The Relationship Between the Workpiece Extension Length/Diameter Ratio and Surface Roughness in Turning Applications by Mr

Volume 23, Number 2 - April 2007 through June 2007

The Relationship Between the Workpiece Extension Length/Diameter Ratio and Surface Roughness in Turning Applications By Mr. John Cooper and Dr. Bruce DeRuntz

Peer-Refereed Article

KEYWORD SEARCH

Machine Tools Manufacturing Materials & Processes Metrology Quality Control Research

The Official Electronic Publication of the National Association of Industrial Technology • www.nait.org © 2007 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org The Relationship Between the Workpiece Extension Length/Diameter Ratio and Surface Roughness in Turning Applications Mr. John J. Cooper Jr. is currently working on his Ph.D. in Workforce Education and Development at By Mr. John Cooper and Dr. Bruce DeRuntz Southern Illinois University - Carbondale. He is a teaching assistant in the Department of Technol- ogy and also serves as an adjunct instructor for The advancement of the lathe and sub- eter, and surface roughness of bar stock the Industrial Technology Outreach Program. His research interests include the implementation and sequent modern technologies was made in both supported and unsupported manufacturing areas of lean manufacturing. possible through research leading to the turning operations in an attempt to re- development of optimization tables that duce setup waste in turning operations. list specific feed rates, spindle speeds, and depths of cut for different materi- Literature Review als. These tables are the standard used Since the invention of the modern in industry as a source of reference, engine lathe, researchers have been when making a change from one job to contributing to the body of knowledge another where the machining param- that will optimize the lathe’s productiv- eters of each may be quite different. ity and quality (i.e. surface roughness). The time, material, and tooling costs Surface roughness plays an important associated with the experimental steps role in product quality and is an espe- needed to find the appropriate machin- cially important design characteristic in ing parameters for each new job are many products that are subject to preci- eliminated, giving the company the sion fits, fastener holes, fatigue loads Dr. Bruce DeRuntz is an Associate Professor in advantage of a reduction in setup costs and aesthetic requirements (Wang, the Department of Technology at Southern Illinois University Carbondale. He provides Six Sigma and improved product quality. 2001, Feng & Wang, 2002, and Kalpak- consulting to Fortune 500 companies and teaches jian et al. 2006). Dimensional accuracy courses in Industrial Technology. His experience includes 12 years in the automotive and defense While there are many machining also, together surface roughness greatly industries as a quality engineer and manager. optimization parameters that have been affect useful part life, especially in Bruce holds a Ph.D. in Workforce Education and Development, a M.S. in Manufacturing Systems developed and put into tables, an area cases in which the components will be and is a CSIT and CQE. He is a Certified Quality that has been overlooked is that of cor- in moving contact with other elements Engineer, Certified Senior Industrial Technologist, and ISO 9000 Auditor. relation between the unsupported work- or materials. piece extension length and the resultant surface roughness. The premise of this Recent research that has identified the research is that the use of a tailstock in relationship between surface rough- lathe work for workpiece support may ness and lesser understood turning sometimes be unnecessary. This idea parameters include; surface roughness is investigated to find out if there are versus lubrication and bed material instances where an unsupported work- (Bruni, Forcellese, Gabrielli & Simon- piece can be machined with equivalent cini, 2006), surface roughness versus results to a supported workpiece in tool wear (Pavel, Marinescu, Deis & terms of surface roughness; so the time, Pillar, 2005), surface roughness versus labor, and materials used to center drill burnishing feed rate, force, and speed the workpiece and setup the tailstock (El-Axir & Ibrahim, 2005), and surface are not used needlessly. Therefore, the roughness versus dry turning (Thomas focus of this study is to extend previ- & Beauchamp, 2003). Kalpakjian et al. ously established research on effects of (2006) illustrates a regressing relation- workpiece elastic deformations (Bena- ship between dimensional tolerance and rdos, Mosialos and Vosniakos, 2006) workpiece length, but does not ad- and examine the relationship that exists equately differentiate the surface rough- between the length, at a specific diam- ness correlation between a supported

