4、（许 娜）_弯扭联合疲劳试验机总体设计_Failure characteristics of some metals subjected to torsion only, and when subjected to combined torsion and compression.docx

《4、（许 娜）_弯扭联合疲劳试验机总体设计_Failure characteristics of some metals subjected to torsion only, and when subjected to combined torsion and compression.docx》由会员分享，可在线阅读，更多相关《4、（许 娜）_弯扭联合疲劳试验机总体设计_Failure characteristics of some metals subjected to torsion only, and when subjected to combined torsion and compression.docx（17页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、INT. J. PROD. RES., 1974, VOL. 12, NO. 2, 179-194 Failure characteristics of some metals subjected to torsion only, and when subjected to combined torsion and compression E. A. PEIRCE* The contents of this paper show that when u metal torsion specimen is subjected to an axially applied compressive s

2、tress, then the resulting angle of twist recorded up to fracture will be greater than if tested under plain torsion conditions. Increasing this longitudinal compression results in a farther increase in the fracture strain. The introduction of CCL4 to the gauge length whilst under test causes a redac

3、tion in the strain to failure. The presence of a longitudinal stress or CCI or a combination of both has little or no effect on tlie fracture strength of the tested materials. Tested tnaterials include pure, two and poly phase and free-machining materials. Introduction This paper re3 rts on the firs

4、t stage of a research intended to study the possible relationships between the failure characteristics of metal specimens tested under known and controlled conditions of loading, and those encountered when machining metals within a specified range of cutting conditions. Many attempts have already be

5、en made to use existing data for this purpose, particularly that information obtained by means of standard materials testing methods or from variations of these, so that more exact explanations of the formation of the chip could be achieved. The results of such work have not been particularly succes

6、sful, partly because the test conditions have not simulated the loading conditions encountered when a chip is being formed and partly because the theories and analyses of the chip formation and forces are not yet complete. There are many difficulties in formulating a theory of chip formation which i

7、s complete, not least those caused by the wide range of variables which create significant differences when machining under modern conditions of high cutting speeds, feed ratest tool angles and tool materials, etc. In view of such difficulties the test conditions chosen and reported upon were design

8、ed to provide a satisfactory basis so that results obtained from a materials test could be satisfactorily compared with those from an appropriate range of cutting conditions. These test conditions approximately simulated the type of loading which established theories of metal cutting predict for low

9、 cutting speeds, where, for example, the relatively high cutting temperatures and strain rates are avoided. The latter could not be reproduced and recorded without excessive difficulty and were beyond the scope of the present work. These simplifications permitted corresponding measures to be adopted

10、 for the subsequent machining tests. The present study is concerned with reporting the work carried out on tubular metal specimens which were subjected to torsion in one series of tests, # Department of Production Technology, Brunei University* Uxbridge, England. Presented at the 2nd Intornational C

11、onference on Production Research (Copenhagen) August 1973. Published by Taylor & Francis Ltd., 10-14 Macklin Street, London WC2B 5NF. 180 E. A. Peirce and to a combination of torsion and specified levels of longitudinal compression in other tests. A further feature of the present work is that two en

12、vironments surrounding the tubular specimen were used, namely ambient air in one case and a combination of carbontetrachloride and ambient air in the other. The reason for the choice for the second environment is associated with the discovery during research in recent years that at low cutting speed

13、s the use of carbontetrachloride causes a significant reduction in the measured cutting forces. By testing the tubular specimens in this environment, as well as in air, a further confirmation of the accuracy of the simulation and comparative procedure might or might not be established. The testing a

14、pparatus Figure 1 (Appendix 1) illustrates the prototype testing equipment mounted upon a Churchill/Redman lathe. This apparatus was designed to accommodate the following requirements : (a) measurement of shear strain (b) measurement of applied torque (c) the provision of a torque/twist curve (d) th

15、e apparatus to be capable of transmitting an applied axial compressive load. The central shaft of the testing apparatus was mounted in self-aligning antifriction bearings. This arrangement minimized rotational resistance of the shaft ; also the shaft was free to move axially. This latter point is im

16、portant especially under the compressive load condition. Strain measuring device Figure 1 shows a large diameter plate made from thin aluminium plate provided with 125 equally spaced in. diameter holes drilled around the periphery. A six volt light source was positioned on one side of the disc whils

