GMTC - Piecing Together GM Proving Ground HistoryA collection of information from 1966; the original date and author are unknown, but it is interesting reading. Select topics, scroll down or print the entire piece.
General Motors Proving Ground-the Milford, Michigan, installation that posted its 40th anniversary in 1964 -- embraces 4,011 acres and 75 miles of roads. On the site are 40 buildings with 689,264 square feet of garages, offices, display areas, test equipment, laboratories, shops and their maintenance and supporting activities. The entire facility during its first four decades has served as an industry prototype, the most comprehensive installation of its kind in the automotive world. Credited to its engineers are many techniques, testing devices, programs, instruments, designs and other innovations now standard throughout the industry. Proving Ground employee number 1,500 men and women. But when annual durability and audit programs begin, the personnel list may rise to as many as 1,650 in this virtually self-contained community. Half the regular employment roll is Proving Ground Staff; the other half is on permanent assignment from car, truck or accessory divisions. More than 65, 000 test miles are driven daily over the blacktop, gravel, dirt, Belgian block, spalled and specially smooth concrete surfaces for an annual total of at least 24 million miles. Altogether, since the spring of 1924, Proving Ground Staff and divisional drivers have amassed more than 335 million test miles. Moreover, this does not include additional millions of miles accumulated by divisional drivers on public roads and highways, both in the Milford area and elsewhere. Today the Milford facility is headquarters of an entire proving ground system. Sixteen miles southeast of Mesa, Arizona, General Motors maintains a 2,548 acre site with a seventeen and a half mile road layout. Originally designated a hot weather testing facility, the Desert Proving Ground now has year-round utility. During the past eight years its activities have multiplied many times over and its personnel now stands at 185, more than a 200 percent increase over the 1957 employment level of 52. Coincident with the 40th anniversary at Milford, the 28 year old Arizona establishment has recently undergone a complete modernization program. Twelve years ago a Pikes Peak Engineering Test Headquarters was set up at Manitou Springs, Colorado, for testing cars and trucks under extreme grade and altitude conditions on Pikes Peak Highway and mountain roads throughout the area. Thus, overall General Motors has 6,559 acres and approximately 92 miles of private roads in its proving grounds system. THE BEGINNING "How true it is that important developments so often flow from accidental circumstances. The idea of a proving ground came in 1923 when we were questioning the desirability of adopting four-wheel brakes on General Motors cars." A group of General Motors executives and engineers were attempting to test them on the public highways near Flint, Mich.-attempting to determine an exceedingly important technical policy under such adverse conditions that essential fact finding was almost futile. The urge to provide a better way of doing such things stood out crystal clear, and the General Motors Proving Ground was the final result." What had happened was that some GM test cars with four-wheel brakes had been demonstrated on a public gravel road. Modifications were suggested. Another test was scheduled as soon as the modifications could be made. On the day of the next demonstration engineers found that the local highway department had paved their test road, nullifying earlier test conditions. GENERAL TECHNICAL COMMITTEE SETS POLICYMeanwhile, in September 1923, the General Technical Committee was formed with Mr. Sloan as chairman. Membership included Charles F. Kettering, then head of General Motors Research Corporation (at Dayton, Ohio); Charles S. Mott, a director and corporation executive; chief engineers of car and truck divisions; Patent Section Director; Henry M. Crane, secretary, and William J. Davidson, assistant secretary. At that time most testing of GM and competitive vehicles was done around Dayton. Late in October 1923, Mr. Kettering and Mr. Davidson discussed some speed tests of a 1915 Cadillac at the old Chicago Speedway. Mr. Kettering suggested laying out a level, one mile concrete road next to the Dayton facilities. Mr. Sloan, Mr. Crane and other committeemen agreed. But they thought Dayton was too far from car and truck divisions for convenience. In November the committee suggested a level piece of land near Detroit where a one mile strip could be constructed. At one end of it would be a modest garage. Mr. Crane explained that a level road would offer only limited answers to engineering questions. He wanted a hill with at least a 10 percent gradient. Mr. Sloan delegated Mr. Davidson to find a parcel of land to meet the committee's requirements. It also was to provide room for expansion and it was to be situated within range of the car divisions-Chevrolet, Oakland, Buick, Cadillac, Oldsmobile-and GMC Truck & Coach. SITE IS SELECTEDLand south of Howell was examined but it lacked elevation. By the middle of November a second area was checked out near Milford in Livingston and Oakland counties, 40 miles northwest of downtown Detroit. Pere Marquette Railway engineers were contacted to help in assessing the topography. On January 17, 1924, the General Technical Committee, meeting at the General Motors Building, unanimously approved the land purchase after most members had reconnoitered the area personally. Wisdom of that choice is obvious today. The Proving Ground is in effect a hub with spokes radiating toward Detroit, Lansing, Flint, Pontiac, Willow Run, Warren and Saginaw-all now within easy driving distance. In the early spring of 1924 the committee met at the 1,143 acre site, walked over the proposed road system, authorized the