile” should be a data-based decision.
My strategy is: Keep it simple. If you can get by with one velocity, use only one, because fill time is easier to reproduce, and it is easier to set up the process.
Some machines profile speed by percentages or in 10 equal segments. Those are nightmares to set up. Profiles must be done properly by position or volume. This is critical if you want to reproduce the setup on a machine with a different size barrel.
Some 95% of the machines that I evaluate are out of calibration, or drift in calibration, for velocity. So you need to duplicate fill times and not machine settings.
In a hydraulic machine, there is no brake and inertia may override valve moment, especially during deceleration. Electric machines do have a brake, but how well does it work? Imagine that you can move a throttle on a speedboat 10 times within 10 meters (~100 ft); but will the boat respond? The same principle applies to injection.
My strategy is: Do not profile unless you have to. Generally about 80% of your jobs will need only one velocity. How do you tell which ones? The parts made while doing the viscosity curve provide the answers. They will also tell you how many profile steps and what speeds you need and at what position/volume you need to change injection rate.
If you have done the viscosity curve correctly, you have eight to 12 short shots that cover the entire velocity range for your machine. Line them up from the fastest to the slowest injection rate, all with the same orientation—e.g., cosmetic side facing you. Inspect each part for cosmetics, warp, blush, jetting, burns, flash, or other problem. Ignore the size changes as those are due to the momentum or changeover issue (remember the no-brake discussion above). This sequence should show the influence of injection velocity on the part. For example, jetting may be seen at the high injection rates but disappear at slower injection rates.
Before I slow down due to jetting, burns, flash, etc., I try to fix the tool. Why not process around the problem, you ask? Well, your job depends on your company making a profit margin. For jetting, flash, and burns, your process fix will be to slow the injection velocity. That adds to cycle time. You cannot take it off the second stage, screw recovery time, cooling, or clamping time. Those were already optimized, so how could you speed them up? (If you can speed them up they were not optimized at the start.) If cycle time goes longer, profits get squeezed.
Take flash as an example. At slow injection rates and longer fill times, the viscosity of the plastic thickens dramatically, so you are fixing flash by forming a gasket around the parting line. So you can slow the injection rate and your flash will go away, but how much time did you add to the cycle?
Jetting, burns, and flash are tool problems, so fix the tool and don’t slow down injection. Remember, we are talking about profiling, not necessarily speed. If the part needs a long fill time, use one slow velocity. I am not saying you must have short fill times.
One problem that will cause me to slow down and suffer the consequences is a blush or some other cosmetic issue that nobody has any idea how to fix. Your sequence of test parts will show at what injection rate the blush is present and at what speeds it disappears. Use the highest velocity that does not give you the problem. If blush arises early in filling the mold, use this speed until you are past the problem. Correlate the flow front’s position after the problem is past and you have the position at which you can increase velocity. Two or three velocities should be all you need, not 10.
You have to worry about other machines in your shop that may run this mold. Can they reproduce this sequence of speeds and fill time? You do not want to limit scheduling so that a certain mold can only be used in one machine. Most of the time I have to profile is because of a tool or design problem. If possible, fix the tool.
Nozzle Leaks: Why They Occur, How to Detect & Fix Them
Click Image to Enlarge
Fig. 2: Pay attention to the contact area between nozzle tip and sprue bushing, as it can cause a variety of problems. “A” shows a full contact area. “B” shows a radius mismatch of nozzle tip to sprue bushing. “C” shows a reduced inner-ring contact area (desirable). These results were obtained by placing a piece of pressure-sensitive film between the nozzle and sprue bushing.
Does Figure 1 bring back any memories of wasted time and money, or conjure up a notion that you could show me something even worse? Nozzle leaks are a serious problem, so let’s take a look at possible causes, what you can do to prevent leaks, and what you can do to detect them as soon as they occur.
The nozzle tip must form a seal against the sprue bushing so as to withstand high injection pressures with no leaks or injection unit push-back. Here are some causes/remedies of nozzle leaks:
Low nozzle contact pressure
Machine specifications usually call out a nozzle contact pressure requirement of 4 to 10 tons of force. This contact pressure must counteract whatever injection pressure the machine can produce. Higher injection pressures require higher nozzle contact forces for any size machine. So for applications that use high injection pressure (greater than 22,500 psi), make sure your machine is not at the low end of this 4-10 ton range.
Be careful, as some in the industry believe that small machines require less nozzle contact force than large machines. Nozzle size and radii are the same for all machine sizes; the injection pressure sets the requirement for nozzle contact pressure, not the size or tonnage of the machine.
Hydraulic machines tend have fewer problems generating and holding this nozzle contact force, if they are specified correctly. And it is relative easy to check nozzle contact force by measuring the hydraulic pressure on the pull-in cylinder and checking it against the machine supplier’s specification.
Most electric machines generate nozzle contact force without the use of hydraulics and some have had issues with inadequate nozzle contact force. Further, there is no easy way to measure the contact pressure as there is with hydraulics. However, there is a quick approximate test with pressure-sensitive film between the nozzle tip and sprue bushing (see Fig. 2 as an example).
Too much nozzle contact area
The common belief is that the nozzle-tip radius should exactly match that of the sprue bushing. While this seems right, the best way to make a seal is with a small contact area, not a large one. The best seal is over a small area; but the forces here are enormous and most processors feel full-radius contact is needed to prevent hobbing the steel. I disagree. First, a better seal is formed with a small contact surface. Second, the larger the contact surface, the more heat loss there is with cold-runner tooling (which causes other problems like stringing). And third, the relatively high crush strength of P-20 steel. Trial a 3/8-in. to 1/4-in. contact ring on the inner surface of the sprue bushing to see if this improves the seal without hobbing or compressing either the tip or sprue bushing. See Fig. 2 “A” and “C”.
Contact force on nozzle-tip perimeter
Figure 2 “B” shows a radius mismatch of the nozzle tip to the sprue bushing. This promotes leakage and injection-unit push-back, as the injection pressure is applied to a large surface area of the nozzle to push the injection unit backward.
Misaligned injection unit & sprue bushing
Anytime you bring the injection unit forward, watch as it touches the sprue bushing. If the injection unit self-aligns (move up, down, or sideways), stop—this is a dangerous situation as it wears the nozzle tip or the sprue bushing. At some point they may not align properly, providing a path for the high-pressure plastic to jet toward you. Get maintenance to align the barrel with the sprue bushing.
