Investigating the Physical Properties of Corn Varieties

Shamma Tasneem Chowdhury; Clairmont L. Clementson; Emmanuel Baidhe

doi:10.13031/ja.15769

Article Request Page ASABE Journal Article
Investigating the Physical Properties of Corn Varieties
Shamma Tasneem Chowdhury¹, Clairmont L. Clementson^1,*, Emmanuel Baidhe¹
Published in Journal of the ASABE 67(3): 631-639 (doi: 10.13031/ja.15769). Copyright 2024 American Society of Agricultural and Biological Engineers.
¹    Agricultural & Biosystems Engineering, North Dakota State University, Fargo, North Dakota, USA.
^*    Correspondence: clairmont.clementson@ndsu.edu
The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License https://creative commons.org/licenses/by-nc-nd/4.0/
Submitted for review on 3 August 2023 as manuscript number PRS 15769; approved for publication as a Research Article by Associate Editor Dr. Kingsly Ambrose and Community Editor Dr. Sudhagar Mani of the Processing Systems Community of ASABE on 20 February 2024. The authors declare no conflict of interest
Citation: Chowdhury, S. T., Clementson, C. L., & Baidhe, E. (2024). Investigating the physical properties of corn varieties. J. ASABE, 67(3), 631-639. https://doi.org/10.13031/ja.15769
Highlights
The effect of moisture on the physical properties of corn varied significantly with variety.
There is no comprehensive moisture model for the prediction of physical properties of corn.
The models derived from experimental data and models posited in the literature provide varying degrees of predictive accuracy.
Abstract. Several corn varieties have been developed to match short maturity seasons and increase yield for specific end use, such as ethanol production and animal feed. However, it is unclear how varietal differences influence the physical properties. Similarly, there is no certainty that these varieties adhere to the moisture engineering properties relationships recorded in the literature. This research investigated the physical properties of 10 corn hybrids at three moisture content levels (13.5%, 15.5%, and 17% w.b.). The influence of moisture content on the different geometric (equivalent diameter, geometric mean diameter, sphericity, aspect ratio, surface area, projected area, flatness ratio) and gravimetric (bulk density, true density, porosity) properties were assessed and compared with the published literature. Further, this study assessed the potential of regression-based moisture engineering properties relationship models for the prediction and description of the physical properties of corn. Predictive and descriptive comparisons indicated that, in some instances, reference models provided better predictions of the geometric and gravimetric properties of a single variety compared to pooled samples. The study models provided better predictions for the pooled samples. However, the study results confirm no one-size-fits-all moisture physical property model exists. Using existing moisture content based regression models for description and predictive purposes should be done with specific variety references.
Keywords.Corn hybrids, Moisture based physical property models, Prediction of physical properties.
Corn (Zea mays L.) is a highly significant crop globally, serving as a vital resource for animal feed, human consumption, and various industrial applications (Revilla et al., 2022). Its fascinating history originated over 6,000 years ago as a wild grass called teosinte in Mexico (Erenstein et al., 2022). Over time, corn cultivation has spread, emerging as an essential crop in production, consumption, and exportation in the United States (U.S.). Presently, close to 40% (6.2 billion bushels) of the corn harvest in the U.S. is processed to produce ethanol (O’Malley and Searle, 2021). Ethanol production is presently the greatest corn industrial application in the U.S., creating a huge demand for the crop. This demand is anticipated to increase as part of the U.S. gasoline ethanol blending strategy under the broad Renewable Fuel Standard program to reduce the effect of conventional gasoline use on the environment (Newes et al., 2022).
Corn also has essential value for feed and food. To quell discontent in the food versus energy debate, corn production must increase, utilizing similar or less land. To meet this steadily growing demand, corn has evolved through genetic engineering, giving rise to hybrid varieties that are high yielding, drought tolerant, disease-resistant, and adaptable enough to mature within the 120-day frost-free season, among other traits. The widespread acceptance of genetically modified grain types was identified as one of the significant changes in the American maize farming industry that contributed to the current high corn output (Hunt et al., 2020). Developing hybrid varieties involves modifying the grain deoxyribonucleic acid (DNA), potentially affecting various grain properties, especially physical properties. Despite the documented increase in crop yields, there needs to be more published information regarding how genetic modification has influenced different grain properties, particularly in the U.S., where almost 90% of corn is genetically engineered (Osborne et al., 2016). A primary concern is whether the currently genetically engineered corn varieties align with the properties described in existing standards and peer reviewed literature.
The development of corn processing systems relies heavily on various engineering properties (Kruszelnicka, 2021; Tarighi et al., 2011). However, conducting individual analyses through experiments can be relatively time-consuming, leading researchers to base their work on available literature. For example, some studies have reported linear and polynomial relationships between corn grain properties and moisture content (MC), enabling the simulation of parameters without further experimentation (Sangamithra et al., 2016; Tarighi et al., 2011). Nevertheless, it remains unclear whether these relationships apply to all the corn varieties in the U.S. Furthermore, whether these relationships adequately account for varietal differences is unknown. Additionally, limited studies have been conducted to evaluate the conformity of physical properties with the information available in the literature. There is a lack of comprehensive understanding regarding how genetic modification affects different grain properties in genetically engineered corn varieties in the U.S. Addressing these questions through further research would provide valuable insights for corn processing systems and contribute to the advancement of genetically engineered crops.
This research aimed to investigate the varietal differences in physical properties of 10 corn hybrids at three different moisture content levels (13.5%, 15.5%, and 17% w.b.). The influence of moisture content on the physical properties was assessed and compared with published literature. The properties considered in this study are grouped into two categories: geometric (equivalent diameter, geometric mean diameter, sphericity, aspect ratio, surface area, projected area, flatness ratio) and gravimetric (bulk density, true density, porosity) properties. The findings of this study will further provide accurate data for simulations and model derivation that give rise to precision agriculture and artificial intelligence tools and solutions. Also, accurate models will account for the variation of corn shape irregularities and inherent properties.
Materials and Method
Sample Collection
Ten (10) varieties of corn hybrids were collected from the North Dakota State University (NDSU) Experiment Station in Casselton, North Dakota, United States (46º52'41.9" N and 97º14'56.9" W). The ten varieties, namely 2B862, LC464 21, 2B851, DS 3203AM, DS 2919AM, 2B863, LC403 22, 9212 10, D31VC23, and 2392VT2PRIB, are coded as V1, V2, V3, V4, V5, V6, V7, V8, V9, and V10, respectively, throughout the document. Each variety was kept separately in a 5-gallon plastic bucket with tight covers. They were stored in a refrigerator at 2°C–5°C until the experiments were carried out.