appropriation for road construction, plus a 200 by 60 foot garage and a clubhouse. The latter item was for rooming and boarding visiting corporation and divisional engineers and executives. In the 1920s a trip from any of the General Motors plant communities to Milford was a long haul. Even today the clubhouse serves as an occasional temporary residence for visiting engineers. F. M. HOLDEN TAKES OVERF. M. Holden was the first director. He developed early engineering and testing techniques, insisting on absolute impartiality. An associate and long-time employee, R. L. McNeal, reported this about Mr. Holden's methods: "Valves were to be ground and carbon cleaned, and carburetor and ignition tuned up previous to every test. It was felt that the only reproducible and truly comparative condition was the best possible shape the Proving Ground could put (an engine) in. Thus all cars were on a strictly comparable basis, both GM and its competition. Right from the start it was decided that there should be no effort to favor any GM car, and this is one ideal the Proving Ground has religiously kept." Indeed, veteran Proving Ground engineers can recall occasional beefs by General Motors divisional engineers that competitive vehicles were favored in test routines. NEW DIRECTOR SPRUCES UP SITEIn May 1925 Mr. Holden went to Oakland as assistant chief engineer and O. T. Kreusser, head of Research Laboratories Technical Data Section at Dayton, took over as Proving Ground Director. He concentrated on new road construction and improving and expanding the physical plant. One of his first purchases was a truck load of shrubs and young trees which Proving Ground employees planted on a Saturday afternoon. CHEVROLET STARTS TESTINGChevrolet was first to use the Proving Ground in May 1924, even before buildings and roads were completed. It set up headquarters and used the surrounding public roads for tests. Buick and Oakland (which became Pontiac in 1926) soon followed, establishing stalls in the new garage. Within another few months crews from Oldsmobile, Cadillac, and GMC Truck & Coach were on the site. Mr. McNeal recalled a "Black Maria" carried Chevrolet drivers to and from Flint daily and they worked two 12 hour shifts. Moreover, they drew gasoline from a pump just inside the garage door and consumed approximately 60,000 gallons before an accounting system was set up to charge the division for fuel. Only Proving Ground vehicle in service the first summer was an aged Chevrolet truck that maintenance crewmen used. Mr. Holden finally obtained a copper cooled touring car for driving back and forth to Detroit. Late in 1924 more of these touring cars with air-cooled engines were turned over to the Proving Ground. Meanwhile, Chevrolet sent five competitive touring cars to the Proving Ground and they were used for practice with new test equipment. Not until late 1924 did regular test vehicles appear. The first 1925 engineering test car, according to Mr. McNeal, was a Jewett Brougham. Actually, the first test car ordered was a Hudson Super 6, but it was used by Buick and didn't arrive until after other test cars were on the site. EQUIPMENT SENT FROM DAYTON By the end of the first year seven miles of road were built, including the 3.8 mile speed loop, east west straight-away, 7.2 and 11.6 percent hills and durability hill route. The clubhouse and first garage were completed. Garage space soon had to be doubled and by 1925 totaled 27,600 square feet. By 1928 a third garage was added, raising the total to 90, 000 square feet. Construction items in the Proving Ground's early history indicated its importance and popularity. As far back as 1926 an auditorium and extra cafeteria were added to the building program to accommodate Sales Section meetings. Then, as now, the Proving Ground was basically an engineering operation, but the property also served as a locale for new model previews, technical society meetings, etc., although for security reasons such gatherings are limited. Additional land purchases likewise indicated steady, vigorous growth. To the original 1,148 acres, 120 were added in 1926, 550 in 1951, 1, 045 in 1953 and 1,148 in 195456 for the current total of 4, 011. In 1929 Mr. Kreusser left to organize the Museum of Science and Industry at Chicago and A. J. Schamehorn, the assistant director, became director. That same year Fisher Body Division organized a Proving Ground headquarters for extensive tests that led to the all metal car bodies of today. BELGIAN BLOCK ROAD IMPORTED FROM CHICAGOAlso in 1929 the Belgian block road was built, duplicating an old road between Brussels and Antwerp that took a brutal pounding in World War I. Smoothly worn granite blocks were imported from Chicago's Loop where they had been in use since the Chicago Fire (1871). In 1945, when the Proving Ground converted from military to civilian vehicle testing, the road was rebuilt with the blocks embedded in concrete so the punishing bumps, irregularities and contours would remain intact-another concession to consistency of test environment. Years of data have verified the Belgian block road is ten times tougher on vehicle bodies, frames and suspensions than any secondary road. Few surfaces like it exist in the United States. It gives a worse beating than it takes. By 1931 GM's private road system totaled approximately 23 miles. In 1934 the "skid pad" was constructed for measurements of car handling and ride characteristics and the first of the 25, 000 mile durability tests was organized, comparing GM and competitive cars. A fundamental annual testing routine, designed to duplicate as nearly as possible the customer's treatment of a car, the mileage limit was raised to 36, 000 in 1962. In 1936 the Acoustical Section from GM Research Laboratories moved to the Proving Ground and a year later the Noise and Vibration Laboratory was built, predecessor of the new laboratory building dedicated in 1961. The original section was headed by