Andy Routsis: August 2009 Archives

A friend emailed me this great question the other day regarding the relationship between fill speed and pressure...

The pressure required to fill will increase because the viscosity change will not completely outweigh the pressure losses... The overall energy consumption does drop considerably during fill since the pressure to fill is being appllied for only one third the amount of time!

It is very likely that a rheology curve would demonstrate that the 1 in/s fill is on the left hand side of the shear thinning transition region and the 3 in/s is on the right hand side of this region.

Think of it in vehicular terms… Your 1 in/s is like pushing a large pickup (higher resistance to movement) at 20 miles per hour, while your 3 in/s is like pushing a compact car (lower ristance to movement) at 60 miles per hour. It takes more gas per minute to move the smaller car at the higher rate, but it takes less time and you consume less gas overall getting to your final destination.

Always keep in mind...the reduced viscosity due to shear thinning will actually save you money... making your processes much more economical since the overall energy consumption to fill the mold as well as the time are reduced.

Additionally, the drop in viscosity will also reduce the pressure required to pack out the mold cavity during 2nd stage. In the long run, all these changes can make a big difference in the productivity and efficiency of your facility.

-Andy

Yes, the goal in PET processing is to (1) maximize output, (2) control shear heat, and (3) minimize melt tempertaure... yet ensure the polymer is fully melted and as homogeneous as possible.

(1) The need to maximize output is due to the fact that most PET injection molding processes are for high-cavitation high-speed molding PET preform machines.

(2) The need to control shear heat is due to the fact that a higher temperature polymer requires more time to cool... and may increase the development of semi-crystalline regions. Also, inconsistent melt temperatures will result in inconsistent cooling and crystallinity... resulting in profit losses due to scrap and troubleshooting.

(3) PET also has very poor thermal conductivity... making it great for drink containers, but terrible for part cooling. The goal in processing PET is to process the material as close to the melting point as possible, while ensuring all the pellets are melted and mixed.

To address all these concerns, different machine manufacturers have devised some great machine and screw designs.

Whether it is a reciprocating screw machine, or a two stage 'shot-pot' machine, most PET single screw designs incorporate a barrier screw. In such a design, the screw channel depth remains the same from the feed into the transition zone, but the channel width becomes progressively smaller. As this channel width decreases... an adjacent shallow channel, separated by a barrier flight begins. Ultimately, the melted material flows over the barrier flight into the shallow channel until it transitions into the metering zone. Such a design reduces shear heating and allows stingy pellets more time to melt.

In most single screw designs, a distributive mixing element is typically incorporated on the screw or a static mixer is placed within the nozzle to ensure a more homogeneous melt... especially if a colorant is added to the polymer.

Some of the more ambitious machine manufacturers are working with twin screw designs for two stage molding machines. These systems show great promise since a good twin screw design would provide the best control over both shear and melt temperature.

PET processing is one of the more advanced and refined fields of injection molding, making it very competitive. I strongly suggest anyone getting into PET molding do some significant research to ensure your strategy is well suited to take advantage of today's advances.

-Andy

Classroom training is not the most effective method available for training your production employees. Some seminars provide 4-8 hours of training a day... which is far too much information to absorb in one week...never mind in one day!

Lengthy classroom training sessions tends to be very disruptive to the production environment. Sending employees out for training can often cost 1-2 times more than the training itself. Likewise, sending a consultant to your plant for on-site training classroom training can be both costly and distracting to your employees. In either case, you still have the underlying issue where the employees do not learn well in the classroom environment.

As an adjunct professor at the Plastics Engineering Department of the University of Massachusetts Lowell, I have successfully blended labs, interactive training, simulation software, product testing, new technology, and outside exercises. Since these classes are comprised of plastics professionals, this helps ensure my participants actually learn the information and understand how it applies to their role as production employees. At Routsis Associates, I have also used this approach successfully in distance learning situations using a variety of media such as interactive training, on-the-job exercises, testing, and short 1-hour training sessions.

As I blogged previously, ask yourself... 'What Do You Hope To Achieve Through Training?'. To paraphrase... training should be purchased to help meet and sustain specific employee development and competency goals. Keeping this in mind will help you evaluate whether a training method develop competent employees, or just give them a bunch of information.

-Andy

I very often hear people justify not training their employees with the age old argument 'If I Train My Employees... They Might Leave.' One of the funniest responses to this I have heard is from Zig Ziglar who says 'If you don't train your employees... they might stay!'.