Incorrect orifice size of nozzle, hot runner
The nozzle orifice on a hot-runner mold should be the same size or only slightly smaller than the hot-sprue-bushing orifice. Any area of the nozzle tip exposed to hot melt will see injection pressure trying to push the injection unit back.
Radius mismatch of nozzle tip, sprue bushing
Make sure you check that the radius on the sprue bushing matches the radius of the nozzle tip for proper mating.
Distorted or gouged nozzle tip
Both the sprue bushing and nozzle tip tend to get abused with use, acquiring gouges, burrs, grooves, or other distortions. As with nearly any aspect of injection molding, pay attention to the details.
Monitor cushion, watch nozzle contact position, and place a dial indicator on the sled to see if there is injection unit push-back during first- or second-stage injection.
Safety first: Nozzle-contact forces can damage the “A” plate if it’s not supported properly. Close the mold to ensure support of the stationary side before bringing the injection unit to nozzle touch. I’ve seen the nozzle-contact force bow or push off the “A” plate.
Establish a True ‘24/7’ Production Process
Recently I had a few calls with the same theme: “We made good parts during the initial mold trial, but now that we are in production the dimensions are not consistent.” In such situations, I have found that the mold startup “process” sometimes consists of quickly shooting a few parts that “look okay,” then walking away believing you are ready for production. On other occasions molders have actually done their homework, conducted a design of experiments (DOE), ran some trial runs, etc., yet the parts still lack consistency.
New or existing mold trials need to produce a production-worthy process that will run 24/7. Yes, I know the mold may have come in three weeks late and somebody wants parts shipped the next day, making everybody frantic. I recognize you do not always have the time to do the full studies—or maybe you do have time and you do the studies and still production is poor or there are customer returns. You have been doing this for umpteen years and things are getting worse. Certainly a case can be made for trying new tactics.
First, get everybody on same page. Explain that you need to put together five critical components for a successful molding application. They are, as formulated by Glenn Beall and Plastics Technology columnist Mike Sepe: piece part design; resin selection and handling; tool design and construction; processing; and testing.
Most molders cannot afford nor need in-house expertise for all of the above elements. Arrange for an outside specialist in the area(s) needed. Remember to have all tool drawings double-checked. The time to find problems is at the drawing stage, not after steel is cut.
Let’s move on to mold delivery and developing a molding process. Most molders would like (and even pay for) someone to write down a stepwise procedure that quickly yields a capable production process. That is the ISO 9000 or TS mentality. Chuck it. The fact is, you are not slinging hamburgers, and no one can write a foolproof procedure for something as complex as molding. Worse, these stepwise procedures kill creativity and critical thinking. Every mold and resin combination has its own oddities. Finding them early is the challenge and following a set procedure is not going to get you there.
FIND & FIX PROBLEMS, DON’T WORK AROUND THEM
Guidelines, not strict procedures, are okay with me. You must allow the operator to think! You also need trained, proficient operators to follow these guidelines. They should be instructed to use their talents to interpret results of various experiments to define any weakness in any of the five components. Their goal is to find and define problems, not process around them. Problems will show up in production and the competitive molder knows that finding them before production will cost less than afterwards.
My guideline experiments include cooling; viscosity curve with visual part inspection at each velocity; velocity-to-pressure switchover response, including momentum; pressure-loss analysis; short-shot study; gate seal; melt temperature; part temperature via infrared imaging (not expensive or time-consuming); second-stage pressure range; screw recovery; and cycle time. All this should take about 2 hr, and if it takes much longer it’s a sign of significant production troubles ahead.
Other experiments may be called for if something odd turns up. I never know beforehand. There is no race to make a part that “looks good.” You need a series of experiments to define the critical parameters for this mold and resin combination. Essentially these experiments are about range-finding—i.e., defining a process window. You want to establish a base process with a list of concerns to be addressed. Do a DOE if necessary. The base process derived from these experiments establishes the levels for the variables (factors), as well as which variables are critical. This often speeds the DOE, as you have data to establish levels and factors rather than guesses.
Developing a process does not start with shooting plastic in the mold. It starts with ensuring the mold is set up properly in an appropriate machine that works. For example, cooling is 90% of your cycle time, and it is often not done right. Most of the time, somebody hooks up a half dozen or more hoses to the mold from a manifold and walks away. Document that each coolant channel is running turbulent flow (Reynolds number of at least 5000) and is regulated such that the flow (gallons or liters/min) can be duplicated from run to run.
Once you have a base process, make some parts for testing to move the process to the center of the part specs. Do not send or show parts to the customer yet, even if they “look good.” What if good parts are scarce when you actually run the first order? Better run for cover; your boss and client will want to know why. If you do make a part that seems pretty good right off the bat, hide it or push it under the forklift. There is testing to be done, including melt flow rate (MFR) on pellets and parts, thermal cycling, and any other tests that show the part meets the critical requirements.
During this testing phase you will learn how to center the process to the part specifications. Often this requires tool modifications. During this phase you will be told these modifications can’t be done, or there is no time or money to do them, so you better find a way around the problem. Don’t do it, or you and the company will be fighting that problem for the life of the job, killing the profit margin and convincing the client you are less than competent.
Once the operator feels this process is stable and initial part testing shows promise, you must challenge the process. Is it truly a stable process? Are fill time, pressure at transfer, cushion, recovery time, cycle time, and part temperature within an acceptable range as you start up, shut down, and have somebody new start it up again?
If so, the next step is to force a viscosity change. This is critical. It will happen anyway sooner or later, so it’s best to know what is going to happen and how to deal with it. Force a viscosity change by testing a different lot, or wet resin, or 100% regrind, or different colors, or whatever you can think of that will mimic the variables in 24/7 production. Are these material variations reflected in the process outputs? Is first-stage consistent as the material variations run through the process? Can you accommodate these changes by changing the second-stage pressure?
This is perhaps the most critical test for a true production process—forcing a viscosity change and seeing if it can be accommodated. It is rarely done. All the work goes into testing one lot of material, but when 24/7 production arrives, you have to process different lots. You cannot blame the resin supplier. As Dr. Deming taught, all processes have normal variations. Your process has to deal with them. Bottom line: If you have tested only one lot of material you do not have a production-capable process.
