Web technology and Six Sigma
How do you best utilise Six Sigma to reduce web defects? It all hinges on applying discrete measures to a continuous product, explains Mark Miller of Coating Tech Services
Regardless of the coated product being manufactured Six Sigma can be applied to web coated products following a road map: Define, Measure, Analyse, Improve and Control the problem.
Derived from the standard deviation (a statistical analysis parameter) it is defined as defect levels below 3.4 defects per million opportunities. It is important that the defect and the opportunity are correctly defined or the measurement is useless.
Six Sigma holds to continuous efforts to achieve stable results that are of vital importance to business success; manufacturing processes have characteristics that can be affected and achieving sustained quality improvement requires commitment from the entire organisation.
Defining your goal
It is important that a goal is defined and ‘entitlement’ – a stretch goal defined by the subject matter experts – understood.
This goal should have a continuous process focus but a discrete product plan. The measurement needs to focus on improving the performance of the product within the confines of the process.
It is important that the human side of the operation is understood. Emphasis on support from management is just as vital as the operator understanding why the project is being worked on. Team buy-in is critical. Never assume the project will work because people ‘do what they are told.’
If we take an existing product and process, individuals who have been involved with this system can be tapped to identify where the weaknesses and strengths are as applied to the product. For a new product on an existing line the borders of the process and materials need to be the constraints for the product. Ultimately, for a new product with an unidentified process the world is wide open. The process map needs to be drawn up regardless of the product constraints.
This takes into account human interaction with equipment, raw material variations and measurement system limitations.
Measuring the process
The key thing to consider in measuring a web process is the final product use. The difficulty in measuring a product that is in roll form is that it cannot be ‘fully’ measured. In theory, each discrete widget can be measured for quality if the test is not destructive. Not so with a continuous web.
Improvements can be made at the main variable locations to reduce defects and improve quality. One simple example is to have idler rolls of different diameters grouped by sections throughout the process. Therefore, if you see a repeating defect you can isolate it to the section of the correlating diameter. The unit operations include:
Fluid and fluid delivery: equipment variables to be monitored include mixers, pumps, flow meters and hoses. The material characteristics to be measured include viscosity. The issues (typically downweb in form) include contamination, air entrapment and temperature control.
Substrate and substrate delivery: equipment variables to be monitored include web cleaners, steering, tension, static and temperature control. Material characteristics to be measured include surface energy and substrate manufacturing data. The issues (typically downweb in form) include wrinkles and debris. Communication with the upstream provider of substrate is critical in understanding what is available to work with.
Coating head: specifically in a slot die the equipment variables to be monitored include attack angle (positioning), temperature control, roll total indicated runout (TIR) and chatter. The material characteristics at this point are a function of the roll throughput yield (failures combined up to this point in the process) of the interfaces.
The issues (typically crossweb in form) include air entrainment and edge effects. In the coating head interfaces are the name of the game. Fluid distribution within the coating station can be simulated prior to manufacturing or operation.
In a production setting the improvement of edge effects needs to be balanced against the overall cost of waste.
Curing: the equipment variables to be monitored include energy into the web, drying and stability. The issues include temperature control, rate of cure and effect on substrate. There are various cure methods and each has different effects. A preventive maintenance schedule helps maintain stability of an operation.
Measurement systems: it is critical to measure the process at the intersection of the unit operations so we have to monitor the incoming substrate, the fluid, coating head and outgoing web while ensuring the measurements capture crossweb, downweb and machine chatter defects. Continuous changes and defects occur over time so steady state operation is important to accurate measurement and reaction.
Analysis of variables
The analysis of the variables (causes) and the effect on the defects is the next step. Once the subject matter experts have identified what to measure to improve product performance it is again time to get management on board. A key issue is the amount of time it will take to accurately measure the process.
For example, a consumable backing roll with a limited lifespan was made to order. Although this Six Sigma project took months to analyse, the root cause was identified and the problem solved.
We need to narrow the scope so the team can identify the unit operation to work on. The team should be focusing on having ‘no wasted effort.’
Experiment for improvement
Two types of experiments are important: designed (DOE) and keeping a process journal. The DOE allows focus on the identified variables. In a 2(5-1) factorial design all five variables can be studied with minimal work. Process journals allow previously unidentified variables to uncover new root causes. A critical idea is roll throughput yield (RTY) defined as the combined variability of the process.
I worked with a coater who had a product within spec for the discrete sample but in roll form had a hard band that made it unacceptable. This caliper variation was not evident until the online gauge and process journal pointed the engineer in the right direction.
Control measures
The goal has been met for product improvement and the changes need to be implemented. Make sure the changes are understood and upheld by having a brief description of the ‘why’ behind these new control measures.
This stage usually means a redefinition of the product and process; ultimately‘control’ means a new beginning.
When utilising Six Sigma for process troubleshooting, approach web coating operations with Define (developing a measureable goal and establishing ‘buy in’); Measure (verify measurement capability and narrow the unit ops); Analyse (cause and effect for the root cause); Improve (experimentation and the process journal); and Control (entitlement and next steps).
Remember to utilise the Six Sigma methodology and identify the important differences when you are considering webs over widgets.
Coating Tech Service
T: +1 (612) 605-6019