1、 PART 7 Measurement of Shaft Power . - - d ,. “r: . ,.- L ASME PTC*197 80 II 0759670 0052723 T II PART 7 MeasureMent of Shaft Power ANSI I ASME PTC 19.7-1980 INSTRUMENTS AND APPARATUS THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS United Engineering Center 345 East 47th Street Ne“ York, N.Y. 10017 ASM
2、E PTC*19-7 80 0759670 0052724 1 No part of this document may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher Date of Issuance: August 31, 1980 Copyright 1980 THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS All Rights Reser
3、ved Printed in U.S.A. ASME PTC*19.7 80 II 0759670 0052725 3 II FOREWORD The Performance Test Codes Supervisory Committee in December 1974 activated a Committee to revise PTC 19.7 (1961) on Measurement of Shaft Powe. This Instruments and Apparatus Technical Committee has prepared an Instruments and A
4、pparatus Supplement which incorporates the latest technology on the Measurement of Shaft Power. The Scope of the work of PTC 19.7 on Measurement of Shaft Power is limited to descriptive material which will enable the user to select an appropriate system or procedure for his applica tion. It includes
5、 criteria for the operating conditions of the equipment whose power is being measured. The Object of this Supplement is to describe the function, characteristics, advan tages, disadvantages and accuracy of equipment and techniques currently available for the measurement of shaft power in rotating ma
6、chines. Only the methods of measurement and instruments, including instructions for their use, specified in the individual test codes are mandatory. Other methods of measurement and instru ments, that may be treated in the Supplements on Instruments and Apparatus, shall not be used . unless agreeabl
7、e to all the parties to the test . This Supplement was approved by the Performance Test Codes Supervisory Committee on july 2, 1979. It was approved and adopted by the American National Standards Institute as meeting the criteria for an American National Standard on April 28, 1980 . iii 0 ASME PTC*1
8、9.7 80 II 0759670 0052726 5 II PERSONNEL OF ASME PERFORMANCE TEST CODES COMMITTEE NO. 19.7 ON INSTRUMENTS AND APPARATUS MEASUREMENTOFSHAFTPOWER Hunt Davis, Chairman and Secretary Robert R. Piepho, Vice Chairman Arthur L. Beaman, Jr., Fellow Design Engineer, Medium Motor and Gearing Division, Westing
9、 house Electric Corp., P.O. Box: 225, Buffalo, New York 14240 Robert Clelland, Senior Principal Engineer, Rotating Equipment Section, Pullman Kellogg Divi sion, Pullman, Inc., Three Greenway Plaza East, Houston, Texas 77046 Hunt Davis, Senior Staff Engineer, Pullman Kellogg Division, Pullman, Inc.,
10、Three Greenway Plaza East, Houston, Texas 77046 Ralph jaeschke, Chief Engineer, Eaton Corp., Industrial Drives Division, 3122 14th Avenue, Kenosha, Wisconsin 53141 -Donald R. Jenkins, Associate Professor, Mechanical Engineering Department, Lafayette College, Easton, Pennsylvania 18042 Irving I. Kahn
11、, President, Kahn Industries Inc., 885 Wells Road, Wethersfield, Connecticut 06109 Douglas C. Falkner, Senior Test Engineer, Cameron TestDepartment, Ingersoll Rand Co., Phil lipsburg, New Jersey 08865 Richard A. Mayer, Senior Consultant, Applied Physics Division, Southwest Research Institute, 6220 C
12、ulebra Road, San Antonio, Texas 78284 Daniel Nobles, Research Engineer,Worthington Pump International, Harrison, New Jersey 07029 Robert R. Piepho, Product Manager, Fuel Preparation Systems, Fossil Power Generation Divi sion, Babcock that is, there are no transients in either torque or angular veloc
13、ity within the time interval required tomake the measurement. 2.02 The measurement of shaft power of rotating rna“ chines in the absence of transients, is accomplished by either direct or indirect methods, The direct methods, utilizing a dynamometer or a torque meter, involve deter mination of the v
14、ariables in the following equation: Physical equation P=wT where P = power w =angular velocity T= torque Power expressed in Sf units P=wT where P =power, watts (W) w =angular velocity, rad/s T torque, newton meters (Nm) Power expressed in English units P = 21rnT 550 where P =power, horsepower (hp n
15、=rotational speed, revolutions/sec (r/s) T= torque (lbfft 2.03 There are cases in which transients in angular velocity and torque occur. Some of the apparatus described herein may, under certain circumstances, be capable of making measurements of instantaneous power, when angular velocity and torque
16、 vary with time; or measurements of 3 average power when angular velocity and torque vary cyclically. 2.04 For these cases of non-steady angular velocity and torque the instantaneous value of power is, in physical terms, P=w T Torque meter systems with appropriate data recording systems may be used
17、to determine the value of Tat any instant. A similar recording of angular velocities, w, then provides a basis for determination of the instantaneous value of P. If the values of T and w vary cyclically the average power may be determined as follows. Let the period of one cycle of torque and speed b
