The winter time in northern China is long, where snow and ice often occur on roads, especially on expressways and living areas, seriously affecting the normal production and daily activities of people.

Snow melting agent as an efficient solution has become a key component of deicing and snow removal, which is mainly divided into manual throwing and institutional throwing ^[1-6]. Manual throwing is labor intensive and harmful to clothing and skin because of the chemical action of the snow melting agent ^[7-12]. Moreover, the poor dynamic performance of the existing throwing mechanisms results in unstable performance and uneven throwing, due to the lack of appropriate research methods and means for mechanism motion characteristics. Thus the corresponding motion parameters cannot be adjusted timely according to the road conditions to improve the throwing efficiency ^[13-17]. Therefore, it is necessary to develop an effective throwing mechanism and a research method of motion performance to improve the throwing efficiency and dynamic performance while reducing the amount of throwing and the environmental pollution.

Nowadays, snow melting agents are evenly distributed on pavement at home and abroad. However, practice has proved that there is no relationship between uniform distribution and snow melting efficiency, and the distribution density of snow melting agent is the main factor that affects snow melting efficiency. The non-uniform distribution of snow melting agent by pulse method is proved to be better than the uniform distribution method through experiments ^[18]. Variable crank length cam-linkage rod combination mechanism can accurately meet any given trajectory requirements, reduce the overall size of the mechanism, and improve the dynamic performance of the machine ^[19-21].

To address the above problems, a snow melting agent throwing mechanism with variable crank length is designed in this paper. By analyzing its motion characteristics and combining the Visual Basic (VB) programming, a parametric model of key parts of the mechanism with human-computer interaction function based on knowledge engineering and computer aided three-dimensional interactive application (CATIA) secondary development technology is established, which improves the throwing efficiency and the realization of snow melting agent, as well as provides an effective mechanism and method for rapid modeling of complex mechanisms.

The variable crank length snow melting agent throwing mechanism is mainly composed of a transmission part, an execution part, and a trajectory control part. The transmission part consists of a large bevel gear and a small bevel gear. The executive part consists of a rotary shaft, a crank, a rocker, a slider, and a throwing frame. The trajectory control part consists of a cam plate. The dispersal device of the snow melting agent and the overall structure throwing mechanism are shown in Figs. 1-2. The working principle of the mechanism is that the big bevel gear transfers power to the small bevel gear, which then drives the rotary shaft fixed to the small bevel gear, and lastly the rotary shaft drives the T-type crank to rotate. Because of the shape locking, the roller at one end of the T-type crank transverse rod moves in the cam groove of the cam plate, and the slider at the other end drives the rocker to swing back and forth to complete the snow melting process.

Fig.1

Fig.1 Dispersal device of snow melting agent

Fig.2

Fig.2 Overall structure diagram of throwing mechanism

As shown in Fig. 3, the kinematics equation of the mechanism was established based on analytic method. The length l_OB of the crank OB varies with the change of the cam diameter, and the angle between it and the x-axis is θ₁. The length l_AC of the rocker AC is fixed, and the angle between it and the x-axis is θ₂. Assuming that the cam is elliptic-like structure, the long half-axis of the ellipse is On and the short half-axis is Om. When the ratio of the length between the long and the short half-axes changes, the profile of the cam will change, and the distance from slider to swing center l_AB will change accordingly. Taking θ₁ as a variable, the top of the rocker is represented by the point C, and its coordinate equation is established as follows:

Fig.3

Fig.3 Motion analysis diagram of throwing mechanism

where

where θ₁=ωt, ω is the angular velocity of the crank, and t represents the time.

In order to verify the correctness of Eq. (1), the lengths of each component in the mechanism were defined as follows: l_Om equals 40 mm, l_On equals 60 mm, l_OA equals 80 mm, l_AC equals 155 mm, ω equals 60 rad/s, and the forward speed of mechanism v₀ equals 18 m/s. By introducing the above values of structural parameters and motion parameters into Eq. (1), the coordinates of point C which varies with time can be obtained, from which the trajectory of the snow melting agent can be obtained. The model of the variable crank length snow melting agent throwing mechanism was established by using CATIA software. The kinematics simulation of the throwing mechanism was carried out by Digital Mockup (DMU) module. The comparison between the theoretical trajectory derived by Eq. (1) and the simulation trajectory obtained by CATIA software is presented in Fig. 4.

