THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING STUDIES IN METAL POWDERS - BEARING. COMPOSITIONS Keshav S Sanvordenker August, 1957 IP-236

This thesis was submitted by Mr. Sanvordenker in partial fulfillment of the requirements for the degree of Master of Science at the University of Bombay, -i-ff

TABLE OF CONTENTS Page LIST OF TABLES o O. o o o o o o o e o o o a o o o o o o o 0 o o o o o o e a o o o o o e o. iv LIST OF FIGURES o O O O a o o o o o o o o o o V SYNOPSI~S O o e o o o o o o o o o o o oo ooo o o oeo o o o o o o o o o ooooo oo o o vi FOREWORD o o o O o.. o o o o o o o. o o..... o o o o o o o o o o o o o vii PART I STUDIES IN METAL POWDERS - ELECTRODEPOSITION OF IRON POWDER Io INTRODUCTION 1...... 1 Io INTRODUCTION - o e o o. o o o o o o e o o o o o o o o o e o o o o o o o 1 Jo a JRIllO oooooooooeooo eoooooooooooooooooo 5 III. RESULTS o o. o o o.. o...... a o o..oe o o o o o o 8 IV. DISCUSSION OF RESULTS...oo..ooo.,,.o o.ooooo 23 V. CONCLUSIONS............0000000000000000000000 27 PART II: BEARING COMPOSITIONS - IRON-COPPER POWDER BEARINGS I. INTRODUCTION o....... o o...o.. o o o o o o o o o o 28 IIo EXPERIMENTAL.. ooo o o. o o o o o o o 00000 0 32 III. RESULTS.. o0 0 0.oooooooooooo..00 0 0 0 0 0 0 0 0 o 0 37 IV. DISCUSSION OF RESULTS o o. o o o o o o o o o o. o 53 Vo CONCLUSIONS.... o... O O O o oo...... o oo.o o o o o 60 LIST OF REFERENCES..oo 0000000o. o0000 00000000000 o000.o 62 - iii -

LIST OF TABIES Table PART I -age I 2 II 9 III EFFECT OF CATHO DE oC GURRT DESITY,.,,.i.O. O - 12 IV E CT OF TIE.OOCEQ AION OF FIROUS CHLO.IDE IN TIE BATHO.,.. o -. * o.,,O... 14 V EFFECT OF T'EMKIEU.. OF THE B ATH* o.,...,.'.-.,,,..'wP.,:. 16 VI EFFECT OF THE VARIATION OF THE pH OF THE BATH.o,.:..,a 18. VII EFFECT OF I ICTTVAL OF POWDER oEM VAL.,., 9. 20 VIII.EFFECT OF ADDITIONGN ASSES T.., e.Q...-..,,...,,.-...,, 22 PA II I.FECT OF BRIQUETTI NGP P ESS2 o UE p~. -.s.,...,, a a. W, 40 II EFFECT OF TMERAThUR OF SINTERING...,.o.,6,,,.:, o. o.. 42 III EFCECT OF TDIE,SIING o *..., o 4 oa ^ IV EFFECT OF COPE CO4TET,.. a... 1 -47 V EFFECT,F GRAPHHITE.E CO1NT o,. o,.:.... o...,:.,..-.. 49 VI EE GCT OF VAIATION. I PATICE.SIZE....., E4 0.,,..o,, 52 < iy -

LIST OF FIGURES Figure PART I Page 1 EFFECT OF TEE VARIABLES ON POWDER CURRENT EFFICIENCY.. 10 2 EFFECT OF CURRENT DENSITY ON POWDER CHARACTERISTICSo.. 11 3 EFFECT OF IRON CONCENTRATION ON POWDER CHARACTERISTICS 13 4 EFFECT OF TEMPERATURE ON POWDER CHARACTERISTICSooo... 15 5 EFFECT OF pH OF THE BATH ON POWDER CHARACTERISTICSo..o 17 6 EFFECT OF INTERVAL OF POWDER REMOVAL ON POWDER CHARACTERISTICS.......oooo.....o oo,oo0o 19 PART II 1 EFFECT OF PRESSURE ON BEARING CHARACTERISTICS.o.o:,0. 38 2 EFFECT OF PRESSURE ON CHANGES AFTER SINTERING..ooo.oo 39 3 EFFECT OF TEMPERATURE OF SINTERING ON BEARING CHARACTERISTICS....ooooo.. oooooooooooooo....... 41 4 EFFECT OF TIME OF SINTERING ON BEARING CHARACTERISTICS.....eo..ooooooooooooo o. o eo*., 43 5 EFFECT OF COPPER CONTENT ON BEARING CHARACTERISTICSo.o 45 6 EFFECT OF COPPER CONTENT ON CHANGES AFTER SINTERING.oo 46 7 EFFECT OF GRAPHITE CONTENT ON BEARING CHARACTERISTICSo 48 8 EFFECT OF PARTICLE SIZE ON BEARING CHARACTERISTICSo... 50 9 EFFECT OF PARTICLE SIZE ON CHANGES AFTER SINTERING.o oo 51

SYNOPSIS With the development of powder metallurgy, iron powder finds increased applications as a raw material, A particular mention may be made of iron base bearings which re replacing their contentional counter. parts in industry due to their self.lubricating properties and their superiority over other porous bearings such as the bronzes due to their higher hadness, strength and a favorable coefficient of thermal expansion which matches closely with that of the steel shafts.. With this in viewy a study has been made about the preparation of iron powder by electrodeposition and the utilization of this powder in the fabrication of the porous bearings. The electrodeposition of iron powder has been carried out in a cell provided with mild steel anodes and stainless steel cathodes, employing ferrous chloride bath. The operating variables with their limits used are as follows: Concentration of the bath.o.o. 25 60 g/1 Temperat-ure#,,... o 0:,., o..22~ - 45~C p.~ o.. 0. o., O o O.0,, o ff O,: a 3. 5.5 Current density.....0 o. O... 40 - 150 amp./sq, ft. Interval of powder removal.. 5 60 mino The effect of the following modifying agents has also been studied: naphthalene sBlphonate, glycerine- glucose and gelatin. The powder is tested for its particle size- apparent density, flow properties and specific surface, This powder mixed with electrolytic copper powder is used as a raw material for the study of the bearing compositions. Bearings from these powder mixtures are pressed in a specially designed die and sintered in hydrogen. The effect of the following:ariables on the properties of the beaings have been studied0. Bri.etting pressureo. o,,. 5 35 tons/sq. Xino Sintering tfemperature. b..... 900 - 13000 Sixterxig time, p o.....o o,..,.oo: *o 8 ho s... Copper content...,................ 30 p4ercent Graphite content..-O. o o... O.. 0 6 percent Particle size o......... o.o, 0.O 100 percent ( 48+1l0 mesh) The dimensional an density Ehanges after.sintering are measured and the bearings tested for their porosity and radial compress — iTe strengtho

FO'REORD The importance of powder metallurgy lies in the ability of this technique to produce complicated metal shapes within a close limit of tolerancel by a process requiring much less skill and time than machining, and much lower temperatures and fewer finishing operations than casting. Moreover the products of powder metallurgy may possess all the physical characteristics of their fused and cast counterparts as well as many that cannot be obtained thereby. Refractory metals and cemented carbides are exmples of materials whose high melting points prohibit the application of fusion methods.o Eren the most modern method.s of centrifugal casting have failqd to produce products in which two or more metals insoluble in each other in solid or l id state (such as copper and lead) are combined in a structur e of unifoly distributed constituentSo Yet another example is the self/-lubricating bearing having a multitude of inter-connecting pores which serve as a storage for lubricating oil.o The new trend.in powder metallurgy is to displace the conventional methods of production of metal parts because of the saving of time and costo A striking example is the gear of an oil. pump. By orthodox method~s only skilled machinists ha've to be employed to prepare these articles whereas by the new technique e"ren a layman may be trusted to operate an automatic pressp Another major consideration is that in cutting a machined gear from a east blank, about 64% of this metal is lost in chips whereas by powder metallurgy technique over 99% of the raw material is fully utilized.. Powder metallurgy is however no panacea.and its applications are s.uject to certain limitations. According to Patch3 these limitations are based on three main factors mate.rial, men and machine so The characteristics of the powder dete.ryine to a large extent the properties of the finished product; Balke has shown the importance of purity and density for the attainment of maximum physical properties. The cost of the powder, though unimportant, in small and intricate articles produ.ed on a large scale, does form an appreciable factor when the parts are larcge0 The equipment used. in powder metallur= gy industry imposes certain other restrictions. For economical production) expensive and large presses of high capacity a4re essential. Another limitation is the design of the dies and punches. Strength and wear resistance of the die and the toughness and fatigue resistance of the punches have to be considered~. Other limitations are imposed by the complicated and X rTli -

expensive equipment for the poroduction of the powder, for conditiontig them and stor i without cotn ation by oxyg or moistwe. ces oeated at high sitering temperae tres ivolve many replaceOver- ming the above mentioned limimttiions calls for a knowledge which can be gained by a systematie investigation. Hence the present wok was undertaken, to study the pod.ution of iron powder by eleetrodeposition on, a semi-pilot plant scale, and to utilize the powder in the prodction of self-lariting beaings and study the different factos which affect their properties. - viii *

PABT I'TUDIES IN:MAJL,POWDRS. ECtBZ O IONl OF IROP: PO. R s:I _'.;-' V -.-~ ~ * * w - 4- ~~ i~

I. INTRODUCTION Iron powder forms a basic raw material in powder metallurgy. When prepared electrolytically, there are three distinctly different methods used.5 (1) A hard brittle deposit is obtained which is subsequently ground to a powder. (2) The metal is deposited in a soft spongy condition and lightly rubbed into a powder. (3) The powder is deposited directly as such from the electrolyte. Ferrous chloride and ferrous sulphate baths have been used. A few preliminary trials showed that coherent deposits are more often obtained with ferrous sulphate bath than with ferrous chloride bath. It was therefore decided to use ferrous chloride bath and the present investigation is concerned with the direct deposition of iron powder from the chloride bath. Literature Survey Several investigators have previously prepared iron powder using varied conditions of bath composition,, current density and temperature. However, it is not possible to say what conditions should be employed. to obtain a powder [having a particular screen analysis, good flow and density properties which are the controlling factors in determining the properties of powder compacts. Kroll6 obtained dendritic iron of 100-300 mesh size by using molten ferrous chloride diluted with an equaimolar mixture of NaCl and KC1 to a concentration of about 20-30 percent ferrous chloride, a current density of 1440-2800 ampj/sq.ft. and a voltage of 1-5 volts. Stoddard' recommends a bath containing ferrous chloride 45-200 g/l., a temperature of 70-90~CG a current density between 50 and 300 amp./sq.ft. and a pH between 1.2 and 2.4. Addition of 0.1-20 g/l of a manganese salt serves to prevent coarsening of the grain size, and to carry out electrodeposition at a lower temperature. 8 Wranglen used a solution containing 15 g/1 of iron as ferrous chloride and 100 g/1 of ammonium chloride and studied the effect of temperature from 20~C to 900C and of pH from 2.5 to 6.5 on the nature of the deposit. He found that higher current densities tend to deposit finer powders and that he p o the bath and hydroxide formation exert an important influence on the nature of the deposit. -1

