Hindawi Applied Bionics and Biomechanics Volume 2021, Article ID 8573897, 16 pages https://doi.org/10.1155/2021/8573897 Research Article Biomimetic Rotary Tillage Blade Design for Reduced Torque and Energy Requirement 1 2,3 1 2,3 2,3 Yuwan Yang , Jin Tong , Yuxiang Huang, Jinguang Li, and Xiaohu Jiang College of Mechanical and Electronic Engineering, Northwest A&F University, Yangling 712100, China College of Biological and Agricultural Engineering, Jilin University, Changchun 130022, China Key Laboratory of Bionic Engineering, Ministry of Education, Jilin University, Changchun 130022, China Correspondence should be addressed to Jin Tong; email@example.com Received 11 April 2021; Accepted 14 September 2021; Published 28 September 2021 Academic Editor: Marco Parente Copyright © 2021 Yuwan Yang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A rotary cultivator is a primary cultivating machine in many countries. However, it is always challenged by high operating torque and power requirement. To address this issue, biomimetic rotary tillage blades were designed in this study for reduced torque and energy requirement based on the geometric characteristics (GC) of ﬁve fore claws of mole rats, including the contour curves of the ﬁve claw tips (GC-1) and the structural characteristics of the multiclaw combination (GC-2). Herein, the optimal blade was selected by considering three factors: (1) the ratio (r) of claw width to lateral spacing, (2) the inclined angle (θ) of the multiclaw combination, and (3) the rotary speed (n) through the soil bin tests. The results showed that the order of inﬂuence of factors on torque was n, r, and θ; the optimal combination of factors with the minimal torque was r =1:25, θ =60 , and n = 240 rpm. Furthermore, the torque of the optimal blade (BB-1) was studied by comparing with a conventional (CB) and a reported optimal biomimetic blade (BB-2) in the soil bin at the rotary speed from 160 to 320 rpm. Results showed that BB-1 and BB-2 averagely reduced the torque by 13.99% and 3.74% compared with CB, respectively. The ﬁeld experiment results also showed the excellent soil-cutting performance of BB-1 whose average torques were largely reduced by 17.00%, 16.88%, and 21.80% compared with CB at diﬀerent rotary speeds, forward velocities, and tillage depths, respectively. It was found that the geometric structure of the ﬁve claws of mole rats could not only enhance the penetrating and sliding cutting performance of the cutting edge of BB-1 but also diminish the soil failure wedge for minimizing soil shear resistance of BB-1. Therefore, the GC of ﬁve fore claws of mole rats could inspire the development of eﬃcient tillage or digging tools for reducing soil resistance and energy consumption. 1. Introduction ing the power requirement [6, 7]. Then, some geometrical optimizations were proposed to reduce the torque require- The rotary cultivator is a primary cultivating machine in ment of the rotary tillage blade. For example, a saw tooth many countries including Bangladesh, India, Nepal, angular blade of the rotary cultivator was designed and tested in ﬁeld . The straight blade required the least Thailand, Japan, Malaysia, the People’s Republic of China, and South Korea [1–4]. It can eﬃciently complete the oper- torque, average power, peak power, speciﬁc energy, and ations of soil mixing, turning, pulverizing, puddling, and eﬀective speciﬁc energy at 375-500 rpm . The optimal leveling and thereby create good seedbeds for crop growing. biomimetic blades based on the geometrical structure of However, the rotary cultivator is always challenged by the one claw tip of mole rats required lower torque than the conventional universal blade during soil-rototilling and large power requirement, since it needs about 80% of the power for the interaction between the rotary tillage blade stubble-cutting operations . In fact, drag reduction can and soil, such as soil cutting and throwing . Studies be achieved by electron osmosis, vibration, magnetization, showed that the blade shape was an important factor aﬀect- or bionic tillage component methods. However, many 2 Applied Bionics and Biomechanics Rotary tillage blade Down-cut rotation Humerus Rotary shaft Five claws (a) (b) Figure 1: (a) Motion diagram of the mole claws according to Scott and Richardson . (b) Motion diagram of the rotary tillage blade. challenges remain in relation to the geometrical parameters of metic rotary tillage blades for eﬃcient working. And the torque requirements of biomimetic blades were researched the rotary tillage blade. In this study, novel rotary tillage blades were designed for minimizing the power requirements. through soil bin tests and ﬁeld experiments at diﬀerent till- Bionic methods based on biological principles have been age conditions. The aim was to achieve an optimal biomi- applied in engineering systems and modern technologies to metic blade (BB-1) and reveal the eﬀects of the geometric improve or create new techniques [11, 12]. The sources of characteristics of the ﬁve fore claws on the soil-cutting performance. successful biomimetic designs include the excellent geomet- ric characteristics of animals, such as the segmented body of earthworms, the nonsmooth morphology of the head of 2. Geometric Characteristics of Five Fore dung beetles, and the strong and sharp claws of mole rats. Claws of Mole Rats For example, a biomimetic bulldozer featuring the wavy body surface of earthworm was able to crush soil clods than 2.1. Five Fore Claws of Mole Rats. Mole rats (Scaptochirus, a traditional smooth bulldozer . A biomimetic rough Talpidae) (Figure 2(a)) were obtained from the northeast curved soil-cutting blade based on the geometrical rough region of China where they are most common and inhabit structures of soil-burrowing animals could experimentally mostly underground. Their broad and strong hands which reduce soil adhesion and friction and thereby considerably consist of ﬁve diﬀerent claws (see Figure 2(b)) were scanned decrease the draught forces . A biomimetic disc based by a three-dimensional laser scanner (Handyscan700, on the proﬁle curves of the second digit of mole fore