A Gecko Inspired Fluid Driven Climbing Robot

From Control Systems Technology Group
Jump to navigation Jump to search

Existing wall-climbing robots are driven by motors and very complicated mechanisms. Most important drawbacks of these types of wall-climbing robots are their low speed and low load capacity. This article describes how these problems are circumvented by designing a rapidly-climbing and heavy-loading robot by applying a simple fluid driven mechanism based on the movement of geckos. Fluid-driven robots own namely a special benefit in comparison to the conventional technologies (wall-climbing robots carrying an on-board pump for vacuum creating). Fluid-driven mechanisms allow the robot to get most of the power from the pressure of the fluid, regardless of the battery level of the control circuit. Besides this, the characteristic fast responses and high power to volume ratio of fluid driven robots are very well-suited for fast climbing and heavy loading.

First the gecko’s movement is studied in laboratory and translated into a two-dimensional kinematic model. Different walking modes of the gecko are considered. In trotting mode (and in quick turns) the gecko moves by means of swinging its middle while keeping two legs stuck on the wall. In normal walking mode the motion of the gecko is initiated by constantly moving one leg while keeping the other three stuck on the wall.

Subsequently, the model of the gecko’s motion is used to design a five-linkage-system and one piston mechanism resulting in the design of a rapid wall-climbing robot. The schematic mechanism of the robot is shown in the figure. It consists of four suckers, two T-shaped linkages (C and G), three straight linkages (D, E and F) and rotating joints, indicated with the green and violet points. The whole five-linkage mechanism is fixed on base plate A by means of the violet rotating joints. The piston mechanism, indicated with G, is also attached on the base plate and allows the robot to climb straight on a wall in the so-called trotting mode. The figure on the right illustrates the movement of the robot and especially the use of the piston mechanism. The legs indicated with red circles mean that they are not stuck on the wall, blue circles imply that those legs are stuck on the wall. The robot has on each side an opposite sucker stuck on the wall, and the piston pulls the five-linkage mechanism to move the other two ‘free’ suckers to its new position. Subsequently, the other two suckers are moved in the same way. This motion is repeated and enables the robot to climb on a wall. If the robot wants to turn, two suckers on the same side are stuck on the wall and the remaining two ‘free’ suckers are displaced by an identical mechanism.

Capture24.JPGCapture25.JPG

An important part of the design of the mechanism is the mechanical analysis of the straight line climbing mode. The centre of gravity of the robot is assumed to be in the center of the base plate and the gravitational force is decomposed into a vertical and horizontal part. The output force of the piston is expressed as function of the gravitational force by means of the geometry between the joints and linkages and the principle of virtual work. Friction forces and absorbing forces of the suckers are included in the mechanical model by means of friction factors and geometrical relations respectively.

As mentioned before, the piston is driven by fluids (pressure water, compressed air, oil). The fluids also enable the robot to create a vacuum for the vacuum suckers. Instead of using a vacuum pump, the fluid flows through the vacuum generator which produces a vacuum for the suckers. Main benefits of this vacuum generator above a conventional vacuum pump is its small volume and light weight. The fluid flows are modelled by means of the venturi effect: when a fluid is accelerated when it passes through a throat, the pressure will decrease according to Bernoulli’s equation. If both the diameter of the throat and the pressure are known, the absorbing forces can be calculated. Eventually it is possible to express the the piston output forces as function of the piston driven function.

Finally, after the modeling the movement and fluid flows, a prototype has been developed and extensively tested. A detachable design is made in order to check and maintain it easily. The experiments were done under different conditions. For example, the water supply from the ground source was changed, and the robot was loaded with different weights. These different conditions influenced the performance (e.g. speed, climb frequency) but in all situations it followed that the robot could indeed climb and turn on a vertical glass wall. The robot was also tested with success on other surfaces: a granite wall, a wooden door, a whiteboard and a glass ceiling. The climbing speed dropped but it was still able to climb on the walls. Another important result was that very little energy was consumed by the control circuit, which allows the robot to work for ten hour continuously. This makes it very suitable for cleaning windows or large surfaces. The last important conclusion states that climbing time of the robot has two main components: the adhering and detaching times of the suckers and the walking time. The former is the bottleneck of the system, it requires time to develop vacuum for creating enough adhesion force. This part takes for the robot approximately 90% of the climbing time while a gecko can perform this in 50% of the climbing time. So therefore it is clear that there are possibilities for improving the efficiency of the robot.