CONFIDENT IALr ENGINEERING RESEARCH INSTITUTE i' SFIEI " UNIVERSITY OF MICHIGAN _ ANN ARBOR DEVELOPENT OF A NEW-TYPE STORAGE BATTERY FOR MILITARY USE September 15, 1947, to December 31, 1953 By L. L. CARRICK Director of Project ERI PROJECT M743 for U.S. WAR DEPARTMENT, ORDNANCE DEPARTMENT Contract W-20-018-ORD-13024, Project 610 ERI PROJECT M875 for U.S. WAR DEPARTMENT, ORDNANCE DEPARTMENT Contract DA-20-018-ORD-4181 RAD NO. 0-11333 ERI PROJECT 2018 for U.S. ARMY, ORDNANCE CORPS Contract DA-20-018-ORD-12112, RAD NO. 1-12748 January, 1954 uL. L- SIFHPIE ' NTIAL

CONFIDENTIAL oHED TABLE OF CONTENTS Page LIST OF TABLES v LIST OF PHOTOGRAPHS vii LIST OF GRAPHS ix SUMMARY xi PERSONNEL OF PROJECTS SINCE INCEPTION xii LIST OF REPORTS ISSUED xiii OBJECTIVE 1 INTRODUCTION 1 PLAN OF INVESTIGATION 3 T1E ALUMINUM GRIDS 4 Storage of Aluminum 4 Type of Aluminum Employed 4 Thickness of Aluminum to Employ in Grids 7 Stamping of Grids 7 Degreasing 12 Inspection of Grids and Removal of Any Aluminum Hairs 12 Annealing of the Aluminum 13 LEAD-PLATING OF ALUMINUM GRIDS 13 Flow Sheet for Production of Lead-Plated Aluminum Grids 13 Alkali Cleaning of Grids (Step 4) 13 Water Rinses and Acid Dips for Alkali-Cleaned Aluminum (Steps 5-12 and 14-15) 16 Conditioner Solution (Step 13) 18 Lead-Plating Bath (Step 16) 25 Agitation of the Bath 25 Temperature of the Bath 25 Hydrogen Ion Concentration 25 Density of the Bath 25 Addition Agents 26 Shields for the Grids in the Plating Tank 26 Voltage 26 Relation of Current Density Distribution to Character of Lead Deposited 26 Filtration of the Bath 35 Anodes 35 Inspection of the Lead-Plated Grid 36 ii CONFIDENTIAL1 -

. i r,'\, '~2 ^ > -^ ~,. a.i,.

CONFIDENTIAL I —.... TABLE OF CONTENTS (Cont ) _ _-. COIFIE Page Relation Between Current Density and Addition Agent 36 Controls used to Maintain Optimum Working Conditions in the Lead-Plating Bath 45 Controls Used with Conditioner Solution No. 2 46 ASSEMBLY OF LEAD-PLATED ALUMINUM PLATES INTO ELEMENTS AND THE PRODUCTION OF BATTERIES FROM THE ELEMENTS 48 Pressing Grids (Step 2) 48 Assembly of Dry Plates into Elements (Steps 6, 7, and 8) 50 Lead Burning of Lead-Plated Aluminum Plates (Steps 7 and 8) 50 Flame 50 Plate Lugs and Plate Strap Assembly 50 Burning Techniques 50 Inspection of Separators (Step 9) 53 Formation of Lead-Plated Aluminum Elements (Step 13) 55 Lead-Burning Cell Connectors and Terminal Posts (Step 18) 56 Charging of Lead-Plated Aluminum-Grid Storage Batteries 56 PILOT-PLANT EQUIPMENT USED FOR THE PRODUCTION OF LEAD-PLATED ALUMINUM GRIDS 58 CHARACTERISTICS OF LEAD-PLATED ALUMINUM-GRID BATTERIES 62 Self-Discharge of Lead-Plated Aluminum-Grid Batteries on Standby 62 Resistance of Lead-Plated Aluminum Grids 65 Comparison of Surface Corrosion of Lead-Antimony and LeadPlated Aluminum Grids 69 Weight Change During Cycling of Lead-Plated Aluminum Grids 69 PERFORMANCE OF LEAD-PLATED ALUMINUM-GRID BATTERIES 74 Comparison of 5-Second Voltage, Battery Voltage and Percent Discharged 74 Recovery of Lead-Plated Aluminum-Grid Batteries on Open Circuit After Discharge 74 Life History of Three 2HN 12-Volt 45-A.H. Batteries 82 Plate Thickness vs. Watt-Hours per Pound of Positive Oxide 86 Failure of Batteries No. 117, 119, and 120 to Charge 88 Battery Inspection and Study of Batteries Returned from Wright Field 100 Design of a 12-Volt 65-A.H. Battery (Battery, No. 152) 103 Specifications for the 12-Volt 65-AH. Chrysler Battery 103 Specifications for Lead-Plated Aluminum-Grid Battery 103 Inspection Report on 6TN and 2HN Batteries Sent to Yuma Test Station 105 Discussion of Conclusions of Yuma Test Station Report of September 29, 19553 108 Fii CONFIDENTIAL; '.^, Il

CONFIDENTIAL - TABLE OF CONTENTS (Conc), i Page Comments on Conclusion A. 108 Comments on Conclusion B 114 Comments on Conclusion C 114 Comments on Conclusion D 115 Comments on Conclusion E 115 iv CONFIDENTIAL

CONFIDENTIAL -~ sI LIST OF TABLES. Table Page 1 Chemical Composition Limits for Wrought Aluminum Alloys (Aluminum Company of America) 5 2 Thermal Conductivity and Electrical Resistance of Aluminum used in Battery Grids 6 3 Flow Sheet for the Production of Lead-Plated Aluminum Grids 15 4 Volume Change of Active Components in a Battery Plate 30 5 Examples of Plating Schedule 35 6 Flow Sheet for Assembly of Lead-Plated Aluminum-Grid Batteries 49 7 Case Formation Cycle of 0.060-Inch-Thick-Positive Plates Assembled with 0.030-Inch-Thick-Negative Plates 55 8 Charging Rate for Lead-Plated Aluminum-Grid Batteries 57 9 Pilot-Plant Equipment used for the Production of Lead-Plated Aluminum Grids 58 10 Standby Record of Six 2HN 12-Volt 45-A.H. Batteries; Standby on Open Circuit 63 11 Change in Specific Gravity of Acid in 2HN 12-Volt 45-A.H. Batteries on Standby for a Period of 45 Days 66 12 Comparison of the Linear Resistance of Lead-Plated Aluminum Grids with Solid Lead-Antimony Alloy Grids 68 13 Comparison of Unit Area Resistance of Lead-Plated Aluminum Grids with Solid Lead-Antimony (8*) Alloy Grids 68 14 History of Three 2HN 12-Volt 45-A.H. Batteries 83 15 Construction Details of Batteries Sent to Wright Field 101 16 Discharge Data, Life Cycle Data, to December, 1953 106 17 Construction of the Batteries Returned from the Detroit Tank Arsenal, Center Line, Michigan, Which had been Tested in Yuma, Arizona 107 18 Inspection Data of Batteries Returned from the Detroit Tank Arsenal, Center Line, Michigan, Which had been Tested in Yuma, Arizona 109 v CONFIDENTI [13.

CONFIDENTIAL I UIFIED LIST OF TABLES (Cont) Table Page 19 Comparison Between Yuma Test Station Data and University of Michigan Data on the Specific Gravity of the Acid in Each Cell-of Each Battery 110 20 Specific Gravity and Cell Voltage Data of the Batteries by Cells on Charging, Following Their Inspection as Recorded in Tables 18 and 19 111 21 Vehicular Mileage, Battery-Water Consumption and Final Specific Gravity of Batteries at Close of Test at Yuma Test Station 112 22 Life History in Cycles insofar as is Known of the Batteries Sent to the Yuma Test Station 113 i r ' ~ -...., vCONFIDENTIAL-. CONFIDENTIAL -~~ L' j

CONFIDENTIAL /: LIST OF PHOTOGRAPHS Photograph Page 1 Grid Die a. Piercing Die Open. 10 b. Piercing Die Closed Ready for Action. 10 c. Blanking Die Open. 10 do Blanking Die Closed Ready for Action. 10 2 Operations in Grid Production a. Operation 1. 11 b. Operation 2. 11 c. Operation 3. 11 3 Photomicrograph of 2S Aluminum, 200 X. a. Aluminum Grid as Received from the Piercing Operation. 14 b. Aluminum Grid Heat Treated at 500~F for 30 minutes. 14 4 Effect of Hydrochloric Acid Concentration in the Conditioner Solution on Heated Lead Electroplate. a. Small Deficiency of Hydrochloric Acid Manifest as Small Blisters. 22 b. A Large Deficiency of Hydrochloric Acid. 22 c. The Blister is Larger then 4b but not as many. A Less Deficiency of Hydrochloric Acid than in 4b. 22 d. A Satisfactory Lead Electroplate on 0,060 inch and 0.030 inch Aluminum Grids. 23 5 Effect of Hydrofluoboric Acid Concentration in the Conditioner Solution on Grids as they come from the Lead-Plating Bath. 24 6 Shields shown in Previous Report. a. Shields used at the Beginning of Project to Help Deposit Lead on Aluminum Grids in 1949. 27 b. Shield used to Plate 6 Grids at a Time Based on Shield Shown in Photograph 6a. 27 c. Modified Shield used in 1951 to Plate 3 Grids at a Time. 27 7 The Effect of Shields on Deposition of Lead Electroplate. 28 8 Ripples in Lead Electroplate Due to Improper Agitation. 37 9 Effect of Licorice and Specific Gravity on Lead Electroplate. 38 10 Effect of Shielding and Licorice Content on Lead Electroplate. 40 11 Different Shields Employed. 43 12 Relation of Plate Lug and Burning Comb. 51 13 Assembly of Plates in Burning Comb. 51 vii CONFIDENTIAL 'DE: pi duj;:EE

CONFIDENTIAL D{:.x -~IED LIST OF PHOTOGRAPHS (Cont) Photograph Page 14 Placing the Steel Mold Around the Plate Lugs Extending Through the Burning Comb 52 15 Puddling the First Layer of Lead from the Burning Stick into the Mold to Build Up the Plate Strap to the Desired Thickness 52 16 Welding the Cell-Post to the Cell Plate Strap 4 17 Removal of the Burned Positive or Negative Cell Element from the Burning Comb 54 18 Lead-Antimony Grids after 660-A.H. Continuous Charge at 0.022 amp/sq in. 70 19 Lead-Plated Aluminum-Grids after 880-A.H. Continuous Charge at 0.022 amp/sq in. 70 20 Photograph Showing Effect of Improper Charging of Batteries a. Edge View of Element Charged over 200 hours at a Maximum Impressed emf of 14.7 volts 98 b. Side View of Same Element Shown in (a) 98 21 Photograph of the Effect of Improper Charging of LeadPlated Aluminum-Grid Batteries 99 22 View of Chrysler Battery Case as Modified a. Side View 104 b. Top View Showing Reduction in Volume of Cell Capacity 104 viii CONFIDENTIAL DE-t''i I. O/

CONFIDENTIAL........ LIST OF GRAPHS -- Graph Page 1 Reduction in Weight of 0.015-Inch Lead-Plated Aluminum Grids Compared to 8%, Lead-Antimony Grids of Equal Area versus Total Grid Thickness 8 2 Relation Between Shields, Current Density per Square Foot, and Concentration of Licorice 44 3 Corrosion of Lead-Antimony and Lead-Plated Aluminum Grids 71 4 Weight Change During Cycling of Lead-Plated Aluminum Grids 73 5 Comparison of Voltage Versus Percent Discharged, Type 2HN 12-Volt 45-A.H. Battery at Various Discharge Rates, at 80 F 75 6 Voltage Versus Discharge Type 2HN 12-Volt 45-A.H. Battery, All Discharges at 80~F 76 7 Open-Circuit Recovery after Ten Minutes for Type 2HN 12-Volt 45-A.H. Battery Discharged at the 150-ampere Rate at -650F. Battery No. 110, Cycle No. 5 77 8 Open-Circuit Recovery after Ten Minutes for Type 2HN 12-Volt 45-A.H. Battery Discharged at the 300-ampere Rate at -40~F. Battery No. 110, Cycle No. 7 78 9 Open-Circuit Recovery after One Hour for Type 2HN 12-Volt 45-A.H. Battery Discharged at the 150-ampere Rate at 80~F. Battery No. 110 79 10 Open-Circuit Recovery after One Hour for Type 2HN 12-Volt 45-A.H. Battery Discharged at the 150-ampere Rate at -40~F. Battery No. 108 80 11 Comparison of Plate Thickness to Watt-Hours per Pound of Positive Oxide 87 12 25 Ampere Discharge at 80~F 89 13 150 Ampere Discharge at 800F 90 14 150 Ampere Discharge at 0~F 91 15 150 Ampere Discharge at -40~F 92 ix CONFIDENTIALDE-: LrjO ilE

CONFIDENTIAL I ~ LIST OF GRAPHS (Cont) Graph Page 16 150 Ampere Discharge at -65~F 93 17 300 Ampere Discharge at 80~F 94 18 300 Ampere Discharge at 0~F 95 19 300 Ampere Discharge at -40~F 96 x CONFIDENTIAL" J

CONFIDENTIAL /^~ -. SUMMARY This report covers the accomplishments during the period June 30, 1953, to December 31, 1953o For the benefit of those who wish to manufacture leadplated aluminum-grid storage batteries, a more detailed account of the manufacturing techniques and production controls required is given in this report, with data based on our pilot-plant controls. Design data are given for the production of a 12=volt 65-A.H. automobile starting battery. Inspection data on batteries which have been tested at Wright Field and Yuma Test Station are included, and summaries of these inspections are listed under their appropriate headings. xi -. CONFIDENT

CONFIDENTIAL SS —S.IFI Personnel of ERI Projects M743, M875, and 2018 ERI Project M743, September 1947, to November 15, 1949. L. L. Carrick, Project Supervisor September 1947, to November 15, 1949 B. Agruss, Full-time investigator February 1, 1948, to November 15, 1949 J. M. Stapleton, Full-time investigator February 1, 1948, to November 15, 1949 H. H. Hicks, Jr., Part-time laboratory December 10 1947, to March 1948 assistant December 10, 1947, to March 1948 Don Church, Part-time laboratory n ati a at December 16, 1948, to November 15, 1949 ERI Project M875, November 15, 1949, to January 1, 1952. L. L. Carrick, Project Supervisor November 15, 1949, to January 1, 1952 B. Agruss, Full-time investigator November 15, 1949, to January 17, 1950 J. M. Stapleton, Full-time investigator November 15, 1949, to January 1, 1952 F. J. Prieskorn, Full-time laboratory assistant January 30, 1950, to January 1, 1952 D. M. Braun, Full-time laboratory assistant February 5, 1950, to January 31, 1951 E. J. Groff, Full-time laboratory assistant November 6, 1950, to January 1, 1952 Carl F. Cooper, Full-time laboratory assistant May 1, 1950, to January 1, 1952 Don Church, Part-time laboratory assistant November 15, 1949, to February 1, 1951 Don E. Kory, Part-time laboratory assistant February 12, 1951, to May 1, 1951 Sam Dreisback, Part-time laboratory assistant June 2, 1951, to October 30, 1951 William Elliot, Part-time laboratory assistant February 13, 1951, to October 30, 1951 ERI Project 2018, January 1, 1952, to December 31, 1953 L. L. Carrick, Project Supervisor January 1, 1952, to December 31, 1953 J. M. Stapleton, Full-time investigator January 1, 1952, to December 24, 1952 Carl F. Cooper, Full-time laboratory assistant January 1, 1952, to June 1, 1953 F. J. Prieskorn, Full-time laboratory assistant January 1, 1952, to December 31, 1953 E. J. Groff, Full-time laboratory assistant January 1, 1952, to December 31, 1953 John Altman, Part-time laboratory assistant June and July 1952 Patrick Doyl, Part-time laboratory assistant August and September 1952 Larry Wilkinson, Part-time laboratory assistant June 15, 1953, to December 31, 1953 Gene Zerlaut, Part-time laboratory assistant September 24, to December 31, 1953 xii CONFIDENTIAL LFiL'D

CONFIDENTIAL NTI Reports Issued During Duration of ERI Project M743, M875, and 2018 and Covered in This Final Report Project No. Date Issued Period Covered M743 March 1948 September 15, 1947, to March 1, 1948. M743 30 June 1948 March 1, 1948, to June 30, 1948. M743 17 August 1948 Special Report. M743 30 September 1948 June 30, 1948, to September 1, 1948. M743 31 December 1948 September 1, 1948, to December 31, 1948. M743 31 May 1949 December 31, 1948, to May 31, 1949. M743 30 August 1949 May 31, 1949, to August 30, 1949. M743 31 December 1949 August 30, 1949, to December 31, 1949. M875 24 April 1950 December 31, 1949, to April 1, 1950. M875 July 1950 April 1, 1950, to July 1, 1950. M875 31 December 1950 July 1, 1950, to December 31, 1950. M875 15 April 1951 December 31, 1950, to April 1, 1951. M875 December 1951 April 1, 1951, to December 1, 1951. 2018 June 1952 December 1, 1951, to June 30, 1952. 2018 July 1952 Special Report on Lead Burning. 2018 January 1953 June 30, 1952, to December 31, 1952 2018 31 December 19553 A final resume of the above reports emphasizing the techniques to be observed in manufacturing of leadplated aluminum-grid batteries. xiii CONFIDENTIAL | -v. C:-ui. C

CONFIDENTIAL -- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY 0O AN THE DEVELOPMENT OF A NEW-TYPE STORAGE BATTERY FOR MILITARY USE OBJECTIVE This work was directed toward the development of a new-type storage battery for military use employing, if such can be discovered or developed, a more satisfactory substance as a substitute for lead and, if none can be found, to find desirable means of improving the performance of lead batteries. The desired battery is to be a 24-volt battery which will have a volume under 2 cubic feet and a weight under 100 lbs, will operate satisfactorily over a temperature range from -70~ to 1650F with a 300-ampere discharge for one minute at -70~F with an end voltage of 1 volt per cell. This battery is to be waterproof for submersion, under a minimum external or internal pressure of 6 psi. INTRODUCTION Since it has become evident during the past few years that lead may become a scarce metal, it is increasingly desirable to find a substitute for lead in storage batteries. It is imperative that battery performance be improved, since the low electrochemical efficiency of the commercial alkaline batteries and their poor performance at low temperatures make them unsatisfactory for military use. Willihnganz, in a paper presented before the Association of American Battery Manufacturers, has shown from theoretical considerations that only a few electrode combinations besides lead are possible. Use of these combinations required considerable research and theoretically appeared very limited. Hence, most improvements have been obtained from studies of CONFIDFNTIAI

CONFIDENTIAL /*ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MI plate thickness and paste mixes. The tendency has been toward thinner plates, more plates per cell, and the use of various extenders or addition of agents in the paste. By these means more porosity and greater active area were obtained. Even though cold performance was improved, the objective set forth above has not been reached. One of the greatest disadvantages of the lead battery has been its weight, which is particularly striking when it is realized that only a small percentage of the lead in a battery is necessary for energy production, the rest being excess in the form of oxides, grids, and conductors. Because of the poor electrical and heat conductivity and low strength of lead, there is a tendency for the grids to buckle and warp under charge of discharge conditions, especially at high rates, and they become brittle when cold. Several investigators have suggested the use of lead-coated light metals such as copper or aluminum for grid materials, but summarily dismissed them as possibilities because of local cell action. It is true that one pinhole in the coating will cause local self-discharge, but in view of the rapid advances that have been made in modern metal finishing and electroplating this suggestion deserves a much more careful examination. Widespread commercial acceptance of electroplates on aluminum was unheard of ten years ago. In addition, the capture of a German battery with lead-coated aluminum plates that were reported to have withstood 1000 cycles substantiates the assertion that such plates should be thoroughly investigated. The long life (1000 cycles) suggests the use of Plante-type plates. It is well known that the Plante" battery had a longer life and a higher capacity than other batteries, especially at low discharge and with regard to current drain per unit plate area. However, it was discarded in favor of the Faure-type plate because of its extreme weight and bulkiness and because, at the same time, the plates were more prone to buckle and warp. It is noteworthy that the modern trend in battery design has been toward the Plante" type in that thinner plates are used to give greater active surface area. Plante plates had thin layers of active material covering large areas. Since it has been conceded that lighter-weight grids would be a boon to the industry and that the trend is toward thinner layers of active material, it would appear that this is the propitious moment for an extensive investigation of lead-coated aluminum plates. The high electrical and heat conductivity of aluminum coupled with its greater strength (compared to lead) should allow the use of much thinner plates without the danger of overheating and buckling, and should provide also a greater active surface area per unit volume or per unit weight. Such an increase in active surface area per unit weight will tend to increase not only the ampere-hour capacity but also the life of the battery. CONFIDENTIAIl

CONFIDENTIAL /.L.. ENGINEERING RESEARCH INSTITUTE UNIVERSITY IU PLAN OF INVESTIGATION Because of the above considerations, it was decided that the primary objective would be the development of a lead-acid storage battery containing lead-coated aluminum grids, and that the development of other electrode systems would be a secondary objective. To some extent, the leadcoated aluminum system may be considered as a substitute for the major portion of the lead in the standard lead-acid battery. The proposed plan of investigation is tabulated below: 1. Develop a suitable technique for plating lead on aluminum, testing each type of plate for continuity of coating, porosity, local cell action, metal diffusion. 2. Develop methods of "forming" active material on lead-coated grids. 3. Assemble test cells and determine their capacity, life, maximum current density, self-discharge, charge and discharge rates, temperature efficiency, etc. a. Correlate test data with methods of plating and with the formation of active positive and negative grid surfaces. b. Ascertain optimum plating, forming, and assembling conditions. c. Specify raw material to meet the aluminum-lead system requirements. 4. Reduce lead-plated aluminum grid elements in the assembled battery so that the final voltage of the fully charged aluminum grid battery will equal the voltage of the fully charged lead-antimony-grid sixelement battery, in order to utilize the existing charging equipment in use today. CONFIDFNTIAI --- -__

CONFIDENTIAL /X ENGINEERING RESEARCH INSTITUTE * UNIVERSITY t HYN Q THE ALUMINUM GRIDS Storage of Aluminum The aluminum stock should be stored in an atmosphere free from acid fumes, since acid fumes tend to pit the surface of aluminum and create an uneven plating surface. The aluminum stock should also be stored in an even-temperature atmosphere where the grease on the surface will not be removed. Excessive oxidation of the aluminum suface may also result in pitting of the grids as they pass through the cleaning solutions, and adhesion is dependent to some extent on the condition of the aluminum surface and the extent of oxide removal. Any atmosphere which tends to increase the thickness of the aluminum oxide film on the stock aluminum should be avoided. Pitting may cause blisters and loss of adhesion of the lead electroplate when it is heated or placed in service, especially at elevated temperatures. Type of Aluminum Employed We have plated all the aluminum alloys listed in Table 1. The composition is quite variable but all accept a coherent, nonporous, and continuous electroplate of lead. Only the alloys EC, 2S, 3S, and 52S have a satisfactory chemical composition. When the copper content exceeds 0.20% we have found that the copper will diffuse through nickel and lead electroplate, migrate from the positive plate, and deposit as an electroplate of copper on the negative plate. Thus copper promotes self-discharge of the battery; hence, it is kept at a minimum. Alloys like 2SO, 3SO, 52SO, have sufficient hardness to pierce satisfactorily in the production of the aluminum grid. The hardness also assists in holding the aluminum rigid during movement in the plating tank. The cold-rolled aluminum seems to possess a nonuniform surface hardness; hence, it has been found advantageous to anneal the aluminum grids after degreasing to produce a uniform surface structure which facilitates alkali cleaning and deposition of a satisfactory lead electroplate. Proper hardness in the aluminum is also conducive to the production of clean-cut edges on the aluminum grids. Depending on the type of piercing die and the C ONFDENIA-. Li r X CONFIDENTIAL -A- I '

