Investigations of great importance are carried out at the National Physical Laboratory at Teddington, Middlesex. Bridge builders, road engineers, aircraft designers and architects are all assisted by data patiently collected by the research worker
A SIDE VIEW of the impact testing machine which is illustrated below. In the centre is the weight which is raised through an arc of a circle and released, in the manner of a pile driver. It strikes a movable arm, to which is attached the part under test, the other end of this part being fixed to the anvil of the machine.
SPECTACULAR feats of engineering are always before the notice of the public. The building of a great bridge, the breaking of a speed record, the construction of an immense skyscraper — all are inevitably surrounded by an aura of publicity. In the background, however, far from the scenes of the engineer’s triumphs over difficulties, is a band of research workers whose quiet, methodical investigations are of as much importance to the engineer as his practical work is to the citizen.
Research work, by its very nature, is unspectacular, sometimes slow, but always intensely interesting. It cannot be hurried; the research laboratory is not invaded by the modern atmosphere of rush and bustle. One single experiment may last for years.
In the Engineering Department of the National Physical Laboratory at Teddington, Middlesex, much research work of immense importance is carried out. Special investigations of extremely varied character are made for firms and other bodies, and the Laboratory is always ready to give consideration to any problem or difficulty submitted to it, when it is thought that the facilities at its disposal may lead to a solution.
The Engineering Department is one of the eight main scientific departments of the N.P.L. The entire Laboratory, founded in 1900, now comprises fifteen large buildings and a number of smaller buildings, occupies some forty-eight acres of land and has a staff of about 700. Each department is divided up into several smaller sections and each section has a number of definite investigations in hand.
Some of the more important classes of work carried out by the Engineering Department during 1936 and 1937 may be described under these main headings: Effect of Wind Pressure on Buildings and Bridges; Effect of Wheel Impact on Roads and Vehicles; Properties of Materials at High Temperatures; and Investigation of Fatigue in Metals. During 1936 some thirty-seven important research items were in active progress, but those mentioned are a representative selection of the more pressing investigations. Researches relating to the pressure of wind on structures have been continually in progress since the formation of the Engineering Department. They have been mostly concerned with the data needed by engineers designing structures likely to be exposed to high winds, whether the structures were bridges, steel-framed buildings or aircraft hangars.
THE GREAT IMPACT TESTING MACHINE in the Engineering Department of the National Physical Laboratory. In the foreground is a chain which is to be tested. The weight or hammer (see illustration above) is in the background, partly raised. It falls with a force of twenty foot-tons on the semicircular arm to which the chain is attached, subjecting the chain to a sudden tensional stress.
The work is carried out for the Building Research Board of the Department of Scientific and Industrial Research, and it includes measurements on small models in a wind tunnel, as well as on full - sized structures in the open. One of the most important branches of this work is investigation of the effects of wind pressure on relatively light buildings, since buildings of heavy and solid construction are less susceptible to the effects produced by winds.
Experiments were carried out for some time on the Severn Bridge at Sharpness, Glos., the main object of the work being to determine the width of sudden gusts of wind. Apparatus was installed at five points on the bridge, the two extreme points being half a mile apart. At each of the five points an automatic record of wind pressure was made, and the average of the five readings was regularly compared with the readings given by one individual recorder — that in the centre.
The object of this experiment was to investigate the degree of variation in pressure intensity during a gale. The widest front over which tests had previously been made was 600 feet. The Severn Bridge tests, extending over some years, gave a valuable series of figures for a wind front of half a mile.
Such data as is made available by these tests are of immense importance to the bridge builder, who must plan his structure in accordance with the stresses which it will be called upon to bear. Hitherto it had not been definitely known whether a gale produced unusual strains over a wide front or merely over relatively small areas.
Another interesting series of tests concerns the effect of adjacent buildings on a particular building under consideration. A tall building standing on an exposed site is normally exposed to wind pressure for its entire height. If, however, a smaller building is placed between the tall building and the direction of the wind, it may so deflect the wind that the upper part of the tall building has to stand a far greater strain than before.
Wind Tunnel Tests
Investigations of this character are carried out in a wind tunnel, using models built to a scale of 1 inch to 20 feet. The models can be placed in various positions with relation to the direction of the wind, and models with stepped roofs or with roof slopes of 30° and 45° have been tested in this way.