Moisture Content Conditioning
In triplicate, the initial moisture content of the samples was determined using a digital moisture meter (Dickey-John Grain Moisture Tester, Model GAC2100) and found to be about 15.7% (w.b.). The digital moisture meter was calibrated with corn data from many varieties and locations and used throughout the experiment for consistency. The samples were then conditioned to the moisture content of about 13.5%, 15.5%, and 17% specified for this study. This involved placing the samples on trays and air drying until the MC reached 13.5% (w.b.) and 15.5% (w.b.). For 17% MC, rewetting of each sample was done by using equation 1 to calculate the amount of water to be added. The specified amount of water was added to the corn samples separately in individually sealed Ziplock bags. The Ziplock bags were thoroughly hand mixed for 3 minutes. The Ziplock bags were then sealed, placed in buckets, and kept in a cold room at about 4°C for 10 days to allow an even distribution of moisture within the kernels (Adeyanju et al., 2022; Coskun et al., 2006; Darfour et al., 2022; Kruszelnicka et al., 2022). Before conducting experiments, Ziplock bags were taken from the cold room and allowed to equilibrate to room temperature. Then samples were randomly obtained for each test.
        (1)
where
Q = amount of water added (g)
W_a = sample initial mass (g)
M_i = initial moisture content of the sample (%, d.b.)
M_b = desired moisture content of the sample (% d.b.).
Determination of Geometric Properties of Corn
Fifty corn kernels were randomly selected from each sample for these tests. The corn kernel dimensions of length (L), width (W), and thickness (T) of each corn kernel were determined using a digital caliper (accuracy of 0.01 mm). The equivalent diameter (D_e) in mm, geometric mean diameter (Dg) in mm, sphericity (f)_, aspect ratio (AR), flatness ratio (FR), surface area (SA) in mm², and projected area (A_p) in mm² were determined using equations 2–8, respectively (Adeyanju et al., 2022; Atere et al., 2016; Brar et al., 2017; Coskun et al., 2006; Darfour et al., 2022; Kruszelnicka et al., 2022; Mohsenin, 1971; Ozturk et al., 2009):
        (2)
        (3)
        (4)
        (5)
        (6)
        (7)
        (8)
Determination of Gravimetric Properties on Corn
Bulk Density, True Density, and Porosity
The bulk density of the corn was determined using the test weight per bushel apparatus (Seedburo Equipment Company, Chicago, USA) with a standard 1 pint test weight kettle (568.26 mm³) with the procedure described by Boac et al. (2023). All weights were taken using a digital weighing scale with an accuracy of 0.01 g. The bulk density was determined using equation 9. Three replicates were tested for each of the varieties and moisture content combinations.
        (9)
where
?_b = bulk density (kg/m³)
M₀ = mass of the empty test weight kettle (kg)
M₁ = mass of the test weight kettle filled with the corn kernels (kg)
V = volume of the test weight kettle (m³).
The true density is the solid density of a corn kernel without the void spaces inside the kernel mass (Chakraverty and Singh, 2014). The true density in kg/m³ was determined by the toluene displacement method with a known mass of kernel in three replicates (Darfour et al., 2022; Ropelewska, 2018). Toluene was used instead of water because it is not readily absorbed by grains and has a low viscosity. 150 ml of toluene was added to the graduated cylinder. A mass of about 80g of corn sample was added to the toluene. The volume of toluene displaced was read from the graduated cylinder. The true density was determined using equation 10 (Darfour et al., 2022).
        (10)
where
?_t = true density (kg/m³)
M = mass of the corn sample
V_d = displaced volume of toluene.
The porosity (%) was calculated based on the values of bulk and true density as given in equation 11 (Atere et al., 2016; Ozturk et al., 2009; Ropelewska, 2018; Singh and Goswami, 1996).
        (11)
where
e = porosity (%)
?_t = true density (kg/m³)
?_b = bulk density (kg/m³).
Statistical and Relational Analysis
Statistical Analysis
The collected data from the experiment was tabulated using MS Excel 2019. Analysis of Variance (ANOVA) at the 5% level of significance was used to perform the statistical analysis. This analysis was conducted with the aid of Minitab Statistical Software Version 21.2. Post hoc analysis was performed using Tukey’s HSD for all parameters. The Tukey HSD test helped to determine which treatments or groups are significantly different from each other.
Simple Regression Analysis and Prediction of Engineering Properties Using Study and Reference Models
A pooled sample (collation of all related physical properties for the specified moisture content) was used as opposed to the variety based analysis that has previously been used by several researchers. This mimicked a corn grain bulk comprising several varieties. The experimental data from the pooled sample was fitted to three (3) regression models (table 1) using nonlinear robust regression with least absolute residuals (LAR) using the cftool in MATLAB 2022b. Robust regression was adopted as it overcomes the effect of outliers and influential observations in the data while minimizing their impact on the regression coefficients (Khan et al., 2021). Curve fitting was adopted for this study as it is the most reported method for developing moisture content based engineering property relationships in the literature. These models were selected based on their previous successful application in describing the relationship between moisture content and physical properties in various grains and seeds. Notably, these models have been used in studies related to flaxseeds (Singh, 2016), corn (Chandio et al., 2021; Coskun et al., 2006; Karababa and Coskuner, 2007; Sangamithra et al., 2016; Sobukola et al., 2013), velvet beans (Adeyanju et al., 2022), pinto beans (Santos et al., 2016), and sesame seeds (Eze et al., 2022; Tunde Akintunde and Akintunde, 2007). The coefficient of determination (R²) was adopted as a measure for goodness of fit, with higher R² values indicating a higher quality of model fitness to the experimental data. The model that provided the best fit was used as the representative study model for the specified property. As used in this study, study models refer to the moisture content based models derived using the pooled samples of experimental data.
Table 1. Models used in the study.^[a]
Model Model Expression
Linear model Y = aMC + b
Quadratic model Y = aMC² + bMC + c
Power model Y = aMC^b + c
^[a]    Y = Engineering property; MC = moisture content (%); and a, b, c = model constants.
Reference models describing the relationship between geometric and gravimetric properties with MC were further applied to this study for comparison with the study models obtained from the pooled experimental data. In this study, reference models refer to the moisture content based models present in the literature that were evaluated in the present study. The reported coefficient of determination (R²) value was the standard criterion for the choice of the reference models, as only those with an R² > 0.9 were adopted in this study.
To establish the potential predictive application, both study and referenced models were used to predict the different physical properties for one random variety (variety V4) and the pooled sample. The percent standard error (SE%) values, as calculated using equation 12 (Phinney et al., 2017), were used to assess the predicting performance of the models. The SE% provides an explanation of how the experimental data differ from the predicted values (Phinney et al., 2017). A lower SE% indicates a better predictive performance. Although several study models were developed, only models whose R² > 0.9 were considered for comparative discussion with reference models. This is because higher R²values explain the degree of dependence of the studied engineering properties on MC.