All humor aside, ignorance is never a good employee retention policy. For example, one company we worked with had a 75% turnover. This meant that 3 out of every 4 employees leave each year. When we discussed employee trianing, it turns out that they never train, nor do they promote from within. They went silent when we asked them... 'So, am I correct to understand that everyone at your company is in a dead-end job'.

Unless your company has a true culture of learning and advancement, sending only one or two people to specialized training provides them credentials not applicable to your production environment. For example, '5S' training provides great information... but sending one employee out for a couple days of training will not transform the attitude and actions of your entire workforce.

Ask yourself two questions... (1) do you have enough 'key' people on your production floor...? and (2) are any of your production employees in 'dead end jobs'...?

The goal here is not to sell...  but to help people understand that you need a plant-wide focus on improving the skills of the entire workforce so ultimately, everyone shares the load.

-Andy

The occurrence of cracking and crazing in this case is most likely due to the factors of molded-in stresses and strain.

Since polysulfone has a high degree of semi-crystallinity and rigidity, the resulting parts have many molded-in stresses... especially in the case of a boss. Many injection molders will anneal the parts with a heat cycle to reduce these molded-in stresses. Doing this before and after the insertion would significantly reduce these stresses.

Reducing the strain during insertion will also reduce the likelihood of cracking and crazing. Using a barbed, threaded, or trilobal insert may reduce this strain. The use of ultrasonic insertion or in-mold insertion will also reduce this.

It may also be a good idea to mold the part in a translucent material so that you can see where the stresses are present under polarized light.

-Andy

The presence of voids in the sprue is very common, but the need to troubleshoot this is unique to sprue-gated parts. Voids in a sprue gate are a little different than a typical sink or void at the gate.

There are three ways to approach this... Cooling, Pressure, and Temperature.

The first way is to improve cooling at the sprue with additional cooling in the sprue area or mold cavity at the gate region. Beryllium copper inserts or a sprue made from an alternative material would improve the cooling in this region.

The second way to reduce this would be to increase the pressure applied during second stage time. As this pressure is increased, the time required to seal the gate will actually increase. Since the increased pressure causes a higher tendency for the polymer to flow back through the sprue.

The third way to improve the situation would be to increase the mold temperature... assuming sinks in the region are acceptable. Since higher mold temperatures increase the tendency to for sinks over voids.

Such a situation is difficult to correct... and typically requires much trial and error to determine a correct process. Again, as with any other process, you really need to document the process thoroughly once the issue is resolved.

-Andy

With I am asked how many vents should be added, I always say more. Each material has design guidelines for vent depth which are typically very helpful.

Ultimately, the more vents you can have on the part, the better. Obviously... strategic venting, such as a porous steel in a boss, can be very important; but getting as much gas out of the mold prior to that boss will minimize the amount of localized venting that is required. Also consider venting the runner, sprue, and cold slug wells.

Since the polymer only flashes as a result of thickness, the width of the vent has virtually no effect. As a result, you can make the vents relatively wide. I have seen large vents taking up high percentages of the parting line with great success.

One commonly overlooked aspect to venting is the vent drop. Behind any well designed vent is a substantive vent drop.  These drops are deep grooves, which channel the gas from behind the vent to the outside the mold. Think of the vent as a gate designed to transfer the gas from the mold to the vent drop.

Lastly, high clamp pressures may reduce the effectiveness of your vents... and may close them completely.

Many mold designers put extensive thought into the getting the polymer into the mold... but pay little attention to how the gas will get out of the way. Most of the molds I see with vent clogging problems often have thin vents, too few vents, no vent drop, or combinations of all three. Give the gas trapped within the mold many wide vents with a big vent drop to help it escape to the atmosphere.

-Andy

I typically ask for more specifics, but in this case it is more about the philosophy behind making the change.

Whenever making wholesale changes to the tool steel, you must take a scientific approach.

An example of this approach would be as follows...

1. Verify the melt temperature

2. Verify the mold temperature

3. Check the vents (you can often reduce the tonnage or place a piece of tape on the parting line to test if the venting is adequate)

4. Check the nozzle and hot runner system for any obstructions... or improper sizing. Depending on the application, you can perform a pressure loss study, actually dismantle the components, or try replacing them with different sizes.

5. Perform an in-mold rheology study (the gate may be too small if the mold cannot fill after shear thinning takes place... but, if there is no shear thinning present on the rheology curve... the gate could even be too large!!!) I strongly recommend performing these in-mold rheology curves at both higher and lower melt and mold temperatures as these conditions can influence shear thinning and viscosity.

