User's Manual for the Combined Stresses for Webs

The Safety Factor

CSW calculates a safety factor for possible tensile damage to your web due to a combination of several insults as described below. Web damage can mean more waste, delay and customer complaints. You can use this program to set tensions, to assess risk or to allocate your error budget into different categories. For example, if roller misalignment is high you might accept higher rates of damage, reduce tensions or reduce drive control variability or some combination. To be safe, the overall Safety Factor should be somewhere between 2 and 4 depending on the application and your tolerance for risk.

So, what if your Safety Factor is too high?

In that case you might have room to increase tension if there were some benefit. However, in most cases your SF may end up less than 2 or even less than 1. So, what does that mean? Well, almost certainly your product is yielding as it goes over rollers. In brittle paper or other materials that would be web break territory. If your product were ductile, such as creped tissue, you may be pulling out a little bit of the stretch, especially in the tight lanes (not baggy portions). On metals you might get strain hardening due to reversed bending. If your product were thick, such as liner or food board, you would end up with curl (inside the wound roll) and working of the outer layers that may crack coatings In any case, you can play 'what if' scenarios with the inputs to see what might be manipulated to bring the SF into a safer place and protect your web product from the risk of tensile damage. It is far easier to try a closer roller alignment or find a less baggy material on the computer than it would be to screen these ideas in the plant.

The inputs

Web tension that is normally 1/10-1/4 of web strength. You can get tension from calibrated load cells, calibrated dancers or other means. Tension is lineal tension in units of N/m or lb/in.

Roller misalignment causes tension to be higher on one side. You can estimate maximum misalignment by initial machine quality specifications provided by the builders or spelled out by contract, maintenance reports or by simply putting a PI (flat tape) around a few roller pairs.

Web bagginess causes tension to be higher in the tight lanes. This cross web multiplier might range from 1.1 (a good but not perfect web) to perhaps 2 (nearly unrunnable by any reasonable standard). Electrical Factor TFDE The electrical drive might not hold tension well as read by a responsive load cell. A good drive might have a 10% variation in tension during a demanding speed change and thus the temporal tension multiplier in this case would be 1.1. A very poor drive might allow tension to double in which case the multiplier would be 2.0. Mechanical Factor TFDM Roller bearing drag (at all times) and roller inertia (during speed changes) will cause the tension to be different in one part of a drive zone than another. This down web tension profile multiplier could be as little as 1.1 for a good design to 2.0 for a challenging tissue or other similar application with light weight webs going over several idler rollers.

Thick websThe last two factors are usually only a concern for thick webs where the top of the web might see more tension than a bottom, especially when going over a roller. These factors are the radius of rollers and the radius of curl.

The Strength of the web

Web strength can be obtained by any tensile tester or from your web supplier as the case may be. Modulus might be reported by the tensile tester or by your web supplier or might be found on the internet. Clicking the ? mark will give you moduli for a few common webs. Beware that variations of strength and modulus can be expected for a nominal grade of material and, as always, be super careful to use correct units as inputs. If you don’t know how to get the correct units, talk to your plant engineer who can do conversions for you.

More details are found in Combined Stresses for Webs Going Over Rollers, 2015, by David Roisum. Email me at DrRoisum@aol.com for a copy of this paper if you wish.