Ernest E. Wilson (who later became assistant director and director of the Proving Ground) and the working space built for it was advanced for its day. The laboratory building included several rooms, two of which were soundproofed. They had separate foundations, walls and ceilings with surrounding air spaces between them and the rest of the building. Also in 1937 the ride and handling test roads were added to the road system and the 27 per cent grade hill was completed. Meanwhile, near Phoenix, Arizona, a laboratory was set up for hot weather and desert testing-the beginning of the Desert Proving Ground. Forty-five and 60 per cent hills were built in 1939 and during that year and 1941 two chassis dynamometer buildings were completed. Mr. Wilson was named Proving Ground Director in 1941 when Mr. Schamehorn was transferred to the B-O-P Assembly Plant at Linden, New Jersey. MILITARY WORK BEGINSEven before Pearl Harbor, Proving Ground engineers were working on ordnance. At Washington the Office of Emergency Management was formed and out of this grew the Office of Scientific Research and Development and the National Defense Research Council. One of the first requests from the council was for a General Motors sound expert, and Mr. Sloan relayed it to the Proving Ground. The result was that instead of sending an expert to Washington, test problems were sent to the Proving Ground's Noise and Vibration Laboratory. Its staff was made available to the War Department. By June 1941 a contract was signed and soon after that military vehicles were a familiar sight on the Proving Ground. A January 1942 "letter of intent" from the War Department accepted General Motors' offer to use the Proving Ground for engineering tests of military vehicles. This meant a complete switch over to wartime operations. Without any new construction, space was allotted for vehicles and office workers as the payroll rose from 165 to 500. The cafeteria was moved back to the clubhouse and the former cafeteria space was converted into stalls for military vehicles. A medical service unit was set up in the remodeled crew's quarters. The main lobby and part of the auditorium were converted into offices. Vehicle tests were planned and authorized by the Office of the Chief of Ordnance at Detroit. Execution was left to the Proving Ground staff in collaboration with ordnance officers. Interested only in getting facts, the Proving Ground staff impartially tested virtually every make of military vehicle. Each vehicle that arrived on the site, regardless of its company or origin, was given a number and that was its only identification in all tests and records. Oldsmobile Division at Lansing had a cold test room which Proving Ground personnel took over to study effect of low temperatures on ordnance vehicles scheduled for Arctic duty. FIELD CONDITIONS WERE SIMULATEDA new and interesting section or tool project was set up in one of the garages. It evaluated standardization and usefulness of various tools for ordnance vehicles. No tools were used that were unavailable in standard or special sets of tools provided by Ordnance Department. No equipment for hoisting or lifting was used other than that provided for troops. The idea was to duplicate as nearly as possible actual field conditions. Gun pits were constructed for Pontiac and Oldsmobile Divisions which were making aircraft and antiaircraft Army and Navy guns. At the same time the Phoenix Laboratory was at Ordnance Department's disposal for testing under heat and dust conditions of Arizona. But the laboratory was only headquarters for test operations on surrounding public roads. A separate area for extensive field tests-five and a half sections of state-owned land-was leased about 22 miles from Phoenix. Level terrain for special air cleaner tests, meanwhile, was used on the Gila River Indian Reservation. Another project operated from Ordnance Department under Proving Ground supervision was the Engineering Standards Vehicle Laboratory, situated close to the Office of the Chief of Ordnance at Detroit. In it were vehicles used by OCOD engineers and draftsmen in making design changes. Practically every type of ordnance vehicle and vehicle equipment was available at the laboratory. Design changes could be made up and installed on these vehicles to check for clearances and fits. They could be made quickly without a trip to the Proving Ground and without tying up vehicles unnecessarily. In December 1942 the Proving Ground was presented the Army Navy Production Award for outstanding accomplishment-singular honor in as much as few non-manufacturing organizations received it. During the first two years of World War II the Proving Ground's effort was directed toward improvement and development of ordnance equipment. Emphasis changed to acceptance checks during the latter two years of the war. By early May 1945 ordnance work was greatly reduced. During the three and a half year ordnance contract period, the Proving Ground issued 1,467 formal reports and 2,277 memorandum reports on 1,555 projects received. In that period 681 vehicles were operated at Milford, 73 at Phoenix and 268 were displayed at the Engineering Standards Vehicle Laboratory. A total of 851,304 miles was posted at Milford and 121,566 at Phoenix. The divisions totaled 906,130 miles in ordnance vehicle development work on the Proving Ground, independent of Proving Ground operation. In January 1944 H. H. (Bill) Barnes was named director following the death of Mr. Wilson. Mr. Barnes had managed the Phoenix Laboratory and served as assistant director of the Proving Grounds. He retired in 1956 and was followed by Louis C. Lundstrom, the present Director of Proving Grounds. At the end of May 1945 the Ordnance contract was terminated. A large building and reconstruction program got under way. The east west and north south straight-aways were widened and re-paved and the large test track or speed loop was resurfaced so automotive