Get the Most From Your Tooling
Click Image to Enlarge
The arrival of a new tool always brings a certain level of excitement. Often it is late and everyone is in a rush and pressured to produce “good” parts. So the history and culture of the molding industry has been to pull out every trick possible to work around design, resin, tooling, processing, and testing issues. Burns? No problem, just slow the injection rate. Flash, color streaks, blush, shrink, warp, etc.? No worries. We spend hours on profiling, fixturing, trimming, and other tricks of the trade to get around these problems.
The result of this approach is often a few good parts but no real process that meets the production level, quality, or profits expected. Even worse, one of these good parts produced “by luck” is often waved around by your boss, who wants to know why, since the mold can “obviously” produce good parts, can’t you get the process to run? This leads to even more experimentation, which wastes time, resin, and money.
Is there a better approach? Sure, but since each tool is unique and has its own peculiarities, it’s tough to have a standardized way of doing things. That said, there are certain requirements and tests that all tools should pass before any parts are sent out.
BEFORE THE MOLD IS BUILT
Let’s start with requirements before the mold shows up:
•Adequate ejection: One of my pet peeves is that if you make a short shot (on purpose or by mistake), you often have to dig the part out. Every shot a mold will see will not be a full and complete shot. So spend the money and put in sufficient ejection for shorts and problem areas like thin ribs.
The production hours lost as processors dig out parts would justify the extra costs. We are probably talking about hundreds of thousands of dollars each week across the country. And that’s not counting the tool damage caused by frustrated processors using screw drivers, pliers, hammers, ejector pins, etc. to yank parts out. If you add that in, maybe even the bean counters would begin to understand why better ejection makes sense. Scratch a mold surface while you’re digging and the problem gets worse. Some simple, safer solutions include adding a vacuum break, an extra ejector pin, a generous radius, polish in the direction of draw, or adequate draft.
(Why don’t we have a special brass tool that heats up like a soldering gun? Pull the trigger and a dull brass blade or probe heats up to melt the plastic. This would be a lot quicker and safer than all those torches heating screws, ejector pins etc.)
•Adequate venting: We have been using vents in molds for some 60 to 70 years. Most people acknowledge that there is air in the cavity that has to escape as the plastic fills the part, but we still see burns on parts, and many a cycle is extended to allow time for the air to escape. True, we are seeing more vacuum venting nowadays, but we have plenty of room for improvement. My favorite example is the mold builder who supplies you a tool with no venting but with a note asking you to indicate where the burn marks occur so he knows where to put the vents! If this happens at your shop, promise me you’ll take your tooling business to another moldmaker.
As general rule, 30% of the cavity perimeter should be vented. Vent depth is resin dependent; seek recommendations from your resin supplier or a good tooling shop. Vents should be ground with a 120 to 240 grit wheel in the direction of air flow. They should not be milled. Vent land should be no longer than 0.090 in. and lead to a wide dump channel 0.015-0.020 in. deep. It’s okay to mill the deeper dump channel.
•Clear labeling: Each water line should be labeled with its required gallons or liters per minute (GPM or LPM). Also, a diagram of the water-line circuits should be affixed to the mold.
•Time-saving hot-runner diagrams: With every hot-runner system, you should receive a diagram of the hot drops and manifold layout so you’ll know which heater is controlling which drop or flow channel. Unless, that is, you actually love puzzling out which temperature controller regulates which drop or runner.
AFTER THE MOLD ARRIVES
Once the new mold arrives, perform the following tests. These also apply to any current tool that consistently causes problems:
•Water-flow check: GPM or LPM of coolant for each channel should be documented for future runs. Ensure that each line has turbulent flow with a Reynolds number of at least 5000 in each channel. And make sure that all lines have open water flow. While not common, it is possible for water flow to be blocked in one or two channels in a new mold: Gun drills stray and plugs manage to get misplaced. It is worth the hassle to find any blockages before you try to develop a production process, since cooling is 95% of your cycle. It is important to get this right from the start and find a way to ensure that it is duplicated on each run. You might also want to investigate the benefits of positive-displacement water pumps over centrifugal pumps, which are more common in our industry.
•Flow-analysis check: If a mold-filling analysis was done, the calculated fill time should be communicated to the molder. You spent good money on this analysis, so you might as well shoot the tool at the targeted fill time. Compare predicted flow pattern and pressure loss to actual, and involve the analyst. This will provide insight into the process.
•Pressure loss study: This is a short-shot study to check the fill pattern for air traps, flow-front acceleration, uniform filling, and balance of fill. Plus you’ll get the pressure loss required for each of the flow-path components.
One such study on a new tool provided the data in the accompanying figure. In this case the machine data indicated the process was in control, yet the reject rate was high. Note the high pressure loss in the sprue and runner. Many molders simply open up the sprue, runner, and gates to solve flow problems. This time, a more targeted approach was required. There was no need to mess with the nozzle or gates—the sprue and runner were the problem. It was a good thing to find this out when the tool was new and under some type of warranty. If discovered later in production, costs would have been much higher. This 20-min test also can be used to compare the actual pressure loss to the prediction of a mold-filling analysis.
Don’t Ignore Nozzle Temperature Control
Click Image to Enlarge
FIG. 1. A spade or butterfly-type thermocouple, placed between the nozzle body and heater band, can help maintain the setpoint temperature of the nozzle tip and body. Photo: Marathon Heater, Inc., marathonheater.com
FIG. 2. A thermocouple situated on the heater-band clamp is unlikely to provide accurate temperature control of the nozzle tip and body. Typically, as shown here, there are gobs of plastic hanging from the wire, which is a safety hazard.
Most molders consider controlling the temperature of the nozzle tip and body a trivial detail. However, Dr. Deming tried to teach us about the importance of details, and ignoring this one will cost you money. Let’s start with some data, another foundation of Dr. Deming’s teachings that too many molders tend to ignore.
Table 1 shows a difference of over 150° F (65° C) between the setpoint and the actual measured temperature of the nozzle tip and body in a PETG molding operation. Reject rate was near 50%. Granted, this may not be typical, but in my experience it is not rare. And, yes, this was with a PID temperature controller, not just a Variac. Table 2 is more typical, in my experience. But both are unacceptable and need to be fixed for quality and consistent production. Why?