17、t on the other side a Milliard O.C.P. 71 phototransistor was mounted. This arrangement provided a direct indication of the angle of twist to one hole, i.e. 2*88. Torque measuring system This was accomplished by making use of a torque arm on which strain gauges were suitably mounted. Specimen geometr

18、y A hollow specimen similar to that employed by Hodierne (1962) was used for all tests. The major modification to Hodierne?s test piece being associated with the method of locating the specimen within the apparatus jaws and to the configuration of the gauge length. With reference to this latter poin

19、t, parallel shoulders were provided thus facilitating the formation of a radiused entry to the gauge portion. A series of tests were carried out with specimens of varying radii. However, the gauge length was maintained constant (straight portion, kept at 0-125 in. for all tests). The radii considere

20、d were 0 005, 0*010 and 0-020 in. respectively. Failure characteristics of some metals subjected to torsion only 181 A radius of 0-010 in. was found to eliminate the cracking caused by localized stress concentration at the junction of the gauge length with the shoulders. Figure 2 illustrates the geo

21、metry of the test piece. Specification of materials tested Table 1 (Appendix 2) shows the analysis of all materials tested. These materials were carefully selected to facilitate practical investigations into their respectively different mechanical properties. With reference to the brasses, only the

22、lead content was accurately determined from a spectro- graphic analysis. From table 1 it can be seen that five different pure metals were considered. To observe any change in the mechanical properties resulting from a second phase in the material, it was necessary to carry out tests on pure 60/40 (C

23、u . Zn) brass. Similarly the effect on the addition of 0-35% carbon to iron was observed by using an E.N.8 steel. The change in the mechanical properties resulting from free machining additives to a material was examined by studying the effect of lead in 60/40 brass. Plain and compressive testing co

24、nditions The calculation of shear strain at fracture is dependent upon the radius, length of gauging section and circumferential displacement. The stability of the gauging section was checked for both plain torsion and combined torsion and compressive conditions. Throughout the preliminary experimen

25、tal stage a hardened steel pin was inserted into the specimen bore, the pin being coated with molybdenum- disulphide grease in order to reduce the frictional effects at the interface. However, under plane torsion conditions the presence of this pin made no discernible difference as the geometrical s

26、tability of the gauging section was maintained without it* Under compressive conditions, the hardened pin was considered necessary in order to maintain specimen stability (Bridgman 1943). To observe the effects of environment, all materials were tested in ambient air conditions, but to study the eff

27、ect of CC14 on the failure characteristics all tests were repeated with the exception of the leaded brasses. The objective of the torsion test was to produce shear strains of a similar magnitude to those encountered in the metal cutting process. The shear stress-strain properties exhibited under tor

28、sion could then be compared with those associated with the cutting process, the average strain rate for the tests being 1-7/sec. A compressive load was applied to the end face of the central shaft, and to minimize the frictional effects the compressive load was transmitted to the shaft via a proving

29、 ring and an anti-friction ball junction. A hydraulic ram was held in a specially machined sleeve (fig. 1) to ensure correct horizontal alignment with the head stock and torsion apparatus. Under compression the torsion specimen was found to shorten, causing a reduction in the applied compressive str

30、ess. To overcome this defect and hence maintain a constant proving ring deflection, the output from the hydraulic 182 E. A. Peirce ram was continuously adjusted. This procedure was found to maintain a reasonably constant level of compressive load. Results A selection of torque-twist curves are shown

31、 in figs. 3 (a) to 3 (m) inclusive. It will be realized that, if existing data on the modulus of rigidity is used as a means of determining the slope of the torque-twist curve for these materials, this portion of the curve should be much steeper ; in fact almost vertical. The actual rotation in this

32、 region is due to relative angular rotation of the measuring rig (Appendix 3) as the torque is increased. From these curves it will also be realized that the experimental apparatus was not sufficiently sensitive to provide the strain in the elastic region. However, the error involved in the determin

33、ation of fracture strain is of the order of 3 per 100 lb in. torque. All results presented in tliis paper have been corrected to allow for this. In the plastic region the actual twist is determined at the point dTldd = 0f i.e. this point is considered the point of fracture. For steel and brass the p

34、oint of fracture is easily identified but this is not so obvious for the other materials. From the torque-twist curves it can be seen that beyond the elastic region, the twist of the specimen, and lienee strain, increases rapidly with increasing torque. In the case of the steel， magnesium and brass