TABLE 1 CHEMICAL COMPOSITION LIMITS FOR WROUGHT ALUMINUM ALLOYS (ALUMINUM COMPANY OF AMERICA) Composition given is a maximum unless shown as a range; the remainder is aluminum. Alloy Copper, Silicon, Iron, Manga- Magne- Zinc, Chromium, Nickel, Titan- Other Elements A4 o % nese, % sium,% % ium, % Each Total EC 99.45% minimum aluminum content 2S* 0.20 ** ** 0.05 - 0.10 - - - 0.05 0.15 2SH14 0.20 ** ** 0.05 0.10 - - - 0.05 0.1 0' 2SH18 0.20 ** ** 0.05 - 0.10 - -- 0.05 0.1 33s0 0.20 0.60 0.70 1.0-1. - 0.10 - - - 0.05 0.15 Z Z - 'T 14s 3.9-50 0.5-1.2 1.0 o.4-1.2 0.2-0.8 0.25 0.10 - 0.05 0.15 nI 18S 3.5-45 0.90 1.0 0.20 0.45-0.9 0.25 0.10 1.7-2.3 0.05 0.05 0.15 Z Z. 224S0 3.8-49 0.50 0.50 0.3-0.9 1.2-1.8 0.10 0.10 - - 0.05 0.15 24ST4 3.8-49 0.50 0.50 0.3-0.9 1.2-1.8 0.10 0.10 - - 0.05 0.15 52SH38 0.10 **** *O 0.10 2.2-2.8 0.10 0.15-035 - 0.15 0.05 0.15 61ST4 0.15-0.40 0.4-0.8 0.70 0.15 0.8-1.2 0.20 0.15-0.35 - 0.1 0.05 0.15 I r 61ST6 0.15-0.40 o.4-o.8 0.70 0.15 0.8-1.2 0.20 0.15-0.35 0.15 0.05 0.15 f~ r7 l*Minimum aluminum content is 99. I. jC**Iron plus silicon is maximum of 1%. / ^ I < j ***Usually contains no titanium. C/ i ****Iron plus silicon is maximum of 0.45%. I J

CONFIDENTIAL rDErpr^^ ENGINEERING RESEARCH INSTITUTE UNIVERSITY hardness of the aluminum stock, the stock aluminum may vary from 1/2 0 to 0 and up in hardness. Aluminum in the grid maintains flexibility of the plate during cold weather. Conductivity or resistance of the aluminum stock is important in maintaining a high discharge voltage (see Table 2), low development of heat during charge and discharge, reduction of internal resistance at all temperatures and especially its effect on cold-temperature operation. TABLE 2 THERMAL CONDUCTIVITY AND ELECTRICAL RESISTANCE OF ALUMINUM USED IN BATTERY GRIDS (Metals Handbook. 1948 Edition) Thermal Conductivity Aluminum Cy Electrical Resistivity at 25~C Type a 5 at 20~C, micro-ohm-cm cal/sq cm/cm/~C/sec EC 0.50 2.6548 2S-0 0.53 2.922 2S-H18 0.52 3.025 3S-0 0.46 3.448 14S-O o.46 3.448 18S-0 0.46 3.448 24S-0 0.45 3.448 24S-T4 0.29 5.747 52S-H38 0.33 4.926 61S-T4 0.37 4.310 61S-T6 0.37 4.310 All aluminum, whether it is 2S or any other designated commercial type of aluminum, is manufactured to meet certain specifications as to minimum and maximum alloy content of iron, silicon, manganese, etc. It is interesting to note that some manufacturers adhere to the lower limit of impurities in their manufacturing operations while others seem to use the 6 iM.. ' CONFIDENTIAL ~

CONFIDENTIAL.... ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF- A AN upper limits as their manufacturing guide. These types of aluminum require a different concentration of conditioning solutions. This will be discussed further in connection with dipping solutions. Thickness of Aluminum to Employ in Grids The thickness of the aluminum is dependent on (1) whether a positive or a negative grid is to be produced, (2) the type of service expected of the lead-plated aluminum grid, and (3) the type of battery desired. The positive grids to withstand 12 S.A.E. overcharge cycles should have an electroplate of 0.015 inch of lead. Thus a 0.075-inch plated grid may carry 0.045 inch of aluminum stock. The negative grids to withstand 12 S.A.E. overcharge cycles need not have over 0.010 inch of lead electroplate. Thus a 0.030-inch-thick grid may carry 0.010 inch of aluminum. We have run 12 S.A.E. overcharge cycles on grids that have 0.0075-inch-thick lead plate on 0.015-inch-thick aluminum. The 0.010-inch-thick lead plate is a safety factor, not a necessity if a good lead plate has been deposited. Since the sloughing of reaction products is primarily on the positive plate, it is advisable to give it a heavier electroplate. We have experimented with electroplates from 0.005-inch to 0.025 inch thick. For starting batteries, we consider 0.015-inch-thick lead electroplate for positive and 0.010-inch-thick lead electroplate for negative grids ample to meet all normal requirements or demands. For standby batteries, which usually must carry much thicker plates, an electroplate 0.020 inch. thick for positive and 0.015 inch thick for negative grids is ample. The balance of the thickness of the grid may be aluminum. Graph 1 gives the relation between the weight of lead-plated aluminuml grids which have a 0.015-inch-thick lead electroplate and a similar grid made of 100% lead-antimony alloy (8%Sb). This saving in weight is dependent on the grid design. Hence it should not be taken as the ratio of weights between lead-plated aluminum grids that we use and leadantimony alloy grids (commercial grids of today) of equal thickness. Stamping of the Grids.n stamping the aluminum it should be remembered that the stamping operation cold-works the aluminum stock. Cold-working will tend to CONFIDFNTIAI ' g C

GRAPH I REDUCTION IN WEIGHT OF 0.015 INCH LEAD-COATED s ______ ALUMINUM GRIDS COMPARED TO 8% ANTIMONY-LEAD __ 80 ~ GRIDS OF EQUAL AREA VS. TOTAL GRID THICKNESS 70 ~~~~ 60 0 " 50 zo z Z 0 FP 040 rl gA A A. A REDUCTION IN WEIGHT OF 0.0075-INCH Z - / LEAD-COATED ALUMINUM GRIDS COM- rl 30_ _ PARED TO 8% ANTIMONY-LEAD GRIDS _ — aQ~~~~~~~ AII IIB. 1 REDUCTION IN WEIGHT OF 0.010-INCH LEAD-COATED ALUMINUM GRIDS COM- re. 20| B E* / PARED TO 8% ANTIMONY- LEAD GRIDS _. 10-INCH'/. l,/ c- ',; C-.':. I rr1 1 40 80 120 160 200 2 0Ci <^IP~ 7i~~~~ FTOTAL THICKNESS OF GRID (THOUSANDTHS OF AN INCH) |,. f J

CONFIDENTIAL,'.. ENGINEERING RESEARCH INSTITUTE UNIVER IC harden the stock to some extent, but such hardening is not necessarily detrimental. We have experienced the greatest difficulty in the type of lubricant used in the stamping operation. Any lubricant which will react with the aluminum under pressure and temperature and form rather stable compounds with the aluminum should be avoided. Many lubricants that form stable compounds are not detrimental, however, as they may be removed by solvent action during the degreasing operation. On the other hand, some of these compounds are difficult to remove in either the degreasing operation or the cleaning operation. Such compounds left on the aluminum are likely to cause lack of adhesion or blistering or both of the electroplate when the aluminum is subjected to heat, as during assembly burning. It has been our experience that a good grade of kerosene free from lubricating oil makes an excellent lubricant for the stamping die. No synthetic turpentine should be used; turpentines do not seem to function as well as kerosene or mineral spirits. The grid die is shown in Photograph 1. Photograph l(a) shows the piercing die with the inside of the two halves exposed, while Photograph l(b) shows the piercing die assembled for action. Photograph l(c) shows the two halves of the blanking die which outlines the grid, and Photograph l(d) shows the blanking die assembled for action. The grid production operations are shown in Photograph 2, depicting the three steps required in using our piercing and blanking die to produce aluminum grids. The three round holes in the upper left-band corner, Photograph 2(a), and the one round hole in the lower left-hand corner are the result of the first operation, which pierces one-half of the grid openings. These two groups of holes will govern the position of the second stamping operation as shown in Photograph 2(b). The grid is advanced one hole in the upper left-hand corner; this pushes the aluminum stock through the die at just the proper distance so that the second operation removes the unpierced area between the existing openings and leaves the grid ribs as shown in Photograph 2(b)o It should be noted that at this stage there are four holes in the upper left-hand corner and four in the lower left-hand corner. When all the aluminum has been pierced as shown in Photograph 2(b), the aluminum ribbon is passed through the blanking die, Photograph l(d). Here the ribbon is guided by the four holes in the upper left-hand corner, Photograph 2(b), and the four holes in the lower left-hand corner of the same photograph and the grid is blanked, outlining the completed grid in the form shown in Photograph 2(c). A die of this type has been used to pierce over 100,000 grids from aluminum stock ranging in thickness from 0.010 inch to 0.030 inch in 9 ~ ae Uf- jSv F1 CONFIDENTIAL^ ~

:iiii:0i 0{:i ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~ i?!:::::::::::::::::: ~:::i:ii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~::::::.....i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~;:::-: z":Z O~~~~~~~~~ z~~~~~~~ z. (a) (b) (c) (d) Photograph 1. Grid Die: (a) Piercing die open; (b) Piercing die closed ready for action. (c) Blanking die open; (d) Blanking die closed ready for action.

-. - t —do mmmin~ --— mm mlmm - - lm --— mm n:ffff~~~~if^0 W:::-i -:iif:iii~i:-l *I _I _ at000000;00 f:0:0: - _ _S_ Z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I ---._ l~-:-iO ~~~~~~z Z.. ---mm — r~~_ __ r,~~i '1'1 men mmmmm 1 mm 0 mm-. —m — Z — I ~ ~ ~~~m ----mlili iiii "~!! i~ m --- —-— m ----m~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i'8i-i-ii~iii-'iBj~i-i mmmin~~~~~~~~~~i -- m m -- ~ ~ ~ ~ ~:::::Lr~mmm I::-:~~~~~~~~~~ni3 o o _ _ b (a) (b) Photograph 2. Operations in Grid Production: (a) Operation 1; (b) Operation 2; (c) Operation 3.

CONFIDENTIAL / - ENGINEERING RESEARCH INSTITUTE ~ UNIVERSiTr -MhUR GI thickness. We also have used this die to pierce copper, iron, and brass 0.010 inch thick. There have been some minor repairs and sharpening, but on the whole the die is in excellent condition. In the construction of piercing dies, the die should be designed to pierce only one thickness of aluminum stock, such as 0.015 inch, 0.030 inch, etc. In this way, less clearance is required and the ribs may be reduced in width, thus reducing the total weight of the grids and the total lead consumed, while maintaining the resistance and conductivity of the grid so that it will be better than a solid lead-antimony-alloy grid. The hardness of aluminum stock to use in each die will depend on the clearance and the sharpness of the die. Degreasing The degreasing operation should come before the removal of aluminum hairs and also before annealing of the aluminum, since annealing of the aluminum may cause decomposition of some of the lubricating compounds, thus depositing carbon in the pores of the aluminum which will 1ce almost impossible to remove and difficult to bridge with subsequent conditioning solutions or electroplates. If degreasing has preceded the removal of aluminum hairs and annealing of the aluminum, it is not necessary to be concerned about any grease that may be deposited on the aluminum grid through handling, especially if the handler will use clean cotton cloths. Any small amount of human grease from the hands is not likely to cause serious harm, as it will be removed by the cleaning solution. Degreasing is done in a vapor degreaser. Trichlorethylene is employed as the degreasing agent. Inspection of Grids and Removal of any Aluminum Hairs The grids that come from the stamping die may have aluminum hairs attached to ribs or edges of the grids. Such aluminum hairs are produced if the die is not sharp or if the second stamping operation is not properly synchronized or spaced with the first stamping operation. Should there be an overlapping in the second operation, this will generally produce aluminum hairs. If the aluminum hairs are excessively fine, they may be removed in the cleaning operation by the action of the alkali arnd subsequent acid dips. Should the hairs be too heavy for such cleaning, however, then they 12 i l Li "i^ L l CONFIDFNTIAI

CONFIDENTIAL ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY O / F' should be removed by some mechanical means such as wire brushing. The wire brushing should be done after the grids have been degreased, to prevent the burning of any of the lubricant into the pores of the aluminum. Annealing of the Aluminum Not all types of aluminum need annealing to be satisfactorily electroplated, but should there seem to be areas on some types of aluminum which do not accept electroplate and these areas cannot be cleaned either by degreasing or passage through the cleaning solution, it is recommended that these grids be annealed at a temperature of 500F for a period of 30 minutes in order to release any stresses or strains resulting from coldworking during the stamping operation. Such areas under stress or strain are likely to receive electrodeposit at a different rate or under different conditions than those which are not subject to such stresses or strains (see photomicrographs in Photograph 3). It has been our experience that it is not necessary to anneal 2S aluminum grids, as the cold-working does not seem to affect the deposition to any serious extent with this stock. LEAD-PLATING OF ALUMINUM GRIDS Flow Sheet for Production of Lead-Plated Aluminum Grids The electroplating of aluminum comprises a total of seventeen steps. A flow sheet for the production of lead-plated aluminum grids is given in Table 3. Steps 1, 2, and 3 of the process have been discussed on pages 7, 12, and 13 respectively. Alkali Cleaning of Grids (Step 4) The first step after the grids have been degreased and annealed is the alkali cleaning, which removes small traces of grease and the film of aluminum oxide on the surface of the grid. This film of aluminum oxide is approximately 0.5 x 10-6 inch thick when the stock is received from the mill. An alkali cleaning solution which has been found to be quite satisfactory is composed of the following ingredients: 13 C-N A CONFIDENTIA IX rJnS'ir^D I

1VIIN]3aIJNOD fT ~socnuTuI o~ aoj oOOG 4 pGa;q' pTrJ uTnuTumTV (q) fuoTq.$Jdo OuTo.9iTd aq; uIOJJ p9ATGaGJ S~ pTJxg nuTumnIV (S) 'XOO0 'mirnumnTmTV SF jo sLdmBoJoTmoloMoqj 'T qdiBo;oqo (q) ^ ^..~'^-i^: * ~ ~, /..^. ^ -:-.: -:. -*:,~,:::'^ ^..::~j ''.,) 1V1-.N]Q4~:IJNO "s, ~~~~~~~~A i~~~~~i: x~~~~~~~

TABLE 3 FLOW SHEET FOR THE PRODUCTION OF LEAD-PLATED ALUMINUM GRIDS ___2 3 4 5 6 | STOCK: 1 r VAPOR^ ~ 1t lALKALINE CLEANER: 2STOCK: VAPOR ANEAL: T.S.P, 2-4i oz/gal i RINSE: RINSE: 2S A|, 3S Al, STAMPING DEGREASER 0 m Soda ash, 1-2 o z/gal S Dip or Spray Dip or Spray or 52S Al, OF GRIDS Trichioro- > 0 ash 0-2 oz/gal 10-20 sec 10-20 any thickness ethylene 2 at 200-10oF at 160-180oF at 6O8 0~F 10 -012 11 |~ 9 8 7 SULFURIC -HYDROCHLORIC NITRIC ACID 0. 0 i 4 RINSE: RINSE: ACID DIP: RINSE: RINSE: Dip or Spray Dip or Spray H2SO04 Dip or Spray Dip or Spray Z7 Ni,110-20 sec 10-20 sec HC sp.gr., 1.050 10-20 sec 10-20 sec -' at l6o-18o0F at 6o-8o~F H20 J at 16o-l80oF at 6o-8OF 1 0-s a 0~^ r~ ~~ sec at 80~F ~ '~ 1^, ___ X 13-5- sec at 80F0~F Z Z u-E^ g13 LEAD-PLATING BATH; 1!^~~~~ i i~l - ^~i Pb(BF4)2. fl oz 18 4 '^ NICKEL CONDITIONER SOLUTION: 1,HBF4 27.0 fl oz.... Conc. HC 7.60 f oz Dip or Spray Dip or Spray Add 0.05 oz powdered S Blast or NiSO4 7H2O 15.40 oz 15-20 sec 10-20 sec licorice-root extract 1-20 Oven / NiCi-6H20 4.70 oz at 16o-18o-F at 6o-8o2F SE 1220 4.70 oz at 16O-18O~F at 6O-8O ~F per gal, dissolved in sec at Distilled H20 to make 1 gal distilled H20. 80 F 5 min at 110-115~F ~^~~ mm1 alO-5F10-63 min at 75790~F, C'. 100-600 amp/sq ft I ~~rr\, ^/ ^ I I ^ I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..... 1 — 1

CONFIDENTIAL; ~ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY FA sJI Trisodium phosphate 2-4 oz/gal Soda ash 1-2 oz/gal NaOH 0.25 oz/gal This cleaning bath operates well if the temperature is between 200 and 2100F. The time required for cleaning will depend on the nature of the 2S aluminum, i.e., whether it meets the highest purity standards or meets only the lowest ASTM purity standards for the manufacturing of 2S aluminum. Aluminum which has been covered with a heavy film of aluminum oxide will need to remain in the alkali cleaner for a longer period, or it may be necessary to dip it in the alkali cleaner more than once. Should there be a large amount of iron in the aluminum, the cleaning period may be reduced to what it is for 2S aluminum which meets the lower limit of manufacturing impurities. Water Rinses and Acid Dips for Alkali-Cleaned Aluminum (Steps 5-12 and 14-15) The grid will turn black due to the precipitation of impurities, and these impurities must be removed in an acid rinse or dip. Should the acid dip be composed of nitric acid, there seems to be a tendency toward pacification of the surface of the aluminum; hence, such a dip is usually followed by a sulfuric acid- hydrochloric acid dip. If the iron content is excessively high, we have found that a dilute solution of hydrochloric acid (5-10/) may be substituted for the sulfuric acid-hydrochloric acid dip or for both the nitric acid and the sulfuric acid- hydrochloric acid dips. In going from the alkali cleaner to the acid dip, the aluminum grid should be rinsed in hot water (160-1800F) for 10 to 20 seconds to remove excess alkali as drag-out from the cleaning tank. This rinse may be either a dip (with agitation) or a spray operation, or a combination of both. The acid dip following the alkali cleaner and water rinse is generally composed of nitric acid 1.270-1.300 specific gravity heated to a temperature of 80~F. The grid is usually allowed to remain in this solution for approximately 15 seconds; this time interval will vary depending on the type of aluminum and the degree of pacification desired. Should little or no pacification be desired, the hydrochloric acid dip of 5-0l hydrochloric acid is quite satisfactory. Following the acid dip tank, the grid is rinsed by dipping or spraying with water at a temperature of 60-80~F to close the pores of the grid and squeeze out any acid that may remain in it. The cold-water rinse, dip or spray, is followed by a hot-water (160-180~F) rinse, dip or spray, in which the pores are opened and water absorbed which will dilute any acid 16. - CONFIDFNTIAI, b t-tLh^IELD

CONFIDENTIAL, ENGINEERING RESEARCH INSTITUTE ~ UNIVERSIT remaining and if allowed to remain 15-20 seconds should reduce the acid retained in the grid to a minimum. Thus, once the aluminum grid has passed the alkali cleaner (Step 4) it is hot-water rinsed (160-180~F). This hot-water rinse opens the pores of the metal and dilutes any alkali that may be in the pores. It is followed by a cold-water rinse (60-80"F) which contracts the pores and ejects the water and alkali. The first acid dip, called the nitric acid dip, is at 80~F. As the metal comes in contact with the 80~F water rinse, the pores tend to remain closed and exclude much of the acid, but sufficient acid enters to neutralize the alkali. The nitric acid dip is followed by a cold-water rinse or spray (60-800F), during which the pores in the aluminum remain closed. This washes off the excess acid from the surface of the grid and suspends further chemical action on the cleaned aluminum. The cold-water rinse or spray is followed immediately by a hot-water rinse or spray, which expands the pores of the aluminum, drawing in water to dilute any acid that may have entered and be retained in the pores, thus neutralizing retained alkali. The hot-water rinse or spray is followed immediately by dipping in a sulfuric acid -hydrochloric acid dip at a temperature of 800F. This procedure squeezes out any water, diluted nitric acid, or salts which may have been retained in the pores and converts any nitric acid salts present to sulfuric acid salts, thus removing all the nitric acid from the grid. The hydrochloric acid in this bath is added to assist in preparing the aluminum for the nickel conditioner by activating the aluminum surface. The hydrochloric acid seems to create considerable effervescence, thus helping to reduce any aluminum oxide left on the grid and form water-soluble salts. The hydrochloric acid also helps to promote adhesion of the nickel conditioner to the aluminum. The sulfuric acid- hydrochloric acid dip is followed by a cold-water and a hot-water washing cycle as described above. The aluminum goes directly from the last hot-water dip in the expanded condition to the nickel conditioning solution at 110-115F. This causes a slight contraction of the pores, squeezing out any acid that may remain but at the same time assisting the deposition of nickel on the surface of the aluminum. Example of Nitric Acid Dip Nitric acid (sp.gr., 1.42 C.P.) 36.2 lb or 11,375 ml, or about 5.17 7-lb bottles. Distilled water 23,245 ml Specific gravity of mixture: 1.165-1.170 (33.9' BHN); operate at 80~F. The addition of nitric acid will be necessary from time to time in 100-mi quantities as the activity is reduced by formation of aluminum nitrate. 17 D- CONFIDENTIAL DE-CL^olJ i EDiu

CONFIDENTIAL r ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN The grids should remain in the nitric acid dip for 20-25 seconds. The time factor depends on the type of aluminum, 2S, 3S, or 52S, etc. The object of this dip is to remove all discoloration from the aluminum surface left by the alkali cleaner. Example of Sulfuric- Hydrochloric Acid Dip Sulfuric acid (sp.gr., 1.84 C.P.) 2.12 lb, or 520 ml Sp.gr., Distilled water 8 gal, or 30,280 ml 1.020-1.025 Hydrochloric acid (sp.gr., 1.19 C.P.) 3.06 lb or 1,165 ml Distilled water 2,320 ml Specific gravity of mixture: 1.030 If the specific gravity is below 1.030, add H2S04 and HC1 in the ratio of 2.12 lb to 3.06 Ib (or 10 ml H2S04 to 22.5 ml HC1) until the specific gravity recovers a value of 1.030. The specific gravity may increase to not over 1.040 without affecting the operation of the acid dip. Operate at 80~F. The time of dip varies with the type of aluminum, from 2-5 minutes. If the bath becomes inactive, the addition of 5 ml of HC1 (sp.gr., 1.19) to a 1-gal bath usually restores its activity. In stubborn cases, 10 ml of HF acid added will probably restore activity. If this fails to renew the activity of the bath it should be dumped. The bath should also be dumped if it becomes cloudy. Conditioner Solution (Step 13) The nickel conditioner solution is very important for good adhesion of the lead electroplate, which is important because a loss of adhesion may cause blistering of the lead electroplate. The conditioner solution to be used on the aluminum grid just following the acid dip and subsequent hot-water rinse may be composed of iron ions as follows: Iron Conditioner Dip FeC12'4H20 757 grams HC1 757 ml H20 3028 ml Operate 30 seconds at 100-120~F Conditioner solutions may also be made from chromium ions or nickel ions. CONFIDENTIAL t L:Lii

CONFIDENTIALA SSIFIED ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI M - We have found that a most effectual dip and one easy to operate is the conditioner solution made from nickel ions according to the formula given below. Conditioner Solution No. 1 Cone. H2S04(sp.gr., 1.84) 1.30 fl oz KBF4 (42%) 1.30 fl oz Cone. HC1 (sp.gr., 1.19) 7.60 fl oz NiS04'7H20 13.40 oz NiC12- 6H20 4.70 oz Distilled H20 To make 1 gal A nickel dip made according to the above formula is generally maintained at a temperature of ll0-115~F. The dipping time will vary in accordance with the chemical composition of the aluminum and whether it has been pacificated in the acid rinses. While we recommend 5 minutes, this may be adjusted downward to as little as 2 minutes. Two cases will illustrate the variations desired in the conditioner solution: (1) with aluminum which meets the lower manufacturing limits for ingredients other than aluminum in 2S, 3S, and 52S aluminum and (2) with aluminum which meets the upper manufacturing limits for ingredients other than aluminum. Conditioner solution No. la, below, is used when the aluminum meets the lower manufacturing limits for ingredients other than aluminum (Case 1). This solution may be used to provide the only barrier coat between the aluminum base and the top coat of lead electroplate. This specific formula may be varied as required to meet changes in the aluminum alloy composition. Conditioner Solution No. la Nickel sulfate (NiS04-7H20) 18.3 oz Nickel chloride (NiC12-6H2) 18.3 oz Sodium chloride (NaCl) 2.8 oz Nickel fluoborate (Ni(BF4)2 - 44% 9.9 fl oz Hydrochloric acid (HC1, sp.gr., 2.2 fl oz 1.18-1.19) Sulfuric acid-water solution (H2S04, sp.gr., 1.270) Ammonium Hydroxide (30 parts NH40H, 1.9 fl oz sp.gr. 0.9, and 70 parts of water) Distilled water To make 1 gal 9.................... CONFIDENTIAl vi l -'..ED