claws Creaform, Canada), and the point cloud of the ﬁve claws performed better in structural strength and cutting eﬃciency (see Figure 3) was created with the reverse engineering using ﬁnite element analysis . software program of ImageWare (version 13, Siemens PLM Mole rats are born diggers that have adapted to a strict software, Germany). After a series of procedures, such as subterranean lifestyle . They possess outstanding digging smoothing, reducing, and simpliﬁcation, the ﬁve claws were performance in digging tunnels over 91 m overnight. A mole reconstructed into a surface, and then, it was generated as an rat exerts the greatest power to the soil following the rotation entity from a surface in SolidWorks software. Then, the char- of humerus around its own long axis and repeats this process acteristics of the contour curves of the ﬁve claw tips and the [16, 17] as shown in Figure 1(a). The principle of the claw- multiclaw combination were described in the following. soil interaction is similar to that between the rotary tillage blade and soil as shown in Figure 1(b). In fact, the claw- 2.2. The Contour Curve Characteristics of Five Claw Tips soil interaction is closely related with the soil-cutting perfor- (GC-1). The claw tip of mole rats reduces the soil penetra- mance of animals . Each of the mole hand has ﬁve ﬁn- tion resistance and makes the digging more eﬃcient and fast. gers, which each possess a large, sharp, and powerful claw. In this study, the curves of the ﬁve claw tips of mole rats Ji et al.  only characterized the geometry of the second were extracted in the reverse engineering software program claw of mole rats. However, when a mole rat cuts soil, the of ImageWare (version 13, Siemens PLM software, ﬁve fore claws always work synergistically with an opened- Germany). The extracted points of the curves of the ﬁve claw up but coplane conﬁguration, which is described as multic- tips were adjusted and plotted in the AutoCAD 2014 soft- law combination, for eﬃcient cutting . Moreover, the ware for obtaining their data information. These point data contour curves of the ﬁve claw tips also signiﬁcantly aﬀect information were imported to the OriginPro 9.1 software the soil-cutting performance of mole rats. for quantitative analysis (Figure 4). The contour curves of In this study, the geometric characteristics (GC) of the the ﬁve claw tips were ﬁtted based on the least square ﬁve fore claws, including the contour curves of the ﬁve claw method, and the Gaussian function equations of the ﬁtted tips (GC-1) and the structural characteristics of the multic- contour curves were obtained simultaneously as shown in law combination (GC-2) inspired us to optimize the biomi- Equation (1). The values of the coeﬃcient of determination Applied Bionics and Biomechanics 3 1 mm (a) (b) Figure 2: (a) A mole rat and (b) the ﬁve fore claws. pattern, and soil forces. Therefore, to clarify the structural and working characteristics of multiclaw combination, we deﬁned the ratio q of length (L) to width (W) of each claw in Equation (2) and the ratio r of width (W) to lateral spac- ing (Δx) in Equation (3). The values of q and r were calcu- lated according to Equations (2) and (3) and presented in Table 2. Moreover, a mathematical model was developed to describe the structural characteristics of the multiclaw combination as shown in Equation (4). Figure 3: The point clouds and contour curves of the ﬁve claws of the mole rat. ð2Þ q = , (R ) were all above 0.95 showing that the ﬁtting curves were close to the contour curves of the ﬁve claw tips. Further- Δx more, the sums of squares for error (SSE) were all less than ð3Þ r = , 0.05 indicating that the Gaussian function equations could accurately describe the contour curve characteristics of ﬁve claw tips. w =0:5W + Δx + Δx + Δx + Δx +0:5W 0 1 12 23 34 45 5 2 2 − x+b /c − x+b /c =0:5W + r W + r W + r W + r W +0:5W ðÞ ðÞ ðÞ ðÞ 1 1 2 2 1 12 2 23 3 34 3 45 4 5 y = a e + a e , ð1Þ 1 2 L L L L L L 1 2 3 3 4 5 =0:5 + r + r + r + r +0:5 , 12 23 34 45 q q q q q q where a , a , b , b , c , and c are the coeﬃcient of the 1 2 3 3 4 5 1 2 1 2 1 2 ﬁtting equations and recorded in Table 1. ð4Þ 2.3. Structural Characteristics of the Multiclaw Combination where L is the length of the claw (mm), W is the width of (GC-2). The ﬁve claws diﬀer in structural characteristics. the claw (mm), Δx is the lateral spacing between adjacent From Figure 2(b), the 3rd claw was considerably longer than claws (mm), q is the ratio of the length (L) and the width the 1st, 2nd, 4th, and 5th claws with the 5th claw being very (W), r is the ratio of the width (W) and the lateral spacing small. The claw dimension and the lateral spacing diﬀerence (Δx), w is the total width of the multiclaw combination between two adjacent claws are important parameters for 0 (mm), L is the length of the i-th claw (mm), W is the width deﬁning the structure characteristics of multiclaw combina- i i of the i-th claw (mm), q is the ratio of the length (L) and the tion as shown in Figure 2(b). The length (L)isdeﬁned as the width (W) of the i-th claw, r is the ratio of the width (W) vertical extension of a claw. The variation of the horizontal ij dimension along the claw direction is typically very small, and the lateral spacing (Δx) of the i-th claw and the j-th and thus, the width (W) is determined at the middle of the claw, and j = i +1, i =1,2, 3,4. claw (L/2). The lateral spacing between two adjacent claws The value of q ranged from 2.63 to 3.41 (Table 2), indi- is labelled as Δx. All of the parameters were recorded and cating the ﬁve claws belonged to narrow tines, and the value presented in Table 2 according to the previous literature of r ranged from 1.12 to 1.60. Clearly, the multiclaw combi- . From the studies by Godwin [21, 22], the working nation also could be regarded as a multitine combination, depth/width ratio and the tine spacing are the important and the lateral space between two adjacent tines was adjust- parameters to determine the categories of blades, soil failure able to allow mole rats to adapt to more circumstances. 