Online surface monitoring during thread whirling
Research Project of IFW for Determining Surface Quality During the Thread Whirling Process.
Thread whirling is a widely used process for manufacturing threads. This paper presents the development of a method for determining surface quality during the thread whirling process. The method is being developed as part of a collaborative project between the Institute of Manufacturing Technology at Hannover University and Bornemann Gewindetechnik GmbH & Co. KG.
Thread drives literally drive the economy. They are used in various applications, such as in machine tools, vehicles, and lifting and conveyor technology. To ensure their functionality, thread drives are subject to strict manufacturing tolerances. In addition to meeting geometric tolerances, ensuring a surface with defined tribological properties is essential for the efficient use of a thread drive.
The geometry of the threads has a strong influence on the application area of the thread drive. For example, trapezoidal threads are used for moving heavy loads, as the deep thread flanks allow for the transmission of large forces. When the load is high in only one direction, as in spindle presses, saw threads are used. The manufacturing of trapezoidal and saw threads is typically done through the thread whirling process. The reasons for this are primarily the often small batch sizes and the favorable surface properties produced by the process.
Figure 2 illustrates the process kinematics of the whirling process. During the manufacturing process, the surface quality during whirling is primarily determined by the wear condition and positional accuracy of the cutting edges. However, monitoring of tool wear or the surface during the process is currently not possible. On one hand, the thread flank is poorly accessible due to its position, and on the other hand, the rotation of both the tool and the workpiece during the process complicates the collection of measurement data.
For these reasons, surface quality control at the machine is currently only performed on a random basis, relying on the experience and knowledge of the personnel. In order to avoid defects and ensure high quality of thread drives, a measurement system for process-parallel monitoring of the surface quality of trapezoidal threads will be developed in a joint project between Bornemann Gewindetechnik GmbH and the Institute of Manufacturing Technology and Machine Tools (IFW) at Leibniz University Hannover.
The Whirling Process
To analyze the measurement task, the kinematics of the manufacturing process must first be considered. In thread whirling, the thread surface is created by a combined movement of the whirling spindle and the workpiece (Figure 2). Typically, there are between three and eight evenly distributed cutting edges on the whirling spindle, which represent the contour of a thread pitch.
In addition to the kinematics of the manufacturing process, the later application also influences the choice of measurement method. Since the contact between the thread nut and the thread spindle ideally occurs along the entire thread flank, it is necessary to capture the entire thread flank surface in order to evaluate its quality.
The rotation of the workpiece during the process excludes both tactile and optical methods with a small measurement range for surface capture, as an additional movement of the sensor would be required to capture the entire flank. A suitable measurement method must allow for the capture of the entire thread flank in a single measurement operation. For this purpose, a monochrome industrial camera is used.
As a functional verification, threads with surfaces within the requirements (OK) and threads with surfaces outside the requirements (NOK) were captured using the camera. The images of the surfaces of a trapezoidal thread TR 65×7 and a trapezoidal thread TR 80×10 are shown in Figure 3. Due to the different surface characteristics, the reflection behavior of the surfaces varies significantly. As a result, the gray value distribution of the image changes with the surface quality.
From the schematically represented histogram, it can be seen that the amplitude of the maximum value for threads with surfaces outside the specifications (NOK) is significantly reduced compared to those with surfaces within the specifications (OK). For different thread sizes, the change in the gray value maximum is expressed to varying degrees. In order to use the altered reflection behavior for surface monitoring, it is therefore necessary to be able to define the boundary between surfaces within the specifications (OK) and surfaces outside the specifications (NOK) even with varying thread sizes.
Challenges Due to Recording Conditions
The application of image processing in the manufacturing process requires a clear and repeatable distinction between surfaces outside the specifications (NOK) and surfaces within the specifications (OK). To enable precise differentiation, it is necessary to identify and analyze the influencing factors during image capture.
The identified influencing factors on the captured image are depicted in an Ishikawa diagram in Figure 4. The target variable (red) for image processing is the roughness. The adjustable parameters (green) during image capture include, among others, exposure time and exposure direction. These parameters must be adjusted so that a clear image of the surface is produced under the given boundary conditions (orange), such as the workpiece rotation speed. Boundary conditions that cannot be compensated for by the adjustable parameters must be addressed by other measures, such as enclosing the camera to protect it from contamination.
The identified influencing factors were investigated in an experimental setup within a machine tool to allow for controlled adjustment of the parameters. A current image processing algorithm is being developed based on the gray value differences of the surfaces using the captured images.
Outlook
Based on the controlled adjustment of the parameters, a process-optimized measurement setup is currently being developed. The setup ensures that the algorithm can also be used during the whirling process. The presented system for surface monitoring is part of a system for quality monitoring during thread whirling. In the next step, the presented system, along with another system for capturing the geometric thread parameters at the whirling machine, will be put into operation. Subsequently, the development of a quality monitoring algorithm for online quality assessment based on measurement data will take place. Using the determined quality state, recommendations for action will later be derived for the operators.
Acknowledgments
The research project ”Online-Qualitätsüberwachung beim Gewindewirbeln – Quali-Wirb” is funded by the Federal Ministry for Economic Affairs and Climate Action (BMWK) as part of the Central Innovation Program for SMEs (ZIM) following a decision by the German Bundestag and is managed by the Association of Industrial Research Associations ‘Otto von Guericke’ (AiF). The IFW and the project partner Bornemann Gewindetechnik GmbH & Co. KG would like to express their gratitude for the financial support provided in this project.
B. Denkena, H. Klemme, N. Klages
Institute of Production Engineering and Machine Tools (IFW)
Leibniz University Hannover