From the Crystal Palace to the pump room.
Today, centrifugal pumps and compressors have reached efficiency levels above 90 percent and are built in sizes ranging from a few watts to megawatts. As machines have evolved to handle a wider range of liquids and gases at higher pressures and temperatures, whole industries have become dependent on them.
The centrifugal impellers in these pumps and compressors make them among the most complex fluid machines ever built. Although the most rapid development of the centrifugal impeller has occurred in the last 60 years, the device has a long history that parallels the rise of modern engineering itself. It is this history - for the perspective it gives to current applications - that we will examine here. The centrifugal impeller is the most widely used turbomachine, and continues to be a major research and development topic.
The first device believed to have functioned as a centrifugal impeller had 10 wooden double-curved blades and has been dated approximately to the fifth century. It was found in 1772 in an abandoned Portuguese copper mine in San Domingos. Nobody knows how it was used.
Centrifugal fans had been used for mine ventilation as early as the 16th century.
Just who invented the centrifugal impeller is a disputed issue. Some would give the credit to Leonardo da Vinci, who suggested the idea of using centrifugal force for lifting liquid, or to Johann Jordan, a less celebrated figure who is reported to have discussed a similar theory around the year 1680.
Most place the origin of the centrifugal impeller with the French physicist and inventor Denis Papin in 1689. Papin's contribution lies in his understanding of the concept of creating a forced vortex within a circular or spiral casing by means of blades.
Following Papin, Kernelien Le Demour, in 1732, and Daniel Gabriel Fahrenheit (the German physicist of temperature scale fame), in 1736, described other designs for centrifugal impellers, but there is no evidence of their practical use.
Leonhard Euler, one of the founders of higher mathematics, presented in his 1754 memoir an idealized theoretical application of Newton's law to centrifugal impellers, based on a conceptualization of a tubular turbine run backwards. This theoretical application gave rise to what is now universally known as the Euler equation. Euler initiated a true mathematical inquiry into the employment of centrifugal force as a means of raising a fluid.
The publication of his equation spurred development of hydraulic turbines in the 18th century, but did little to influence the development of centrifugal impellers, which had to develop gradually through tedious cut-and-try methods.
Just prior to Euler's publication, in 1752 John Smeaton introduced the use of models in the study of turbomachinery. He also defined power as equivalent to the rate of lifting of a weight, a concept that is still fundamental in thermofluids.
The construction of a centrifugal impeller, even of metal and in fairly large sizes, would have been tedious but possible in the 18th century. Why then was a successful pump or compressor with a centrifugal impeller not commercially developed before the 19th century? One hindrance was probably that the centrifugal machine needs high speed, and until the end of the 19th century there was no form of motivational power well suited to it. Another possible obstacle was the popularity of reciprocating pumps and compressors, which satisfied the technical requirements of those times.
The development of the centrifugal impeller has been helped because the device can be used in both pumps and compressors. The earliest development occurred in pumps, which were more advanced than compressors until early in the 20th century. Since World War II, the leading knowledge has come from the area of compressors.
The year 1818 was a landmark in the history of the centrifugal impeller, because commercial production of the centrifugal pump started in the United States at the Massachusetts Pump factory. The Massachusetts Pump, which was based on Papin's theory, had an open impeller with straight tangential vanes that revolved in a casing of rectangular section, and was roughly of volute form.
In 1831 the Blake Co. of Connecticut brought out a vertical pump, which employed a horizontal disc with a series of radial vanes attached to the underside and running just clear of the casing, thus introducing the half-shrouded centrifugal impeller design.
Following Papin's theory, Combs presented a paper in 1838 on curved vanes and the effect of curvature, which subsequently proved to be an important factor in the development of the centrifugal impeller. In 1839, W.H. Andrews introduced the proper volute casing and in 1846 he used a fully shrouded impeller.
W.H. Johnson constructed the first three-stage centrifugal pump, also in 1846, and James S. Gwynne constructed a multistage centrifugal pump in 1849 and began the first systematic examination of these pumps.
At about the same time, John Appold conducted an exhaustive series of empirically directed experiments to determine the best shape of the impeller, which culminated in his discovery that efficiency depends on blade curvature.
James Thomson, in about 1850, suggested the use of a whirlpool chamber, which enabled the single-stage centrifugal pump to be used economically for higher lifts.