Casey9 prepared iron powder using a cell having stainless steel cathodes and iron and carbon anodes such that about 5 -10 o of the current passed through the carbon anodes. The bath was 30-90 g/l ferrous chloride, 10-80 g/l ammonium chloride, pH 4o 0-6 3 and current density 40-200 amp/sq.fto Cudd and Freeman0 produced iron powder using aqueous ferrous chloride or sulphate solution containing 250-300 g/1 of the salt, a temperature 6f below 40~C, a current density from 25-140 amp/sqft. and a pH between 182.0o Modifying agents like gelatin saponins or peptones were incorporated in the bath and were found to effect a reduction in the particle sized 11 Primavesi has described a method to produce iron powder using iron or aluminum cathodes, separated from graphite anodes by a diaphragm. The anodes are covered with iron scrap to prevent the generation of chlorine. The operating conditions are: electrolytic bath - 10-100 g/l ferrous chloride and 50-200 g/l ammonium chloride free from organic impurities, pH 330o-6o5, temperature below 70~C and a cathode current density of 50-200 amp/sq.fto The following table briefly summarizes the main conditions eployed by the various workers. TABLE I Investigator Extra Concentration pH Current Density Tempo ~C Additions g/1 ___po /sq Kroll W fa & K 200-300 a. 1500-3000 chlorides fused bath Stoddard Manganese 40-240 1. 2-26 25-300 70-90 salt Wranglen Ammonium 15 2 5-6.5 100-200 20-50 chloride Casey 1 - 30-80 4.0-6.3 40-200 Cudd and Gelatin, 250-300 1 &8-2. 0 25-140 Below 40 Freeman peptone saponin Primavesi Ammonium 10-100 3, o6. 5 50-200 Below 70 salts The object of the present investigation was to study quantitatiely the effect of the different controllable variarbles on the deposition of iron powder from ferrous chloride bath and hence obtain conditions for the deposi-s tion of iron powder suitable for use in powder metallurgyo ^ ~2 -

ltectrod.epositioan Chemistry and. Mehanism of the Process Chemistry Electrodeposition is the most significant method for powder production of high melting point metals. The method is precisely the reTersed application of the well-known electroplating processo Iron cathodes can be used although stainless steel or monel metal serves better due to the bright surface polish and resistance to chemical attack. Mild steel is generally used for anodes though pure iron would be ideal In,sme- cases when chemical attack is appreciable, insoluble anodes such as graphite or lead are used together with iron. The principal cathode reaction is the discharge of ferrous ions in the form of loose powder. However. when the supply of metallic ions from the solution becomes inadequate, hydrogen evolution invariably occrs, fue to the depletion of hydrogen ions in the cathode film, the pH is usually higher than in the bulk of the solution and hydroxide precipitation is the subsidiary reaction. The electrodeposition in powder form occurs necessarily at less than 100% cathode efficiency, whereas the anode efficiency is generally above 100% due to anodic solution and chemical attacks Mechanism Special conditions in the electrolyte cause the deposit to be in the form of loosely adherent fine crystals, The electrodepo-sition of the metal causes a reduction of metallic ions in the vicinity of the cathode; this is restored by mechanical movement and.onvection of the solution, by diffusion.and by ionic migration in the electric field, The film of liquid for a distance of about 0.5 m.m. from the cathode is not readily disturbed by agitation and cvetion and ioic migration is also slow (Circ. 5 x lO14 m/see/V/a..c:12 Thus, diffui trgh the station. wsy cathode layer is the. principal cause of the supply of ions. When the rate of' deposition exceeds the rate of supply, a part of the eCrrt is. diverted to an alternate cathode process viz., hydrogen evolution,. The metal ion concentration and the cathode potential change very rapidly across the relatively thin film of cathode.solutiony the result being the formation of excrescences on the cathode. These receive a higher current density, contributing to the growth of numerous tree-like, dendritic deposits which may be easily separated from the massie tcathode by a slight distu ban e. -. 3 -

8 Wranglen discusses the factors important for the formation of powdery depositso These are: (1) Low metal overvoltage (2) Depletion of the metal content of the cathode film (3) Formation of basic material in the cathode film. It is doubtful whether low metal overvoltage per se, is responsible for powder deposition, and it may as well be claimed that low recrystallization temperaturet3, low melting point or low tensile strength promote powder deposition. However, the fact remains that metals with low metal overvoltage are much more easily obtained in powder form than the high overvoltage metalso Furthermore the other properties are found to coincide with low metal overvoltagee Under conditions favoring a high depletion of metal ions in the cathode film, the deposited metal cannot form a coherent coating, but tends to grow away from the cathode and out into the bulk of the solution where the metal ion concentration is higher. Chiefly, these conditions may be enumerated as an initial low concentration of the metal, a high current density and a low temperature. Precipitated flocks of hydroxide present in the film of the cathode may settle down on the cathode surface and prevent it from growing, whereas the free parts of the cathode continue to grow, Precipitation of hydroxide is promoted by the use of a neutral or only slightly acidic solutions and by hydrogen evolution which causes a depletion of hydrogen ions in the vicinity of the cathodeo Lower temperatures form a contributing factor due to the slower rate of diffusion. It must be noted that only precipitated hydroxide (gel form) causes the formation of powdery or spongy depositso Basic material in the colloidal form (sol form) in the cathode film often has a contrary effect fostering the deposition of fine-grained, smooth and hard deposits In the case of iron powder, depletion and a low cathode current efficiency are favored by the same conditions of electrolysis,vizo, low temperatures, high current densities and low initial concentration of the batho Thusdepletion and hydroxide formation are the principal causes of deposition of iron powder. - 4

Io EXPERIMENTAL Preparation and Analysis of the Solution The electrolyte used was an aqueous solution of ferrous chloride C.Po The normal strength of the solution was 30 g/1 FeCl2o4H20o On standing, and similarly after electrolysis, the solution becomes turbid due to the slight hydrolysis of the ferrous salt. However, this does not affect the composition appreciably, and the precipitate can be easily filtered. The iron content was estimated by titration against potassium dichromate solution using potassium ferricyanide as an external indicator. Aparatus and. Experimental Procedure The apparatus used in these experiments was similar to that used by Joshi et al.15 The electrolytic cell consists of a glazed porcelain rank with a bottom outlet. Stainless steel cathodes and mild steel anodes (6" x 4") spaced 2" apart are suspended from two threaded parallel rods and kept in position by nuts. The rods are mounted on brackets: which are screwed to two spindles. The turning of the spindles enables the entire assembly to be raised or lowered. The bath liquid is circulated by a pump; a heat exchanger incorporated in a circuit controls the temperature of the bath. The current is supplied by a motor generator set operating.at 20 volts. A proper resistance included in the circuit enables the control of the cathode current densityo The volume of the solution used in the experiments is about six litreso The experiment is carried out for one hour, the powder being gently tapped off the cathode at intervals of ten minutes. The powder is removed from the bottom of the tank, washed free of the electrolyte and dried, Recovery of the Product As it is essential to obtain a pure and unoxidized product, considerable attention has been given to the washing and drying of the powder. This is particularly important in the case of iron powder owing to its reactivity in the presence of moist air. Thus, a freshly deposited sample kept under water for a week is found to oxidize to the extent of 80%, - 5

Among the numerous methods tried, particular mention may be made of Vacuum drying, drying in a reducing atmosphere, washing,the pow~ der with antioxidants like tartrate, stearic acid, etc.> incorporation of modifying agents like gelatin in the bath, Though soe of these prevent oxidation to a certain extent, none of them was quite satisfactory. A method recently suggestedl2 inTolves washings with too many solutions and was therefore not followed. It was observed that the powder, when heated in a reducing atmosphere, loses its reactivity and can be stored without any further oxidation taking place. The electrodeposited powder is hard2 and in order that it may be useful in powder metallurgy, it is essential to decrease its hardness- by an annealing treatment in hydrogen at a temperatue varying from 600-700oC even though the powder might have been washed and dried without mch oxidation taking place during drying. Hence, the procedure followed was to heat the wet powder, wash. ed free of the electrolyte, to a temperature of about 650~C in a tn osphere of hydrogen, special care being taken to remove the final traces of oxygen from hydrogen. Thus, both drying as well as annealing was achieved in one operation. It must be mentioned that the temperature used in drying also affects the particle size. If the temperatures are those at which caking takes placey then mch coarser particles result, and powders which are apparently non-flowing may have an appreciable flow rate. At the temper. ature used for drying (650C), aking of the powder was not noticeable, Powder.Chactristc.s The iron powder so prepared was in each ase tested for purity particle size distribution, apparent density, flow rate and specific surface: i..l ty The purity of the powder was determined by estimating the pert centage of metallic iron in the sample,.The powder is treated with mer. ur'ic chloride solution whereupon the metallic iron reacts to form ferrou. *x' 12 A procedure referred here consists of the following scheme. The product is washed first with tap-water and the supernatant liquor decanted, then with 1% sulphuric acid and then again with tap-water. A small quantity of 10% citric acid solution is mixed and allowed to stand for a few Minutes. The mixture is again washed with tap-water and then dilute ammonia added. The wet powder is thoroughly washed by decantation with tap-water until the supernatant liquor is colorless, and finally with distilled water after which it is rapidly vacuum filtered, ringed with acetone and rapidly dried either on a hot plate or preferably in vacuum or dry hydrogen or coal gas. This involves too many washings and was therefore not followed., 6

chloride, the mercuric salt forming calomel or metallic mercury. Any other impurity including iron oxide remains unaffectedo The mercurous salt or any mercury so formed must be filtered off before the solution is titrated against standard dichromateo The metallic iron content was usually found to be above 96% and the total iron content about 98.5 to 99%. Particle Size Distribution Particle size distribution is important in all powder utilizations, particularly in molding practiceo A mechanical sieve shaker with Tyler screens was employed. The results are represented on the basis of median particle size which is the average particle size of the sample. Apparent Density Apparent density is important from the aspect of die construction and its influence on the characteristics of the finished producto 17 It was determined according to AoSoToMo specifications. Flow Rate Flow rate is the time required (sec.) for a powder sample of standard weight to flow through an orifice in a standard instrument according to a specified procedure. The apparatus used was the same as the one employed to find the apparent density. The results are based on the time required in seconds for the flow of 25 go of the dried powder, Specific Surface Specific surface is the surface area of one gram of the powder usually expressed in, sq. cm, Powder metallurgy technique is fundamentally based on this factoro That it. is related to particle size is.obvious. Finer the particle size, the larger the surface area.. Apart from this, it depends on the history of the powder. The one prepared electrolytically has the largest specific surface due to its dendritic porous structure o The apparatus used for the determination of the specific surface was a modification of the air permeability method, the one designed by Lea and Nurse. The specific surface is calculated using Carmen' s equation-9, 7

