Preliminary investigation on performances of a bernoulli gripper in grasping leather products.

Siketova, Katarina ; Liska, Jan ; Horvath, Stefan 等

1. INTRODUCTION

In leather industry there are many problems related to labour: the environment is critical because of the presence of different chemical products, the leather plies could be heavy and their manipulation for loading the machines could be tiring for operators.

Nevertheless, the automation of these processes presents many difficulties. In this phase the leather plies are very flexible, delicate and their surface is liable to imprint, if grasped by unappropriatetools. In particular, the irregularity of their borders makes impossible the grasping of each single ply from the top of the stack. All these characteristics in practice avoids the use of standard grippers to automate the manipulation of leather plies by robots.

Interesting contributions in the field of flexible object manipulation is given in [Seliger, 2003]. Furthermore, in [Failli, 2004], a specific gripper based on special vacuum cups disposed as a grid has been developed. This technique assures the required grasping forces, their correct distribution on the surface of the ply, and a sufficient delicacy in approaching the object, due to the cup material and to its special internal shape.

However, in some conditions (e.g.: very delicate leather surfaces, wet conditions, etc.) some imprints remain on the surface after manipulation.

[FIGURE 1 OMITTED]

Therefore, the study of a non-contact gripper based on the Bernoulli effect seems to be very attractive for this kind of problem.

2. BERNOULLI GRIPPER

Bernoulli grippers (or air flow grippers) are typical non-contact grasping devices [Monkman, 2007]. Some applications have been investigated in literature, such as in [Erzincanli, 1996] and [Davis, 2006], but the use of such devices in grasping leather has to be deeply demonstrated.

2.1 Working principle

The working principle is based on the well-known Bernoulli's law. The air flow moves between the internal gripper ring surface and the deflector. The vacuum for grasping the object is created by both the turbulent flow in point X (Fig. 1) and the increasing of air speed in the gap between the gripper and the object.

2.2 Bernoulli gripper configurations

The design proposal of the Bernoulli gripper includes two basic parts: the gripper body and the deflector. Three air inputs on the external cylinder surface have been realized by reason of obtaining a stable and constant air flow. The aim of the design is to investigate on the influence of: i) different air flow angles; ii) different disk surfaces (Fig. 2). In particular, two different deflectors having an air flow angle respectively of 30[degrees] and 60[degrees] have been designed and realized. As far as the disk surfaces, two different types have been considered: a simply flat surface and a radially slotted surface, including eight Venturi channel used to increase the vacuum effect (a detail of these channel is given in the lower part of Fig. 2).

In this way, four different gripper configurations result: i) 30[degrees]-flat, ii) 60[degrees]-flat, iii) 30[degrees] slotted, iv) 60[degrees]-slotted.

[FIGURE 2 OMITTED]

3. EXPERIMENTAL TESTS

[FIGURE 3 OMITTED]

The aim of the experimental tests is to identify the influence of different process and design parameters on the measurement of the maximal suction force exerted by the gripper.

It is very important to emphasize that these tests are preliminary and concern only the evaluation of lifting force on leather objects in some specific conditions.

3.1 Experimental setup

Experimental tests have beencarried out in the laboratory of Department of Mechanical, Nuclear and Production Engineering of the University of Pisa. The experimental setup is schematically illustrated in Fig. 3 and is mainly formed by a Scara robot, a Kistler dynamometer model 9257B, a charge amplifier and a PC.

Three different objects have been grasped:

* leather A: leather with a thickness of 1 mm (very flexible material);

* leather B: leather with a thickness of 2 mm (flexible material);

* plastic (rigid and non-porous material).

The lifting force has been measured using 3 different values ofair flows (100, 200 and 250 LPM), using the following repetitive cycle:

1. the gripper goes down at a given distance from the object surface (0.5, 1 and 1.5 mm);

2. itstops at the previous distance for 5 s;

3. it comes back to the start position.

3.2 Results

Fig.4 to 7 show some results obtained during experimental tests. Each graph reports the average value of the lifting force (F-average) measured during the working cycle in function of the distance between the gripper and the object (gap) for 3 different materials.

The following preliminary considerations can be made:

* generally, lifting force increases with gap, with the exception of using the 60[degrees] deflector in grasping rigid material;

* the maximum lifting forces measured in grasping leather object are lower than those ones obtained for rigid material (from 25 to 50%);

* the best performances in grasping leather have been obtained using the 60[degrees] deflector with a maximum value of about 5 N using a flat disk and about 7 N using a slotted disk. Unsatisfactory values have been obtained using a 30[degrees] deflector.

4. CONCLUSIONS

The preliminary experimental tests performed in this study have shown that the performances of the grippers are different using different design parameters and different materials. The presence of the Venturi channels and a 60[degrees] deflector gives a positive effect, but further investigations should be made in the future.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

5. REFERENCES

Davis, S.; Gray, J.O.; Caldwell, G. (2006) An end effector based on the Bernoulli principle for handling sliced fruit and vegetables, Robotics and Computer Integrated Manufacturing.

Erzincanli, F.; Sharp, J.M. (1996) Development of a non contact endeffector for robotic handling of non-rigid materials, Robotica, Vol.15, pp 331-335.

Failli, F.; Dini, G. (2004), An innovative approach to the automated stacking and grasping of leather plies, Annals of the CIRP, Vol.53/1, pp 31-34.

Monkman, G.J.; Hesse, S.; Steinmann, R.; Schunk, H., (2007), Robot grippers, Wiley-Vch.

Seliger, G.; Szimmat, F.; Niemier, J.; Sephan, J. (2003), Automated handling of non-rigid parts, Annals of the CIRP, Vol.52/1, pp 21-24.