        (12)
Table 2. The effect of variety and moisture content on the geometric dimensions of corn.^[a]
Variety MC L W T D_e D_g f AR FR SA A_p
V1 13.5 12.0^bcd 8.5^abcde 4.8^defg 8.1^bcdef 7.8^bcdef 0.65^defg 0.71^efghij 0.59^ef 193.4^bcdefgh 80.0^bc
15.5 11.1^efgh 8.2^cdefg 4.9^abcdefg 7.8^cdefgh 7.6^defgh 0.69^abcd 0.75^abcdefghi 0.60^cdef 183.0^defghi 72.0^efghi
17 11.7^cde 8.6^abcd 5.0^abcdefg 8.1^bc 7.9^bcd 0.68^abcdefg 0.74^abcdefghi 0.58^cdef 197.8^bcd 78.6^bcde
V2 13.5 10.7^h 7.8^gh 5.4^abc 7.7^gh 7.6^defgh 0.71^ab 0.73^cdefghi 0.69^ab 183.0^defghi 65.9ⁱ
15.5 10.8^h 8.4^abcde 4.9^abcdefg 7.9^cdefgh 7.7^cdefgh 0.71^ab 0.79^abc 0.58^cdef 186.1^bcdefghi 72.1^efghi
17 10.9^gh 8.2^abcd 4.9^bcdefg 7.9^cdefg 7.7^bcdefg 0.71^ab 0.77^ab 0.57^def 187.6^bcdefgh 74.1^cdefg
V3 13.5 10.9^fgh 7.6^h 5.4^ab 7.7^gh 7.7^cdefgh 0.69^ab 0.69^ghij 0.72^a 184.9^cdefghi 65.8ⁱ
15.5 12.3^b 7.9^efgh 5.2^abcde 8.1^bcde 7.9^bcd 0.64^efg 0.65^j 0.65^abcd 198.1^bcd 76.7^bcdefg
17 10.8^h 7.9^fgh 5.5^a 7.8^defgh 7.7^bcdefg 0.72^a 0.73^bcdefghi 0.69^ab 187.3^bcdefgh 66.7^hi
V4 13.5 11.6^cdef 9.0^a 4.9^abcdefg 8.3^ab 8.0^ab 0.69^abc 0.78^abcd 0.58^f 201.7^ab 81.8^b
15.5 12.9^a 8.9^ab 4.9^abcdefg 8.5^a 8.3^a 0.64^fg 0.69^ij 0.56^ef 217.1^a 90.9^a
17 11.2^efgh 8.7^abc 5.1^abcdef 8.1^bcd 7.9^bcde 0.71^ab 0.78^abcd 0.59^cdef 197.0^bcdef 77.3^bcdef
V5 13.5 10.9^fgh 8.4^bcdef 4.7^efg 7.8^efgh 7.5^fgh 0.69^abcd 0.76^abcde 0.56^ef 179.3^ghi 72.6^defgh
15.5 11.2^efgh 8.6^abcd 4.6f^g 7.9^cdefgh 7.6^efgh 0.68^abcdef 0.77^abcd 0.54^f 181.9^efghi 75.7^bcdefg
17 10.8^h 8.5^abcde 4.7^efg 7.8^fgh 7.5^gh 0.69^abcd 0.78^abcd 0.55^f 177.8^hi 72.1^efghi
V6 13.5 11.2^efgh 8.1^defgh 5.3^abcd 7.9^cdefg 7.8^bcdefg 0.69^ab 0.73^defghi 0.68^abc 190.6^bcdefgh 71.4^fghi
15.5 11.5^cdefg 8.3^cdefg 5.1^abcdefg 8.0^bcdefg 7.8^bcdefg 0.68^abcde 0.73^defghi 0.61^cdef 193.4^bcdefgh 75.1^cdefg
17 11.1^efgh 8.1^defgh 5.2^abcd 7.9^cdefg 7.8^bcdefg 0.70^ab 0.74^abcdefghi 0.65^abc 190.9^bcdefgh 71.2^fghi
V7 13.5 11.3^efgh 7.9^fgh 4.6^g 7.6^h 7.4^h 0.65^cdefg 0.69^fghij 0.58^cdef 171.5ⁱ 70.2^ghi
15.5 12.1^bc 8.3^cdefg 4.6f^g 7.9^bcdefg 7.7^bcdefg 0.63^g 0.69^hij 0.55^f 186.7^bcdefghi 78.9^bcd
17 11.3^efgh 8.5^abcde 4.6f^g 7.8^cdefgh 7.6^fgh 0.67^bcdefg 0.8^abcdefgh 0.54^f 181.4^fghi 75.1^cdefg
V8 13.5 11.6^cde 8.4^bcdef 4.9^bcdefg 8.0^bcdefg 7.8^bcdefg 0.67^bcdefg 0.73^cdefghi 0.58^cdef 191.3^bcdefgh 76.9^bcdef
15.5 11.3^efgh 8.5^abcde 4.9^bcdefg 7.9^cdefg 7.7^bcdefg 0.69^abcd 0.76^abcdef 0.57^cdef 188.5^bcdefgh 74.9^cdefg
17 11.4^defg 8.4^bcde 5.3^abc 8.1^bc 7.9^bc 0.69^ab 0.74^abcdefghi 0.64^abcde 199.3^bc 75.7^bcdefg
V9 13.5 11.3^efgh 8.5^abcd 5.2^abcde 8.1^bcd 7.9^bcde 0.70^ab 0.75^abcdefg 0.61^bcdef 197.7^bcde 75.9^bcdefg
15.5 11.5^defg 8.5^abcde 4.9^abcdefg 8.0^bcdefg 7.8^bcdef 0.68^abcde 0.74^abcdefghi 0.59^cdef 193.9^bcdefg 76.6^bcdefg
17 11.2^efgh 8.6^abcd 5.2^abcde 8.1^bcd 7.9^bcde 0.71^ab 0.77^abcd 0.61^bcdef 197.1^bcdef 75.8^bcdefg
V10 13.5 11.4^defg 8.8^abc 4.8^cdefg 8.1^bcd 7.9^bcdef 0.68^abcd 0.77^abcd 0.56^f 194.7^bcdefg 78.9^bcd
15.5 11.1^efgh 8.5^abcd 4.9^abcdefg 7.9^bcdefg 7.7^bcdefg 0.69^ab 0.77^abcde 0.58^cdef 189.3^bcdefgh 74.4^cdefg
17 10.9^fgh 8.7^abc 5.0^abcdefg 8.0^bcdefg 7.8^bcdefg 0.71^ab 0.79a 0.57^cdef 191.6^bcdefgh 74.9^cdefg
Mean 11.35 8.39 4.99 7.97 7.77 0.69 0.74 0.6 190.48 74.95
SD 0.92 0.80 0.75 0.45 0.46 0.06 0.09 0.12 22.49 9.96
SE 0.02 0.02 0.02 0.01 0.01 0.002 0.002 0.003 0.58 0.26
CV 8.17 9.58 15.10 5.72 5.86 8.65 12.08 19.33 11.81 13.29
P_v <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
P_m <0.0001 0.001 0.007 0.017 0.078^ns <0.0001 <0.0001 0.002 0.072^ns <0.0001
P_vm <0.0001 <0.0001 0.043 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
^[a]    MC = moisture content, L = length (mm), W = width (mm), T = thickness (mm), D_e = equivalent diameter (mm), D_g = geometric mean diameter (mm), f = sphericity, AR = aspect ratio, FR = flatness ratio, SA = surface area (mm²), A_p = projected area (mm²), SD = standard deviation, SE = standard error, CV = coefficient of variation, P_v = p value of variety, P_m = p value of moisture content, and P_vm = p value of interaction. Values are means of 50 replicates of corn seeds. Means with different letters in a column are significantly different at 0.05 level of significance. ns = not significantly different at 0.05 level of significance.
where
SE (%) = percent standard error (%)
SEE = standard error of estimate (unitless)
? = average of the experimental values.
Results
Geometric Properties
Table 2 presents the interaction effect of variety and MC on different geometric properties, where the interaction effect was found to be statistically significant at the 5% level of significance. The kernel length varied between 10.7 mm (variety V2 at 13.5% w.b.) and 12.9 mm (variety V4 at 15.5% w.b.), with the mean length of corn kernels being 11.35 ± 0.92 mm. The length of variety V4 at 15.5% w.b. was significantly higher than the length of all the other treatment combinations (p < 0.05). A significantly higher length was observed at MC of 15.5% w.b. compared to 17% w.b. for varieties V3, V4, and V7, as shown in table 2. The length of some varieties, such as V10, generally decreased with MC, although the variation was not significant (table 2). The kernel width varied between 7.6 mm (variety V3 at 13.5% w.b.) and 9 mm (variety V4 at 13.5% w.b.). The mean kernel width was 8.39 ± 0.8 mm. The kernel thickness varied between 4.6 mm (variety V7 at 13.5%) and 5.5 mm (variety V3 at 17%), with a mean thickness of 4.99 ± 0.75 mm. The kernel thickness increased within the MC for V1 and V10. However, kernel thickness did not change even with the increase in moisture for varieties V5 and V7, as shown in table 2.
The geometric dimensions of the varieties evaluated were in the range of values reported in the literature at lower MC. Kruszelnicka (2021) reported that the length, width, and thickness ranged between 7.24–12.6 mm, 3.4–9.7 mm, and 4–7 mm, respectively, at 12.68% for corn varieties grown in Poland. A study on corn kernels grown in Turkey by Coskun et al. (2006) reported a range of mean kernel length, width, and thickness of 10.56 mm, 7.9 mm, and 3.45 mm, respectively, at 11.54% d.b. (10.4% w.b.). A similar range of values for the geometric dimension were reported for corn varieties grown in India within a similar range of MC of 9%–17% w.b. by Sangamithra et al. (2016).
The equivalent diameter varied between 7.6 mm (variety V7 at 13.5% w.b.) and 8.5 mm (variety V4 at 15.5%). The geometric mean diameter varied from 7.4 mm (variety V7 at 13.5% w.b.) to 8.3 mm (variety V4 at 15.5% w.b.). The mean equivalent and geometric mean diameter for the pooled sample were 7.97 ± 0.45 mm and 7.77 ± 0.46 mm, respectively. A similar range of values for equivalent diameter was reported for corn kernels by Kruszelnicka (2021) at a low MC of 12.68%. Darfour et al. (2022) reported a similar range of geometric mean diameters of 6.65 mm and 7.55 mm at a low MC of 10%–12.55%. Sangamithra et al. (2016) reported a similar range of values for 8.7%–21.7% d.b. MC.
The kernel sphericity varied between 0.63 (variety V7 at 15.5% w.b.) and 0.72 (variety V3 at 17% w.b.). The mean kernel sphericity for the pooled sample was 0.69 ± 0.06. It is observed that sphericity reduces with MC from 13.5% to 15.5% w.b. except for V1, then generally increases at 17%. This is unique, though, as most literature reports a linear increase in sphericity with MC. Coskun et al. (2006) reported sphericity values of between 0.61 and 0.64 for corn at MC between 11% and 16.7% w.b. For the same range of MC of 13%–17%, Karababa & Coskuner (2007) reported sphericity values of 0.66–0.69 for corn in Turkey. A similar range of sphericity values was, however, reported by Darfour et al. (2022) and Kruszelnicka (2021) in the low MC range of 10%–12.7%. The kernel aspect ratio varied between 0.65 (variety V3 at 15.5% w.b.) and 0.79 (variety V10 at 17% w.b.), with a mean of the pooled sample of 0.74 ± 0.09. The study findings follow the range of values obtained by Kruszelnicka (2021) at 12.68% MC. The kernel flatness ratio varied between 0.54 (variety V5 at 15.5% w.b.) and 0.72 (variety V3 at 13.5% w.b.). Mousaviraad and Tekeste (2020) reported aspect and flatness ratio values of 0.6 ± 0.07 and 0.64 ± 0.13, respectively. Sphericity, aspect, and flatness ratios define the rolling ability of the grains. They are critical in the design of conveyance and bulk storage systems for grains.