IIIo RESULTS General Plan of Work The general plan of work and the experimental procedure adopted in these investigations are as"follows~ A. Preliminary studies on the range of concentration and current density giving satisfactory deposits Bo Detailed study of the influence ofS (1) Current density (2) Iron concentration (3) Temperature of the bath (4) pH of the bath (5) Interval of powder removal (6) Addition agents on. (a) Powder current efficiency (b) Particle size distribution (c) Apparent density (d) Flow rate (e) Specific surface C. Studies of the drying conditions on the characteristics of the powdero Preliminary Studies Preliminary studies showed that the deposition range for iron directly in the powder form is narrow. Powder deposition was found to be unsatisfactory at low current densitieso Low iron concentrations gave a powder of a spongy nature, whereas high concentrations deposit a mixture of coherent and powdery iron, Detailed Studies The following table gives the values of the conditions held constant during each series of experiments and the extreme plausible limits of the operating variables. 8

TABLE II I.. _ ~,,, ~.:....! Variable Fixed Value Extreme Limits of..._____________________ _____________________Variation Current density 90 amp./sq. ft. 40-150 amp./sq. ft. Iron cone. as FeC12 * 4H2o 30 g/1 25-60 g/l Temperature 350C I 20-45~C pH 4.5 3 - 5.5 Interval of powder removal 10 min. 5-60 min. The results obtained by employing the different operating variables are given in Tables III to VII and the results are shown graphically in Figures 1 to 60 Experiments have been carried out in the presence of addition agents like glucose. glycerine, sodium P-naphthalene sulphonate, and the results are given in Table VIII. - 9.

100..100 z z w I I0 a: n- o GT a. a. S 80 -- -90.. I — I — Z Z: 60 80 o U o 0 a- 40 _______. 70_ 20 30 40 50 60 15 30 45 60 IRON CONCE.NTRATION g/l INTERVAL OF POWDER REMOVAL, min. IL 90 0 5 70 7 —- -13-5 CURRENT DENSITY, amp/ sq. f -t. FIG. EFFECT OF THE VARIABLES ON POWDER CURRENT EFFICIENCY a 50 o 50 70.90 110 130 150 CURRENT DENSITY, omp/ sq. ft. FIG I. EFFECT OF THE VARIABLES ON POWDER CURRENT EFFICIENCY

115 105 1 1500 w 90 -' i I;w 1300 75::,~o co W 1100\ 60 W' 0 a. 45 " 90C! d o - 0 N I 13 — I10 1]3 z W W _100 I —- -Z 4 W a. I3~~~~~~~~~~~ a: ___~~a. 90 a. 80- 0-9 _______ 40 60 80 100 120 140 40 60 80 100 120 140 CURRENT DENSITY, omp /sq. ft. FIG 2 EFFECT OF CURRENT DENSITY ON POWDER CHARACTERISTICS.

TABLE III EFFECT OF CATHODE CURRENT DENSITY Concentration of the bath - 30 g/l FeCl2.4H20; pH - 4.5;Temperature - 350C Interval of powder removal - 10 min.; Temperature of drying - 6500C under hydrogen Current Powder Median Apparent Flow Specific density current particle density rate surface Remarks amp/s q.ft efficiency size secs,> sq..cm/g percent microns 40 45iO.0 103.2 1.30 7.75 1112 Codeposition of J coherent and loose H powder ro i 1 ~50 67.7 113.2 1.10 8.J7 1471 Podery 6o 81.9 116.4 1.29 5.4 1183 tt 70 83.5 120.2 1.56 4.8 933 80 855.5 105.2 1.24 6.5 11Q00 90 85.1 95.2 1.10 8.3 1460 100 83*3 89.8 1.00 9.4 1530 n 120 87.5 84.8 o.94 10.3 1550 Excessive evolution of hyd r og e n 150 80.0 8392 0.92 11.8 1590 and heat The results are presented graphically in Fig. 1 and Fig. 2.

8 0 -- - _2000 -- E w U. "L w 6-0^~ ~ 1200 " --- I40! 04.00.w 400.'.__ a. 240 2-i5 —- — 25 - 200- 2-0 160 z -5! I I I / I I I — I IF<1: z w< 80 0 —--.5-2 —— 0 —4 - 20 30 40 50. 60 CONCENTRATION OF THE BATH Fe CI2 4 4H20 g/l FIG 3 EFFECT OF IRON CONCENTRATION ON POWDER CHARACTERISTICS

TABLE rv EFFECT OF THE CONCENTRATION OF FERROUS CHLORIDE IN THE BATH pH - 4.5 Temperature 350C Cathode current density - 90 amp./sq.ft. Interval of powder removal - 10 min. Temperature of drying 6500C under hydrogen Powder Median Conc. current particle Apparent Flow Specific g/l efficiency size density rate surface Remarks percent microns g/c.c. sees. sqcm./g H 5prcn F5s Apret Fo spR ~a~ ~ 25 77.5 85.0 0.715 xx 2183 Spongy 30 85.1 95.2 1.10 8.3 1460 Powdery 35 92.5 121.5 1.15 6.6 1200 Powdery 40 1 74.5 161.2 1.8 3.6 618 Mixture of 50 60 210 2.0 3.7 1 407 coherentand 60 55 233 2.17 3.5 359 loose powder The results are presented graphically in Fig. 1 and Fig. 3.

1-25C.. - - -- 2500 E i 1-15O ( 2 __ —- -— 2200 O\ 1-05 CI 1900 - /s\ <-"~~~~~~~~~~~~~~~~~i U<0-95 1600 a. 0n5 -i 1000 a. u~ iL w -U90 f 85 80 ----— / —----—, —75,,,___8 3; a 20 25 30 35 40 45 20 25 30 35 40 45 TEMPERATURE ~C FIG 4 EFFECT OF TEMPERATURE ON POWDER CHARACTERISTICS

TABLE V EFFECT OF TEMPERATURE OF THE BATH Concentration of the bath - 30 g/l FeC12.4H2O Current density - 90 amp./sq.ft. Interval of powder removal - 10 min. Temperature of drying - 65o00C under hydrogen pH - 4.5 Powder Median Apparent Flow Specific e Temperature current particle Remarks oc.effdiciency size density rate surface percent microns g/cIc. sees. sq.cm./g Spongy deposit 22 77.5' 85,2 0.875 xx 2379 73% Fe metalic r<~ ~30 82,5 93.3 104 9.3 1620 Powdery tj, ~~35 85.1 95.2 1.10 8.3 1460 4$ 92.3 92.-1 1.22 6.6 1288 55 xx xx xx xx xx Coherent The results are presented in Fig. 4

-LT - MEDIAN PARTICLE SIZE MICRONS APPARENT DENSITY g/cc - O O O _ _ _ ~(O O0 c- 01 0 1 0 1 0 co 0o _ d0 (A 6A 0 O1 66 - -0 —--------- ) ( / 0 — I. / — CA 0 ___ ___ _______ W / -O _ N O o co o oJ "r' —'1 o \.. I,' c-I -, __ __ m> CA 03 ~A S - ^- ---------— / —

TABLE VI EFFECT OF TEE VARIATION OF THE pH OF TEE BATH Concentration - 30 g/1 FeC1.4H20 Current density - 90 amp/sq.ft Interval of powder removal - 10 min. Temperature - 350C Temperature of drying - 650~C under hydrogen Powder Median Apparent Flow Specific current particle density rate surface Remarks efficiency size g/Coc. secs. sq.cm./g percent microns 2.5 xx xx xx xx xx Mixture of coherent and | I I f 1 I I loose powder co 3.0 80.3 99.0 lo22 6.1 891 Powdery 4.~0 84o 0 93.5 114 7.8 1400 O 4.5 85.1 95.2 1.10 8.3 1460 Addition of acid not necessary 5.5 75 1 83.5 0 795 xx 2951 Basic material deposited 6Q8 Fe(met,) The results are presented graphically in Figo 5

9 0 - -- - -- 1-4 - -— 0 7'0 w 1'4 z 0 ___ ____I _____ _ _ _ _ Ub - d 160C -- - -- - — 1400 z E O 140 100 1 2C0 6 I0 d 120 -- - - - 2-00 100 — 10, —- 1-200C 0 15 30 45 60 0 15 30 45 60 INTERVAL OF POWDER REMOVAL min. FIG 6 EFFECT OF INTERVAL OF POWDER REMOVAL ON POWDER CHARACTERISTICS

TABIE VII EFFECT OF INTERVAL OF POWDER REMOVAL Concentration of the bath - 30 g/l Current density - 90 ampo/scqoft. Temperature of the bath- 350C Temperature of drying - 650C under hydrogen pH - 4.5 Interval Powder Median Apparent Flow Specific of powder current particle density rate surface Remarks removal efficiency size g/c. c. secs. sq. cm/g min. percent microns 5 1 82 93.3 0.93 xx 2603 Spongy iX l: ~10 85.1 95.2 o 10 8.3 1460 Powdery ~i 15 87o 5 99- 2 11 8.0 1 395 30 90.5 112. 9 1.29 6.2 1123 45 94.0 126.5 o 47 5.0 964 60 93.0 16606 lo48 4o5 620 120 xx xx xx xx xx Hardvery coarse 180 xx xx xx xx xx Coherent The results are presented graphically in Fig. 1 and Fig. 6

Addition Ageits In electroplating practices of various metals, Various organic as well as inorganic additions have been found to improve the plating from a coarse to a fine-grained structure. Their effects, however are yet to be thoroughly understood so that their use is, purely empirical. The effect is attributed to their influence upon the polarization or cathode potential and structure of the metal deposit. There are basically no fixed rules for addition agents, They may be specific in their action, i.e,, an addition agent that works with one metal does not necessarily work with another. Various other factors like the pH of the bath have also some influence. Instead of a single addition agent, two or more are sometimes needed in combination to produce the desired effect. The subject is thus a wide one and the practice is one of trial and error. In the present series of investigations, the action of glucose: glycerine and sodium naphthalene sulphonate has been studied in details and qualitative observations made on the influence of gelatin, saponin and peptone. The results are presented in Table VIII on the following page, Drying Conditions It must be mentioned that the temperature used in the drying,and annealing the powder also affect the particle size. eWhen the tempersatures are those at which caking takes places much coarser powders result and powders which are apparently non-flowing may have an appreciable flow rate. An atmosphere of hydrogen is preferable because it presents a reducing atmosphere, imparts a bright appearance to the powder and on the whole improves the powder characteristics.. City gas and nitrogen were tried but the appearance of the powder was dull and the flow properties poor. 21

TABLE VIII EFFECT OF ADDITION AGENTS Concentration of the bath - 30 g/l,FeC12.4H20 Current density - 90 amp./sq. ft. Temperature - 35C; pH 4.5 Interval of powder removal - 10 min.. Temperature of drying-650~C under hydrogen Addition Addition Powder Median Apparent Flow Specific Agent Agent current particle density rate surface concentration efficiency size g/c.. o secs. sq. cm. /g percent percent microns Ni1 Nil 85.1 95. 2 1. 10 8.3 1460 Sodium 0.05 91-5 122.8 1.56 5.0 545 |naphthalene 010 94o 0 114.1 1.305 5.9 76 sulphonate 0.25 81.0 87.2 1.03 12.0 890 |,: 0, 50 73..2 77.3 0.87 14.3 1040 0. 05 91.6 142.4 1.664 5.2 575 Glycerine o 10 90.5 150.2 1 517 57 6 639 0.25 81.0 110.2 1 142 xx 1390 0.50 80.0 104.0 0.892 x x 1970 0.e05 92.3 133.0 1.215 8.8 1170 Glucose 0.10 80.7 116.2 0.950 xx 2020 0.25 78.3 80.3 0, 72 xx 2220 0. 50 xx xx xx x x xx i I i I ~ i m I I I II I I I I.,....