A highly visible test for the centrifugal impeller came during England's Great Exhibition of 1851 at the Crystal Palace, where several designs for pumps were presented and compared. Appold's pump with curved blades showed an efficiency of 68 percent, more than three times better than any of the other pumps present. Appold's design attracted much attention because of its simplicity, compactness, and high efficiency. From then on, the development of the centrifugal pump was rapid.
Following Osborne Reynolds' patent of a vaned diffuser in 1875, pumps with this type of diffuser were built in 1887. Widespread manufacturing of centrifugal pumps by Mather & Platt, Sulzer Brothers, and other companies began after 1893.
Soon centrifugal impeller design using one-dimensional flow theory became routine. Every manufacturer developed its own approach and, although each had a slightly different method of calculation, the broad underlying principles were similar.
None of the early pumps or compressors employed a diffuser, and all were used for low heads and pressure ratios. Most developed efficiencies of 45 to 65 percent and were driven directly by a reciprocating steam engine. The development of the centrifugal compressor impeller is directly related to the development of the gas turbine, as well as to the steel and mining industries.
By late in the 19th century the centrifugal gas compressor was able to deliver a head of water only 1 meter high. In 1899, A. Rateau showed that a centrifugal compressor impeller could achieve a tip speed of 250 meters per second or higher. He first designed and tested a centrifugal compressor with an impeller that was 250 mm in diameter. This compressor developed a head of water 5.8 meters high and an inlet flow of 2,000 cubic meters per hour, while operating at a tip speed of 264 meters per second and a rotational speed of 20,200 rpm.
Following the success of this centrifugal compressor, Rateau designed two more. The first, in 1903 for the steel industry, developed a head of water 3.4 meters high and an inlet flow of 10,000 cubic meters per hour while operating at a rotational speed of 14,500 rpm. The second was a similar compressor in 1904 for a sugar refinery.
Rateau designed the first multistage centrifugal compressor, which was built in 1905. With five stages on the same shaft, it was designed for an inlet flow of 2,500 cubic meters per hour, a 4-meter head of water, and a 4,500 rpm rotational speed.
It took several decades to discover the superior qualities of the axial compressor. Following Rateau's success, many compressor manufacturing companies licensed his patent. From 1910 to 1920, as industrial applications of the centrifugal compressor increased, impeller designers focused on producing higher flow capacity, efficiency, and tip speed.
About the same time as Rateau was working, Charles Parsons investigated axial compressors for gas turbine applications, but results showed that this compressor was poor in terms of efficiency and stability. Parsons investigated some 30 different compressors, but ultimately stopped his work in dissatisfaction.
The theory and functioning of a gas turbine were known long before the necessary materials and the detailed knowledge of flow mechanisms were available. The gas turbine concept was patented in 1791 by John Barber, based on a reciprocating compressor.
But the development of a useful gas turbine did not occur until Brown Boveri demonstrated the first sizable gas turbine power plant at the Swiss National Exhibition in Zurich in 1939. The first turbojet-powered flight also took place the same year.
A compressor with a centrifugal impeller was the first type to attain a range of pressure ratio and efficiency useful for turbojet engines. Such a compressor was used in the von Ohain engine in 1939 and the Whittle engine in 1941. Many of today's jet engines have axial rather than centrifugal compressors because the latter would have to be impractically large to deliver the required flow.
EARLY FLOW PHYSICS
The theory of hydrodynamics and aerodynamics of centrifugal pumps and compressors has developed slowly in this century. The major initial task was to establish the relationship among the performance parameters of the impeller, the form of the flow, and the form of the fluid passages. One of the most important of these relationships was that between the torque transmitted from the impeller to the fluid and the change of the properties of the fluid as it passes through the impeller. This relationship is the Euler equation.
However, there were many intrinsic functions of the impeller that are not covered by the Euler equation and yet have equal significance.
Early in this century, the centrifugal impeller designer developed a moderately efficient pump or compressor by considering only gross effects by a one-dimensional theory and then resorting to empirical development based on many years of experience in order to obtain acceptable designs. The designer did not pay much attention to details of the impeller flow.
The first intense research into centrifugal impellers occurred between 1910 and 1930. Using the ideal assumption of two-dimensional irrotational flow of an inviscid and incompressible flow in the impeller, A. Busemann, W. Kucharski, F. Konig, W. Spannhake, and others attempted a theoretical analysis of the flow. A. Stodola, D. Thoma, E. Kearton, W. Kucharski, E. Grunagel, and W. Stiess experimented to gain an understanding of the flow mechanism in a centrifugal impeller. They were able to observe secondary flows but, owing to their lack of adequate measuring and analytical tools, they were unable to extend their observation with quantitative structure.
