Taguchi experimentation is often used as an aid in calculating the effects of changing processing variables on manufacturing production processes. This method creates an orthogonal array where the effect of each variable can be measured in combination with other variables. This avoids having to carry out costly experiments identifying the effect of changing one processing individually.
The method involves refined changes of the inputs to a process or an activity in order to observe the corresponding outputs. Taguchi stresses simplicity in its approach and avoids statistical complexity in the analysis of its results.
The Taguchi approach allowed British Steel to focus on the adhesive bonding process used in the manufacture of access flooring panels. Floor panels (ca. 1 metre x 1 metre sq.) for mezzanine floors etc. are made by bonding coated steels to fibreboard using single pack polyurethane adhesives. There is a move towards solvent-free adhesives due to regulatory requirements. Due to the large surface area for adhesive bonding, either adhesive or cohesive failure of the floor panel occurred in different areas of the 'sandwich'.
Taguchi also allowed the examination of production variables on access flooring manufacture. The design of experiments treatment of the process indicated the linked dependency of certain parameters on each other e.g. adhesive and cure temperature. The consolidation pressure being the most critical for the process
The analysis also allowed a reduction in the experimentation from the full matrix of tests which would require 512 experiments to 16 experiments describing the process, a considerable cost saving in time and materials.
The objective was to decrease the incidence of 'delamination' and optimise process conditions. The orthogonal array measures the effect of each variable in combination with other variables. Properly designed experiments assessing each variable at typically two levels, enable the array to separate the effect of each variable within a matrix of manageable size. Nine process variables were identified and the associated test levels determined are given below.
|
Experiment |
Factor |
Level 1 |
Level 2 |
|
A |
Adhesive type |
Standard |
Solvent free |
|
Steel coating |
Zintec |
Hot dip Galv. | |
|
C |
Fibreboard condition |
normal ambient |
added moisture |
|
D |
Adhesive application |
Coated steel |
Fibreboard |
|
E |
Amount of adhesive |
High |
Low |
|
F |
Open Time |
Long |
Short |
|
G |
Joining Pressure |
High |
Low |
|
H |
Cure Temperature |
10oC |
25oC |
|
I |
Cure Time |
12 hours |
48 hours |
Coated steel
Fibreboard
Single pack polyurethane adhesives
Standard or solvent free
Zintec or Hot dip Galvanised coating of steel
Normal ambient or added moisture for fibreboard
Single pack polyurethane adhesive applied to fibreboard or alternatively to coated steel surface. Two substrates placed in contact and defined joining pressure applied.
Heat cure. Cure temperature and time varied in study.
Heat cure 10 to 25ºC
Cure time 12 to 48h
Heat cure
The bonding was carried out at the factory location.
Results have been analysed on the basis of chipboard tear as the failure criteria, the higher the level of board tear the greater the quality of the bonded structure.
The optimum process conditions from the analysis is given below (unranked)
|
Factor |
Setting |
|
Adhesive Type |
Standard |
|
Steel coating/Fibreboard condition |
Hot dip Galv./Normal |
|
Adhesive application/Amount |
Steel/High |
|
Open Time |
Short |
|
Joining Pressure |
High |
|
Cure Temperature/Cure time |
25oC/48 hours |
The confirmation trial indicated that the most significant factor by far was the pressure applied to consolidate the bond. The standard, solvent-based adhesive also gave slightly better results. Due to the poor resolution of results it was not possible to identify major differences between adhesives.
Further tests would aid the resolution of this process.
DTI MTS programme adhesives – Dissemination Guides ADH5CS3.DOC
NPL/ British Steel / ESR Technology Limited
Analysis was conducted by British Steel
Photo courtesy Steelcase
