Allan Griff, independent consulting engineer and plastics extrusion expert, has agreed to extend his relationship with PlasticsToday. Already the host of our well-received Extrusion Expert webinar series, Allan now also authors this weekly column. The formula is simple: you send questions, and Allan chooses the best or most intriguing and answers them.

Our aim is that this column becomes the leading source of extrusion information on the web, but don't shy away if your question is focused more on materials than extrusion. Have a question for Allan? Send it to him at [email protected]. We will not publish your name or company name unless you expressly ask us to do so, but also will not accept anonymous questions as this opens the opportunity for a supplier to send a "baiting" question for which the answer might highlight his products.

Catch Allan "live" this Thursday, Oct. 13, at 14:00 Eastern when he offers another spirited presentation, followed by a live Q&A, to help you better understand melt viscosity, why it matters, how it can be misleading, and how you can test and influence it so that you and your company can produce more saleable parts. More than 300 attendees from Allan's previous webcasts already are registered; join them and us and you'll realize a very positive return on your 60-minute investment. Register for the free webcast here.

To this week's Q&A:

Q: Why does Delrin/acetal have more percentage of shrinkage when injected than nylon 6/6?

A: I don't have the figures, but I suspect that the difference between melt and solid density for Delrin is larger. This also depends on melt temperature and molding conditions (mold temperature, injection pressure, packing).  Du Pont, who supplies both materials, should have the actual figures to compare.

 
Q: Is there any epoxy that bonds to PBT?

A: Permabond has developed an epoxy adhesive ES5741designed to bond PBT. The web link is http://source.theengineer.co.uk/production-and-automation/welding/joining/adhesives/bonding-equipment/permabond-introduces-pbt-bonding-grade-structural-epoxy-adhesive/2000325.article

The company is based in the U.K., with U.S. agency in NJ, phone  (732-868-1372). My mention of them should not be considered a preference; I don't know anyone there, but had reason to deal with them for another tech problem.

You may also find some interesting information in a paper given at SAE 2007, http://papers.sae.org/2007-01-1517/. Two of the authors are from Denso, a huge automotive parts maker, with operations in USA and elsewhere.

Q: How does a black speck from molecular breakdown due to heat look different than a black speck caused by contamination?

A: Heat will first make stagnating resin yellowish or brown, and that may stay discrete and become a round "gel" in the extruded surface. Eventually, it will degrade resin that is stuck in slow-moving areas, and turn it to carbon, which eventually flakes off and gets into the product. Such flakes come in "bunches" and typically have sharp edges. If they are larger than the screen meshes, they were formed after the screen.

Contamination is a general term that could include undispersed colorant or filler, fibers from paper bags or other fibrous material that got into the feed, as well as the carbon flakes mentioned above, and even incompatible resin left in from a prior extrusion. In the latter case, it will start out at maximum and eventually go away. If the contamination is there from the beginning and stays there, it may well be coming from the supplier. A small extruder used to test incoming raw material can be very useful in this regard, as well as providing viscosity data to aid later production.

 
Q: For crystalline thermoplastics, can the degree of crystallinity be manipulated during processing (molding/extrusion)? What are benefits of more/less crystallinity?

A: Firstly, we assume that the material is completely amorphous as a melt (there are rare exceptions). The primary determinant of crystallinity is the molecular structure of the resin itself; HDPE is more crystalline than LDPE, for example. However, the degree of crystallinity of the finished product can be manipulated by the rate of cooling, as well as the presence of nucleating additives. Remember that it is not only the total percentage of crystallinity that counts, but also the size and shape of the crystallites - many smaller ones will produce different properties than fewer large ones.

Benefits of crystallinity include ability to lock in orientation, greater rigidity, and often the realization of greater heat resistance and tensile strength. The rigidity may be either a plus or a minus, depending on the application. With PET, crystallized feed is needed to avoid sticking and clumping at dryer temperatures.

The detriments include the cost and operation of the equipment needed to crystallize and (if done) to orient, and usually more haze (except where crystallite size gets below the wavelengths of visible light). This counters the image of the word "crystal," which one would expect to refer to a clearer product, however, the refraction of light as it enters and leaves the crystallites causes haze that is absent in amorphous resins such as PVC and unmodified PS.

