A Force-Based Flute Breakage Detection Algorithm for a Peripheral End Milling Process
 
 

Daniel Joseph O'Brien | UIUC | MS | 1990

ABSTRACT:

Commonly encountered faults in the end milling process include accelerated tool wear, tool chipping, failure of the cutter, and excessive runout. Although all are serious problems, catastrophic tool failure poses the greatest economic threat as well the greatest hindrance to automation of the process.This thesis presents a force-based flute breakage detection algorithm for a peripheral down milling operation.

End Milling cutters fail in primarily two ways: shank breakage and flute breakage. Shank breakage is defined as failure of the cutter across its entire cross sectionand is the most severe form of tool failure. Shank breakage is readily evident from the cutting force signature, after the failure, as the forces drop to a near zero level. Shank breakage can result in a marred or gouged workpiece and is hazardous to the machine tool, surrounding equipment, and personnel. Flute breakage, failure of a portion of the cutting edge, may be more subtle in nature and might go undetected for a period of time. With a portion of the flute missing, the peak cutting forces are increased, which may lead to additional flute fractures and possibly to shank breakage. Often times, flute breakage results in tool material lodging between the workpiece and the cutter which produces extremely high forces. Undetected flute breakage not only increases the potential for further tool damage but also affects the surface error/finish of the machined part. Chipping may be considered to be a minor form of flute breakage in which a portion of the cutting edge breaks off, yet, unlike flute breakage, this cutting edge still maintains the ability to cut material. Like flute breakage, chipping can result in increased forces, accelerated tool wear, degraded surface error/finish, and can lead to additional fractures of the tool.

Catastrophic tool failure in the end milling process can prove to be extremely costly in terms of damage to the workpiece, the machine tool itself, and surrounding equipment and additionally poses a significant safety hazrd to operating personnel. Traditionally, to combat the threat of tool fracture and to secure conformance t o design specifications, overly conservative cutting parameters have been chosen yielding suboptimal material removal rates, and consequently, low production output. In the move from manual control of machining processes toward fully automated, unmanned machining centers, more sophisticated than currently available methods of monitoring and controlling processes are required. The successful automation of machining processes require effective and reliable monitoring and control techniques that: 1. ensure high quality parts with respect to dimensional tolerances and surface texture, 2. allow for high material removal rates, and 3. provide the ability to safeguard the machining system (machine tool, tools, and workpiece).

Online identification and timely response to flute or shank breakage in the end milling process addresses all of the requirements enumerated above. The general requirements for a practical, on-line tool breakage detection scheme in ther end milling process include: 1. efficient algorithms providing practical real time application, 2. early detection, 3. detection schemes independent of the cutting conditions, workpiece materials, and tool materials, 4. insensitivity to changes in the cutting conditions such as those that occur upon entry and exit, 5. low incidence of false alarms, i.e. indications of tool failures, when none have occured, and 6. reliable detection of tool failures, i.e. low incidence of missed failure events.

This thesis presents a force-based flute breakage detection algorithmcapable of rapidly and reliably detecting flute breakage. The flute breakage detection scheme employs efficient algorithms providing practical real-time application and is robust to the effects of cutter runout, noise in the force signal, and changing cutting conditions such as those that occur upon entry and exit. In addition, the algorithm can be apploied independent of the cutting conditions and workpiece/tool materials and hence requires no trial or "learning" cuts to tune the algorithm. The algorithm is developed and verified via physical experimentation for a peripheral down milling operation although it is expected that algorithm can be extended to other types of cuts, e.g. up milling cuts or cornering cuts.

Chapter Two of this thesis presents a review of the currently available literature on cutter breakage strategies for machining processes. The review focuses upon acceleration, acoustic emission, and force-based monitoring techniques for turning, drilling, and milling operations. This review will demonstrate that although a degree of success has been achieved in the case of single-point cutting operations (e.g. turning), the more difficult problem of cutter breakage detection in multi-point cutting operations (e.g. face/end milling) still remains largely unsolved.

Chapter Three presents a review of a computer-based mechanistic model capable of characterizing the effects of the cutting conditions, entry/exit transients, and flute breakage itself on the cutting force system.

In Chapter Four, the experimental work performed to provide verification for the flute breakage detection algorithm is documented. In total, four sets of tests were conducted in order to: 1. verify the mechanistic cutting force model, thus allowing for use of the model in algorithm refinement and veri fication, and 2. provide actual cutting force data with which the flute breakage detection algorithm could be tested

Chapter Five provides verification of the cutting force model presented in Chapter 3 via physical experimentation. Three sets of tests demonstrate the ability of the cutting force model to predict cutting forces during steady state cutting, upon entry and exit, and under the influence of flute breakage. For the case of flute breakage, tests were performed with a portion of a single flute ground down so as to effect a similar tool geometry to that of a cutter with a broken flute.

Chapter Six presents the flute breakage detection algorithm with the aid of a cutting force signal predicted by the mechanistic model. The algorithm is based on the periodicity of the steady state cutting force signal and utilizes a two- step signal processing approach to filter out the dynamic components of the raw force signal. Flute breakage is detected when an exceedingly large force, generated at a specific angular position of the cutter, causes the error series to exceed the detection limits. Graphs are used through out the development to depict, step by step, the various stages of signal processing. The hardware and computational requirements of the algorithm are also discussed.

Chapter Seven provides verification of the flute breakage detection algorithm. The algorithm is demonstrated to be effective in detecting flute breakage through further model simulations and, additionally, through a series of down milling exoeriments including three actual flute breakages: a "natural" flute breakage occuring during steady state cutting, a breakage on entry, and a breakage on exit. The algorithm is also demonstrated to be insensitive to the effects of cutter runout, entry/exit cuts, and noise in the force signal through application of the algorithm to simulated and actual cutting force data.

Chapter Eight summarizes the results of this work and provides recommendations for future work.

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