One of the primary variables in molding is using the resin within its proper temperature range, and just as important is that the melt be uniform in temperature. With temperature variations you will get rejects. Possible types of rejects include the following:
2. Gloss differential,
5. Color variation,
6. Texture variation,
7. Inconsistent properties,
9. Cold slug in the nozzle tip,
10. Drooling or Stringing,
11. Burn marks at the gate,
12. Asymmetrical filling,
13. Shot-to-shot variations,
Bottom line: This is a significant problem that costs molders tens of thousands of dollars a month, including hours of wasted time trying to adjust a perfectly good process. I do not have a complete fix, but the following will provide melt temperatures much closer to setpoint and significantly better consistency between machines.
First, perhaps to convince yourself you have a problem, take a walk through the production floor and check where the thermocouple is located on the nozzle body and/or tip on a number of machines. You will notice a lack of consistency. Sometimes the controlling thermocouple is on the heater band, others have it in the hex of a nozzle body. If the nozzle body is longer than 2 in., the hex is not the place to put it. Still others will have the thermocouple on the clamp around the heater band. You can bet money that one is not giving the temperature controller the right information. In fact, all of the above have significant problems.
Ideally, we would have a magical thermocouple that we’d put on the outside of the nozzle body and, through ultrasonic sound waves, microwaves, infrared, or some new technology, we would get an accurate temperature. Some might think the thermocouple must be in the melt, but that will not work. For a thermocouple to provide an accurate temperature, I am told it has to be at least six diameters into the melt. There is not enough room to do this. Further there are more high-pressure machines being used—with 25,000 to 40,000 psi plastic pressures—and I am concerned about the safety of a hole drilled in the nozzle.
To date, my best compromise is to first clean up the nozzle body. Get all of the charred plastic, etc. off until you have a clean steel surface. Obtain a spade, also known as a butterfly thermocouple (see Fig. 1). Place this directly on the nozzle body, using some thermal paste to help with heat transfer. Placement should be one-third the length of the nozzle body back from the nozzle tip. Put it on top and wrap it with three or four turns of glass tape. The heater band goes over the spade. The heater band should cover as much of the full length of the nozzle body as possible, and ensure that it is the right voltage, 110V or 240V.
Take another look at Table 1 to see the before and after results of this technique. Fig. 2 shows the molder’s setup before the fix, with a thermocouple on the heater-band clamp. Seriously, why do you buy a $50,000 to $250,000 piece of equipment and then put up with Neanderthal temperature-control technology? Also note that there is some plastic on the wire in Fig. 2. This is common, with globs of it hanging from the thermocouple and heater-band electrical wire. These wires should not be left to hang below the barrel; they should be routed on the side of the barrel out of harm’s way.
Coping with Weak Weld Lines
Weld lines are where two flow fronts come together. Often the result is a “witness” line, also referred to as a flow line, knit line, meld line, etc. Weld lines cause significant reject rates and are a common problem that all molders face. Weld lines significantly weaken the structural integrity of a molded part and have resulted in liability issues. Part cosmetics are also usually compromised; weld lines can look like a fine scratch, a gloss differential, blush, haze, or discoloration steak.
The reasons weld lines are weak has nothing to do with the temperature of the flow fronts bumping into one another. The flow fronts do not have an opportunity to cool; plastic flows like lava—it rolls out, exposing fresh hot melt. So the hottest material is constantly being exposed as the flow front progresses. (It is this phenomenon that allows in-mold labeling.) The main reason for weld-line weakness is a lack of polymer chain entanglement across the junction of the two flow fronts. A major contributing factor is air entrapped at this junction; the flow-fronts do not actually meet due to gas trapped between the fronts. Sophisticated microscopic analysis has shown air entrapment even in “good” parts.
Weld lines are one of the most difficult defects to eliminate. Their cause, strength, and appearance are influenced by each of the five “Key Components of a Successful Plastic Application.” These components are:
•Tool design and construction,
•Material selection and handling,
Since minimizing or solving weld-line issues deals with all five, let us address each individually and learn its role in the root causes.
To minimize effects of weld lines, the designer must consider part performance and filling or flow pattern of the plastic as it enters and flows through the mold, as well as tooling issues. Plastic part design is a challenge to marry the product performance and marketing demands with manufacturability. For example, large parts often need multiple gates, and multiple gates create weld lines and possibly a cosmetic blemish. And we often design plastics parts with features to aid in assembly, so parts have projections off the nominal wall, like bosses and ribs, as well as holes or depressions.
Projections, depressions, and changes in nominal wall thickness create flow interruptions, which also lead to weld lines. All of these conditions must be considered with respect to the flow pattern upon filling the cavity. The flow pattern plus performance requirements are further complicated by mold manufacturing restrictions. (When you are designing the part you are also designing the mold!) Tough decisions and hard compromises have to made at this stage if the part is to perform as expected.
With so much to consider, where does one start? First, minimize flow interruptions and place them such that the flow fronts are allowed to meet and flow some distance after the interruption. Ribs should be oriented in the direction of flow for ease of filling and venting. Those are easy design criteria to state, but they often require difficult compromises between the customer, designer, and moldmaker. Communication between these functions is critical but often nonexistent.
TOOL DESIGN & CONSTRUCTION
As stated, tool design must be considered concurrent with part design. A large part may have multiple gates, resulting in weld lines. If the cosmetics and performance are critical, the weld lines can be eliminated by employing a valve-gated hot-runner system with sequential filling. In multi-gate, cold-runner molds, there will be weld lines. The customer, designer, and moldmaker have to work together to ensure the weld lines will be in areas of least stress or cosmetic importance. If that’s not possible, it may be best to use a post-molding operation, like drilling a hole rather than molding it in.
One bright spot is that today’s mold-filling analysis programs do a great job of pointing out where the weld lines will be; there is no longer any reason for surprises in finding a weld line when the tool is trialed. Optimize strength by picking a gate location that allows flow for a long distance after the flow interruption. This allows more of the chains to entangle across the weld line.
Sometimes it is worth the hassle of adding a flow tab to provide a place to allow the flow fronts to “knit” a bit better. This flow tab has to be cut off after molding, which is another operation and expense but does provide a valuable vent for trapped air as the flow fronts meet. All weld lines trap air, and it is a critical tooling issue to properly vent the area near a weld line. This includes venting core pins, blind holes, and ribs. Venting is one tooling issue not to be compromised: If necessary, provide for vacuum venting.
MATERIAL SELECTION & HANDLING
Different resins can have drastically different weld-line strengths. Tensile testing of dual-gated tensile bars can provide information on weld-line strength retention. Some soft-touch materials are shear sensitive and can provide a weld-line appearance even when there is no flow-front interruption; a “weld line” can appear with a change of shear rate.
