An earlier study introduced a formula for determining the holding force of a toroidal hydrostat gripper when gripping an object of constant radius, considering a linear stress/strain relationship. However, beyond the holding force, understanding the grasping disturbance force exerted on the object due to the rolling behavior during grip adjustment is crucial to ensure the objects are not damaged or displaced, and are successfully picked up. This study aims to extend the existing model by incorporating the disturbance force for objects with variable radii and using a non-linear stress/strain relationship for the membrane material.
A parametric model was developed to estimate these forces as a function of its geometric and material properties. The model was validated through experimental testing and refined using a tuning parameter obtained via optimization algorithms to address parameter uncertainties.
The model approximated the holding forces well, with a maximum deviation of 15\% at the maximum swallow distance. It overestimated the peak disturbance forces by a factor of 2.5. But when normalized, the force curves showed good resemblance.
Surprisingly, the obtained tuning parameter reached a high value of $1.45$. as the parameter uncertainties are thought to be less than 20%. This overestimation could be the result of an additional vacuum force, as evidenced by a clear difference in holding force and gripper behavior is observed in pulling tests for objects with smooth and rough surfaces.

This thesis presents the state-of-the art gripping literature and the implementation and extension of an existing Pseudo-Rigid Body Modelling (PRBM) method for modelling initially-curved compliant grippers out of which a circular object is extracted. From the existing literature, challenges are found in the design and evaluation of concepts of initially-curved compliant grippers, mainly involving the relevant design parameters, such as the thickness and enclosing angle of the finger.

The main goal of this thesis is therefore to present and validate the pull-out force modelling for these fingers within defined load conditions to provide comprehensive insights into the relation between important design parameters, such as the enclosing angle and thickness of the finger.

This goal is accomplished by extending an existing 3R PRB-model for initially-curved beams with a fifth link that represents the object and analysing the kinematics and kinetics to determine the relation for the pull-out force of a gripper finger with defined dimensions and a known load case.

This model is validated by designing and building an experimental test setup in which the reaction forces of a initially-curved testpiece of PLA and stainless steel material are measured. Errors for the kinetics between 12 and 32 percent were determined, consisting primarily of systematic errors.

The validated model is used for a parametric study, where relations between relevant design parameters, such as the enclosing angle and thickness of an initially-curved compliant gripper finger, are determined and visualized in a design chart that are be applied in the design of an initially-curved enclosing gripper prototype.

This research presents a measurement method for a compliant gripper finger, whereby the location of the contact with an object and the exerted grasp force are measured indirectly, instead of directly on the contacting surface. For this purpose, the deformation is measured using strain gauges at two locations on the inside structure of the finger, where high deformations occur. The finger is modeled using a Finite Element Analysis, whereby strain data at the two sensor locations is gathered for a set of different contact locations. Now, by matching the measured data to a modeled scenario, both the actual contact location in this scenario and the accompanying contact force can be determined. For the experimental validation, the compliant finger embedded with the two strain gauges is manufactured and an experimental setup has been built. The results show that the contact location can be inferred from strain measurement and contact force is determined accurately up to order of magnitude of 10^-1 newton. Comparing the force-deformation behavior of the original and the embedded finger, a maximum difference of 6% in actuation force was observed. Hence, the sensing method is useful to indirectly measure contact location and force for a compliant finger in contact with an object, while having a minor influence on the grasping performance of the original design.

Fin Ray Effect (FRE) grippers have proven versatile and effective in pick & place automation applications. However, spherical and lateral round objects are still a challenge for the state-of-the-art, as solutions are unstable when placing those objects off-centre from the longitudinal line. Resulting in large and heavy FRE gripper assemblies. The object often shoots out of the grip, since the gripper cannot counteract the lateral forces from the object with normal forces from the structure, but instead has to rely on friction forces. In this paper, a new spatially designed FRE based gripper is proposed that form-closes around the spherical object, providing a stable grip. The design variables, such as structural variations and different thicknesses of segments, are modelled using LiveLinks between SolidWorks, COMSOL and MATLAB, resulting in an optimal design. The optimal design is then compared to a basic FRE gripper design, using the model, and the stability of both is validated in the real world by an experiment. The new design shows improved performance compared to the Basic FRE grippers, and even shows centering capabilities for frictionless objects.