IV, DISCUSAION OF RESULTS,,* I_,- -..,' -.- v Powder Current Efficiency Considering the process of electrodeposition, at low current densities, the discharge of ions occurs slowly, and so the rate of growth of nuclei exceeds the rate at which the new ones form. This gives a coarse-grained deposit. As the current density is raised, the rate of formation of nuclei will be greater and the deposit will be fine-grained. At very high current densities, the rate of discharge of ions is very fast, so that in the vicinity of the cathode, there is depletion of the ions required for discharge and, as a result, the crystals tend to grow out into the bulk of the solution. It is at this stage that the metal deposits in the form of dendritic, tree-like powder. It may be noted that below a current density of 50 amp./sqoft, the operation is not plausible due to the codeposition of a coherent and powdery deposit. Below this value of the current density, it appears that the rate of discharge of ions is not fast enough to cause sufficient depletion of cations in the vicinity of the cathode so that dendritic out)d growth is not very much favoredo With an increase in current density above this value, conditions favor depletion and hence facilitate powder deposition, until the powder current efficiency reaches an optimum value of 85.5 percent at a current density of 80 amp./sqft. In this range, hydrogen bubble evolution, by obstructing crystal growth, has the same effect as depletion, in promoting powder deposition. Thereafter, evolution of hydrogen and. heat is excessive so that the powder current efficiency decreases to 80% at 150 ampo/sq. ft. The change in the concentration of the electrolytic bath has the most deciding influenceo It provides a very limited range for the proper deposition of the powdero At low concentrations,vizo 25 g/l, the current.carrying capacity of the solution is low and the secondary reactionyviz hydrogen evolutionis encouraged: This increases the pH of the cathode film solution which, with the high depletion of cations, deposits a very spongy powder with basic material and with a low current efficiency of 77.5 percent. Increasing the concentration increases the rate of diffusion and accelerates powder deposition process giving an optimum value of 92.5 percent powder current efficiency at a concentration of 35 g/lo At a concentration of 40 g/l, the deposit is a mixture of coherent and loose powder and any further increase in the concentr.ation increases the amount of the coherent plating, and decreases the amount of loose powder (Figo, ). X 23 -

The influence of concentration on the rate of formation of new nuclei is uncertain. Certain workers hold the opinion that the presence of a large number of ions in a concentrated solution favors the formation of fresh nucleio Certain experiments by Glazunov21 however, indicate that the rate of formation of nuclei is actually decreased by increasing concentration, but the plating is due to an increase in the rate of crystal growth across the cathode surface, combined with a decrease in the rate of growth in a perpendicular direction. Increase in temperature has two effects which oppose each other, In the first place, diffusion is favored so that depletion is counteracted and coherent deposits encouraged; on the other hand, hydrogen overvoltage is decreased so that'gas evolution is encouraged which facilitates powder formation. The ultimate effect depends upon which factor is predominanto At moderate temperatures, it is the first factor that has the deciding influence, whereas the second factor is important only in electrolytic baths of fused salts 8 At 22~C the powder is very spongy and the current efficiency is 77.5 percent It increases to 92.3 percent at 45~C (Fig, 4d ) above which the deposit attains a coherent nature. The pH of the bath has a peculiar effect on the powder current efficiency. As can be seen from Fig. 5d, an optimum value of the electrolytic bath maintains itself without the addition of any acid, At a higher pH, the deposit is spongy, basic material is precipitated in a large quantity and the efficiency falls to 75,1% at 5.5 pHo At a value of pH below 3.0, coherent powder is deposited and the operation is not plausible. The effect of interval of powder removal on the powder current efficiency is shown in Fig 1.. It:can be noted that as the powder formation continues undisturbed, the current efficiency increaseso More energy seems to be spent on the formation of fresh nuc-ei than in the growth of the previously formed powder crystals. The efficiency increases to a peak value of 94 percent at 45 mino interval and is very nearly the same at 60 mino interval-o On the whole it is found that the powder current efficiency is low for:a spongy deposit and increases as the powder becomes coarser in nature and suddenly decreases when the deposit tends to be coherent, 24

Paticle Size and Apparent Density The increase in current density from 40 amp./sqf.ft. to 70 amp,/sqoft. increases the particle size of the deposit. Further increase in current density causes a deposition of fine-sized powder (Fig. 2b), In most powder deposition processes, the current density is above the limiting value for the given electrolyte and hydrogen evolution is the side reaction, As the current density is increased, the hydrogen evolution is increased, and the bubbles interfere with the crystal growth, thus giving a fine-sized powder. Secondly, due to the discharge of the hydro-: gen ions, the cathode film solution tends to be alkaline and deposit basic material which is the second factor in the formation of fine-sized crystals. This is illustrated by the current density range above 70 amp./sq. ft. Up to this value, the hydrogen evolution it not sufficient to interfere with crystal growth and the particle size increases between 40 amp /sqo.ft, and 70 amp./sq ft. With increase in iron concentration, the rate of formation of new nuclei is decreased and there is an increase in the rate of growth of crystals.21 This is borne out by the experimental results by the increase in particle size with increase in concentration (Fig. 3b) With increase -in temperature, the effective thickness of the diffusion layer is decreased, and this favors an increase in particle size (Fig. 4b). With increase in the pH of the bath, the cathode film solution tends to go far in the alkaline region and deposits hydroxides. This causes the deposit to be fine-grained and spongy (Fig. 5b)o It may be mentioned that in all the ecases, as the particle size increases, the powder becomes hard and compact and ultimately becomes coherent. This is well illustrated when the interval of powd.r removal is steadily increased. Up to an interval of powder removal of one hour the powder falls freely from the cathode. At the end of two hours, it tends to cling to the electrodewhile if allowed to grow undisturbed for three hours, it forms an adherent plating. Being closely related to particle size and shape, (the dendritic or tree-like shape is peculiar to all electrolytic powders), the changes in apparent density are similar to those in particle size. However, when the powder is codeposited with an adherent plate, and similarly when the particle size is large; the particle shape becomes rouaded instead of the characteristic dendritic shape. In these cases, there are apparent irregularities in the nature of the curves of particle size and apparent density. Figs, 2b and 2d show that at 40 amp./sq.ft. current density. the powder is codeposited with a coherent deposit, and though the particle size is smaller than that at 50 amp./sq.gft., the apparent density is larger. A similar effect can be observed in Figs. 4;a and 4b. At 45~C, although the particle size is smaller, the apparent density is larger than at 35~C. 25 -

Specific Surfa-ce and Flow Rate Since flow rate and specific surface are related to apparent density, it may be noted that specific surface and flow rate curves are similar (Figs. 2c, 3c, Ic, 5c and 6d, and Figs. 2a, 3a, 4a) and the cure characteristics are just the reverse of those for apparent densityo ~Effect of Addition Agents Among the various addition agents tried, gelatin, saponin and peptone were found to have similar influences on the deposition of thepowders. These effects may be briefly summed up as follows: (a) The solution becomes viscous. (b) A large amount of stable froth is formed at the cathodes with the powder getting entrapped in hydrogen bubbles,collecting at the surface and thereby getting oxidized to a certain extent. (c) Hydroxides separate from the solution and washing the powder is difficult and slowo (d) The range of the powder is increased with the use of these reagents, Thus a good powdery deposit is obtained from a solution of 60 g/l iron content with the use of 0.05 percent gelatin.Similarly, a powdery deposit is obtained by using 0,25 percent gelatin in a solution containing 120 g/l of iron chloride. The particle size also becomes in — creasingly finer with increased additions of these reagents7 if other conditions are maintained constant, The addition agents are surface active substances and are adsorbed on the crystal nuclei, thus preventing their growth; the discharged ions are consequently, compelled to start new nuclei and the result is the fine —grained deposit. Sodium naphthalene sulphonate. glycerine and glucose were found to be quite suitable and produced no froth. The effect of these additipon' agents on the powder characteristics is given in Table VIII. Special mention may be made.of sodium naphthalene sulphonate. A concentration of 0. 05 0,1 percent improVes the flow cha-acteristics the aveage part.i le s ize and the' apparent density. Within- this range, the powder current efficiency is also high and increases as the pe-centage of the addition agent increases. g' 26

Vo CONCLUSIONS The following conclusions may be drawn from the present investigations (1) The range of conditions for direct deposition of iron in powdery form from ferrous chloride solution is narrow. For the conditions employed it was found that below a current density of 50 amp./sq.ft., the operation is not plausible because of a coherent deposit. An optimum current efficiency of 85.5 percent is obtained at 80 amp./sq.ft, Above this the current efficiency decreases due to an increase in the evolution of heat and hydrogen. (2) The important factor is the concentration of the bath. The optimum concentration is found to be 35 g/1 (though 30 g/l was also quite satisfactory). More dilute bath gives a spongy powder which presents difficulty in processing while a more concentrated bath gives a mixture of coherent and powdery deposit. (3) Coherent deposits are favored by low cathode current density, high temperature and high acidityo Lesser depletion of ions in the cathode film solution and higher rate of diffusion or diminishing of the thickness of the diffusion layer are factors leading to coherent deposits (4) The maximum value of powder current efficiency is obtained at 450C at a pH of 4.5 (the pH of the- electrolytic bath without any addition of acid), and when the interval of powder removal is long enough above 30 minm (5) The median particle size and the apparent density increases with iron concentration, temperature, intezral of powder removal and with current density up to an 80 amp, /sqofto beyond which these decrease with a further increase in current density. The flov rate and the specific surface decrease in close conformity with increase in particle size and apparent density. The powders codeosited with coherent platings show abrupt changes in all their characteristics. (6) A low concentration such as 0.05 - 0.1 percent of sodium naphthalene sulphate, glycerine and glucose, improves the flow properties, increases the average particle size, apparent density and the powder current efficiency. Gelatin, saponin and peptone, though they give better powdery deposits and increase the range of deposition are not suitable due to the formation of froth and due to washing difficulttes on account of the higer viscosity of the solution. - 27