2 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org and unsupported workpiece in relation The third interview was with a setup process. The experiments, involving to its length. man in a job shop, whose area of seven different workpiece lengths and expertise was lathe work. He gave an only on the diameter of 1”, focused on The practice of choosing appropri- estimate of 1:4, which he felt would a straight cut of a half inch that was ate process parameters can be quite be appropriate for a straight cut, using made from the end of each workpiece. difficult. To make this determination low carbon steel. When asked about The resultant surface roughness was currently requires time consuming trial the unsupported workpiece extension measured for each of the test incre- and error experimentation which is length/diameter ratio reference chart, ments. The surface roughness values costly in time and material resources. he said he felt there was a definite need reflected the difference in surface finish The solution to this dilemma, according for this type of reference in industry as the workpiece extension length to to Wang (2001), is to develop a chart to because of the amount of trial and error diameter ratio changed. Since previ- serve as a quick reference for industry necessary to optimize the machining ously published research on this spe- to determine pre-chatter conditions, so process (Western Machine Tool & Die, cific subject is very limited, the results poor surface roughness can be avoided. personal communication, December 20, of the study were based on empirical Cutting condition values could be put 2005). evidence obtained during the course of in a formula to evaluate whether the the experiments performed during this cutting conditions would produce a The fourth interview was with an exec- research. This research project was de- chatter-free workpiece. Work in this utive of manufacturing operations. He signed to address the issue in a general area has already begun with the devel- was also unaware of such a formula and manner, focusing on straight turning opment of a knowledge-based system stated “The lack of a formal procedure of low carbon steel using a finishing for the prediction of surface roughness was a problem, especially for the new cut. The experiment was carried out in in turning process (Abburi & Dixit, machinists who had no past experience a setting which closely replicated an 2006). This would eliminate guesswork with which to base a judgment call on.” actual manufacturing facility. The re- by optimizing cutting parameters and He felt there was a very real need to es- search laboratory used is situated in the controlling the quality required for tablish some form of reference material College of Engineering, Department of desired surface finishes. relevant to the unsupported workpiece Technology, Manufacturing Processes extension length/diameter ratio (Kelm Laboratory, at Southern Illinois Uni- This study attempted unsuccessfully to Accubar, personal communication, De- versity Carbondale. The SIUC labora- find existing literary evidence of unsup- cember 28, 2005). Although there is no tory equipment chosen for use in the ported workpiece extension length/di- clear agreement on what an appropriate experimental process are all commonly ameter ratio studies, where the surface ratio should be, there is consensus as to available at industrial manufacturing roughness values were equivalent to the need for such a standard. suppliers. They include the following: those of supported turning. In addition • Bridgeport EZPATH SD CNC Lathe to the past research conducted in this Purpose of the Study (ROHM No. P5354 three jaw chuck) area, the researcher sought out experts The purpose of this study is to identify • Brown & Sharpe Pocket Surf Surface explicitly involved with work utilizing a region along the length of the work- Roughness Tester, Range: Ra - 1µ” to lathe applications. An interview with the piece where the surface roughness val- 250µ” first expert, who specialized in tooling ues of the unsupported workpiece were • Brooks Rockwell Hardness Tester applications, revealed no knowledge statistically comparable to those of the • Kennametal Tool Holder: Part No. of previous formulas for determining supported workpiece. A reference chart MCLNR-124B (Right Hand Tool the unsupported workpiece extension was then developed to illustrate where Holder – Combination lock pin and length/diameter ratio in turning op- an operator could forego the use of the top clamp) erations. He cited a rule of thumb of tailstock during setups. This chart, like • Kennametal Coated Carbide Inserts: “3-5:1” (3 to 5 inches length in ratio to the commonly used feed rate, spindle Part No. CNMG 432 FN, Grade every 1 inch in workpiece diameter) for speed, and depth of cut charts, would 9125 a straight cut (Kennametal Inc., personal serve as a reference to operators who • Kool Mist Formula No. 78 (Synthet- communication, October 10, 2005). routinely perform lathe setups to elimi- ic Coolant Concentrate) nate uncertainty in the setup procedure, • 1018 Cold Rolled, Low Carbon The second interview was with a highly minimize required setup time, maxi- Steel: 1” diameter experienced engineer who was involved mize setup and operational efficiency, in the design of “Steady Rests”. A rule and reduce overall operating costs. Due to the differing opinions among of thumb that specified approximately industry professionals about a common “7-10:1” was given as their guideline Methodology workpiece extension length/diameter for an unsupported workpiece exten- An experimental design was created to ratio (Kennametal Inc., 2005; Aro- sion length/diameter ratio (Arobotech test the resultant surface roughness of botech Systems, Inc., 2005; Western Systems, Inc., personal communication, varying workpiece extension length/ Machine Tool & Die, 2005; and Accu- December 16, 2005). diameter ratios in a straight turning bar, 2005), various workpiece exten-