operations could be resumed when 1946 models arrived in the spring. The gasoline station and a large garage building were begun in the spring of 1946. In the fall of 1949 another new building got under way, covering the areaway between the first and second buildings constructed at the Proving Ground, and an addition was built for Fisher Body Division. By late summer 1949 a maintenance building was started to house buildings and road maintenance, carpenter shop, paint shop and electrical groups. Meanwhile, on August 31, 1948, Proving Ground test miles passed the 100,000,000 mark. MILITARY TEST AREA RESERVEDIn February 1951 management set aside 800 acres adjoining the Proving Ground site to provide testing facilities for GM manufactured military combat and transport vehicles. In connection with this an operations building, specially designed to service and repair military vehicles of all kinds, was constructed. In June 1951, 2,280 acres of desert land near Mesa, Arizona, were purchased, 34 miles southeast of Phoenix. By the beginning of 1952 work started on a five mile circular track and a one and three-fourths mile level engineering test straight-away across the diameter of the track. The General Motors Desert Proving Ground was formally dedicated April 22, 1953, and in 1955 the DPG track was rebuilt and expanded. TEST TRACK GETS STEEP ANGLE PAVINGAs a part of the postwar building and construction boom at the Milford facility, the "super" or 3.8 mile test track was resurfaced in the spring and summer of 19520 This was the high speed track originally built in 1926. Its level and semi-level sections were re-paved in 1944. But the outside high banked turn lanes were untouched because no paving contractor had equipment or experience for such a super-elevated job. Proving Ground engineers not only modified paving machinery for the assignment, they also built new equipment. They worked out a technique that involved paving a 77 per cent slope (equivalent of a 37 degree angle) with an 11 ton finishing roller. First, they built and tested scale models of the equipment. Then a 200 foot long concrete replica of a section of a high-banked turn of the loop was built to prove whether the scale model theories would work. Further experiments were carried out with various equipment items before actual paving began. The project became an engineering landmark. The technique, or modifications of it, have been used on other test tracks with high-banked turns CHEVROLET GETS NEW GARAGE AS ROADS INCREASEIn April 1953 a new garage for Chevrolet was started. The next year the test track grade separation overpass, a traffic control circle and track access road were completed. To the road system was added a new 7.2 per cent engineering test hill loop and access road. At the base of the hill near the traffic circle three fuels laboratories storage buildings were constructed. The new administration building was completed in 1956; work on the new Noise and Vibration Laboratory began in 1960, and construction of the new 4.5 mile circular test track was begun in 1962. Its completion and dedication was a feature of the Milford Proving Ground's 40th anniversary. In four decades the Proving Ground has grown from 1,143 to 4,011 acres, from a garage and clubhouse to a 40 building complex and from a layout of about six miles of roads to the present 75 mile network. And this physical expansion is only a minor part of the growth story. An even more impressive record would include the Proving Ground's growth in personnel, in engineering technique and knowledge, in influence on product improvement. Now under way are a new 50,000 square foot building for Pontiac and Oldsmobile Divisions, a new corrosion test facility in quarters now occupied by Hydra-Matic, a two-story 40,000 square foot addition to the Chevrolet facility, and an addition to the Military and Heavy Vehicle Test Department building. The present Oldsmobile and Pontiac facility will be remodeled and taken over by Fisher Body, and Hydra-Matic Division headquarters will be moved to the space now occupied by Fisher Body. PHILOSOPHY"It is always sound philosophy to recognize that the most effective attack is the determination of facts without prejudice and with an open mind. "That was the fundamental concept of the General Motors Proving Ground. That, I believe, always will be its function-to test General Motors engineering under the most favorable scientific conditions and with the aid of the latest technical apparatus, first by identifying points of weakness as well as strength, and then by measuring them. Certainly in no better way probably in no other way, could we achieve the essential objective of giving constantly increasing value in General Motors Products." This 1944 comment by Alfred P. Sloan, Jr., then Chairman of the Board, epitomizes the engineering philosophy of General Motors' Milford, Mesa and Pikes Peak Proving Ground facilities. It prevails even in this era of sophisticated electronic instrumentation and computers. For the simple fact is that laboratory tests notwithstanding nothing produces such meaningful results as road testing. It involves the total vehicle, not separate components. And it reflects with almost brutal accuracy what happens to an automobile once it is in the customer's hands. Road testing has been a fetish since the earliest days of the Milford facility. Road testing also verifies and supplements the continuous laboratory experimental work in the corporate staffs, the car and accessory divisions. Long before a new device or component becomes part of a production vehicle it is exhaustively road tested as standard procedure, after it has been thoroughly evaluated in the laboratory. This notion of customer oriented tests figured in the Proving Ground's first driving schedules. In a 1926 booklet explaining its functions and facilities, this paragraph appeared: "From these tests the Proving Ground is gathering with impartiality and