Author investigated austempering of Si-Mn spring steel and considered whether austempering can be practically applied for heat-treatment of springs instead of quenching and tempering by comparing mechanical properties of the steel heat-treated with both processes. The results obtained are summarized as follows; (1) Yield point and elastic limit are lower by austempering than by quenching and tempering at the same tensile strength. (2) Hardness is lower by austempering than by quenching and tempering at the same tensile strength of more than 125kg/mm2, i.e., tensile strength is higher by austempering than quenching and tempering at the same hardness when steel was austempered at temperature of 350-400°C. (3) Elongation is inversely proportional to tensile strength when quenched and tempered. When austempered, its value shows peak at temperature of 350-375°C, but the relation between elongation and tensile strength is not clear. (4) Shock value is higher by quenching and tempering than austempering so long as tensile strength is lower than 125kg/mm2, but this relation is reversed at the tensile strength of more than 125kg/mm2.
Now a days there are many investigations on the S curves of steels, and it is generally recognized that mechanical properties of Bainite structure are suitable for spring. We examined the effect of austempering process for several kinds of spring steels and investigated whether this process is suitable practically or not. The results obtained are summerized as follows:- (1) Suitable size of steel for austempering process may be very small, especially for carbon steel. (2) In carbon steel, Eutectoid composition may be suitable for austempering process. (3) In Si-Mn steel, the effect of austempering process may be better than in carbon steel. (4) We recognized that Cr-steel may be unsuitable for this process.
The author investigated the durability of plain carbon and Si-Mn spring steel by Matsumura's repeated impact testing machine. The results obtained are as follows: Si-Mn steel is considered superior to plain carbon in durability when impact energy is beyond a certain limit: the opposite is true when the impact energy is below that limit. Tensile strength and yielding point versus equal hardness is lower in Si-Mn steel than in plain carbon steel but elongation, reduction and durability are higher in Si-Mn steel. Tensile strength. and yielding point versus equal hardness is lower in Si-Mn steel with higher Si and Mn content, but elongation, reduction and durability are higher in that case. The durability varies with elongation and reduction.
It is well known that surface conditions have much influence on the fatigue strength of steel. This fact cannot be too much emphasized for springs, because nearly all springs have an as-rolled surface condition and function under repeated stress. The author has investigated the relation between the thickness of decarburized layer and reduction in fatigue strength of Si-Mn spring steel. The results obtained are as follows: (1) Reduction in fatigue strength is not proportional to the thickness of decarburized layer. Only slight amount of decarburized layer causes sudden reduction in fatigue strength to a certain value and no distinct reduction is observed even when its thickness increases further. (2) Fatigue strength of this steel when decarburized in surface is 50 to 60 per cent of nondecarburized steel. (3) This reduction is considered to due to notches which break out in the decarburized layer.
With reference to the improvement of carbon steel for springs the hardenability of SUP 3 has been studied as follows: (1) Discussion on the effects of the chief chemical components and some hardness distributions, from the statistical data of quality control and other several investigations. (2) Tests on mechanical properties, hardenability and microstructure. From the study, it is urgently necessary to improve the hardenability of SUP 3, and is recognized that the practical mass effect is to be considered in every case.
In these days, one of our troubles in spring production is that spring leaves whose sizes are 13×100×250-1250mm are not hardened enough to the center of the section in usual oil quenching because of low hardenability characteristics of plain carbon spring steels (JIS SUP 3, SUP 4). It is, therefore, desirable to improve hardenability of these steels. For this purpose, authors firstly consider to increase Mn content, and then have experimented and calculated the effect of C and Mn content upon the hardenability of these steels. From these results, authors have determined the minimun limit of Mn content needed for the desired hardenability, and also observed the eftect of C content at constant Mn content. The results obtained are as follows: (1) Hardenability of SUP 3 and SUP 4 is too low to get the desired hardening effect. (2) To obtain the desired hardenability, it is necessary that the ideal critical size D1 practically determined by Jominy End-Quench Test is 2 inches or over. From our experiments, it is found that the ratio of practically determined D1 to calculated D1 from chemical analysis and grain size is in the range from 0.75 to 1.0 and mean value is about 0.9. (3) It is necessary that needed minimum content of Mn for JIS SUP 3 and SUP 4 is at least 0.9-1.0%. (4) Improvement of hardenability is expected by increasing C in 1% Mn series, but not in 0.4% Mn.