The surface area of corn kernels varied between 171.5 mm² (variety V7 at 13.5% w.b.) and 217.1 mm² (variety V4 at 15.5% w.b.). The kernel projected area varied between 65.8 mm² (variety V3 at 13.5% w.b.) and 90.9 mm² (variety V4 at 15.5% w.b.). The mean surface and projected area were found to be 190.48 ± 22.49 mm² and 74.95 ± 9.96 mm², respectively. Darfour et al. (2022) reported surface and projected area values of 139–180 mm² and 52–73 mm², respectively, for corn at 10%–12.55% MC. Karababa and Coskuner (2007) reported surface area values of 151.8–182.8 mm² for corn kernels between 13.24% and 17% db. Coskun et al. (2006) reported a projected area of 59.7–75.6 mm² for corn at a low MC of 10.4% w.b.
Table 3. The effect of variety and moisture content on the gravimetric properties of corn.^[a]
Variety MC Bulk Density
(kg/m³) True Density
(kg/m³) Porosity
(%)
V1 13.5 763.3^abc 1224.8^ab 37.67^defgh
15.5 759.1^bcd 1237.4^ab 38.59^bcdefgh
17 726.4^ghi 1212.1^ab 40.07^bcdefg
V2 13.5 749.9^de 1212.1^ab 34.13^bcdefgh
15.5 741.9^ef 1205.5^ab 38.44^bcdefgh
17 715.9^hij 1212.9^ab 40.93^abcde
V3 13.5 755.1^cd 1237.4^ab 38.96^bcdefgh
15.5 765.3^abc 1212.1^ab 36.91^gh
17 715.5^ij 1224.8^ab 41.57^ab
V4 13.5 762.3^abcd 1200.2^ab 38.91^bcdefgh
15.5 769.7^ab 1212.9^ab 36.03^gh
17 705.8^jk 1200.2^ab 41.18^abc
V5 13.5 772.5^a 1237.4^ab 37.56^efgh
15.5 772.5^a 1224.8^ab 36.87^gh
17 729.8^fg 1176.5^ab 37.97^cdefgh
V6 13.5 750.2^de 1224.8^ab 38.73^bcdefgh
15.5 753.8^cde 1212.1^ab 37.81^cdefgh
17 707.5^jk 1200.2^ab 41.03^abcd
V7 13.5 699.8^k 1176.5^b 40.52^abcdef
15.5 715.7^ij 1184.1^b 39.56^bcdefgh
17 679.7^l 1212.1^ab 43.92^a
V8 13.5 753.9^cde 1237.4^ab 39.06^bcdefgh
15.5 770.2^ab 1218.7^ab 36.81^gh
17 716.3^hij 1212.1^ab 40.90^abcde
V9 13.5 772.8^a 12125.5^ab 36.89^gh
15.5 759.4^bcd 1213.9^ab 37.44^fgh
17 735.9^fg 1200.2^b 38.67^bcdefgh
V10 13.5 771.4^ab 1224.8^ab 37.00^gh
15.5 764.5^abc 1266.7^a 39.51^bcdefgh
17 728.9^gh 1176.5^b 38.04^cdefgh
Mean 742.83 1213.9 38.79
SD 26.05 26 2.00
SE 2.75 2.14 0.21
CV 3.51 2.14 5.16
P_v <0.001 0.019 <0.001
P_m <0.001 0.002 <0.001
P_vm <0.001 0.002 <0.001
^[a]    SD = standard deviation, SE = standard error, CV = coefficient of variation, P_v = p value of variety, P_m = p value of moisture content, and P_vm = p value of interaction. Values are mean of three replicates for properties. Means with different letters in a column are significantly different at 0.05 level of significance.
Gravimetric Properties
The interaction effect of variety and MC on all gravimetric properties was statistically significant at the 5% significance level (table 3). The grain bulk density varied between 679.7 kg/m³ (variety V7 at 17%) and 772.8 kg/m³ (variety V9 at 13.5%). The mean bulk density was 742.83 ± 26.05 kg/m³. The true density varied between 1176.5 kg/m³ (variety V7 at 13.5%) and 1266.7 kg/m³ (variety V10 at 15.5%). The mean true density was 1213.9 ± 26.0 kg/m³. Seven of the ten varieties evaluated showed lower true density at 17% MC compared to values at 13.5% MC (table 3). Darfour et al. (2022), however, reported bulk density values of 709.5–746 kg/m³ and a true density of 1250 kg/m³ at a lower MC of 10%–12.55%. The porosity varied between 38.91% (variety V4 at 13.5%) and 43.92% (variety V7 at 17%), with a mean value of 38.79 ± 2.0%. The same range of porosity values was reported by Darfour et al. (2022) for corn kernel at a lower MC of 10%–12.55%. There was a significant difference in porosity when MC increased from 13.5% to 17% w.b. for all varieties examined (table 3).
Prediction and Description of Engineering Properties Based on Their MC Relationship
MC-based engineering properties regression relationships are relevant for the description and prediction of the different engineering properties and the simulation of grain behaviors during processing. The predictive indices of the study and reference models are presented in table 4. The reference model gave a much better prediction for kernel length for both single varieties (< 1.8% error) and pooled samples (< 3.18% error) compared to the study model at 2.8% and 14.8%, respectively. A comparable degree of accuracy was observed for both the study and reference models in assessing the geometric mean diameter of variety V4 and the pooled sample, with marginal differences in SE%. The percent standard error was found to be 0.39% and 0.4% with variety V4 and 1.03% and 1.08% with the pooled sample for the study and reference models, respectively (table 4). The study and reference models predict the width for variety V4 grains with a very large error (> 14%), although it reduces greatly for the pooled samples. To the contrary, the study and reference models gave better predictions for the thickness of variety V4 (< 1.2% error) compared to the pooled sample at about 4.7% error (table 4). The reference models predicted the sphericity of variety V4 grains and pooled sample grains with an error of over 100%, despite the reported very high R² greater than 0.95 (Karababa and Coskuner, 2007; Kazarian and Hall, 1965; Sangamithra et al., 2016). Additionally, the study model for surface area (R² = 0.969) predicts the surface area of variety V4 and pooled sample grains with an error of 53% and 102.9%, respectively (table 4). The reference model better predicted the projected area of V4 than the study model, although the study model gave better results for the pooled sample (table 4).
Table 4. Performance of the study and reference moisture based prediction models for engineering properties of corn.^[a]
Property Study Relationships Between Moisture Content and Physical Properties Prediction Performance
of Models
Study Model
Equation R² Reference
Model Variety V4 Pooled Sample
SE₁
(%) SE₂
(%) SE₁
(%) SE₂
(%)
Geometric
properties L 0.038 ^[b] 2.8 1.8 14.8 3.18
W 0.991 ^[b] 15.4 14.3 3.13 2.96
T 0.996 ^[b] 1.15 1.15 4.62 4.8
D_e 0.003 ^[c] 2.52 2.52 4.86 4.86
D_g 0.969 ^[b] 0.39 0.4 1.03 1.08
f 0.974 ^[d] 0.06 8.1e4 0.21 2.5e5
AR 0.970 ^[e] 0.47 0.36 1.58 1.05
FR 0.975 ^[e] 0.22 0.22 0.89 0.99
SA 0.969 ^[b] 53.5 55.4 102.9 111.2
A_p 0.012 ^[f] 1.4e3 62.4 5.62 90.9
Gravimetric
properties ?_t 0.960 ^[d] 1.84 7.2 5.2 16.1
?_b 0.460 ^[d] 0.59 6.49 5.01 17.2
e 0.339 ^[d] 0.49 2.46 0.92 5.26
^[a]R² = Coefficient of determination for the fitted study model, SE₁= percent standard error for prediction based on study model, SE₂ = percent standard error for prediction based on reference model, MC = moisture content, L = length, W = width, T = thickness, D_e = equivalent diameter, D_g = geometric mean diameter, f = sphericity, AR = aspect ratio, FR = flatness ratio, SA = surface area, A_p = projected area, ?_b = bulk density, ?_t = true density, and e = porosity.
^[b]    Sangamithra et al. (2016).
^[c]Sobukola et al. (2013).
^[d]Karababa and Coskuner (2007).
^[e]Kruszelnicka et al. (2022).
^[f]Coskun et al. (2006).
Discussion
Relationship Between Engineering Properties and MC
Geometric Properties
Geometric properties are fundamental for designing engineering processes such as screening/sieving, milling, threshing, and heat transfer (Darfour et al., 2022). The interaction effect of variety and MC on all the different geometric properties was statistically significant at the 5% level of significance. Adebowale et al. (2011) also found the combined effect of variety and MC significant (p < 0.05) for all engineering properties of paddy rice except bulk density.