accuracy data on all cars to provide comparative information that will reflect the customer's point of view, making it available as permanent record for all General Motors divisions." PERFORMANCE TESTS ARE ORGANIZEDBy 1925 a clear-cut technique of posting car performance items was set up, such items as acceleration, hill climbing ability, deceleration, fuel economy, etc. In the Proving Ground's new office quarters a large chart room was established, an engineering sanctum in which General Motors Technical Committee met. Row upon row of hinged, graphic performance and specification charts lined the walls, an intimate biography of every General Motors and competitive vehicle to undergo the Proving Ground routine. Early instrumentation included a fifth-wheel speedometer, an accelerometer and a decelerometer for brake testing. Fuel economy tests were run with five gallon gasoline tanks and burettes. The equipment came from Dayton headquarters of General Motors Research Corporation, fore-runner of GM Research Laboratories. During the latter half of 1924 the testing routine was mostly dry runs to familiarize Proving Ground personnel with equipment. Robert L. McNeal, now a retiree, had much to do with establishing test procedures and maintaining records. "The early 1925 test cars arrived in December (1924) and January 1925," he recently recalled. "Considerable 1924 time was spent in working out a test procedure. As I recall it, a considerable number of tests were made at odometer 1, 000, especially hill climbing. "It was soon decided that the test program should be tests at 2,000 and 5, 000 miles, and every 5, 000 miles thereafter. Some of the early cars went 20, 000 miles. The 2,000 mile figure was a compromise between getting the data as fast as possible and getting the engine run in, so it would not stick up on maximum speed and acceleration tests. It was felt that at 5, 000 miles the engine would be thoroughly run in and that this test would represent the best that the car was capable of. "According to acceleration test data sheets, a large number of these tests were run in the latter part of March 1925. A large number of preliminary tests were run, setting up a definite procedure-especially the hill climb tests at 1, 000 miles which was a big jamboree, testing a large number of cars in succession. The Chandler was the only one that would climb the hill on this test..." At 2,000 miles, Mr. McNeal noted, several of the higher powered 1925 cars went over the top, but they took such a long time that they barely made it. The Chandler indicated "definite reserve power." CARS AUDITED ANNUALLYThis was the beginning of the annual engineering "audit" of General Motors and competitive cars, one of two major testing routines. The other is the 36, 000 mile durability run. The "audit" now includes some 80 individual tests and measurements on each car, following a 2, 000 mile break-in routine. As many as 150 vehicles may be included-if necessary, one of each model and series produced by General Motors and its competitors. It may cover every engine transmission combination or option a car manufacturer offers in a given model year. In essence, the "audit" is a quantitative, comprehensive record of performance and dimensional absolutes. It wraps up everything engineers need to know about automobiles, their own and the other fellow's. But the data are not collected for the esoteric pleasure of engineers alone. They relate directly to factors that influence the customer's choice-performance, ride, handling, control, etc. DURABILITY PUT TO TESTThe other major test routine is the annual 36,000 mile durability run. From 40 to 50 General Motors and competitive cars are put through it. Each in a single month totals as many miles as the average vehicle travels in a year-at least 12,000. This intensive round the clock fitness course is run as early as possible in the model year. It gives divisional engineers-from Chevrolet, Pontiac, Oldsmobile, Buick and Cadillac-a quick and thorough reading on how their cars will endure in service. The 36,000 miles are packed on at a more accelerated pace than any customer would set. Experience has shown that if a car survives the durability run intact, its service problems will be insignificant. It provides a correlation between testing technique and actual service wear and tear. It amounts to more than sheer punishment of running gear and engines or shakedown of chassis and body components. In addition to drives over every type of surface from erratic Belgian blocks to smooth concrete, the test includes all kinds of turning, parking, braking and reversing exercises. Horns, windshield washers and wipers, heaters, starting motors, radios and other accessories are switched on and off a prescribed number of times. When a driver takes over at the beginning of his shift-drivers are rotated to minimize driver effect-he not only checks tire pressure, fuel and oil levels, he also opens and closes doors, deck lid and hood latch, runs windows up and down. Any deviation from standard is noted. Operating costs of each test car are figured down to the penny-for repairs, maintenance, gasoline, oil and whatever replacement parts are needed. As if all this weren't enough, the 36,000 mile durability run extends through the miserable months, usually when Proving Ground pavements are well salted. Yet each car gets a salt spray along its flanks and underside. THE AUTOPSYAt the end of a durability test every car is disassembled. Parts are systematically laid out on long tables. Each part is checked and tagged with various colors denoting degree of wear. Failed parts are singled out on separate tables. As many as 6,000 General Motors and divisional engineers examine this huge display of impartial mechanical evidence together with figures covering replacement, repair, maintenance, fuel, oil and other operating cost items. What does such an expensive and exhaustive durability program achieve?