A comparison of US and Japanese produced Automotive springs was made regarding the following: 1. Material used 2. Quality and Grade of steel a. Incidental alloying elements b. Non-metallic inclusions c. Grain size 3. Scale and Decarburizing Results of comparison were follows: 1. US springs were made of Silicon-Manganese steel, Chromium steel and Molybdenum steels, but Japan mades were chiefly of Silicon-Manganese steels. 2. The quality and grade of steels used for US springs were superior to those of Japan make. 3. Scale and decarburizing on surfaces of US spring could not be observed, but they were usually observed on surfaces of Japanese springs.
It is well known that. among the copper alloys, phosphorous bronze and nickel silver are the best used for springs but their severe cold forming is fairly difficult, so that for special purposes they have been replaced with other alloys, for example, brass. However, plain brass has a few unfavorable properties for springs of high class. Therefore, the authors investigated first to improve the present brass and then to explore the other alloys being substituted for above alloys. Furthermore, the authors examined the so-called “seasoning-change” in spring properties of various copper alloys in order to illustrate the phenomenon. The results were as follows: (1) The plain 60/40 brass could be improved by adding small amount of third elements, such as Fe, Mn, Al or Sn: When any one of these elements is selected, it is better to increase its amount until the other practical processings become difficult. (2) In the field besides brass, Cu-Mn-Ni alloys displayed peculiar properties as spring with good heat-resisting one. (3) Cu-Fe alloy containing only about 4% of Fe would be enough brought to the future usage as spring material, if the suitable heat-treatment for it is taken into account throughout the manufacturing process. (4) The “seasoning-change” in copper alloy springs was observed to occur more intensely as the Zn content of alloys increases and the process was considered to be as a sort of ageing phenomena.
In 1947, J. E. S. which specified slight concave flat bar for leaf springs was established at first. From that time, after several revisions, present slight concave spring leaves have been decided, and are being used in both railway and vehicle springs. We selected some sorts of bar with different form, examined the properties of them, manufactured the spring by using the best bars of them, and investigated the results statistically. The results obtained are summarized as follows:- (1) With respect to the form of concave section, the flat bar with tapered section, which has 4mm parallel parts in both edges, and 1/1000×width slight concavity in the middle part of the section, may be suitable. (2) The deflection of the springs, which are composed of slight concave leaves is larger than that of the old type. (3) In the case of mass-production of springs, camber and deflection of several kinds of railway springs, which we treat in this paper, are more suitable for present specification than those of old type.
A new formula calculating the deflection of a leaf-spring is introduced under the assumption that the leaves contact each other at their free ends. The formula is more precise than the one which has been used in general and coincides well with the practical cases. By the formula the forces which are transmitted to the leaves are also calculated and the stress distributions are obtained. This is useful especially on the case of a three-leaf-spring.
The vibration of the rolling stock has a close connection with the rigidity and frictional force of the laminated spring. But most of the formulae hitherto published for calculating the deflection of the laminated spring neglect the influence of the “inter-leaf” friction or deal with it only experimentally. The writers have already presented a new theory for calculating the deflection of the laminated spring, in which the influence of the “inter-leaf” friction is taken into consideration, and have shown how to obtain the hysteresis curve of the spring thereby. This paper describes the outline of the theory and shows several examples of numerical calculations to obtain the hysteresis curve in the case of any cyclic change of the load. In another paper, the writers will show the results of the experiments made with a model spring, which fairly agreed with the calculation.
The amount of nip in laminated springs has been determined in most factories by mere experience due to lack of knowledge. The writer has already published a method for its calculation in order to obtain the total force necessary for fastening up the nipped plates, the pull or increase in camber at the fastening, and the initial stresses in each plate due to the fastening. (See the Trans. Jap. Soc. Mech. Engrs. Vol. 17 No. 63, 1951). This paper describes an application of the theory, setting several examples of numerical calculation with various kinds of actual truck springs. The results of the calculation agreed with the experiments with an error of less than 10%, and it was found that unexpected high stresses are produced in the shorter plates even by a slight nip in the shorter plates. The writer, therefore, recommends that the amount of nip should be limited so far as the stress in the shortest plate does not exceed 30kg/mm2.
The author introduced an equation to find stress due to nips of laminated springs, assuming that each leaf is laminated with equal curvature. He found that this equation coincides with experiments better than the method assuming each leaf is cantilever, at the center of leaves. He also, with this method, introduced an equation for the practical design to find the curvature of each leaf with simplified stress calculation.
Making leaves of 20mm×3mm section from a spring steel, we assembled model laminated springs of cantilever. The spans were 300mm, off-set were uniform steps and the number of leaves were 3, 5 and 7. The frictional coefficients of the laminated springs were 0.017, 0.028 and 0.040, respectively. We studied the characteristics of the vibrating systems consisted of these springs and the mass attached to the springs. Exciting the sprung mass, we observed the exciting frequency and the full amplitude; experiments were done at the frequency up to 400 cpm. and the full amplitude of the mass up to 2mm. The results: At certain exciting forces, there were differences in the vibration characteristics when the exciting frequency was increasing and the case when it was decreasing, as it usually occurs in the non-linear vibrating systems. Considering that the area of static load-deflection hysteresis curve means the energy loss during 1 cycle, and that resonance occurs when the loss becomes equal to the input energy, we can estimate the resonant amplitude of the system. It becomes clear that difference between the observed and the estimated amplitude is below 20%.
