Pneumatic linear actuator

Abstract

The invention relates to a pneumatic actuator. According to one embodiment, the actuator has a housing with a pressure chamber, a rod inserted from the outside into the pressure chamber of the housing, a rod seal arranged around the rod, which seals the pressure chamber, and a rod guide mounted on the housing, which is configured to guide the rod along its longitudinal axis. There is no piston in the pressure chamber. Instead, inside the pressure chamber, there is an unsealed annular gap between the rod and the inner wall of the pressure chamber, so that the gas pressure present in the pressure chamber can propagate throughout the pressure chamber to the rod seal.

Description

Technical Field

[0001] This invention relates to the field of pneumatic actuators, and more particularly to a low-cost design of a pneumatic linear actuator that can replace a double-acting pneumatic cylinder in certain applications. Background Technology

[0002] Pneumatic actuators come in many different types. Pneumatic cylinders, in particular, are used in a wide range of applications. Pneumatic actuators are also used for force control, such as in automated, robot-assisted surface machining, or generally in applications where the robot should make “gentle” contact with the surface (without collision) with the tool. Examples of pneumatic handling devices for industrial robots are described in US10,906,177. Known devices and systems include bellows cylinders, pneumatic muscles, or double-acting pneumatic cylinders, which makes such devices complex and expensive to manufacture. The requirement for such handling devices for robot-assisted surface machining is the ability to withstand bending moments.

[0003] The inventor's aim is to create a low-cost pneumatic linear actuator, particularly suitable for applications in robot-assisted surface machining. Summary of the Invention

[0004] This document describes a pneumatic actuator. According to one embodiment, the actuator has: a housing having a pressure chamber; a rod inserted into the pressure chamber from the outside of the housing; a rod seal disposed around the rod, sealing the pressure chamber; and a rod guide mounted on the housing, configured to guide the rod along its longitudinal axis. There is no piston in the pressure chamber. Instead, an unsealed annular gap exists between the rod and the inner wall of the pressure chamber inside the pressure chamber, allowing the gas pressure present in the pressure chamber to propagate throughout the pressure chamber to the rod seal.

[0005] Other embodiments relate to a processing device or a linear actuator having one or more of the above-described pneumatic actuators. Attached Figure Description

[0006] The various implementations will now be explained in more detail with reference to the examples shown in the figures. The representations are not necessarily to scale, and the invention is not limited to these aspects. Rather, the focus is on illustrating the basic principles of the proposed embodiments.

[0007] Figure 1 This is an exemplary schematic diagram of a robot-assisted grinding device having a grinding machine coupled to an industrial robot by means of a force-controlled linear actuator; the linear actuator partially decouples the industrial robot and the grinding machine.

[0008] Figure 2 An example of a pneumatic linear actuator handling device is shown in a side view.

[0009] Figure 3 This is a longitudinal sectional view of the first embodiment.

[0010] Figure 4 This is a longitudinal sectional view of the second embodiment.

[0011] Figure 5 This is a longitudinal sectional view of the third embodiment under moderate deflection.

[0012] Figure 6 It shows Figure 5 Example of minimum deflection. Detailed Implementation

[0013] In robot-assisted surface machining, machine tools (e.g., grinders, drills, milling machines, polishers, etc.) are guided by manipulators (e.g., industrial robots). During this process, the machine tool can be coupled via a so-called end effector flange in various ways, the position of which determines the tool center point (TCP) coupling of the manipulator; typically, the manipulator can adjust the position and orientation of the TCP almost arbitrarily to move the machine tool along a trajectory, for example, parallel to the workpiece surface. Industrial robots are typically position-controlled, allowing the TCP to move precisely along the desired trajectory. However, the following applies not only to robot-assisted surface machining but also to general robotic applications where the robot must make more or less gentle contact with the surface with the tool (without collisions). This might also apply to pick-and-place applications, for example.

