502
GRAY-STUART ET AL .
TABLE 3
Box lifetime results for each trial in this study
Trial
Preparation Median life time (days)
Mean life time (days)
CV (%)
Number of boxes Number of failures Failure rate
A
Control
11.4
11.4
14
14
2
14%
Filled
12.7
12.3
26
16
12
75%
Foil
n/a
n/a
n/a
16
0
0%
B
Control
12.4
12.4
45
6
6
100%
Filled
5.5
5.6
6
8
8
100%
Foil
16.0
15.9
32
6
5
83%
C
Control
7.2
9.6
70
8
4
50%
Filled
3.4
5.4
101
8
7
88%
Foil
11.1
10.4
36
7
5
71%
D Control
n/a
n/a
n/a
8
0
0%
Filled
n/a
n/a
n/a
8
0
0%
Foil
n/a
n/a
n/a
7
0
0%
TABLE 4 30%BCT
Mean secondary creep rate for all trials with boxes at
rate of the filled boxes is significantly different from the control and foil boxes, while there is no difference between the control and foil box preparations. These results show that the internal pressure from the box fill can have a significant negative impact on box performance and creep rate under certain conditions. At constant 70% RH, the box fill had less of an influence on creep rate, there was a significant dif- ference between the filled and control boxes but not the filled and foil boxes. The reason why the filled boxes performed worse at higher humidities could be due to the mechanical properties of paper decreasing with increasing moisture content. A likely mechanism is that at higher moisture contents the stiffness of the panels decrease to a point where the internal pressure from the box fill can increase the rate of panel bulging relative to the control and in turn increase the creep rate. These results give merit to having different box prepa- rations in creep tests as box performance is significantly affected when presented in a way that is more representative of supply chain conditions. There is no significant difference in the creep rates of the foil and control boxes within the same trial. In constant RH conditions, the foil does not prevent the box panels from reaching the same moisture content as the uncoated panels so the creep rate will be similar. How- ever, under cycling conditions, the foil has a significant effect on the moisture content of those panels as the 12 h cycle time is not long enough for them to reach equilibrium. While the difference in creep rate is not statistically significant, Figure 5 shows that the maximum creep rate observed for the foil boxes is lower than the median for the control. Extrapolation box performance from creep tests on nor- mal boxes could result in creep rate and box lifetime being under- predicted as box performance is influenced by uneven distribution of moisture content. When boxes are palletised, boxes on the side of the pallet have only one panel exposed to the ambient conditions and boxes within the pallet can have all surfaces in contact with other boxes and no ambient exposure. The findings here demonstrate that the creep rate and box life- time are highly dependent on how the box is prepared. Notably the boxes with product performed considerably worse, having a shorter
Trial
Condition Creep rate per day CV (%)
Log creep rate
2.82E 04 a 8.90E 04 b 1.31E 04 a 8.12E 04 a 1.57E 03 b 5.23E 04 a 9.36E 04 a 2.88E 03 b 7.05E 04 a 1.78E 05 b 2.92E 05 a 2.28E 05 a,b
A Control
74%
8.17 7.02 8.94 7.12 6.45 7.56 6.97 5.85 7.26
36%
Filled
24%
Foil
B Control
24%
Filled
20%
52%
Foil
88%
C Control
Filled
56%
61%
Foil
D Control
25%
10.9 10.4
Filled
32%
Foil 10.7 Note : For each trial, the secondary creep rate for each box preparation was compared with a post hoc Tukey HSD test. Means that do not share a letter are significantly different. 22%
3.4
Secondary creep rate
|
The mean creep rate for each trial is shown in Table 4, and the creep rate data are presented in a box and whisker plot in Figure 5. A post hoc Tukey HSD test was performed to determine if there were signifi- cant differences in the mean creep rate of the box preparations within each trial and to compare the effect the RH conditions had on each box preparation (Table 5). The coefficient of variation (CV) for creep rate is higher than found in most previous studies 3,26,32 ; however, it is worth noting that CV was positively correlated with %BCT in the study by Hussain et al. 3 and the load used in this study is at the upper end of this range. The control in trials A and C have a notably higher CV of 74% and 88% and the failure rate of boxes in these tests was 14 and 50%, respectively. This tends to skew the variation in creep rate as the boxes which fail, especially if they fail early on, usually have a considerably higher creep rate. For trials A, B, and C, the creep
Made with FlippingBook flipbook maker