18、e the time b., and the ro tational travel for one cycle be the angle (), in radians. Also let !f be defined as !f = fo E 1-L-,-RANDOM ERROR DISTRIBUTION TEMPERATURE, T FIG. 1 SYSTEMATIC AND RANDOM ERRORS ILLUSTRATED FOR CASE OF A THERMOCOUPLE CALIBRATION (Ref. 1) confidence interval statement is giv
19、en by power= 386 6 kW; (95%) where 386 kW is the estimated value, based on averaging a number of measurements, 6 kW is the confidence interval; and the probability of the true value ofpower being in the range 38.5 to 392 kW is 95%. For a given set of observa: tions, as the confidence interval is mad
20、e larger, the prob ability becomes greater. 2.09 Measurements of shaft power involve determination of multiple physical quantities. The complexity of the ap plication of the foregoing statistical concepts, and the pro cedures of References 1 and 2, Appendix D, varies greatly depending on the method
21、and apparatus to be used. 2.10 The contribution of systematic errors is minimized by careful calibration of individual components. The con tribution of random errors is minimized by increasing the number of readings of output for fixed values of the con trolled operating parameters. 2.11 The term pr
22、obable error refers to the confidence inter val around the estimated value for which the probability is 500;0. 2.12 In the succeeding parts of this Supplement, numerical values are assigned to the “errors“ or “overall errors“ of the various systems. Each of these is to be considered as the probable
23、error of a single determination, having a 50% probability that the given range includes the real value. These overall errors are usually dominated by random sources because the procedures require calibrations be ASME PTC*19.7 80 II 0759670 0052734 4 II ANSI/ASME PTC 19.7-1980 SECTION 2 _, X x x POOR
24、 PRECISION -POOR ACCURACY POOR PRECISION GOOD ACCURACY x x BIAS GOOD PRECISION POOR ACCURACY GOOD PRECISION -GOOD ACCURACY FIG. 2. ACCURACY AND PRECISION (Ref. 1) performed where appropriate. The nature of the systems which are subject to calibration is such that the confidence interval (500) of the
25、 calibration data is usually consider ably smaller than the stated overall error. This fact can be understood when one realizes that calibrations are usually conducted repeatedly and a history of consistency in data is obtained; furthermore the calibrated component often provides only one of numerou
26、s inputs required fora deter mination of shaft power. 2.13 Causes of errors given for the various measurement methods are categorized as predominantly random or sys tematic, where possible. These listings provide guidance about a the potential for improvement in results attain able by calibration, a
27、nd b) by maldng multiple determina tions of the shaft power for each value of the independent operating parameters. 2.14 If an individual instrument has from. experience and numerous calibrations a known probable error, signified 5 by C/; (50%), the confidence interval for higher probabil ities may
28、be assumed to be related as follows: Cl; (90%) 1.645a _ 2 44 C/1 (50%) 0.674a- Cl,(95%) “ 1.960a = 2 91 C/j50% - 0.674a . Cl; (99%) ,_ 2.576a = 3 82 C/1 (50%) - 0.674a These ratios may be useful in relating historical “prob able error“ descriptions of uncertainty to confidence inter vals having high
29、 probability. The latter are used in Refer ences 1 and 2, Appendix D. 2.15 The application of statistical procedures to deter mine the confrdence interval statements for shaft power measurements is a matter for judgment of the parties con cerned with the testing of equipment under one of the ASME PT
30、C*19.7 80 II 0759670 0052735 6 II SECTION 2 ANSI/ASME PTC 19.7-1980 Performance Test Codes. The basis for statistical treatment of data, calibration reference standards, number of data points, probability levels desired and computational pro-cedures should be agreed in advance. Table 2 outlines typi
31、cal values of probable error for the methods described. TABLE 2 Summary of Typical Probable Errors for Different Shaft Power Measurement Methods Reaction Torque Systems Cradled Dynamometers Uncradled Dynamometers Shaft Torque Measurement Surface Strain Systems Angular Displacement Systems Mechanical
32、 Electrical Optical Calibrated Motors pneumatic load cells or mechanical scales. The force measuring device is attached to the torque arm at a known distance from the centerline of the dynamom eter. 3.07 Strain gage load cells are available in various con figurations. Most models can be used for eit
33、her load direc tion in the line of force. Maximum deflection when loaded is approximately O.Dl in. (0.25 mm). Most models are temperature compensated for a temperature range up to 100F (55C) above normal ambients. Strain gage load cells are used with analog or digital torque indicators. Error of str