If so, look for an alternative resin. Weld-line strength for plastics can vary from 20% to nearly 100% of the strength of the plastic itself. Neat polypropylene retains nearly 90% of its strength on a properly designed and molded weld line. But add 30% glass and weld-line strength retention drops to only 34% as high as the compound with no weld line. Glass filler improves strength over the neat resin but reduces weld line strength; it does not improve it.
Processing can affect the strength and cosmetics of weld lines, but it cannot eliminate the root causes. Processing variations in temperature and pressure usually can provide only marginal results. The root causes are poor polymer chain entanglement across the weld line and air entrapment. Both of these root causes must be resolved in part design or tooling.
While many processors like to raise melt temperature to improve weld-line strength, the increase of volatiles coming off the polymer often actually decreases weld-line strength.
First, if at all possible, make sure the weld line is formed during the first stage of fill. Creating a strong weld line during pack and hold is often problematic. Along with getting the weld line formed during first-stage fill, it often helps to increase the injection velocity, thus decreasing the fill time and increasing the shear rate. This lowers the viscosity of the polymer during fill, which in turn allows for better chain entanglement and better packing.
Occasionally increasing pack or hold pressure helps. If it is a cosmetic issue, a slower injection velocity may help, but often a higher mold temperature provides better results. Vacuum venting is a rare but a powerful tool that is even more effective and aids in cosmetics and strength. It should be planned for any anticipated weld-line issues with a part or tool.
Screw Decompress Before Screw Rotate
When developing a molding process that will run 24/7, most details are unique to the mold, machine, and part. But there are a few that should be applied to all processes. One is when to start screw rotation to build the charge for the next shot size. While not a critical topic, it is one of those pesky details that deserve some thought.
Doing this thoughtfully can reduce the wear and tear on the screw tip, check valve, and screw motor. Less wear, especially on the check valve, will provide more consistent parts and a more robust process. This equates to better profits.
Many molders don’t pay all that much attention to when they start screw rotation. “Just let it start when the second stage (pack and hold) times out. It’s no big deal anyway, and we don’t want to extend the cycle.” Does that sound like you?
My advice is to think it through. Take a typical process, and think of the typical second-stage pressure you are using. Then, if it is hydraulic pressure, convert it to plastic pressure by multiplying your set hold pressure by the machine/screw intensification ratio. (If it is an electric machine the pressure setting will already be in plastic pressure, so there is no need to convert.)
Let’s take a look at a process using 8000 psi plastic pressure as your second-stage pressure. This is the pressure on the molten polymer directly in front of the screw that the check valve is pushing against. This is the pressure pushing the check ring closed. Consider the sequence of the cycle. When the second-stage or hold timer times out, what does the controller tell the machine to do next? Within a heartbeat it tells the screw to start rotating.
Plastic is compressible, and even though the hold pressure will start dropping immediately, think about the torque required to turn the screw with something near 8000 psi pushing on it. Think about the grinding action between the check-valve seat and the check ring. Both are known to be high-wear items to begin with. Don’t add to the probability of significantly more wear for no reason.
Instead, program the controller to have a screw-rotate delay time of 0.1 to 0.5 sec to allow the pressure to bleed down. Or, program the controller to pull the screw back a fraction of an inch before screw rotation to accomplish the same pressure drawdown. Then start screw rotate.
While I am not one to design my process around being kind and gentle to the molding machine, doing this will help the check valve last a bit longer. I can’t quantify this, and it is not a variable to be computer simulated. But what you and I both know is how often these check valves need to be replaced or checked during preventive maintenance. Changing the check valve is a time-consuming and costly repair.
What are the benefits of this approach? If less torque is required to turn the screw, there will be less wear and a longer life for the screw motor. Decompression also allows the sliding ring to pull off the seat, and when the screw does start to turn you’ll get much less grinding action between the ring and seat. This translates into less wear on a high-wear, problematic piece of the screw assembly. The real benefit would be better cushion repeatability and more consistent parts—equalling a few more dollars in your pocket.
Take a Scientific Approach to Troubleshooting
Many consider injection molding to be an art form. While certain aspects of molding do remain somewhat mysterious, I believe that taking a scientific approach minimizes the “art” of our craft, which is particularly helpful when troubleshooting. My attempt here is be to establish a first step in troubleshooting the following part or processing problems:
•Non-return valve issues,
•Blushing, or other surface-finish problems,
•Sticking in cavity, core, or sprue,
•”Pin push” or ejector marks,
•Machine switchover response.
This is a fairly long list of issues, and with all of them you can get a significant amount of information to start a scientific approach toward a resolution by turning off the second-stage or pack-and-hold pressure. It just might be worth your time to try it. What have you got to lose?
For this first step to be of value it has to be done correctly. There are three methods commonly used for taking off second-stage or pack/hold pressure. You can take hold pressure off by either taking the time off the timer, reducing the hold-pressure setpoint to near zero, or doing both. It seems like you should get the same results with each of these approaches; but due to machine response and other factors, you will not get the same result with each method. Check it out.
On a tool that you are sure will eject the part if you make a short shot, do one trial with taking the hold time off, another leaving the time on and taking hold pressure down to near zero, and a third trial where you reduce both hold time to zero and hold pressure to near zero. You should see a different in the size of the parts. All three should be visibly short by volume (not by weight). If they are not all short it almost surely indicates that you have set the transfer point from first-stage to second-stage injection incorrectly. Further, if you are molding the same part in both an electric press and a hydraulic press, you need to check out both. You may be surprised at the difference between them.
You might have a few molds that require a full shot at the end of first-stage injection, but that should be rare. Extremely small or micro-molded parts are exceptions to the rule, but that is because they are so small that it is impossible to use a typical two-stage molding process. (Yes, you can still practice Scientific Molding and use only one stage. But that’s another topic.)
There is a preferred method for troubleshooting: Take the hold pressure down to a low level, perhaps 5 to 25 psi on a hydraulic press; and for an electric press set pressure as low as it will allow or 100 to 250 psi plastic pressure. It is important to leave a minimum of 0.3 sec on the second-stage or pack/hold timer. Why? Refer to the previous experiment. It should show you that the part made with time on the second-stage timer is larger than the others. This means you get the see the machine’s response on switchover and the momentum involved in transferring from first to second stage. Another way of stating this is that we get to see what happens during transfer if you do have second stage on. Most jobs run with some time on the hold timer and you need to understand the over-travel that often occurs in the first stage.
