Numerical investigation and experimental verification of effect of the radial lip seal geometry on sealing performance

Sealing is the control of fluid interchange between two regions, which share a common boundary. In order to control this boundary, sealing elements are used in machine industry. There are various type of sealing elements for different conditions but for different applications, the sealing elements generally considered as in two groups. These groups are static and dynamic stations. In dynamic situation, there is relative dynamic motion between material couples but in static situations, there is not relative motion. Lip seals, labyrinth seals and piston seals are example of the dynamic seals on the other hand gasket, O-ring and backup ring area static sealing elements. Radial lip seals are machine elements, which are often needed in industry, especially in machine design. Sealing elements have improved considerably with machine technology changing and evolving day by day. The function of the sealing elements must be determined very well for accurate sealing. Radial lip seals has three main parts. These are garter spring, steel frame, and elastomer part. Steel frame maintain the rigidity of seal and support elastomer part. Garter spring provide diametrical distributed load with equally. Elastomer part contains primary lip to maintain the sealing function also in dirty working areas, it contains secondary lip or auxiliary lip to exclude contaminants. It is known that sealing performance of the sealing elements are influenced by many factors. Some of these are; surface roughness, pressure, temperature and viscosity. In this study, it was investigated how the geometry parameter would affect the sealing performance. It is known that when the geometry of the lip is symmetrical, it affects the sealing and it causes leakage. For this reason, optimum range of α and β angles is determined and it is recommended that they remain these ranges in order to provide the sealing. In addition, secondary lip or auxiliary lip provide protection against dust that come from outside and also it is know that secondary lip creates extra friction due to crashing surfaces. Besides, influence of the vertical distance between garter spring and seal lip or in other words effect of the lip thickness is now well known and also it is hardly known how the horizontal distance between garter spring and seal lip. Only the appropriate range values for the effect of these two parameters are known, but the effect on the sealing performance of these parameters within valid ranges are not yet known. In addition, lip seals should have a certain pretension and tightness to ensure a good sealing. Pretension and tightness causes stress concentration of the seal lip and it causes the deformation of the seal lip. In addition, this situation causes friction torque on the lip. Researchers are trying to reduce friction losses in mechanical systems by designing new seal geometries. For this reason, lip geometry was changed in this study and it has determined which geometric parameters could affect the frictional torque at contact area with pre study. In pre-study some parameters were studied which include different lip geometry. In pre-study it was found that most effective geometric parameters of the lip which affect frictional torque, were lip thickness and eccentricity between lip and garter spring. Therefore, five new models were created by changing lip thickness, eccentricity, numerical model was established for these models, and experimental verification was done. In the machine industry, various type of elastomers were used as seal materials. Nitrile rubber (NBR), acrylic rubber (ACM), fluoroelastomers (FKM) and polyurethanes are some of the most popular seal materials in machine industry. In this study, Thermoplastic polyurethane was chosen as radial lip seal material. Thermo plastic polyurethane has many advantages compared to other materials. It has bigger working temperature range than other materials, which is between -30 and 110 0C.Also it has high abrasive resistance, strong mechanical properties and high strain value. It is necessary to understand the mechanical behaviors of rubber like materials, which are seal material in order to make numerical model. Because this type of materials have non-linear behavior unlike metals, the modelling is very difficult but when the studies are examined up to now, many models have been developed for modelling rubber-like materials like Mooney Rivlin, Yeoh, Neo Hookean, Ogden… etc. In this study, these models were examined and it is seen that some models are suitable for larger than %700 strain values. Mooney Rivlin model makes some assumption to make easier non-linear materials modelling which has large strain values. Therefore due to this assumptions Mooney Rivlin model was chosen as the most suitable material model for the polyurethane used in the study. Mooney Rivlin coefficients were obtained because of a detailed literature study. After selection of material model, boundary conditions were determined. In finite element analysis, firstly inner pressure was chosen 0 bar and outer pressure was chosen 0,01 Mpa. Also secondly, due to tightness and garter spring normal force was determined as 1 Newton. Thirdly, position some edge of the seal and shaft was fixed and then friction torque at the seal lip were calculated by using boundary conditions and forming the 2D asymmetric model by finite element analysis (FEA). In order to make finite element analysis, mesh of the geometry should be defined. Therefore, three types of mesh were defined to obtain most accurate results. When the various type of meshes were compared, number of elements and nodes, quality and skewness values were examined. After that, best mesh is chosen and modelling is finished. In finite element analysis part, firstly, reference model, which is taken by SKF, was modeled in ANSYS with 2D and after defining boundary conditions average normal stress at the contact area was found and frictional torque at the contact are was calculated. When the FEA result and experimental result were compared results are very close to each other. The purpose of this thesis is investigate the effect of radial lip seal geometry of sealing performance with examining characteristic of frictional torque at contact area. For this purpose after the finite element analysis experiment system has been developed. In this system, a cylinder is supported with four load cell and so that frictional torque at the contact area could be measured and this moment occurs between lip of seal and shaft. During the experiment, frictional force values are measured and these measurements were transferred to the computer with data acquisition system. After that, frictional torque values were calculated After the finite element analysis (FEA), experimental verification carried out. For the experiment 5 different seal models which include different lip geometry, were manufactured by SKF. All models have same inner diameter have 40 mm and outer diameter 62 mm. Two parts of experiments were done and 18 specimens were used for these experiments. After the test conditions are done, fist part of the experiment was done to verify the FEA results as follows; • Test speed are chosen as 1 m/s. • The seals are tested under 0 bar for a short period. For the second part of the experiment, frictional torque was measured for different speeds and inner pressure. The effect of speed, pressure and geometry were investigated but the other conditions are stayed constant. For the second part of the experiment test conditions were determined as follows; • The speed are chosen as 1, 3, 5 m/s. • The seals are tested under 0, 0,3, 0,5 bar for a short period. As a result, it was seen that the result of finite element analysis and experimental results were confirm the each other and it was determined which geometrical parameters of lip geometry had a positive effect on the sealing performance and which had a negative effect. Because of this work, new seal models will emerge with optimization studies.