CONFIDENTIAL -; ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY M IGNJ D This conditioner is operated at a temperature of 80-90~F. After degreasing and removal of surface oxides, the article is placed in the conditioner for 2 to 3 minutes and on removal from the conditioner any adhering solution is removed by rinsing in hot and cold running water. The nickel electroplate is deposited directly on the nickel container deposit, which serves as a barrier coat for the top coat of nickel, chromium, lead, nickelchromium, or nickel-lead electroplate. A conditioner solution suitable for aluminum which most nearly approaches the upper manufacturing limits of ingredients other than aluminum (Case 2) is Nickel Conditioner Solution No. 2. Conditioner Solution No. 2* Nickel chloride (NiC12 6H20) 9 lb **Hydrochloric acid (sp. gr., 1.19 C.P.) 1400 ml Distilled water to make 39,000 ml ~***Hydrofluoboric acid added as needed Temperature 110-115~C. It has been determined by calculations and verified in practice that 10 grams of nickelous chloride salt (NiC12-6H20) will coat 700 2HN aluminum grids of 0.030-inch thickness. The coating of nickel is approximately 0.0001 inch thick or 0.24 g of nickel per grid as calculated by the nickel removed from the bath. A typical cleaning operation prior to the nickel conditioner solution is as follows: Alkali cleaner: 2 to 3 minutes. *See p. 46 for controls to be applied. **Hydrochloric addition depends on the aluminum stock; more than 1400 ml may be required. The hydrochloric acid is added until the bond to the aluminum is satisfactory (i.e., until it melts and flows on the surface of the aluminum without loss of adhesion or the formation of blisters). Should blisters or loss of adhesion occur during the operation of the bath, hydrochloric acid is added in 5-ml amounts (in a 1-gal-capacity bath) until the defects are eliminated. ***Should pinholes occur in the lead surface during lead plating, 25 to 50 ml of 42% hydrofluoboric acid (HBF4) is added per gallon of bath, or until pinholing stops. This addition may have to be made each day, as the HBF4 and HC1 are lost during operation. The quantitites of hydrochloric and hydrofluoboric acids suggested above are only approximate; the correct amounts depend on the operation of the conditioner solution. Usually pinholes do not occur in electroplate on new aluminum. --- 20 ~ j ~-1~ ' 'm CONFIDENTIA DFIED

CONFIDENTIAL r ENGINEERING RESEARCH INSTITUTE * UNIVERSIt O"'MtC+I J D Hot- and cold-water rinse in running water. Nitric acid rinse: 15 seconds or until aluminum is cleaned. Cold- and hot-water rinse in running water. Sulfuric-hydrochloric acid cleaner: 2 to 5 minutes or until aluminum is cleaned. Cold- and hot-water rinse in running water. The hydrochloric acid and hydrofluoboric acid generally need replacement every day, as the concentration of each acid is changed due to (1) air agitation, (2) drag-in, (3) drag-out, and (4) temperature of the bath operation, etc. Photographs 4a, b, and c, illustrate the effect of imperfect concentration of the above acids in the conditioner solution in which the aluminum grids are dipped, while Photograph 4d shows a satisfactory lead plate. The loss of these acids has been traced and found to produce certain definite undesirable characteristics of the aluminum surface. Thus, a deficiency in hydrochloric acid and chloride ions in the conditioner solution causes blistering. The size of the blisters is roughly proportional to the deficiency of hydrochloric acid. Thus in Photograph 4a there is a deficiency of hydrochloric acid but the blisters are small. In Photograph 4b there is a greater deficiency of hydrochloric acid and the blisters are larger and more widely distributed. In Photograph 4c there is not such a deficiency of hydrochloric acid, even though a large blister has occurred; this blister seems to be the result of oil acquired from handling. Fine blisters indicate a slight deficiency of hydrochloric acid. Photograph 4d shows a satisfactory electroplate in which the hydrofluoboric and hydrochloric acids are properly balanced. The amount of hydrochloric acid required to be added each day is dependent primarily on the volume of the bath, the temperature of the bath, the rate of aeration or agitation, and the number of grids conditioned. What has been said above in regard to the addition of hydrochloric acid to the conditioner solution will apply equally as well to the addition of hydrofluoboric acid to the lead-plating bath. In the 34-liter conditioner solution we are using, it has been found necessary to add approximately 100-150 ml of hydrochloric acid (sp.gr., 1.19) for each 24 hours that the bath is operated and maintained at l10-1150F. Photograph 5 illustrates the formation of pits in the surface of lead-plated aluminum grids which have been attributed to a deficiency of hydrofluoboric acid in the conditioner solution. A deficiency of hydrofluoboric acid in the 'lead-plating bath may also cause pits, but the primary cause of pitting is found in the conditioner solution. We have found that the addition of approximately 100 ml of 42% hydrofluoboric acid, to a 21 CONFIDENTIAL LaLicHD i

::.., -.,: X E -. te~(a) (b) (-) deficiency of hydrochloric acid manifested as small bliste.-,.; large deficiency of hydrochloric acid; (c) The blister is larger than 4(b), but there are not. as many. A less deficiency of hydrochloric acid than in 4(b); (d) See page 23. _ rT, 1 1 r r i _1 (I _ I. -Photo h v Effecof Hydrochloric Acid Concentration in the Conditioner Solution In Heated Lead JA, ' ^ d ^^ciency of hydrochloric acid than in 4(b); (d) _e e 2 /: I / i

yy "" '"'i "" ", ^.~.......,.. ^,...... 0. 060":4 0.0.50" i1 I 0 ~.. -'.1 I '! I i "' _- r~.. g.....................j.... Poga A Stsaor Led Elcroi o'.o ^^ l"'^'._'L fti ){ i'**r j^~stinii iiiiiiiiii iiii- i:.. V;^-^^ ' ['_ i '.i Z': ~^ t..................~......... ' ~.:...-..-......................iii..i......i......ii.~.."ii..i......i...,,........!.:. ---.:.....,.;.-~.llll.......ll. *.,,.. g m $.'^!' t^^1__^_____IL. ii~~~~~ llllllllll:^__._l^~l~~l~, ' 1.. 'CI^^X^I^- ' " ^ I.., I -- i 3 ' >:'::i" 'i 1 '1 ~-~_-.... i3~"

CONFIDENTIAL rt i v W -E - i b-grga......:.. '"'**3'Hm~ -.I i 'i i...'*- t W..-I. =1- 1=1::: =1=1 Photograph 5. Effect of Hydrofluoboric Acid. Concentrations in the Conditioner Solution on Grids as They Came from the Lead-Plating Bath. 24 _~o _ _ _~~~~~~~~~~~~~~~~~~ _, __ _i

CONFIDENTIAL F57 ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIG 34-liter conditioning bath is required for each 24 hours of operation at a temperature of 110-115F. Lead-Plating Bath (Step 16) The lead-plating bath is a lead fluoborate bath which has been modified slightly from that generally used in the industry. The composition of the bath as we have been using it is given below: Pb(BF )2 (5o%) 485 ml/l IBF4 (42%) 92 ml/l Distilled water To make 1 liter Specific gravity 1.320-1.390 Agitation of the Bath. The agitation of the grids attached to the movable cathode in the bath has been found helpful in reducing the treeing of lead deposits, especially with high current densities. In addition to the agitation of the cathode, it has been found that additional air agitation of the bath in and around the anodes will help prevent channeling between the anodes and cathodes. The air bubbling may be at such a rate as to produce a slight foam on the surface of the bath. The foam does not seem to interfere with the deposition of the lead. Temperature of the Bath. A temperature of 75-90~F is recommended. We have plated at higher and lower temperatures, but the higher and lower temperatures were not intentional, but rather were the result of environmental conditions We would recommend that the tanks be equipped to maintain the temperature indicated (75-90~F). Hydrogen Ion Concentration. The pH or acidity of the bath is regulated by addition at proper intervals of hydrofluoboric acid as the physical condition of the electroplate may require. This is necessary because some of the hydrofluoboric acid is lost by air drag-out or evaporates as the bath stands and even during operation. The pH is controlled by regulating the acidity as directed under controls for lead-plating bath, page 45. It seems that if the acidity drops below the value indicated, the deposit of lead is not as fine and free from imperfections as when the acidity is maintained at 92 ml/l of 42% HBF4. Density of the Bath. For ordinary lead-plating of grids to a thickness of 0.015 inch of lead in 20 to 30 minutes, a lead-plating bath having a specific gravity of 1.320 to 1.390 is recommended. Should the current density employed be in the range of 300 to 500 amperes per square foot, we have found it advisable to maintain the specific gravity of the lead fluoboric bath at approximately 1.380, and also to increase the number of anodes which are in the bath. CONF 25IDENTA-. -~ 1 CONFIDENTIAILX...!

CONFIDENTIAL / -. ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIG Addition Agents. Water-soluble Canadian protein is recommended as the best addition agent. Since this type of protein is difficult to obtain, we have, by a series of extractions with water-immiscible solvents such as toluene and benzene, been able to isolate a water-soluble protein from the usual American proteins. However, we have found the processed protein manufactured by Canadian Packers Ltd. to be the best protein for our use; enabling us to plate the grids at a current density of between 200 and 500 amperes per square foot. We have used many other types of additives but the one that comes the nearest to giving as satisfactory a lead plate as the Canadian protein is licorice-root extract, which is prepared by dissolving 5 grams of powdered extract of licorice root in 100 ml of water and filtering the licorice-root water extract. The licorice solution has to be added to the tank at periodic intervals to maintain a fine-grained, uniform deposition. The rate of addition depends primarily on the drag-out and drag-in of the tank in operation Shields for the Grids in the Plating Tank. It has been found necessary to use shields to reduce the treeing of the lead deposit on the grids. This, coupled with mechanical agitation, permitted us to lead-plate at 300 to 500 amperes per square foot without any visible trees. It has also aided us in securing a more uniform deposition of lead electroplate over all the surface of the grid. Pictures of the shields which we have used are shown in Photographs 6 and 11. Satisfactory operating current densities and the temperatures and times of deposition of a lead plate on O.O15-inch-thick aluminum stock are indicated. Photograph 7 shows the effect of shields on the deposition of a continuous even-thickness electroplate without trees. The grids to the right in each view show the effect of using no shield. Voltage. The voltage that is maintained in the lead-plating bath is between 3 and 6 volts. This produces a uniform deposition of lead. Relation of Current Density Distribution to Character of Lead Deposited. Up to July 1952, lead-plating had been done at a constant grid current density from the beginning to the end of the lead deposition period for a given grid. This method of lead deposition favors the formation of spongy lead on the surface of the aluminum and a much less spongy lead deposit near the surface of the lead electroplate. The dense lead near the surface of the electroplate promotes sloughing of peroxide, due to expansion of lead when it changes to lead dioxide, while the spongy lead near the aluminum grid favors penetration of the lead plate by ions of the electrolyte and provides room for expansion as the lead dioxide is formed. On discharge the formation of lead sulfate requires additional volume over and above that provided by the change from metallic lead to 26 CONFIDENTIAL DE-....ii

CONFIDENTIAL (a) (b) (c) -: api 6. Shields Shown in Previous Report. Shi elds Used at the Beginning of Project to Hel lre!., i.,,.. AlL'inu.ni Grids in 1949. b. Srlield Used to!:late:.;I:I-.. ~-a Ti.me, Based on Shield Shown in Photograph 6a.l ' Used in 1951 tC Plate 3 Grids at a Time. CONFIDENTIAL

CONFIDENTIAL (a) WITH SHIELD WITHOUT SHIELD LEAD PLATED AT 300 AMP. SQ. FT. TIME 8 MIN. (b) Photograph 7. Effect of Shields on Deposition of Lead Electroplate: (a) Note trees on lead grids held by the operator in left hand; (b) A closeup of one of the grids in 6(a). CONFIDENTIAL

CONFIDIENTIAL L -~s^SIUIEp ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OFMICHI t lead dioxide; hence, any spongy lead on the surface would help reduce sloughing and provide better penetration and thus greater capacity. In the following table are listed the active ingredients of a lead storage battery and the calculated volume and diameter of a molecule of each substance. The volume of one molecule was obtained by dividing the molecular weight in grams by the density to give the gram molecular volume, and dividing that value by Avagadro's number (6.06 x 1023). The diameter was calculated from the volume of one molecule by assuming that the molecule was a sphere. These calculations are reasonably correct for this theoretical discussion. As a check, the x-ray data for lead show that its unit cell is a face-centered cube with parameter ao = 4.92 A. From these data, the closest approach of any two atoms (which is the atomic diameter) is 3.48 A. That value is reasonably close to the value (3.862) given in the table. First let us assume, -for the sake of simplicity, that the pores in the active material are cubes, (Any simple geometric figure will give the same results.) Although the pores are not simple geometric shapes, they must be figures with plane surfaces, since all the active materials crystallize in figures with plane surfaces. Hence, a cube may be considered as one of the pores. If the cube side is given the dimension "a", then the volume of the pore, Vp, is V = a5 Vp = a3. The area Of any one face, A, is A = a2 and the total surface area = 6a2. Now let us assume that the reaction proceeds through only "t" molecular layers at the surface of the pore. This is justified since it has been established that only one-third of the active material is used in any plate.,Then the volume of the molecules, VR, reacting in this reaction layer is VR = 6a2tD, where D is the diameter of the molecule. '29~>- 29 ~~ I I.... CONFIDENTIAL ~

TABLE 4 VOLUME CHANGE OF ACTIVE COMPONENTS IN A BATTERY PLATE Substance Molecular Density, Volume of one Dia Name Formula Weight g/cm3 molecule x 1023 molecule x 10 -v, cm3 D, cm Lead Oxide PbO 223.21 9.5 3.93 4.22 Lead Dioxide PbO2 239.21 9.375 4.2105 4.3163 0 0 Z Lead Sulfate PbSO4 305.27 6.2 8.0728 5.3619 Z -n - Metallic Lead Pb 207.21 11.337 -.0160.8620 0~~~~~~~~~ rTI ~ Water (40C) H20 18. 1. 2.9703 358424 z z ""I Ice (-200C) H20 18. 0.88 3.754 4.0097 Sulfuric Acid H2SO4 98.08 1.834 8.825 55259 -^"^1r (-200C) 1:rIrr7 ft Ir.:- - f - ^ ^ i

CONFIDENTIAL rF/.-l ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY- 0 The number of molecules involved in the reaction is vR n = -- Vm ' where vm is the volume of the molecule reacting. Let us first consider the positive plate. The pore here is surrounded by PbO2 molecules. Then Vpb02 = 6a2 t Dpb02 VPbO Vp 6a2 t Dpbo 0 VPbO 6a2 t DTb, n = VPbO PbO0 2 b02 The reactions at the positive plate as given by Vinal are: PbQO + 2H20 Pb (OH) 4 Pb+++ + 40HPb++++ + 2e - Pb++ Pb++ + S04 — _ PbS04 40H" + 4H+ - 40. l The sum of these reactions is: PbO2 + H2S04 + 2H+ + 2e =PbS04 + 2H20 For each molecule of PbO2 consumed, there is produced one molecule of PbS04 and two molecules of H20, and one molecule of H2S04 is consumed from the electrolyte. The H+ for the reaction is supplied from the accumulated H+ at the plate. Internally, the positive plate is the cathode and there will be an accumulation of.H+ there. The H+ consumed will be rapidly replaced, since the H+ has an extremely high transference number or migration velocity under a potential gradient. The modern concept is that the H+ does not actually move through the solution but that a Grotthus-type transfer takes place between H30+ That is, H+ forms H30+ with water and the H+ jumps or transfers from one water molecule to the next, thus accounting for the rapid migration velocity. Hence, there will always be a preponderance of H30+ or H+ at the positive plate for the reaction to proceed. The S04- associated with the 2H+ will be used in the reaction at the negative plate, to which it migrates. 31 D-E-NA L -E CONFIDENT I L b I --- t I

CONFIDENTIAL! DL WD ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN The volume changes of the pore itself will involve only the Pb02 and the PbS04 molecules. The Pb02 is replaced by PbS04 and the change in volume of these moleucles will determine the change in volume of the pore in the active material. Electrolyte volume alterations will not influence the pore size, since the electrolyte is a liquid within the pore. However, the change in electrolyte volume in relation to the change in pore volume will dictate whether electrolyte flows into or out of the pore. The relation of electrolyte volume and pore volume will therefore be taken up after the change in pore volume is discussed. Since only Pb02 and PbS04 molecules are involved in the change in pore volume, this change for n molecules of Pb02 reacting is: AV V -V VP PbS04 PbO2 Vpbs04 n VPbS04 nPbSO4 4 VPb02 = n VbO2 therefore AVp = n (VbSO4 vpb n x 10-23 (8.0728 - 2.3105) (2) = 3.8623 n x 10-23 This means that the pore is being closed rapidly by the accumulated PbS04 molecules and that small pores are completely closed off from reacting further before all the available PbO2 molecules have reacted. Thus available capacity has been reduced. That small pores may be completely closed is seen from the following: Vpf = Vpi - AVp = Vpi - n (vm) where Vpf = final pore volume and Vpi = initial pore volume = a3. 32 CONFIDENTIAL — __ _

CONFIDENTIAL DE-CL ssFIE ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICH- - Vpf = a - n (2vm) 100 Vf = change = 100 a -n (v) Vpi a 3 = 1- (Vm) 100 (3) a- J 100 6a2 t DpbO2 = K2 t Vpb02 f100 I [ Ka2 t (Zvm) ioo10 a50 j o LI 0 100, o 100 pi ' L a3 since vm is constant. Thus, as the pore size gets smaller the percent change becomes increasingly larger, so that small pores may be closed. Another factor reducing capacity may be seen from the molecular dimensions given in the table. The PbS04 molecule is so much larger than the PbO2 molecule that it blocks off or covers unreacted Pb02 molecules, and prevents them from reacting, thus reducing capacity. Hence, besides the closing of the pore, the available reactive material is blocked off. It is conceivable that many pores large enough to remain open are not reactive because the PbSO4 molecules have covered the unreacted PbO2 molecules. These are the reasons why only about one-third of the active material of the positive plate is utilized. Another factor governing capacity and the reason for a spongy lead near the surface of the lead electroplate is acid concentration in the pore or at the reactive surface. The change in volume of electrolyte, AVE, is AVE = n vH2S4 - 2n vH0 (4) = n x 10-23 (8.825 - 2 x 2.9703) = 2.884 n x 10-23 As is observed in practice, the electrolyte volume shrinks. Thus, if a pore did not change in volume, fresh electrolyte would be forced into the pores, thereby aiding the process of diffusion. However, if the pore closed faster than the electrolyte volume shrank, acid would be forced CONFIDENTIAL JE Wi.L

CONFIDENTIAL I EN CII ^miErD ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OFMICRTGANout of the pore, inward diffusion would be hindered, and the acid density in the pore would decrease at an abnormal rate due to the accumulation of water. Theoretically, then, what does happen in the pore being studied? From the above, it has been calculated that AVp = 3.8623 n x 10-23 (2) but AVE. = 2.8844 n x 10-23 (4) Hence, the pore is closing about 1-1/2 times faster than the reduction in acid volume and the second category discussed above is operating, i.e,, electrolyte is being forced out, inward diffusion is hindered, and abnormal acid dilution is taking place due to accumulated water. This means that capacity again is being abnormally reduced. These calculations may be used to explain the difference in capacity at high and low current densities at room temperature. It has already been pointed out why such excess of active material must be present at low current densities. At high current densities, however, the pores are closing so rapidly that practically pure water is adjacent to the reactive surface, not only because diffusion is too slow to replace the acid, but also because any acid in the pore is being forced out so rapidly that inward diffusion is prevented. The new plating technique deposits dense lead on the surface of the aluminum and spongy lead near the outer surface of the lead electroplate. The spongy lead does not slough off after it is changed to lead peroxide and the dense lead near the aluminum reduces ion penetration from the electrolyte. The spongy lead near the surface of the negative plate favors high negative- and positive-plate capacity, and the dense lead near the aluminum helps prevent ion penetration as indicated by the above study. The initial lead deposition is done at a low current density, while the final lead deposition is done at a high current density. The time factor is not increased by using this plating technique. For an example of the current density used to plate one of the grids in 55 minutes, see Table 5. CONFIDENTIAY- ^ I

CONFIDENTIAL DE.-C ENGINEERING RESEARCH INSTITUTE * UNIVERSITYL IeH AN - TABLE 5 EXAMPLES OF PLATING SCHEDULE Total We ight Thicknes Area of Weight of Plating Schedule Weigh Area f Thickness ^^^ ^ ~ ^,.,Area of Lead of Aluminum uminum nplated Amperes Time, Plated Surface of Grid. to be Grid, per Grid Minutes Lead-lated Grid, in. Plated, Grams Grid, Grid, sq in. Plr~~atr~edc, ~Grams sq in. 0.015 16.61 4.44 5.5 25 38-42 31.36 11.0 15 16.5 15 0.030 20.39 8.9 15 25 68-72 34.81 36 15 60 15 Such a plating schedule will deposit the same amount of lead on the grid as if the current density were set at 11.5 and 19.5 amperes per grid, respectively, for the duration of the entire plating cycle. Filtration of the Bath. The bath should be filtered regularly to remove all specks of dirt which are the result of the solution of lead anodes, contamination from the air, or the introduction of grids, (drag-in) and grid holders in the lead-plating bath. The filtration is done by means of a stainless-steel filter pump which is equipped with felt filter pads. The bath should be filtered whenever the solution shows the presence of specks of any kind, at least once a week. The filtered solution should be clear and sparkling. Anodes. The anodes in the lead-plating tanks are prepared by casting bar electrolytic lead into shapes 1 inch thick, 4 inches wide, and 12 inches long. These anode shapes are drilled and monel metal hooks are placed in the tops to suspend them from the anode. We use 12 anodes in each 10-gal tank. They are suspended in 3 rows of 4 each and so placed that their corrosion will give the most uniform deposition of lead on the movable cathodes. There are two movable cathodes in each bath, suspended inside the shields, and the shields and cathodes move in unison between rows of anodes. This means that the middle row of anodes furnishes lead for cathodes on either side. 35IE AV: CONFIDENTIAL — _ I

CONFIDENTIAL: - ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MI During the casting of the lead some lead is oxidized and some foreign matter may be included in the cast anode. The cast anodes require a sheathing of vinon cloth, either in the form of a sack or wrapped around the anode. The vinon cloth withstands the actions of the acids in the bath. We have used some of these vinon cloths for over two years. Due to the processing of the anode it may become contaminated with a layer of grease before it is ready to be placed in the lead-plating bath. Jence, we advise that it be degreased and handled subsequently with gloves until it is placed in the lead-plating bath, as grease does not function well in the lead bath. Inspection of the Lead-Plated Grid. For Blistering and Adhesion of the Lead Plate: The lead-plated grid should be inspected as it comes from the lead-plating bath for trees, for uniformity of coat, and for tendency to blister when subjected to an elevated temperature. We have found that if the grids, after coming from the leadplating tank, are rinsed in water at 80'F and then placed in an electric oven at a temperature of 220-2500F for 2 to 5 minutes, they will develop blisters if blisters are likely to occur under the electroplate. This step is important because if they do not blister under the conditions cited above, they will take the lead-burning operation without failure. For Ripples in the Lead Plate: Ripples in the lead plate on aluminum grids are illustrated by Photograph 8. If the acidity and all other conditions of the bath are normal, ripples like these may occur if the agitation is not adequate. In order to correct this condition, we have found it necessary to agitate the lead-fluoboric bath with a current of compressed air. This agitation need not be too vigorous,but it should be sufficient to keep the lead-fluoborate thoroughly mixed. Otherwise there seems to be channeling of the lead from the anode to the cathode in spite of the presence of shields or motion of the cathode. This is especially true when plating with high current densities. Relation Between Current Density and Addition Agent. The effect of licorice and current density on lead plating of battery grids was investigated. It was found that the specific gravity was not a critical factor, although at 150 am/ft2 a specific gravity of 1.380 was slightly more conducive to lead treeing than a specific gravity of 1.300, as illustrated in Photograph 9. At each of the above specific gravities, holding the current density at 100 amp/ft2, a decrease in lead treeing resulted from an increase in the licorice concentration. The effect of the licorice can be appreciated more by examining Photograph 10, where different types of shields were used. 36CONFIDENTIAL CONFIDENTIAL......,,!