PART II. - *~~..ARX!^ COI.POITIO.g IRON]&lT~CO PMWR BKAHGS

I. INTRODUCTION The production of porous bearings is a natural application of powder metallurgy since all its important processing methods lead to 3aaterials which are porous per se. The two principal uses of pores in metals are as storage space for a liquid auxiliary substance such as oil or grease lubricants, or as separating cells for impurity-containing liquids. To the latter category belong the porous filters, diaphragms and similar products, while the porous bearings and bushings constitute the former. To fulfill the purposes, the pores must be intercamiunicating in character as well as of carefully controlled shape, size and distribution. Pressing and sintering have been found to suit ideally to produce these very desirable properties. The pore volume incorporate in the bearings varies from about 15-40 by volume, but is usually in the neighborhood of 25%. Depending upon the bearing composition and the nature of oil, between 4-6% by weight of oil is needed for a pore volume of 25% to be satisfied. Although in certain cases provision has to be made to furnish lubrication from outside, in most applications the lubrication is produced from within the bearing itself. Mechanism of Self-Lubrication The underlying principle of self-lubrication is briefly as follows. When the motor is started, the rotating shaft brings the oil to the surface by a direct as well as an indirect way. Firstly, the rotating motion of the shaft causes a pumping actiono Secondly, as Bowden and Ridler22 have found out, the bearing surface temperatures shoot to surprisingly high altitudes,in some cases between 600~C to 1000C, even though the mass of the metal remains coolo Such frictional heat is the second aid to the flow of oil from pores to the surface. At the interface between the shaft and the bearing, the oil forms a protective film preventing direct metal*to-metal contacto A steady supply of oil is furnished from the: pores to the entire bearing surface, resulting in a continuous and uniform oil film, even at low rotating speeds. The thickness of the film is automatically regulated by the surface temperature which in turn is governed by the load and the rotating speed. Upon completion of the running cycle, the oil is re-absorbed into the interior of the bearing by capillary forces and the cooling of the bearing and the housing. However, a thin film is retained by surface tension, thus preventing metal-to-metal contact at all times. This alternate action is repeated every time the motor starts or stops and, during running, the bearing furnishes adequate lubrication for a long time without losses from dripping or leakage, - 28 -

Bearing Compositions Formerly, it was believed that in a bearing alloy, there should be a fine distribution of hard compounds in a relatively soft matrix* The basic theory for this was that the hard components would support the load and the softer material, wearing away slightly, would form a reservoir for oil. This has since been proved incorrect and today it is felt that any metal or alloy can function as a bearing so long as it possesses certain properties to a degree as outlined by Mougey23, viz. mechanical strength, bonding characteristics, high melting point, fatigue resistance, anti friction properties, nonscoring characteristics, conformability, embedability, thermal conductivity and corrosion resist."ee, Tin and lead form the bases of two important groups of bearing alloys; other metals such as copper, cadmium and aluminum are used in other types. Among the porous bearings and bushings, copper-tin compositions were originally used exclusively and are still used most commonlyA Their main applications have been in small motors of fractional horsepower'capacity. Hovweyer, they have an inherent weakness in their low strength so that frequent replacements are necessary on appliances of heavy loads. As a remedy to this, the iron base bearing is a more recent improvement and forms the subject of investigation in this thesis. These have distinct merits in their higher hardness and strength as compared to the bronzes and possess a favorable coefficient of thermal expansion which matches with that of the steel shaft. Langhammer compares the properties of bronze and iron-copper bearing materials produced by Amplex Division, Chrysler Corporation, U.S.A., and the results show that for the same specific gravity, the iron-copper bearing material has two and a half times the ultimate strength, one and a half times the Brinell hardness, three times the resistance to permanent deformation, and much smaller coefficient of thermal expansion although the porosity was about 5% less - than the bronze bearing material. Literature Survey For the manufacture of bearings by the powder metallurgy method, iron and copper powders are combined in various ways. Several workers have studied the methods of combination of these powders. their compressibility and the physical and mechanical properties of the bearing metals, 25 SchwartzkopDf pressed iron skeletons from powder and sintered to 85% density. Molten copper was then infiltrated in the skeleton and thermally treated to secure hardness. - 29

26 Tormyn patented a process wherein iron powder is pressed to the general shape and size, sintered at 20000F, copper powder applied to the surface and the whole repressed to compact the copper powder and density of the iron, and then resintered at 1900~F in a non-oxidizing atmosphere. 27 Truesdale in another patent employs a method to coat indti vidual iron powder particles with copper: Finely divided Cu20 is intimately mixed with iron powder and reduced at 150^600~C by hydrogen until the reduction is complete. Additions of about 1% CuCl was found to aid the complete envelopment. 28 Langhammer and Calkins prepared porous metal bearings using a mixture of 1.% copper and 90% iron such that the particles of iron are on the average ten times the particles of copper, and sintering the briquette in a non-oxidizing atmosphere at a temperature between the melting points of the two metalso Bosch29 found that the compressibility of iron-copper mixtures is improved by using copper powder such that its lumetric weight is smaller than that' of iron powder. 30 Kelly studied the influence of copper content and sintering time and temperature on the properties of iron-copper compacts and found that the tensile strength of a compact containing 10% copper sintered at 1200~C for ten minutes compared favorably over other copper contents with similar treatment, and was greater than that of pure iron sintered at any temperature for any length of time. The strength and the ductility of 10% copper compositions increase with time for a sintering temperature of 1100~C. At constant sintering conditions of 1100~C and one-half of an hour, both the strength and the ductility increase with copper content up to 30% the values being three times at 30% Cu-.Fe than at 5% Cu-Fe compositions. Northcott and Leadbeater3 engaged in similar investigations established a relationship between tensile strength and porosity of the compacts and found that for constant porosity, maximum strength was obtained with 10%o copper contents However the best physical properties and highest densities were obtained for 25% copper-iron ratio sintered in hydrogen for one hour at 1100OC when copper was in the liquid phase, 32 Chadwick, et al. working with different kinds of iron powders obtained results analogous to those of Northcott and Leadbeater. The optimum tensile strength and elongation values were obtained with 25% copper contents. It was found that iron powders of low compressibility (reduced) gave better physical properties than the readily compressible (electrolytic) iron pbwderso. - 30 *

24 Langhammer2 compared the physical properties of bronze and iron-copper bearing metals and found that for the same specific gravity, the iron.copper bearing material. has two and one-half times the ultimate strength thrice the resistance to permanent deformation, and one and one half times the Brinell hardness and a much lower coefficient of thermal expansion, although the porosity was 5% lower than that of the bronze bearing material. 33 Arataa found that iron-copper compacts compare very favor'ably in all physical properties such as ultimate strength, Brinell hardness, PV factor etc. over both pure iron and bronze compacts, - 31 -

IIo EXPEREr\NTAL The Raw Material The iron powder used in these investigations was prepared electrolytically as described in Part I of this thesis. The copper powder was also prepared electrolytically using acid sulphate bath and electrolytic copper electrodes. Except in the series of experiments where the particle size distribution is the ar:iable, the iron powder mix had the following size distribution: -48+ 65 mesh 11.1% -65 + 100 " 27.7% -100 + 140 " 27.7% -140 + 200 " 11.1% -200 + 325 " 11.2% -325 " 11.2% In general, the powder mix had 10% of copper and 2% graphite except in the series of experiments where these were the variables. These extra additions to iron powder were restricted to -325 mesh size. The size distribution of iron powder was chosen arbitrarily on the basis of the powder prepared electrolytically, as described in the previous experiments. The Die The die is fabricated from alloy steel containing: C 0.32% N. 4.10% Cr 1.30% M'o 0.20. After machining to a slight oversize, the die was hardened by the following treatment. The die was introduced in a furnace at 820~C and maintained at that temperature for 45 minutes, On removal from the furnace, it was rapidly cooled in an air blast. It was then tempered at 250~C and air cooled. This treatment gave the die material a tensile strength of about 100 tons/ sq..in. The die surface was then ground to a mirror-like finish to a close - 32'

limit of tolerance. The figure shows the setup consisting of five partso The upper punch A is the movable part, which transmits the pressure and forms the upper end of the briquette and ejects it from the die after compression. The stationary parts are the die B which forms the outer diameter of the bearing, the core rod C which forms the inside diameter, and the base D which in conjunction with the ring E constitutes the lower punch and forms the lower end of the bearing. Briquetting The core rod C is screwed to the base D, the ring E slid over the core rod onto the base and the die B placed on it. The powder is filled in the cavity so formed and the punch A brought down to compress the briquette. The guide pins insure that the punch is in alignment with the assembly. After compression, the base is unscrewed and the punch brought down further to eject the bearing compact. The green bearing had an inside diameter of 1.256" and an outside diameter of 1.763" compared to the diameter of the core rod of 10253" and the die diameter of 1,760". Iepending on the conditions of the experiment, the length of the bearing varied between 0.53" to 0,75" for a weight of about 60 g. Sintering Sintering of the green compact was carried in an atmosphere of hydrogen stripped of all its oxygen content by passing over copper turnings at 600~Co Bearing Characteristics Green Density Green density or the pressed density is the ratio of grams to cubic centimeters of a briquette compacted to specified dimensions by a designated pressureo Sintered Density It is the density of the briquette after the sintering treatment 33' ~

Oil sImpreat_.ed Density It is the density of the briquette fully impregnated with a lubricant such as.oil. The oil impregnation may be carried out with the help of either heat or pressure. As specified by AoSoToMo341 the following procedures may be employed: (1) The specimen shall be immersed for at least four hours in oil (viscosity approximately 200 secso Saybolt Universal at 1000F) held at a temperature of 180 + 10~F and then cooled to room temperature by immersion in oil at room temperatureO (2) The pressure over the specimen after immersion in oil at room temperature shall be reduced from atmospheric pressure to not more than 2" Hg after which the pressure shall be allowed to increase to atmospheric in 10 minutes, the specimen remaining immersed in oil for the whole period. The bearing can be fully impregnated with oil by any of the procedures. However, the second method was employed because of its convenience. The oil impregnated density is calculated as follows:34 B B-C D Density g/c.c. B - Weight of the oil impregnated sample C Weight of the sample suspended in watero A.SoT.M. standards specify that the density calculated as above should lie between 508 and 602 g/c.C' for the iron-copper bearings.34 34 -