These observations are attributed to the varied reduction and increase in length and width with an increase in MC for the different varieties (table 2). Equivalent diameter, geometric mean diameter, sphericity, aspect ratio, flatness ratios, surface, and projected area are all derived from kernel length, width, and thickness. Additionally, the study results reveal that different grain varieties can be within a similar range of moisture but have varying geometric properties. This could be attributed to the varietal differences in water absorption and diffusion behavior, influenced by the structural and internal composition of individual grain kernels. Further studies may be necessary to ascertain if these variations are affected by intraparticle behavior.
Gravimetric Properties
The interaction effect of variety and MC on all gravimetric properties was statistically significant at the 5% level of significance, and the mean separation results are presented in table 3. On the contrary, most varieties had significantly lower bulk density at 17% compared to values at 13.5% and 15.5% (table 3). However, the variation in true density with MC was not significant for all varieties evaluated (table 3). This agrees with the general reduction in bulk density of corn with an increase in MC reported by Karababa and Coskuner (2007) and Kruszelnicka et al. (2022). Sangamithra et al. (2016) reported an increase in bulk and true density with MC. Coskun et al. (2006) observed a general increase in true density with MC. The porosity had a quadratic relationship with MC with R² value of 0.339. This varied from the reported positive linear relationship observed for porosity by Coskun et al. (2006) and Karababa and Coskuner (2007). True and bulk density had linear and power relationships with MC with R² values of 0.96 and 0.46, respectively (table 4). Sangamithra et al. (2016) reported a quadratic relationship for both bulk and true densities with R² > 0.99. The study indicated that porosity at 17% was higher than values at 13.5% and 15.5% for all varieties, depicting a relative increase with MC (table 3). A similar general increase in porosity with an increase in MC was reported by Coskun et al. (2006) and Karababa and Coskuner (2007), although Sangamithra et al. (2016) and Sobukola et al. (2013) reported a reduction in porosity with MC.
Analysis of Potential to Predict and Describe Engineering Properties Based on Their MC Relationship
The observed inconsistent predictive power for both the study and reference models could be attributed to the documented limitations of simple regression models and the likely intraparticle behavior of a biological material. Simple regression assumes that the data is independent. This may not be possible with biological systems such as grains, as they always have intra relationships between different components. Negative effects are further worsened with cumulative intra relationships for pooled samples (with several varieties), as used in this study. For instance, absorption and diffusion of moisture is highly dependent on the structural and interstitial makeup of the grain, factors that are not catered for in the model, yet different grains act differently with MC. The significant interaction between moisture content and variety for most engineering properties confirms this dependence. This could be the reason for the high percent error observed for width and surface area with variety V4 in the study model.
Additionally, regression models return the mean value of the dependent variable (in this case, physical property) for a given value of an independent variable (in this case, MC) (Casson and Farmer, 2014). The higher variability among
the different varieties as observed for the geometric properties is responsible for the high prediction percent error for the surface area with the pooled samples. The model cannot satisfy all the data points but rather only those in good approximation to the mean value of the pooled samples at the respective MC levels.
The principle of curve fitting allows iterative analysis of the available data, developing term coefficients that create a best fit relationship with the algorithm of least square error or least absolute error (Motulsky and Ransnas, 1987). For instance, a linear relationship with the thickness was observed for both the study and reference model (table 4), despite the different term coefficients. The variability in term coefficients results from differences in data sets applied to the development of these models. Thus, the application of these models requires a proper understanding of experimental conditions. However, the referenced studies do not define limitations and boundaries within which their model is applicable for predictive purposes with minimal error. Furthermore, simple regression through curve fitting abandons the random error term. In fact, all studies and referenced models do not account for the error term. Therefore, in actual practice, the variation in MC will not result in an equivalent change in the physical properties (Ali and Younas, 2021), adding to the already observed errors. Prediction, on the other hand, encompasses already developed model term coefficients with fixed directional relationships to estimate the outcome of the predicted variable. Thus, prediction has to be done within the experimental limitations for minimal error. Hence, the use of existing physical properties MC-based regression models for description and predictive purposes should be done cautiously, as they are prone to greater error if similar experimental conditions to which they were developed are not met.
Conclusions
This research investigated the varietal differences of the geometric (equivalent diameter, geometric mean diameter, sphericity, aspect ratio, surface area, projected area, flatness ratio) and gravimetric (bulk density, true density, porosity) properties for ten (10) corn hybrids and evaluated the influence of moisture content. The study evaluated the potential use of regression based moisture physical properties relationships in predicting these physical properties. Predictive and descriptive comparisons indicated that reference and study models provided varying degrees of accuracy for prediction of the geometric and gravimetric properties for single variety and pooled samples. The study models provided better predictions for most properties with the pooled samples compared to reference models. The results suggest general linear and quadratic descriptive trends in physical properties for reference and study models, respectively. The study findings provide evidence that no ‘one size fits all’ moisture physical property model exists. This study advises researchers, practitioners and grain handlers to be cautious when using existing MC-based regression models of physical properties for descriptive and predictive purposes.
Acknowledgments
The author would like to thank Dr. Clair Keene and Darin Eisinger from the Department of Plant Science at North Dakota State University for providing the corn samples used for this study. The authors would also like to thank Dr. Niloy Chandra Sarker, Dr. Ibukunoluwa Ajayi-Banji, and Tayler Johnston for their supportive role during the experiments.
This research was supported in part by the intramural research program of the U.S. Department of Agriculture, National Institute of Food and Agriculture, Hatch #7005304. The findings and conclusions in this publication represents those of the author(s) and have not been formally disseminated by the U.S. Department of Agriculture and should not be construed to represent any agency determination or policy.
Nomenclature
A_p = Projected area
AR = Aspect ratio
d.b. = Dry basis
De = Equivalent diameter
Dg = Geometric mean diameter
f = Sphericity
FR = Flatness ratio
SA = Surface area
SE = Relative standard error