𝛥x L 4 Applied Bionics and Biomechanics 6.4 6.2 6.2 6.15 6.1 5.8 6.05 5.6 6 5.95 5.4 5.9 5.2 5.85 5.8 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 1.2 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 x (mm) x (mm) st nd (a) 1 claw (b) 2 claw 9.2 4 9 3.9 3.8 8.8 3.7 8.6 3.6 8.4 3.5 8.2 3.4 3.3 7.8 –1 –0.5 0 0.5 1 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 x (mm) x (mm) rd th (c) 3 claw (d) 4 claw 2.7 2.6 2.5 2.4 2.3 2.2 2.1 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 x (mm) th (e) 5 claw Figure 4: The ﬁtting curves of the ﬁve claw tips of mole rats. 3. Design of Biomimetic Rotary Tillage Blades cutting edge, at the maximal rotary radius of the blade, and along the end of the blade, respectively. The conﬁgura- tion of the bionic rotary tiller blade is shown in A conventional rotary tillage blade was composed of the holder, lengthwise surface, transition surface, and scoop sur- Figure 5(b). The contour curves were magniﬁed to ﬁtthe face as shown in Figure 5(a). When a tillage operation is per- dimension of the corresponding claw. The values of q and formed in the ﬁeld, the lengthwise surface of blade will touch r referred to that of the prototype of the multiclaw combina- soil ﬁrstly and cut open the soil, cut oﬀ, or push aside the tion. And the whole width (w ) of the bionic structure was ground stalks and weeds; and then, the scoop surface of set to meet the dimension requirement of the rotary tillage blade will cut soil transversely and mix, turn, pulverize, blade. and throw the tilled soil . Consequently, more energy is required by the scoop surface than the lengthwise surface 4. Soil Bin Tests during the tillage operation of blade . According to the geometric characteristics of ﬁve-claw combination described 4.1. Test Preparation. The rotary tillage tests were conducted above, the ﬁve claws were arranged along the scoop surface in an indoor soil bin (Figure 6(a)) at the Jilin University, of blade for reducing the soil resistance. The ﬁrst two claws, China. The soil bin (40 m long, 3 m wide, and 0.8 m deep) the 3rd claw, and the last two claws were located along the was used to provide a repeatable soil condition for the y (mm) y (mm) y (mm) y (mm) y (mm) Applied Bionics and Biomechanics 5 Table 1: Coeﬃcients of ﬁtting equations of the ﬁve claw tips. (Figure 7). The section (0.8 m wide and 20 m long) was set with two transition parts at both ends and a 5 m long transi- Coeﬃcient of Claw tion part which well met the test conditions. The 10 m long equation 1st 2nd 3rd 4th 5th stable part remained enough for collection of the experimen- 1.932 4.251 8.196 0.5015 1.657 1 tal data. Two rotary tillage blades were ﬁtted at 180 out of phase by the holders. The forward speed was maintained at b -0.6218 -0.7791 -0.6644 0.6707 0.5445 -1 3kmh and the operating depth at 80 mm following the 0.7628 0.864 1.569 0.4673 0.5248 Government Standard of Rotary Tiller in China, the same 5.789 5.658 5.318 3.937 2.473 for all sections. 0.4466 0.6178 1.061 -0.1576 -0.2982 4.2. Regression Test. Regression design could eﬃciently 1.563 1.178 1.165 1.673 0.7749 obtain the suﬃcient and accurate information of the SSE 0.03953 0.01268 0.006721 0.003073 0.001469 response variable by selecting the values of the regression R 0.9965 0.9803 0.9986 0.9975 0.9985 variable . Box-Behnken design is a common way to 2 combine the diﬀerent factors to develop a mathematical Adjusted R 0.9959 0.9738 0.9983 0.9968 0.998 relationship between the targets and the factors with cost- RMSE 0.03899 0.02907 0.01789 0.01272 0.009897 eﬀectiveness and fewer tests . In order to simplify the tests and ensure the accuracy at the same time, these tests used Box-Behnken design to select factor values, conducting Table 2: Geometrical parameters and structural elements of the the test arrangements in soil bin. Due to the adjustability of multiclaw combination. the lateral spacing between adjacent claws, r was varied and L W Adjacent Δx regarded as one inﬂuential factor aﬀecting the working per- L Δx Claw q = r = W W (mm) (mm) claws (mm) formance of the rotary tillage blade. And the inclined angle st st nd (θ) of the multiclaw combination, deﬁned as the angle 1 6.47 2.34 2.76 1 and 2 3.86 1.60 nd nd rd between the ﬁve claws and the vertical line (Figure 5(b)), also 2 7.83 2.42 3.24 2 and 3 2.89 1.12 rd rd th aﬀected the soil-cutting performance of the biomimetic 3 8.82 2.59 3.41 3 and 4 3.04 1.17 rotary tillage blades and was selected as an inﬂuence factor. th th th 4 7.99 2.46 3.25 4 and 5 3.22 1.31 Moreover, rotary speed (n) also played an important role on th 5 5.43 2.06 2.63 the working performance of rotary tiller blades  and was as another inﬂuence factor. Here, the ratio (r)of width to lat- eral spacing was set at an interval of 0.25 from 1.25 to 1.75, the experiment. The soil used for the experiments (46% sand, ° ° inclined angle (θ) was changed from 50 to 70 at an interval of 33% silt, and 21% clay content) is a loamy soil which is rep- 10 , and the rotary speed (n) ranged from 160 to 320 rpm at an resentative of a large proportion of crop-growing regions in interval of 80 rpm. In total, nine bionic rotary tiller blades were northeast China. The soil preparation involved adding a pre- designed for this study. For the experiments, all the bionic determined amount of water to reach the targeted moisture blades were wire-electrode cut, tapered, and sharpened to content and in the following day loosening by the rotary cul- achieve the bionic structures (Figure 8). tivator (1GKN-125 made by Lianyungang Weidi Machinery A mathematical relationship of torque with the three fac- Co., Ltd.), leveling by a scraper blade, and compacting by a tors (r, θ, n) was established in the Design-Expert Software roller. Finally, the soil bed was prepared with an average soil -3 (Equation (5)). The factor-level coding was given in moisture content of 18.23% d.b., a bulk density of 1957 kg m , Table 3. The test scheme included twelve factorial points and a soil compaction of 0.68 MPa, which was suitable for and ﬁve zero points as shown in Table 4. In order to identify rotary tillage operation. the inﬂuence of the above parameters on the torque, a qua- The test equipment consisted of