[0014] In many applications, control of process forces (such as forces upon contact with a workpiece or contact forces during surface machining, such as grinding forces) is essential, and this is often difficult to achieve with sufficient precision using conventional industrial robots. The large, heavy arms of industrial robots have too much inertia, making it difficult for closed-loop controllers to react quickly enough to fluctuations in process forces. To address this issue, a smaller (lighter) linear actuator, less powerful than the industrial robot itself, can be placed between the end effector flange of the robot and the machine tool, coupling the end effector flange to the machine tool. During surface machining, the linear actuator controls only the process forces (i.e., the contact forces between the tool and the workpiece), while the robot, including the linear actuator, moves the tool or machine tool along the desired trajectory under position control. Through force control, the linear actuator can compensate for inaccuracies in the position and shape of the workpiece being machined, as well as inaccuracies in the robot's trajectory (within certain limits). However, some robots can adjust process forces through force / torque control without using the aforementioned linear actuator, but this method is relatively complex and expensive.

[0015] Before detailing the various embodiments, a general example of a robot-assisted grinding apparatus is first described. It is understood that the concepts described herein can also be applied to other types of surface finishing (e.g., polishing, milling, drilling, etc.), and are not limited to grinding. As mentioned above, the embodiments described herein can be used as linear actuators (processing devices) in various applications and generally represent a low-cost alternative to linear actuators driven by pneumatic cylinders.

[0016] according to Figure 1 The robot-assisted grinding device includes a manipulator 80, such as an industrial robot, and a grinding machine 50 with a rotating grinding wheel 51. The grinding machine 50 can be coupled to the end effector flange of the manipulator 1 via a linear actuator 20. The position (position and orientation) of the end effector flange also determines the TCP. Strictly speaking, the TCP is not a point, but a vector that can be described by three spatial coordinates (position) and three angles (orientation). In robotics, the position of the TCP is sometimes described using generalized coordinates (usually the six joint angles of the robot) in configuration space. The position and orientation of the TCP are sometimes also referred to as the "pose". The position (including orientation) of the TCP, as a function of time, defines the motion of the grinding wheel, called the trajectory. The center point of the robot's end effector flange is usually defined as the TCP, but not necessarily. The TCP can be any point (theoretically, it can also be located outside the robot), and its position and orientation can be adjusted by the robot. The TCP can also define the origin of the tool coordinate system.

[0017] In the case of a six-degree-of-freedom industrial robot, the manipulator 80 can consist of four segments 82, 83, 84, and 85, which are connected by connectors G. 11 G 12 and G 13 Connection. In this case, the first segment 82 is typically rigidly connected to the base 81 (however, this is not always the case). Connector G 11 Connecting sections 82 and 83. Connector G 11 It can be dual-axis, and allows segment 83 to rotate about a horizontal rotation axis (elevation) and a vertical rotation axis (azimuth). Connector G 12 Connecting segments 83 and 84, and allowing rotational movement of segment 84 relative to segment 83. Connector G 13 Connecting sections 84 and 85. Connector G 13 It can be 2-axis, therefore (similar to connector G) 11 The end effector flange and TCP have a fixed relative position to segment 85, which typically also includes a swivel joint (not shown) that allows the end effector flange 86 located on segment 85 to rotate about the longitudinal axis A of segment 85. Figure 1(Seen as dashed lines, which in the example also correspond to the rotation axis of the mold). Each axis of the joint is assigned an actuator (e.g., an electric motor) that can rotate about its respective joint axis. The actuators in the joint are controlled by robot controller 70 according to the robot program. Various industrial robots / manipulators and associated controllers are known in themselves and will not be explained further here.

[0018] Robotic arms 80 are typically position-controlled, meaning the robot controller can determine the pose (position and orientation) of the TCP and move along a predefined trajectory. Figure 1 In the diagram, the vertical axis of segment 85, where the TCP is located, is denoted by A. When actuator 100 is connected to the end stop, the pose of the end effector flange (or TCP) also defines the pose of the grinder 50 (and the tool / grinding wheel 51). As previously described, linear actuator 100 is used to adjust the contact force (process force) between the tool and the workpiece 60 to the desired value during grinding. For grinding applications, direct force control by robot 80 is generally too inaccurate because the high inertia of segments 83 to 85 of robot 80 makes it nearly impossible to quickly compensate for force peaks (e.g., when placing the grinding wheel on the workpiece 60) with conventional robots. Therefore, robot controller 70 is configured to control the pose (position and orientation) of the TCP of robot 80, while force control is typically achieved solely by actuator 100.

