The draw resonance phenomenon

Thickness variation in extrudates can be caused by many trivial factors such as inhomogeneous materials, fluctuation in the feeding rate of the extruder or in the melt temperature, surging of the extruder and lack of mechanical stability and stiffness of the equipment. A more fundamental reason for surface roughness or thickness variations is melt fracture, which appears suddenly as soon as a certain critical extrusion rate is exceeded. 

The thickness variations due to draw resonance have a different character. The thickness may vary by a factor of approx. 2, but there is no surface roughness. The transition from thick to thin r ons is very gradual and occurs at regular distances there is a certain wave-length , which may be as large as one metre or more. This defect has been called draw resonance by Bergonzoni and Dicresce because it is only observed when the draw-down ratio exceeds a certain value, usually approx. 1 4. Draw resonance also plays a role in paper coating and in the manufacture of fibres and monofilaments. It is therefore remarkable that so far this technically important subject has only received scant attention in scientific and technical literature. 

Polymers with a low molecular weight and a high melt temperature, such as the polyamides and polyesters show much less tendency towards draw resonance than polyethylene or polypropylene. 

It is also evident that draw resonance hardly occurs if at all during chill-roll quenching or film blowing. Chill-roll quenching prevents spontaneous lengthening of the melt mechanically, while film blowing results in a two-directional orientation, which prevents a preferential lengthening in one direction. 

On the other hand, Bergonzoni and Dicresce also observed that draw resonance can occur in polystyrene extrudates. If this observation is confirmed, it would be in disagreement with this explanation. Hence, the agreement of Schuur s explanation with a number of experimental facts does not prove that this hypothesis is complete or even correct, and 

---