CONFIDENTIAL {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t _ +.t. '> ': ____ ______ __....._.._,.. I_..... '.WM^ ^>'~X"~!"'~* -i..::.::...^-.^ — ~~~~nll~~ ~allll:ll~"-t..1.~1.~11 i _- _ T _ ]-~" ]]]: ~ 1:r,~..,.], _,=" — _ _ _:-:- _..... i~^~^~~ ~ -MW ~__ _Q_ _ _ _L, '~-" "...l^~. tl~lm 111111... WM U M W i ~ M - ' I l~:-~, -~-...... -— ~ ^ __ C.... _______.........^-... -., ',:- —., _ =_ _1 I a.~l.~!~. W W H H U K H U M H ~ ~i "= 'l '...... _ Photograph 8. Ripples in Lead. _ectropiate Due to Improper Ag-i.tation. 7|_ a_ _ _ _ f ~~~~~~~~~~~ssmr Z~n~ t;>rmo. 1al,n= _- _ 1 | 1 37 iCONFIDENTIAL..,,oS"i. r, _ _. t~~~t.=sow= /!.ov> XtwL:

0.0625 g/l Licorice-Root No Licorice-Root Extract in Plating Bath Extract in Plating Bath Cdd 0 0 C+ ct-ct 0~~~~ (CD 0A (D z z, U,, H~~~~~~C HO - ___ 0cC tjM \' \;-r^ ~ ^^^^ Om d- ~O L<OK O~~~D H HH cC-' C6 (D m rrI CnO C 0 CD 1 n> a) p ~rn., F H-P.^ ^1 ^^ Cf 0^^~~~~~~i I- TlD^ Pi~~~~~~~~~~~~~~i ^7 << O H) ^ I a - ~ ^ ^ -I f (D~~~~~ (D JC~ ~ ~~ ^i= ^ )1 ^ e.-~~~~~~~t ~~~ ct~0 IB ~LJ oo g-^ rrrr rr ~ r1 11I~1 YIIcC II~IJr I I I \ ^1 1^ ^1^ _,_ ^ _:.^........-.^pll* >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ap.. \ -? ~ ~ ~ ~ ~ ~ H \ r ^oCd ~~~~~~~~~~~~~~-~~~~~~~~ "^B^ ~ ~ ~ ~~~~ JBB. l Cl &8 ^ mC ' " ' ' ' ^ '~ " " ~ " ', 1^ i ( ^^ ^......... -^ i ^ -. ^ ^ 1^ ^~

0.219 g/l Licorice-Root 0.141 g/l Licorice-Root Extract in Plating Bath Extract in Plating Bath 0 - - Z o IZ rn mC l - - - JI /~'/ o0~~~~~~~~~~~~~~~~~~~~~~t 0 P — i YI ~...~ --- —--- I I1 0VoHPW^' i ' r \ f ~<~~~~~~~ ~~~~ ~ Ji^ ^S^ ^ ' '~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ~

CONFIDENTIAL r: X Lj^^^,,,w -9ED B TYPE OF SHIELDING "B" "C" "D" 0 I1.D woo s 4 — _C d i, 'H-H _ 00 b.0c6gS g ~ m J* r 1.. 1. -— I - 1^ ^-: I I ~~ -- -- '; t -- --; I ~. --. i'H Photograph 10. Effect of Shielding and Licorice Content on Lead Electroplate. Cond.itions: Room Temperature, Specific Gravity 1.320, Current Density 200 amps/ft2. 4D,~~~~~~ eter:w) X ~~~CNIET~ b I i '

CONFIDENTIAL SFIED TYPE OF SHIEDLING "B" - "C" "D" 0i a g E g I g d g 1 g C Pli bf) t e0 rAX 0pPA bO o, r ~ o![ O,,,t TDype of Type of Trees Trees Formed in Formed in Licorice-Root Plating Bath Free with 0.219 g/, Plating Bath Licorice CONI I DENT....FIENfA ' J! oJ ~~~ ~ ~.L-u-'^^ [ \,

CONFIDENTIAL c,~; ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI It should be noted at this point that the shield used for Photograph 9 is of the type A shown in Photograph 11 and even though the current density (150 amps/ft2) is lower than that used for Photograph 10 (200 amps/ft2), the treeing was more prevalent. As a result of this, shields of the type B, C, and D are recommended. One disadvantage of B, C, and D shields is that the location of the grid with respect to the shield is more critical than with the type A shield. For production work, locating blocks can be used to attain the proper location. For the grids plated in this investigation the location was judged visually, as evidenced by some instances in Photograph 10 where treeing is predominant on one edge and absent from the opposite edge, e.g., shield C with no licorice. As shown in Photograph 10, shield D is the most effective in preventing treeing. This is because the shield overlaps the profile of the grid. If excessive overlapping by the shield is permitted, however, the edges of the grids will be deprived of sufficient lead plate to protect the aluminum base from the battery acid. On the other. hand, if the shield does not overlap the profile of the grid, the edges will have a greater tendency to tree as illustrated by the results with shield B in Photograph 10. Various addition agents such as Canadian protein, protein hydrolysate, gelatin, sassafras-root extract, and licorice-root extract have been added to the lead-plating baths. The Canadian protein seemed to give the best results but it is no longer on the market. In lieu of this, the powdered extract of licorice root was used for this investigation. In order to prepare the powdered extract for addition to the plating bath, a 5% solution of licorice extract powder in water was first boiled and then filtered. The effect of the licorice on the types of lead trees can be observed at the bottom of Photograph 10. In the absence of licorice the lead crystals grew to lengths of about 3 mm, whereas in the bath with licorice the crystal growth was stunted to about 0.3 mm. The size and shape of the lead trees might be used as a rough check of the concentration of the addition agent. Graph 2 shows the maximum permissible current density per square foot allowable with various shields at different concentrations of licorice. It appears that the addition of more than 0.141 g/l of licorice has no effect on the lead treeing, but more addition agent may be used to insure its continued presence for a period of time. The co:nsumption of the licorice has not yet been determined. The aluminum grids used for this examination were grids rejected due to a faulty piercing operation to the extent that one or more ribs in the gridwork are missing. Some treeing occurred on the ends of the broken 42 _ \ CONFIDENTIInrl;i iA l 9 )-' \

CONFIDENTIAL l jj5 K-i HIELD C;" ISHIELD _S I I I I L-~~11 /~^/~> Ouri/^ i 5 CONFIDENTf~t~jth~c, Gk -

CONFIDENTIAL GRAPH 2 250 TYPE D SHIELD (1.320 Sp. Gr. BATH) IL ~. |^ 1TYPE C SHIELD (1.320 Sp. Gr. BATH) 200 w z / I w 0 I L _ / / TYPE B SHIELD (1.320 Sp.Gr. BATH) TYPE A SHIELD (1.380 Sp.Gr. BATH) I-'. R10O0 l S RELATION BETWEEN SHIELD, URRENT DENSITY AND CONCENTRATION OF LICORICE. 50 6.. 0 0.0625 0.141 0.219 LICORICE CONTENT, g/l. 44 CONFIDENTIAL, L.' u — ^'^J

CONFIDENTIAL.... ENGINEERING RESEARCH INSTITUTE ~ UNIVERS jY f: " 0 ribs; therefore, the treeing at these points should b liscounted as irrelevant to the investigation herein involved. Controls Used to Maintain Optimum Working Conditions in the Lead-Plating Bath. The nominal composition of the lead-plating bath was as follows: Pb(BF4)2, 50% 485 ml/liter HBF4, 42% 92 ml/liter Specific gravity 1.380 Tank capacity 64.0 liters Total Acid: The total acid i.e., ml/l of HBF4, may be controlled by titration with standard normal sodium hydroxide: Dilute a 10-ml sample of the bath to approximately 100 ml in a 250-ml Erlenmeyer flask. Titrate with iN NaOH to the first trace of turbidity which remains on shaking. Record ml of IN NaOH required and calculate the total acid as the ml/l of HBF4: A = B x 18.4 where A = HBF4, mi/l, and B = ml, lN NaOH required for titration. Analyses should be made only when the tank is at standard working volume. A low tank will give high results, while a diluted tank will give low results. Under full operation, a daily analysis of the total acid as HBF4 is recommended. Any deficiency in HBF4 should be corrected by equivalent additions. Calculate this addition as: C = (92-A)D, where C = ml HBF4 to add, A = ml/l HBF4 found by titration, D = capacity of tank, and 92 = ml/I of 42% HBF4 of unused tank. The major loss in HBF4 seems to be due to dilution from drag-out and drag-in, in proportion to the number of grids plated. There may be some loss due to evaporation upon standing. CONFIDENTIAL

CONFIDENTIAL...ENGINEERING RESEARCH INSTITUTE UNIVERSIT Lead Fluoborate (Pb(BF4)2): The lead fluoborate may be conoled by hydrometry. A specific gravity between 1.320 and 1.390 is satisfactory if taken when the HBF4 concentration is at 92 ml/l and the tank is at standard working volume. Our observations have shown that over a period of sixty days of operation, the specific gravity remained between 1.384 and 1.388. This may be accounted for by the fact that as the lead is plated from the solution, an equivalent amount of acid is liberated. This excess acid attacks the lead anodes, returning to the solution an amount of lead equivalent to that lost through plating. Thus, the balance is maintained as long as the total acidity (HBF4) is kept constant (92 ml/1) by replacing that, lost through dilution and evaporation. Licorice: The licorice concentration, as mentioned on pages 26 and 42 of this report, can be roughly controlled by observation of the type of "trees" formed when plating at high current densities, i.e., 200 amp/ft2. An excessive amount of licorice will cause black spots to appear on the surface of the plated lead. Special Controls used with Conditioner Solution No. 2. The nominal composition of Conditioner Solution No. 2, page 20, is NiC12 *6H20 100. g/l HC1, sp gr. 1.19 35 ml/l HBF4, 42% as needed; see below pH 0.30 Specific gravity 1.055 (fresh tank) Tank capacity 40.0 1 Analysis of Nickel as Nickel Chloride (NiC12-6H20): Pipette a 25-ml sample into a 100-ml volumetric flask. Dilute to the mark with distilled water and mix thoroughly. Pipette a 5-ml sample of this solution into a 400-ml beaker, and dilute to 200 ml with warm H20. Dissolve 2-3 grams of either tartaric or citric acid in the solution; render ammoniacal with NH40H and heat to 150~F. Add with stirring 15 ml of 2% dimethylglyoxime (dissolved in equal parts of acetone and ethyl alcohol). Allow the precipitate to separate and test the supernatent liquid for complete precipitation. When precipitation is complete, digest for 1/2 hour just below boiling. DO NOT BOIL. Cool and filter through a tared Gooch crucible. Wash 10-12 times with small portions of warm 1% NH40H. Dry in an oven at 115~C for 46 ~. CONFIDENTIAAl DbUL"oriLt

CONFIDENTIAL /.- -- ~ ENGINEERING RESEARCH INSTITUTE UNIVERSI1 HIIA 30-40 minutes. Cool and weigh as Ni(C4H702N2)2. Calculate as NiC12-6H20 in g/l as follows: A = 0.823 x B x C where A = NiC126H20, g/l, B = weight of precipitate, C = 1000/ml in sample* When adding dimethylglyoxime, care must be taken to insure only a slight excess. Dimethylglyoxime is insoluble in water, and an excess will crystallize out, producing high results. Analyses should be made no less than once a week during full operation. The nickel chloride concentration should not be allowed to drop below 80 g/l. To determine the amount of nickel chloride to add, use the equation: D = (100-A)E where D = grams nickel chloride to add, A = g/l nickel chloride analyzed, and E = tank capacity, 1. Tank makeup and HC1-HBF4 Control: The tank is made up without initial hydrofluoboric acid. The correct amount of nickel chloride (NiC12'6H20) is dissolved in distilled water with heat and filtered before addition to the tank; 1 liter of water will dissolve about 2500 grams of NiC12 *6H20. With the initial 35 ml of HC1 per liter present, the pH of the conditioner solution was found to be approximately 0.50. Our studies have shown that this pH is high for successful deposition. By adding to the tank both HC1 and HBF4 until a pH of 0.30 was reached, a satisfactory deposition was obtained. During the first seven days of operation of a new tank, during which 370 grids were plated, 850 ml of HC1 and 755 ml of HBF4 were added. The pH, after the above-mentioned additions, was found to be at 0.25. As stated in procedure, a 25-ml sample is diluted to 100 ml, and 5 ml is taken for analysis. Thus 1000 4000 C = == - = 800. 25/20 5 CONFIDENTIAJ _L u;l I

CONFIDENTIAL ^^ubrl:DENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF A It is recommended that HC1 and HBF4 be added in nearly equal quantities until a pH of 0.30+0.05 is reached; and thereafter, at intervals necessary to keep the pH within the range 0.25-0.35. Due to drag-out, evaporation, and other factors depending on local conditions, considerable loss of acid concentration will occur. It is therefore necessary that daily pH readings be made, and in all probability it will be found advantageous to make small daily additions of both HC1 and HBF4. Dissolved Aluminum: The deposition of nickel on aluminum is essentially a displacement reaction, and equivalent quantities of aluminum are displaced into solution by the deposited nickel. There seems to be a point at which the displaced aluminum makes the satisfactory deposition of nickel very difficult if not impossible. It appears that as this point is reached, further additions of H11 and HBF4 do not compensate for the diverse effect of the high aluminum concentration. Observations have shown that when the specific gravity approaches 1.130, the nickel deposition becomes unsatisfactory indicating an excess of aluminum ions. At this time the only solution to this problem appears to be substitution of a fresh conditioner bath. We have conditioned over 5000 aluminum grids in a dip tank which has a capacity of 40 liters. ASSEMBLY OF LEAD-PLATED ALUMINUM PLATES INTO ELEMENTS AND THE PRODUCTION OF BATTERIES FROM THE ELEMENTS In Table 6 is outlined the sequence of operations which we follow in the assembly of lead-plated aluminum grid elements and their subsequent assembly into various-sized batteries. Some of these operations will not be discussed, as they are rather general and any type of equipment may be used to conduct the operation indicated. The only steps on this flow sheet to be discussed here are Steps 2, 6, 7, 8, 9, 13, 18 and 20. The other steps are selfexplanatory. Pressing Grids (Step 2) In Step 2 the grids are pressed to a uniform thickness. If it is desired to produce a 0.060-inch-thick lead-plated aluminum grid from 0.030 inch-thick aluminum stock through the use of 0.015-inch-thick CONFIDENTIAI I lED

TABLE 6 FLOW SHEET FOR ASSEMBLY OF LEAD-PLATED ALUMINUM-GRID BATTERIES Receive from Drying Rejects to Lead Oven and Inspect >Melt for Reclaiming 1 2 3 4 6 7 Cut Lug Press to Paste Positives Stack and Dry Burnish Assemble Dry Lead-Burn to Length Thickness ~ and Negatives ~ Oxide Paste Lugs Plates in Jig Plate Strap O O Z 13 Z In^~ ___12 11 Form at 0.02-amp/sq in. 10 8 Positive Area- Connect Place Cells in |Positive Area; PCells Insert Inspect Lead-Burn nm | Discharge in 1.100-sp.gr. in 1.100-sp.gr. H Separa r Posts Z HH2S04 at End Form; Forming Tani Separa s Ss Cl P s Reform in 1.100-sp.gr. H2SO4 - ' I 14 15 17 18 19 20 Change from Assemble in Insert Lead-Burn Cell Charge *'; [l.5O:~i~~r~oHS BteyAdd 1.283- Top | 1.100- to ~ Battery Cell Connectors and. for 1.300-sp.gr. H.S Case sp.gr. H2O4 Covers Terminal Posts T Ho _j( ~~~~~~~~~~~~~~~~~~~I I C' ' rr, I -,- - I CI 0_ I.

CONFIDENTIAL ENGINEERING RESEARCH INSTITUTE * UNIVERSIT GA electroplate, the lead is plated approximately 0.016-0.017 inch instead of 0.015 inch. This procedure produces a uniform thickness for 0.015 inch and a slight excess of 0.001-0 002 inch on the edges of the ribs of the grid. The pressing of the grid to a 0.060-inch thickness by use of a spanking press set for a thickness of 0.060 inch will tend to form the edge of the ribs so that they have a V-notch on them, which will assist in holding the active material in place. Such a spanking and reduction in thickness does not seem to affect the adhesion of the lead electroplate. Assembly of Dry Plates into Elements (Steps 6, 7, and 8) The assembly of the dry grids in the burning jig (Photograph 12, Step 6) and the lead-burning of the plate strap and the cell posts into position (Steps 7 and 8 of the flow sheet) are discussed as one topic. Lead Burning of Lead-Plated Aluminum Plates (Steps 7 and 8) The burning techniques for lead-plated aluminum plates are little different from the burning techniques employed in the burning of cast antimony-lead alloy plates, Flame. The burning may be done with either a gas-oxygen, hydrogenoxygen, or acetylene-oxygen reducing flame. It is important that a reducing flame be used to prevent surface oxidation, i.e., formation of lead oxide. The molten pure lead flows and welds easily under a reducing flame. Regulate the size of the reducing flame to correspond to the size of the burning operation. Plate Lugs and Plate Strap Assembly. Photographs 12 and 13 show the relative position of the plate lug and the burning comb. It has been found advantageous to use a mold of steel to form the plate strap (see Photographs 14 and 15). The extension of the plate lug (Photographs 12 and 13) above the top of the comb should not exceed half the total final thickness of the plate strap. If the plate lug extends closer to the top of the plate strap, blisters or pinholes may occur in the plate strap which are difficult to remove. Burning techniques. Now that the plate lug extension above the burning comb has been fixed (Photograph 13) and the steel mold used to form the plate strap is in position (Photograph 15), use a reducing flame and puddle lead from a pure-lead burning stick into the mold (Photograph 15) C5OFD0 j ~ --- E CONFlnFNTIAI D b^^ ^O;iED

CONFIDENTIAL:: Photograph 12. Relation of Plate Lug and Burning Comb.: 5 V_..: Photograph 13. Assembly of Plates in Burning Comb. 51 CONFIDENTIAL

CONFIDENTIAL Photograph 14. Placing the Steel Mold Around the Plate Lugs Extending through the Burning Comb. Photograph 15. Puddling the First Layer of Lead from a Burning Stick into the Mold in Building Up the Plate Strap to the Desired Thickness. 52 CONFIDENTIAL

CONFIDENTIAL li- B ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI so that the thickness of the molten lead extends between 4/5 and 5/6 of the distance from the top of the burning comb to the top of the plate lug which extends above the burning comb. This is called the first layer of lead. Melt the top of the first layer of lead about the lugs and add. additional lead from the burning stick to form a continuous lead strap to the top of the mold (Photograph 15). (Note: If desired, molten lead may be poured into the mold to a depth of 4/5 to 5/6 of the distance from the top of the burning comb to the top of the lug which extends above the burning comb. The top of this first layer of lead should then be melted and the thickness of the strap completed by running lead from a lead burning stick to the top of the mold, 'Photograph 15 ). The first layer of lead in the burning mold lends mechanical strength to the plate lugs and helps reduce running of the lead on the plate lugs when the balance of the strap is welded or sealed to the first layer of lead in the mold. An expert lead burner may produce the strap in one operation. Set the cell post on the plate strap in a predetermined position and pass the reducing flame around the base of the cell post to weld it to the plate strap (Photograph 16). In some battery designs the attachment of the cell post and the production of the plate strap may be combined into one operation, as in the use of a combined aluminum cell post and plate strap which has been leadpplated. The hot, completed element may be removed from the comb as indicated (Photographl7) or left to cool in the comb until it may be handled. Inspection of Separators (Step 9) Since the spacing between the assembled positive and negative plates is usually of the order of 0.030 to 0.040 inch, it is necessary to inspect the separators; we have found that as many as 10 or 15 percent of separators of these thicknesses will be imperfect, i.e., contain one or more small openings. These elements can be inspected by passing them over a 250-watt electric light enclosed so as to focus the light on the separator as it passes. These openings, while small, would not be of such great importance if there were a larger separation between the positive and negative grids. However, with such a small clearance for acid circulation to remove sloughing positive oxides, it is necessary that the separators be free from small visible openings. This inspection step makes possible the running of 12-15 S.A.E. overcharge cycles on the batteries before failure occurs. 55.CONFIDENTIA iIED CONFIDENTIAIL UE

Photograph 16. Welding the Cell Post to the Cell Plate Strap. Photograph 17. Removal of the Burned Positive or Negative Cell Element from the Burning Comb. CONFIDENTIAL -.. m, i,. J,, SIFIFP

CONFIDENTIAL - ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHI @F. Formation of Lead-Plated Aluminum Cell Elements (Step 13) The following method is being used to case-form the elements of 2HN 12-volt 45-A.H. batteries. The 6TN 12-volt 90-A.H. batteries are formed at double the rates indicated for 2HN 12-volt 45-A.H. batteries. At the present time we are using 0.060-inch-thick positive plates and 0.030-inch-thick negative plates assembled with wrap-around pormax separators 0.030 inch in thickness. The temperature of formation is maintained between 80 and 90'F by setting the unsealed battery case with the elements connected with temporary pure-lead straps in water. The formation is done in 1.100-sp.gr. sulfuric acid. The formation cycle may be summed up as shown in Table 7. TABLE 7 CASE-FORMATION CYCLE OF 0.060-INCH-THICK-POSITIVE PLATES ASSEMBLED WITH 0.030-INCH-THICK NEGATIVE PLATES Specific Gravity of Acid, 1.100; Temperature, 80-90~F Current Current for Eight Operation Type of Time, Operation YB Density, Positive Condition Battery Amp/sq in. Plates per Element, Amp 1. On charge 2HN 0.015 5 30 Sulfate cleared. 12-volt 45-A.H. 2. Discharge 2HN 0.009 3 3-3-1/2 To end voltage 12-volt of 10.8 volts. 45-A.H. 3. On charge 2HN 0.015 5 21 Positive a uni12-volt form chocolate45-A.H, brown color at close of formation period. CONFIDENTIAL II ^bF

CONFIDENTIALFEN ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY IG Dump the 1.100-sp.gr. acid, add 1.300 sp.gr. acid, and charge at 3 amperes for 1-1/2 hours; measure the specific gravity of the acid and continue the charge until the acid does not gain in specific gravity during a l-hour interval between specific-gravity readings. Now adjust the specific gravity of the acid while charging at the 3-ampere rate. The mixing of the acid is slow due to (1) the thinness of the separator ribs and (2) the wrap-around separators employed. This formation is conducted in the case in which the battery is to be housed. The battery is completely assembled except for the covers and tarring and burning of the connectors and terminal posts, which is done after the battery has been formed. Lead-Burning Cell Connectors and Terminal Posts (Step 18) In step 18 the cell connectors and terminal posts are burned. The cell connectors are burned as in the production of ordinary leadantimony batteries except that pure-lead connectors are employed. The terminal posts are built up of pure lead in a mold which has the dimensions required of the final terminal posts. The terminal post is built from the top of the cell cover to the exposed top of the terminal post, and the lead sleeve of the cover and post are sealed together with molten lead. Charging of Lead-Plated Aluminum-Grid Storage Batteries The charging procedure for type 2HN 12-volt 45-A.H. and 6TN 12-volt 90-A.H. lead-plated aluminum-grid batteries is given below: 1. After the 2HN 12-volt 45-A.H. battery has been discharged to an end voltage of either 10.8 or 6.0 volts, the specific gravity of the acid and the number of ampere-hours obtained from the battery during the discharge cycle are recorded. This number of ampere-hours obtained from the discharge cycle is then multiplied by 1.1 to give the approximate ampere-hours to be charged into the battery on the subsequent charging cycle. 2. The next battery is charged at 2.5 amperes until the number of ampere-hours returned to the battery equals the number of ampere-hours obtained during the discharge cycle. At this point the charging rate is reduced to 1.5 amperes and charging continued at this rate until the specific gravity of the acid is between 1.285 and 1.290 and the voltage of the battery measured across the terminal posts, while being charged at 1.5 amperes, lies between 16.20 and 16.50 volts. When the specific gravity of the acid and the battery potential reach these values, the battery is completely charged. CONFIDENTIAL..