Fig.4

Fig.4 Trajectory contrast diagram of throwing mechanism

From Fig. 4, it can be seen that the simulation trajectory and the theoretical trajectory basically coincided with each other, and the error was very small. Moreover, it was found that the theoretical deduction results in this paper are accurate and reliable.

In order to study the influence of different parameters on the trajectory of throwing, the parametric analysis method was used to investigate the effects of cam structural parameters, rotary shaft angular velocity, and the overall forward velocity of the throwing mechanism on the trajectory and the acceleration of throwing. The results are shown in Figs. 5-7.

Fig.5

Fig.5 Diagrams of effects of structural parameters

Fig.6

Fig.6 Diagrams of effects of angular velocity

Fig.7

Fig.7 Diagrams of effects of forward velocity

It can be observed in Fig. 5 that when a/b (i.e., the ratio of the lengths of the cam's long and short half-axes) increases gradually, the throwing trajectory of the snow melting agent presents pulsation characteristics, the peak value becomes larger, the wave peak becomes sharper, the spreading width increases, the spacing changes insignificantly, and the acceleration curve of the snow melting agent throwing gradually becomes gentle. When the ratio exceeds 2.5, the acceleration curve exhibits sharp points, which indicates that the motion becomes unstable.

According to Fig. 6, when the angular velocity of the rotary shaft increases, the spreading width remains unchanged, the spraying interval increases, and the amplitude of the throwing acceleration increases gradually, while the increase is not large and the period of speed and acceleration decrease.

From Fig. 7, it can be found that under the same conditions, with the change of the overall forward speed of the mechanism, the peak value of the throwing trajectory changes obviously. With the increase of the forward velocity, the spreading width remains unchanged, the spacing becomes larger, and the throwing acceleration does not change significantly.

In order to facilitate designers to modify the structural parameters of the throwing mechanism according to the road conditions and adjust the matching relationship between dynamic parameters in real-time, a human-computer interaction simulation system for the variable crank length throwing mechanism was established by using VB6.0 software as shown in Fig. 8, in which the left side is the input box of structural parameters and dynamic parameters, and the middle is the trajectory display area.

Fig.8

Fig.8 Human-computer interaction simulation system for throwing mechanism

According to the provisional provisions of the national urban planning quota indicators, the maximum width of the main road is 60 m, and the number of two-way motor lanes is 4. If the width of the middle isolation zone is removed, 2 L₂, which is the spreading width of the snow melting agent, should reach 3.5 m. Considering the abilities of different snow melting agents, the spreading interval of L₁ should not be greater than 1.5 m. Therefore, according to the requirements of snow melting agent distribution, three-level and four-factor orthogonal experiments were adopted to design the spraying width and spacing of the snow melting agent, respectively. The results are presented in Tables 1-2.

表 1

Table 1 Orthogonal test of throwing width Numbers A a (mm) B b (mm) C ω (rad/s) D v (m/s) Width yi (mm) (yi-1750)/100 1 500 600 40 20 1452 -2.98 2 500 1200 60 40 1325 -4.25 3 500 1800 80 60 959 -7.91 4 1000 600 60 60 1497 -2.53 5 1000 1200 80 20 1664 -0.86 6 1000 1800 40 40 1052 -6.98 7 2000 600 80 40 1768 0.18 8 2000 1200 40 60 1730 -0.20 9 2000 1800 60 20 1576 -1.74 yj1 -15.14 -5.33 -10.16 -5.58 $\sum\limits_{i = 1}^9 {({y_i} - 1750)/100 = - 27.27} $ yj2 -10.37 -5.31 -8.52 -11.05 yj3 -1.76 -16.63 -8.59 -10.64 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j1}}$ -5.05 -1.78 -3.39 -1.86 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j2}}$ -3.46 -1.77 -2.84 -3.68 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j3}}$ -0.59 -5.54 -2.86 -3.55 Rj 4.46 3.77 0.55 1.82 Optimal Combination A3 B2 C2 D1