a power control unit, a dratic mathematical regression method was used to establish data collection unit, a power transmission unit, and the the regression equation between each factor and the torque. rotary tillage blades as shown in Figure 6(a). The power con- Each test was repeated three times, and the average value of trol unit mainly included the soil bin trolley, which provided the torque obtained from the data collection unit was used the blades the forward and rotary power. The data collection for result analysis. unit included a torque sensor (CYB-803S, 0-±100 N m) and a computer. The power transmission unit consisted of a uni- versal joint coupling, a gear box (1 : 1 transmission ratio), a T = b + b x + b x + b x + b x x + b x x + b x x 0 1 1 2 2 3 3 4 1 2 5 1 3 6 2 3 ð5Þ synchronous belt wheel (1 : 1 transmission ratio), a rotary 2 2 2 + b x + b x + b x , 7 8 9 1 2 3 shaft, and two blade holders and transmitted the power to the blades (Figure 6(b)). In this study, the selected blade (IT245) had a working radius (R) of 245 mm and a width where T is the torque (N-m); x , x , and x are the codes; 1 2 3 (b) of 60 mm, which were typical with rotary tillers. Accord- and b is the coeﬃcient of the codes. ing to the mathematical model as shown in Equation (3), the Here, the purpose of this test was to obtain a set of opti- whole width (w ) of the bionic structure was set to be mized bionic rotary tiller blade (BB-1) parameters for 60 mm. The soil bin was divided into three sections for reduced torque requirement and analyze the inﬂuence of avoiding interactions between consecutive test runs each factor in the regression model on the response value. w0 6 Applied Bionics and Biomechanics O O Scoop surface Blade holder Cutting edge Lengthwise surface b = 60 mm b = 60 mm Transition surface (a) (b) Figure 5: Conﬁgurations of rotary tillage blades: (a) the conventional; (b) the biomimetic. 8 7 (a) (b) Figure 6: Soil bin tests: (a) soil bin and test equipment and (b) test toolbar: 1: soil bin trolley; 2: torque sensor; 3: universal joint coupling; 4: gear box; 5: synchronous belt wheel; 6: rotary tillage blade; 7: rotary shaft; 8: blade holder. 5 m 10 m 5 m Transition part Section 1 Stable part Transition part Section 2 Stable part Transition part Transition part Section 3 Transition part Stable part Transition part 20 m 40 m Figure 7: Scheme of three sections in the soil bin. Further, a comparison of the soil-cutting performance of the controls in the soil bin, aiming to investigate how the torque optimized bionic blade (BB-1) with a conventional blade requirement was inﬂuenced by the bionic geometric charac- (CB) and a reported biomimetic rotary tillage blade  teristics. The controls were the conventional rotary tillage (BB-2) would be conducted in the soil bin. blade (CB) and the optimal biomimetic rotary tillage blade (BB-2) in Ref.  which had three arc concave teeth equally 4.3. Comparative Test. The optimal biomimetic blade (BB-1) arranged on the front cutting edge with a central angle of obtained from the regression test was compared with two 60 . Figure 9 shows the three types of the blades used in R = 245 mm R = 245 mm 0.8 m 0.8 m 0.8 m 3 m Applied Bionics and Biomechanics 7 Figure 8: Nine biomimetic rotary tillage blades designed for the soil bin tests. Table 3: Factor-level coding. Table 4: Test scheme and result of coded test. Level Number r θ ( ) n (rpm) x x x T (N-m) 1 2 3 Factor Code -1 0 1 1 1.25 60 160 -1 0 -1 23.84 rx 1.25 1.50 1.75 2 1.50 60 240 0 0 0 18.27 ° ° ° ° ° θ ( ) x 50 60 70 3 1.25 70 240 -1 1 0 17.55 x 4 1.25 50 240 -1 -1 0 17.26 n (rpm) 160 240 320 5 1.50 70 320 0 1 1 22.31 6 1.25 60 320 -1 0 1 20.56 7 1.50 60 240 0 0 0 18.33 the comparative test. During the comparative test, the rotary speed changed from 160 to 320 rpm at an interval of 40 rpm, 8 1.75 50 240 1 -1 0 19.05 and each test was repeated three times. 9 1.75 60 160 1 0 -1 25.81 10 1.50 50 320 0 -1 1 22.06 11 1.50 60 240 0 0 0 18.21 5. Field Experiments 12 1.75 70 240 1 1 0 19.33 5.1. Experimental Site and Equipment. The experiments were 13 1.50 60 240 0 0 0 18.44 conducted in a tested ﬁeld of Jilin Agricultural University, 14 1.50 50 160 0 -1 -1 24.85 China. The soil was a loamy soil with a bulk density of -3 15 1.50 70 160 0 1 -1 25.51 1040 kg m measured by the oven-drying method, a water 16 1.50 60 240 0 0 0 18.31 content of 20.82% measured using a TDR-300 type Soil Moisture Meter (RGB Spectrum Equipment, USA) with 17 1.75 60 320 1 0 1 22.58 12 cm probes, and a soil compaction of 0.74 MPa measured using a SC-900 type Soil Compaction Meter (RGB Spectrum Equipment, USA) with a 1/2 00 diameter cone tip. As the BB-1 and CB (IT245) operated at diﬀerent rotary speeds, shown in Figure 10(a), a tractor (70.8 kw, KUBOTA-M954 forward velocities, and tillage depths. According to the Gov- made by Kubota Agricultural Machinery (Suzhou) Co., ernment Standard of Rotary Tiller in China, the rotary Ltd.) with a rotary tiller (1GQN-230 made by Tianjin Trac- speed, forward velocity, and tillage depth were, respectively, -1 tor Manufacture Co., Ltd.) moved forward to cut soil, and a set in the ranges of 150 to 350 rpm, 1 to 5 km h , and 80 to sensor testing system recorded the data during the cutting 200 mm. A four-factor random complete block design with process. The blades were installed in the rotary tiller to con- three replications was used, as follows: nect the blades with the tractor power take-oﬀ shaft. The torque sensor (CYB-803S, 0-±1000 N m) was placed on the (i) Factor A (blade geometry)—2 treatments: CB and tractor power take-oﬀ shaft (Figure 10(b)) so that the torque BB-1 of the blades could be transmitted to sensors. The detail of (ii) Factor B (rotary speed)—2 treatments: 254 and the experimental methods and processes could be found in 267 rpm (according to the tractor power output our former work . shaft speed) 5.2. Experimental Treatment and Data Processing. The (iii) Factor C (forward velocity)—5 treatments: 1, 2, 3, 4, -1 experiments compared the torque requirements between and 5kmh 8 Applied Bionics and Biomechanics (a) (b) (c) Figure 9: Three types of the blades used in the comparative