To understand the mechanism of inter-leaf friction in leaf springs, the author tried the vibration test for three kinds of springs. By these tests it has become clear that damping factors and natural frequencies are variable with amplitude. It is considered that variation of damping factor shows complex of friction mechanisms, and variation of natural frequency is due to hysteresis characteristics. Now, the author intends to investigate variation of forced vibration characteristics by amplitude and relation of vibration test to static test and, moreover, influence of nip, lubrication and liner between leaves on vibration characteristics of leaf spring.
In this report, we studied the effect of the lowering the stress concentration of the hole of center bolt by the change of the hole form of the leaf spring. In the bending fatigue limit, αk>βk, and ηk is not constant, where ηk=(βk-1)/(αk-1), αk: stress concentration factor by the calculation of the elasticity theory and βk: stress concentration factor by the fatigue test. Next, we experimented the effect of the lowering the stress concentration of the hole of center bolt by changing the section form of the leaf spring. The form of Jeep Type is not good in this test. Si-Mn steel used is, one as rolled (polished), and the other after quenched and tempered (sandpaper-finishing). 12 test pieces were selected.
Recently in Japan some reports concerning nips for spring leaves are published, and in our factory this problem has also been taken up. In this report nips for bearing spring AC 3397 have been treated. We calculated the proper nips by “the same curvature method, ” and we are manufacturing springs with newly determined ones. Owing this calculation and manufacturing process, following facts became clear:- (1) Number 10-14 leaves of the springs with old type nips are overstressed, then they are dangerous of breaking or setting. (2) From this reason, it is desirable that there are no nips in these leaves. (3) The springs above mentioned are able to be manufactured by the careful operation in manufacturing process.
The leaf spring is one of the important constituents of automobiles. So its improved quality and durability will be vital factors for increasing car working efficiency and reducing maintenance cost. In order to be used for long time in good condition the leaf spring is not only required to be designed suitably but also attention should be paid to the quality of material and processing. Of course, the leaf spring is apt to snap earlier than usual when the working conditions on road, speed and load etc. are not suitable. We suppose, however, that it is apt to snap earlier owing to the various defects on the quality of materials used. Accordingly, we wanted to investigate what defect will force the leaf spring to snap earlier. And so we picked out about 170 leaves snaped within their guarantee period, and we investigated each their snapped position, surface aspect, fracture, hardness, depth of decarburized layer, micro-structure etc. In addition, we investigated on characteristics of vibration when automobiles run through various roads, and on the effect of shock absorbers which were used on a bad road.
In order to investigate the cause and existing conditions of the failure in the automobile leaf springs, we have performed the test on the front leaf spring of the Toyopet trucks. We found that there are two kinds of failure, the one being impact failure and the other fatigue failure. Spring leaves that have been heat treated to comparably high hardness are often failed by impact, and fatigue failure is properly related with the hardness, surface conditions and quality of steel. Means to improve the quality of automobile leaf springs were, therefore, concluded as follows: (1) Improving the quality of spring steel, (2) Improving surface conditions of spring leaves, (3) Improving uniformity of hardness in each spring leaf, (4) Minimizing marks of hammering on the surface resulting from correcting the shape of leaves with hammer, (5) Rationalizing amounts of nip.
In order to improve the design and characteristics of coiled spring, the author and his cooperators have been investigating formulae, effects of surface condition and other stress problems. These investigations are now going on, and this paper is only a brief description of the results hitherto obtained. 1. On formula In conventional formula, number of active turn is assumed to be constant. But in our investigation, it is variable with deflection of the spring as follows; nt-na=aδ+c=(a dD/p)(δ/nt p/dD)+c where, nt: number of total turn, na: number of active turn, δ: deflection of spring, d: bar diameter, D: mean coil diameter, p: pitch of spring a, c: constants and nl-na=1.11-1.56 a dD/p=6.9-8.1 are obtained experimentally. 2. On surface conditions and stressing, investigation of the following subjects are now going on: a). Decarburization of Si-Mn spring steel. b). Theoretical and experimental investigation of fatigue strength of case-hardened spring. c). Improvement of torsional fatigue strength by shot-peening.