CONFIDENTIAL| ENGINEERING RESEARCH INSTITUTE ~ UNIVE HI 3. The completion of the charging cycle will frequently be reached when the ampere-hour input equals the product obtained by multiplying the ampere-hours of discharge by 1.05, providing the above recharging procedure has been followed. 4. Since the plate thickness is the same in both cases, the 6TN 12-volt 90-A.H. batteries are charged in the same manner except that the initial charging rate and the finishing charging rate indicated in Table 8 below are used for 16- and 17-positive-plate assemblies. TABLE 8 CHARGING RATE FOR LEAD-PLATED-ALUMINUM-GRID BATTERIES Charging Rate, Current Density, Total Area, Type of Lead-PlatedAmp Amp/sq in. sq in. Aluminum-Grid Battery 2.5 0.00864 289 2HN with 7 positives 1.5 0.00501 289 2HN with 7 positives 2.85 0.00864 330 2HN with 8 positives 1.5 0.00454 330 2HN with 8 positives 5.5 0.00833 660 6TN with 16 positives 3.0 0.00oo4 660 6TN with 16 positives 5.85 0.00606 701 6TN with 17 positives 3.2 0.00454 701 6TN with 17 positives 5. The above charging procedure is based on a positive plate thickness of 0.060 inch. If the positive plates are 0.040 inch thick, and the plate area is the same, then the initial charging rate is 4/5 of that given in Table 8 and the finishing rates are the same as those given in the table. In the charging of the lead-plated aluminum-grid batteries it is important to remember that the voltage of a 12-volt battery at the end of the charging cycle should be between 16.2 and 16.5 volts, while the voltage of a 12-volt battery made from lead-antimony grids is not as high. Thus a charging unit which is used to charge a lead-antimony grid battery is usually set to cut out at 14.7 volts, but such a cut out voltage will leave the ordinary lead-plated aluminum-grid battery only about two-thirds charged, or with a final specific gravity of 1.250 to 1.260 (the specific gravity, if the cell is charged to an end voltage of 16.2 and 16.5 volts, will be 1.285 to 1.290). This means that the line voltage near the end of the charge should be such that it will be between 16.2 and 16.5 volts if the battery is to reach full charge. The above charging rates are based on the voltage of 16.2 to 16.5 volts, ^ ~~~., ~...., ~ CONFIDENTI

CONFIDENTIAL ~ 1. iFED ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY As an example, the charging of a lead-plated aluminum-grid 12 -volt battery which contains 0.060 inch'-thick positive plate and 0.030 inchthick negative plates separated with pormax separators of 0.030 inch in thickness will be presented in the report to be issued June 30, 1954. PILOT-PLANT EQUIPMENT USED FOR THE PRODUCTION OF LEAD-PLATED ALUMINUM GRIDS In outlining the equipment required to lead-plate the aluminum battery grids, only that equipment actually needed for the plating is listed in Table 9' and no attempt is made to give actual size of tanks, rheostats, filters, c-c generators or rectifiers, or any other equipment pertinent to the production of lead-plated aluminum battery grids. The choice of this apparatus naturally is dependent on the production desired. However, it is important to set up the controls as outlined and to plan for adequate facilities to handle the job. This equipment list is subject, of course, to change as improvements may be made which delete or add to the existing requirements. TABLE 9 PILOT-PLANT EQUIPMENT USED FOR THE PRODUCTION OF LEAD-PLATED ALUMINUM BATTERY GRIDS Description Number Required Vapor degreaser One Basket for holding grids during degreasing One Steel tank with drain, water, steam coil, temperature One regulator, and overflow; for alkali cleaning Steel tank treated inside to inhibit rust formation, with hot-water facility to operate at 160-180~F, inlet and' overflow dam or water spray; for hotwater rinse Stainless-steel tank or equivalent with drain; for 1.270-sp. gr. HN 3 CONFIDENTI A - = 1

CONFIDENTIAL LA'SSIFIE ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI TABLE;9 (cont) Description Number Required Steel tank treated inside to inhibit rust formation, water inlet, and overflow dam or spray; for cold- One water rinse Steel tank treated inside to inhibit rust formation, water inlet, and overflow drain or spray with heat One regulator to operate at 160-180OF; for hot-water rinse Korseal-lined steel tank or equivalent with heat Two regulator to operate at 110-115~F; for conditioning dip Steel tank treated inside to inhibit rust formation, water inlet, and overflow dam or spray with heat regu- One lator to operate at 160-180"F; for hot-water rinse Korseal-lined steel tank or equivalent with drain; for Depending on lead fluoborate plating solution production desired Steel tank treated inside to inhibit rust formation, water inlet, and overflow dam with heat regulator to operate at 160-180~F; for final rinsing D-c rectifier or d-c generator; size optional Depending on production desired Plating racks, constructed of cast No. 12 Al alloy; Depending on size optional production desired Drying oven, to operate at 220-250~F One Balance scales; for weighing plated grids and Two; one for dry chemicals dry chemicals, one for plated grids Plastic shields designed to block off path of Depends on lead ions and equalize deposit of lead plate plating instalon grid; design dependent on plating installation lation 59............ CONFIDENTIALi.,siFIED I

CONFIDENTIAL L - ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI TABLE 9 (cont) Description Number Required Rheostat with step controls, ammeter, shunt, and One for each voltmeter; Capacity required dependent on time cycle lead-plating and area of grids plated in each tank tank Cathode work rod, with plating-rack holders attached Depends on and with reciprocating motion or equivalent; design de- plating pendent on design of plating installation installation Anode rods Three per tank when designed in tandum. Filtering equipment, for filtering lead-plating One of each solution: (1) Alsop Disc filter or equivalent, stainless-steel construction with motor, acid-resistant hose, stainlesssteel nipple, and strainer (2) Cuno Filter or equivalent for rough filtering of solutions during carbon treatment Cutoff machine, for cutting grid lug to length Depending on production, usually one Press with motor, plate and anvil; for pressing One grids to thickness; size optional, but not less than 50-ton Lead melting pot, oil- or gas-fired; for melting lead to cast anodes; size optional, but not less than Two 30*lb capacity Lead ladle; for pouring lead; size optional, but not less than 12-lb capacity One Anode molds; for making lead anodes; size dependent Two on plating installation Motorized cutoff saw or shears; for cutting lead anodes to length; size dependent on plating instal- One lation and size of lead anodes 6O - CONFIDENTIALi Ut-UL-S6HlED I

CONFIDENTIAL li-;:.F. ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHI TABLE 9 (cont) Description Number Req.uired Motorized drill and taps, for miscellaneous work One Exhaust hood and fan, for exhausting fumes from plating line One Work bench, with vice, hammer, pipe wrenches, assorted files, assorted crescent wrenches, screwdrivers, pliers, One and like tools Hydrometer, range 1*100 to 1.300 Two: one for ENO0, one for leaa fluobor4 -ate bath Burette and stand Two; one for titration of conditioning bath, one for titration of lead-plating bath Pipettes, for obtaining sample of baths; 1- to 10-ml TWO capacity pH paper, acid range 0-1.5, for lead baths Interval time clock; for timing cycles Depends on plating installation Plating-rack loading jig; for alignment of grids Depends on in rack plating installation Bench rack-holding fixture, for holding loaded plating Depends on racks plating installation Air line, equipped with filter and connected to all tanks except alkali cleaner and water rinse tanks CONFIDENTIALi

CONFIDENTIAL L1.CA' SSIF1p ENGINEERING RESEARCH INSTITUTE * UNIVERSI,f GAN. TABLE 9 (Conc) Description Number Required Dolly or truck for handling lead pig One Oven to operate at 500-600"F; for heat-treating aluminum grids CHARACTERISTICS OF LEAD-PLATED ALUMINUM-GRID BATTERIES Self-discharge of Lead-Plated Aluminum Grid Batteries on Standby In the January, 1953, report, pages 1 and 2, the standby discharge capacity of three batteries that stood 37 days is reported. This report indicates that these batteries lost of the order of 1 percent or less of their capacity, which confirms previous standby experiments we have conducted on other lead-plated aluminum-grid batteries using different thicknesses of grids. No special study was made of the specific gravity of the electrolyte during standby. Special work had been done by private battery companies and it was reported that the loss in gravity was, in general, less than that of the lead-antimony batteries for corresponding lengths of time. In order to re-evaluate the above findings, six 2HN 12-volt 45-A.HI batteries were charged and set aside for various periods as indicated in Table 10. The batteries employed in this standby experiment were batteries which were produced in our laboratory and shipped to the Detroit Tank Arsenal on February 6, 1953. These batteries remained in the possession of the Detroit Tank Arsenal until approximately April 1, 1953, when they were returned for us to charge and condition for other experiments. Of this lot, batteries 129, 132, 137, 139, 140, and 141 were selected at random for these experiments. The elements in these batteries contained 8 positive 0.060-inch-thick grids, and 9 negative 0,030-inch-thick grids assembled with 0-030-inch-thick pormax separators. The aluminum grids used in these elements had-3/8-inch-tall feet which rested on the bridges in the bottom of the cells. In Table 10, column 2 lists the discharge capacities in minutes at the 25-afmpere rate for each of these batteries just prior to their shipment to Detroit. Column 3 gives the capacities in minutes at the 25 -ampere discharge rate which were obtained from these batteries on recharging ~~.....~r........ A",~ i CONFIDENTIAiL L-iLASSIFIED

TABLE 10 STANDBY RECORD OF SIX 2EN 12-VOLT 45-A.H. BATTERIES; STANDBY ON OPEN CIRCUIT Discharged at 25 25 amp Dis25-amp 25-amp amp Just Prior to Acid at 82F charge at Variation Capacity loss Discharge Discharge Start End Time Number of Standby Period at End of Stand- from Expected Due to ate Time Before Time After of of of Previous 80~F by Period Capacity be- Change in sp.gr. ry Shipment to Return from Standby Standby Standby Discharge Cycle Time, Amp- Start of End of End of Time, Amp- fore Stand- of Acid, o Detroit*, Detroit**, Period Period Days Cycles No. min hr Standby Standby Discharge min hr by period amp-hr fI..___ ain ain _ __ Period Period Cycle () 0 ^ ~129 70 60 1 June 15 July 45 8 9 55 22.9 1.290 1.285 1.175 58 + 55 - 9.65 l ~ 132 72 60 1June 15 July 5 8 9 60 25.0 1.290 1.290 1.175 60 25.0 0.0 0.9 1T 137 75 195 19 0 11 12 70 29.1 1.285 1.260 1.125 70 29.1 0.0 -38.2 rnr ^^^153 r15 139 72 78 1 June 9 Sept 11 11 12 70 29.1 1.290 1.260 1.125 70 29.1 0.0 -45.8 1953 1953 5 1 June 93 Dec 140 72 75e 3 e 186 9 10 75 31.2 1.280 1.225 1.150 40 16.6 -46. -78.5 141 741 June 7 186 9 10 75 31.2 1.285 1.225 1.125 50 20.8 -5.8 Z 1953 1953 1- I *Batteries sent to Detroit Tank Arsenal on February 6, 1953. g**Batteries returned to University of Michigan laboratory on April 10, 1953. - These Batteries were reconditioned and the results of their discharge cycles are ~r I t7 Iu given in Colum" 7 with capacity in inutes.2ovm in column 9. F^rn c..! C'~ -~

CONFIDENTIAL l -. ENGINEERING RESEARCH INSTITUTE ~ UNIVERSI CHI ED and discharging following their return. All six of the batteries were reconditioned just prior to June 1, when each was placed on standby. In this experiment, six batteries were fully charged by our usual charging technique (see page 56), caps were screwed in tightly, and the units were set on a bench in the laboratory. No special precaution was taken to secure them against contamination or accidental contact with any outside sources. No addition of acid or water was made during the standby period. The temperature of the laboratory was approximately 75-85~F. On July 15, after a standby period of 45 days, batteries 129 and 132 were discharged at the 25-ampere-hour rate and the number of-minutes required to reduce the voltage to 10.8 volts was recorded for each battery. It will be noted that the discharge time in the case of No. 129 increased by 3 minutes, while that for No. 132 remained constant at 60 minutes of discharge time. Batteries 137 and 139 were allowed to stand until September 9, 1953, a total of 101 days. Neither of these batteries decreased in capacity during this standby period. The discharge times for these batteries were identical to those before standby, Batteries 140 and 141 were allowed to stand on open circuit until December 3, 1953, a total of 186 days. These batteries gave a discharge of 40 and 50 minutes respectively at the 25-ampere discharge rate, This would indicate a loss in capacity of 46.5 and 33^33 percent respectively. The information obtained on batteries 129, 132, 137, and 139 duplicate previous experiments, but the loss in capacity indicated by batteries 140 and 141 is greater than we had experienced in previous experiments. We have been of the opinion that a loss in specific gravity of the acid in the batteries during the standby period is not an accurate criterionof the discharge capacities of the batteries. The data on batteries 129, 132, 137, and 139 show no reduction in their discharge capacities as compared with the discharge capacities just prior to the standby period, but all these batteries, except No. 132, showed a reduction in the specific gravity of the acid. The amount of reduction in capacity as based on the specific gravity of the acid for these batteries is given in column 17. We have had similar experience with batteries of this type in previous standby periods in excess of 100 days. If we are to credit the figures given in Table lO for batteries 129, 132, 137, and 139,::there seems' to' be no direct correlation, between reduction in capacity of the batteries on standby and the reduction in specific gravity of the acid A similar analysis of the data on batteries 140 and 141 also shows no direct correlation, since the decrease in specific gravity in the acid indicated at least 64 i CONFIDENTIAL"

CONFIDENTIAL ENGINEERING RESEARCH INSTITUTE ~ UNIVER IT RH a 75 percent loss, while the actual loss of discharge capacity was less than 50 percent. In general, a loss in specific gravity after 101 days does indicate some loss in capacity, but there seems to be no direct relation between the two quantities. In Table 11 the specific gravity of each of these batteries is reported week by week during the first 45 days. The specific gravity was measured by the usual battery hydrometer. It is understood that more exact measurements could have been made, but these are more nearly in accord with the values any layman or industrial user of the batteries would obtain. These measurements indicate little or no loss in capacity due to a change of specific gravity, which is borne out by the capacity at the 25ampere-hour rate indicated in Table- 10. In the January, 1953, report, a study was presented on the effect of possible puncturing of the lead coating on one of the lead-plated grids. On page 9 this study is summarized as follows: "At the end of the standby of 37 to 41 days each battery was discharged at the same rate as it had been discharged 37 to 41 days earlier. The data tabulated in the tables referred to in this report indicate that exposed aluminum on the leadplated aluminum grids immersed in the battery acid does not affect the time capacity of these batteries at high discharge rates. The percentage of aluminum in the battery acid increased, but the extent of solution of the aluminum under the lead electroplate is retarded by the formation of aluminum sulfate and aluminum oxide, which help to retard further action of the battery acid on the aluminum grid." The aluminum ion has only one valence; hence, it does not assist standby discharge. If any aluminum were to be deposited on the battery plate, it would soon be converted to the sulfate and to the oxide, which would become practically inactive. The main possible objection to aluminum in the electrolyte would be the ultimate collection of aluminum oxide on the surface of the plates, reducing their active surface areas and thus tending to reduce the capacity of the battery in proportion to the area of the plates inactivated by the shielding of the aluminum deposit. Resistance of Lead-Plated Aluminum Grids The resistance per unit length at 20OC of the grid stem composed of Pb-Sb alloy and lead-plated aluminum grids has been calculated and is given in Table 12. The resistance per unit area at 20~C to the center of the grids at right angles to the plane of the grid is given in Table 13. CONFIDENTIA UL0 I

CONFIDENTIAL UD-LJnSOirjEOD TABLE 11 CHANGE IN SPECIFIC GRAVITY OF ACID IN 2HN 12-VOLT 45-A.H. BATTERIES ON STANDBY FOR A PERIOD OF 45 DAYS Date Cell Cell Cell Cell Cell Cell Average Days on _No,.1 No.2 NO.3 No.4 No.5 No.6 _ Standby Battery 129 Started June 1, 1.290 1.290 1.290 1.290 1.290 1.290 1.290 0 1953 June 8 1.290 1.290 1.290 1.290 1.290 1.290 1.290 8 June 16 1.290 1.290 1.290 1.290 1.290 1.285 1.289 16 June 23 1.285 1.285 1.285 1.285 1.285 1.285 1.285 23 June 29 1.285 1.285 1.285 1.285 1,285 1.285 1.285 29 July 15 128 1.285 1.285 1.285 1.285 1.285 1.285 45 Battery 132 June 1 1.290 1.290 1.290 1.290 1.290 1.290 1.290 0 June 8 1.290 1.290 1.290 1.290 1.290 1.290 1.290 8 June 16 1.290 1.290 1.290 1.290 1.290 1.290 1.290 16 June 23 1.290 1.290 1.290 1.290 1.290 1.290 1.290 23 June 29 1.290 1.290 1.290 1.290 1.290 1.290 1.290 29 July 15 1.290 1.290 1.290 1.290 1.290 1.290 1.290 45 Battery 137 June 1 1.285 1.285 1.285 1.285 1,285 1.285 1.285 0 June 8 1.285 1.285 1,285 1.285 1,285 1.285 1.285 8 June 16 1,280 1.285 1.285 1.285 1.285 1.285 1.2841 16 June 23 1.285 1.285 1.285 1.285 1,285 1.285 1.285 23 June 29 1,280 1.283 1.283 1.283 1,283 1.285 1.2841 29 July 5 1,280 1.285 1.285 1.285 1.285 1.280 1.2833 45 CONFIDENTIAL DL-LoiLel, iLD I

CONFIDENTIAL ULUL TABLE 11 (Cont.) Date Cell Cell Cell Cell Cell Cell A Days on No.1 No.2 No.,.. No..5 No6 5 verage Battery 139 June 1 1.290 1.290 1.290 1.290 1.290 1.290 1.290 0 June 8 1.290 1.290 1.290 1.290 1.290 1.290 1.290 8 June 16 1.285 1.285 1.285 1.285 1.285 1.285 1.285 16 June 23 1.285 1.285 1.285 1.285 1.285 1.285 1.285 23 June 29 1.285 1.285 1.285 1.285 1.285 1.285 1.285 29 July 5 1.280 1.2880 1 1.28 81.20 80.8 1.280 3 July 15 1.28 1.280 1.280 1.280 1.280 1.280 1.280 45 Battery 140 June 1 1.285 1.285 1.285 1.285 1.280 1.280 1.283 0 June 8 1.285 1.285 1.285 1.285 1.280 1.280 1.2833 8 June 16 1.290 1.285 1.290 1.290 1.285 1.285 1.288 16 June 23 1.290 1.285 1.290 1.290 1.285 1.285 1.288 23 June 29 1.290 1.285 1.285 1.285 1.285 1.285 1.2841 29 July 5 1.2 1.285 1.28 185 1.285 1.285 1.285 1.285 35 July 15 1.280 1.285 1.285 1.285 1.280 1.280 1.2833 45 Battery 141 June 1 1.285 1.285 1.285 1.285 1.285 1.285 1.285 0 June 8 1.285 1.285 1.285 1.285 1.285 1.285 1.285 8 June 16 1.285 1.285 1.285 1.285 1.285 1.285 1.285 16 June 23 1.285 1.285 1.285 1.285 1.285 1.285 1.285 23 June 29 1.285 1.285 1.285 1.285 1.285 1.285 1.285 29 July 5 1.280 1.280 1.280 1.280 1.280 1.280 1.280 35 July 15 1.280 1280 1.280 1.280 1.280 128 1.280 1.280 1.0 10 0 45 67 CONFIDENTI l- Fi lIDi

CONFIDENTIALl D-b SSIFWD ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OFMI TABLE 12 COMPARISON OF THE LINEAR RESISTANCE OF LEAD-PLATED ALUMINUM GRIDS WITH SOLID LEAD-ANTIMONY ALLOY GRIDS Description Width, Thickness, Length, Resistance per of material inches inches cm Unit Length ohm/cm 0.030-inch Al 0.5 0.030 1 30.3 x 10-6 0.030-inch Al coated with 0.015-inch of Pb 0.5 0,060 1 26.1 x 10-6 (total, 0.060-inch) 0.015-inch Al 0.5 0.015 1 60.5 x 10-6 0.015-inch Al Coated with 0.0075-inch of Pb 0.5 0.030 1 52.7 x 10-6 (total 0.030 inch) 91% Pb - 9% Sb 0.5 0.060 1 132 x 10-6 91% Pb - 9% Sb 0.5 0.030 1 264 x 10-6 TABLE 13 COMPARISON OF UNIT AREA RESISTANCE OF LEAD-PLATED ALUMINUM GRIDS WITH SOLID LEAD-ANTIMONY (8%) ALLOY GRIDS Total Traversed Resistance per Description Area, of Material TThickness, Thickness, c2 Unit Area, _____of Material inches inches m__ ohms/cm2 0.030-inch Al coated 6 with 0.015-inch of Pb 0.015-inch Al coated 0.030 0.015 1 0.450 x 10-6 with 0.0075-inch of Pb 91% Pb - 9% Sb 0.060 0.030 1 2.06 x 10-6 91% Pb - 9% Sb 0.030 0.015 1 1.03 x 10-6 68 f

CONFIDENTIAL Li. AiLD] ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Comparison of the Surface Corrosion of Lead-Antimony and Lead-Plated Aluminum Grids To compare the rate of surface corrosion of cast lead-antimony battery grids and lead-plated aluminum grids, a cell consisting of one positive and two negative lead-antimony unpasted grids was connected in series with a like cell containing lead-plated aluminum grids. Each grid was accurately weighed before starting the surface-corrosion test. The weights of the lead-antimony grids were recorded as one group and the weights of the lead-plated aluminum grids recorded as a second group. Between the positive and negative grids of each group were inserted microporous rubber separators having a thickness of 0.042 inch. Rubber bands were put around the outside of the grid assembly to keep the grids in place. They were then placed in 1.283-sp.gr. H2S04 and charged at 0.022 amp/sq ino of positive grid surface area. The grids were charged continuously for 168 hours, after which time they were removed from the acid, carefully dipped in three washings of distilled water, dried for 4 hours at 110~C, and then accurately weighed; the process of charging was then repeated for subsequent cycles of 168 hours at 0.022 amp/sq in. Graph 3 shows the percentage loss in weight of the positive elements of the two types of grids. The lead-antimony grid, as shown in Graph 3, had less surface corrosion during the first three weeks of the test, but after the three-week period the surface corrosion rate of the leadantimony grid increased rapidly with the overcharge and at the end of the fourth week it was more than 1.5 times as fast as the surface corrosion of the lead-plated aluminum grid. At the end of the fifth week of surface corrosion, the positive lead-antimony grid had corroded 1.7 times as fast as the positive lead-plated aluminum grid, while at the end of the sixth week or 660 ampere-hours of charge, the lead-antimony grid had corroded 2 times as rapidly as the lead-plated aluminum grid under the same conditions of acid concentration, current density, and temperature. At the end of the sixth week the lead-antimony positive grid had lost 42.24 percent of its original weight, while the lead-plated aluminum positive grid had lost only 20.75 percent of its original weight. Photograph 18 shows the condition of the positive and negative lead-antimony grids after 660 ampere-hours of charge. Photograph 19 shows the positive and negative lead-plated aluminum grid after 820 ampere-hours of charge. This surface-corrosion test was conducted at a current density during the charging cycle equivalent to-the regular overcharge rate specified for 2HN-type batteries. Weight Change During Cycling of Lead-Plated Aluminum Urrids The purpose of this cycling test was twofold; (1) to determine the number of charge and discharge cycles the lead-plated aluminum grids 69 — ~- ~~.. CnNFinFrMTiA K ^-D

CONFIDENTIAL DU -..;i, -r;r ^^R _ HIS= -i I., l b, _ =-== _ = l- s~~~~~I:4:C: _"... - '_ _. L _ = ':. 1*"..... _",::-:~," _.^ L_: _r.:.., =- = =,.....ir =.=..... t = | _, _ _ _ _ ~~~~~~~..........._. Photograph 18. Lead —Antimony Grid.s after 660 -A.H. continuous Charge at 0.022 amp/sq in. _.~- ~. ~; _- ~ ~ WW WW.....-' _ _ _9__ _~~~~~~~~~~~~~~~~~~~. ___..... '...... [.:...,:i,:..::::- ':.. — '...... W_...... WW........,} = _ -L _ w __ _ __ _ __ WA6LL._i4i=I _W W............ L...... W.::::::::::::::: M_ ---T V.]... _~" I C __ _ __ _ i i _ _ W ~~L '. I;,;'.; 1_;, i....::::.:'::.......::::........... __W W. _ _ -- _ _ _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --- -- Photograph 19. Lead-Platimn Aluminm-Grids a fte 660 A80-. 1::1. conti nuou.s Ch b.arge~ at, 0. O;22 a m::/s cj in. - "" _"" "* * ^ " -* *~ t * * " " ~ ~ ~ * ---- — ^ -"" - - ^*^: 2'. _CNIENTI,,.............. J..~.,..: ' - *.'" """ "'" " "" "'.. 1P.' I ~~~' "M$^ " ~**"~~~" MM. W.:.....~^ ' ^^1 =i t-j ir t 2 88.. Continuous Charge t 0.2 amps n 1 X x - X.. CO FDNIL ^ _ ___, D:

GRAPH 3 660 - / A AM 550 ~~~_ _~ ~,40 0 0 Z 330 ____ Z =.? ^ CORROSION OF LEAD-ANTIMONY '1!~ I /^ AND LEAD-PLATED ALUMINUM! Irm GRIDS. C.D., 0.022 AMP/SQ IN. m Z 220 ^^~20 TEMP 800E Z v PERCENT LOSSI I (. I a. rr"I rn I C' 0 20 30 40~ ( i C,' I ->!~5a l l" r

CONFIDENTIAL -,.sSIFIED ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY1 would stand and (2) to determine the change in weights of both positive and negative grids during the cycling. A description of the cycling is given below. Graph 4 shows the change in weight of both the positive and negative lead-plated aluminum grids at the end of 50, 100, 200, 300, and 400 cycles, respectively. In this test the grids were accurately weighed before cycling. They were then assembled unpasted into a cell consisting of one positive and two negative elements with microporous rubber separators 0.042 inch thick and were placed in 1.283-sp.gr. H2S04 and given a 15-minute charge at 80F and a current density equivalent to 0.011 amp/sq in. of positive surface area. The charging rate selected is approximately equivalent to the regular charging rate used for 2HN-type batteries. At the end of a 15-minute charging cycle the cell was discharged at the same rate as the charging rate. When the cell voltage during discharge dropped to 0 volts, the time was recorded and another 15-minute cycle was started. Graph 4 shows the completion of 500 cycles. The test was continued to breakdown of the lead electroplate at 520 cycles. During the first 200 cycles the time ratio of charge to discharge averaged approximately 60 to 1. Since the charging time was held at a constant of 15 minutes, the discharge time gradually increased, changing the time ratio of charge to discharge to approximately 30 to 1 at the end of 300 cycles, approximately 3.3 to 1 at the end of 400 cycles, and approximately 1.25 to 1 at the end of the 500th cycle. Graph 4 clearly indicates that during a series of cycling under normal charging and discharging conditions, the positive grids show an increase in weight which does not change materially during the first 300 cycles. The negative grids during the cycling show a loss in weight after the first 50 cycles. The opposite condition is revealed in Graph 6, where cycling does not take place and the current is always in one direction. During cycling the increase in weight of the positive grid is due to the formation of PbO2, which is probably held intact by the presence of small amounts of PbSO4 in the positive grid. The negative grid forms PbSO4 during discharge, which is quite insoluble in the H2S04. This PbS04 then sloughs off, resulting in a decrease in weight during cycling of the negative grids. This clearly indicates that a 0.015-inch lead deposit. is adequate for more than 300 cycles. _ 72 ~ 1 ~-. CONFInFNTIAi- d-^