To illustrate the value of taking hold pressure off to obtain information about a part problem, let’s use an example from our list. Let’s say you have a part that is flashing, either sometimes or all the time. Most troubleshooting guides suggest reducing injection pressure. Sorry, this is simply wrong! With Scientific Molding you must have an appropriate Delta P on the first stage, so reducing first-stage pressure is not an option.
In troubleshooting flash, the first question is: Is it occurring in the first or second stage? How do you tell? Taking off the second stage via by eliminating time on the second-stage timer may not provide an answer, because you will not see the overtravel after the first stage. Note in the first experiment described above that the part with time on the timer is larger than the others. However, taking the hold pressure down to a low value and leaving some time on the second-stage timer will tell you what you want to know: Is it the first or second stage that is causing flash?
Ram overtravel could be causing the flash. If you see flash and the part is filled and packed with second-stage pressure set low, you are going in too far in the first stage. You need to re-establish your stroke transfer position to fill less material during the first stage of injection.
If you have a short shot by volume and there is flash, you have a parting-line or clamp problem—in other words, a tooling or press problem—but not a processing problem. There is very little pressure on the parting line when the part is short by volume. You cannot build much pressure on the parting line until the part is completely full for most parts. Thin-wall molding can be an exception.
If the part is about 90% full, with no flash, by ending the first stage on position with low second-stage pressure and some time on the second-stage timer, the problem causing flash during molding is too much pressure in the second stage, or you have a weak parting line. Try reducing second-stage pressure.
If this does not resolve the flash, get the parting line checked out, and do not be fooled by using bluing compound. A good parting line needs 3000 to 7000 psi to hold. Bluing will transfer from one side to the other with just 5 lb of touch force.
What’s more, the parting-line test has to be done with the mold mounted in a press, because molds flex with the platens. This phenomenon will not show up in a spotting press.
Pay Attention to Heat Transfer
Click Image to Enlarge
A mere 2 mm of lime, calcium oxide, hydroxide, carbonates, etc. in a cooling channel will nearly triple the cooling time required, dramatically extending cycle times and draining profits. Source: Optimizing Injection Moulding by Thermal Images by Willi Stenko.
Infrared images of a molded part before and after cleaning of heating/cooling channels in the mold. The increased productivity was equivalent to 2500 production hours annually. Source: Optimizing Injection Moulding by Thermal Images by Willi Stenko.
One of the most significant reasons molders have difficulty in reproducing a process, from run to run and machine to machine, is that the mold cooling is either not done properly, not done consistently, or is inadequately documented.
Cooling and coolant flow through cooling lines are not subjects for debate. This is a well-defined science that is taken seriously and used without question in every other industry that deals with heat transfer. It is a required course for many engineering degree programs, and if you find someone who ever took a course in fluid dynamics, watch their face as you bring up the memory. Nearly everyone who had the class hated it and could not wait to sell their textbooks. Why? It was a dull class where you get told what to do, were provided the equations, and then challenged to do all the calculations as a drill and proof of mastery of the science.
There was no debate of the validity of the equations or concepts, or whether they work or are practical. These same rules apply in a molding shop. People may not get hurt physically if cooling is not done right, but they certainly may be injured financially. Cooling is about 95% of your cycle time, so it deserves proper attention.
Let’s start with some of the basics on heat transfer:
1. Heat transfer requires a difference in temperature between the fluid and the metal cooling channel.
2. Maximum change in temperature, in vs. out, is 2° C (4° F).
3. You must have turbulent flow, as defined by the Reynolds Number; my target is 5000.
4. Because cooling rate can influence part dimensions and properties, you must duplicate the cooling rate from run to run, machine to machine, for the life of the mold.
The first step with any mold, new or old, is to find out the diameter of each water line, often easily determined by the NPT thread size. Once you know the cooling line diameter and the water temperature, calculate the minimum gallons per minute (GPM) or liters per minute (LPM) need to achieve a Reynolds number of 5000. There are free calculators on Google, though most are more complicated than necessary. I suggest setting up the Reynolds Number Equation using a spreadsheet with a viscosity table vs. temperature of coolant. Now sum up the required GPM for all the lines on each mold half. This will be the minimum GPM need for this mold. With a good spreadsheet this should take you fewer than 30 minutes. Remember, there are no ifs, ands, or buts about this minimum GPM required.
The real problem is ensuring that each cooling line gets the required GPM, and it has to be the same every time the mold is mounted in a press. It has to be regulated. This is not such an easy task, since coolant takes the path of least resistance on different size water lines, especially if you have baffles or bubblers.
How close is your shop to having appropriate GPM, and is it regulated for each line? If you do indeed have everything in order, what happens if there is calcium, lime, rust, or some biological goo built up the inside of one or more of the lines? Let’s not pretend this does not happen, as some of you have seen lines completely blocked. Lines are sometime so badly fouled they have to be gun drilled out to clean them. What does this buildup on the inside of the cooling line do for heat transfer? The difference in in/out coolant temperature, or Delta T, drops astronomically, hence longer cycles and poor profits.
Figure 1 shows how a mere 2 mm of lime, calcium oxide, hydroxide, carbonates, etc., will nearly triple the cooling time required. Figure 2 shows thermal images of a mold before and after cleaning the water channels. In researching his book, Optimizing Injection Moulding by Thermal Images, Dr. Willi Stenko investigated more 700 molds and found 70% of them had extended cycle times due to fouled water lines, resulting in a revenue loss of about $60,000 per mold.
Dare I say that instead of spending money on the next Six Sigma Training, spend it on cleaning your molds. That would be a better return on investment. Our industry does not pay proper attention to cooling, water treatment, and keeping cooling channels clean. No ifs, ands or buts about that, either.
Level Your Press in an Hour or Less
Your injection molding machine is leveled on installation, but how long does it take for it to drift out of level? How often should it be checked or re-leveled? For that matter, why bother leveling it in the first place? Isn’t the machine’s frame strong enough to keep things aligned?
Unfortunately, regardless of age, your machines need to be level to keep the platen parallel and the parting line of the mold in good shape—in other words to keep producing consistent, high-quality parts.