This research proposes and validates a novel grasping principle to be able to set caging grasps in cluttered environments. A prototype is designed making use of manoeuvering fingers which propagate along the object surface and are covered with everting belts to minimize slip and thus friction forces on the object and environment. Using several lab setups, it is shown that the designed gripper fingers closely follow spherical objects, and do so while exerting only negligible normal and friction forces on the object or surrounding clutter. Furthermore, the gripper is validated with a robotic system in a practical test case of grasping tomatoes from a dense cluttered pile. The gripper successfully grasped 100% of the objects without any control effort in terms of clutter avoidance and grasp pose control. This shows that this novel strategy for setting a caging grasp makes grasping difficult objects in cluttered environments robust and highly successful.

ContextIn a society where sustainability is becoming more and more relevant, coal-burning kamado barbecues are more popular than ever in the Netherlands. From an ignited barbecue a significant amount of energy is lost through heated air. This brings up an interesting design opportunity; “Is it possible to design a product which proposes a more efficient use of the energy of a kamado barbecue?” This project investigates the possibility to design a product powered by energy lost during the process of a barbecue. It has been executed for the company 200Fahrenheit, a relatively young company that is specialised in ceramic barbecues known as kamados. The company is growing rapidly and where most competitors seem to barely change, the company is eager to invest in innovative products and sustainability in general to stay relevant in the future.Approach The foundation of the process was the classic double diamond design approach, consisting of Research & Analysis, Conceptualisation, Embodiment and Final Product. The outcome of the Research & Analysis phase brought valuable insights from extensive desk research to qualitative interviews with users and vendors. It led to a defined persona with valuable needs, preferences and frustrations. From the embodiment part of the process, the aim was to deliver answers to the high risk/high reward assumptions. The emphasis lied on the functionality and feasibility of the implemented technology, which was crucial for the company to know whether or not it was fruitful to continue developing. Research by design, design by doing and prototyping proved to give insights in a time efficient manner.ResultThe result of the project is a product that is able to keep food warm in a container at the right temperature after it has been grilled. The temperature of the container can be set to match the needs of the user. It was found that most heat energy was lost via the chimney of the kamado, which made this a suitable location to harvest energy. The energy that powers the product is completely recovered from the hot air that exits the kamado when it is ignited, in the form of heat energy collected by water. The water transfers the heat to the container, warming up the container. The electricity used by the system is generated by this residual heat as well using Peltier elements. Peltier elements are small electronic plates that are able to convert a temperature difference to electricity. This temperature difference is provided by the water which is always under 100 degrees Celsius, and the outgoing air of the kamado which when ignited, virtually always exceeds 100 degrees Celsius.The product empowers the user to take advantage of the normally wasted energy. Where kamado users were already putting an incredible amount of effort in their dishes, the product now enhances their dish until the last moment where it is presented to the table, under perfect conditions.

The detachment of a product from a suction cup gripper is a challenge that emerged in recent years in the high-speed case packing of packaged food. In this industry suction cups are used to temporarily attach a product to a case packing robot. The detachment of products from the suction cup gripper threatens throughput. In, some cases, the industry notes robots to operate at only 40 picks per minute. This is 30 % of their maximum throughput.
This study aims to gain a better understanding of the dynamic grasping strength
of suction cups. This was done by reviewing the state-of-the-art from industry and state-of-the-practice from literature. It was found that suction cups are the best overall gripper class, and that actuation time poses the main limiting factor for other grippers to perform well in this industry.
Secondly, a six-axis force moment sensor that passes airflow for gripper actuation and without limiting the pull-out load measurement performance was designed, fabricated, and validated. With the current sensor technology, the vacuum hose must be placed over the sensor, leading to the generation of parasitic loads. Passing the airflow through the senor structure is essential to eliminate loads otherwise induced by the stiff vacuum hose. The sensor was designed using strain gauges and a Maltese cross sensor structure. Finite Element Modeling and optimization using sequential quadratic programming was used to determine the dimensions of the sensor. The sensor validation showed a maximum measurement error of 8% in the z direction.
Lastly, with the validated sensor, the dynamic grasping strength of compliant and stiff suction cups was measured using motion paths and products that are typically seen in the packaged food industry. For both the stiff and compliant suction cups, the moment around the axis perpendicular to the plane of motion showed to be a leading factor in detachment of a product from a suction cup gripper. The failure of stiff suction cups is explained by impulse loading, this results in peaks in the load in the force in horizontal direction and the moment perpendicular to the plane of motion. For the compliant suction cups, the detachment was found to be caused by accelerating downwards while the product was not in the center of the gripper. These results have led to the recommendation of two gripper designs. The first, is a suction cup-underactuated hybrid gripper. The second design intently provides rotation freedom to reduce the moment loading on the suction cup. The second design
requires input-shaping to reduce the vibrations at the end of the motion cycle.