test: (a) CB; (b) BB-2; (c) BB-1. (iv) Factor D (tillage depth)—3 treatments: 80, 120, and cessing, the average value of torque from three replications 160 mm was used in the analysis. Treatment Factor A was randomly followed by a random 6. Results allocation of two levels of Factor B at a forward velocity of -1 3kmh and a tillage depth of 120 mm, randomly followed 6.1. Results of Soil Bin Tests by a random allocation of ﬁve levels of Factor C at a rotary speed of 254 rpm and a tillage depth of 120 mm, randomly 6.1.1. Result Analysis of Regression Test. The orthogonal test followed by a random allocation of three levels of Factor D of soil cutting by the biomimetic blades was conducted in at a rotary speed of 254 rpm and a forward velocity of the soil bin according to the test scheme (Table 4). The tor- -1 3kmh . In order to eliminate the eﬀect of the acceleration que ﬂuctuated from 17.26 to 25.81 N-m as the factors both of the tractor, the data were recorded continuously after varied, which expressed the importance of the optimization the tractor had moved 10 m. During this period of data pro- of the tests. The data obtained in Table 4 was processed Applied Bionics and Biomechanics 9 Torque sensor Tractor power take off sha ft Tactor Sensor testing system Rotary tiller (a) (b) Figure 10: (a) Equipment used in the ﬁeld experiment; (b) sensor testing system. Table 5: ANOVA table for the regression model. Source of variation Quadratic sum Degree of freedom Mean sum of square FP Model 139.12 9 15.46 1090.96 <0.0001 7.14 1 7.14 504.23 <0.0001 0.27 1 0.27 19.32 0.0032 19.53 1 19.53 1378.49 <0.0001 x x 2:50E − 5 2:50E −51:76E − 3 1 0.9677 1 2 x x 6:25E − 4 6:25E − 4 1 0.044 0.8396 1 3 x x 0.042 1 0.042 2.97 0.1287 2 3 x 0.26 1 0.26 18.54 0.0035 1 1 x x 0.23 1 0.23 16.45 0.0048 2 2 x x 111.03 1 111.03 7836.71 <0.0001 3 3 Residual 0.099 7 0.014 Lack of ﬁt 0.07 3 0.023 3.25 0.1426 7:22E − 3 Pure error 0.029 4 Cor. total 139.22 16 according to Equation (5) by the least square method, and other factors of x x , x x , and x x were the insigniﬁcant 1 2 1 3 2 3 the quadratic regression model was developed as follows: terms due to their smaller F values and larger P values. Therefore, the eﬀect of each factor on the torque would be analyzed as follows. T =18:31 + 0:95x +0:19x − 1:56x − 0:0025x x 1 2 3 1 2 ð6Þ 2 2 2 In order to obtain a single-factor model, the other two +0:012x x − 0:1x x − 0:25x +0:24x +5:14x : 1 3 2 3 1 2 3 factors were ﬁxed at 0 levels so that the submodels of the ratio of claw width to interval, the inclined angle, and the The ANOVA of the regression model is shown in rotary speed were Table 5 to analyze and judge the reliability of the regression model. Based on the F distribution table, the value of F 0:05 ð9, 7Þ was 3.68, which was much less than the F value of > T =18:31 + 0:95x − 0:25x , x 1 > 1 the above regression model. And the P value of this model was less than 0.05, indicating that this regression model ð7Þ T =18:31 + 0:19x +0:24x , x 2 2 was reliable and signiﬁcant. It was also highlighted by the : 2 T =18:31 + 1:56x +5:14x : smaller F value (3.25) and larger P value (0.1426) of the lack x 3 3 of ﬁt, showing that the lack of ﬁt was not signiﬁcant and conversely the ﬁtting results of the quadratic regression were outstanding. Moreover, the factors of x , x , x , x x , x x , The single-factor models of the above formula are 1 2 3 1 1 2 2 and x x were the signiﬁcant terms to inﬂuence the model derived from their respective single factors to the marginal 3 3 due to their larger F values and smaller P values, while the equation of the torques of each factor at diﬀerent levels: 10 Applied Bionics and Biomechanics –4 –8 –1.0 –0.5 0.0 0.5 0.1 –1.0 –0.5 0.0 0.5 0.1 x x x x 1 1 x x 2 2 x x 3 3 (a) (b) Figure 11: (a) Regression curves of the submodels and (b) the derivation curves of factors. 19.5 18.5 17.5 70 1.75 1 320 0 1.75 1 60 60 1.50 1.50 240 240 1.50 1.50 𝜃 (°) 𝜃 (°) rr rr n n (r (rp pm) m) 50 50 1.25 125 160 160 1.25 125 (a) (b) 320 70 240 60 n (rpm) 𝜃 (°) 160 50 (c) Figure 12: Response surface diagrams of torque: (a) n = 240; (b) θ =60 ; (c) r =1:50. T (N-m) Torque (N-m) Torque (N-m) T (N-m) Torque (N-m) x Applied Bionics and Biomechanics 11 > T x =0:95 − 0:5x , 1 1 ′ ð8Þ T x =0:19 + 0:48x , 2 2 > 25 T x =1:56 + 10:28x : 3 3 The regression curves of the submodels and their corre- sponding derivation curves of factors are as shown in Figure 11. When x increased from -1 to 1, the change of x had a positive eﬀect on T (Figure 11(b)); so, Tx would x 1 increase all the time with the increasing of x (Figure 11(a)). When x > −0:5, the change of x started to positively eﬀect on T ; inversely, when x < −0:5, the change of x had a negative eﬀect on T (Figure 11(b)); as a result, 2 x Tx would get the maximum or minimum near x = −0:5 160 200 240 280 320 (Figure 11(a)). When x >0, the change of x had a signiﬁ- Rotary speed (rpm) cant positive eﬀect on T ; oppositely, when x <0, the change of x had a signiﬁcant negative eﬀect on T (Figure 11(b)); CB 3 x thus, Tx would get the maximum or minimum near x =0 BB-2 BB-1 (Figure 11(a)). T was most inﬂuenced by x than x and x 3 1 x , while the impact of x was slightly higher than x . When 2 1 2 Figure 13: The torques of the three rotary tillage blades aﬀected by the factors were at diﬀerent levels, the torque would change rotary speed. in diﬀerent degrees. Therefore, it was necessary to optimize the geometries of the blade operating at a suitable condition Among the three blades, the BB-1 had the lowest torque for meeting a lower torque requirement. requirement in the soil tillage process. On average, the BB- 1 reduced the torque requirement by 13.99% than that of 6.1.2. Optimization. The torque aﬀected by the interactions the CB. And the torque requirement of the BB-2 was of the ratio of claw width to lateral spacing (r), the inclined 3.74% less than that of the CB which was paralleled to the angle (θ), and the rotary speed (n) was displayed by the result of the literature . It indicated that the geometric response surface methodology as shown in Figure 12. The characteristics of the ﬁve fore claws of mole rats have