GRAPH 4 Z WEIGHT CHANGE DURING CYCLING OF LEAD-PLATED ALUMINUM GRIDS. C.D.- 0.0011 AMP/SQ IN. 0.10 I ^ POSITIVE 0.9 CZ OS lNEGATIVE lJ 0.302 0 0Z I ___ ___15 PERCENT LOSS 0.50C AT 500 CYCLES 1"^~ ~ 0 100 200 300 400 500 & CYCLES rCtI C. ~o 0

CONFIDENTIAL F^^,^^:^~-^ ENGINEERING RESEARCH INSTITUTE ~ UNIVERS G I PERFORMANCE OF LEAD-PLATED ALUMINUM GRID BATTERIES Comparison 5-Second Voltage, Battery Voltage, and Percent Discharged Study of battery voltage versus percent discharged at various discharge rates has shown that the lead-plated aluminum-grid 2HN 12-volt 45-A.H. battery maintains a very high average voltage. This is essential for a high discharge power factor, whether it is figured on the weight of the battery or per pound of active material. Graph 5 shows the 5-second voltages at the 20-, 5-, 1-, and 1/2 -hour discharge rates for a type 2HN 12-volt 45-A.H. battery made of leadplated aluminum grids, plotting the voltages as ordinate against the percent discharged condition of the battery as abscissa for the various rates. A high 5-second voltage is indicated for all rates. There is also a tendency for a high voltage to be maintained until near the end of the discharge period, when the voltage drops rapidly. This type of discharge insures a high power factor. Graph 6 shows the same information on the same type of battery as Graph 5, except that measurements were taken at 10-, 8.5-, 3.5-, 2- and 1-minute discharge rates. The 1-minute rate curve is dotted because since there was no equipment available to observe the battery being discharged at the I-minute rate accurately the 1-minute rate was calculated by the Peukert equation based on the observed data for the 3.5- and 2-minute rates. The curves on Graph 6 show that a high voltage is maintained until the battery has delivered about 75 percent of its total energy at the 10-, 8.5-, 3.5-, or 2-minute discharge rate. The power factor is high and the battery will serve as an almost constant source of power. Recovery of Lead-Plated Aluminum-Grid Batteries on Open Circuit After Discharge Since the lead-plated aluminum-grid 2HN batteries are designed for low-temperature discharges and at the same time to deliver the proper rating at the 20-hour rate, it has been thought advisable to determine the recovery of capacity at various temperatures, such as -65~F and -400F, and compare these with the recovery of the same type of batteries discharged at similar rates at room temperature. The results of these tests are shown in Graphs 7, 8, 9, and 10. Graph 7 is a study of the recovery rate (Battery No. 110) of the type 2HN 12-volt 45-A.H. lead-plated aluminum-grid battery discharged CONFIDFNTIAI~

5-SECOND VOLT. 12.60.5-SECOND VOLT. 12.52 12.5 z.0 30 NRAE5120-HR. RATE, 2.25 AMPERES 5-HR. RATE, 7.0 AMPERES ( GR^ __ ro30-MIN. RATE.51 AMPERESH 5 rrCOMPARISON OF VOLTAGE VERSUS PERCENT DISCHARGED, \ ' 5 COMAIO o-EoND VOLT.,11.4 +ER"N. 0l 0 10 20 30 40 50 60 O 801 PERCENT DISCHARGED C TYPE 2HN I2-VOLT 45-A.H. BATTERY, AT VARIOUS DISCHARGE C RATES, TEMPERATURE 80F Z ^~~~ ~~~2~ 'r'", I '' ' C^~~1.... -. —i. —, -. — " ~~~~~0I0_ 103 05 0 6 0801 0

5-SECOND VOLT. 11.7 12 5-SECOND VOLT. 11.4 0 _= _~ _^ -_oIO-MIN. RATE, 110 AMPERES i 5-S-SECOND VOLT. 10.8 ^^^, - I 38.5-MIN. RATE, 150 AMPERES /-3.5-MIN. RATE, + E 300 AMPEREST SG-SECOND VOLT..ID T E 0.0 0 -. - _MIN. RATE, 380 AMPERES """"" 1 oL 3.H UNI5-SECOND VOLT. 8.1MICHIGAN I-MIN. RATE 600 AMPERES r= 0 10 20 30 40 50 60 70 80 9010 ~~Cf'~~~~~ ~PERCENT DISCHARGED!~ ~GR-A-PH NO. 2. VOLTAGE vs. PERCENT DISCARGED TYPE 2HN 0C-~~ - I~~12-VOLT 45-A.H. UNIVERSITY OF MICHIGAN 0^e~~~13~ 1 BATTERY ALL DISCHARGES AT 80~ F r^ i rn/..I.

OPEN-CIRCUIT 12.8 VOLTS OPEN-CIRCUIT RECOVERY AFTER 10 MIN.,12.4 VOLTS (1 w l - ~~~~~- C) 0 0 _ _ _ 0 Z~ J0-1 5-SECOND VOLT. 8.7 VOLTS -T1 1 INITIAL DISCHARGE I I I_ S 1 _______ I Z Z ~~ I I AFTER OPEN-CIRCUIT RECOVERY - 1 rF~~~~~~ \ ~TIME TIME [. * I 7 \ | S19 SECONDS 75 SECONDS ~ 6L~~~1~ I_ _ I__ _ _ _7 00 10 20 30 40 50 60 70 80 Fc.' - TIME IN SECONDS J r. i - OGRAPH 7 I F OPEN-CIRCUIT RECOVERY AFTER TEN MINUTES FOR TYPE & 0 2HN 12-VOLT 45-A.H. BATTERY DISCHARGED AT THE 150- ~;::: iAMPERE RATE, TEMPERATURE -65FE BATTERY NO. 110, CYCLE r -~ NO. 5. rO I,

13 +-OPEN-CIRCUIT, 12.8 VOLTS OPEN-CIRCUIT RECOVERY 12 AFT R 10 MIN..12.2 VOLS 10 ^ - ^\\ Q IIIII C o 0 ^ o -T o ~\\ -n IIII z _ _ _ _ _ _ _ _ _ _ Z -1, 0 ~ ~-.^ > \5 SECOND VOLT., 8.2 VOLTS, BOTH INITIAL DISCHARGE AND AFTER STANDING 20 HRS. AT -40"F rn 8 \ _______ ________ _______ 1m1' z I S '' ^INITIAL DISCHAzGE 5- \^SEC ND VOLT, AFTER "^ I1 V ^IOMIN. RECOVERY, 7.0 VOLTS I 7_ _____ 13 SECOND TIME 15 SECONDS L___ 50 SEC. 0 10 20 30 40 Cd 4 TIME IN SECONDS ' ^~~~~"~~~~ ~GRAPH 8 F^r~ ~OPEN-CIRCUIT RECOVERY AFTER TEN MINUTES FOR TYPE 2HN 12-VOLT 45-A.H. BATTERY DISCHARGED AT THE 300- AMPERE RATE, TEMPERATURE -4F BATTERY NO. 110, CYCLE NO. Z

OPEN CIRCUIT, 13.2 VOLTS 13 5-SECOND VOLT., 11.8 VOLTS dt DISCHARGE 12 ^5-SECOND VOLT., 11.6 VOLTS 2 DISCHARGE sof ^^^ _ — ~1^eIt DISCHARGE 10 U a ^ 3\ DISCHARGE 0 _ _ __ _ _ - z~~ _n -1 - n 4t hDISCHARGE m _SCH._______ m -I \ W^ ^ "6th 5th \~ ^ J^8^\\ th DISCHARGE DISCH. D 7 9th ~ DISCH. \ 10 I \ \^ \ TIME \ TIME TIME \TIME TTIME 1 5 \ 9\\65 15 SEC. \25 SEC. \30 SEC. \45SEC. \5MIN. ZJ SEC. SEC. lSEC. V19SEC. 1 0 1 2 3 4 5 rMIN.- 1' 2 0 10 20 30 40 50 EC. TIME t iGRAPH 9 r I OPENWCIRCUIT RECOVERY AFTER ONE HOUR FOR TYPE 2HN 12-VOLT 45 A.H. BATTERY DISCHARGED AT THE 150- AMPERE RATE, TEMPERATURE 80eF BATTERY NO 110.

13 I OPEN-CIRCUIT, 12.7 VOLTS, It DISCHARGE OPEN-jRCUIT RECOVERY AFTER ONE HOUR, 12.4 VOLTS, 2- DISCHARGE 12 - OPEN CIRCUIT RECOVERY AFTER ONE HOUR, 12.3 VOLTS 3r DISCHARGE 5-SECOND VOLT.,9.9 VOLTS _i ^ jL__I0o o __ ^01 0 TIm 5-SECOND VOLTAGE, 8.2 VOLTS Z 8 z\r-~~~~~X~. $rd 1-2.DISCHARGE r. ~ H..M 3 DISCHARGE ~ I.. TIME 5 SECONDS \ TIME -TIME 7 SECONDS 8 SECONDSCONDS Fr"0" 0 0 20 30 40 50 60 70 80 90 00."7 TTIME IN SECONDS Ce^./-. '| iGRAPH 10 I _- - OPEN-CIRCUIT RECOVERY AFTER ONE HOUR FOR TYPE 2HN 7 f r- 1' 12-VOLT 45-A.H. BATTERY DISCHARGED AT THE 150-AMPERE RATE, TEMPERATURE -40*F BATTERY NO. 108. 12 ~~~~~~.. j~,.., I 0~~~~~ C, J...

CONFIDENTIA L Fy(7 -ENGINEERING RESEARCH INSTITUTE * UNIVE si FFI IAN at the 150-ampere rate and at a temperature of -65~F. The battery was allowed to stand for 10 minutes after it had been discharged at 150 amperes at -65oF to an end voltage of 6 volts and then discharged again at the same rate. As shown in Graph 7, the initial discharge was for 1 minute and 15 seconds at -650F at the 150-ampere rate to an end voltage of 6 volts, and the battery recovered in 10 minutes to an open-circuit voltage of 12.4 volts as compared to an original open-circuit value of 12.8 volts. The second 150 -ampere drain at -65iF gave a 5-second voltage of 8.4 volts and a discharge period of 18 seconds to an end voltage of 6 volts. Data in Graph 8 are for a study of the recovery capacity of the same battery as that shown in Graph 7 at -40~F but discharged at the 300-ampere rate to an end voltage of 6 volts. At -40~F the battery had a time capacity of 50 seconds at the initial 300-ampere discharge rate and a 5-second voltage of 8.2. When allowed to stand for 10 minutes at -400F and then discharged under the same conditions, its 5-second voltage was 7.0 and the time required to drop the voltage to 6 volts was 13 seconds. It was then held at -40~F for 20 hours without recharging and again discharged at the 300-ampere rate. On this discharge, the 5-second voltage was the sam as during the first discharge, but the capacity dropped from 50 seconds to 13 seconds. This reveals that the lead-plated aluminum battery plate construction will function at low temperature and high rates of discharge and that it has a recovery characteristic which is very desirable in subzero weather. At a temperature of -650F, there may be temporary ice formation in the pores or on the surface of the plate, resulting from the first discharge at -65 or -40~F, but the ice is readily converted into water and diffuses through the battery, leaving the oxide exposed. The failure to function after two or three strikes at -40 or -65F is probably due to surface discharge effects involving the lead oxide. One such effect is surface sulfation. Graph 9 is a study of the recovery capacity of a type 2HN 12-volt 45-A.H. battery (Battery 110) discharged at the 150-ampere rate at 80~F to an end voltage of 6 volts. It was allowed to stand without charging at a temperature of 80'F for a 1-hour interval and then discharged again at 150-ampere rate to 6 volts. This procedure was repeated until the 5 -second voltage dropped to 6 volts at the 150-ampere rate, which required nine such discharge cycles. The recovery capacity and the discharge time are shown in Graph 9. Since the oxide in the plates had yielded less than half the ampere-hour capacity it does at the 20-hour rate, the failure to function after the ninth strike is considered to be due- in large measure, to the exhaustion of the surface oxides or their conversion to lead sulfate, which insulated or sealed in the remainder of the active oxide from contact with the electrolyte, CONFInFNTIAI __l_-__II

CONFIDENTIAL p57771 ENGINEERING RESEARCH INSTITUTE UNIVER ITY OF IHrG. Graph 10 shows the effect of discharging a 2HN 12-volt 45-A.H. battery (Battery 108) constructed of pasted lead-plated aluminum grids which was discharged at 150 amperes and a temperature of -40~F to an end voltage of 6 volts at 1-hour recovery intervals while being held at -40~F. The results of Graph 10, as compared with the data of Graph 8, on first glance might seem to conflict; but on analysis the difference is quite understandable. In Graph 8, the ampere discharge rate is double that in Graph 10; hence, the acid on the surface and in the surface pores is exhausted readily, with an early drop in voltage due to surface discharge effects (ice formations sulfation, exhaustion of acid, etc.). In this case ice is readily dissipated due to solvation and diffusion, while the sulfation constitutes only a thin layer which is still porous enough to permit acid diffusion. In the case of Graph 10, the initial discharge lasted about twice as long, thus permitting a deeper penetration of the sulfate and of ice formation. While the ice was undoubtedly removed during the 1-hour standby period, the initial sulfation covered the surface of the active material more effectively, thus reducing its contact with the acid electrolyte. Here a state of exhaustion was soon reached on the second strike at -40~F. As a consequence, the ampere-hour efficiency at -40~F favors discharging at the 300-ampere rate rather than the 150-ampere rate by a ratio of almost 3 to 2. Life History of Three 2HN 12-Volt 45-A.H. Batteries Batteries No. 108, 109 and 110 were assembled with 8 positive 0.060-inch-thick plates, 9 negative 0.030-inch-thick plates and 0.030-inchthick pormax separators 5-1/2 inches long by 4-1/2 inches wide. Each grid had two lugs on the bottom 3/16 inch high. The elements were assembled in the cell volumes of a 2HN case and case-formed in 1.100-sp.gr. H2S04 at 7 amperes per element or a current density of 0.02 amp/in.2. At the close of the formation period the 1.100 -sp.gr. acid was dumped and new 1.285-sp.gr. acid added. The specific gravity while charging was adjusted to 1.285. The cycling. of these batteries is shown in Table 14. At the close of the overcharge cycle the batteries were dismantled and the apparent cause of failure noted. Battery No. 108. Battery No. 108 did not pass the 8th overcharge cycle on February 20, 1953. It was dismantled and it was observed that: (1) the positive grids were sloughing freely; (2) the negative plates were shorting across to the positive plates at the bottom and also through the separators; and (3) one cell was shorted across at the top between the plate straps which was the only observable puncture of the lead electroplate. CONFIDENTIA DL

TABLE 14 HISTORY OF THREE 2HN 12-VOLT 45-A.H. BATTERIES Cis - End Temperature Battery 108 Battery 109 Battery 110 -Secon. Voltage Cycle charge Voltage OF No. Rate, per Time ATp- Amp- Amp- Bat. Bat. Bat. amp Battery initial FTime Time T ime amp Battery Initial Final hrs hrs hrs 108 109 110 1 25 10.8 78 -- 1.16 hr 29.0 1.16 hr 29.0 0.91 hr 22.8 2 25 10.8 78 -- 1.28hr 52.0 1.5 hr 53.2 0.91 hr 22.8 - - 25 1o.8 78 -- 1.3 hr 2.5 1.35 hr 55.6 0.91 hr 22.8 - - 4 2.25 10.8 78 - 21.33 hr 480,o 21.45 hr 48.35 Lost 0 5 150 6.o0 -65 - 1.2 min 3.0 1.1 min 2.75 1.9min 8.9 r^ ^ 0.3 min 8. 0^6 25 10.8 8o - 1.25 hr 31.2 1.38 hr 34.5 0.91 hr 22.8 -- Z 7 150 6.0 -4o0 - 2.l6min 534 1.5 1.03 min 2.5 -~ ^ -- 9.6 -- -- 8 0.83 min 8.2 co 7 300 6.0 -4o0 0- -..21 min. 7.0 m 0.08 min 6.0 Z 0.25 min 8.2 8 2.25 o10.8 74 71 22.75 hr 51,18 21.08 hr 46.8 19.5 hr 43.9 - - - z 9 150 6.0 0 8 4.1 min 10.25 2.49 min 6.96 5.0 min 12. 10.5 11.2 10.5 1- *After an initial discharge of 1.9 minutes at 150 amperes at a temperature of -653F, Battery 110 was allowed to stand 10 minutes at 65oF and again discharged under the same conditions, to an end voltage of 6.o afte 0.5 minute. C * After an initial discharge of 2.16 minutes at 150 amperes at a temperature of -40OF, Battery 108 was allowed I +to stand for 20 hours at -40F and then d is charged at 150 amperes to an end voltage of 6.0 after 1.03 minueT i C/~ ' This battery was subjected to a series of discharges at -40oF. The first discharge was to an end voltage fC - C/10 6.0; the battery was allowed to stand 10 minutes between successive discharges to an end voltage of 6.0( After the 5-second voltage reached 6 volts, the battery stood 24 hours at 40F and then gave a O.5-minut discharge at 150 amperes. ^ Fi ___^ ^n~~~~~~~~~~~~~~~~~~~~~~~~~~p~

TABLE 14, continued Dis- End Temperature, Battery 108 Battery 109 Battery 110 5-Second Voltage Cycle charge Voltage F OF ~~No. ~~ Rate, per Time Amp- Amp- Time Amp- Bat. Bat. Bat. No. Rate, per time Time Time N_ amp Battery Initial Final hrs hrs hrs 108 109 110 10 150 6.0 78 105 6.07 min 15.20 6.0 min 15.0 7.28 min 18.0 10.8 11.4 11.4 11 8 10.8 78 -- 4.5 hr 36.0 4.8 hr 38.4 4.75 hr 38.0 - 12 75 6.0 72 80 19.3 min 24.12 19.7 min 24.6 17.28 min 21.5 11.8 11.7 11.9 13 7.5 10.8 78 -- 4.08 hr 30.6 4.66 hr 34.85 4.75 hr -- 14 75 6.o 0 12.5 10.66 min 13.3 6.11 min 7.63 1.1 min -- 11.4 11.4 11.3 15 7.5 10.8 80 -- 4.0 hr 30.0 3.45 hr 25.87 4.45 hr 16 75 6.0 -42 -40 3.48 min 4.3 2.23 min 2.80 4.43 min - 10. 10.6 11.4 Z 17 7.5 10.8 78 - 2.18 hr 16.3 2.92 hr 21.9 4.5 hr - - - - z 1.63 min 4.07 0.8 min 2.0 9.9 9.8 -- c 0 18 150 6.0 -40 -40 0.11 min 27.8 in 0.20 8.2 6.0 -- cm 1 5-0 o0.08 min 0.20 6.0 1z.5.3 min -- 11.8 1.1 min -- 11.6 ""i 0.75 min -- 10.4 -0 >~ ' 0 5 -0.5 min -- 9.9 t i 18 150 6.0 80 80 * 0.41 min -- 9.8 0.25 min -- 8.9 r-,F 1.... 0.16 min -- 8.2.... 0.15 min -- 7.7 0.08 min -- 6.0 19-1*150 6.0 74 82 4.0 min 10.0 2.05 min 5.03 4.58 min -- 11.2 11.311.9 After an initial discharge of 5.3 minutes at 150 amperes and 80~F, Battery 110 was allowed to stand at 80~F ) CGIE and then discharged at 1-hour intervals at 150 amperes to an end voltage of 6.0 until the 5-second voltage and the terminal voltage reached 6 volts. At the close of the 18 cycles, these batteries were placed on overcharge cycling. _I '; U) -

TABLE 14, concluded Dis- End Temperature, Battery 108 Battery 109 Battery 110 5-Second Voltage Cycle charge Voltage ~F No. Rate, per Time Amp- Time Amp- Amp- Bat. Bat. Bat. amp Battery Initial Final | hrs hrs hrs 108 109 110 19-2 150 6.0 75 80 3.13 min 7.8 3.13 min 7.82 4.28 min -- 11.1 10.5 11.4 19-3 150 6.0 74 80 1.73 min 4.3 3.25 min 8.15 3.91 min -- 10.6 11.311.2 19-4 150 6.o 62 78 1.41 min 3.51 2.75 min 6.87 3.65 min -- 8.4 11.06 11.05 19-5 150 6.0 76 78 0.65 min 1.63 2.83 min 7.08 3.05 min -- 8.10 11.05 11.1 19-6 150 6.0 76 82 0.85 min -- 1.2 min -- 2.28 min -- 8.4 10.85 10.9 () 19-7 150 6.0 72 76 0.50 min --.88 min -- 1.98 min -- 8.0 10.7 11.6 O 19-8 150 6.0 75 -- 0.30 min Failed 2.0 min 5.0 2.85 min -- 8. 11.2 11.0 O Z 19-9 150 6.o 78 -- -- 1.43 min -- 1.98 min -- -- 11.8 11.8 -I 19-10 150 6.o 72 -- 1.75 min -- 2.41 mn -- -- 120 11.5 S u 19-11 150 6.0 72 0.88 min -- Run not recorded -- 11.0 -- rn rn Z 19-12 150 6.o 72 -- 1.3 min -- 3.5 min -- -- 10.5 11.5 z "" 19-13 150 6.0 73 0-.80 min -- 2.5 min -- -- 11.0 11.5 -" -. 19-14 150 6.0 74 - -- 1.33 min -- 2.9 min -- -- 11.0 11.2 *"" 19-15 150 6.o 76 -- -- - 0.55 min -- Run not recorded -- 10.0 -- r 19-16 150 6.0 83 ~ ~ o0.88 - 2.6 min -- 11.4 11.4 ^ 19-17 150 6.0 82 -- -- -- 0.33 min Failed 1.5 min -- -- 8.3 11.3 | 7; r 19-18 150 6.o 82 -- -- -- 0.91 min - -. 11.0 C/: 19-19 150 6.0 80 -- -- -- -- -- 0.3 min -- -- --.0 C^ -,1 r ) i _ i, I,',,, CXL l ' 7~. ^^ i ~ ~-f~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

CONFIDENTIAL l ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHI Battery No. 109. Battery No. 109 failed to pass the 17th overcharge cycle on June 15, 1953. This battery was assembled in the normal fashion with 0.030-inch-thick pormax separators. The lugs of the plates which rested on the bridges in the bottom of the cell were separated by only 0.030 inch. It is not surprising that the lugs shorted when the grids began to slough. No punctures of lead electroplate were observed. Battery No. 110. Battery No. 110 passed the 18th overcharge cycle, on June 19, 1953, but failed on the 19th overcharge cycle due to excessive sloughing of the negative and positive plates, which caused bridging of the 0.030-inch-thick gap between the separators on the bridges at the bottom of the cell. There was some shorting on the edges of the grids where the separators were improperly assembled. No other causes of failure were observed. The rise in temperature, excessive gassing, and nonacceptance of charge may be controlled by observing the charging voltage. Another method of controlling these factors, which may reduce the ampere-hour capacity of the battery, is to increase the thickness of the ribs on the separators with a corresponding decrease in positive-plate area. A better method is to eliminate the feet on the bottom of the plates and use a separator which is wrapped around the negative plate so that the negative plate is completely shielded from positive-plate sloughing. Plate Thickness vs. Watt-Hours per Pound of Positive Oxide The watt-hour efficiency per pound of positive oxide shown in Graph 11 for various thicknesses of grids was obtained from the discharge curves at 150- and 300-ampere rates of 2HN 12-volt batteries. Each of the different thicknesses of grids shown in the graph was assembled into the same cell volume, and consequently each assembly differed in its weight of positive oxide. A plainometer was used on the discharge curves of the 2HN 12-volt batteries, using 9.3 volts as the end voltage. The total watt-hours obtained for each of the respective assemblies were then calculated to a one-cell equivalent..This gave the watt-hours from a known weight per cell of positive oxide material, which was converted to the watt-hours per pound and recorded in Graph 11, At the 150-ampere discharge rate the 0.040-inch-thick grid gave 20 percent more watt-hours per pound than the 0.060-inch-thick grid, while the 0.025-inch-thick grid gave 125 percent more watt-hours. At the 300 -ampere discharge rate the 0.040-inch-thick grid gave 300 percent more watthours per pound that the 0.060-inch-thick grid, as compared to 490 percent more watt-hours for the 0.025-inch-thick grid. CONFIDENTIIA-