Q: Why are some knit lines stronger than others?

A: In both molding and extrusion, the knit depends on time in contact (the longer the better), the melt temperature (hotter the better), and the pressure (more = better).  The mold design has an effect as the flow is constantly being cooled and if the surfaces are too cool on contact, they won't knit well.  

In extrusion, this is an issue with all hollow items such as pipe and film. In pipe, the die must be long enough and the supports of the core (spider arms) small enough to promote rewelding. Some specialized designs are available. In film extrusion, most dies have spiral mandrels, which effectively solve the problem, although these still may leave visible lines (port lines) down the film.  

Q: Can you please comments on HIPS, specifically its physical and thermal properties?

A: I'm glad you asked this question, as I have a lot of experience with HIPS (high-impact polystyrene) that has gone unused recently, as there has been so much emphasis on the polyolefins and, lately, bioplastics. HIPS is one of the easiest materials to extrude and form, as it has a broad softening range and its melt viscosity changes less sharply with temperature than some of the more crystalline plastics.  

Unfortunately, its price has climbed faster than PP in the last decade or so, and some of its former uses have been lost to that resin. Physical properties depend very much on how much rubber is included (that's what gives it the high impact strength). From 6-8% rubber content is typical for standard HIPS, but lower and higher percentages are sold, and they are usually compatible with each other, so you can get any impact strength you want by blending.  

Boiling water is close to the limit for HIPS; we used to test heat-resistant grades by pouring boiling water in a thermoformed cup. Heat resistance was achieved by keeping softening additives (including residual monomer) and even the rubber to a minimum, and using a co-monomer if necessary. The usual comparative tests are Heat Deflection Temperature (D648) at 264 psi and Vicat (D1525).

HIPS is flammable, and like all styrenics it gives off a sooty black smoke when burned.  It can be made flame retardant with additives.

The odor depends mainly on the residual monomer, which can vary with supplier, and also can be further minimized by vacuum venting in processing.

As for physical properties, this is a tradeoff of rigidity with impact strength - less rubber is more rigid but less tough. Some additives have an effect as well, notably certain fillers which raise rigidity, but the rubber content, rubber type and method of polymerization all matter, too, so just comparing at equal rubber level isn't enough. Mechanical properties are shown by the tensile/elongation/modulus test, D638, and toughness by the notched Izod, D256. In these tests, remember that the results depend on the orientation of the molecules; injected specimens will therefore look tougher/stiffer than extruded ones, and in extruded sheet, it is important to test specimens cut in cross as well as machine direction. Thickness counts, too - 1/8" specimens will look better than 1/4" specimens.


Q: How would you define "polydispersity ratio" and how it affects processing?

A: The polydispersity ratio is the ratio of the weight-average molecular weight to the number-average molecular weight. It is sometimes referred to as molecular weight distribution, but is not quite the same. The weight average considers the weight of each molecule, somewhat like the House of Representatives in the U.S. Congress considers population. Molecules that are bigger (longer, more weight) count more than the smaller ones.

The number average just counts molecules, considering them all equal, like the U.S. Senate has two senators for each state no matter what the population. The little molecules have more effect on the number-average figure than on the weight-average. For typical film-grade polyethylenes, the polydispersity ratio is usually between 1.5 and 2.0.  It is often said that narrow distribution (low ratio) is better for clarity and tensile strength, and broad distribution is easier to extrude (less sensitive to temperature variation, more stable bubble) but there are so many other factors such as blow-up ratio, melt temperatures, cooling rate and air temperature, and line speed, that it is hard to compare materials on this basis alone. Furthermore, it is a ratio, so that two materials with very different molecular weights can still have the same polydispersity ratio but behave quite differently.

Finally, the ratio may not consider the bimodality. Bimodal resin (with two peaks of molecular weight) may have the same values as a monomodal resin, but may draw more easily or have other advantages. It is best to have a small test extruder available and compare resins by running actual samples wherever possible, under conditions which relate to production conditions, rather than rely mainly on such basic molecular data.