The results already obtained have made it possible to deduce general coefficients for the design of buildings in open or built-up areas, and attention is now being given to a large composite model representing a portion of London covering about half a mile of frontage on the Thames Embankment. Tests will make it possible to determine the distribution of wind pressure, in various circumstances, on a building situated in the middle of this area.
The method of measuring the pressure on different parts of these models is interesting. A scale model of an aircraft hangar, for instance, has a number of small perforations at different points on the roof. Nipples on the underside of the roof are coupled, by a group of rubber tubes, with a series of pressure gauges. The variation of pressure on each point is watched as artificial gales of varying intensity are produced in the wind tunnel.
The maximum pressure differences which are likely to occur over a portion of a roof or wall may thus be determined. Further, the pressures set up on the full-sized building exposed to the wind may be compared with those observed under test in the wind tunnel.
FOR THE INVESTIGATION OF FATIGUE OF METALS — one of the many types of machine used for this work. This particular type of apparatus uses alternating current to cause rapid vibration of the metal part under test. Failure of metal parts under repetitions of load, commonly known as “fatigue”, is responsible for nearly all breakages of moving parts of machines or structures subject to repeated and vibrating loads.
Wind-tunnel tests are indispensable, since gusts of wind steady enough to give reliable results do not often occur naturally. This is a good example of the patience of the research engineer — that he may have to wait for several months before suitable weather conditions make it possible for him to take observations on a building which has been fitted with recording apparatus.
Wheel Impact Research is a subject which bears no resemblance at all to the foregoing, but probably is equally important. For many years it has been possible to install in a private car a small piece of apparatus which will make a permanent record of the movement of the car over a given road. The “signatures” of various London roads recorded in this way were given certain publicity during 1936.
The N.P.L.’s work is to tackle the problem from the opposite angle. Instead of assessing the performance of a given vehicle over various roads, the engineers record the performance of a vehicle, in different loading and tyre inflation conditions, over a known road.
Road Tests at 45 m.p.h.
The first measurements of wheel impact forces made in Great Britain were carried out by the N.P.L. in 1931. A trailed vehicle was then used, but in 1933 experiments were started with self-propelled vehicles. Four vehicles are now used for the purpose — a heavy six-wheeler, a heavy four-wheeler, a medium four-wheeler and a private car.
The object of the experiments is to determine, in varying conditions, the impact force imposed upon the road, and the equal and opposite force on the car. The load on one wheel of an axle — either on the near or off side — is determined by four electrically recording gauges connected to the axle. Two of these gauges measure the load imposed by the weight of the chassis and body, and the other two, known as accelerometers, measure the vertical acceleration of the wheel as the vehicle passes over an irregularity in the road.
Thanks to the electrical method of recording, it has been possible to interconnect the four instruments so that the total vertical wheel load can be continuously recorded on a strip of cinematograph film. Simultaneous records of time and distance are imposed on the film. Thus, the impact load at the speed at which the vehicle was travelling may be seen by reference to the film and to the known positions of the irregularities in the road.
Most of these tests have been carried out on a private road on which speeds up to forty-five miles an hour could be attained in safety. A standard irregularity is used, and it has been designed from experience of the conditions of roads in a bad state of repair. Experiments have been made to determine the effect of such factors as speed, unsprung weight, type and size of tyre, inflation pressure, adjustment of shock absorbers and so on. The results obtained have proved extremely interesting. One of the most striking is that the maximum load on the road may be imposed, in certain circumstances, when the vehicle is unladen. The effects of variation in speed have proved somewhat complex with the six-wheeler because of interaction between the two axles of the bogie.
The two small accelerometers designed, for use with a private car have been successfully used to measure wheel impact on the driving axle of a railway locomotive. Tests were made with the locomotive in normal service, and leads were taken from the accelerometers on the axle-boxes to recording apparatus housed in a guard’s van behind the tender. The tests extended for more than four days and included 1,000 miles’ running.
A considerable amount of research has also been carried out on the endurance and wear of concrete roads and the skidding of vehicles on different types of road surface.
The section dealing with Properties of Materials at High Temperatures is of more importance than its title would suggest to anyone but an engineer. It has a vital bearing on every branch of steam engineering and the generation of electric power is widely dependent upon steam engineering.
A ROTATING-BEAM FATIGUE-TESTING MACHINE. The metal rod under test is rotated at high speed by an electric motor and is at the same time subjected to bending stresses by the two weights on the right of the illustration. Fatigue-testing machines may be arranged to impose complicated combinations of tension and shearing stresses, simulating the complex vibrations to which the rapidly moving parts of modern machinery are often subjected.