R² = Coefficient of determination
w.b. = Wet Basis
?_b = Bulk density
?_t = True density
ANOVA = Analysis of Variance
DNA = Deoxyribonucleic acid
MC = Moisture content
NDSU = North Dakota State University
References
Adebowale, A.-R. A., Sanni, L. O., Owo, H. O., & Karim, O. R. (2011). Effect of variety and moisture content on some engineering properties of paddy rice. J. Food Sci. Technol., 48, 551-559. https://doi.org/10.1007/s13197-010-0173-3
Adeyanju, J. A., Abioye, A. O., Adekunle, A. A., Ajala, A. S., Oloyede, A. A., & Afolayan, E. T. (2022). Assessment of physical and thermal properties of velvet bean (Mucuna pruriens) as potentials for development of processing machines. FUOYE J. Eng. Technol., 7(2), 122-125. https://doi.org/10.46792/fuoyejet.v7i2.778
Ali, P., & Younas, A. (2021). Understanding and interpreting regression analysis. Evid. Based Nurs., 24(4), 116-118. https://doi.org/10.1136/ebnurs-2021-103425
Atere, A. O., Olalusi, A. P., & Olukunle, O. J. (2016). Physical properties of some maize varieties. J. Multidiscip. Eng. Sci. Technol., 3(2), 3874-3880.
Boac, J. M., Casada, M. E., Pordesimo, L. O., Petingco, M. C., Maghirang, R. G., & Harner III, J. P. (2023). Evaluation of particle models of corn kernels for discrete element method simulation of shelled corn mass flow. Smart Agric. Technol., 4, 100197. https://doi.org/10.1016/j.atech.2023.100197
Brar, I. S., Dixit, A. K., Khurana, R., & Gautam, A. (2017). Studies on physical properties of maize (Zea mays L.) seeds. Int. J. Curr. Microbiol. Appl. Sci., 6(10), 963-970. https://doi.org/10.20546/ijcmas.2017.610.116
Casson, R. J., & Farmer, L. D. (2014). Understanding and checking the assumptions of linear regression: A primer for medical researchers. Clin. Exp. Ophthalmol., 42(6), 590-596. https://doi.org/10.1111/ceo.12358
Chakraverty, A., & Singh, R. P. (2014). Postharvest technology and food process engineering. Boca Raton, FL: CRC Press.
Chandio, F. A., Li, Y., Ma, Z., Ahmad, F., Syed, T. N., Shaikh, S. A., & Tunio, M. H. (2021). Influences of moisture content and compressive loading speed on the mechanical properties of maize grain orientations. Int. J. Agric. Biol. Eng., 14(5), 41-49.
Coskun, M. B., Yalçin, I., & Özarslan, C. (2006). Physical properties of sweet corn seed (Zea mays saccharata Sturt.). J. Food Eng., 74(4), 523-528. https://doi.org/10.1016/j.jfoodeng.2005.03.039
Darfour, B., Ayeh, E. A., Odoi, K. M., & Mills, S. W. (2022). Physical characteristics of maize grain as influenced by varietal and moisture differences. Int. J. Food Prop., 25(1), 1351-1364. https://doi.org/10.1080/10942912.2022.2077756
Erenstein, O., Jaleta, M., Sonder, K., Mottaleb, K., & Prasanna, B. M. (2022). Global maize production, consumption and trade: Trends and R&D implications. Food Secur., 14(5), 1295-1319. https://doi.org/10.1007/s12571-022-01288-7
Eze, P. C., Eze, C. N., & Ide, P. E. (2022). Effect of moisture content and variety on the some physical and aerodynamic properties of sesame seeds relevant to its processing. ARRUS J. Eng. Technol., 2(2), 78-91. https://doi.org/10.35877/jetech678
Hunt, E. D., Birge, H. E., Laingen, C., Licht, M., McMechan, J., Baule, W., & Connor, T. (2020). A perspective on changes across the US Corn Belt. Environ. Res. Lett., 15, 071001. https://doi.org/10.1088/1748-9326/ab9333
Karababa, E., & Coskuner, Y. (2007). Moisture dependent physical properties of dry sweet corn kernels. Int. J. Food Prop., 10(3), 549-560. https://doi.org/10.1080/10942910601003981
Kazarian, E. A., & Hall, C. W. (1965). Thermal properties of grain. Trans. ASAE, 8(1), 33-37. https://doi.org/10.13031/2013.40420
Khan, D. M., Yaqoob, A., Zubair, S., Khan, M. A., Ahmad, Z., & Alamri, O. A. (2021). Applications of robust regression techniques: An econometric approach. Math. Probl. Eng., 2021, 6525079. https://doi.org/10.1155/2021/6525079
Kruszelnicka, W. (2021). Study of selected physical-mechanical properties of corn grains important from the point of view of mechanical processing systems designing. Materials, 14(6), 1467. https://doi.org/10.3390/ma14061467
Kruszelnicka, W., Chen, Z., & Ambrose, K. (2022). Moisture-dependent physical-mechanical properties of maize, rice, and soybeans as related to handling and processing. Materials, 15(24), 8729. https://doi.org/10.3390/ma15248729
Mohsenin, N. N. (1971). Physical properties of plant and animal materials: v. 1: Physical characteristics and mechanical properties. New York, NY: Routledge. https://doi.org/10.4324/9781003062325
Motulsky, H. J., & Ransnas, L. A. (1987). Fitting curves to data using nonlinear regression: A practical and nonmathematical review. FASEB J., 1(5), 365-374. https://doi.org/10.1096/fasebj.1.5.3315805
Mousaviraad, M., & Tekeste, M. Z. (2020). Effect of grain moisture content on physical, mechanical, and bulk dynamic behaviour of maize. Biosyst. Eng., 195, 186-197. https://doi.org/10.1016/j.biosystemseng.2020.04.012
Newes, E., Clark, C. M., Vimmerstedt, L., Peterson, S., Burkholder, D., Korotney, D., & Inman, D. (2022). Ethanol production in the United States: The roles of policy, price, and demand. Energy Policy, 161, 112713. https://doi.org/10.1016/j.enpol.2021.112713
O’Malley, J., & Searle, S. (2021). The impact of the US renewable fuel standard on food and feed prices. Technical report. International Council on Clean Transportation. Retrieved from https://theicct.org/wp-content/uploads/2021/06/RFS-and-feed-prices-jan2021.pdf
Osborne, S. L., Lehman, R. M., & Rosentrater, K. A. (2016). Grain and biomass nutrient uptake of conventional corn and their genetically modified isolines. J. Plant Nutr., 39(14), 2047-2055. https://doi.org/10.1080/01904167.2016.1187753
Ozturk, I., Kara, M., Yildiz, C., & Ercisli, S. (2009). Physico-mechanical seed properties of the common Turkish bean (Phaseolus vulgaris) cultivars ‘Hinis’ and ‘Ispir’. N. Z. J. Crop Hortic. Sci., 37(1), 41-50. https://doi.org/10.1080/01140670909510248
Phinney, D. M., Frelka, J. C., & Heldman, D. R. (2017). Composition-based prediction of temperature-dependent thermophysical food properties: Reevaluating component groups and prediction models. J. Food Sci., 82(1), 6-15. https://doi.org/10.1111/1750-3841.13564
Revilla, P., Alves, M. L., Andelkovic, V., Balconi, C., Dinis, I., Mendes-Moreira, P.,... Malvar, R. A. (2022). Traditional foods from maize (Zea mays L.) in Europe. Front. Nutr., 8, 683399. https://doi.org/10.3389/fnut.2021.683399
Ropelewska, E. (2018). Effect of grinding on thermal properties of wheat grain. J. Consum. Prot. Food Saf., 14(2), 139-146. https://doi.org/10.1007/s00003-018-1200-y
Sangamithra, A., Gabriela, J. S., Prema, R. S., Nandini, K., Kannan, K., Sasikala, S., & Suganya, P. (2016). Moisture dependent physical properties of maize kernels. Int. Food Res. J., 23(1), 109.
Santos, M. M., Devilla, I. A., Lisboa, C. F., de Carvalho Melo, P., & Antunes, A. M. (2016). Determining and modeling the physical, thermal and aerodynamic properties of Pinto beans with different water contents. African J. Agric. Res., 11(46), 4735-4741. https://doi.org/10.5897/ajar2016.11690
Singh, A. K. (2016). Moisture dependent thermal properties of ground flaxseed (Linum usitatissimum). Agric. Eng. Int.: CIGR J., 18(1), 256-262.
Singh, K. K., & Goswami, T. K. (1996). Physical properties of cumin seed. J. Agric. Eng. Res., 64(2), 93-98. https://doi.org/10.1006/jaer.1996.0049
Sobukola, O. P., Kajihausa, O. E., Onwuka, V. I., & Esan, T. A. (2013). Physical properties of high quality maize (Swam 1 variety) seeds (Zea mays) as affected by moisture levels. African J. Food Sci., 7(1), 1-8. https://doi.org/10.5897/AJFS11.167
Tarighi, J., Mahmoudi, A., & Alavi, N. (2011). Some mechanical and physical properties of corn seed (Var. DCC 370). Afr. J. Agric. Res., 6(16), 3691-3699. https://doi.org/10.5897/AJAR10.521
Tunde-Akintunde, T. Y., & Akintunde, B. O. (2007). Effect of moisture content and variety on selected physical properties of beniseed. Agric. Eng. Int.: CIGR J.