[0019] As mentioned earlier, during the grinding process, the contact force F between the grinding wheel (grinding machine 50 with grinding wheel 51) and the workpiece 60 can be adjusted by means of the linear actuator 100 and the force controller (e.g., implemented in the controller 70). K This results in a contact force F between the grinding wheel 51 and the workpiece 60. K (In the direction of the longitudinal axis A) corresponds to the predetermined set point. Contact force F K The actuator force F applied by the linear actuator 100 to the surface of the workpiece. A The actuator 100 responds to the lack of contact between the workpiece 60 and the tool 51. Due to the lack of contact force on the workpiece 60, the actuator 100 rests against the end stop (not shown, as it is integrated into the actuator 100) and presses against the end stop with a defined force. The force controller can always be active. In this case (no contact), the actuator deflection is therefore maximum, and the actuator 100 is in the terminated position. The defined force of the actuator 100 pressing against the end stop can be very small, or (theoretically) even adjustable to zero, to allow for the smoothest possible contact with the workpiece surface.

[0020] The position control of the robotic arm 80 can operate completely independently of the force control of the actuator 100 (or it can be implemented in the controller 70). The actuator 100 is not responsible for the positioning of the grinding machine 50, but only for adjusting and maintaining the desired contact force F during the grinding process. K It also detects the contact between the tool 51 and the workpiece 60. The contact can be detected in a simple way, for example, if the actuator has moved out of the end position (the actuator deflection is less than the maximum deflection on the end stop).

[0021] It should be understood that the direction of motion of actuator 90 and the rotation axis of grinding machine 50 do not necessarily coincide with the longitudinal axis A of segment 85 of robot arm 80. In the case of pneumatic actuators, force control can be achieved in a known manner through regulating valves, regulators (e.g., implemented in controller 70), and compressed air reservoirs or compressors. Since the tilt relative to the vertical direction is related to gravity (i.e., the weight of grinding machine 50), actuator 100 may include a tilt sensor, or this information may be determined based on the joint angle of robot arm 80. The determined tilt is taken into account by the force controller. The specific implementation of force control is well known and is not important for further explanation, therefore no further details are provided. Actuator 100 not only allows for a certain mechanical decoupling between robot arm 80 and workpiece 60, but also compensates for inaccuracies in TCP positioning.

[0022] In another type of robot-assisted surface machining, the machine tool is mounted on a fixed base via linear actuators, while a conventional industrial robot positions the workpiece onto the machine tool (e.g., a grinder). Process force control is performed sequentially by the linear actuators, while the robot's position is controlled in a conventional manner. This means that during surface machining, the linear actuators press the machine tool against the workpiece (supported by the base), and the workpiece is held in a predetermined position by the robot.

[0023] The linear actuator 100 is also referred to below as a handling apparatus. Figure 2 An embodiment is shown in a side view. According to... Figure 2 The device has two opposing mounting plates 101 and 102 (mounting flanges), wherein the first mounting plate 101 is configured to mechanically couple the device to a cutting tool (e.g., a gripper) or machine tool (e.g., a grinder, polisher, etc.), and wherein the second mounting plate 102 is configured to mechanically couple the device to the end effector flange 86 of a robot (see...). Figure 1 For example, the second mounting plate 102 can be mounted to the end effector flange 86 using screws. Similarly, the machine tool can be mounted to the first mounting plate 101 using screws. Other alternative mounting options (clamps, bayonet locks, etc.) are also possible.

[0024] In the example shown, the interior of the processing unit located between mounting plates 101 and 102 is covered by a bellows 105. This is essentially to keep dust and other impurities away from the internal components of the unit. Other covering designs are also possible.