The use of superheated steam and of constantly increasing pressures in boilers and turbines is the chief reason for the importance of this branch of research. The “creep” or flow of metals which occurs at high temperatures is studied in closely controlled conditions, and the creep-resisting properties of different kinds of steel are compared.
Steel bars placed in an electric furnace are subjected to a constant tensile stress for long periods and measurements are taken at regular intervals. The apparatus developed by and used at the N.P.L. will detect a change in length of one-millionth of an inch on a steel bar some five inches long.
The loading equipment has a capacity of five tons, and the electric furnaces are designed for a maximum temperature of 800° centigrade. They are controlled by thermostats which keep the temperature constant within extremely fine limits. The changes in length of the test pieces of steel are measured by instruments known as extensometers, and readings are taken by observing the reflection of an illuminated scale in a telescope.
Another branch of research with a bearing on steam engineering concerns the design of pipe flanges and bolted connexions. This has been initiated at the request of the turbine and associated industries, because of the rapid trend towards the use of even higher steam pressures and temperatures than those which have become general during the past few years.
Complete assemblies of stud bolts, nuts and washers have been subjected to prolonged loading at a temperature of 975° Fahrenheit, to determine the contribution which each component makes to the total deformation due to creep. The influences of such factors as the size and material of the nut, the diameter of the bolt shank, the length of the threaded portion of the bolt,
have been examined in detail. The elastic flexibilities of different types of flanges for 8-in. pipes, one being screwed and welded, the other welded only, have been determined. Tests were also made on a complete steam pipe assembly in high temperature and pressure conditions. At a pressure of 1,400 lb. per sq. in. and a temperature of 950° Fahrenheit, the joint failed after twenty-three days. The total dimensional changes occurring during the test were recorded.
At the N.P.L. a wide variety of machines is used for producing almost any conceivable kind of stress in a metal part. One machine — the largest of its type in the world — imposes a single impact upon the part under test. A huge hammer, moving round the circumference of a circle, may be raised to a predetermined height and allowed to fall on a movable arm, to which is attached the part under test, the far end of the part being fixed. Railway couplings and chains of all kinds may be subjected to the influence of a single impact of this kind. Here the visitor to the N.P.L. will see chains stretched far beyond their normal length, with the links solidly jammed into one another so that the chain resembles a solid steel bar and may be picked up by one end, remaining perfectly rigid.
Breakages of Moving Parts
The Investigation of Fatigue of Metals has, perhaps, the widest application of all the research groups mentioned. Failure of metal parts under repetitions of load, commonly known as “fatigue”, is responsible for nearly all breakages of moving parts of machines or of structures and components subject to repeated and vibrating loads. The various branches of this important subject have been studied at the National Physical Laboratory for many years and numerous ingenious machines have been made for laboratory fatigue tests.
Some of these machines apply simple tension or bending stresses to metal test pieces. Other machines impose complicated combinations of tension and shearing stress, simulating the complex vibration to which the rapidly moving parts of modern machinery are often subjected. The work has gradually developed into a comprehensive study of fatigue phenomena in its relation to the molecular structure of metals.
MODEL OF AN AEROPLANE HANGAR in the small wind tunnel used for investigating the effect of wind pressure on buildings. Small perforations on different parts of the model are coupled by nipples and tubes to the pressure gauge illustrated below, and the distribution of pressure over different parts of the model building is studied as artificial gales of varying intensity are produced in the wind tunnel.
Subsections into which it is divided vary from data for the design of high-duty crankshafts for aircraft engines to research into the fundamental aspects of the fracture and deformation of metals. The latter branch of the work is particularly connected with an investigation of the crystalline structure of metals. To determine the effect of the orientation of the crystal axes on the fatigue resistance of the metal, single crystals are tested under alternating stresses. To study the changes in crystalline structure caused by fatigue stresses, the specimens are by an X-ray process.
Crystalline Fractures
It has been established researches that whatever the the stress may be the progress of deformation of the crystalline structure is identical. Failure is caused by progressive breakdown of the “grain” to a certain limiting size. Specimens of mild steel are subjected to five stress systems — tension, torsion, alternating tension and compression, repeated tension and alternating torsion.
