Modelling of a motorcycle/mountain bike suspension and digitization of a cam drum motor control
Sammanfattning: This Master’s thesis was done in collaboration with Öhlins Racing AB, a Swedish suspension system-manufacturer. For Öhlins, the quality of their products is highly important and they are therefore devoting many resources for the development and testing of their products. Because testing is such a big part of what Öhlins as a company is doing, it is important to continuously strive to improve the testing methods used within the company. Two popular methods for improving testing methods are through automation of the machine control and by simulating the test with mathematical models. Both methods have the potential to reduce the time consumed during testing. This project focuses on these two methods and is therefore split in two parts. The first part focuses on digitizing the motor control of a rolling road test bench called the Cam Drum, which is used to do life cycle tests of suspension assemblies, to allow for automated control. In the second part the rolling road test bench has been modelled as a suspension system to simulate tests prior to production. The goal of the digitization is to enable more advanced tests while simplifying usage of the Cam Drum, thereby reducing the time necessary to operate the machine. The goal of the suspension model is to get validation results that point towards the model being good enough to use as a tool when developing new products. A programmable logic controller was connected to the existing frequency drive that controls motor rotational speed and an HMI screen was used to control the controller. Communication between the controller and frequency drive used the serial protocol Modbus RTU. The hardware with which the new motor control system was built was primarily supplied by Siemens. Controller and HMI programming was carried out in Siemens’ software SIMATIC Step 7 using programming languages LAD and FBD. The digital motor control system was live tested with great results and good feedback from the technicians. The only functionality missing is being able to send webserver data over the buildings industrial network due to IT related security reasons. Future work should focus on solving this problem. A front fork suspension model and a rear swingarm suspension model have been modelled in Matlab Simulink. Both models are designed to simulate motorcycle or mountain bike suspension however the front suspension model has only been validated against mountain bike data and the rear suspension model against motorcycle data. An alternative tire model was developed to handle problems linked to conventional 1-dimensional tire models. The new model estimates the area of compressed air in the side view plane and scales the force output accordingly. New values for tire spring stiffness and damping coefficient for this system was freely estimated during validation. Validation was done using camera recorded position signals and position signals recorded with a position sensor. The front suspension model was tested against two different front fork models, but validation finally focused on several test runs done with one of the forks due to insufficient recorded data with the other fork. The result was a correlation between the behaviour of the real and modelled suspension however further tweaking of the tire parameters should give better results. The result should however be sufficient for making estimations. Validation of the rear suspension was done against a camera recorded position signal but as evidence from the front suspension validation shows this is insufficient. The rear suspension validation still requires more work before being utilized as a development tool.
HÄR KAN DU HÄMTA UPPSATSEN I FULLTEXT. (följ länken till nästa sida)