“The Seven Stages of Rotomolding” Questions and Comments

As part of ARM’s on-going commitment to member service, we are ramping up our webinar program for 2020.  The latest webinar was presented on March 19 by Dr Gareth McDowell of 493K.  I thoroughly recommend that, if you missed the live show, you catch up with it on the ARM website.

Many of you will be familiar with Dr Gareth’s lively presentation style from his many presentations at ARM conferences. He was able to bring this approach to the very different format of an on-line event and, as a result, we saw a high level of reaction from the live audience, in terms of comments and questions.

Unfortunately, we simply ran out of time to address everyone’s needs, so I’m doing a wrap-up via my Technical Director’s Blog.  The length of my Blog reflects the number of questions, but feel free to dip in and out of it, if you don’t have time for a long read!

Thanks again to Gareth for a really excellent webinar.  

Comment: Additional Benefits of Temperature Measurement Traces

Gareth showed a typical plot of temperature versus time, see below:

The thick red line shows the trace of the internal air temperature (IAT), which illustrates the seven stages of rotomolding that he was describing.

A molder who regularly uses this technology commented:

“As well as the IAT trace, we also find the continuous temperature measurement OUTSIDE the mold to be of considerable value”

This is the thinner, purple colored, trace shown on Gareth’s chart.

The molder continued:

“ from experience, we know what this outside air trace will typically look like for our particular oven.  When we see any deviations from expectations, the outside air trace gives us a clear idea that something has gone wrong.  Recently we were able to spot a minor burner glitch that, if not addressed, could potentially have given us scrap parts”.

Question: Effects of Over- & Under-weight Shots

One audience member asked how under- or over-shot weights would affect the IAT trace.

If shot weight varies, you will notice several issues related to processing:

1. Shot weight over target.  At the same oven/cook time conditions, the part will be less well cooked.  If you are monitoring internal air temperature (IAT), it will take longer to reach your target IAT than in previous runs.
2. Shot weight under target.  At the same oven/cook time conditions, the part will be more cooked.  If you are monitoring internal air temperature, it will take less time to reach your target IAT than in previous runs.
3. Well formulated roto grades (e.g. most of the materials available in North America) offer a reasonably wide processing window.  Effectively this provides a degree of freedom for the molder; cook conditions can vary and the results (in terms of aesthetics and physical properties) will still be OK.  However, the width of this processing window can vary considerably, even for grades that are similar in MI and Density.
4. In summary, it makes lots of sense to keep everything as consistent as possible and not to rely on a wide processing window in day to day operations.
5. The first thing you’ll probably notice with an over-weight shot (resulting in a relatively under-cooked part) is an increased incidence of surface porosity (aka “pinholing”).  As Gareth explained in the webinar, eliminating bubbles is a function of both time and temperature (see also Section “Where do the Bubbles Actually Go?”).
6. The first thing you’ll probably notice with an under-weight shot (resulting in a relatively over-cooked part) is discoloration of the inside surface.  This doesn’t necessarily mean that the part is degraded; many heat stabilizers continue to protect even when there is a noticeable color change.  However, you are getting closer to the point where degradation does occur, after which you can expect sudden and catastrophic loss of impact strength.  My R&D team used to describe the onset of degradation as “falling off a cliff” – a bit melodramatic, but descriptive.
7. Modern weighing scales are accurate, so under- / over-in manual weighing is more about carelessness than lack of equipment specification.  Despite digital units with a tare facility, weighing errors are surprisingly common. Check-weighing your finished parts is a simple way of checking that your weighing procedures are working well.
8. Variations in shot weight may also affect cooling time.  If you’re trying to demold at a specific IAT (see Section “Ideal Demold Temperature”), you may notice this; if you’re not, you may well not notice.
9. Under-weight parts will be thinner, over-weight parts will be thicker.  ARM low-temperature impact strength is significantly affected by part thickness.  In terms of stiffness, remember that a 10% increase in thickness will, theoretically, result in a 33% increase in bending stiffness.

So – in summary – variable shot weight is a variation that is unnecessary and can affect process and part properties quite significantly.

Question: Rotation Speed

Gareth made the point that the optimum rotation speeds for the mold can depend on the oven temperature.  Several audience members asked for clarification.

A high oven temperature will mean that the inside mold surface will heat up quicker and therefore powder will start to stick and lay down earlier than with a lower oven temperature.  In order to make sure that the powder inside the mold experiences the same number of rotations as it would at a lower temperature, it would be logical to speed both rotations up.  It would also be logical that you should maintain the same rotation ratio as you had previously.