[0025] Figure 3 A first example of the processing apparatus described herein is illustrated schematically. It should be emphasized that this is not a conventional combination of piston and (pneumatic) cylinder, but simply a rod 110 guided in a rod guidance 112 inserted inside the housing 130. A standard rod seal 113 seals the interior space of the housing 130 along the circumference of the rod 110. That is, the rod seal 113 and the rod guidance 112 are axially spaced from each other (along the longitudinal axis B of the rod). The rod 110 is mounted in the rod guidance 112 along its longitudinal axis and is movable along the longitudinal axis. In the example shown, the rod guidance 112 is disposed in a recess 131 in the housing 130. For example, the rod guidance 112 can be pressed into the recess 131. Other techniques for securing the rod guidance 112 to or onto the housing 130 are also possible. The rod guidance 112 may be or include a recirculating ball guide. The rod guidance is typically made of stainless steel. Different rod guides are known and commercially available, so they will not be discussed further here.

[0026] The rod guide 112 allows the rod 110 to move only in the longitudinal direction (along the longitudinal axis B) and is particularly effective at absorbing bending moments, i.e., torques on the normal / lateral axis about the longitudinal axis B. The rod guide is also called a shaft guide, and in particular, a linear bearing. In one embodiment, a linear ball bearing is used as the rod guide. Linear ball bearings, also known as ball bushings, have the advantage of generating relatively small (virtually none) static friction between the bearing and the rod, thus largely avoiding stick-slip effects.

[0027] When rod 110 is displaced, the volume of the internal space of housing 130 changes. The internal space of the housing can provide compressed air (see...). Figure 3 The internal space (compressed air inlet / outlet 115) is referred to below as pressure chamber 114 (pressure p1). The end of rod 110, located outside housing 130, is connected to one of the mounting plates (in the example shown, it is connected to mounting plate / flange 101). Housing 130 is mounted on another mounting plate (in the example shown, mounting plate / flange 102). Mechanical connections between rod 110 and mounting plate 101, and between housing 130 and mounting plate 102, can be made, for example, by screws. However, other connection techniques are also possible (e.g., gluing, crimping, etc.).

[0028] When the air pressure in pressure chamber 114 is p1, the force F1 acting on rod 110 (along its longitudinal axis B) is equal to p1·A1, where A1=d1 2 π / 4. The pressure p1 is typically overpressure, i.e., greater than the atmospheric pressure outside the pressure chamber. Here, parameter d1 represents the diameter of the rod 110 in the pressure chamber 114, particularly the diameter of the rod 110 in the region of the rod seal 113. In the example discussed here, the pressure chamber 114 is shaped as a cylinder with an inner diameter d1', wherein there is an (annular) gap δ (i.e., d1' = d1 + 2·δ) between the circumference of the rod 110 and the inner wall of the pressure chamber 114. The force F1 generated by the compressed air pushes the two mounting plates 101 and 102 apart against a restoring force F, which may be generated, for example, by the spring 150. R The function of the spring 150 is as follows. In the example shown, the spring 150 also acts between the two mounting plates 101 and 102, and produces a displacement ΔL relative to the rod 110 (see [reference]). Figure 2 The relevant restoring force F R (F R ≈k·ΔL, where k represents the spring constant. The minimum distance between mounting plates 101 and 102, for example, the minimum distance defined by the end stop is L0 (ΔL=0, see Figure 2 The maximum distance between mounting plates 101 and 102 can also be determined by the end stop. Figure 3 The end stop is not shown. Other reset elements can be used instead of spring 150.

[0029] Unlike conventional piston / cylinder combinations, the effective pneumatic area is equal to the cross-sectional area of ​​the rod 110 within the rod seal 113 region. In the example shown, there is no equivalent of a piston seal (which would move with the piston), only the rod seal 113 mounted in the housing 130 (and not moving with the rod). Since there is no piston with a piston seal in this described example, the gas pressure p1 present inside the housing 130 can propagate throughout the pressure chamber 114 (and thus into the annular gap δ) up to the rod seal 113. In contrast, a piston would divide the interior space of the housing 130 into two pressure chambers, which is not the case in this described example. The housing 130 includes only one (single) pressure chamber 114. Meanwhile, the rod guide 112 (linear bearing) ensures reliable absorption of bending moments in a compact and cost-effective design. In conventional actuators using ordinary pneumatic cylinders, linear guides capable of accommodating larger bending moments are arranged separately next to (i.e., parallel to) the pneumatic cylinder.