Table 1 Orthogonal test of throwing width

表 2

Table 2 Orthogonal test of throwing spacing Numbers A a (mm) B b (mm) C ω (rad/s) D v (m/s) Spacing zi (mm) (zi-1500)/100 1 500 600 40 20 1657 1.57 2 500 1200 60 40 1651 1.51 3 500 1800 80 60 1816 3.16 4 1000 600 60 60 1708 2.08 5 1000 1200 80 20 1687 1.87 6 1000 1800 40 40 1688 1.88 7 2000 600 80 40 1768 2.68 8 2000 1200 40 60 1673 1.73 9 2000 1800 60 20 1622 1.22 zj1 6.24 6.33 5.18 4.66 $\sum\limits_{i = 1}^9 {({z_i} - 1500)/100 = 17.70} $ zj2 5.83 5.11 4.81 6.07 zj3 5.63 6.26 7.71 6.97 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j1}}$ 2.08 2.11 1.73 1.55 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j2}}$ 1.94 1.70 1.60 2.02 ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j3}}$ 1.88 2.09 2.57 2.32 Rj 0.2 0.31 1.18 1.14 Optimal Combination A3 B2 C2 D1

Table 2 Orthogonal test of throwing spacing

In the tables, A, B, C, and D represent the factors of orthogonal test, respectively. The subscript in A₃, B₂, C₂, and D₁ indicate the level of the specific factor. v represent the travel speed of the throwing mechanism. y_j and z_j represent the width and spacing that can be achieved under each combination. The subscripts in y_j1, y_j2, and y_j3 indicate the sum of each lever for width. The subscripts in z_j1, z_j2, and z_j3 indicate the sum of each lever for spacing. ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j1}}$, ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j2}}$, and ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over y} _{j3}}$ represent the mean value of each lever for width. ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j1}}$, ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j2}}$, and ${\mathord{\buildrel{\lower3pt\hbox{$\scriptscriptstyle\rightharpoonup$}} \over z} _{j3}}$ represent the mean value of each lever for spacing. R_j represents the range of each lever. The equation (y_i-1750)/100 and (z_i-1500)/100 represent the standardization result for width and spacing, respectively. The equations in the lower right corner of Tables 1-2 represent the sum of the standardized results, respectively.

The rules and checking attributes to meet the requirements of specific throwing trajectory were determined based on the principle of knowledge engineering. When the parameters fail to meet the requirements, the interactive system prompts the parameters to make mistakes, which provides guidance for the subsequent improvement design.

According to the orthogonal test design of throwing width and spacing, the optimum parameters for winter snow melting agent throwing in Harbin are as follows: the length of cam long half-axis is 2000 mm, the length of cam short half-axis is 1200 mm, the forward speed of the whole machine is 20 m/s, and the angular speed of the rotating axis is 60 rad/s.

In this paper, a snow melting agent throwing mechanism is designed, and its trajectory, velocity, and acceleration are deduced by theoretical analysis. The correctness of the theoretical results was verified by CATIA motion simulation. The effects of different structural parameters and motion parameters on the trajectory and motion characteristics of the throwing mechanism were compared using parametric method. The human-computer interaction simulative system of the throwing mechanism was established by VB programming. Based on knowledge engineering and combined with the parameters of main roads in Harbin, a three-level and four-factor orthogonal test table was constructed. The human-computer interaction simulation system was used to obtain the optimum structure and working parameter combination of the throwing mechanism suitable for Harbin, which lays a theoretical foundation for the use of variable crank length snow melting agent throwing mechanism. The specific conclusions are drawn as follows:

1) With the increase of the length ratio of the long half shaft and the short half shaft of the cam, the spreading width increased, and the spreading interval did not change significantly. When the ratio exceeded 2.5, the acceleration curve presented a sharp point, and the spreading motion became uneven and stable;

2) When the rotary angular velocity increased, the spreading width remained the same, the spreading interval increased, and the amplitude of the spreading acceleration increased;

3) With the increase of the overall forward speed of the throwing mechanism, the spreading width remained the same, the spreading interval became larger, and the spreading acceleration did not change significantly, which means the movement was stable.