The idea of durability runs dates back to the Proving Ground's first decade. They were part of the first over-all testing program. By 1934 the first of the 25,000 mile runs were formalized and this mileage limit continued until 1962. The reason for raising the total to 36,000 that year was simply that the total wear at 25,000 miles wasn't significant enough to interest engineers. This is as good an index as any of the evolutionary progress of automobiles. OTHER SERVICE PROBLEMS STUDIEDCold weather operation with incomplete warm-up and short trip driving involves wear, cold starting and fuel economy problems, as contrasted with durability driving with the engine operating at nearly constant temperature. But short trip service is what many city and suburban car owners demand of their vehicles. So within the area surrounding laboratories, garages and shops, another Proving Ground car fleet moves at low speed, starting and stopping at each corner and intersection. This is short trip service with a vengeance. And when the guinea pig cars are not in motion, they stand out in the weather. Men who drive them need outdoor clothing. They're never in the cars long enough to take advantage of the heaters. As soon as they've finished one short trip, they step into another cold car. HOW THE MISSION IS CARRIED OUTIn carrying out the testing and engineering mission that Mr. Sloan outlined in his 1944 comments, the Proving Ground has three main activities:
SAFETYGeneral Motors Proving Ground is the safest place in the world to drive. This statement, often made by GM Automotive Safety Director Louis C. Lundstrom, is well supported by the record. Moreover, he and his staff originate much automotive safety data. They cooperate with government, industry and technical groups in studies and surveys, experiments and tests affecting public safety policies. As for the Proving Ground environment, the accident record on this 75 mile private road network near Milford is estimated to be at least 25 times better than on any public highway so far as injuries are concerned. This is no minor achievement. As many as 66,000 miles may be logged in a single day by test drivers. Although traffic isn't comparable to Detroit Lodge Freeway at 5 p.m., enough vehicles usually are traveling around the site to make drivers acutely aware of one another. Another built-in safety factor is a road system that is predominantly one-way-and, for the most part, the major test roads are limited access. DRIVERS ARE NONPROSOn the other hand, "accident" ratings under the Proving Ground system are far more stringent than on public highways. For instance, on the Proving Ground an "accident" occurs when a car leaves the pavement edge or when the slightest dent or scratch appears on a car body. Such low intensity mishaps on public roads or parking lots would go unreported. While Proving Ground management doesn't hire chronic traffic offenders or accident-prone drivers, the fact is that most drivers on the site are not trained professionals. Predominantly in the 21 to 40 year age bracket, they come from the surrounding area. Occupationally they include farmers, teachers, tugboat captains, cowboys, golf and baseball pros or any other individuals who turn to test driving as temporary or off-season work. Their amateurism is a virtue because it correlates with customer behavior in an automobile, not the techniques of a track or road virtuoso. Their training is informal. Their safety attitude is excellent. In 1960 a trophy was offered to the first driver group accumulating one million "scratchless" or accident-free miles. One group of 90 men promptly posted 1,750,000 before an "accident"-the strict Proving Ground definition-occurred. Another group of 30 covered more than 3,500,000 miles, best record to date. ROADSIDE HAZARDS REMOVEDMost recent safety program at the Proving Ground is roadside hazard removal. It makes off-the road crashes into trees, culverts, steep ditches, sign bases, overpass abutments, telephone or lamp poles a virtual impossibility. Mindful that the Proving Grounds' private roads were engineered by 1920 standards, the Proving Ground staff in the middle fifties began a redesign program. It was based on the plant safety engineer's philosophy: Design hazards and accidents out of plant equipment. Proving Ground engineers believed the same idea could be applied to an outdoor automotive testing laboratory. As a result, many potential injury accidents have been prevented. Even with experience and skill, veteran test drivers do run off the road once in awhile. They may skid on icy pavement. They may doze. Or they may swerve onto the roadside to miss a stalled vehicle or other obstruction. At the Proving Ground they can leave the road in complete safety. Not only have trees and other obstacles been cleared 100 feet either side of the roadbed, the roadsides themselves have been recontoured and sloped. It is impossible to crash into a culvert or steeply graded ditch within that spacious safety corridor. STANDARDS ARE ESTABLISHEDA set of design standards has evolved from this hazard removal program, It's available to any highway or safety organization and several are utilizing it in present and future road building programs. The work has been twice recognized nationally-in 1961 by the Highway Research Board of the National Academy of Sciences, National Research Council, and in the Metropolitan Life Awards for Research in Accident Prevention at the 1963 National Safety Congress. In both instances, Kenneth A. Stonex, GM executive engineer, automotive safety engineering, accepted in behalf of the Proving Ground staff. The studies presented by Mr. Stonex indicated from both Proving Ground and other accident data that at least 80 per cent of single car off highway accidents could be prevented with only a 30 foot clear corridor either side of a public highway. In 1964 such off highway crashes killed 13,700 persons. GUARDRAILS ARE STUDIEDAn urgent need in the roadside hazard removal program was reliable information about guardrails. No comprehensive test program ever had been reported. On the Proving Ground were 14 miles of guardrail, convex steel beam type mounted on wooden posts. Some of it had deteriorated, but the question was