When you look at the IAT trace, you can often see the effect of the movement of powder inside the mold, especially at an early stage in the cook cycle.  The “noise” that you see in the IAT trace (ie the small fluctuations) is an indicator of the current position of the powder relative to the thermocouple being used for IAT measurement.  When the relatively cold powder draws close to the thermocouple, it momentarily cools the air and the IAT drops.  As the powder moves away from the thermocouple, this cooling effect diminishes and the IAT rises.

The behavior described above is especially apparent in “rock & roll” machines; there is often a series of discernible “humps” in the IAT graph; these can be directly linked to the rocking motion.  I know one kayak manufacturer who uses this to set up his cook conditions for new molds; experience has taught him that he needs to see at least 8 “humps” in his IAT curve to make sure that his thickness distribution is on target.

Question: Over-cook & Degradation

Gareth discussed avoiding under- and over-cook, aka correct cure during the webinar.  Coincidentally, I had previously received a member query on the same topic.

Quite a bit of the detail on this topic has been given in the section “Effects of Over- & Under-weight Shots”, but there are a few more points that may be useful:

If you don’t regularly monitor IAT in your mold, I doubt that you can really be sure that you’re achieving correct cure.  You certainly shouldn’t totally rely on the fact that your material supplier is providing you with a wide enough processing window that you can be cavalier about this.  Things change and you need to avoid getting caught out.

In my several decades of observing industrial rotomoulding operations, I would say that the majority of molders tend to under-cook, rather than over-cook, their parts.  This is understandable; we all need to get as many parts out of a shift as possible.  Just bear in mind that the aesthetics of your parts may be worse and that you may not be getting full value from your material, in regard to physical properties.

There is also a sizeable minority of molders who tend to over-cook; mostly because they are trying to get a pinhole-free surface and they know that increasing the cook will help drive out bubbles from the part (see section covering “Where do the Bubbles Actually Go?”).  These molders need to bear in mind that they may be approaching the point where their material properties (especially impact strength and brittleness) suddenly drop in a catastrophic manner.  You can’t always rely on the color of the inside surface and the smell to guide you here; the reaction of different (but similar) polyethylenes to over-cook can vary considerably.

Incidentally, molders have often asked me what that characteristic “over-cook” smell actually is.  In the past, I’ve told them that it’s the reaction products from degradation of the polyethylene.  But what actually is it composed of?  Quite recently I read a (non-rotomolding) technical article that said that these off-gases were various types of ketone, especially aldehydes.  So now you know!

Question: Ideal Demold Temperature

This was another question from the seminar that we didn’t have time to answer.

It’s well worth trying to initiate your demolding at a particular IAT.  This will help you avoid too much shrinkage variation.  Unfortunately, you can’t entirely rely on a simple time measurement for this.  If the ambient temperature of the surroundings is relatively high (eg a summer’s day), it will take longer for the mold to cool to a specific temperature.  In the middle of a Minnesota winter, you would expect molds to cool faster!

If you can’t measure IAT, you could use an infra-red gun to check the mold temperature and demold when it reaches a set temperature each time.  This won’t be as good as IAT, but it’s certainly better than nothing.

In terms of setting a target IAT for demolding, I would try a value around 200°F.  If you have a more complex shaped part, particularly one with undercuts and cores, you may need to go lower.  Trial-and-error will show you the best value for your particular set-up.  The key is to try to be consistent.

Question: Where do the Bubbles Actually Go?

Disappearance of the entrapped air bubbles during cooking (what Gareth called the “consolidation” step) is a function of temperature and time.

When I was young and foolish (many years ago), I assumed that the air bubbles somehow migrated to the inside surface of the part, while the surrounding polymer was still molten.  Perhaps I had spent too much of my time staring at the rising bubbles in my beer glass and it seemed logical enough!

However, when you take account of the extremely high viscosity of molten polyethylene, you will quickly appreciate that this process would take literally years to complete.

So what’s really happening?

The trapped air bubbles contain… literally… air.  A mixture of gases; predominantly nitrogen, some oxygen and some trace gases like carbon dioxide.

At the temperature of the melted polymer, these gases will dissolve into the polymer matrix.  I would expect the carbon dioxide to dissolve first, the oxygen to dissolve next and the nitrogen more slowly.  However, given time, they will dissolve and disappear.

The higher the temperature of the melt, the faster the rate of dissolution of gases.

If you stick a decal or a label on a finished part, as soon as it’s cooled, you’ll probably see air bubbles accumulate under the surface in the first few days.  This is caused by these dissolved gases coming out again, because the environment is now much colder and the polymer matrix can no longer accommodate the same level of dissolved gas.

Dr Nick Henwood serves as the Technical Director for the Association of Rotational Molders. He has 25 years-plus experience in rotomolding, specializing in the fields of materials development and process control. He operates as a consultant, researcher and educator through his own company, Rotomotive Limited, based in UK.