Let’s start with why the machine needs to be level. Upon clamp-up, the mold halves must mate perfectly and clamp evenly to keep the platen parallel and prevent flash, parting-line wear, and mold damage. If the machine is not level, the two halves of the mold will actually grind a wee bit on clamping, as they mate and reach tonnage. In fact, there will be a twisting motion as the two halves mate, and you’ll get parting-line wear. When the parting line wears and/or vents get peened shut, you’ll get flash, cosmetic blemishes, and/or burns on the part. Parting lines wear in normal production, but a machine that’s not properly leveled will accelerate the issue.
Part problems caused by an out-of-level machine often are not seen right away and are misdiagnosed when they are observed. Most processors will assume they are having an issue with the mold. They’ll bring the mold to the tool shop for redressing the parting line or for some other repair, only to put it back in the press and see the same part problems before the first or second production run is completed.
Do we agree now that it’s wise to have the machine level? OK, then how do we get it there…the first time? Usually it’s done this way: Someone places a machinist’s level across the tiebars near the stationary platen, and then levels the machine. Then the level is moved near the die-height platen, and the leveling feet are adjusted once more. Then the level may be placed parallel on a tiebar, with the leveling feet once again adjusted. Then each leveling position is rechecked, only to find that more leveling is needed. This approach can take hours, and still not yield the desired result.
Try another method: First, round up three machinist’s levels (yes, three), one torque wrench, and two people.
My suggestion is to buy a torque wrench and just one of the levels you’ll need. These things are expensive and most of the time just sit around. Try to find two other molders nearby with levels and borrow them when needed (and of course share yours).
Here’s my method:
1. Go to each leveling foot and clean and oil each of the leveling screws.
2. Of the two persons assigned to this task, pick the one with the most mechanical ability. Give him or her the torque wrench and get each of the leveling feet at the same torque. I wish I could tell you what number to shoot for but it varies with the type of leveling foot, weight of the machine, etc. The main idea is to get each foot to support about the same amount of weight. One needs a bit of mechanical sense to find the number for each machine/location. You might be surprised how level the machine is at this point.
3. Place one level across the tiebars (or guide ways, if they are supporting the platen), against the stationary platen. Place another level near the die-height adjustment platen, and the third parallel along one of the tiebars.
4. The person in the best position to see all three levels should now instruct the one who did the initial adjustment which feet to slowly move up or down until all three levels are level.
5. This will take 30 to 45 min, though big machines take a bit longer. Move on to the next machine on the list.
How often should this be done? I suggest at least once a year. There is no readout on the computer screen to provide information on the machine level; you have to get maintenance to check this
‘Auto Compensate’ Your Press To Handle Viscosity Variations
Most agree that the goal in molding is to produce identical parts, a difficult task indeed due to the hundreds of process variables.
Let’s discuss one that many complain about: material viscosity variations. Then let’s see what we can do to attenuate or normalize this variable from a processing point of view.
The viscosity of a resin can change for a number of reasons:
1. Moisture content. Certain hygroscopic resins, like polycarbonate and PET, will suffer chain degradation due to hydrolysis if not dried properly. It’s not just splay you need to worry about; the amount of residual water can greatly influence the viscosity of the material.
2. Lot changes. Though certified to have the same specifications, even the same melt-flow rate, resin lots can differ in viscosity when processed on an injection press. In fact, even though they have the same average molecular weight, the molecular-weight or chain-length distribution can be different.
3. Screw-rotate time. Since most of the energy to melt the plastic comes from the mechanics of the screw rotation and backpressure, changes in screw-rotate time cause viscosity variations.
4. Additive type and percent. Usage of color, fillers, mold release, flow aids, antistats, antioxidants, regrind, etc. can have a huge impact on viscosity. How well the mix is blended can also play a role.
5. Resin temperature. Higher temperatures result in lower viscosity, lower temperatures in higher viscosity. Nothing new here, but this can vary during processing due to hot-runner oscillations, heater-band function, etc.
6. Mold temperature. The resin behaves as if the viscosity changed if you have a hotter or cooler mold.
7. Injection velocity. Plastics are shear-sensitive and change viscosity dramatically with changes in injection speed. This is the “viscosity curve” and nearly all long-chain molecules exhibit this behavior.
It would be interesting to know which of the above has the largest influence on viscosity. If I had to pick one, it would be the injection velocity, or shear rate or fill time. Even though the resin can start out at a different viscosity, if driven to the same shear rate you can attenuate some of the viscosity variations and get a more consistent process. Simply stated—run to run, shot to shot, summer to winter—keep fill time the same and your process will be more consistent. Pick a fill time for the “life” of the mold.
So if you concur that fill time needs to be kept constant for all production runs, how do you accomplish this? Certainly there are differences of opinion in the industry. Some processors feel it is their job to adjust for these changes. However, is it really possible, plausible, or the right strategy to expect an operator to stand at the machine adjusting it as different variations arise? I prefer the strategy where the machine automatically adjusts, something like a car on cruise control. It can be done during first-stage filling, provided you set up the machine correctly.
Do not run pressure-limited, and make sure the machine is properly load-compensated. We will concentrate on load compensation for this article and assume you’re running under velocity-control conditions—that is, the machine’s first-stage injection has an “appropriate” amount of extra power (pressure, usually not maximum pressure), so that velocity is controlled by allowing the pressure to vary in response to viscosity changes. This will allow the machine to keep fill time the same, providing it is “load-compensated.”
What does “load-compensated” mean? Just that no matter what the load (viscosity), the machine will use the appropriate pressure to drive the distance from the start of injection to the stroke cutoff position in the same amount of time, just like on cruise control. So whether you have polymer, water, air, or ice cream in front the screw, you will get the same fill time. Testing this is simple: Shoot plastic into the mold and, using the identical stroke, drive air into the mold. That is a huge difference in “load,” and your machine should provide an identical fill time for both air and plastic.
CONDUCTING THE EXPERIMENT
With all safeties in place, get the machine up to operating temperatures and a steady-state condition. With second-stage hold and/or pack pressures reduced to near zero, adjust your shot to provide a 90% to 99% full part by volume, not by weight. It is critical that the machine is not pressure-limited and operating with an appropriate “Delta P” and that the part be visibly short. Ideally, you are using about 75% of the barrel. If using under 25% of barrel volume, results may not be of value.
Under semi-automatic or automatic conditions, mold a part or two. Record the fill time to hundredths of a second and the peak pressure during injection; it may or may not be the pressure at transfer—you have to know your machine. Ensure you get peak pressure. Now set up the machine to shoot air into the mold instead of plastic. In manual mode, bring the screw to the zero position, pushing all of the plastic in front of the screw out through the nozzle. You do not have empty the entire barrel, just ensure there is as little resin in front of the screw as possible.