[0030] The housing 130 can be made of plastic, for example, by injection molding or by additive manufacturing (3D printing). The material used to manufacture the housing 130 is more flexible (less rigid) than the material used to manufacture the rod guide (typically steel). In another example, the housing 130 is manufactured by die casting of aluminum. Machining (e.g., by milling) is only required in the area of ​​the socket 131 and on the surface that may be connected to the mounting plate 102. Overall, Figure 3 The linear actuator described herein is generally easier and less expensive than linear actuators that use conventional pneumatic cylinders as actuation elements.

[0031] exist Figure 4 In the example shown, as a reset element, in order to generate a reset force F R It also provides a device having a rod 120, a housing 140, a rod guide 122, and a rod seal 123 to replace the spring 150. Figure 4 The left side of the middle and Figure 3 The examples in [the document] are the same and refer to the description above. Figure 4 The right side of the device is constructed similarly to the left side, but is connected to the mounting plates 101 and 102 in the opposite direction (upside-down). Two actuators with longitudinal axes B and B' (each actuator includes a housing and a rod with rod guides and rod seals) are arranged antiparallel to each other.

[0032] The illustrated twin-bar arrangement is more stable in absorbing bending moments and can generate greater forces. The two actuators can also be connected in parallel (rather than antiparallel). In this case, Figure 4 The right side of the device will be designed identically to the left side and connected to the mounting plates 101 and 102 in the same manner. In some embodiments, two or more combinations of housings having pressure chambers, rods, and rod guides are provided to increase the maximum actuator force (the sum of the pneumatically effective end faces of the rods under the same pressure in the pressure chamber) and the maximum possible bending moment.

[0033] The housing 140 of the second actuator is connected to the mounting plate 101 (e.g., by screws), and the end of the rod 120 located outside the housing 140 is connected to the opposite mounting plate 102 (also by screws). The housing 140 has a socket 141 in which a rod guide 122 is disposed. A rod seal 123 (axially spaced) is disposed adjacent to the rod guide 122 in the housing 140 (similar to housing 130 and rod seal 113). A pressure chamber 124 is formed inside the housing 140, the volume of which depends on the position of the rod 120. Compressed air (pressure p2) can enter the pressure chamber 124 through the inlet / outlet 125. The force F2 acting on the rod 120 is related to the pressure p2 and the pneumatic effective area A2 = d2. 2 It is directly proportional to π / 4 (i.e., F2 = p2·A2).

[0034] exist Figure 4 In the example, when the gas pressure p2 in the second pressure chamber 124 is suppressed, the combination of the second rod 120 and the second pressure chamber 124 disposed in the second housing 140 can act as a reset element. Pressure refers to the pressure p2 below the atmospheric pressure outside the device. The suppressed pressure chamber causes the rod / housing / pressure chamber combination to actually act like a spring, thereby generating a reset force. In this case, the force is adjusted by setting (over)pressure in another pressure chamber, because in practice, overpressure is easier to adjust than low pressure.

[0035] like Figure 3 In this example, the effective pneumatic surfaces A1 and A2 are also equal to the cross-sectional areas of rods 110 and 120 in the regions of rod seals 113 and 123. The areas of surfaces A1 and A2 can be equal. Unlike conventional solutions, this example does not require a piston seal. Rod seals 113 and 123 are respectively housed in their respective housings 130 and 140 and are immovable relative to their respective housings 130 and 140. A spring is not required in this example because the return force F2 is pneumatically generated. However, a spring (similar to...) can also be provided separately. Figure 3 ).