The first comprehensive study of impeller flow is that of K. Fischer and D. Thoma ("Investigation of the Flow Condition in a Centrifugal Pump," Transactions of the ASME, Vol. 54, 1932), in which they showed that practically all flow conditions for a real fluid were actually different from those theoretically derived for an ideal frictionless fluid. The actual velocity distribution along circles concentric with the axis of rotation of the impeller was not the same as generally assumed.
In 1915, R.L. Daugherty's book Centrifugal Pumps appeared, and became the design reference of the 1920s. However, the significant authoritative reference books on the complete design of the centrifugal impeller were produced later by C. Pfleiderer and A.J. Stepanoff.
Until the end of the 1950s, due to the lack of proper knowledge concerning radial thrust in volute pumps and compressors, designers faced the serious problems of rapid wear in the bearings, leakage from the glands, and failure of the shaft due to fatigue.
Currently, a typical centrifugal impeller is designed and optimized with the help of a combination CFD and other programs. One code automatically generates input files for another code, and output data are displayed graphically.
Centrifugal impellers are currently found in a wide range of pumps, small gas turbine engines, turbochargers, and refrigerators, and are used extensively in the petrochemical and process industries. Pump applications may be classified into four major areas: power generation, water supply, process and environmental uses, and semi-industrial and general industrial applications.
The wide range of demands placed on centrifugal impellers brings many design requirements that must be met. These requirements fall into two categories: stress and fluid dynamics. The stress problems are caused by material strength limitations. These problems make it necessary to accurately predict blade and rotor steady state and vibrational stress for complex impeller shapes and high rotational speeds.
The fluid dynamics problem is to efficiently accomplish large deflections and diffusion at high velocity, with the added difficulty of very small flow passages required for efficiency and high pressure.
Even though the individual components of the compressor or pump are capable of achieving high efficiency, it is the efficiency of the whole stage that is of great importance. Thus, component matching is an essential aspect of design. Often, engineers must redesign one or more compressor components due to improper matching, and sometimes they sacrifice the efficiency of a component to achieve good matching.
The gas turbine centrifugal compressors are usually the most demanding from the standpoint of performance and mechanical factors. The requirements for these machines include medium- to high-pressure ratio, higher efficiency demands at diverse speeds, and restricted overall diameters. However, range demands are usually not as great as for the other applications.
Today, only low-flow gas turbines use centrifugal compressors. In aircraft, centrifugal compressors are found in turboprops, helicopters, and executive jets. In land-based gas turbines used for power generation and as drives, centrifugal compressors are found only in applications producing under 1 MW.
In process applications, the molecular weight and distribution of component gases may vary significantly during the lifetime of a given compressor. It is not uncommon to find applications that require 30 or 40 percent stable operating range (design flow to surge flow) to meet the diverse changes in possible operating conditions.
Refrigeration compressors are quite similar to process compressors except that their stage pressure ratio and flow conditions are defined by a unique thermodynamic cycle. Due to the thermodynamic properties of refrigerants, one is frequently operating quite close to the top of the vapor dome, and so design calculations with real gas properties often become mandatory. Perhaps the most distinguishing feature of the refrigeration compressor is the requirement for operation in very wide climatic conditions through summer and winter, thus requiring loads anywhere from 10 to 110 percent of the design load.
Today, a centrifugal-impeller pump with a head of 850 meters per stage, corresponding to a tip speed of about 135 meters per second, is commercially available. It is unlikely that the head will exceed 1,000 meters per stage any time soon, but the average head per stage will likely continue to increase. High-performance centrifugal compressor impellers can achieve pressure ratios above 10 per stage and tip speeds up to 600 meters per second.
This article is adapted from a technical paper (98-GT-22) presented at the 1998 International Gas Turbine & Aeroengine Congress & Exhibition in Stockholm, Sweden.
Abraham Engeda is an associate professor of mechanical engineering and director of the Turbomachinery Lab at Michigan State University in East Lansing.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||centrifugal pumps; facility in Hyde Park, London, England|
|Date:||Feb 1, 1999|
|Previous Article:||Upgrading the old turbine.|
|Next Article:||Touching the right nerve.|