Mechanical analysis and distance optimization of t

Mechanical analysis and distance optimization of a robot support arm

Abstract: ADAMS virtual prototype simulation technology is used to model the main components of a robot, and the force conditions of the support arms before and after the model are analyzed; Seven models of different distances between the front and rear arms are established respectively. Through data comparison, the distance between the two arms that minimizes the power output of the motor is optimized

key words: robot; Support arm; Adams modeling; With the development of industry, robots are gradually replacing people in all walks of life to engage in various dangerous operations. Colleges and universities and relevant scientific research institutions are actively developing robots with various functions []. Cheng Jun [3], Du Zhijiang [4], and Liu Yanwu [5] respectively analyzed and simulated heterogeneous legged robots, biped robots, and robot carrying capacity. Harbin Institute of technology and Beijing University of Aeronautics and Astronautics have made many achievements in the research of robots, and have made high-tech robots such as football robots and Bionic Hexapod robots

the robot studied by the author is a multi-functional experimental robot, which can complete many operations, such as crossing ditches, climbing low walls, dredging pipelines, removing obstacles and so on. The problem encountered in the original design is that in the actual obstacle crossing process, due to the influence of design and other objective factors, the support arm cannot successfully support the body. Therefore, in the improved design, on the premise of neither changing the length of the support arm nor using a higher power motor, the dynamic model of the anti-terrorism robot is established in the dynamics software ADAMS to seek to optimize the distance between the front and rear arms of the robot, so as to reduce the motor torque by changing the length of the force arm, so that it can successfully lift the robot and realize the obstacle climbing function

1 data flow of numerical calculation of dynamics

the dynamic analysis of multi-body system is to give the variational equation of motion of free objects according to Newton's theorem, and then use Lagrange multiplier theorem to derive the dynamic equation of multi-body system based on constraints. Adams, a dynamic software integrating constraint equations, can automatically establish the dynamic differential algebraic equations of the system, such as formula (1). For dynamic differential algebraic equations, Adams selects different integration algorithms according to the characteristics of mechanical system []:

this kind of mathematical model is differential algebraic equations, also known as Euler Lagrange equations. It has a large number of equations, but the coefficient matrix is sparse, which is suitable for computer to automatically establish a unified model and process it

for the rigid system, ADAMS software adopts the backward differential formula (BDF) rigid integration program with variable number in this year's K exhibition. It is an automatic prediction and correction method with variable order and step size, and adopts the modified Newton Raphson iterative algorithm at each step of integration in the form of index3, SI2 and SI1 integration respectively

for the high-frequency system Adams, the coordinate separation method is used to reduce the differential algebraic equation to a pure differential equation expressed in independent generalized coordinates, and then ABAM (Adams bash forth Adams Moulton) method or Runge Kutta (rkf45) method is used to solve it respectively

2 working principle and mechanical modeling of robot

in the process of climbing over the wall, the anti-terrorism robot mainly uses four support arms to support the whole body. First, the motor installed on the first two support arms drives the two forearms to cross the fence, and then turns on the drive motors on the second two support arms to lift the whole body of this kind of light composite wire launched by Celanese and southwire at the beginning of this year, and then crawls on the fence through the crawler on the body, so that the whole body passes through the fence. The robot physical model is shown in Figure 1

Figure 1 physical model of anti-terrorism robot

the actual anti-terrorism model must be simplified before modeling with ADAMS. Since Adams only considers the centroid and mass of the part, and does not consider the external shape of the part, it is of little practical significance to accurately describe the complex shape of the part in the model. The key is that the connection between the four support arms of the robot and the body should be as practical as possible. The simplified model of the robot is shown in Figure 2, and the mass and size data of each part are shown in Table 1