Porosity The quality resulting from a multiplicity of pores distributed in a compact and measured as the percent voids per unit volume is called porosity. Porosity is the basic property in powder metal bearings. The pores, being intercommunicating, serve as a reservoir for the lubricant. Obviously, the physical properties of the compact are impaired considerably by an inherent porosityb The effect is not proportional to decrease in cross-sectional area, but is largely increased by the multitude of pores, each acting as an individual source for stress concentrations and notch effects. Hence, for each application a balance must be struck between the physical properties necessary and the maximum porosity possible. Thus, for example, a bearing which is to be subjected to heavy loads at high temperatures requires a high porosity of about 35% and a specific gravity of about 5.6. Here a coarse powder has to be employed with the resultant large average pore size. For a bearing to be used for high speeds, light loads and low temperatures, a high porosity and large pore si-e would result in excessive losses of lubricants In such a case, a low porosity and.comparatively a smaller pore size is the correct choice and the powder will have to be of a finer size. No doubt, the lubricant plays an important part. A heavy oil with a high viscosity will be the most appropriate in the former and a light oil with a lot viscosity will suit the latter applications 34 The porosity is calculated as follows: p^ oi^B-A x 100 (BtC)x s P - Porosity percent A Weight of the sample B = Weight of the oil impregnated sample C = Weight of oil impregnated sample suspended in water s = Specific gravity of the lubricant AoS.TM, Standards specify that the porosity calculated as above must not be below 18% by volumeo - 35 -

Radial Crushing Strenth For satisfactory working, a bearing must possess a certain compressive strength. A measure of this is the radial crushing strength which is determined by compressing the test specimen between two flat surfaces, the direction of the load being normal to the longitudinal axis of the specimen. The point at which the load drops due to the first crack is the radial crushing strength.34 According to A.S.T.M. specifications34 the radial crushing strength determined as above shall not be less than the value calculated as follows: p KLT2 D-T where P = Radial crushing strength in pounds D = Outside diameter of the bearing in inches T Wall thickness of bearing in inches L = Bearing length in inches K = Strength constant. For iron-copper bearings, the value of K is 40,00Oo In the present series of investigations, the length of the bearing varied from 0.55" to 0.75, depending upon the operating conditions like the briquetting pressure and the composition of the powder mix. The radial crushing strength of a bearing varies directly with the length. Thus, comparison of the strength of the bearings is possible only if the bearing length remains the same in all cases. Hence, in the representation of re-. suits, the value of the strength is corrected to that corresponding to a bearing length of 0.57'". This value of the length was chosen because it gives a projected area (outside diameter times length) of'one square incho 36 -

III. RESULTS General Plan of Work A detailed study was made of the influence of (1) Briquetting pressure (2) Sintering temperature (3) Sintering time (4) Copper content (5) Graphite content (6) Particle size on (a) Density (b) Porosity (c) Strength (d) Changes after sintering in density, length and diameter. The following table gives the values of the variables kept constant and their extreme limits of variation. VARIABLE FIXED VALUE LIMITS OF VARIATION Briquetting pressure 20 tons/sq.oin 5-35 tons/so inD Sintering temperature 1100 oC 900-1300 ~C Sintering time 1 hour. -8 hours Copper content 10% by wt. 0-30o% Graphite content 2% by wt, 0-6% Particle size 35%(-48+100 mesh) 0-100% - 37 -

1-7 --— _700_ 550 0 I II ^400 --- -------- o z _j ~~~~~~~~~~w 0*9 I-1 0 0.~9 -- ^ —------- /1rr 0 ^2 0 --- -----------— 250 0 -0100 —04Q 4.5 5.0 5.5 60O GREEN DENSITY b d 6_5_-__ 40 -d______ OIL IMPREGNAT.ED - 6.0 ---- --- aJ^^ — 35 ISNTERED 5.50 —------- as 30 —GREEN. W 4-0 I -- -- ---------------— ^ 0O 10 20 30 40^O 10 20 30 30 BRIQUE'TTIN6 PRESSURE. tons / sq in FIG I EFFECT OF PRESSURE ON BEARING CHARACTERISTICS

I100.0w 9 w 10 w w 0 0 z w >-8 w 9 z I w 0 W0 7 8 4 Q. 3'0 " 2-8 - Or) I1 Id w 2-5 -0- 2'6 bJw C.. 2-0 2-4 0 10 20 30 40 0 10 20 30 40 BRIQUETTING PRESSURE, tons /sq in FIG 2 EFFECT OF PRESSURE ON CHANGES AFTER SINTERING

TABLE I EFFECT OF BRIQUETTING PRESSURE Copper 10 percent Graphite 2 percent Sintering time - 1 hour Sintering temp. 11000C under hydrogen Briquetting Density g/c.oC Porosity Radial Changes after sintering percent pressure Green Sintered Oil percent ushing Density Length Diameter Volume tons/sq. in. nated strength increase shrink- shhrinklbs o age age age 5 4.10 4,49 4.83 40.36 103 9.90 2,51 3.31 10.7 10 4.61 5.03 5.33 33.5 230 9.11 2.56 2.95 10.5 15 4.90 5.35 5.63 28.30 385 9.20 2.69 2.85 10.0 20 5,14 5.61 5.98 25.36 510 9.16.2.70 2.65 9o91 25 5.35 5.81 6.o 04 22. 84 580 8.68 2. 78 2.94 9.37 30 5.50 5.94 6.14 21.o06 630 8.00 2.85 2.25 9-10 35 5.62 6.06 6.26 20.1 700 7.83 2.91 2.15 8.90 The results are presented graphically in Fig. 1 and Fig. 2

d 6.1 _ 28 z 5.9.........,iw 26 ___ _ 5.7 24 ________ w 0'z 5.5....... -22.. bc 5~0 72 51 3C 5 62 5 H a: \ /S C 30 52 5 LENGTH 2 5 DIAM TER 2 2'5 4251 900 1000 1100 1200 1300 900 1000 1100 1200 1300 TEMPERATURE ~C FIG 3 EFFECT OF TEMPERATURE OF SINTERING ON BEARING CHARACTERISTICS.

TABLE II EFFECT OF TEMLPERATURE OF SINTERING Briquetting pressure, 20 tons/sqoin, Copper, 10 percent Sintering time, 1 hour in hydrogen Graphite, 2 percent Temperature Density g/c.c. Porosity Radial Changes after sintering percent of sintering Green Sintered percent c|shing Density Shrinkage Shrinkage ~C istr th increase in length in diamet 900 5,14 5o71 24.5 330 10.9 3.3 6304 1000 5.14 5o68 25,02 355 10.05 2.94 2,32 1100 5o.14 5.61 25.36 510 9.45 2.76 2.65 1200 5.14 5.63 26.16 625 9.5 2,98 2.78 R3 1300 5.14 5.91 24,2 715 14.83 3.60 3-18 The results are presented graphically in Fig. 3

0' 5.87- -------— 5751 z 5-7 - _____ —-------- _ 525 0 IU I — w w z b d 4-0 28 z 35 27. 0 H 2 Z 0 SO 2 4 6 860 2 4i6 FIG 4. EFFECT OF TIME OF SINTERING ON BEARING CHARACTERISTICS 2'0~0 2 4 6 826 TIME2~ HOURS FIG 4. EFFECT OF TIME OF SINTERING ON BEARING CHARACTERISTICS

TABLrE III EFFECT OF TIME SIITERIIIG Briquetting pressure, 20 tons/sq.oin. Copper, 10 percent Sintering temperature, 11000C under Graphite, 2 percent hydrogen Sintering Density g/Co.C Porosity Radial Changes after sinterng pe.nt time Green Sintered percent crushing Density Diameter Length hours strength increase shrinkage shrinkage lbs 2 514 5.71 25.12 440 10.9 3.25 3.61 -I" 1 5.14 5o61 25.536 510 9 15 2.98 2,65 2 5.14 5.61 26o3 535 9.02 2.52 2.70 4 5.o14 5.52 27.8 560 7o4 2-22 2.80 8 5.14 5059 26.5 480 8.75 3.12 3-3 The results are represented graphically in Fig. 4

o - -I 8OC — - C - - I 6-19 6700 ~~z _1 ~~~~~~~ y /^t~~~~~~~~~~~~~~~~~~~~~~~~-.f z:o b I — z 5-7.- I-7 500 b' H z >-./HLi 2 5 2 ~ 22 rJ3 a. Z 0 1 I i I I 0 C,) Z 51'- 10 15 20 25 30 -5 10 15 20 2 COPPER CONTENT % FIG 5 EFFECT OF COPPER CONTENT ON BEARING CHARACTERISTICS

17 ---- 0 ____ I -------- ------- 17 0 w 15'o " 15 <[C/)~~~~~~ w w 13 w 1 ^ c/i. w 5C ____ —b__ ___-C-__-___ - b d Z/ 40Z 45 w H w / z,- ( 3. 4- 0. w z 250 5 10 15 20 25 30 5 10 15 20 25 30 COPPER CONTENT % FIG 6 EFFECT OF COPPER CONTENT ON CHANGES AFTER SINTERING Ing ~ III

TABLE IV EFFECT OF COPPER CONTENT Briquetting pres:sure, 20 tons/sqoin. Graphite, 2 percent Sintering time, 1 hour Sintering temp., 1100~C under hydrogen Copper Density g/c. c. Porosity Radial Changes after sintering perent content Green Sintered Oil Impreg- percent crushing Density Length Diameter Volume percent nated strngth increase shrink- shrink- srink lbs. age age age 0 50P035..603 5.82 24.8 400 11.2 3.57 3.4 12.7 5 5.11 5.60 5.87 25.1 450 9,7 2.80 2.85 10 5 10 514 5.61 5.98 25.36 510 9.16 2.70 2.65 9.91 15 5,20 5..77 6.01 23.7 565 10.9 3.2 3.15 1.0 20 5.27 5.85 6.o4 23.2 615 11.1 3.40 3.57 11.6 25 5.31 6.05 6.23 21.3 715 13.8 4.10 4.10 13.9 30 5.38 6,28 6.44 17.6 840 16.6 4.85 4.76 15.9 The results are plotted in Fig. 5 and Fig. 6

0CCr~~~~~~~~~~~~~~~C) w I ^ 5-3 ----- -— 350 z C'' 5-i- 250 - co 4.30 0 2 og~ 28 LI DIAMETER z 0:] IE ILENGTH o -2 26 0 2 4 6 0 2 4 6 GRAPHITE CONTENT % FIG 7 EFFECT OF GRAPHITE CONTENT ON BEARING CHARACTERISTICS

TABLE V EFFECT OF GRAPHITE CONTENT Briquetting pressure, 20 tons/sq.in. Copper, 10 percent Sintering temperature, 11000C Sintering time, 1 hour under hydrogen Graphite Density g./e.c. Porosity Radial Changes after sintering percent content -----— percent crush — percent Green Sintered ing Density Diameter Length strength increase shrinkage shrinkage lbs.. 0 5.i4 5554 29-6 545 7.7 2,53 2.9 1 5.14 5.60 27.0 525 8.9 2,84 3.35 %lo 2 5.14 5.61 25.36 510 9.15 2.70 2.65 3 5.14 5.59 25.8 440 8.7 3.30 3.00 4 5.14 5.43 26.4 340 5.6 2.94 2.40 6 5.14 5.24 26.7 270 1.9 -0.6 2.05 The results are presented graphically in Fig. 7

6.0 1200 C o5.6 400 O SINTEREGN TED 5.4 __... I I. >5.1 ____ i i-2625 5.3 28 5 d ~ 0 PARTIE 4 _____.1 _____ _____. 225 _____ _2.5 20 15 10 5 0 -. 25 20 15 10 5 75 60 45 30 15 -200 75 60 45 30 15 0 PARTICLE SIZE FIG 8. EFFECT OF PARTICLE SIZE ON BEARING CHARCTERISTICS.