GRAPH II COMPARISON OF PLATE THICKNESS TO WATT- HOURS PER POUND OF POSITIVE OXIDE. TEMP 80" F uJ 9100 X 0 9 Z 0os Z ^ 1z60~^^ LL o g 8 m. L EA 40 RALTTE t~ 20 cu~ Qo2 I I > / 1 MPERES_^ ^- - r * ctd 125 200 300 ^~C'h~~~~ j ~DISCHARGE RATE, AMPERES m E~~~~~~i~ -n G

CONFIDENTIAL LDECLiED ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICH The weight of the finished batteries with the three types of grid construction is recorded as approximately 35 pounds each. The watthours per pound of battery are higher for the 0.040-inch- and 0.025-inchthick lead-plated aluminum grids than for the 0.060-inch-thick lead-antimony grids at discharge rates of 150 and 300 amperes. The data from which Graph 11 was constructed are shown in Graphs 12, 13, 14, 15, 16, 17, 18, and 19. Failure of Batteries No. 117, 119, and 120 to Charge University of Michigan Batteries No. 117, 119, and 120 are shown on the Detroit Tank Arsenal records as Batteries XXI, XX3, and XX4 respectively. These batteries and elements are assembled in 6TN 12-volt 90 -A*H. cases. The plates are produced from the lead-plated aluminum type of grids, pasted with lead oxides. Each element of batteries No. 117 and 119 consisted of 18 positive 0.060-inch-thick plates and 19 negative 0.030 -inch-thick plates assembled with 4-1/2-inch-wide by 5-inch-tall 0.030-inchthick pormax separators. Battery No. 120 has the same assembly except it has 17 positive and 18 negative plates. These batteries were sent to the Detroit Tank Arsenal on December 19, 1952. They were returned to the University of Michigan on May 30, 1953. The Arsenal reports (1) that these batteries did not accept charge, (2) that the specific gravity of the acid does not return to its origianl value even after 200 hours of continuous charging at constant voltage, (3) that the batteries also gassed excessively, (4) that the acid showed above the top of the grids during charging, then dropped on standby to a level which could not be seen, and (5) that the ampere-hour capacity did not approach that shown in the graphs which accompanied the batteries at the time they were delivered. The batteries were therefore dismantled and the following observations made: (1) The batteries do not accept charge because the upper voltage limit of the charging equipment employed was set to finish a leadantimony acid battery (14*7 volts). A lead-antimony acid battery is usually completely charged when the cell voltage reached 2.55 volts, or 15.3 volts for a 12-volt lead-antimony acid battery. The lead-plated aluminum-grid batteries, however, require a voltage of 16.5 volts to complete the charge. Since the lead-antimony acid battery charging voltage is usually set at 14.7 to 15 volts so as not to 88sIJLUoiCC )Hi[ED I CONFIDENTIAijJ^

CONFIDENTIAL ~ ~t-&v5F VLL~ ~S1T~1 I -' r..... m X wi WH tX g ^ ^ t4Xtw w g -tt, Xt tl,0lf l ' ' -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~;. I ~I 1i I1 -illiililXX iM ill tliiiiit iitl ~ — < '- 1 1 '1 1 W X wm S Xt H<?L$ I- -tTi "ItS ssm t t lllll^ I';i 1 1 1 — - 1^::~ - - =i ffi 1XL1 'j! t ~~~~~~~~~I ]ttl,t t 1;..__ g 'i J............ 1 |~~~~~~~~~~~~~~~~~~~~~; > -:t --— w@t-v*- U 1: ''1 ' / @* 0-B8 " |zg N i: s,*T -~ + _c TJ '1 1 *- -~~- -~ -- - -'1' *- 7'~ - '~ ~~t I....................:f:..~ I 1 it:j: 5 i 1 1 'i. 1 i... i...,.....:.......,.....;...-.. i ' _ _. _ _ t ~...... _... i _...*- ^**- - * ~^ * * l ~- | ~ ~~~@ -~ - ~...,..,........................ ~... _l-1' ' t.ll '' I:' L.__r^~~j _ " r ~.M................,. '..f=^ ~ _....::...T;::..:::... 1 *.;.*- 1 1 1 ".: I::::. I::: '. l 1...::;: ~~~'...,^ ~f - -.. _,f.... +-L-~ --- 1 '.. - -...._.... -L ~tr-1 -f 1: a~j t I: f i l -T...;. Iili / l 1 f lr:: i i i t f l......r..., -. -- z......1.:1: 1 - ': ' ': 1 1 ' ': -.................._ 1.^ - s A. ~;..... ^lll. ^1~'1''1'1 1..-.:1....... — ii -t:i i:;: _ _: 1 * i::1: i - i::1 - i::l:is ii: ^: 1!'~~~~~~~~~........ _ L e' >,-,,^, * —1-;;;:1 T: 1 t —t a S t t.: 1f1 0,::: t 1:1 1:i-:Iw.-l-l. H -.,1:.' <.0:01 j: Iq 10 - r.'1.'.. t j. 1... tt IT..........:i: [i I'-iGRtH 12 1 - I: I 1: 4- T. 1 i II '..T - - t X-r +; f!- I- I i. t. t i tF ) j r | l i Ut 1 X 1 XX 01 '' I'' i ' t ' 'I t I -I i1 ' I i i I | gi: | I'~ I:: | |i 0 il.t04411,Itttt 1 ', CONFIDETIAL ii | at tllittlF~pftlF~..trll.l m -. l+ t. State~~~~~~~~~~~~~~~~ GR^PH 12 VTi -4OFDNI - D /4

I I I I t~ ~ ~ ~ ~~ ~ ~~ ~ ~~ ~ ~~ ~ ~ ~~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 44 - 1 -4 i T —t?TT ~~~~~ I- + lil~~~~~F —:k).. Sd rI r~rT rI T i ~lrl i r IT I M I If —t-~~f~-+i —t —t- l fill II ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~I;t; r ~~~~~~~~rrl ~~~fW II r t~~~~~~~~~~~~~~~~~t 1-;~~~~~~~~~~~~~~~~~~- r r r _e,-tf -~:_1;-I- -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c+ f~~~ i i-ec L,- 1C~f i Ii —iL C —CCCi;-lCC~- -CC;- -— C-~ —C CC~i LIC I1 1I U I?i 1i 1I I 1 II I1 II i77m 1I T1 1 l Ir rI r r Ir rI If Ililk~) I I I I I. 1 i i ir i-t i re~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~IIii — III~~~~~~~I z z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 o i-t o~~~~~~~~~~~~H H

III till~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~cr i- L r I"' CONFIDENTIAL[ tI I r T R M F E 0 H | TR U 00 F THUM m r 4mu gj A i-f4-H — Hi r _r- -,T_ _ ~ _ _ _ _ _ _ F _ S i ' _ '_. t ti- 4L A~~~~ r r. _. T _ X. r 4 XE - T T T 4 ~~~~~~~~~~~~~~~~~ T 4 F F t; i 0H;0 0t] iiii44 f - lli01 E:tf A~ tAttt.- lttt L, 1- t ~- tli it~t WttEMW~mX~t~iltil Ir tl t tX T W0 m tt S 1~~~~~~~~~~~~' 't';'1 '' —' bm~ti mtit iM 7 1- dr t7 W.1.'..1.. '..1..1...............I i,........... I t.. I Xi Q g t Xitl tt. '..; -i T FT iT IiT I Iiii I( l sF t v,............. @ @ * * I * * * %,~~~~~~~~~~~~~ ~~~~~~~~...,.-.'... l~~~~~~~~~~- _ _.,..........I.i;.1 r.....t.t t. 1................. ~itll r-~1-t02t tft 4..!.::::::. 1.; '1 1; t 0 j ' i i.t.4.... ^~~~~~~~~~~~.........t... t:::::::I.;:::::::::|::|::1::1 S:::I: _ _.. _............. F... ', i _,-....~~~~~~~~~~~~~~................. so,,,;,,.- 9I _....'...-... F s ~~~~~~~~~~~~~. F....................................,..... T '7 r................. i1~I.. -7-+ 1...........;R~~~~~~~~~~~~~~~~~~~., -. - i....... _, _..I:.:.:..t..,.............,. I, i T~~~~~~~~~~~tt ti ~.ii~ t.,.,....,.irl '.' _,...;; l::l;::'..l 11.. H... l.....r t t 14:i L~ ~~in;zti L~~~~~ri tff i r i t. ~, I,~~~~~~~n &;m s t - -.-.-....;Itj..l....J.;;L...;l - -- - -;- -i t f..,..,.....VII I LE I, -14I I -t1- H+H-tti(, '.,..::.: —;;-::: '::1'::'::1:' <::t':1:::::'1 ':::::ls t - - i i< _,~ _1 t '-( ~1~. S;. z 1-1; _~~~~~~~~~~~~~~~~. - I.+;*|;t | *I 1J I 1!*t* 1*I.;**t r L,t,; _:i. * i'*;..;t 1t. ~ I J 1 1 11*1J-* -*;rt x r1r~.,, 'I.V|* i.:|,:|*|-.|: --- 0 ~~. f v 1 Wff' 'Il ~ ~(1;; ~(:i::1::t::::M::I::~::'::: 1:: I::i ~ c~ ~ RAP ~~ ~4.. CO NFIDEN...,...lil-...

T11 Ht.~ ~ 14 -L: I r: S II II I SSS rM _ T i S II SI I I IIIII - I t: lt~ if I I I: I I I...1 Li I 1 1 + 4 S I I~~~~~~~~~~~~~~~~~~~~~~~ f4 - -4~~ tt i St:,: if~. 1................ bji +-t t 1~ tr~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ e1 LII!II ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ C C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S4 Z~~~~~~~~IS I S IS S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Z a:~~~~~~~~~fl O Z~~~~~. F -L ]if....... S if ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I -— i ---A~~~~~~.11 fill ISIS —: — — S [if I III I S! ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ --- ——: ---SI S mS S I S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~mi I — - - -~- - I S I II I I ii S S I S I I I I I I I I I S I................. SI I fill 4-4;~~~~~~~~~~~~~~~~~~~~~ii~l I S I IS~I i~ -:-:I; m. i': i:!I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:~ fill I ff~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ';.. '. 1.. I - 1,

.- ~... ~+........... lllll —ll- ll —1ll"ll -'' 'TI I I l lll^ 1 < - | — ~ 1 --- ~~ -1~ -!~ 1~~ 1 -i;~r ~ ~:l11- (l~~l~ltiltttlit~it~ ~~8! lw! 1-r~ | I<<-r~~~~ —~LL-r~..-:~ 1~.- 1:~I~~ ~ 1Zf I:iIlA~k~Ig I I L L `..~~~~... ~ ~.......... n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- m I TX 1- - ~ - ~ —t-~-~~ ~ ~ ~ ~I~~~- 1~- ~~I~~~~,~~ ~ - r ~ < C ~ iK E Ullill~~lkilil;l l l 1 1. 41;1 li s X I tS A tt' Wg s-' -~ ". |L w: 0 Tir I E n I I I i 1 1E S S; Y - | c,. u~~~" -'.: I I. (1 i MM S~~~~~~:^-Q -=- = —^_~~~~ — ==.~~~ —==~ ~^^;rir~ ~=-^='=^^= ~=L==~=-==~=^=-~^ ir^ M ^ ^~ ~ ~ ~ ~ jf ~ ^ ^f --- —t -~-1- -~- ~~ I~~I~~ I~~~ (( 1 ''' I;~;(~ —(,rn~ I-n 'ow -4-4 _^ ^~ - ~ l~~~~~ ~^ ~~ ~ --- —- ~^ nc~ — ~ —~~- ~~ — I ~ ~ -- -- ~ ~ ~~-~ ~ -~ -~'~~~ ^~r~ — 'i-~^*-~ ~~- (n -i —"~ -' -" - ~- ~-~~ ~ ~ ~ ~ ~~ ' ~ ~ —~ ~-* - t i *- ~ + — I~~~~ t*^ ] r 1 *.. ~~ - ~ ___ ___ ~ ~~ & __ __ ~.. - -- -.. -. - -........., 1... i. ~ l I * ' ~ ~ - - - - ~ ~-..-..1...~..-.~ ~ ~ ~.~..- ~<..~.,.. ---^ —.-.. --- —-- c ---- l.-.-,-.-~-.................................. - T7 ---. J C*t —+ --- —---- -, ----c — --- — ~I~ -~~~ -~~- I~~~ ~Ilu+14I ------

I~~~~~~~~~~~~~~~ l7q0( IN LHNO Al NdVf. FfH211: V1:lll~Pl~l4i,|ll~tdll~l1121l~l|!itl~it:4 i t~~~~ ~ -

Z~~~~~~~~~ ~-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I"M~~~~~~~~~~~~~~~~ -r ~~~t i........ -~ - - - - - - - - - - - - - -- q~~~~~~~~~~~~~~~~~~ ------ ----- - ----- - -—. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ H I Ill~I I~~~~~ 11~~~~~~~~~~~~~~7

I ))~~~~~~~~~~~~~~~~~~~p ~'~~~~~4:lX L ^U ^ SUJ ^^\ ":_~::!: _-rbi: r^;. i^~ ll^: j~~~~rt: ^: 1^ "!+il ^::: 4^ T'pt'- 1-^- -T?^ ^-t^ ^^^ ^~^- ~:!i.::.'.-:.4:::::l:!-~-?_':-:~-:'-..:-:-:.:~,.:.-'LT'.-:':-.:^:^,i^ ^^~~~~~~~-4 A ^ ^ ^^. ---..-~ - l-____- -?I I j l>. ' -{ — { - *{ ' j ~ ~~^'' ~}*~^ ^^^ - I' -~~^- - -- ~ ~-;~:~~:- ~ ~:~ ~~;- ~~ ~ -.... -. -__ _ ^. ^f- ^-_^-^ _ - -- - ~ - T P ^}iT ^ f ^ l t T i ^T F ' ' " T ^ r F t ~ r ^ r " ^ "^ ~ " ~ "~ " - '.- - ~- - ~ -- ~ -- ^ ^t^~ ' i -P1- F " ^*~: -r~"" ~'-^ ~;; j l"''"'r ^^' 1" '' T^T "1*" "TT' '^ i "T^""'T " i^ "T ~" ^T " ~^~ ""~" ^"' "" ~ " ~ " ~ ~ ~ " ""~~".;' "T " "~^ T"TT * """ '^T "tiT -~-~~* — ~~~~ ~ ~~-~~~-~ --- l^!r — h^' ^ ^ ~^4 LL- ^^ C i"T~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~ ^ ~...^ ^ n: TT T "p;i T? -— 4 -;! ~? r - -f t - T ^ ' " ^ ' *^ ~ - - ~f^ ~ ~~~ " - ^ —: ^^ ~- t^ ^ - ~~ ~~ -~ - - - --:.^;. - ~ ^ i -- -i^ -;~I ^ ^ ~ " ^ P ~ - ] -— v.- --. - -.v --- -..- - I:i-T f f A ^^^ ~-~ —r~-^-^j~S^^ ~ — ~:~-^-~m~-^-^~^~- ^~- ^ ^T-^^~T ---'-^ -^ —^ — ^~ — ^ —^ -~^4 —r"^ "5-B~" —' —^~^-^^ ---j:^ ----~-~" ---~~ —~-~-^ ^^^~r",L+-. ^ ^ ^ ~ ~.^......4 -~ 4-i ~ i r-1::::^ ^ ^ ^ ^r-L-:::::::::::::::^::::::: ^ ^ \^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~fAt \ ^^ i^~~~~4 v~i 'T " ''Qi '^~ "Ti" "TF '^ """^ "F^ ^T^-==}4T'-'-''- ''-'-' '-' —'- ''-' — '-: ^-: ~- ^"^ 4 1^1^ ^:^:^~ -::::; ^1^:^ E:: ^1311^1':^ ^:^ — ^ ^ '" i^ -:E-:E:E- -:E?::E;= =:-: "=E ~:?::: ~E E:E: ^ ^:^ ^S-~~~~~~~~~~~~~~~~~I 4

CONFIDENTIAL DE-Li ~SlFIED ENGINEERING RESEARCH INSTITUTE * UNIVERSITY O - overcharge the battery seriously, it is apparent that a lead-plated aluminumgrid battery will reach a specific gravity of only about 1.250-1.260 at the voltage of 14.7-15.0, and will begin to function like a battery in an automobile where the generator cuts out at 14.7. The specific gravity of the acid will not increase because the back emf of the battery balances the emf impressed on the terminals by the charging unit. If the current flowing through the battery is too greatly reduced, the battery will sulfate on the charging line if the back emf is as great as the impressed emf. To charge the lead-plated aluminum-grid batteries the impressed emf must be greater than the back emf of the battery, namely 17 volts. It has been observed that when the impressed emf has reached 14.7 volts the battery begins to float on the line as the back emf reaches 14.7 volts. It is obvious the battery cannot charge, as the emf is too low to cause the removal of sulfate from the surface and convert it into lead dioxide on the positive grids or sponge lead on the negative grids. To charge the pure-lead surface of the lead-plated aluminum-grid plate and to convert the desired amount of lead sulfate to active positive PbO2 and active negative sponge lead requires an emf of 16.7 volts. The specific gravity of the electrolyte did not increase above 1.250-1.260, as the emf was too low to change the remaining PbSO4 to H2S04 and PbO2 and/or Pb. (2) The acid did not circulate in battery No. 117, 119, and 120 because (a) the channels between the ribs on the separators became imbedded in the active material of the positive plate due to expansion of the active material of the negative plate, and (b) the cells were not filled with electrolyte to a line above that indicated in Photographs 20a and 20b. The lower edge of the white line across each element in the lower portion of each photograph indicates the level of the acid in the cells. The lack of electrolyte caused sulfation of the top two-thirds of the plates (Photographs 20a and 20b). Since the electrolyte wet only the lower onethird of each plate, the normal charging current density for the battery was virtually tripled as only the lower half of the plates were being charged. This high charging current density accelerated gassing for a normal charging rate. There was also a loss of electrolyte and a rise in temperature, which caused the separators to become welded to the upper two-thirds of the plate and caused an overcharging of the lower one-third of the plates (see Photograph 21). There was excessive sloughing of the positive active material on the lower one-third of the plates. In spite of this abuse the lead electroplate was intact, and there were no shortings. These cells were removed, equipped with new separators, and reassembled; on cycling they were found to have over half of their rated 97 } -n CONFIDENTI

CONFIDENTIAL (a) (b) Photograph 20. Photograph Showing Effect of Improper Charging of Batteries: (a) Edge view of element charged over 200 hours at a maximum impressed emf of 14.7 volts. (b) Side view of same element shown in (a). 98 CONFIDENTIAL

CONFIDENTIAL DEnCL E Photograph 21. Photograph of the Effect of Improper Charging of Lead-Plated Aluminum-Grid Batteries. 99 CONFIDENTIAL. __ _J

CONFIDENTIAL iDE-CLSSIFIED ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MI capacity on the first discharge following the reassembly. If the cells are worked sufficiently, the sulfate may be partially removed and almost normal ampere-hour capacity restored. Should the plates in the heavily sulfated condition be overcharged, however, they may slough excessively before the original balance between PbSO4 and PbO2 or Pb is restored. (3) The gassing was due to (a) excessive current density on the bottom third of the plates as a result of overcharging, (b) the closing of part of the channels between the separator and the positive plates, and (c) a lack of general circulation of the electrolyte. (4) On charge the gas pumped what little electrolyte there was to the surface of the cell element. When charging was discontinued the electrolyte settled to a level two-thirds the height of the element below the top of the element. (5) The ampere-hour capacity could not be expected to equal that of a normal battery, as only one-third of the plates were active, i.e., covered with electrolyte. Battery Inspection and Study of Batteries Returned from Wright Field. The following lead-plated aluminum-grid batteries were returned for our inspectiQo and study on January 8, 1953, after a series of tests conducted at Wright Field, Dayton, Ohio. M7-B-46, a 12-volt 6TN 90-A.H. battery M7-B-45, a 12-volt 2HN 45-A.H. battery M7-B-69C, a 12-volt 2HN 45-A.H. battery M7-B-71C, a 12-volt 2HN 45-A.H. battery Construction of the batteries is shown in Table 15. Batteries M7-B-46 and M7-B-45 had an electroplate system composed of zinc, copper, nickel, and lead, while batteries M7-B-69C and M7-B-71C had an electroplate system composed of zinc, chromium, nickel, and lead. When these batteries were dismantled and an inspection of the elements made, the following deficiencies were noted: Battery M7-B-46: 1. Four of the six cells were dry. Either the electrolyte had been poured out or boiled out, or no water had been added during tests on the battery. 2. Most of the lugs were eaten through and shorted. 3. The principal cause of failure was the diffusion of copper through the lead electroplate of the positive plate and its subsequent depositjn-the-negative plate. CnOFIPnFN -TI

TABLE 15 CONSTRUCTION DETAILS OF BATTERIES SENT TO WRIGHT FIELD Positive Plates Negative Plates Thickness of Number ' Number Microporous Battery Per Thickess, er Thickness, RicbP Date Manu- Electroplate Type Per Per Rubber No. inches inches factured System Cell Cell Separator, ) __________ _._________ __ inches _____ o 0 z M7-B-46 6TN 15 0.040 16 0.030 0.042 5/11/51 usual -1 g M7-B-45 2HN 8 0.040 9 0.030 0.042 5/11/51 partly Cu 1m M7-B-69C 2HN 8 0.040 9 0.030 0.042 11/2/51 partly Cr Z z -.J M7-B-71C 2HN 8 0.040 9 0.030 0.042 11/2/51 partly Cr. w - -. iS ' ' i... c- i I.! '. i _ _, J ~^.J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~9

CONFIDENTIAL l,; ENGINEERING RESEARCH INSTITUTE ~ UNIVER F Battery M7-B-45: 1. Five of the six cells were dry. No acid remained in the cell due to lack of addition of water, spillage, excessive gassing, or boiling on charge. 2. All the plates were sulfated, possibly due to lack of sufficient electrolyte or a low charging voltage. 3. Most of the positive plates were shorted due to solution of the lead. 4. Copper diffused through the lead electroplate and deposited on the negative plates, leaving a very porous lead deposit on the positive plates. Battery M7-B-69C: 1. Four of the six cells were dry. The other two were damp, but the acid was low. 2. The lugs on the positives were all corroded to the aluminum and had shorted. 3. The plates were badly sulfated. Battery M7-B-71C: 1. The cells which were covered with acid were in good condition. 2. The top lugs of the plates were blistered. 3. The top lugs of one or two positive plates of each cell were eaten through and the plates had dropped from the plate strap. The plating job must have been imperfect. 4. The plates were sulfated, possibly due to a low charging voltage. These four batteries were not of the best design nor the best type of construction: 1. No copper should be used under a lead electroplate, as it will diffuse through the lead electroplate and be deposited on the negative plate. Nickel electroplate of 0.001 to 0.002 inch is not sufficient to prevent copper diffusion. 2. The chromium substituted for the copper in the electroplate system did not seem to improve the life cycle of the electroplate system materially. This may have been due to the interelectroplate coat pealing at the chromium-electroplate interface. 3. Copper which diffused through the lead electroplate seems to have left it sufficiently porous to accelerate the sulfation of the lead electroplate and also produce passages through the lead electroplate which admitted acid to the aluminum. Until the passages were greatly enlarged, they were sealed by A1203 or/and A12(S04)3 x H20. This seems to account for the excessive lug corrosion and shorting. 102 _ coNFlnFNTIA!LPp 7' /jlFfl

CONFIDENTIAL - ENGINEERING RESEARCH INSTITUTE * UNI ITMILCHIGA - Design of a 12-Volt 65-A.H. Battery (Battery No. 152) The purpose of this battery design was to determine if a leadplated aluminum-grid battery could be produced with a 65-A.H. capacity based on the knowledge obtained in the production of 2HN 12-volt 45-A.H. and 6TN12-volt 90-AH. batteries. Specifications for the 12-Volt 65-A.H. Chrysler Battery. A 12-volt battery case such as is used in the design of lead-antimony batteries was secured; it occupies 718 cubic inches and has the following dimensions: Dimension Width, Height, Length, _~ l- ~ inches inches inches Outside 7 8-3/8 12-1/4 Inside 6-1/4 6-1/4 11-5/8 Inside of Cell 6-1/4 6-1/4 1-5/8 Each elements is composed of 13 plates: Number and Thickness, Height, Width, Total Plate Area Type of Plates inches inches inches per Element, in.2 6 Positive 0,063 5.25 5.667 356 7 Negative.0063 5.25 5.667 514 Specific gravity of the acid is 1.285, and the separators are the conventional type, approximately 0.060 inch thick. This battery is required to produce 3.25 amperes (at the 20-hour rate) at 80~F to an end voltage of 10.8 volts and 150 amperes for 3.71 minutes at -20~F to an end voltage of 6 volts. Specifications for Lead-Plated Aluminum-Grid Battery. The 12 -volt Chrysler case was blocked off so as toaccommodate the smaller 2HN grids. Side and top views of the case may be seen in Photographs 22a and 22b. The volume of the battery case was reduced 117 cubic inches to a final volume of 601 cubic inches, with the following dimensions: Dimension Width, Height, Length, inches inches inches Outside (Proposed) 5-1/4 8-3/8 12-1/4 Inside 4-1/2 6-1/4 11-1/8 Inside of Cell 4-1/2 6-1/4 1-5/8 --- - 103 NIDT.... _.b.iPIED CONFIDFNTIAI~

