From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Study on air-springs and vehicle riding-quality, detailing test results and measurement instrumentation.
Identifier | ExFiles\Box 43\3\ Scan018 | |
Date | 28th September 1923 | |
10 AIR-SPRINGS AND RIDING-QUALITY TABLE 3—RELATIVE REDUCTION OF AXLE-MOVEMENT Date of Tests, Sept. 28, 1923 Mason Laboratory Course, 5 Miles; Yale University Test Car, Lincoln Chassis, 1923 Design Mason Laboratory Instrument Used Car Weight, lb. Front 1,880 Rear 2,060+640 Total 4,580 Tire Size, in. Front 33x5 Rear 33x5 Inflation-Pressure, lb. per sq. in. Front 60 Rear 60 Wheelbase, in. 130 Speed, m.p.h. 20 to 25 Number of Passengers as Load 3 Duration of Test, min. 30 Readings | Steel Springs Only | With Westinghouse Air-Springs ---|---|--- A | 484.68 | 583.68 | 563.00 C | 7,114.00 | 6,274.00 | 6,287.00 E | 5,864.00 | 1,717.00 | 1,890.00 Riding-Constant, per cent | 46 | 85 | 85 in. is the increase in compression and 3/16 in. the increase in rebound, a total increase of 13/16 in. in maximum axle-movement. This is the highest individual movement recorded; it is not the average maximum-movement, which of course is considerably less. A rough Belgian-block road generally will show more vibrations and accumulated axle-travel, as both the steel springs and the air-springs must work continuously at faster average-intervals with longer average axle-travel. Generally speaking, the air-spring suspension is there at its best and the snubbers at their worst. As all our tests at present are made to secure data pertaining to the operation of a certain car with and without air-springs, we leave out certain factors that affect riding-quality, such as the center of gravity, balance and the ratio of sprung to unsprung weight. The influence of the air-spring at the axle is contrasted best at a speed of TABLE 4—STUDY OF ROAD TEST OF FRONT-AXLE MOVEMENTS* Percentage of Improvement in Riding Quality over That of Steel Springs Alone. | Snubbers No. 1 | Steel Springs Alone | Air Springs and Snubbers No. 1 | Air Springs Alone ---|---|---|---|--- Front-Axle Movements | Constant at 91.0 | Constant at 88.0 | Constant at 70.0 | Constant at 40 | Percentage Increase 34 | | Percentage Increase 21 | Percentage Increase 40 A—Total Up-and-Down Movements, in. | 77.34 | 144.37 | 99 | 176.34 C—Total Number of Movements | 765 | 1,674 | 823 | 1,386 E—Number of Times the Axle Passes the Normal Line | 967 | 1,724 | 891 | 332 F—Maximum Axle-Travel Above the Normal Line, in. | 1 3/8 | 1 3/4 | 1 3/8 | 1 3/8 G—Maximum Axle-Travel Below the Normal Line, in. | 1 3/8 | 1 3/4 | 1 3/8 | 1 1/2 *This test was made at a speed of 15 m.p.h. on a rough cobblestone and Belgian block course 1 mile in length. 15 m.p.h., as the degree of difference in action between the two cases seems to be at its maximum at this speed. The compression curve of the air-spring alone is practically isothermal, and, as the steel spring shows a straight-line curve, the combination curve reaches a point where the spring scale of the combination equals that of the steel spring alone, but, in the ordinary working zone, the combination-spring scale is considerably weaker than that of the steel spring, which explains the difference in axle-movement. Therefore, the air-spring increases the total axle-amplitude, the vertical, up-and-down flexibility, approximately 28 per cent and reduces the amplitude that is above the normal line approximately 29 per cent; but it increases the same total amplitude below the normal line 105 per cent, thus pushing the theoretical normal line or axis of the axle vibration away from the chassis under running conditions only. The duration of the axle vibration is reduced. Reference is made to Figs. 10, 11, 12 and 13 and also to Table 2. The net reduction over a period of time is approximately 16 per cent, or say a reduction in axle-movement from 5 vibrations per sec. without air-springs to 4.12 vibrations per sec. with them. This reduction is a natural function and still leaves the axle free and flexible in both directions; yet it is capable of very fast action in absorbing shock and making contact with the ground in depressions. The general effect is to smooth out the body line of travel and reduce the objectionable vibrations in the chassis and the body. MEASURING RIDING-COMFORT Before pointing out the weaknesses of instruments, let us consider the other factors that are important in making a road test; and let us keep in mind the fact that, primarily, we want to measure the general riding-quality of the car. Then we can specialize and find out what a certain type of suspension will do on a certain type of road. The two constant factors in making any road test are (a) the course and (b) the instrument. A good course should have a variety of types of road surface, grades both up and down and some turns. It should be of sufficient length to average the human error that is bound to occur in driving. The variable factors are (c) the speed and (d) the ability of the driver to retrace his course in identically the same path at identically the same speed. The two variable factors, (c) and (d), are not so easy to control, especially on a rough course that winds and turns; but this is the type of course on which to compare the general riding-quality of any car. Test runs on particular types of road on straight courses are also of value in determining the performance of any certain car on that particular type of road; generally, they give an indication of what one would expect the car to