3 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org sion lengths that would encompass all currently examined industry guidelines from one extreme to the other, and beyond, were tested. This research challenged the traditionally accepted boundaries by exceeding the previous maximum workpiece extension length/ diameter ratio of 10:1 to nearly double that figure, by using a ratio of up to 19:1 in the experiments. The reason for using this approach was to find the safe limit of process functionality for a given workpiece diameter. However, attempts to effectively machine an unsupported workpiece in the range of 17:1 to 19:1 resulted in severe workpiece deflection which rendered the cutting action of the insert ineffective, leaving a surface of pronounced chatter. Attempts to machine workpiece material in the range of 7:1 to 15:1, resulted Figure 1. Illustration of workpiece extension length/diameter ratios in fractured cutting inserts. Finally, workpiece surface roughness produced in the range from 4.5:1 to 6.5 exhib- from a group of supported workpieces and also because it would be beyond ited extreme chatter that at one point as well. The experiment was repeated the scope of this research to involve was beyond the operating range of the as previously stated, with the exception all materials at this level. The mate- surface roughness tester. Since the early of the use of a tailstock as a means of rial was a standard 1” diameter cold experimental trials illustrated a lack of workpiece support. The use of a tail- rolled unmachined bar. The bar stock machinability in these workpiece exten- stock with a live center for workpiece consisted of 20 individual pieces, each sion ranges, this research focused on support is the typical method in which being 6” in length. The additional two what was determined to be the effective turning operations are performed. inches in length allowed for chucking cutting range of this particular material, Therefore, this method was used to es- of the bar stock. The seven different bar of 1:1 to 4:1. tablish a control condition with which lengths, as illustrated in Figure 1, were to determine the statistical significance tested by machining 10 pieces at each Therefore, a series of workpiece exten- of the experimental results. Test values length unsupported as the treatment and sion lengths which include the follow- for both the supported and unsupported 10 pieces at each length supported by ing 1”, 1 ½”, 2”, 2 ½”, 3”, 3 ½”, and 4” trials were evaluated for statistical means of a tailstock for establishing the served as the treatments or independent significance. control condition. The control condi- variables. In Figure 1, an unsupported tion (supported) surface roughness workpiece is shown and the various Procedure represents a comparison against which workpiece lengths and ratios are illustrat- An approach that focused on common the experimental (unsupported) surface ed by the lines spaced every half inch. applications was used. For the initial roughness was compared. research, it was important to select a The experiment consisted of the previ- material that would be applicable to a Operating Parameters ously mentioned group of workpiece broad range of manufacturing applica- Next, the operating parameters for the lengths that were chucked in the lathe tions so as to have the greatest benefit lathe and tooling were identified using and turned one-half inch from the to industry. A description of the steps the Kennametal Lathe Tooling Cata- unsupported bar end using a straight necessary to perform the study was log 1010 (2001). This step referenced cut. Also noted is the uniform length of developed. They are as follows: existing knowledge of turning opera- workpiece material of two inches that tions to optimize the cutting process, was clamped in the chuck for each of Material Selection and Preparation with guidelines specific to this material the seven trials. As noted, 1018 cold rolled, low carbon and tooling application. The guidelines steel was used for the workpiece mate- listed were as follows: In order to verify differences in surface rial. This particular material, while not The insert chosen for this research was roughness measurements between the representative of all workpiece materi- the CNMG432FN Grade KC9125. It supported and unsupported workpieces, als, was chosen specifically because is a TiN coated carbide insert that is it was necessary to examine trial results of its wide-spread use in industry, designed for finishing cuts and has a

