Document BymL8E0MK18kZvYK79Je3j09o

First step In making pip* model It to drow on tiomotric sketch of piping, o* oulhor does hero, to fix model slie KJjembly of elements In boiler superheater mode). The rod M Closeup of modal showing' tuperheoter-oloment, section of Kitmenti ore silver soldered to Insure complete fixation .main tleom piping attached to electrical measuring units How Model Tests Cut Piping Design Ibrication Costs Use of scale models to predetermine stress conditions in high*temperoture piping assures economic design, good performance of modern power plants By JOHN t O'ROURKE, Methodical Engineer, Tfct M W Kellogg Co Recent trends toward higher pressures and temperatures in power-plant installation* have increased the importance ot adequate engineering for thermal expansion. Proper design and layout.of piping is essential for satisfactory service per formance. Inadequate analysis in this respect has resulted in the shutdown o( entire plants due to fatigue failure of piping or distortion of connected equipment Mathematical analysis of the expansion stresses and reactions \% broadly covered by the General Analytical Method presented by The M W Kellogg Co in Design of Piping Systems, 1941. In employing this method, the solution of up to 16 simultaneous equations, which represent, in general, a system of four points of fixation, can be handled effectively. With an increase in the number of fixed ends or intermediate restraints, the mathematical approach becomes extremely laborious and time-consuming. The model test solution of piping systems has been evolved to solve problems too intri* cate for calculation or those where the mathematical solution is not economically feasible. Modol feats have additional appeal to designers of hightemperature piping in that they permit visual evidence of the interaction of oil mombera in the system. Addition of guides or ohanges in configuration can be made directly on the model. and results from alternate tests readily obtained. Since alii piping costs ore a significant item in high-pressure p1' the model tests can be used with considerable advantage toward the elimination of excess piping. Model tests have been conducted for hundreds of | petroleum and chemical plants, and for marina installation for she United States and British Navies. Most recently, tl have been employed for atomic-power plant design. Prindplo utilised in the model test method ia that of d plocing the ends of the raodcl^imulated pipeline with r to an assumed fixed point, and measuring the end readied* created by this springing of the model. Displacement*'*| plied arc a percentage of the calculated free thermal oipsa,-. sion of the ends from the fixed point, including the expsnsiott of connecting equipment. The forces and moments at the wdt of the model are proportional to those occurring in the a ` pipeline when brought up to operating temperature. Model Construction. Solid oold-rollod steel rods, fig* f are used in building the model. They are bent to shape cold, and are braced together. Model ends are silver-soldered i ^ 1-In.-square steel blocks having topped boles to focHitaiCjl fastening to the measuring units. The length scale used fsrC the model depends on the dimensions of the testing apparstols and its arrangement. Moat models oome within a range sjff !4*in.--1 ft to 1 in.si ft, though some tests'have been rapji on full-acale models of short connecting lines where operating conditions were critical. Where the system has the same size pipe throughout, wfjl rod diameter from Vs through Vi In. is used. If two or irwryi pipo sizes are used in the actual system, their stiffness ratio ia incorporated in the model. This Is done by obtaining 1 product of the moment of inertia and the modulus of elssth city of the actual pipe and choosing rods in (his relsIfooValves or similarly stiff members ara represented In 1 90 ENOINHIINO AND MANA0EMIN1 SECTION R,-rIjwgc^hy solid steel blocks or other heavy sections. In recent fjmsiU has been e practice to include the terminal elements Ktffhe.superheater as part of the piping to increase the gSterjaU flexibility of the system. These complicated lube ffmanigements ore reproduced by bending drill rod of correct tiffness Id represent the tubes running to tho header from Us/point of exit through the boiler arch, Fig. 3 and 4. gpsi apparatus consists of an adjustable tubular framefyerk: for supporting the displacement units, which apply fmavtiaenti to the ends of the model. Machined tubular-steel VpMts.exlead 'from floor to ceiling of the test room, Fig. 6. Kurthese pans, tubular arm* may be attached to give access SftjsnV point within a 7-foot radius. The arms are splined to Qttceive the micrometer diaplocement units. ffi'lho displacement units combine three mutually perpenfgicular slides, each operated by a micrometer lead screw fjwjup.t thimble graduated in thousandths of on inch. sfiftThe alideways arc hand-scraped to ovoid any possibility undesirtd motions during application of the displacements Sjtdea arc locked In position after movements hove been made. _$!Jhe electrical measuring unit* resolve the force and moment jst.eseh end of the model and meosure the three coordinate xmponenii of each. This combined action Is accomplished S^epplying the force and moment, by means of a threeidjineRiionsI cross, through six cantilevers aligned will) the jjtoree coordinate planes. pairs of strain gages are cemented to the cantilevers, resistance varies in proportion to the cantilever deflec- fi-r*' by measuring the change in resistance of the K&>> "rain gages, the component forces and moments may IsSyound. Each pair of gages is permanently wired into a 9^h^ststone bridge circuit. A self-balancing potentiometer, frJft-.'T' gives the readings directly in hundredths of a pound. ? .'> (Continued on page 02} 91 sw--rmnaronTw-- 11 -----------------------Turbina leads from main steam piping attached to electri cal measuring units for determining system end reoctrons General view of model test apparatus showing moin tteom lino running from tho superheater to turbine-lead takeoff