[0036] As described above, housings 130 and 140 can be made of a more flexible (less rigid) material than the material used to manufacture rod guides 112 and 122 (typically steel). For example, housings 130 and 140 could be made of plastic (injection molding) or aluminum (die casting). As previously mentioned, additive manufacturing processes (3D printing) are also possible. The relatively more flexible housings allow for compensation (within certain limits) of deviations in the perfect parallelism of the longitudinal axes B and B' of rods 110 and 120, respectively, and prevent the linear actuators from jamming. The parallelism deviations of the longitudinal axes B and B' could be due, on the one hand, to manufacturing tolerances, and on the other hand, to bending moments during operation.

[0037] Figure 5 Another embodiment is shown, which is similar to Figure 4 The examples in [the text] are very similar. In terms of functionality, Figure 5 The embodiments and Figure 4 The embodiments are basically the same, wherein two anti-parallel actuators are arranged between mounting plates 101 and 102. In addition, a spring 150 is arranged between the two mounting plates 101 and 102 as a reset element, so that the linear actuator / processing device can be in a specified end position even when the pressure chambers 114 and 124 of the two actuators are not compressed.

[0038] Similar to the previous example, housings 130 and 140 have insertion holes 131 and 141 for rod guides 112 and 122, respectively. Rod seals 113 and 123 are arranged coaxially with rod guides 112 and 122 within their respective housings. The two rods 110 and 120 are guided antiparallel within rod guides 112 and 122. The volumes of pressure chambers 114 and 124 within the housings change depending on the position of the rods (i.e., depending on the deflection ΔL of the linear actuator). Figure 5 In the case shown, the deflection ΔL is in a moderate range. The maximum volume of pressure chambers 114 and 124 is determined by end stops (not shown). Rod 110 is rigidly connected to mounting plate 101 by screw 111. Similarly, rod 120 is connected to mounting plate 102 by screw 121. The associated housings 130 and 140 are rigidly connected (e.g., screwed) to their respective other mounting plates.

[0039] In the retracted state (i.e., minimum deflection ΔL = 0), the ends of rods 110 and 120 located inside pressure chambers 114 and 124 cannot completely rest against the walls of the pressure chambers; otherwise, surfaces A1 and A2 would no longer be available for pressure application, generating corresponding forces p1A1 and p2A2 respectively. Figure 5 The screw 129 shown is screwed into the end side of the rod 120, forming an end stop for the retracted state. The screw 129 extends from the end side of the rod 110, thus also forming a spacer, such that at least a portion of the cross-sectional area A2 remains pneumatically effective, so that even in the fully retracted state, force can be applied to the rod due to the pressure p2 in the pressure chamber 124. The spacer on the rod 110 is not absolutely necessary due to the mechanical coupling of the two rods 110, 120 and the associated housings 130, 140 to the mounting plates 101, 102.

[0040] As a magnetic displacement sensor, Figure 5 and Figure 6 A portion of the permanent magnet 118 (not shown) can be secured to another rod (left rod 110) with screw 119. A displacement sensor is used to measure the displacement ΔL of the linear actuator. Various types of suitable magnetic and other displacement sensors are known in themselves and therefore will not be discussed further. Relevant to the example shown is the magnet 118 mounted on rod 110 and moving with the rod, which allows for simple displacement measurement.

[0041] Figure 6 It shows Figure 5The device is in its fully retracted state (ΔL = 0). It can be seen that the screw head of screw 129 forms an end stop and is connected to the wall of pressure chamber 124 (the side opposite the end of rod 120). Only the screw head of screw 129 contacts the wall of pressure chamber 124, not the end of rod 120. Rod 120 is a distance x from the opposite wall of pressure chamber 124 in the position shown (ΔL = 0). It should be understood that screw 129 can also be screwed into the opposite wall of pressure chamber 124, instead of the end of rod 120. Instead of screw 129, a spacer can also be formed directly onto the housing wall or rod. The spacer and housing 140 can be a single component (e.g., a casting).