20 20 0Ne w II c 1-5 -- 15 w w _ o H 10 -- 10 5 5 b d 8 1 8 \ ___ ___ i — - - - 7 s \^0. 0 6(,'0 z 5 - 5 z, 3 Z- 3.I -J l — I —----- 0 20 40 60 80 1004. 48+100 0 20 40 60 80114 25 20 15 10 5 0_ -100+200 25 20 15 10 5 PARTICLE SIZE FIG 9. EFFECT OF PARTICLE SIZE ON CHANGES AFTER SINTERING

TRjLOIAZ Vi EFLFECT OF VARIATION IN PARTICLE SIZE Briquetting pressure, 20 tons/sq. in. Copper, 10 percent Graphite, 2 percent Sintering time, 1 hour Sintering temp., 11000C under hydrogen Percentage Density g/c.C. Porosity Radial Changes after sintering percent.48+0oo Green Sintered Oil Im- percent crush- Density Diameter Length Volume mesh iron pregnated ing increase shrink - shrink- shrinkpowder strength age age age lbs. 0 4.81 5,78 6,00 22,5 1000 20.1 6.27 7,27 19.95 20 5.00 5.62 5.86 25.8 760 12.4 4.90 5.35 16.05 40 5o07 5.57 5.83 26.8 600 9.85 3-95 4,0 12.73 60 5.l8 5.57 5.82 27.0 440 9.65 3.24 3.1 10.0 80 5.15 5.50 5.76 28.05 330 6.8 2.32 2.02 8.9 100 5o22 5-50 5.75 28,55 310 5.36 1.92 1.51 76 The results are presented graphically in Fig. 8 and Fig. 9.

IV. DISCUSSION OF THE RESULTS 1. Green Density The effect of the different variables on the densities of the compacts is shown in Figures lb, 5b and 8b. From Fig. lb, it may be seen that the density steadily increases with increase in briquetting pressure. This is natural since the closeness and the area of contact of the powder particles are increased by the applied pressure and this is an index of the density of the compact. It may also be noted that as the briquetting pressure increases, it tends to exert a lesser and lesser influence on the green density. Thus, for a 10 tons/sq. in. rise in pressure from 5-15 tons/sq. in. the density increase is 0.8 g/c.c. from 4.1 to 4.9 g/c.c. Between 15-25 tons/sq. in. increase in pressure, the density increases by 0.45 g/c.c. from 4.9 to 5.35 g/c.c. and correspondingly between 25-35 tons/sq. in. The increase in density is only 0.27 g/c.c. from 5.35 to 5.62 g/c.c. This happens because at low pressures, densification occurs more by redistribution of the particles and reorientation into strata, while the frictional resistance is comparatively low. At high pressures, the free space in the compact decreases, and densification takes place by a smoothing of the crevices and filling of the voids through plastic deformation. In this phase, the die wall friction becomes an appreciable factor, and the energy expended is comparatively very large and increases beyond proportions as the density of the compact approaches the theoretical value of the metal. From the relationship between briquetting pressure and the corresponding decrease in the volume of the compact, Balshin35 arrives at the conclusion that if the height and the relative volume of a compact decrease in arithmetical progression, the pressure must increase geometrically. In other words, expressed as an equation: log p = Ld + C where d is the density and L and C are constants depending upon the nature of the powder and conditions of compacting. Hence, if log p be plotted against d, the nature of the curve should be a straight line. Fig. la illustrates the linear relationship and endorses Balshin's equation. - 53 -

Figo 5b illustrates the increase in density with increase in copper content of the compact at constant briquetting pressure. The density increases from 5.03 g/c.c. at 0% to 5o38 g/c.o. at 30% copper content. Copper being a metal with a face-centered cubical lattice is easily deformable and is not subject to work-hardening at room temperatures. Iron, on the other hand, being a metal with a body-centered lattice has a greater resistance to deformation and has a great work-hardening tendency. Accord-: ingly, lower pressures are required for coherent compacts of copper than for those of ironand correspondingly for the same applied pressure, copper compacts are denser than those of iron. Thus, greater the copper content, greater is the density of the iron-copper compacts. The effect of the particle size on the density is plotted in Fig. 8bo It is seen that the density decreases as the particle size decreaseso This is because the pressure is absorbed primarily by the shifting of particles and deformation is only secondary. Thus, the finer sizes will give a lower density than the coarse sizes for the same pressure. However, although the density is low in the fine-sized compacts, the number of contact points is larger as can be seen from their higher strength. 2o Sintered Density and the Changes After Sintering On sintering there is an increase in the density of a powder compact which is brought about by a reduction in the pore volume as well as the over-all volume. With the application of heat, sintering starts at the metal-to-metal contacts and the surface tension forces operate in a direction to minimize the surface areas. The intensity of the surface energy being less. at the smooth surfaces than at the crevices, transfer of the metal takes place, thus decreasing the over-all volume and increasing the density. Figo 2 shows the effect of the compacting pressure on the changes after sinteringo With increasing pressures, the rate of increase in density and the percent of shrinkage continually decrease. With increasing compacting pressures, the microscopic, and perhaps molecular, irregularities are lessened, and consequently during sintering,the surface energy differences being lower, the dimensional changes are also smaller With increase in copper content, the sintered density increases the same way as green density (Fig. 5a). However, the percent increase in density after sintering is 11.2% for 0% copper content, - 54 -

decreases to 9.16% for a 10% copper content and then continuously increases to 16.6% increase for a copper content of 30%. The same trend is seen in all the other changes after sintering, viz., that there is a minimum shrinkage in compacts having a 10% copper content, The increase in shrinkage with increase in copper content may be explained by the fact that the liquid phase largely removes the surface films and other irregularities which would otherwise obstruct:sintering and crystillization. The rounded grains of the solid metal form closer bonds with the liquid phase because of the increased mobility. The surface tension forces of the liquid also contribute to the formation of a closely knit structure. The effect of particle size on the sintered density and on the changes after sintering is interesting; whereas, with increasing particle sizes the green density increases, the sintered density increases with decreasing particle size. This is illustrated in Figs. 8a and 8b. The change in density is particularly sharp between 0% to 20% of 100 mesh size content. The shrinkage in the diameter and the length also show the sa;me trend as may be seen from Fig. 9. The increased shrinkage with increasing fineness of the powder mix is obtained, because, in the case of the fin sizes, during pressing, densification takes place more by redistribution of the particles than by def ormation., so that during sintering there is a greater allowance for dimensional changes. With coarse powders there is a large amount of deformation and cold working.during deformation itself. Secondly, with the greater surface area of the finer particles, the surface tension forces are more active. The third, though perhaps a minor factor, is that finer powders adsorb larger amounts of gases than the coarse grades. During sintering, as the gases force their way out, many new "clean" surfaces are produced. which again takes place on a larger scale in fine-sized compacts. 3. Porosity and -Strength Porosity and strength are inter-related properties in a porous powder compact. Increase in porosity causes a decrease in strength. With increase in compacting pressure, the porosity decreases uniformly from 40.436% at 5 tons/sq.in. to 20.1% at 35 tons/sqoin. pressure. Correspondingly, the radial crushing strength increases from 103 lbs. at 5 tons/sq.in. to 700 lbs. at 35 tons/sq.in. pressure. The pressure affects the strength by increasing the number of contact points so that the sintering progress is accelerated and recrystallization and grain growth proceeds faster (Fig. 1 cd). -55

Increase in copper content of the powder mix decreases the porosity. Copper being more soft and deformable, it facilitates particle redistribution and during sintering, the liquid phase produces a wellknit structure. Strength increases with increased copper content. This is due to the higher degree of densification of the structure by close approach to individual crystallites and due to the increased grain growth, because of the removal of surface films by the liquid phase (Fig. 5c,d, ) With graphite additions, there is a decrease in the porosity from 29.6% at 0% to 25.36% at 2% graphite content, and again there is an increase to 26.4% at 6% graphite content. However, the strength decreases steadily from 545 lbso at 0% to 270 lbso at 6% graphite content (Figo 7c,do)o Graphite additions improve the wear characteristics of the bearing and act as a lubricant during compaction of the powder. With increased fineness of the powder mix, the porosity decreases and the strength increases. The increase in strength may be attributed to the larger surface area of the fine powders and the greater metal-to-metal contact (Fig. 8 c,d)o 4. Effect of the Temperature and the Time of Sintering on Bearing Characteristics Figs. 3 and 4 illustrate graphically the effects of the temperature and the time of sintering respectively on sintered density, the shrinkages in length and diameter, the strength and the porosityo It is known that neither the temperature nor the time affects markedly the "sintering forces" so that their main function is to minimize the effects of the factors obstructive to sinteringo Jones5 points out that the most important effect of temperature in aiding sintering is the increase in the plasticity of the metal. In the same way, the influence of time on completeness of sintering is a matter of plastic flow.. More plastic the metal, the more it will flow in a given time, and in the same way, more of the metal of any given plasticity will flow for a greater length of time. Accordingly, for monometallic compacts, it is usually found that with increasing temperature and time of sintering, there is increased shrinkage and a reduction of porosity in a powder compacts In the present investigation, however, it is found that shrinkage decreases with increase in temperature between 9000C to 11000C, and increases with further increase in temperature Porosity increases with temperature from 24.5% at 900~C to 26.16% at 1200~C, and decreases to 24.5% as the sintering temperature -56