whether it was effective. If not, how should new guardrails be designed? Preliminary tests showed old structures were inadequate for cars hitting them at speeds as low as 35 miles an hour. Remotely controlled cars were steered into various redesigned structures from zero to 33 degree angles at speeds of 30 to 68 miles an hour. Laboratory tests were devised to supplement the full scale crash tests. Later, several tests were run with drivers in control of crash cars. Out of this series of more than 60 crashes, including one with a transit bus and several with trucks, came a new set of recommendations for guardrail structures. The project involved frequent conferences with guardrail manufacturers and highway officials. Although data for redesigned Proving Ground guardrail systems are not intended for application everywhere-an important variable, for instance, is soil condition-they have been offered to public highway engineers for whatever use they can make of them. At best guardrails are poor substitutes for a well cleared and contoured roadside. But they are necessary at bridge approaches, sharp cuts, mountain cuts, overpass piers and roadside signs. They can be designed so that when hit they won't pocket and trap a car with violent deceleration or bounce a car back onto the road. They can be ramped at the end so that a vehicle hitting a guardrail end on merely straddles it. On public highways guardrail end sections often are not firmly anchored, and when a car hits them end on it may be "speared" end to end. OTHER SAFETY BYPRODUCTS STUDIEDOther interesting byproducts have evolved from roadside hazard removal. One is a lightweight lamp or utility pole with a tripod base. When hit by a car it shears at the ground level. Damage to the car is only slight and both driver and passengers are likely to be unharmed. Ordinary highway marker signs are hazardous when mounted at the 42 inch height If a car crashes into them, the sign usually shears from its stem and falls into the windshield. With a higher mounting the sign shears off and the car passes beneath it, with little hazard to its occupants. The latest manual on Uniform Traffic Control Devices recognizes this and now requires mounting heights of five feet on rural highways and six feet on freeways. STRUCTURAL BREAKDOWNS VIRTUALLY ELIMINATEDProving Ground records verify reliability of modern automobiles. Time was when parts failures or mechanical breakdowns caused frequent accidents on the site. But they are no longer a serious factor thanks to improved design, testing, metallurgical, manufacturing and engineering practice, plus good maintenance. Even tire failures are not menacing because drivers know how to react when they occur. Only seven accidents (by Proving Ground standards) have resulted from blowouts in the last ten years. Yet driver caused accidents must still be reckoned with, although GM safety policies have reduced them effectively and steadily. Drivers receive only informal instruction and from three to five days on the job training, but they must obey rules and regulations. They operate their vehicles in safe, congenial surroundings as compared with many drivers on public highways. But not everything is in their favor. Test drivers must contend with monotony in an eight hour shift. At the same time, they must maintain definite mileage schedules. And they may encounter foul weather, foul enough to keep most people home. Biggest Proving Ground safety problem is drowsiness at the wheel mostly during hours near daybreak. To counteract this tendency, night, shift men must sign in every hour. In fact, any driver who feels sleepy is urged to return to the garage. Various experimental keep awake gimmicks and alarm gadgets have been tested, but none has yet solved the drowsiness problem. So even in an "Utopian" driving environment, human lapses occur. In the past ten years Proving Ground pilots have strayed unwittingly or inadvertently from the road at least 300 times. In terms of vehicle mile totals, this rate undoubtedly is very low. On public highways the incidence is much higher and certainly more significant. It accounts for as much as one-third of the nation's total annual traffic deaths. CRASH TESTING DATES BACK 30 YEARSFor more than 30 years General Motors engineers have run crash tests of one kind or another at the Milford Proving Ground. In some years as many as 40 vehicles are destroyed or damaged in rollovers, head-on, barrier impact or car to car crashes. In fact, even a 41 passenger inter-city bus has been crashed into a barrier by GMC Truck & Coach Division-the only known test of its kind. In this overall crash research the Proving Ground also has developed nondestructive and laboratory techniques, much of them original in the automotive industry and some of them now used by other testing organizations. Such techniques include every component that figures in driver and passenger safety. Rollovers of experimental and production vehicles began in the early 1930s, at the time the all steel turret top was introduced. The object was to get information under controlled conditions that could be correlated with actual highway accident damage. With a spiral ramp at the top of a hill it was possible to start an automobile rolling over and over down the hillside. Another technique was simply to drive a car into a skid on a level field. With its comparatively high center of gravity, the car could be overturned easily. Over turning today's models with their low center of gravity must be more adroitly stage-managed. A modern car in a skid on level ground won't overturn unless "tripped" by some obstruction. As early as 1934 General Motors ran barrier impact tests. They were both simple and straightforward; the driver aimed the test vehicle at the barrier at low speed so he could jump out of it just before the crash. Today such tests are run at any desired speed with the test cars remotely controlled. CAMERAS WITNESS CRASHESMoreover, modern crash tests are witnessed in minute detail by high speed cameras and are precisely detailed by decelerometer readings. In such instrumented crash vehicles are fully instrumented life-size dummies heading toward synthetic doom. From countless of these staged catastrophes General Motors