Claims

1. An apparatus for robot-assisted surface processing, characterized by include: A first housing (130) having a first pressure chamber (114); The first rod (110) is inserted from the outside into the first pressure chamber (114) of the first housing (130). A first rod seal (113) is provided around the first rod (110) to seal the first pressure chamber (114), wherein there is an unsealed annular gap (δ) between the first rod and the inner wall of the first pressure chamber (114) inside the first pressure chamber (114), such that the gas pressure (p1) present in the first pressure chamber (114) can be propagated to the first rod seal (113) throughout the first pressure chamber (114). A first rod guide (112) mounted on the first housing (130) is configured to guide the first rod (110) along its longitudinal axis, wherein the first rod guide (112) is spaced apart from the rod seal (113); and A reset element (150) is arranged to counteract the force applied to the first rod (110) by gas pressure (p1); A second housing (140) having a second pressure chamber (124) is arranged side by side and parallel to the first housing (130); A first mounting plate (101) and a second mounting plate (102), wherein the first mounting plate (101) is firmly connected to the first rod (110), and wherein the first housing (130) is firmly connected to the second mounting plate (102); The reset element is configured to generate an attractive force (F R , F2) between the first mounting plate (101) and the second mounting plate (102) outside the first housing (130). The reset element (150) is a spring.

2. The apparatus according to claim 1, characterized in that, The first housing (130) has a single pressure chamber, and therein, the pressure is substantially the same throughout the first pressure chamber.

3. The apparatus according to claim 1 or 2, characterized in that, The first housing outside the pressure chamber (114) includes a socket (131), and the first rod guide (112) is disposed in the socket (131).

4. The apparatus according to claim 1, characterized in that, The first rod seal (113) is disposed in a groove in the first housing (130) that extends around the first rod (110).

5. The apparatus of claim 1, wherein, Also includes: The second rod (120) is inserted from the outside into the second pressure chamber (124) of the second housing (140); A second rod seal (123) is provided around the second rod (120) to seal the second pressure chamber (124), wherein, inside the second pressure chamber (124), there is an unsealed annular gap (δ) between the second rod (120) and the inner wall of the second pressure chamber (124), allowing the gas pressure (p2) present in the second pressure chamber (124) to propagate throughout the second pressure chamber (124) to the second rod seal (123); and A second rod guide (122) is mounted on the second housing (140) and is configured to guide the second rod (120) along its longitudinal axis (B').

6. The apparatus according to claim 5, characterized in that, The reset element is formed by a combination of the second pressure chamber (124) and the second rod (120), wherein the gas pressure (p2) present in the second pressure chamber (124) is negative.

7. The apparatus according to claim 5 or 6, characterized in that, The second mounting plate (102) is securely connected to the second rod (120), and wherein the second housing (140) is securely connected to the first mounting plate (101).

8. The apparatus of claim 1, wherein, Also includes: The spacer (129) used as an end stop is located on the end side of the first rod (110) in the first pressure chamber (114), or on the wall of the second pressure chamber (124) opposite to the end side of the first rod (110), and the spacer (129) in the retracted end position of the device ensures the distance (x) between the end face of the first rod (110) and the opposite wall of the first pressure chamber (114).

9. The apparatus according to claim 8, characterized in that, The spacer (129) is screwed into the first rod (110) or the first housing (130); or The spacer (129) is a component of the first housing (130) or the first rod (110).

10. The apparatus according to claim 1, characterized in that, The first housing (130) is made of a material that is more elastic than the material used to manufacture the first rod (110).

11. The apparatus according to claim 1, characterized in that, The device is used as a pneumatic linear actuator; the first rod guide (112) is configured to absorb bending moments.

12. The apparatus according to claim 11, characterized in that, The rod guide (112) is a linear bearing.

13. The apparatus according to claim 11 or 12, characterized in that, The rod guide rail (112) is disposed in the insertion hole of the housing (130), located outside the pressure chamber, and spaced apart from the rod seal (113).

14. A system for robot-assisted surface processing, characterized by include: robotic arm; The device mounted on the end effector flange of the robot according to any one of claims 1 to 10, and cutting tools or machine tools mounted on the device.

15. A system for robot-assisted surface processing, characterized in that include: Robotic arms used for holding and positioning workpieces; The device mounted on the base according to any one of claims 1 to 10, and cutting tools or machine tools mounted on the device.

Images