CONFIDENTIAL (a) (b) Photograph 22. View of Chrysler Battery Case as Modified: (a) Side view; (b) Top view showing reduction in volume of cell capacity. 104 CONFIDENTIAL

CONFIDENTIAL... ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OT -MttH CiGAbL&._ - Each element is composed of 21 plates: Number and Thickness, Height, Width, Total Plate Area Type of Plates inches inches inches per Element, in.2 10 Positive o.060 5-1/16 4-1/8 420 11 Negative 0.030 5-1/16 4-1/8 462 The separators are 0.030-inch-thick pormax (plastic) wrapped around the negative plate. The specific gravity of the acid is 1.285. A total of 579.5 grams of positive oxide and 414.5 grams of negative oxide was used in each element. The total weight of the element (10 positive plates, 11 negative plates, 22 separators, acid, and lead plate strap) was 6.19 pounds. Performance requirements are the same as for the lead-antimony Chrysler 12-volt 65-A.H. battery. The performance data of the lead-plated aluminum-grid battery is given in Table 16. By reference to the table giving the discharge data of a Chrysler 12-volt 65-AoH. battery, it may be observed that this battery is nearly living up to specifications. Sometimes it exceeds specifications while other times it is just under specifications, apparently because of variation in discharge techniques and the personnel used in conducting the discharges. It should be noted that this battery can be built with a reduction of 117 cubic inches in overall volume as compared to the existing 12-volt 65-A oH battery. If we were to use the present case with lead-plated aluminum-grids of the height and width of those used in lead-antimony 12-volt 65-A.H. batteries, the capacity would exceed those of the proposed lead-plated aluminum-grid batteries and the existing lead-antimony batteries as follows: the 3.25-ampere rate (20-hour rate) could be raised to 4.5 amperes, and the specified 150-ampere-hour rate of 3.71 minutes at -20~F could be raised to 5,1 minutes at the same temperature, -20~F, Inspection Report on 61"4 and 2HI Batteries Sent to Yuma Test Station The batteries included in this test were four 6TN 12-volt 90-A.H. batteries, Nos. 126, 128, 157, and 158 and eight 2HN 12-volt 45-A.H. batteries Nos. 135, 136, 142, 143, 147, 148, 149, and 151. Their construction is detailed in Table 17. The batteries were sent to the Detroit Tank Arsenal Laboratory on February 2, 1953, and remained there until April 20, 1953, when they were 105 CONFIDENTIAL

TABLE 16 DISCHARGE DATA AND LIFE-CYCLE DATA, TO DECEMBER, 1955 Cycle Discharge End 5-Second Time of Ampere- Temp., Rate, eak No. Voltage Voltage Discharge hrs F emars amp 1 25 10.8 -- 1 hr, 50 min 45.7 72 First discharge is generally low. 2 3.25 10.8 -- 21 hr, 50 min 69.8 74 Met low-rate requirement. 5 150 6.0 11.6 12 min, 4 sec 30.2 76 No specification. 4 3.25 10.8 -- 22 hr, 5 min 71.7 80 Met low-rate requirement. 5 150 6.o 11.0 8 min, 20 sec 20.8 0 No specification. 6 3.25 10.8 -- 25 hr, 30 min 82.8 80 Met specification. 7 150 6.0 10.2 2 min,48 sec 7.0 -41 No specification. 8 3.25 o10.8 -- 19 hr, 30 min 64.2 83 Not fully charged. 9 150 6.0 7.0 18 sec 0.75 -65 No specification. 0) 10 150 6.0 7.0 18 sec 0.75 -65 No specification. Q 11 150 6.0 10.2 2 min,21 sec 5.9 -40 No specification. Q 12 3.25 10.8 -- 20 hr 65 76 Met specification. L j13 150 6.0 10.6 3 min,55 sec 9.8 -20 5-sec voltage under specification. 'El H 14 3.25 10.8 -- 20 hr 65 78 Met specification. 'l1 g 15 150 6.0 10.85 3 min,22 sec -- -20 16 150 6.0 10.85 3 min,30 sec -- -20 rn 17 5.25 10.8 -- 25 hr, 45 min 83.69 81 'E Z 18 150 6.0 10.65 3 min,54 sec -- -20 Z ~~1 19 5.75 10.8 -- 17 hr, 30 min 67.5 83 Met specification at 3.75 amperes.~ 20 150 6.o 9.5 1 min, 17 sec -- -4o 21 3.25 10.8 -- 19 hr, 30 min 63.4 82 Not fully charged. (1 ~ 22 150 6.0 6.95 18 sec -65 23 3.0 10.8 -- 20 hr, 30 min 61.8 78 Not fully charged. -, 24 25 o10.8 -- hr, 20 min -- 80 25 Not recorded ~ ~ 26-1* 150 6.0 12.10 11.7 sec -- 80 26-2 150 6.0 11.60 5 min,43 sec -- 82 3 C 26-3 150 6.0 11.90 10 min -- 80 C 1r 26-4 150 6.0 11.50 4 min -- 80 Stood 21 days before dischargi.f 26-5 150 6.0 11.50 3 min,50 sec -- 80 o 0o 26-6 150 6.0 10.8 1 min,40 sec -- 80 Stood 21 days before dischargi C0 26-7 150 6.0 11.1 3 min -- 80 Regular ASTM cycle. *Number after dash indicates number of overcharge cycle. r' C^~~ i~-.- ~ ^

TABLE 17 CONSTRUCTION OF THE BATTERIES RETURNED FROM THE DETROIT TANK ARSENAL, CENTER LINE, MICHIGAN WHICH HAD BEEN TESTED IN YUMA, ARIZONA Date Number of Number of Number of Discharges Battery Manu- 0.060-inch 0.030-inch Separator Prior to Being Sent No. Type factured Positive Grids Negative Grids to Yuma 126 6TN 12-29-52 17 18 0.050-inch-thick pormax 126 6TN 12-29-2 17 i8 0^~7 feet on the grids 128 6TN 12-29-52 17 18 0.030-inch-thick pormax feet on the grids 157 6TN 3-23-55 17 18 0.030-inch-thick pormax wrapped around negative 0.030-inch-thick pormax Li^ 158 6TN 5-25-55 17 i8wrapped around negative 0 135 2HN 1-12-55 8 9 0.050-inch-thick pormax I"']~ 1^ 2HN 1-12.^ 8 9 ~~~~~feet on the grids 16 2HN 1-12-55 8 0.030-inch-thick pormax 136 2HN 1-12-539 FTIrn~~~~~~~~~~~ ~feet on the grids r Z 12 2HN 1-26-3 8 9 0.030-inch-thick pormax ~^^ ~ 4 2N 16-58feet on the grids 145 2HN 1-29-55 8 9 0.030-inch-thick pormax 6 feet on the grids I 14 1~7 2HN 1-26-553~ 8 9 ~ 0.030-inch-thick pormax 147 2HN 1-26-53 9 I^~~~~~~~~~~~~ ^ I ~~feet on the grids r- 1 148 2HN 1-26-53 8 9 0.030-inch-thick pormax 1k*'1 \ ~~~~~~~~~~~feet on the grids I I- 1489 2HN 1-26-53 8 9 0.030-inch-thick pormax feet on the grids 151 2HN 1-26-55 8 9 Q0.030-inch-thick pormax 6 _,"~ 15 \H 1165 0 9 feet on the grids r^11^ c, ~~p

CONFIDENTIAL 'ENGINEERING RESEARCH INSTITUTE * UNIVE returned to the laboratory at the University of Michigan. The record of their treatment at the Arsenal is not available. Each battery was given one or two cycles (cycle No. 6 or 6 and 7'of Table 22) before they were returned to the Detroit Tank Arsenal about June 8, 1953, to be sent to the Yuma Test Station. After the completion of their tests at Yuma Test Station about September 29, 1953, the batteries were returned to the laboratory of the University of Michigan for inspection and study. The results of the inspection of the batteries as-received are given in Table 18. At the Yuma Test Station, the vehicular operation characteristics and the standby characteristics were tested. A comparison of the specific gravity of the acid in each cell as-received with the specific gravity listed in the September 29, 1953, report from the Yuma Test Station is given in Table 19, while the specific gravity after recharging is given in Table 20. In most instances the batteries regained their original specific gravity, and the values are generally higher than those reported at the Yuma Test Station and recorded in Table 19. Further data on the condition of the batteries at the close of the Yuma tests are presented in Table 21, and the available cycling data in shown in Table 22. Discussion of Conclusions of Yuma Test Station Report of September 29, 1953 It should be recalled that the 6TN batteries contained 17 positive plates 0.060 inch thick and 18 negative plates 0.030 inch thick, assembled with 0.030-inch-thick pormax plastic separators, while the 2HN batteries contained 8 positive plates 0.060 inch thick and 9 negative plates 0.030 inch thick, assembled with 0.030-inch-thick pormax plastic separators. The elements fit rather tightly in the cell. This means that the active material was crowded between part of the ribs on the separators and there was not very much space for acid circulation. These plates had feet which rested on the bridge at the bottom of the cell. The clearance between the edges of two positive plates was of the order of 0.090 inch. This crowding of the elements left little opportunity for (1) circulation of acid (2) sloughing of positive material, and (3) the washing away of any positive material from the surface of the bridge which might cause shorting between the positive and negative plates. Comments on Conclusion A. It is stated in this section of the report that the water consumption of the experimental aluminum-grid batteries is considerably greater than that of standard lead-antimony batteries under high ambient temperatures as experienced in the desert. This conclusion seems quite valid for the batteries tested. lo08 CONFIDENTIAL

TABLE 18 INSPECTION DATA ON BATTERIES RETURNED FROM THE DETROIT TANK ARSENAL, CENTER LINE, MICHIGAN, WHICH HAD BEEN TESTED IN YUMA, ARIZONA Corroded Lugs Positive and Holes Battery Battery Accepted Retained Dead the De Chare Charge (Positive or Negative in the Voltage No. Type Charge Charge Negative) Oxide Intact Separators on Return 126* 6TN No No Both Yes Yes* 1 2.0 128 6TN Yes -- No Yes No 2 2.0 () 157* 6TN No No Both Yes Yes* 2 2.0 O 158** 6TN No No Both Yes No 2 2.0 Z 135 2HN Yes Yes No Yes No 0 2.0 Z In -n D ^ 136 2HN Yes Yes No Yes No 0 2.0 r 1142 2HN Yes Yes No Yes No 0 2.0 r 3 Yes Yes No Yes No uI 1147 2HN Yes Yes No Yes No 0 2.06 148 2HN Yes Yes No Yes No 0 2.0 149 2HN Yes Yes No Yes No 0 2.06 - 151 2HN Yes Yes No Yes No 0 2.0 Leaky case: the battery had apparently been dropped. ' j Complete battery was dismantled. M 01i I...........

CONFIDENTIAL ljLi irl ID TABLE 19 COMPARISON BETWEEN YUMA TEST STATION DATA AND UNIVERSITY OF MICHIGAN DATA ON THE SPECIFIC GRAVITY OF THE ACID IN EACH CELL OF EACH BATTERY Bat- Dead tey Data Cell Cell Cell Cell Cell Cell Aver- Cls No. 1 2 3 4 5 6 age Discharge 1261 Yuma 1.240 1.292 1.252 1.232 1.292 1.166 1.245 Nos. 5 (6TN) UM 1.275 1.150 1.135 1.190 1.150* Dry* 1.180 and 6 1282 Yuma 1.292 1.280 1.286 1.242 1.232 1.242 1.260 Nos. 3 (6TN) UM 1.235 1.175 1.160 1.185 1.235 1.185 1.195 and 4 1573 Yuma 1.280 1.252 1.262 1.242 1.262 1.262 1.260 Nos. 2 (6TN) UM 1.150 Dry* ** 1.150 ** 1.150 and 3 1584 Yuma 1.282 Dead Dead 1.236 1.266 1.246 1.205 Nos. 1 (6TN) UM 1.050 1.100 1.050 1.050 1.200 1.200 1.108 and 4 135 Yuma 1.266 1.232 1.292 1.312 1.343 1.272 1.285 All ood (2HN) UM 1.215 1.270 1.230 1.215 1.120 1.200 1.208 136 Yuma 1.252 1.246 1.292 1.312 1.286 1.272 1.280 (2HN) UM 1.210 1.225 1.240 1.225 1.175 1.185 1.210 All good 143 Yuma 1.252 1.276 1.312 1.322 1.312 1.282 1.295 (2HN) UM 1.235 1.260 1.250 1.245 1.240 1.220 1.242 All good 151 Yuma 1.272 1.272 1.292 1.292 1.272 1.272 1.280 (2HN) UM 1.235 1.235 1.260 1.260 1.235 1.235 1.243 Stored Batteries 142 Yuma 1.282 1.292 1.292 1.302 1.292 1.302 1.294 All oo (2HN) UM 1.260 1.260.26260 1.260 1.260 1.260 147 Yuma 1.262 1.282 1.312 1.302 1.292 1.302 1.292 (2HN) UM 1.235 1.235 1.235 1.235 1.250 1.235 1.238 go 148 Yuma 1.300 1.297 1.290 1.300 1.300 1.272 1.293 (2HN) UM 1.265 1.240 1.260 1.260 1.235 1.210 1.245 Allgood 149 Yuma 1.300 1.300 1.300 1.312 1.300 1.300 1.302 (2HN) UM 1.250 1.250 1.250 1.250 1.225 1.250 1.246 Case was broken. Too low to read. Notes: 1. Case had been dropped and broken, plates bent and shorted; acid leak. 2. Cells 3 and 4 shorted because they were too tight in the case. 3. Cells 2 and 3 blistered on top of positive lugs; too tight in the case. 4. Cells 1 and 4 shorted on sides of grids; separators too narrow; shorted on negative lugs; separators too short. 110 CONFIDENTIAL

TABLE 20 SPECIFIC GRAVITY AND CELL VOLTAGE DATA OF THE BATTERIES BY CELLS ON CHARGING, FOLLOWING THEIR INSPECTION AS RECORDED IN TABLES 18 AND 19 Specific Gravity of the Acid in Each Cell Recharge Voltage of Each Cell and the Corresponding Positive of Each Battery When Fully Charged and Negative Voltage against Cadmium as Observed Battery Cell Cell Cell Cell Cell Cell Aver- Voltage Cell Cell Cell Cell Cell Cell Terminal No. 1 2 3 4 5 6 age of: 1 2 3 4 5 6 Voltage Cell 2.64 2.58 2.58 2.60 10.40 126* - -- -- Broken Broken -- Pos.cad. 2.40 2.40 2.0 236 cae oken r 4 Neg.cad. -0.22 -0.22 -0.22 -0.22 cells Cell 2.68 2.58 2.58 2.62 2.66 2.62 128* 1.265 1.235 1.235 1.235 1.260 1.235 1.244 Pos.cad. 2.44 2.42 2.32 2.44 2.42 2.40 15.74 Neg.cad. -0.24 -0.20 -0.20 -0.22 -0.22 -0.20 Cell 2.60 2.62 2.64 2.64 10.50 157' 1.265 Broken Shorted 1.265 1.275 1.275 1.270 Pos.cad. 2.40 Broken Shorted 2.38 2.42 2.40 for 4 Neg.cad. -0.20 -0.24 -0.22 -0.24 cells Cell 2.08 2.64 2.66 2.14 2.66 2.68 158' 1.185 1.260 1.270 1.170 1.270 1.260 1.236 Pos.cad. 2.16 2.44 2.46 2.22 2.44 2.44 14.86 (') 158* 1.185 1.260 1.270 1.170 1.270 Neg1. cad. +0.10 -0.20 -0.20 +0.08 -0.20 -0.24 Cl) Cell 2.62 2.68 2.68 2.64 2.60 2.68 0) 135 1.285 1.300 1.300 1.285 1.285 1.275 1.288 Pos.cad. 2.40 2.50 2.50 2.50 2.40 2.40 15.90 Z Neg. cad. -0.20 -0.16 -0.18 -0.16 -0.18 -0.24 z Cell "1 136 1.285 1.285 1.300 1.300 1.265 1.265 1.283 Pos.cad. Neg.cad. rm Cell 2.46 2.48 2.46 2.46 2.46 2.40 143 1.275 1.300 1.300 1.300 1.275 1.260 1.285 Pos.cad. 2.34 2.34 2.34 2.34 2.34 2.34 14.72 Z 145 1275 1.0Neg. cad. -0.10 10.20 -0.10 -0.10 -0.10 -0.10 -.10 z -""" Cell 2.66 2.64 2.62 2.66 2.58 2.60 21 151 1.270 1.285 1.285 1.285 1.275 1.275 1.279 Pos.cad. 2.50 2.48 2.50 2.52 2.46 2.42 15.76 Neg.cad. -0.18 -0.16 -0.16 -0.16 -0.20 -0.18 Cell 2.70 2.64 2.70 2.70 2.70 2.70 142 1.235 1.285 1.285 1.285 1.285 1.285 1.276 Pos.cad. 2.50 2.50 2.50 2.50 2.50 2.50 16.14 '; j Neg.cad. -0.20 -0.18 -0.20 -0.20 -0.20 -0.20 Cell 2.62 2.60 2.64 2.64 2.72 2.64 r 147 1.300 1.300 1.285 1.300 1.285 1.300 1.295 Pos.cad. 2.44 2.42 2.46 2.46 2.48 2.48 15.86 ( Neg.cad. -0.20 -0.20 -0.24 -0.24 -0.24 -0.18 Cell 2.60 2.64 2.64 2.64 2.60 2.64 148 1.260 1.290 1.290 1.260 1.280 1.300 1.280 Pos.cad. 2.30 2.46 2.46 2.40 2.40 2.46 15.76 ( Neg. cad. -0.20 -0.20 -0.20 -0.20 -0.20 -0.14 ~C Cell 2.42 2.66 2.66 2.64 2.66 2.66 ': 149 1.280 1.280 1.285 1.280 1.280 1.280 1.281 Pos.cad. 2.46 2.46 2.46 2.46 2. 46 15.70 1 Neg.cad. +0.02 -0.20 -0.20 -0.18 -0.10 -0.20 *Type 6TN batteries; all others were type 2HN. '. r- It. _ _.

TABLE 22 LIFE HISTORY IN CYCLES INSOFAR AS IS KNOWN OF THE BATTERIES SENT TO THE YUMA TEST STATION Cycles Battery at Yuma Cycles at the No. Cycles at the University of Midhigan Test UniversityofMichigan Station Cycle History of Battery from Records of the Discharge to 10.8 volts at University of Michigan Laboratory the 2-ampere rate, follow| |,ing a conditioning charge Cyc CCycl e Cycl e ycle Cycle Cycle Cycl20) after their inNo. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 spection (Tables 18 and 19) 6* 180 mi 18 min 154 min Run not 226 hr 170 min 245 hr 180 min for good cells 128* 186 min 165 min 160 min Rnot 22.6 hr 175 min 25.3 hr 180 min for 4 goo cells 'TZ ^ 157* 195 min 195 min for 4 good cells s- > 158* 195 min | 180 min for 4 good cells 135 78 min 25.6 hr 77 min 70 min 21 hr 80 min | 75 min Z C 136 60 min 21 hr 55 min 22 hr 45 min 45 min 75 min | 80 min - - | 142 73 min 24.7 hr 77 min 80 min 26.0 hr 80 min 80 min 143 73 min 24.6 hr 77 min 80 min 26.2 hr 85 min m~ 75 min 147 74 min 25.0 hr 76 min 73 min 22.0 hr 80 min F 85 min H 148 74 min 24.3 hr 76 min 70 min 22.0 hr 80 min | 80 min c I 149 71 min 25.0 hr 80 min 70 min 26.0 hr 85 min 75 min 151 70 min 25.2 hr 80 min 80 min 26.0 hr 85 min 85 min. Type 6TN batteries; all others were Type 2HN. |

CONFIDENTIA iSUSSIlliD ENGINEERING RESEARCH INSTITUTE ~ UNIV 1" -IGAN The spacing between the grids in the lead-antimony batteries is greater than that between aluminum-grid batteries. This means not only that there is less electrolyte present, but also that there should be greater gassing of the electrolyte as its specific gravity is decreased through sulfation of the plates in the aluminum-grid battery. The boiling point of low-specific-gravity acid more nearly approaches that of water than does that of high-specific-gravity acid. Considering the high temperatures involved, then, it is not surprising that the water consumption was greater than in the standard batteries. The comments on Conclusion B are also pertinent to this problem, Comments on Conclusion B. In Conclusion B the experimental aluminum-grid batteries are said to be inferior to the standard lead batteries in maintaining charge during vehicular operation under desert conditions. It is not known what voltage was delivered by the generator on the experimental vehicles. If this voltage was in the neighborhood of 14 or 14.7 volts, the batteries could not be expected to maintain a specific gravity above 1225 to 1250. The aluminum-grid batteries require a minimum of 16.5 volts for complete charging. A voltage below this value on the vehicle would mean that these batteries were floating on the line at partial charge and never could attain full charge, nor could they utilize fully the charge delivered to them at the low voltage. At the low voltage and the low specific gravity of the acid resulting from the low voltage, the plates would sulfate. Once sulfation started the low voltage would not be sufficient, in connection with the current delivered, to break down this sulfation layer. Hence, the battery would not maintain its full charge during vehicle operation. If the voltage delivered by the generator of the vehicle was 17 volts, then the batteries should accept charge during vehicle operation. Comments on Conclusion C. It is not surprising that the 6TN batteries experience permanent loss of capacity as the result of normal vehicle operation. This would be a natural consequence of several factors such as (1) low generator voltage, (2) sulfation resulting from low generator voltage, and/or (3) possible loss of some electrolyte through excessive packing of the plates in the cell, which would tend to reduce circulation of the electrolyte and retard loss of heat. The difficulty with the packing of the cell is being eliminated by removing one positive and one negative plate. This will afford better circulation, greater utilization of the active material, and reduction of heat evolved during operation. It should at the same time reduce water__ CNINT114! CONFIDENTIAI

CONFIDENTIAL DE-CUhSSIFIED ENGINEERING RESEARCH INSTITUTE * UNI ICI O m consumption. The first two factors listed are at present, however, beyond control of the manufacturer. Comments on Conclusion D. The conclusion that the experimental batteries appeared to have a normal self-discharge rate as compared to the lead-antimony batteries is surprising. We have had 2HN batteries from the same group as those tested at Yuma on standby for 90 days with only a 12 percent loss in capacity. It would be interesting to know if the electrolyte had evaporated from these batteries and also if water was added just prior to the determination of their capacity after standby. It would also be interesting to know the condition of the separators below the vent, i.e., if they have been indented by the hydrometer and thermometer used to measure the gravity and temperature respectively. We have not in the past been placing a gauze under the vent opening, but are now doing so to prevent undue deformation of the separators and shorting of the plates through rough handling of hydrometers and thermometers. Comments on Conclusion E. We concur with conclusion E that the terminal posts lack the necessary construction to withstand rough handling. This is a matter that was beyond our control. We have known for a long time that these terminals should be larger in view of the softness of lead as compared to antimony-lead. The determination of what constitutes a dead cell seems to be part of the discrepancy between the Yuma Test Station data and our findings. If a cell will no longer give its rated capacity, it is looked upon with suspicion, and is expected to affect the terminal voltage of the battery profoundly. If a cell goes out before its fellows but the terminal voltage is above or equal to 1.8 volts at the end of the 20-hour rate or 1 volt at the 150-ampere rate for a 2HN-type battery, the cell is considered satisfactory. Generally a dead cell is one that cannot function even though it has sufficient acid to cover the plates. It is a weak cell if it gives only a fraction of its rated capacity. The batteries that are being constructed today are of a different design from those tested at Yuma. The new 6TN batteries contain 16 positive and 17 negative grids, while the new 2HN batteries contain 7 positive and 8 negative grids instead of 8 positive and 9 negative grids. We have found that this decrease in plate area is more than offset by the greater efficiency of active material in the remaining plates. We have also modified the construction of the newer batteries by using vinylite separators (not the exact type we desire, but the only CONFIDENTIAL

r ~ -~~~~~~~~~~ ENGINEERING RESEAC I STITUTE UNI kind y Or MICHIGAN kidwecan-obttain from the poueswihw timtes. rvn hota;O he bridge at the bottom of the ive t l -KIIes v^~ap Callnd nh+ -f~ara all I ^

-c% I: io _!N| I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ TI____ Cl D -: 0o) l - h