is further raised to 1300~Co It is interesting to note that from 1100~C to 1200~C, there is an increase in shrinkage as well as a small increase in porosity. In the same way, porosity which is 25,1% for - hr. sintering time increases to 27.8% for 4 hrso and again decreases to 26.6% for an 8-hour period. 36 Schecht et al. studied the density changes of loose iron powders of various origins and found that there is a marked drop in density starting near the o(- Y transformation point and reaching a minimum at 1000~Co Libsch,et a137 did similar work with iron powder compacts and obtained analogous results, The second factor which influences the density changes and the porosity and shrinkage is the influence of gases and vapors. In a powder compact, the potential sources of gas are: (1) adsorbed films; (2) gases evolved by chemical action on heating; (3) mechanically entrapped airo It is unlikely that the third source plays any decisive part in the changes in porosity after sinteringo Ruer and Kuschmann38 have shown that the amount of gas adsorbed by metal powders even at room temperature is considerable. Working with copper powder prepared from cupric oxide by reduction with hydrogen at 750~C followed by evacuation at 440~C, they found that it gained 5o44 mg/100 g? within one hour. Similarly, reduced iron powder gained 20.43 mg/100 g, in the same time, The second source of gas,vizo that produced by chemica-l reaction or decomposition(Ex: by reduction of oxide films by hydrogen)may be the major factor at higher temperatures. If these gases escape before sintering and consolidation take place, it will be harmless, whereas, if they are evolved after consolidation, the gas pressure being considerable, an increase in the number and the size of the pores will be the result. This affords a satisfactory explanation for the observations in the present investigations. At 900~C and onwards, sintering has progressed sufficiently and the shrinkage decrease takes place due to either c - Y transformation or the gas evolution or both in conjunction. Above 1000~C when the decrease in shrinkage and increase in porosity still continue, the only affecting factor is the gas evolution. By 1200~C the gas evolution is complete so that, due to the removal of oxide films, and due to the increased plasticity, there is a decrease in porosity and an increase in density at 1300~Co An additional factor is the presence of copper in the liquid phase which densifies the structure. The steady increase in the radial crushing strength of the bearing compact, in spite of the irregular changes in porosity and density, focuses ones attention on the liquid cement theory of sinteringo Theoretically, an increase in porosity should cause a decrease - 57

in strength, and this is observed true for monometallic compactso However, in the present case, the increased plasticity of copper due to temperature. more than counterbalances the negative effects of-increased porosity. The explanation for the apparent irregularities in shrinkage with increasing temperature applies equally well to explain the irregularities in shrinkage due to increased time of sintering It appears that four hours is the optimum time, a point when the gas evolution is complete, the shrinkage minimum and the strength maximum, If the time of sintering be increased to eight hours, there is a decrease in porosity, accompanied by a decrease in strength, which is more probably due to grain growth taking place when the specimens are heated for a long time at a high temperature as 1100~C. 5 Operating Conditions to Obtain Bearings Satisfying A, SToMo Specifications The density limits specified for iron-copper bearings fully impregnated with oil or lubricant are 5.8 and 6.2 g/c.c. All the bearing compositions tried have densities lying within this range except those obtained by using a briquetting pressure below 20 tons/sqOin., copper content of more than 25%, graphite content above 3% and having more than 60% of -48 + 100 mesh size, The porosity need be above 18% by volume, All the bearing compositions tried satisfy this condition, except those having more than 25% copper contentso The radial crushing strength should not be below the value calculated as follows: P- EKLT2 D-rT P = radial crushing strength in lbso. L = bearing length in inches T = wall thickness of bearing in inches D = outside diameter of beating in inches K = constant the value being -40,000 for Fe-Cu bearings, 58 -

For the standard length of 0.57" of the bearing, the radial crushing strength calculated from the above formula is 950 lbs. The highest value of 1000 lbs. for the radial crushing strength is obtained when the bearing composition has 0% -48 + 100 mesh particle size content. In this case, the density is 6.00 g/c.c. and the porosity 22.5%. With the particle size content of 20% -48 + 100 mesh, the strength is 760 lbs. with a density of 5.86 g/coc. and a porosity of 25.8%. In these cases, the briquetting pressure was 20 tons/sqoin. Using a higher briquetting pressure (30-35 tons/sq.in.), the bearing strength can be increased sufficiently to obtain a value much higher than the minimum necessary, meanwhile maintaining the density and porosity within proper limits. It has also been found that higher sintering temperature increases the strength without decreasing the porosity (or increasing the density) to an appreciable extent. Thus, increasing the sintering temperature from 1100~C to 1300~C increases the strength by 40%, while the density changes only by about 6% and porosity decreases by about 1%. Similarly, increase in the time of sintering up to 4 hours increases the strength accompanied by a small increase in porosity. Increase in the copper content also increases the strength. Thus, from 10% copper to 25% copper, the strength increases by about 40%o, density by about 4%, while the porosity decreases by about 4%. Thus, a combination of these operating variables can be utilized favorably to obtain bearings satisfying AoS6T4Mo specifications. It may then be said that to obtain bearings satisfying AoSoToMo specifications, the following conditions may favorably be employed. Briquetting pressure 20-35 tons/sq.in. Sintering temperature 1100~ 1200~ C Sintering time 1 - 4 hours Copper content 10 - 25% Graphite content 0 - 2% Particle size distribution: Fine sized with as little +100 mesh particles as would be possible. - 59 -

VO CONCLUSIONS The following conclusions may be drawn from the present investigations: 1L Pressure Increase in pressure causes an increase in green and sintered density of a powder compact. With regard to green density, Balshints relation between pressure and relative volume of a compact that log p = LV + C is found to be true within limits of experimental error for the pressure range studied in these investigationso The shrinkage after sintering decreases with increasing pressures. Porosity decreases and strength increases with increasing pressures so that a compromise must be effected between the strength necessary and the porosity possible for any given compacting pressure. 20 Sintering Temperature The strength increases with increased temperatures of sintering. From 9000C to 1200~C there is a decrease in the density of the bearing and an increase in porosity. The minimum shrinkage is obtained in the temperature range of 1100~C to 12000C. The increase in porosity with increasing temperature is most likely due to the gas evolution as the temperature increaseso At 1300~C, there is a marked increase in shrinkage and a decrease in porosityo 30 Sintering Time The time and the temperature of sintering act in the same direction. The optimum time seems to be 4 hours, when there is a maximum porosity and maximum strength. 4. Copper Content Increase in copper content causes an increase in density, a decrease in porosity and an increase in strength. The shrinkagehowever, shows a minimum for a copper content of 10% and is larger on either side, Copper in the liquid phase removes the surface films and promotes sintering which effects in the higher strength and lower porosity. - 60 -

5. Graphite Content Addition of graphite rapidly decreases the strength,. The porosity decreases up to 2% additions and increases with further ad. ditionso The shrinkage characteristics also show a maximum between 2 to 3% graphite contents. 6. Particle Size The finer the particle size, the smaller the green density. This is because the pressure is accommodated more by redistribution in the fine powders than in the coarse powders where deformation is a large factor. The same thing causes a much larger shrinkage for fine-sized compacts. A compact having 0% + 100 mesh powder has a 20% increase in density and the same decrease in volume, whereas a compact containing totally + 100 mesh powder shrinks only by 7.46% in volume and increases by 5.36% in density. This is of great concern in the matter of die design. The porosity decreases with increased fineness of the powder. The finer the particle size, the greater the strength of the compact. This is particularly noticeable between 40% and 0% (+ 100 mesh) contents. Hence, it is most desirable that a compact should have as little of + 100 mesh powder and as much of the finer sizes as would be possible. To obtain iron base bearings satisfying AoSoToMo specifications a the following conditions may be favorably employed. Briquetting pressure: 20-35 tons/sq.ino Sintering temperature 1100-1200~ C Sintering time 1 4 hours Copper content': 10-25% Graphite content 0 0 2% Particle size distribution: Preferably fine sized with as little + 100 mesh as would be possible. 61

REFERENCES 1. Crane, E. V., and Bureau, A. G., Trans. Electrochem. Soc., 85, 63 (1944) 2. Baeza, W. J., A Course in Powder Metallurgy, Reinhold Publishing Corp, New York (1943) 3. Patch, Eo S., Iron Age, 146, No. 25, 31 (1940) 4. Balke, E. S., Symposium on Powder Metallurgy, ASToM., Philadelphia, (1943) 5. Jones, W. D., Principles of Powder Metallurgy, Arnold, London (1943) 6. Kroll, W. J., UOS.Po, 2,413,411, Dec, 1946. 7. Stoddard, W. B., U.S.P., 2,420,403 8. Wranglen, G., Jo Electrochem. Soc., 97, 353 (1950) 9, Casey, H. M., U.S P., 2,481,079, Sept, 1949 10o Cudd, H. Ho and Freeman, F. J., U.S.P., 2,533,096, December, 1950. 11. Primavesi, D., Swiss Po 266,686, May, 1950 12. Selected Government Research Reports, Powder Metallurgy, Vol. 9; Report No. 4, H. M. Stationery Office, London, (1951) 13. Williams, R, S. and Homerberg, Vo 0., Principles of Metallography, 4th Edition, McGraw Hill, New York (1939). 14. Tamann, G., and Jaachs, H., Z. Anorg. Uallgen Chem., 227, 249, (1936) 15. Joshi, M. V., Modi, H. J. and Tendolkar, G S., J. Sc. Ind. Res., 10B, 48 (1951) - 62 -

16. Merck, Z,, Anal. Chem., 41, 710 (1902) 17. AoSoToMo Standards Specification, B212-48, Part 2, 943 (1952) 18. Lea, F. M., and Nurse, R. W., J. Soc. Chem. Ind., 58, 277T (1939) 19. Carman, P. C., J. Soc. Chem. Indo, 57, 225T (1938) 20, Graham, A. K., Trans. Electrochem. Soco, 52, 157 (1927) 21. Gluzunov, Z. Physik. Chem., 167, 399 (1934) 22. Bowden, F. P. and Ridler, K, E., Proc. Roy. Soc. London, 154, 640-56 (1936) 23. Mougey, H. C., Proc. Am. Insto Petr., III17, 42-49 (1936) 24. Langhammer, A. J., Machinery, N.Y,, 51, No. 8, 152 (1945) 25. Schwartzkopf, P., Metal Progress, 57, 64-68 (1950) 26. Tormyn, H., UoSoP. 2,299,192, Oct. 1943 27. Truesdale, E., U.S.P. 2,483,075, Sept. 1949 28. Langhammer, A. J., and Calkins, W., U.S.P. 2,229,330, Jan. 1941 29. Bosch, Ro, Br. P. 491, 415, Sept. 1938 30. Kelly, F. C., Iron Age, 158, No. 7, 57-60 (1946) 31. Northcott, L., and Leadbeater, C. J., Symposium on Powder Met. Iron and Steel Inst. London, Sp. Report No. 38 42 (1947) 32. Chadwick, R., Broadfield, E. R., and Pugh, S.F., ibid., p. 151 33. Arata, W. H., Product Eng., 15_, 561 (1944) 34. A.S.T.M. Standards Specifications, B202-51T, Part 2, 931 (1952) - 63 -

35. Balshin, M. Yu., Vestnik Metalloprom, 18, No. 2, 124 (1938) 36. Schecht, L., Schubart W., and Duftschmid, F., Z. Electrochem., 37, 485 (1931) 37. Libsch, J., Volterra, R., and Wulff, J., Powder Metallurgy, Am. Soc. Metals, Cleveland (1942) 38. Ruer, R. and Kuschmann, J., Z. Anorg. Uallgen. Chem., 154, 69-78 (1926) 166, 257-74 (1927) 173, 233-61 (1928) - 64 -