4 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org negative rake geometry. Its shape is been positioned to act as a stop to also Brown & Sharp test equipment required diamond with 800 nose angle and a 70 allow exactly two inches of the bar to no calibration before each measurement. relief angle. Using the Kennametal In- be chucked. The lathe was then pro- However, to eliminate human error as a sert Selection Guide, which illustrates grammed, using the aforementioned factor, the workpiece and tester were set how to choose the correct machining machining parameters, to machine the up in a manner that standardized the pro- specifications for a given application, last half inch of material from the end cess, allowing for a hands-off technique. the chosen parameters were depth of of the bar. All ten bars in the respective The measuring process for surface cut 0.035 inch, with a feed rate of 0.10 set being processed (experimental or roughness consisted of first placing each inch per revolution. These two criteri- control) were machined consecutively workpiece sample in a vise. Then, using ons were chosen because they represent at the particular bar length being evalu- the B&S Pocket Surf surface roughness average insert usage as represented in ated (ex: 4”; 3.5”; 3”; 2.5”; 2”; 1.5”; tester which was attached to a base unit, the guide. However, for this particular or 1”) and then measured for surface taking three separate measurements vi- research application, the surface feet roughness. Next, the machined sections sually spaced approximately 1200 apart per minute rate, or cutting speed, as of each bar in the set were cut off on around the perimeter of the bar on the listed in the Kennametal guide proved the lathe and the experiment repeated machined surface. This was to insure ac- to be too high, which resulted in ex- at the new bar length, one-half inch curacy of the readings by averaging the treme chatter and tool breakage. The shorter than before. This procedure pro- natural variations of the surface rough- Machinery’s Handbook 25 (1996) was vided 10 samples at each bar length and ness values. also consulted for appropriate machin- was repeated until all seven of the pre- ing parameters with regard to surface scribed lengths had been machined and Findings feet per minute. The data from their recorded for each of the respective sets. The surface roughness values for each table was similar to that of Kennametal. The hardness of each of the 20 bars was of the 10 length values for both sets However, the text stated “Although the also checked, on an unmachined sur- were arranged and are reflected graphi- accompanying tables provide recom- face. The measurements were obtained cally in Figure 2. mended cutting speeds and feeds for by using a Brooks Hardness Tester. The many materials, experience in ma- Rockwell “B” scale was used to per- The differences between average surface chining a certain material may form form the tests. The tests revealed a very roughness values for the supported the best basis for adjusting the given uniform hardness, ranging only from 93 and unsupported workpiece treatments cutting speeds to a particular job” to 95, throughout all of the specimens. were then statistically analyzed using a (Industrial Press, Inc. 1996, p. 977). This eliminated inconsistency of hard- t-test. Using a critical α level of .05, a So, by evaluating the surface roughness ness as a factor related to the surface Two-Sample t-test (Assuming Unequal characteristics to find a point at which roughness results. Variances) statistical evaluation was tool breakage had been eliminated and performed to test for a significance level the workpiece was free of chatter, the Measurement process between the supported and unsupported maximum surface feet per minute rate To take measurements in a consistent workpiece experimental results. Addi- of 327 was determined to be the best manner that eliminated machine and tionally, the final analysis included the rate for our experiment. Therefore, the human error, the measuring process use of descriptive statistics for construc- surface feet per minute rate was revised was examined for inconsistencies. The tion of a graph, for a visual representa- for use in this research. The machine RPM was 1250. Figure 2. Comparison of Surface Roughness Mean Values Material Processing There were two ten bar sets. One set was labeled E to represent the experimental set and the other labeled C for the control set. One new insert was used to process the experimental set and one new insert was used to process the control set. This was done in order to eliminate tool wear as a factor in the final surface roughness analysis. The unsupported bars were chucked precisely two inches from the labeled end by aligning the chuck jaw with a mark on the workpiece. The supported bars were placed securely by hand against the tailstock, which had

5 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org tion of the results. The results from this Table 1. t-Test: Two-Sample Assuming Unequal Variances test are summarized in Table 1. (Note: Su = Supported; Un = Unsupported) Analysis Results of Test Data