Figure 2 the virtual prototype of the anti-terrorism robot

simulates and measures the force and torque at the hinge of the front and rear support arms of the robot. The results are shown in Figure 3 to figure 6

it can be seen from Figure 3 that although the trolley support arm did not contact the ground 0~2.25 s ago, due to the certain mass of the car body and support arm relative to the motor, it is easy to cause the vibration of the whole model when the motor rotates, which is reflected in the fluctuation of the curve between 0~2.25 s. At about 2.25 s, the front support arm contacts the ground, causing a large sudden change in the force on the hinge, which is shown as a straight line rising vertically in Figure 3. After 2.25 s, the force on the front support arm decreases slowly with the increase of time. This is because when the support arm contacts the ground and the trolley is fully lifted, the friction between the wheel and the ground decreases, making the force required to support the trolley relatively small

it can be seen from Figure 4 that the force on the rear support arm at the hinge is very similar to that on the front support arm, but the change trend of the force is opposite to that on the front support arm

it can be seen from Figure 5 that at about 2.25 s, the front support arm contacts the ground, causing a large sudden change in the torque at the hinge, which is shown as a vertical rising straight line in the figure. After 2.25 s, the torque at the front support arm decreases regularly with the increase of time. This is because the main factor in measuring torque is the z-axis. When the trolley just touches the ground, the distance between the force action point of the support arm and the measured hinge is the largest. According to t = fs (f represents the force on the hinge and s represents the distance from the Z axis), it can be seen from Figure 3 that although f changes, it is not very large, while the change of S is relatively obvious, so the curve presented is similar to a quadratic function curve

it can be seen from Figure 6 that the performance of the waste foam granulator measured at the hinge between the rear support arm and the car body is higher than the domestic GDP growth rate in the same period. The torque change is similar to that of the front support arm. Because the center of gravity of the car body is backward, the torque value is significantly greater than the front support arm, which corresponds to the situation that the force value of the rear support arm is greater than the front support arm reflected in Figure 3 and Figure 4

3 position optimization of support arm

analyze the situation that the starting distance between the two arms is 235mm and the forearm position is not moving, and compare the force and torque of the front and rear support arms respectively. The results are shown in Table 2 and table 3 (because the robot is symmetrical left and right, only one side of the hinge is taken for analysis). In the table, - 10 represents that the distance between the two support arms decreases by 10 mm, and 10 represents that the distance between the two support arms increases by 10 mm. The following measured data are the average between 0~8 s

based on the strength of the front and rear arms, it can be seen from table 2 that the force on the front and rear hinges is the smallest at point 3, that is, when the distance between the two arms is 245mm

similarly, comprehensively consider the torque at the hinge of the front and rear support arms. It can be seen from table 3 that when the distance between the two arms is 245 mm at point 3, the torque at the hinge of the front and rear support arms is the smallest

therefore, it is considered that when the distance between the two arms is 245 mm, the torque and force at the hinge of the front and rear support arms are the smallest, so the distance between the two arms that minimizes the power output of the motor is obtained

4 conclusion

a virtual prototype model of the robot is established by using ADAMS software, and the support arm of the robot is analyzed and optimized, and the distance between the two arms that minimizes the motor power is obtained. This method can be used for dynamic visual simulation and optimization of general mechanical structures, and can provide reference for solving similar problems

References:

[1] Wang Haiying, you Bo, sun Xiaobo, etc. Optimal design of mobile robot dexterous hand [j] Journal of petrochemical colleges and universities, 2007, 20 (3):

[2] Chen Xu, Gao Feng Central air conditioning ventilation duct cleaning robot [j] Light industrial machinery, 2007 (8):

[3] Cheng Jun Joint simulation of heterogeneous two legged walking robot [j] Journal of system simulation, 2007 (21): 4955

[4] duzhijiang Research on motion simulation of Biped Robot Based on Virtual Prototype Technology [j] Journal of system simulation, 2007 (19): 4456

[5] Liu Yanwu Simulation of bearing capacity of parallel robot based on ADAMS [j] Manufacturing automation, 2007 (7):

[6] Caselles V, Kimmel R S. Geodesic active contours[J]. International Journal of Computer Vision, 1997, 22(1):.

[7] Ambrosio L, Tortorelli V M. Approximation of functional depending on jumps by elliptic functional viaconvergence comm[J]. Pure and Appl Math, 1990, 43(8): 036.