The Prediction and Effects of Flute Breakage on the End Milling Process

  Michael Scott Wagner | UIUC | MS | 1988

ABSTRACT:

In the search to improve productivity, manufacturing industries have turned toward automation as a means to achieve this important goal. The push towards factory automation has manifested itself in several ways, but all of them do have one thing in common. Automation seeks to remove the operator from performing dangerous, repetitive, or exceedingly difficult tasks, essentially removing him from performing any tasks. For example, numerical control was developed in an effort to achieve complicated tool paths that would not be possible to achieve manually. Industrial robots are well suited for the hazardous environment of automobile painting where human operators would be exposed to dangerous fumes. It would also seem reasonable to save a man for more creative tasks and let an automated guided vehicle continually supply a process with material. The implementation of automated manufacturing systems, as well as the side effects of their implementation, introduce new challenges to industry.
 

Once a manufacturing system is automated, i.e. the operator is removed, the complexity of the entire system increases with the addition of new elements. For instance, consider the process of automated workpiece fixturing. Fixtures must be designed to allow for quick, easy, and very accurate workpiece positioning. In an automated environment, no longer is the operator available to make minor adjustments in fixturing. Next, consider an automatic tool changing system. If a specific
tool is required to perform an operation, and an operator knows what it is, the probability of the wrong tool selection is low. If, however, the system is unmanned, and some environmental disturbance, such as a voltage surge, interferes with the electronic selection of the proper tool, the entire operation may fail. Finally, consider a problem that may arise in an automated machining process. Recently, there has been quite a bit of interest in improving the performance of unmanned machining centers. There are obviously many actions occurring at such a center, but the one of principal importance involves the actual metal cutting operation. As with any metal cutting process, tool wear and/or severe cutting conditions may cause catastrophic tool failure, a situation which is detrimental to both the workpiece and to the machine. In an automated system such a failure may never be detected.
 

The capability to sense and process information concerning an automated system is readily available. The problems mentioned above may be solved with the aid of sensors transmitting information to computer controllers which act on the information. For example, the position of a workpiece can be determined with low powered lasers or ultra-sonic transducers.. If a workpiece is incorrectly fixtured and positioned, it could be sensed, and action taken to correct its location. The insurance of the correct tool may be obtained through the use of optically sensed bar coding. The problem of tool breakage is not quite as simple to solve. Sensor information output by the process is not difficult to obtain. Presently, systems to monitor the cutting forces produced by metal cutting operations are being used. The problem is what to do with the information.
 

Useful tools in solving the problem of what to do with information concerning the condition of the process are models that simulate process performance. If a model was available, for example, that could simulate the cutting force system produced by the end milling process, then it could be used to predict the cutting forces generated by a broken tool. The model predictions could then be compared to the actual cutting force signal produced by the process in an effort to monitor process condition on-line. Deviations of the actual and predicted cutting forces indicate a change in tool condition which would trigger an automated system to take corrective action. Another use of such a process simulation model is the off-line selection of cutting conditions. Cutting conditions may be selected through repeated model simulations used in conjunction with some optimization technique. In other terms, what input conditions will avoid a tool failure? If this function of the model performs well, possible problem situations may never arise.
 

In order to utilize a model for the prediction, of the cutting force system and tool breakage in end milling, model inputs and outputs must be known. Model inputs would include the complete specification of the tool used as well as the current cutting conditions. Model outputs include the prediction of catastrophic tool failure as well as the regions of the tool that break. Also, the cutting force system required for on-line comparison with actual forces produced by the operation must be output. If such a model was developed, detecting on-line tool failure through force signal monitoring would be one step closer to attainable, and another type of automated system problem would be closer to being solved.
 

In this thesis a model for the prediction and the effects of flute breakage on the end milling process is developed. Chapter 2 is a detailed review of the literature describing a model that accurately predicts the cutting force system of the end milling process. This model serves as a building block to Chapter 3, in which the steps involved in developing a flute breakage model and its effects on the cutting force system are described. Chapter 4 reports the results of a series of experiments performed to condition the model and verify its accuracy. In this chapter the cutting forces are compared for model predicted and actual data. Chapter 5 provides the results of a series of simulations created through repeated use of the model, and Chapter 6 draws conclusions based on the results of both chapters 4 and 5. Chapter 6 also provides explanation of the results, and suggests topics for which a more intensive study is necessary.
 

It is important to understand where this work stands in relation to the broader topic of detecting tool failures. This work lays the groundwork for a more complete understanding of the cutting force system in end milling, and how it is impacted by flute breakage.

If you have any comments or suggestions please e-mail jwsuther@mtu.edu.