Addition of CVD and CVL to the TEAM Prediction Formulae
The TEAM-2 formula for the prediction of Hauteur included variables for Fibre Diameter (D), Staple Length (L), Staple Strength (S), Vegetable Matter (V), and Adjusted Mid Breaks (M*). Following the analysis of the TEAM-3 database it was found that, with the exception of Vegetable Matter, these variables were still important in the prediction of Hauteur. One of the early Hauteur prediction equations that was evaluated by the TEAM-3 Committee was as follows:
Hauteur = 0.58L + 0.45S + 1.00D – 0.17M* + 0.03V
This equation had a Standard Error of 2.6mm and a R2 of 82%. The only coefficient that was not statistically significant was Vegetable Matter.
Following this initial analysis, the Committee commenced evaluating further equations where variables such as Coefficient of Variation of Diameter (CVD), Coefficient of Variation of Length (CVL), and Mean Fibre Curvature (MFC) were added to the above model in different combinations.
Following this evaluation it was shown that the addition of CVD and CVL to the prediction equation resulted in a small improvement to the model with the Standard Error dropping to 2.5mm and the R2 increasing slightly to 83%. The coefficients for CVD and CVL were both statistically significant indicating an influence on processing performance.
The equation including these two variables was as follows:
Hauteur = 0.41L + 0.33S + 1.45D – 0.18M* + 0.08V – 0.63CVD – 0.32CVL
It is worth noting that when CVD is added to the model together with the TEAM-3 variables that the coefficients for both SS and CVD are significant. It has been previously suggested that CVD could replace SS as a predictor of processing performance. This analysis suggests that SS is still important for the processing prediction of consignments and cannot be replaced by CVD alone.
The addition of CVL to the TEAM-2 Hauteur formula was considered by the TEAM-2 committee in 1988. Early analysis of the database indicated that CVL did have a small influence on processing performance. However, analysis of the total database at the conclusion of the TEAM-2 trial did not confirm the significance of CVL in any prediction formulae. It was suggested at the time that CVL may be significant for a mill specific formula. As such, it is not surprising that CVL could be considered as an addition to the processing prediction formulae following the TEAM-3 analysis.
Following a further adjustment for Mid Breaks and Vegetable Matter, the TEAM-3 Committee published the TEAM-3 Hauteur formula as follows:
Hauteur = 0.43L + 0.35S + 1.38D – 0.15M – 0.45V – 0.59CVD – 0.32CVL + 21.8