168
To determine if hand held near infrared spectroscopy can be used to measure corn particle size, corn particle distribution and corn moisture
Tuesday, March 15, 2016: 2:00 PM
316-317 (Community Choice Credit Union Convention Center)
Alejandro Criado
,
Aunir, Towcester, United Kingdom
Chris Piotrowski
,
Aunir, Towcester, United Kingdom
P. Wilcock
,
AB Vista Feed Ingredients, Marlborough, United Kingdom
C. L. Bradley
,
AB Vista Feed Ingredients, Marlborough, United Kingdom
Ben Haberl
,
Iowa Select Farms, Iowa Falls, IA
Benjamin C Smith
,
Iowa Select Farms, Iowa Falls, IA
Abstract Text: In swine production the effect of corn particle size on performance in pigs fed pelleted diets has shown that in general the finer the grind size the better the feed efficiency. Currently, to determine particle size, the most widely used methods are the onsite 3 pan sieve analysis because of its simplicity, cheapness and ease of interpretation or the 13 pan sieve analyses which is often tested off site at a laboratory. Near-infrared (
NIR) spectroscopy is an analytical technique used for grain quality assessments due to its versatility and speed. Although the ability of NIR to estimate particle size is well documented, obtaining the full particle size distribution profile has not been studied in depth. The objective of this trial was to determine if a hand held near-Infra-red (HHNIR) spectroscopy could be used to measure corn average particle size, moisture and particle distribution. In order to develop this calibration, 95 corn samples resulting in 141 scans across two instruments (from roller and hammer mills) were collected, split and either tested using the corresponding particle size references by a 13 stage sieve analysis derived from the US standard Sieve Series (
SV) using a sieve agent or scanned in triplicate on a hand held NIR (950-1650 nm) spectrometer (
NIRS). For average particle size and moisture there was a strong correlation between SV and NIRS (r
2 = 0.79; RPD = 2.2 and r
2 = 0.91; RPD = 3.3 respectively). The final prediction model on the NIRS shows the accumulation of sample which will remain in each sieve and the calibration comparison compared to the SV methodology (Table 1) with all sieves showing a strong correlation and RPD. Overall, the results confirm that a HHNIR can be used on site to instantly determine corn particle size, corn moisture and corn particle distribution.
Table 1.Calibrations for individual sieve pans.
Sieve Pan
|
Mean
|
SD
|
RSQ
|
SECV
|
RPD
|
d6
|
0.13
|
0.07
|
0.67
|
0.07
|
1.7
|
d8
|
0.89
|
0.60
|
0.80
|
0.39
|
2.2
|
d12
|
3.99
|
2.71
|
0.80
|
1.69
|
2.2
|
d16
|
18.43
|
11.66
|
0.62
|
8.24
|
1.6
|
d20
|
40.06
|
13.72
|
0.80
|
8.57
|
2.2
|
d30
|
57.46
|
9.96
|
0.81
|
6.15
|
2.3
|
d40
|
68.28
|
6.94
|
0.81
|
4.59
|
2.3
|
d50
|
74.97
|
5.30
|
0.80
|
3.29
|
2.3
|
d70
|
78.98
|
4.65
|
0.80
|
2.82
|
2.2
|
d100
|
82.92
|
3.80
|
0.80
|
2.41
|
2.3
|
d140
|
86.12
|
3.18
|
0.82
|
1.95
|
2.3
|
d200
|
88.18
|
2.67
|
0.82
|
1.69
|
2.3
|
d270
|
89.54
|
2.43
|
0.83
|
1.45
|
2.4
|
Keywords: Particle Size, Corn, Near Infrared Spectroscopy