signif- response surface diagrams intuitively reﬂected the variation icant eﬀects on the torque requirement of rotary tillage trends of torque aﬀected by factors. The trend of torque blades. aﬀected by r and θ was opposite to the trends aﬀected by r and n and θ and n. The variation of torque aﬀected by r 6.2. Results of Field Experiments. A comparison of the torque and θ was small (Figure 12(a)); by contrast, with n, the value requirements between the BB-1 and CB conducted in ﬁeld variation of torque was relatively large (Figures 12(b) and experiments at diﬀerent tillage conditions is shown in 12(c)), indicating the factor n had the most signiﬁcant eﬀect Figure 14. It could be seen that the variation trends of torque on the torque than r and θ. In Figure 12(a), the torque chan- of the optimized bionic blade were similar to the trends of ged more intensely aﬀected by r and θ, which indicated that the conventional blade regardless of tillage conditions. the eﬀect of r was slightly higher than θ. These results based When the rotary speed increased from 254 to 267 rpm, the on the response surface diagrams of torque were basically torque signiﬁcantly reduced due to the reduction of the bite similar to the above results obtained from Figure 11. The length of blades which was observed in previous study . torque could get the minimum value within the range of r, -1 As the forward velocity increased from 1 to 2 km h , the tor- θ, and n from the comparison of Figure 12(a). Consequently, que was also enlarged rapidly due to the increase of the soil when taking ðr, θ, nÞ as (1.25, 60 , 240), the torque attained retillage as mentioned by the literature , whereas the soil the minimum value, which meant that the optimized param- retillage was reduced with further increases in forward veloc- eters were the ratio r =1:25, the inclined angle θ =60 , and -1 ity from 2 to 3 km h resulting in a decline of the torque; but the rotary speed n = 240 rpm. the torque increased once again when the forward velocity -1 6.1.3. Results of Comparative Test. Figure 13 shows the com- was from 3 to 5 km h due to the increase of the bite length parative test results of the three rotary tillage blades aﬀected of blades as suggested by Salokhe et al. . Therefore, the -1 by rotary speed. The torques of three blades showed a simi- forward velocity of 3 km h was a more suitable operating lar trend with the varied rotary speeds. The torque decreased condition for the rotary tiller blades to require a lower tor- as the rotary speed increased from 160 to 240 rpm because of que. Additionally, the tillage depth was also an outstanding the reduction of the bite length and a smaller volume of soil eﬀect on the torque requirement of blades. The torque per bit. And then, the torque increased when the rotary increased all the time with the tillage depth varied from 80 speed increased from 240 to 360 rpm due to the soil retillage to 150 mm due to the increase of soil volume in per bit. and an increasing in soil acceleration. Therefore, there was a Overall, the optimization of rotary tiller blade in this study minimal torque at the rotary speed of 240 rpm. This varia- made no impacts on the variation trends of torque with tion trend accorded with the results of the literature . tillage conditions. Torque (N-m) 12 Applied Bionics and Biomechanics 600 600 500 500 400 400 300 300 200 200 100 100 0 0 254 267 –1 Forward velocity (km·h ) Roatry speed (rpm) Conventional rotary tiller blade Conventional rotary tiller blade Optimized bionic rotary tiller blade Optimized bionic rotary tiller blade (a) (b) 80 120 160 Tillage depth (mm) Conventional rotary tiller blade Optimized bionic rotary tiller blade (c) Figure 14: Torques of rotary tiller blades at diﬀerent tillage conditions: (a) rotary speed; (b) forward velocity; (c) tillage depth. 𝜔 𝜔 O V V m B B J 2 2 I G H 2 2 G 1 H 2 J 1 1 (H ) F 1 (G ) E 2 (a) (b) (c) Figure 15: Soil wedge formed in front of the rotary tillage blade: (a) the conventional (CB); (b) the optimal biomimetic (BB-1); (c) the comparison between the CB and the BB-1. Note: ABEF—soil surface cut by the blade; CDEF—previously soil surface cut by the preceding blade; ABI1J1G1H1—soil wedge formed in front of the CB; ABI2J2G2H2—soil wedge formed in front of the BB-1. Torque (N-m) Torque (N-m) Torque (N-m) Applied Bionics and Biomechanics 13 3.5 3.5 3 3 2.5 2.5 2 2 1.5 1.5 1 1 0.5 0.5 0 0 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 1.2 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 X (mm) X (mm) (a) (b) 1.8 2.5 1.6 1.4 1.2 1.5 0.8 0.6 0.4 0.5 0.2 0 0 –1 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 X (mm) X (mm) (c) (d) 3.5 2.5 1.5 0.5 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 X (mm) (e) Figure 16: The curvature analysis of the ﬁtting curves of the ﬁve claw tips: (a) the 1st; (b) the 2nd; (c) the 3rd; (d) the 4th; (e) the 5th. On the other hand, the optimized bionic blade always far in the torque requirement of the whole blade so that it was surpassed the conventional blade in the torque require- necessary to optimize its geometries to achieve lower energy consumption. ments. It was found out that the torques of the optimized bionic blade were averagely 17.00%, 16.88%, and 21.80% lower than those of conventional blade at diﬀerent rotary 7. Discussion speeds, forward velocities, and tillage depths, respectively. It could be concluded that the soil-cutting performance of In general, when the rotary tillage blade cuts soil, the length- the optimized bionic blade was better than that of conven- wise surface, the transition surface, and the scoop surface tional blade due to the fact of reduced torque requirement. successively touched the soil. Thus, the sliding cutting per- In summary, these results further conﬁrmed that the formance of the cutting edge played an important role in scoop surface of rotary tillage blade played an important role the soil cutting . Upon the soil entrance of blade, the soil –1 –1 Curvature (mm ) Curvature (mm ) –1 Curvature (mm ) –1 –1 Curvature (mm ) Curvature (mm ) 14 Applied Bionics and Biomechanics failure wedge in the soil-cutting process based on the GC-2 was gradually broken into small clods in conformity to the Mohr-Coulomb failure criterion [30, 31]. In each circle of of the multiclaw combination, thus largely reducing the tor- this broken process, a soil wedge  was formed in front que of BB-1. The smaller torque requirement made BB-1 more eﬃcient. Therefore, the GC of ﬁve fore claws of mole of the blade as shown in Figure 15(a), and the required shear force followed the Coulomb formula. rats is more applicable to the geometric optimization of rotary tillage blades. 