34、ain gage force measuring systems is typically 0.2% of full scale reading if digital indication is used, and 0.5% of full scale for an analog indicator. The advantages of the strain gage load cell systems are their compactness, low error, and provision for auxiliary instrumentation and controls from
35、the electrical output signal. The disadvantages are the effects from stray vibra tions produced by other operating equipment, the limited fatigue life of the gage bondings, and the need fortem perature compensation. 3.08 Hydraulic load cells have a piston and cylinder assembly which converts a force
36、 into a proportional hy draulic pressure signal. Maximum deflection under load is approximately 0.01 in. (0.25 mm). Some models are avail able with a temperature compensator and a signal damping device. Hydraulic load cells are used with precision Bour don tube pressure indicators. The pressure gage
37、 may be calibrated in torque units for a particular dynamometer torque arm length. The hydraulic system must be com pletely filled with fluid and any air must be purged. The overall error of the force measuring system is between 0.2% to 0.5% of full scale reading, depending on the load capacity of t
38、he system, the mechanical condition of the pressure indicator, and the cylinder and piston assem bly. Generally, large systems with high load capacity have lower error. 9 SECTION 3 The advantages of this type are its ruggedness, freedom from temperature effects, and low error. Its disadvantages are
39、its bulkiness, the need to convert pressure to electrical signal for auxiliary control, and the hydrostatic correction needed for difference in elevation between the load cell and the indicator. 3.09 Pneumatic load cells have essentially the same design features, operating characteristics and errors
40、 as hydraulic load cells. Bourdon tube pressure gages or mer cury manometers are used for force (or torque) indication. A continuous supply of clean, dry, regulated air at about 100 psig (690 kPa gage) is required. The advantages and disadvantages are similar to those of the hydraulic load cells, ex
41、cept that the elevation consideration is negligible. 3.10 Mechanical scales are weighing systems containing levers with knife edge fulcrums or other low friction link ages which attach to the dynamometer housing at a known torque arm distance from the machine centerline. These systems may require di
42、splacement when loaded, which should be small in order that the effective torque arm radius is not affected by the displacement. 3.11 Most mechanical scales have a reversing linkage so that the force scale will read in positive direction regard less of the direction of torque. If a dynamometer is so
43、 equipped, the housing should be preloaded slightly in the direction in which it will tend to move under running conditions. This prevents the reversing linkage from causing a scale error as it passes through its neutral position. Pre loading of horizontal dynamometers should be done by attaching a
44、counterweight from the proper checking torque arm; or by attaching a counterweight to the frame so that the center of gravity of the frame is shifted horizontally, but not vertically. The scales can be reset to zero to com pensate for the preload weight. The reversing linkage con tains a lever havin
45、g a 1 to 1 ratio and any error in this ratio will cause the scale calibration in one direction to differ from that in the other direction; hence, the calibration should be checked in both directions. The preload should be reversed whenever the direction of torque is reversed. 3.12 Friction in the sc
46、ale, linkage or knife edges, or cradle bearings will show up as a difference of scale read ings for the same calibrating point dependent upon the direction from which the point is approached. For this reason it is necessary to calibrate the scale by adding weights carefully so as not to overshoot th
47、e readings, and then to remove them in the same manner. The difference between the increasing and decreasing curves is an indica tion of possible error due to friction. 3.13 If the dynamometer weighing system uses a dash pot to damp the torque arm fluctuations, its adjustment ASME PTC*197 80 II 0759
48、670 0052739 3 II SECTION 3 should be checked periodically to determine that the dashpot effect is equal on the “in“ and “out“ stroke. This check should be performed with the weighing system free of influence from the dynamometer. The natural frequency of the dashpot should be at least twice, and pre
49、ferably five times, the highest rotational speed of the dynamometer. 3.14 The overall error of power measurement by a cradled dynamometer results from several contributing factors: (a) Trunnion bearing friction. (b) Force measurement system error. (c) Torque arm and calibrating arm length error. (d) Static unbalance of dynamometer housing. (e) Restriction of free movement caused by water lines, lubrication lines and electrical leads connected to the dynamomet