35、specimens the strain curve in the plastic region is of a non-linear characteristic, and when it reaches the maximum torque value, i.e. when dTjdd = 0, drops rapidly towards zero torque. In the case of aluminium, copper, zinc and iron, the straining in the plastic region commences with a non-linear p

36、ortion, but then assumes a characteristic which is roughly linear up to the maximum torque value, then again drops towards zero torque. However, in the case of aluminium, zinc and iron materials the extent of the straining is much greater than that for steel, magnesium or brass. Method of determinin

37、g shear stress and strain values is shown in Appendix 4, the effect of change in specimen geometry being also considered. Discussion of results Tests in an air-environment To clarify the discussion it is convenient to analyse these results by first comparing properties under the following sub-sectio

38、ns : (i) single-phase materials (ii) two-phase materials (iix) free-machining materials (iv) effect of axial compressive stress on shear strain and shear stress at fracture. (i) Single phase materials Table 2 shows the mechanical properties exhibited for single phase materials at zero compressive st

39、ress. Failure characteristics of some metals subjected to torsion only 183 Table 2. Variation in strain and stress at fracture Materials Fracture strian Fracture stress Ibf/in2 Aluminium 5-3 13 400 Copper 4*38 33 200 Iron 3*95 41 200 Zinc 1-78 18 200 Magnesium 0*28 18 000 From table 2 it can be seen

40、 that the greatest strain to fracture is exhibited by aluminium, with copper and iron showing similar but slightly lower levels, of these three materials aluminium and copper have a F.C.C. structure, whilst iron has the H.C.C. type structure. These structures exhibit the characteristics of high duct

41、ility when compared with the C.P.H. structures of zinc and magnesium. The shear strengths of these single phase materials judged by the stress at failure do not collate with any trend in the ductility, for example aluminium shows the lowest shear stress for the highest strain, but iron shows the hig

42、hest shear stress even though it exhibits high ductility. Magnesium, on the other hand, exliibits the characteristics of both low shear strain and shear stress at failure, (ii) Two-phase materials Table 3 shows data on properties of two phase materials at zero compressive stress, but also includes i

43、nformation on copper and iron for the purpose of comparison. Table 3. Effect of second phase on the strain and stress at fracture Material Fracture strain Fracture stress lbf/in2 Copper 4.38 33 200 Brass (60/40) 0*97 41 200 Iron 3.95 41 200 Steel (E.N.8) 107 63 900 From table 3 it can be seen that t

44、he effect of the second phase additives are to produce significant changes in the mechanical properties of both copper and iron. The addition of zinc to copper to produce the brass in the ratio 60 (copper) 40 (zinc causes an increase of 24% in the maximum shear strength but a reduction of approximat

45、ely 80% in the shear strain. This gain in shear strength is accomplished by a substantia loss in ductility. By comparison with the brass an addition of only 0-35% of carbon to iron, to give steel to Kn. 8 specification, produces a 55% increase in shear strength TiJaP R|i U 184 E, A. Peirce but a red

46、uction of 73% in the shear strain at fracture. This effect is similar to that of brass, namely the gain in shear strength is achieved at the expense of the ductility. (iii) Free-machining materials Table 4 shows data on the 4 free machining 1 brasses, but includes the data on copper and on 60/40 pur

47、e 7 brass. Table 4. Effect on mechanical properties caused by the addition of lead to produce free-machining brass Material Fracture strain Fracture stress lbf/in2 Copper 4.38 33 200 Brass (t being given by : 4 = L where $ is twist in radians, or = rxe 0-1875 x 0 or 57*29L 57*29 x 0-140 = 0*0232%. A

48、 gauge length of 0-140 in. is considered to be an approximate mid-position point between possible minimum and maximum gauge lengths. At this point the curvature 1 of the changing specimen diameter is small but above which is large. It should be noted, however, that the instantaneous diameter and len

49、gth of the specimen is necessary to compute the actual stress and strain values. The testing apparatus was designed to allow free axial movement thus enabling length changes to be determined. For a few limited tests on copper specimens this increase was measured under plain torsion conditions as sliown in fig. 6, 194 Failure characteristics of some metals subjected to torsion only The value of shear stress, based on the original dimensions of tubular crosssection is given by : Tmax 3MT 2ll(r-r22) 1.5MT n(0-0066-0*00196) where 3/T = applied