American Society of
Agricultural and Biological Engineers

2950 Niles Road, St. Joseph, MI 49085
Phone: +12694290300 Fax: +12694293852
Copyright © 2026 American Society of Agricultural and Biological Engineers

Table 1. Models used in the study.^[a]
Model	Model Expression
Linear model	Y = aMC + b
Quadratic model	Y = aMC² + bMC + c
Power model	Y = aMC^b + c
^[a] Y = Engineering property; MC = moisture content (%); and a, b, c = model constants.

Table 2. The effect of variety and moisture content on the geometric dimensions of corn.^[a]
Variety	MC	L	W	T	D_e	D_g	f	AR	FR	SA	A_p
V1	13.5	12.0^bcd	8.5^abcde	4.8^defg	8.1^bcdef	7.8^bcdef	0.65^defg	0.71^efghij	0.59^ef	193.4^bcdefgh	80.0^bc
	15.5	11.1^efgh	8.2^cdefg	4.9^abcdefg	7.8^cdefgh	7.6^defgh	0.69^abcd	0.75^abcdefghi	0.60^cdef	183.0^defghi	72.0^efghi
	17	11.7^cde	8.6^abcd	5.0^abcdefg	8.1^bc	7.9^bcd	0.68^abcdefg	0.74^abcdefghi	0.58^cdef	197.8^bcd	78.6^bcde
V2	13.5	10.7^h	7.8^gh	5.4^abc	7.7^gh	7.6^defgh	0.71^ab	0.73^cdefghi	0.69^ab	183.0^defghi	65.9ⁱ
	15.5	10.8^h	8.4^abcde	4.9^abcdefg	7.9^cdefgh	7.7^cdefgh	0.71^ab	0.79^abc	0.58^cdef	186.1^bcdefghi	72.1^efghi
	17	10.9^gh	8.2^abcd	4.9^bcdefg	7.9^cdefg	7.7^bcdefg	0.71^ab	0.77^ab	0.57^def	187.6^bcdefgh	74.1^cdefg
V3	13.5	10.9^fgh	7.6^h	5.4^ab	7.7^gh	7.7^cdefgh	0.69^ab	0.69^ghij	0.72^a	184.9^cdefghi	65.8ⁱ
	15.5	12.3^b	7.9^efgh	5.2^abcde	8.1^bcde	7.9^bcd	0.64^efg	0.65^j	0.65^abcd	198.1^bcd	76.7^bcdefg
	17	10.8^h	7.9^fgh	5.5^a	7.8^defgh	7.7^bcdefg	0.72^a	0.73^bcdefghi	0.69^ab	187.3^bcdefgh	66.7^hi
V4	13.5	11.6^cdef	9.0^a	4.9^abcdefg	8.3^ab	8.0^ab	0.69^abc	0.78^abcd	0.58^f	201.7^ab	81.8^b
	15.5	12.9^a	8.9^ab	4.9^abcdefg	8.5^a	8.3^a	0.64^fg	0.69^ij	0.56^ef	217.1^a	90.9^a
	17	11.2^efgh	8.7^abc	5.1^abcdef	8.1^bcd	7.9^bcde	0.71^ab	0.78^abcd	0.59^cdef	197.0^bcdef	77.3^bcdef
V5	13.5	10.9^fgh	8.4^bcdef	4.7^efg	7.8^efgh	7.5^fgh	0.69^abcd	0.76^abcde	0.56^ef	179.3^ghi	72.6^defgh
	15.5	11.2^efgh	8.6^abcd	4.6f^g	7.9^cdefgh	7.6^efgh	0.68^abcdef	0.77^abcd	0.54^f	181.9^efghi	75.7^bcdefg
	17	10.8^h	8.5^abcde	4.7^efg	7.8^fgh	7.5^gh	0.69^abcd	0.78^abcd	0.55^f	177.8^hi	72.1^efghi
V6	13.5	11.2^efgh	8.1^defgh	5.3^abcd	7.9^cdefg	7.8^bcdefg	0.69^ab	0.73^defghi	0.68^abc	190.6^bcdefgh	71.4^fghi
	15.5	11.5^cdefg	8.3^cdefg	5.1^abcdefg	8.0^bcdefg	7.8^bcdefg	0.68^abcde	0.73^defghi	0.61^cdef	193.4^bcdefgh	75.1^cdefg
	17	11.1^efgh	8.1^defgh	5.2^abcd	7.9^cdefg	7.8^bcdefg	0.70^ab	0.74^abcdefghi	0.65^abc	190.9^bcdefgh	71.2^fghi
V7	13.5	11.3^efgh	7.9^fgh	4.6^g	7.6^h	7.4^h	0.65^cdefg	0.69^fghij	0.58^cdef	171.5ⁱ	70.2^ghi
	15.5	12.1^bc	8.3^cdefg	4.6f^g	7.9^bcdefg	7.7^bcdefg	0.63^g	0.69^hij	0.55^f	186.7^bcdefghi	78.9^bcd
	17	11.3^efgh	8.5^abcde	4.6f^g	7.8^cdefgh	7.6^fgh	0.67^bcdefg	0.8^abcdefgh	0.54^f	181.4^fghi	75.1^cdefg
V8	13.5	11.6^cde	8.4^bcdef	4.9^bcdefg	8.0^bcdefg	7.8^bcdefg	0.67^bcdefg	0.73^cdefghi	0.58^cdef	191.3^bcdefgh	76.9^bcdef
	15.5	11.3^efgh	8.5^abcde	4.9^bcdefg	7.9^cdefg	7.7^bcdefg	0.69^abcd	0.76^abcdef	0.57^cdef	188.5^bcdefgh	74.9^cdefg
	17	11.4^defg	8.4^bcde	5.3^abc	8.1^bc	7.9^bc	0.69^ab	0.74^abcdefghi	0.64^abcde	199.3^bc	75.7^bcdefg
V9	13.5	11.3^efgh	8.5^abcd	5.2^abcde	8.1^bcd	7.9^bcde	0.70^ab	0.75^abcdefg	0.61^bcdef	197.7^bcde	75.9^bcdefg
	15.5	11.5^defg	8.5^abcde	4.9^abcdefg	8.0^bcdefg	7.8^bcdef	0.68^abcde	0.74^abcdefghi	0.59^cdef	193.9^bcdefg	76.6^bcdefg
	17	11.2^efgh	8.6^abcd	5.2^abcde	8.1^bcd	7.9^bcde	0.71^ab	0.77^abcd	0.61^bcdef	197.1^bcdef	75.8^bcdefg
V10	13.5	11.4^defg	8.8^abc	4.8^cdefg	8.1^bcd	7.9^bcdef	0.68^abcd	0.77^abcd	0.56^f	194.7^bcdefg	78.9^bcd
	15.5	11.1^efgh	8.5^abcd	4.9^abcdefg	7.9^bcdefg	7.7^bcdefg	0.69^ab	0.77^abcde	0.58^cdef	189.3^bcdefgh	74.4^cdefg
	17	10.9^fgh	8.7^abc	5.0^abcdefg	8.0^bcdefg	7.8^bcdefg	0.71^ab	0.79a	0.57^cdef	191.6^bcdefgh	74.9^cdefg
Mean		11.35	8.39	4.99	7.97	7.77	0.69	0.74	0.6	190.48	74.95
SD		0.92	0.80	0.75	0.45	0.46	0.06	0.09	0.12	22.49	9.96
SE		0.02	0.02	0.02	0.01	0.01	0.002	0.002	0.003	0.58	0.26
CV		8.17	9.58	15.10	5.72	5.86	8.65	12.08	19.33	11.81	13.29
P_v		<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
P_m		<0.0001	0.001	0.007	0.017	0.078^ns	<0.0001	<0.0001	0.002	0.072^ns	<0.0001
P_vm		<0.0001	<0.0001	0.043	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
^[a] MC = moisture content, L = length (mm), W = width (mm), T = thickness (mm), D_e = equivalent diameter (mm), D_g = geometric mean diameter (mm), f = sphericity, AR = aspect ratio, FR = flatness ratio, SA = surface area (mm²), A_p = projected area (mm²), SD = standard deviation, SE = standard error, CV = coefficient of variation, P_v = p value of variety, P_m = p value of moisture content, and P_vm = p value of interaction. Values are means of 50 replicates of corn seeds. Means with different letters in a column are significantly different at 0.05 level of significance. ns = not significantly different at 0.05 level of significance.