Two particular trends appear in the Pair Sample t-Statistic T Critical two-tail Significance graph, one expected and the other Pair 1 1” Su/Un -3.046 2.119 Significant rather unexpected. As expected, as the workpiece extension length gets pro- Pair 2 1.5” Su/Un 0.077 2.144 Insignificant gressively longer, the mean chart values Pair 3 2” Su/Un 1.421 2.109 Insignificant for the unsupported workpiece begin to move in an upward trajectory, sug- Pair 4 2.5” Su/Un 6.065 2.178 Significant gesting an increasing instability within Pair 5 3” Su/Un 13.557 2.200 Significant the system as expected. Surprisingly however, the pattern suggests that at Pair 6 3.5” Su/Un 4.010 2.100 Significant distances of less than 1.5”, an undesir- Pair 7 4” Su/Un 5.831 2.200 Significant able element leading to system instability has been introduced to the supported workpiece. This indicates stability tion. The unacceptable regions rep- explore the following: issues, possibly linked to the cutting resent areas that will produce surface • Different workpiece diameters tool’s proximity with the lathe chuck. roughness values that are inferior to • Different workpiece materials typical supported turning operations. • Larger sample sizes Conclusions Table 2 illustrates an abbreviated • Further detailed examination at turn- Although this plan looks only at the Unsupported Workpiece Comparison ing lengths from 1-1.5” relationship of one diameter of bar Chart. stock in varying lengths, the results Using the same workpiece material, validate the original premise of equal Implications the chart can be broadened to include a surface roughness in some regions of The method used in this work can be wide range of diameters and lengths ap- workpiece length when comparing effectively expanded, through addition- plicable to this particular material. An supported and unsupported turning and al research, over a wide range of work- example of one possible configuration have been used to develop a reference piece diameters, lengths, and materials. is the Future Unsupported Workpiece chart that defines the boundary in ques- Comprehensive data can be compiled to Reference Chart, illustrated in Table 3. tion. The results illustrated in Table develop a series of charts which can be 1, show statistically significant differ- used to eliminate an element of setup Further testing, beginning with the ences in surface roughness at work- waste related to turning operations 2” and 3” diameter intervals, may piece extension length/diameter ratios within the manufacturing industry. be enough to reveal a pattern where of 1:1, 2.5:1, 3:1, 3.5:1, and 4:1. This prediction becomes possible through leaves the ratios of 1.5:1 to 2:1 as the Recommendations means of a model which uses an algo- region where workpiece support played This study has addressed an area that rithm to create a viable chart without an insignificant role in surface rough- has been largely neglected in the past. physically performing “hands-on” test- ness for workpieces of 1” in diameter. Now that the initial research has now ing at each individual increment. However, even though the 1:1 ratio was been conducted for this specific topic, statistically significant, the unsupported surface roughness equivalencies of the References method produced superior surface supported vs unsupported workpiece, Abburi, N.R. and Dixit, U.S. (2006). roughness values at this length. So, the a methodology is available to those A knowledge-based system for the applicable range of workpiece exten- who wish to continue with this effort. prediction of surface roughness in sion lengths can be safely extended Future work should include studies that turning process, Robotics and Com- from 1:1 to 2:1 without loss of product surface finish quality. Table 2. Unsupported Workpiece Comparison Table The development of the reference chart included regions that were labeled to Range of Acceptable / Unacceptable Regions for: 1018 Cold Rolled Steel represent workpiece extension length/ diameter ratios which are acceptable Diameter Lengths and unacceptable respectively. The acceptable regions represent areas able Acceptable Unacceptable to produce surface roughness values 1” 1” - 1.5” - 2” 2.5” - 3” - 3.5” - 4” equivalent to a supported turning opera-

6 Journal of Industrial Technology • Volume 23, Number 2 • April 2007 through June 2007 • www.nait.org

Table 3. Future Unsupported Workpiece Reference Table for Surface Roughness Prediction in Finish Turning. The International Range of Acceptable / Unacceptable Regions Journal of Advanced Manufactur- ing Technology, 20, 348-356. Diameter Lengths Industrial Press, Inc. (1996). Machin- 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 ery’s Handbook 25, New York: Author .5 Kalpakjian, S., and Schmid, S. (2006). .75 Manufacturing Engineering And Technology, 5th, PEARSON Pren- 1.0 A A A U U U U U U U U tice Hall. 1.25 Kennametal Lathe Tooling Catalog 1.5 1010. (2001). Manufacturing Tech- nology, 20, 348-356. 1.75 Pavel, R., Marinescu, I., Deis, M., and 2.0 Pillar, J. (2005). Effect of tool wear on surface finish for a case of con- 2.25 tinuous and interrupted hard turning, 2.5 Journal of Materials Processing 2.75 Technology, 170, 1-2, December, 341-349 3.0 Thomas, M., and Beauchamp, Y. (2003). Statistical investigation of modal parameters of cutting tools in puter-Integrated Manufacturing, 22, material on the workpart surface fin- dry turning, International Journal 4, August, 363-372 ish and tool wear in finish turning of of Machine Tools and Manufacture, Benardos, P.G., Mosialos, S. and AISI 420B. International Journal Volume 43, 11, September, 1093- Vosniakos, G.-C. (2006). Predic- of Machine Tools and Manufacture, 1106 tion of workpiece elastic deflections 46, 12-13, October, 1547-1554 Wang, Z. (2001). Chatter Analysis of under cutting forces in turning, El-Axir, M.H. and Ibrahim, A.A. Machine Tool Systems in Turning Robotics and Computer-Integrated (2005). Some surface characteristics Processes. Ph.D. Thesis, National Manufacturing, Volume 22, 5-6, due to center rest ball burnishing, Library of Canada, Acquisitions and October-December, 505-514 Journal of Materials Processing Bibliographic Services, 395 Wel- Bruni, C., Forcellese, A., Gabrielli, F., Technology, 167, 1, August 2005, lington Street, Ottawa, ON K1A and Simoncini, M. (2006). Effect of 47-53 ON4 Canada the lubrication-cooling technique, Feng, C. X. (Jack) and Wang X. (2003). insert technology and machine bed Development of Empirical Models