The superiority of the quality of materials is essential for a good time piece spring. Accordingly the quality of raw materials, composition, heat-treatment, micro-structure, surface finish etc. must be fully examined for making good springs. The author has investigated influence of impurities, relationship between non-metallic inclusions and cleanliness, change of hardness and micro-structure with a rate of cold working and intermediate annealing, heat-treatment and mechanical properties on the spring steel sheet which had been made of good Masago iron sand. The author has also investigated for purposes of reference, foreign and Japanese time piece springs which were obtained on the open markets.
In order to improve spring steel made in Japan we have investigated fundamentally the following properties of spring steel sheet made in Sweden and England as to the chemical composition, micro-structure, allowance of thickness and behaviors on quenching treatment.
The authors examined the relation between the damping capacity and fatigue strength of time piece spring. The results obtained are as follows: (1) The damping capacity of those springs which are oil quenched and tempered is the lowest at the tempering temperature 300-400°C. At this tempering temperature the structure of the spring is most stable and strongest for repeated stress. This low damping capacity and high fatigue strength are due to the fine tempering structure. (2) When the spring is tempered at constant temperature, the damping capacity rapidly decreases and then slowly increases as the tempering time. This co-ordinates the relaxation of internal stress and the decomposition of residual austenite. (3) About the Austempering, the experimental results were so irregular that no conclusion was secured.
To obtain the fundamental conditions for practical heat treatment of spring steel (thickness 0.4mm) containing 0.94 and 1.19% carbon, the following researches are performed: on scaling and decarburisation, grain growth characteristics, softening speed on tempering and hardness change on isothermal transformation by heating in usual resistance furnace and salt bath.
In this report, the ideal form of spiral mainsprings was calculated, so as the spring to be of uniform strength, and it was found that, if the original form of the spring is one of the modified Cornu's-Spiral, the spring will suffer uniform bending moment either at full or at no wind.
Now-a-days, on account of shortage of Swedish Steel which had been imported prior to the War, time piece springs are produced from high carbon steel in this country. There were many impurities in this steel (e. g. Cu, Cr, P, S, etc.) and caused breakings very often in rolling and employment. The. study of time piece spring became necessary. There are a few references relating to the effect of austempering on the properties of steel. However little seems to have been done on the effect of austempering or martempering on time piece springs. The authors, taking samples of two kinds of thin high carbon steel plates and rod bars, examined the mechanical properties due to austempering and martempering. Below 300°C the mechanical properties due to austempering were superior to martempering in thin high carbon steel plates.
In the “Metal Progress” 1948, it is reported that the fatigue limit is improved by Martemper and temper. It seems to differ from martemper so-called in this country, but it means to hold a temporary time just above the Ar″ transformation and to cool gradually in this critical range of the Ar″ transformation and temper. The fact that the fatigue limit is improved by Martemper and temper is likely due to being free from Micro-cruck by quenching, but the problem is not yet solved. Authors reported on the comparison of the mechanical properties in Martemper and Austemper at the low reaction temperature (tempering) in the 1st report. The result of the comparison of the mechanical properties in Martemper and temper, Austemper, and Oil quenching and temper on the Hardness, the Tensile Strength, the Elongation, the Fatigue Limit and the Simple Torque Test is given in the 2nd report.
Referring to the time piece spring it may be classified in two according to its size: watch and clock. In general the former is about 0.95-1.2 per cent carbon and has the sorbite structure having a few spheroidal cementite. On the 3rd report the anthers examined what effect the size of the spheroidal cementite will give to the mechanical properties of the main spring. (1) At the B heat-treatment, when the cementite size is below 2.8, the Fatigue Breaking Number, the Torque Test Value, the Proportional Limit and the Young Modulus are at the max. value while the Permanent Strain and the Damping Capacity are at the min. value. (2) At the D heat-treatment, namely below about 4.1 of the cementite the mechanical properties become worse abruptly and above this size they become the worst. (3) At the A heat-treatment, namely the spheroidal cementite being absorbed in the austenite. The mechanical properties are inferior to the B heat-treatment.
Authors devised the torque and fatigue testers for main spring of clock and gramophone, and investigated on testing, heat-treatment and characteristics of these springs. Authors also investigated the hardness test of these thin spring sheets and determined the thickness limit to be measurable in a reasonable accuracy by Shore, Rockwell and Vickers hardness testers.
On revising Japanese Industrial Standard of spring lock washer, the auther was in charge of drawing up a revision plan as a member of spring lock washer sub-committee. Many weak and irrational points were found in the present standard by calculating fibre stress and load for solid compression, and these have been revised in the new standard. Attention has been given to have the size of sections of washer conform as much as possible to the SAE standard to reduce the number of calibers at factories where washers of SAE-and JIS-size are to be produced.