7.1. Eﬀect of the Contour Curve Characteristics of the Five Claw Tips (GC-1) on the Torque of BB-1 in the Soil-Cutting 8. Conclusion Process. The reported BB-2 reduced the torque requirement Biomimetic rotary tillage blades were designed to reduce the by 3.91% because the three concave arcs could improve the torque requirement according to the geometric characteris- penetration performance . Similarly, BB-1 also had a tics (GC) of the ﬁve fore claws of mole rats, including the better penetration performance than CB, since the cutting contour curves of the ﬁve claw tips (GC-1) and the structural edge could reduce the contact area with soil when the blade characteristic of the multiclaw combination (GC-2). Results touched the soil. Moreover, the contour curves of the ﬁve of soil bin tests and ﬁeld experiments showed the following: claw tips had an excellent sliding cutting performance. The curvatures of the ﬁve claw tips were calculated as shown in (1) The order of inﬂuence on the torque was as follows: Figure 16. The curvature of the 1st claw tip showed one peak -1 the rotary speed (n), the ratio of claw width to inter- of 3.5 mm at x = −0:15 mm, then declined rapidly to val (r), and the inclined angle (θ), all of which have x =0:2mm, and continued to decrease at a lower rate from obvious impacts by further analyzing regression x =0:2 to 1.2 mm, indicating that the 1st claw tip suﬀered curves of the submodels and their corresponding so severe soil wear that it had a better sliding cutting perfor- derivation curves of factors. Moreover, for the pur- mance. The curvatures of the other four claw tips each had pose of the reduction in torque requirement of the two peaks and were small between the two peaks for each rotary tiller blade, the optimal working combination tip, indicating that the middle part between the two peaks was set by the response surface methodology: for each tip mainly cut soil, especially for the 2nd claw tip. r =1:25, θ =60 , and n = 240 rpm For the 3rd claw tip, the narrow middle part between the two peaks contributed to a better penetration performance. (2) The new optimal biomimetic blade (BB-1) in this Interestingly, the peaks of the curvatures decreased from study and the reported optimal biomimetic (BB-2) the 1st to the 3rd claw and then increased to the 5th claw. of the literature  averagely reduced the torque The smaller peak of the curvature in the 3rd claw tip could requirement by 13.99% and 3.74%, respectively, avoid quick wear by soil when it penetrated into soil. On compared with the conventional blade (CB) the whole, the contour curves of the ﬁve claw tips had a bet- (3) Field comparison experiments also showed that the ter penetration and sliding cutting performance. As a result, torques of BB-1 were averagely 17.00%, 16.88%, the sliding cutting action was performed at a small dip angle and 21.80% lower than those of CB at diﬀerent to reduce friction force and insure emergence from the soil, rotary speeds, forward velocities, and tillage depths, which was similar to the slide cutting performance of the respectively fore claw of Cryptotympana atrata nymph . Also, the tips of the ﬁve claws could penetrate in to soil with small force Overall, the geometric structure of the ﬁve claws of mole requirement of BB-1. rats may play an important role in reducing the torque and energy requirement of rotary tillage blades. This research 7.2. Eﬀect of the Structure Characteristics of the Multiclaw also provides a new method to design other soil-engaging Combination (GC-2) on the Torque of BB-1 in the Soil- tools for achieving a minimum soil resistance and energy Cutting Process. The rotary tillage blade started to break consumption. the soil after the whole body of the blade entered the soil, and then, soil wedge was formed in front of the blade as Data Availability shown in Figure 15. Based on the previous study , the soil rupture distance ratio of the ﬁve-claw combination was The data used to support the ﬁndings of this study are about 19.6% smaller than the predicted values of simple available from the ﬁrst author upon request. blades. It indicated that the soil failure wedge of the BB-1 was signiﬁcantly diminished so that a lower force was Conflicts of Interest needed for soil shearing. Figure 15(c) illustrated the diﬀer- ence in the soil failure wedges formed by CB and BB-1 sep- The authors declare that there are no conﬂicts of interest arately. By the comparison of the soil failure wedges, it regarding the publication of this paper. conﬁrmed that the ﬁve claw structure could change the soil failure pattern of the rotary tiller blade to form a smaller soil Acknowledgments failure wedge for minimizing the torque requirement largely. In sum, BB-1 ﬁrstly enhanced the penetrating and slid- This work was supported by the National Natural Science ing cutting performance of the cutting edge based on the Foundation of China (Grant No. 51475204), the Fundamen- GC-1 of the ﬁve claw tips and secondly diminished the soil tal Research Funds for the Central Universities of China Applied Bionics and Biomechanics 15 (Grant No. 2452019205), and the China Postdoctoral  M. Li, D. H. Chen, S. J. Zhang, and J. Tong, “Biomimeitc design of a stubble-cutting disc using ﬁnite element analysis,” Science Foundation (Grant No. 2020M683577). Journal of Bionic Engineering, vol. 10, no. 1, pp. 118–127,  R. G. Scott and R. C. Richardson, “Realities of biologically References inspired design with a subterranean digging robot example,” in Proceedings of the 6th IASTED international conference on  S. Chertkiattipol and T. Niyamapa, “Variations of torque and robotics and applications, pp. 226–231, Cambridge, MA, speciﬁc tilling energy for diﬀerent rotary blades,” International USA, 2005. Agricultural Engineering Journal, vol. 19, no. 3, 2010.  