Now retract the screw, without rotation, to the exact same start point as your shot with plastic. If the start point with plastic was 4.35 in., make sure this “air” shot is starting at 4.35 in. too. In semi-automatic mold, shoot air into the mold, again noting fill time and peak pressure. Without getting into equations, the fill times should be nearly identical; a general rule of thumb would be ± 0.04 sec for normal fill times. Results can be influenced by length of stroke, amount of pressure difference and long or very short fill times. The bottom line is that both shots, plastic and air, should have the same fill time and a large difference in peak pressures. Do this at four different velocities.
If the machine passes this test, you can be assured it will provide consistent fill times as the resin viscosity varies for any reason. Make this a requirement for the annual maintenance check of the machine.
How to Stop Flash
Click Image to Enlarge
Testing the parting line, using pressure-indicating paper: The image on the left is flash-free. The part on the right side is flashed; note the location of flash where the paper is mostly white at the top and bottom. The white color indicates the parting-line touch force is below 1500 psi. The paper does not start to turn pink until there is at least 1400 psi of touch force. So when I raised the hold pressure to really pack out the part, it flashed at the top and bottom. (You do not see the flash on top due to the angle of the shot.)
Flash has to be one of the top 10 problems molders face, causing rejects and a significant dent in profits. Troubleshooting flash is complicated, and often there is a lot of tail-chasing on the shop floor in trying to get a solution. Here is a strategy that should lead you to the solution via a data-driven procedure.
Flash can appear on the part’s edge along the parting line of the mold, or anyplace where the mold has metal meeting metal to form a boundary of the part. Flash can occur for several reasons, including variations in the process, mold issues, machine problems, or material issues.
First, I’ll assume: 1) You have checked the parting line carefully and that it is clear of any foreign material; and all slides, etc. are clean and clear of flash or buildup of residue. 2) You are running the process under velocity control and not pressure-limited.
To start, find out if the flash is occurring in the first or second injection stage. Here’s how: Take off second-stage or pack and hold pressure by reducing the pressure to a very low level—shoot for no more than 500 psi plastic pressure. It is important that you do not take off the second-stage or hold time. Check out the part after injection under these conditions: It should be visibly short at the fill time you established and at the normal transfer pressure. If it isn’t, readjust the cut-off position to make the part short.
Once you’ve made a short shot, look for flash. If the part is flashed you have evidence that the parting line is most likely the problem and needs repair. It is difficult, if not impossible, to process around a weak parting line. If you’re running a multi-cavity tool, make a short shot where all parts are short. Be careful—with some tools, shorts may not eject. Use mold release if necessary. If the short shot shows that cavity filling is non-uniform, this means packing cannot be uniform. This is why parts from multi-cavity molds may experience flash in one cavity and sink marks in another on the same shot. A previous column (t)see Editor's Picks top right) discusses this problem.
Is the parting line is OK? To check this, don’t use bluing compound. Instead use a special pressure-sensitive paper that develops a red color relative to the amount of pressure developed. The parting line may be perfect on a bench press but not tight when clamped by the molding press. Clamping in most machines exhibits something known as platen wrap. The platens literally bow around the mold. The photo shows how bowing manifests itself in the part.
The picture on the left is flash-free, the parting line is mating all the way around and the vents are open. The indicator paper tells us the side parting lines provide a touch force of 5000 to 6500 psi, as indicated by the deep red color. The part on the right side is flashed where the indicator paper is mostly white at the top and bottom. The white color indicates the parting-line touch force is below 1500 psi. The paper does not start to turn pink until there is at least 1400 psi of touch force.
So when I raised the hold pressure to really pack out the part, it flashed at the top and bottom. The mold works perfectly on the bench, but when clamped in a machine it cups 0.001–0.003 in., and it needs preloaded support pillars at the top and bottom to prevent the development of flash at higher second-stage or hold pressures.
Clamp misalignment can also cause flash. Check machine leveling and clamp parallelism, which can also produce flash.
If you get no flash on the shorts from first-stage injection only, you have established that flash is occurring in the second stage. It still may be a parting-line issue, as shown in the photo, but the indicator paper provided the data to answer that question. We have reduced the complexity of finding the root cause—it has to be something related to hold pressure. There are only a couple of exceptions to this conclusion: thin-wall molding, where cavity pressures can be very high in filling to cause flash; and micro-molding, where parts are so small that you cannot separate fill from pack (i.e., there are no first and second stages).
Going back to the point where you have shorts and no flash, now start adding hold pressure, starting low, with about 500 psi plastic pressure in the nozzle. As you raise the second-stage pressure, watch for when and where flash shows up. If flash develops with low pack and hold pressures, this again suggests the culprit is the parting line, clamp, or press leveling.
If the flash is in the center of the mold, it may be due to inadequate mold support. Molders should consider whether the mold has enough support pillars in the right places for the cavity and core plates. Molds deform upon clamping, and to combat this they are often built “preloaded” so that the center support pillars are slightly taller than others.
The sprue bushing is another possible source of flashing. Nozzle contact forces can range from 5 to 15 tons. If thermal expansion causes the bushing to “grow” far enough past the parting line, the nozzle contact force can be enough to push on the moving side of the mold, trying to open it. For non-sprue-gated parts, molders should check the length of the sprue bushing while it is hot.
Next, let’s examine clamping. If the flash is concentrated toward the center of the mold, it could be caused by too-high clamp pressure. If a small mold is mounted on a large platen, the forces on the mold can be greater on the four outside corners than in the center. The excessive clamp pressure may tend to “wrap” the platens around the mold. To solve the problem, be sure that the mold takes up about 70% of the distance between the tiebars.
Flash can also be caused by too little clamp tonnage. If the part passes the first-stage short-shot test without flash, then the pack/hold phase may be pushing the parting line apart. Increasing the clamp tonnage may be the solution. That’s especially true of high-speed filling in a thin-wall application, which can require 35,000 to 60,000 psi plastic pressure in the nozzle.
If material viscosity is too low, flashing can result. Resin can become too “runny” for a variety of reasons: melt temperature too high, excessive residence time that causes degradation, too much moisture in moisture-sensitive resins like nylon, or excessive amount of colorant or other additive that contains a lubricating vehicle. You can take the temperature of the melt at the nozzle with an infrared detector.