A PRESSURE GAUGE IN WORKING CONDITIONS, showing the wind pressure on different parts of the model illustrated above. The great variation in pressures is clearly visible. The glass tubes contain coloured alcohol, and each of them is connected by rubber tubing to one particular perforation on the model in the tunnel. Wind tunnel tests are of great value, since gusts of wind steady enough to give reliable results do not often occur naturally, and if they do, the engineers may not then be conveniently placed for making observations.
New apparatus has been devised whereby it is possible to study, during the progressive stages of a test, the same identical crystals. The crystalline structure of steel is well known. Examination of a broken back axle or of any fractured steel bar is sufficient proof. Specimens that have broken under fatigue invariably show distinct signs of a crack. Part of the fracture exhibits a clean crystalline surface, but the part where the crack originated will generally appear dull, since the two surfaces have been rubbing together under the influence of varying stresses.
The repeating stresses are produced by ingenious apparatus in which alternating current is used to produce rapid vibration of the part under test. Another machine rotates the metal rod under test, in a horizontal plane, but with a downward load on one end. The bar is thus twisted downwards, although it is rotating at a high speed. This produces a rapidly alternating load on the surface of the bar. Practical applications of this work are innumerable.
One interesting branch of it concerns the causes of failure of lifting gear such as the hooks, chains and rings used for the suspension of elevators and cranes. One particular series of experiments has been carried out to ascertain the effect of annealing a chain at intervals. Chains tested at just above the normal safe limit, with annealing at intervals, however, showed no increase in endurance over unannealed chains. Fatigue tests on wrought-iron rings of circular section and machined mild-steel rings of square section have also been carried out.
The entire arrangement of the N.P.L. Engineering Department is flexible. It must be so, from the nature of its work. Small experiments with individual pieces of apparatus must be carried out from time to time, but in the background are investigations —such as that into the behaviour of metals at high temperatures — which continue for years.
TYPICAL MODELS OF BUILDINGS, with a scale of one inch to twenty feet. That on the left represents an aeroplane hangar; in the centre is a representation of a small factory building, and on the right is a warehouse, with the dimensions of the building which it represents marked on it. A group of connexions, to which the rubber tubes from the pressure gauge are attached, may be seen projecting below the model on the right.
Tests are undertaken, for instance, to determine the efficiency of engines and gears, of agricultural implements and tractors, of steam pipe coverings and pressure gauges, of pumping apparatus on fire engines and on all kinds of machinery. Lengthy investigations have also been carried out into the mechanical properties of steel springs. Lubricants, bearings and bearing metals are tested in running conditions, and special tests of various kinds have been carried out on such widely varying objects as gas cylinders, pipe couplings, tiles, bricks, paper, safety glass and electrical insulating materials.
Investigation of service failures of all sorts has been undertaken on behalf of industry. Failures dealt with during 1936, for instance, included alternator shafts, lifting hooks, turbine blading, connecting-rod bolts, a rudder stock, a bicycle steering column, boiler tubes, axles and front axle beams of motor cars.
All work of this kind is undertaken on a strictly confidential basis, and no reference may be made to the results obtained on individual tests.
Testing New Aircraft
The results of general work are published from time to time in papers presented by members of the N.P.L. staff to technical institutions and scientific societies, and in Annual Reports. Engineering is only one field covered by the research work of the National Physical Laboratory. Other fields include all branches of physics, electricity and magnetism, radio communication, metallurgy and ship design. The routine testing of instruments and materials involves much general research work, of which examples are found in all departments. As an instance, the investigation of unnecessary noise of all kinds may be quoted.
Investigations carried out for the Air Ministry have included such fundamental matters as the effect of wing thickness and surface finish on drag and on maximum lift. During 1936 fourteen designs of new aircraft were studied in the wind tunnels, often in detail. The almost universal adoption of the monoplane has necessitated research on the stability of flight, most of the older stability data having been obtained on biplanes. Extensive work is also carried out on radio problems which arise in civil aviation.
Tests of all classes of measuring instruments are undertaken, and special problems are investigated on behalf of Government Departments. Payment is received for the work done for outside bodies, and the charges for special investigations are based on the time occupied in carrying out the work. Other phases of the activities of the N.P.L. will be described in later chapters.
ELECTRIC FURNACES for determining the creep of metals at high temperatures. Steel bars are placed in these furnaces and subjected to a constant tensile stress, at temperatures up to 800° C., for long periods. The changes in length of the steel bars are measured by instruments known as extensometers, and readings are taken by observing the reflection of an illuminated scale in a telescope. The apparatus used will detect a change of length of one-millionth of an inch in a steel bar five inches long.