A. C. Reed, “Locomotion and appendicular anatomy in three  K. H. M. Siddique, C. Johansen, N. C. Turner et al., “Innova- soricoid insectivores,” American Midland Naturalist, vol. 45, tions in agronomy for food legumes: a review,” Agronomy for no. 3, pp. 513–671, 1951. Sustainable Development, vol. 32, no. 1, pp. 45–64, 2012.  B. Li, Y. Chen, and J. Chen, “Modeling of soil-claw interaction  M. Veerangouda, E. R. Sushilendra, and M. Anantachar, using the discrete element method (DEM),” Soil and Tillage “Development and evaluation of multipurpose tool carrier Research, vol. 158, pp. 177–185, 2016. for power tiller,” Karnataka Journal of Agricultural Sciences, vol. 24, no. 5, pp. 704-705, 2011.  W. F. Ji, D. H. Chen, H. L. Jia, and J. Tong, “Experimental investigation into soil-cutting performance of the claws of  M. Mairghany, A. Yahya, N. M. Adam, A. S. Mat Su, mole rat (Scaptochirus moschatus),” Journal of Bionic Engi- W. Aimrun, and S. Elsoragaby, “Rotary tillage eﬀects on some neering, vol. 7, no. S4, pp. S166–S171, 2010. selected physical properties of ﬁne textured soil in wetland rice cultivation in Malaysia,” Soil and Tillage Research, vol. 194,  Y. W. Yang, M. Li, J. Tong, and Y. H. Ma, “Study on the Interaction between Soil and the Five-Claw Combination of a Mole Using the Discrete Element Method,” Applied Bionics  C. P. Grupta and R. Visvanathan, “Power requirement of a and Biomechanics, vol. 2018, Article ID 7854052, 11 pages, rotary tiller in saturated soil,” Transactions of the ASAE, vol. 36, no. 4, pp. 1009–1012, 1993.  R. J. Godwin, “A review of the eﬀect of implement geometry on  G. S. Zhang, Z. Q. Zhang, M. H. Xiao, P. Bartos, and A. Bohata, soil failure and implement forces,” Soil and Tillage Research, “Soil-cutting simulation and parameter optimization of rotary vol. 97, no. 2, pp. 331–340, 2007. blade's three- axis resistances by response surface method,” Computers and Electronics in Agriculture, vol. 164, p. 104902,  R. J. Godwin, G. Spoor, and M. S. Soomro, “The eﬀect of time arrangement on soil forces and disturbance,” Journal of Agri- cultural Engineering Research, vol. 30, pp. 47–56, 1984.  J. M. Beeny and D. C. P. Khoo, “Preliminary investigations into the performance of diﬀerent shaped blades for the rotary  V. S. Saimbhi, D. S. Wadhwa, and P. S. Grewal, “Development tillage of wet rice soil,” Journal of Agricultural Engineering of a rotary tiller blade using three-dimensional computer Research, vol. 15, no. 1, pp. 27–33, 1970. graphics,” Biosystems Engineering, vol. 89, no. 1, pp. 47–58,  G. U. Shinde and S. R. Kajale, “Design optimization in rotary tillage tool system components by computer aided Engineerin-  J. Sakai, H. L. Van, K. Iwasaki, and Y. Shibata, “Tillage resis- gAnalysis,” International Journal of Environmental Science, tance characteristics of Japanese rotary blades distribution of vol. 3, pp. 279–282, 2012. tilling resistance to the tipless (straight) blade and bending tip portions of a rotary blade,” Journal of the Japanese Soci-  M. A. Matin, J. M. Fielke, and J. M. A. Desbiolles, “Torque and ety of Agricultural Machinery, vol. 46, no. 1, pp. 593–598, energy characteristics for strip-tillage cultivation when cutting furrows using three designs of rotary blade,” Biosystems Engi- neering, vol. 129, pp. 329–340, 2015.  G. E. P. Box and D. W. Behnken, “Some new three level designs for the study of quantitative variables,” Technometrics,  J. Tong, W. F. Ji, H. L. Jia, D. H. Chen, and X. W. Yang, vol. 2, no. 4, pp. 455–475, 1960. “Design and tests of biomimetic blades for soil-rototilling and stubble-breaking,” Journal of Bionic Engineering, vol. 12,  J. G. Hendrick and W. R. Gill, “Rotary tiller design parameters, no. 3, pp. 495–503, 2015. V: kinematics,” Transactions of the ASAE, vol. 21, no. 4, pp. 0658–0660, 1978.  J. Y. Sun, Y. M. Wang, Y. H. Ma, J. Tong, and Z. J. Zhang, “DEM simulation of bionic subsoilers (tillage depth >40 cm)  Y. W. Yang, J. Tong, Y. H. Ma, X. H. Jiang, and J. G. Li, “Design with drag reduction and lower soil disturbance characteris- and experiment of biomimetic rotary tillage blade based on tics,” Advances in Engineering Software, vol. 119, pp. 30–37, multiple claws characteristics of mole rats,” Transactions of 2018. the Chinese Society for Agricultural Engineering, vol. 35, no. 19, pp. 37–45, 2019.  J. F. Sun, H. M. Chen, Z. M. Wang et al., “Study on plowing performance of EDEM low-resistance animal bionic device  V. M. Salokhe, M. Hanif Miah, and M. Hoki, “Eﬀect of blade based on red soil,” Soil and Tillage Research, vol. 196, type on power requirement and puddling quality of a rotavator p. 104336, 2020. in wet clay soil,” Journal of Terramechanics, vol. 30, no. 5, pp. 337–350,  R. Zhang, B. Chen, J. Q. Li, and S. C. Xu, “DEM simulation of clod crushing by bionic bulldozing plate,” Journal of Bionic  W. M. Ding and S. Z. Peng, “Research on design of sidelong Engineering, vol. 5, pp. 72–S78, 2008. edge of rotary blade,” Transactions of the Chinese Society for Agricultural Machinery, vol. 26, no. 4, pp. 56–61, 1995.  L. Q. Ren, Z. W. Han, J. Q. Li, and J. Tong, “Experimental investigation of bionic rough curved soil cutting blade surface  C. E. Johnson, R. D. Grisso, T. A. Nichols, and A. C. Bailey, to reduce soil adhesion and friction,” Soil and Tillage Research, “Shear measurement for agricultural soils: a review,” Transac- vol. 85, no. 1-2, pp. 1–12, 2006. tions of the ASAE, vol. 30, no. 4, pp. 0935–0938, 1987. 16 Applied Bionics and Biomechanics  W. R. Gill and G. E. Vanden Berg, “Soil Dynamics in Tillage and Traction,” in Agricultural Handbook No. 316, pp. 23–45, U.S. Government Printing Oﬃce, Washington, DC, USA,  E. McKyes, Soil Cutting and Tillage, Elsevier science publisher, Amsterdam, 1985.  Z. Y. Chang, W. Liu, J. Tong et al., “Design and experiments of biomimetic stubble cutter,” Journal of Bionic Engineering, vol. 13, no. 2, pp. 335–343, 2016.
Applied Bionics and Biomechanics – Hindawi Publishing Corporation
Published: Sep 28, 2021