6. After obtaining all this data, and is the problem does not resolve itself in the process of discovery, you should review the facts with the designers and processors to determine the correct course of action.

As a engineer... I always prefer taking a scientific approach to the resolution to any defect. This prevents rash and costly decisions from being made.

-Andy

A material such as this is a flame retardant grade of PC/ABS. By itself, PC/ABS can be a difficult material to process as it tends to have a very narrow processing window. What many people do not realize is the thermal instability of flame retardants.

It may seem counter-intuitive... but flame retardants tend to have extremely high degrees of thermal instability! These flame retardants which were added to the material often degrade easily due to shear, temperature, and moisture.

Initially, the material should be dried, even if it comes in sealed bags... and you should be molding the part in a machine which is using 40-80% of the shot size to ensure a very low residence time. 

Next, measure the actual melt temperature and the dewpoint of the material at the feedthroat and then compare these to the recommendations. You should also reduce the screw speed & back pressure, and increase the screw delay to reduce shear and barrel residence time.

You can always increase the number of gates on the parting line... since more gates don't cause flash. If you instill a schedule for the routine cleaning of the mold, this will also help prevent the gas buildup.

An in-mold rheology test will best determine when shear thinning occurs during fill, as well as visually demonstrate the fill time where the rate of filling degrades the polymer.

When your process is stabilized and functioning properly... you really need to document the process based on the process outputs such as melt temperature, fill time, part weight at end of fill, plastic pressures, etc.

The science behind today's polymers can create some materials which really perform great in their applications, but require very delicate processing. When the correct process is reached it is critical that you document the process and not just the machine settings so it can be repeated.

-Andy

In a recent webinar, I received this question form one of our participants...

It depends... but typically larger machines should be levelled every 6 months. The smaller machines tend to be more stable and rigid, requiring levelling every 12 months.

What is even more important is the fact that you need to measure any machine that is new to your facility every 3 months for the first year. Many molders assume that newer machines require less attention, but it may take up to a year for the machine to settle in.

Keep in mind, you should always level the machine by placing the level on the tie bars. Although bubble levels can be used, a laser level is significantly more effective at measuring the levelness of the machine. Clean off the tie bar before measuring and use a grooved level so that it properly rests on the tie bar.

-Andy

A customer of ours asked this great question the other day...

Assuming you are not 100% full during first stage… the 10,322 psi calculation would be more correct.

(10,322 lbs/in*in) x (19 in*in) / (2000 lbs/ton) =  ~98 tons.

This is because the pressure is not distributed across the mold cavity until the mold cavity is full... assuming mold filling is completed during 2nd stage packing.

Ultimately, the pressure losses that occur during fill actually reduce the actual pressure the mold cavity realizes so the 98 tons calculation would actually include a fudge factor for safety.

If you were to take a more exacting approach to this calculation, you could preform a pressure loss study to determine the actual pressure loss through the nozzle and sprue as well as during fill. From this pressure loss data you could estimate the average pressure distribution across the mold cavity and relate it to the 2nd stage packing pressure distribution.

In most cases, the simple approach used above would be satisfactory for the typical custom molder.. especially since it would accommodate a small fudge factor to compensate for variability and machine inconsistency.

Molders who perform fewer mold changes... or are purchasing a machine specifically for an application should perform a more detailed pressure study.

-Andy

A contributor at engineering.com posted an optimistic blog entry entitled ‘Time to Learn; Time to Save’. This entry addresses employment and professional development during recessionary times.

Corporal Willy (excerpt) 

The entire entry can be read here:

-Andy

During the NPE 2009 trade show, an attendee asked me an insightfull question...

I answered, if you’re serious about improving the skills of your production workforce, you need to begin by asking yourself a few questions:

1. What training methods have you used in the past? Were any of these methods effective? Which metrics did you use to measure their effectiveness?

2. What about now...how much time do you devote to training? Which training methods are you currently using and are they working?

3. More importantly, what effect has the current training plan had on your bottom line?

4. Lastly, what are your training goals? What do you hope to achieve by training your workforce? 

Training needs to take place regularly, and time needs to be allocated for this purpose. Proper training requires resources, so it is essential to have the support of management when establishing a training system for your workplace. Make sure that management, as a whole, understands the scope of the training plan - and which metrics can be used to verify its effectiveness.

In the end, training should be purchased to help meet your company's specific employee development goals. 

-Andy