do on an entirely different type of road, but often they are misleading. A good example of this is that certain snubbing devices are really of negative value on a certain type of road at a given speed, but on an entirely different character of road they will improve the riding-quality of the vehicle considerably. Everyone who has worked long enough on spring-suspension has at some time given some thought to the instrument to be used in measuring riding-comfort, and everyone appreciates that a certain amount of error is unavoidable. The problem is to keep the percentage of error low. The simplest and easiest way to accomplish this is to use an instrument that is an automatic unit-recording machine, making all tests a matter of relative value; instrument error will vary slightly under different AIR-SPRINGS AND RIDING-QUALITY 11 conditions, but the comparison includes the error in any number of cases and, if the other factors in making the test, such as the course, the speed and the driving have average values, the comparative result will be an average. AXLE-MOVEMENT RECORDER We have experimented with a number of different devices and, after considerable preliminary work, have decided that the spring-suspended weight-type is best suited for our work. We sidetracked the seismograph principle for the same reason that we gave up trying to analyze long graphical axle-movements, and developed a mechanical axle-movement recorder. We appreciated the accuracy of the records as traced graphically in both cases, but the work of analyzing hundreds of different test runs could not be considered. The spring-suspended weight-type undoubtedly is open to criticism because it requires inherent damping action to remove the effects of its own vibration; moreover, some of the instrument vibration will be accumulated with the car vibrations and cannot be separated from them. However, this type can be depended upon to tell the riding story accurately enough to be of immense value in comparing the results obtained with various spring-suspensions. The design may have a vertically operated weight or one that swings on the arc of a circle. The latter type, though complicated and limited in action, is attractive, as a low-period flat spiral-spring can be used easily to support the weight, and the accumulated vibrations can be recorded easily. The sensitiveness can be controlled by the amount of the weight that is supported by the spring. Just how sensitive the instrument should be is an important right now, but at some future time this question must be answered because we must decide what acceleration is objectionable for both the passengers and the car. The period and the amplitude of any given vibration govern the acceleration and, likewise, the sensitiveness of the instrument. The instrument we are using now is yet in a crude state, yet we consider it reliable enough to give fair average comparisons as to the relative merits of different spring-suspensions. Fig. 18 is a photograph of an early instrument. It is nothing more than an “over-grown” pedometer. The recording apparatus of a pedometer is impelled by an oscillating weight that is supported by a lever-arm, the whole mass being nearly counterbalanced by an adjustable spiral spring. The pedometer records the vertical movements of the human body, and then attempts to translate these movements into terms of dis- FIG. 18—THE VERTICAL VIBRATION-RECORDER THAT OPERATES ON THE SAME PRINCIPLE AS A PEDOMETER In Its Present State of Development the Device Is Considered Reliable Enough To Give Fair Average Comparisons as to the Relative Merits of Different Spring-Suspensions. Its Action as a Pedometer Is Impelled by an Oscillating Weight Supported by a Lever Arm, the Entire Mass Being Nearly Counterbalanced by an Adjustable Spiral Spring tance covered. The translation is the difficult part but, for the purpose for which we use our device, the record needs no translation into terms of distance covered and so the problem is much simpler. Fig. 19 shows the instrument we are now using. It is compact, easily handled and can be used by any skilled mechanic. It is simply a refinement in both design and workmanship of the crude device shown in Fig. 18. The weight is spring-suspended through the radius-arm, the spring being located in a central position on the outside of the friction drum that actuates the tape. This drum is floated on ball bearings and rotated by pawls that connect to the radius-arm, so that it gives motion only in one direction. METHOD OF RECORDING VIBRATIONS Regular ticker-tape is used to record the vibrations. This tape is pulled by the rotation of the friction drum as the weight goes up; so, when the weight returns, the tape is stationary, and a small pin in the weight perforates and leaves a permanent record on the tape for each throw. A long throw will show considerable space between the perforations. Short, snappy, fast vibrations condense the perforations. The entire story, therefore, is written on the tape. Measuring the length from the first perforation to the last, we have the accumulated weight travel. By counting the perforations, we have FIG. 19—THE PRESENT VIBRATION INSTRUMENT This Is a Development of the Instrument Shown in Fig. 18 and Involves the Same Principles. This Instrument Weighs 15 1/2 Lb., Is Totally Enclosed, as Shown, but with Non-Breakable Glass Windows, So That Its Action Is at All Times Visible to the Tester. It Sits on the Knees and Wrists of the Passenger, or the Passenger as the Shocks Occur | ||