AN INVESTIGATION OF MEASUREMENT METHODS TO TEST CUTTING FLUID EFFECTS ON TOOL WEAR IN THE PROCESSING OF ACME THREADS

Open Access
Author:
Shtub, Shai
Graduate Program:
Industrial Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
None
Committee Members:
  • Edward Demeter, Thesis Advisor
  • Dr Christopher Saldana, Thesis Advisor
Keywords:
  • cutting fluids
  • measurement methods
  • ACME threads
Abstract:
ABSTRACT A single point threading process is used to cut threads for pipe designated for the oil drilling and gas drilling industries. Thread geometries are complex and tightly toleranced. Threading insert wear is the primary reason for threads going out of tolerance. The geometric errors of each machined pipe thread are measured and tracked. If the errors reach a threshold, the tool offsets are changed in order to accommodate loss of tool geometry. If the tool wear is too great, the insert is flipped or replaced. Regardless, the machining process must be stopped for any adjustments or insert replacement to occur. This stoppage negatively affects threading process lead time, throughput and cost. Currently there is significant interest from the academic community, pipe manufacturers, cutting fluid suppliers and cutting insert suppliers in developing new technologies and processing methods that will reduce the rate of threading insert wear for pipe threading processes. PROBLEM TO BE SOLVED: In general, technology-process development projects require an assessment procedure in which alternatives are tested in a controlled, designed experiment and important dependent process variables are measured. These experiments are replicated in order to provide statistical samples and allow the use of statistical analysis tools to filter out the effects of random variation. The data is subsequently analyzed and alternatives are identified as leading to the greatest overall improvements in the process variables. A well designed experiment uses dependent process variables that are highly sensitive to changes in the process (e.g. differences in the alternatives) and/or that are subject to minimum random variation. In turn this implies that the measurement system should have sufficiently high accuracy and resolution. In general, a well designed experiment leads to the identification of superior alternatives, having the least amount of time and cost. The problem currently encountered by those seeking to improve gas pipe threading processes, is that there are currently no prescribed methods for how to carry out tool wear experiments for single point threading processes in general especially for ones to be used to machine complex threads such as those found on drilling pipe.   RESEARCH GOAL AND OBJECTIVES The goal of this research was to develop experimental procedures that would effectively identify changes in threading insert wear rate in response to changes in the threading process. The objectives of this research were: • Design a set of simple experiments that can be carried out off-line of the production equipment ordinarily used to thread drilling pipe • Identify experimental design procedures necessary to take into account the systematic variation observed in threading pipe stock • Identify process variables that show the greatest sensitivity with regard to changes in threading process • Identify measurement systems that are effective at measuring the critical process variables RESEARCH METHODLOGY The basis for this research was an experimental study used to characterize the impact of different cutting fluids on the wear rate of cutting inserts used to machine multi-vector threads into pipe made of API L80 and API P110 steel. This study was a collaboration between TMK-IPSCO (pipe manufacturer), Quaker Chemical (cutting fluid supplier), and Penn State. Despite its limited scope, it is believed that the results of this investigation can be generalized to studying the effects of all process variables on threading insert wear. RESEARCH RESULTS The threading cycle used in this investigation was a simplification of the threading process carried out by TMK-IPSCO. Specifically, a 3” long, external ACME 8P thread was machined into the straight end of a 2-3/8” OD pipe. This eliminated the need to form the ends of the pipes prior to machining. In turn, it significantly reduced the time and cost necessary to carry out the investigation. It also eliminated the need for TMK-IPSCO to divulge proprietary information regarding their thread geometry and threading process. The threading inserts used in this investigation were supplied by High Tech Tool. They are comprised of the same substrate and coating as the threading inserts supplied to TMK-IPSCO. The cutting speeds and in-feeds were varied depending on pipe grade. Each set was chosen from ranges published to be suitable for generating the ACME 8P threads into the respective pipe grades. Specific values were selected to induce a measurable amount of tool wear within a reasonable rate of time. Despite the simplifications discussed, the experimental design focused on assuring that the threading processes and thread geometry used in this investigation were close enough to those used by TMK-IPSCO to capture the critical interactions between the cutting fluid, threading insert and pipe material and therefore data and conclusions from this thesis are a good representation of the effects each cutting fluid would have if used in the TMK-IPSCO process. Three measurement processes were developed in order to estimate the effects of the different cutting fluids on the wear rate of the threading insert. Each process was used to monitor data collected from the change in different variables. The three methods are: 1. Part dimension changes as a correlation to tool wear – a method that monitors the geometric changes in the final machined part. This method was not efficient in this experiment due to inconsistency in the shape of the raw pipe. This shape affected the finished part dimensions as well as the stresses induced on the threading insert. 2. Tool weight loss method – a method that monitors weight change in the threading insert. This method resulted in inconsistent results, due to the fact that the number of data points collected using this method is smaller than in other methods. This brings forth a less accurate statistical analysis. Also, there are several different phenomena which need to be addressed in order to ensure this method is accurate. An example of this is built up edge which would directly affect the weight of the threading insert and skew the results. 3. Image processing tool wear measurement system – this method monitors the changes taking place with the threading insert itself. Eleven wear variables were developed and monitored using this measurement method. The results found using this method showed statistical differences between the different cutting fluids. Some of the developed variables proved to be more valuable in the analysis than others due to the fact that the results shown were different using the image processing equipment available for analysis. The variables which have not resulted in good results are Rake Face – Tool Wear – Crater Area and Rake Face Leading Edge – Tool Wear – Wear Width. Overall, it is recommended to monitor all the variables in conjunction in order to get an overall final picture which is clearer due to the fact that different sides and aspects of the insert are being monitored. A major factor has affected the entire experiment development process. Egg shaped pipe directly affects the wear rate of the threading inserts and the overall efficiency of the process. The unique shape induces higher stresses than designed and therefore cause the threads to be out of specifications and the inserts to wear faster than designed. The discovery of the inconsistency in the pipe shape directly affected the data collected in the first developed measurement method. Monitoring changes of the machined part dimensions is affected by the lack of consistent shape of the raw material. In order to overcome the effects of the raw material shape, a randomization system was developed which takes the shape of a specific pipe out of the overall data. Randomizing consisted of using each raw pipe with all different cutting fluids. This step ensures that any original shape the raw pipe has, affects the results of all cutting fluids and all measurements and therefore the egg shape does not affect the conclusions of the thesis. Based on the experiments conducted and data collected, it is concluded that the Image Processing Tool Wear Measurement System is the most efficient and accurate method of identifying the wear rate induced on threading inserts.