engineers have assembled an imposing backlog of simulated physiological and engineering data. It figures in the design of car roof and body pillars, door locks and interior furnishings. Also, it has revealed valuable information about energy absorbing capability of a car's sheet metal in certain types of crashes, particularly head-on into fixed objects. CARS ARE SNUBBED TO STOPSObviously, crash testing is expensive. Thus, in recent years engineers at the Proving Ground have installed various simulation devices. One of these is a heavy-duty hydraulic snubber first used in 1955. It was good for relatively precise control of a full-size car in 3 to 35G stops. A cable tied to the test car "snubbed" it to a stop. The interior test results were similar but less destructive to the car's exterior than a crash into a barrier. The snubber test was good for evaluating seat belts, seat belt anchorage, seat mountings, steering wheel structures, crash padding, etc. It had an added advantage over outright crash testing; tests could be better controlled and they were repeatable with the same vehicle. Its only disadvantage was that outdoor testing depended on weather and good lighting which was necessary for high-speed photography. IMPACT SLED DESIGNEDTo bypass weather and other outdoor testing variables, Proving Ground Experimental Engineering Department in 1963 activated an Impact Sled. It simulates almost all types of auto crashes in the laboratory-head-on, side or angle impacts. The device gives engineers greater control over repetitive tests than was possible with previous apparatus, enabling them to build up a large volume of test data in a short time. It subjects a test vehicle or component to an acceleration pulse instead of a deceleration shock, and the pulse can be varied over a wide range. Recording oscillographs monitor and record the various stresses, accelerations and contacts that occur to any structure undergoing test. Two banks of floodlights in the main test area produce more than 10,000 foot candles of illumination so that "crash" sequences can be filmed in color at rates as high as 3,000 frames per second. Basically, the Impact Sled is a platform mounted on 100 feet of parallel rails. A compressed air mechanism "fires" it with a thrust as high as 300,000 pounds. Test specimens-from complete automobile bodies with life-size dummies to such items as seat belts, steering assemblies, door latch mechanisms, etc.-are bolted to the sled. Slippers on the bottom of the platform guide and support it on the rails. Once the structure reaches its highest acceleration, brakes gradually decelerate it within the length of the 100 foot track. Since being checked out the Impact Sled has been operating almost full-time evaluating crash padding, seat mounts, seat belts and anchorage, windshield glass breakage, etc. Chances are it may become a standardized testing device for testing all types of automotive body components. SKID TRAILER MEASURES SURFACE FRICTIONIn the last seven or eight years the Michigan and other state highway departments have been measuring "coefficient of friction" of their road surfaces, using a so-called "skid trailer" towed behind a car or truck in the regular traffic stream. Prototype of this device was designed and developed by the Proving Ground Experimental Engineering Department. It was assembled from salvage materials so any highway department could duplicate and operate it at low cost. The trailer idea was proposed when Proving Ground engineers were planning new roads in the early 1950s. Before embarking on the road building program they wanted to check comparative road surface coefficients, various types of concrete and asphalt materials. They developed a somewhat crude technique with a large tank truck and a standard automobile. The auto braked only with its rear wheels whenever the tank truck doused a section of the road with water. Strain gauges measured forces developed between the wetted road surface and skidding rear tires. The technique indicated comparative slipperiness of various road surfaces, particularly when wet. It was the coefficient of friction between tire and road. They found that friction coefficients varied not only with the type of road materials but also from constant polishing action of auto and truck tires on heavily traveled surfaces. Obviously, condition of the road surface was an important safety factor. In their early experiments, for instance, Proving Ground engineers tested their friction coefficient measuring technique on a nearby county road which they knew had a high accident rate. They found the road surface when wet was about as slippery as ice. Another reason for knowing as much as possible about the interaction of tire and road surface was that the comparative roughness or smoothness of pavement affected brake tests. For some time engineers had noticed certain obvious brake improvements weren't reflected in test results. The reason was that the road surface had changed; it was smoother. Thus, any braking improvements were "washed out" by a lower coefficient of friction between tire and pavement. From this experimenting came the more compact "skid trailer," the present model now used by several highway departments as well as the Proving Ground. It looks like an ordinary two-wheel trailer that any vacationer or camper would pull behind his car. Water is piped to a pair of spouts in front of the two trailer wheels. As the unit is towed along the highway, the operator in the towing truck merely flicks a switch. Water squirts under the trailer wheels and the brakes lock. Coefficient of friction is instantly recorded on a dial in front of the operator. Thus highway inspectors are able to spot stretches of pavement where the coefficient of friction is low. Here, they know, dangerous skids might occur in wet weather.
|
Origin
of General Motors Proving Ground was modest. Twenty-two years
ago, when the facility observed its 20th anniversary and 89,000,000
miles of testing,
In September 1924 two loads of test equipment came from General
Motors Research Corporation at Dayton. Developed by GM Research,
they included a fifth-wheel speedometer, an accelerometer, a
decelerometer for brake testing and a fuel economy measuring
device.