David Roisum’s blog posts are always worth reading. Here’s his latest:
The most useless general winding question you can pose is what is the best ‘curve’ for a winder. By this I presume someone is struggling with winder troubles and has the Polyannish thought that some magic (tension) curve will make life better (sometimes it might). They may even be more specific as to ask for a taper, or whether or not a constant tension, constant torque or linear taper is best. So, here is the skinny.
First, which defect are you trying to reduce? Note that most winding defects are not even sensitive to any of the winder TNT knobs because they are machine design or maintenance or operational defects. Second, presuming that we do have a tightness related defect, it is unreasonable to expect that one curve would work well for two or more different defects. Third, it is unreasonable to expect that one curve would work well for two or more different materials.
So, here is the short of what might take two days of explanation, such as my video-on-demand course in winding. You must first be absolutely clear which defect and subspecies you are trying to fix. Next, you must know whether that subspecies is a tight, loose, roll-structure or other class of defect. Then, you move ALL of your knobs (not just tension or taper) in the appropriate direction as far as possible. You must also consider whether the winding machine is making this worse than necessary (such as not being in good mechanical condition or not calibrated so that settings can drift). You must also consider how much web quality (i.e., profile variation cross the width) makes this worse. ALL tightness related defects (tight, loose and roll structure) are sensitive to machine and web quality.