Table 3. The effect of variety and moisture content on the gravimetric properties of corn.^[a]
Variety	MC	Bulk Density (kg/m³)	True Density (kg/m³)	Porosity (%)
V1	13.5	763.3^abc	1224.8^ab	37.67^defgh
	15.5	759.1^bcd	1237.4^ab	38.59^bcdefgh
	17	726.4^ghi	1212.1^ab	40.07^bcdefg
V2	13.5	749.9^de	1212.1^ab	34.13^bcdefgh
	15.5	741.9^ef	1205.5^ab	38.44^bcdefgh
	17	715.9^hij	1212.9^ab	40.93^abcde
V3	13.5	755.1^cd	1237.4^ab	38.96^bcdefgh
	15.5	765.3^abc	1212.1^ab	36.91^gh
	17	715.5^ij	1224.8^ab	41.57^ab
V4	13.5	762.3^abcd	1200.2^ab	38.91^bcdefgh
	15.5	769.7^ab	1212.9^ab	36.03^gh
	17	705.8^jk	1200.2^ab	41.18^abc
V5	13.5	772.5^a	1237.4^ab	37.56^efgh
	15.5	772.5^a	1224.8^ab	36.87^gh
	17	729.8^fg	1176.5^ab	37.97^cdefgh
V6	13.5	750.2^de	1224.8^ab	38.73^bcdefgh
	15.5	753.8^cde	1212.1^ab	37.81^cdefgh
	17	707.5^jk	1200.2^ab	41.03^abcd
V7	13.5	699.8^k	1176.5^b	40.52^abcdef
	15.5	715.7^ij	1184.1^b	39.56^bcdefgh
	17	679.7^l	1212.1^ab	43.92^a
V8	13.5	753.9^cde	1237.4^ab	39.06^bcdefgh
	15.5	770.2^ab	1218.7^ab	36.81^gh
	17	716.3^hij	1212.1^ab	40.90^abcde
V9	13.5	772.8^a	12125.5^ab	36.89^gh
	15.5	759.4^bcd	1213.9^ab	37.44^fgh
	17	735.9^fg	1200.2^b	38.67^bcdefgh
V10	13.5	771.4^ab	1224.8^ab	37.00^gh
	15.5	764.5^abc	1266.7^a	39.51^bcdefgh
	17	728.9^gh	1176.5^b	38.04^cdefgh
Mean		742.83	1213.9	38.79
SD		26.05	26	2.00
SE		2.75	2.14	0.21
CV		3.51	2.14	5.16
P_v		<0.001	0.019	<0.001
P_m		<0.001	0.002	<0.001
P_vm		<0.001	0.002	<0.001
^[a] SD = standard deviation, SE = standard error, CV = coefficient of variation, P_v = p value of variety, P_m = p value of moisture content, and P_vm = p value of interaction. Values are mean of three replicates for properties. Means with different letters in a column are significantly different at 0.05 level of significance.

Table 4. Performance of the study and reference moisture based prediction models for engineering properties of corn.^[a]
	Property	Study Relationships Between Moisture Content and Physical Properties		Prediction Performance of Models
		Study Model Equation	R²	Reference Model	Variety V4		Pooled Sample
		Study Model Equation	R²	Reference Model	SE₁ (%)	SE₂ (%)	SE₁ (%)	SE₂ (%)
Geometric properties	L		0.038	^[b]	2.8	1.8	14.8	3.18
	W		0.991	^[b]	15.4	14.3	3.13	2.96
	T		0.996	^[b]	1.15	1.15	4.62	4.8
	D_e		0.003	^[c]	2.52	2.52	4.86	4.86
	D_g		0.969	^[b]	0.39	0.4	1.03	1.08
	f		0.974	^[d]	0.06	8.1e4	0.21	2.5e5
	AR		0.970	^[e]	0.47	0.36	1.58	1.05
	FR		0.975	^[e]	0.22	0.22	0.89	0.99
	SA		0.969	^[b]	53.5	55.4	102.9	111.2
	A_p		0.012	^[f]	1.4e3	62.4	5.62	90.9
Gravimetric properties	?_t		0.960	^[d]	1.84	7.2	5.2	16.1
	?_b		0.460	^[d]	0.59	6.49	5.01	17.2
	e		0.339	^[d]	0.49	2.46	0.92	5.26
^[a]R² = Coefficient of determination for the fitted study model, SE₁= percent standard error for prediction based on study model, SE₂ = percent standard error for prediction based on reference model, MC = moisture content, L = length, W = width, T = thickness, D_e = equivalent diameter, D_g = geometric mean diameter, f = sphericity, AR = aspect ratio, FR = flatness ratio, SA = surface area, A_p = projected area, ?_b = bulk density, ?_t = true density, and e = porosity. ^[b] Sangamithra et al. (2016). ^[c]Sobukola et al. (2013). ^[d]Karababa and Coskuner (2007). ^[e]Kruszelnicka et al. (2022). ^[f]Coskun et al. (2006).

Investigating the Physical Properties of Corn Varieties

Published in Journal of the ASABE 67(3): 631-639 (doi: 10.13031/ja.15769). Copyright 2024 American Society of Agricultural and Biological Engineers.

Materials and Method

Sample Collection

Moisture Content Conditioning

Determination of Geometric Properties of Corn

Determination of Gravimetric Properties on Corn

Bulk Density, True Density, and Porosity

Statistical and Relational Analysis

Statistical Analysis

Simple Regression Analysis and Prediction of Engineering Properties Using Study and Reference Models

Results

Geometric Properties

Gravimetric Properties

Prediction and Description of Engineering Properties Based on Their MC Relationship

Discussion

Relationship Between Engineering Properties and MC

Geometric Properties

Gravimetric Properties

Analysis of Potential to Predict and Describe Engineering Properties Based on Their MC Relationship

Conclusions

Acknowledgments

Nomenclature

References