December 1993
Contents
1. a single jointhas no restrictions on its motion but except for hoop stress it also has no pressure containing capabilities In exchange for providing the most freedom this joint requires special care in overall design The piping around the joint must be well guided to prevent any squirm in the joint that would be further aggravated by pressure thrust on the joint Simple expansion joints also require axial stops or anchors COADE Mechanical Engineering News somewhere up and down the line to absorb the pressure thrust load So with all the freedom in this joint the pressure containment requirements reduce its flexibili ty in most applications to only the axial direction A Double Expansion Joint is a related configuration This is an assembly of two single joints separated by a short run of pipe The pipe separating the two joints is usually restrained from motion by an anchor The use of a double joint is the same as a single joint but it can share the total axial deflection between the two joints TIED Single Tied Expansion Joint Pressure thrust on the joint may be contained without adding anchors and guides to the attached piping Instead the axial thrust is checked by the addition of tie rods around the joint These tie rods drastically alter the nature of the joint When untied the joint provides flexibility in the axial direction Now tied the joint is essentially rigid in the axial direction but lateral flexibility is2. edition differs only slightly from the 1988 8 edition Appendix P which deals with nozzle flexibilities corrects all of the mistakes in the 8 edition and modifies the equation for nozzle rotation The main objective of incorporating API 650 nozzle flexi bilities into a piping analysis is to ensure that the proper loads can be computed at this point and that these loads will not exceed the maximum allowed loads also determined by API 650 Although there are three global forces and three global moments which can be applied to the tank opening the radial force the longitudinal moment and the circumferen tial moment are the only three normally considered signifi cant enough to cause shell deformation The stiffnesses associated with these three forces and moments are comput ed and automatically inserted into the CAESAR II analysis when the user specifies an API 650 nozzle The figure below shows the orientation of these loads P i e IOT Ee a D Se Longitudinal Moment M Circumferential Moment Mo Figure 1 December 1993 The API 650 Appendix P also provides equations to com pute the radial growth and longitudinal rotation of the shell at the nozzle location These values are computed by CAESAR II and provided as additional information These values are not automatically applied as pre defined dis placements If these values are desired as pre defined displacements the user must manually speci
3. 0 2 Where u first derivative of axial strain with respect to x y distance of point of interest from neutral axis 6 first derivative of angle of curvature with respect to x Total elemental strain energy U considered over the entire length and the entire cross section is calculated as U ZZ E g dAds U ZL E w y amp 12 dAdx Where E material modulus of elasticity A cross sectional area of element Noting that for a prismatic member ZA A Zi dA 0 Zz dA I Liu dA P Where I moment of inertia of element P axial force acting on element cross section The elemental strain energy can be written as U ZAE 2u dx ZQII2 P dx Z 2 P dx The first integral in the expression above provides the standard stiffness matrix for a truss element or the axial terms of the stiffness matrix for a beam element The second COADE Mechanical Engineering News integral provides the lateral and rotational stiffness matrix for a beam element These are the only stiffness terms that are included in CAESAR II s default stiffness matrix formula tion as well as that of virtually all other stiffness method programs The third integral however represents the work done when differential elements of size dx are stretched an amount of 6 dx 2 by an axial force P From this integral then the stress stiffening matrix K_ which should be added to the default S element stiffness matrix can be deriv
4. NOZZL3 This model is identical to the original except that pressure strain effects often called the Bourdon effect have been included This effect considers the elongation of pipes under pressure tensile stresses as a loading condition For example the piping codes generally require that pressure be considered as having not a strain component but only a stress component which should be simply added to the calculated sustained stresses S PD7 D 7 D iM Z sl Where S total sustained stresses P internal pressure D internal diameter of pipe D_ external diameter of pipe o i stress intensification factor M moment on cross section due to non pressure sustained loads Z section modulus of pipe 17 COADE Mechanical Engineering News For straight pipes the pressure strain can be calculated in a straight forward manner Longitudinal pressure stresses cause longitudinal strains approximately equal to Pr 2tE which are reduced by the Poisson contraction due to the hoop and radial stress approximately equal to 2vPr 2tE for a net longitudinal strain of 1 20 Pr 2tE Where longitudinal pipe strain due to pressure V poisson s ratio P internal pressure r internal radius of pipe t pipe wall thickness E modulus of elasticity of pipe material For elbows the calculation of the end displacements is more complex Assuming that the pressure strain is constant through
5. The flexibility of the joint cannot limit the axial deflection of the pipe due to pressure and so this thrust must be held elsewhere upstream and 13 COADE Mechanical Engineering News downstream from the untied joint Any supports on the line must be examined to confirm their ability to withstand the pressure thrust load CAESAR II as most any other pipe stress analysis program does not automatically incorpo rate structural analysis of pressure loading It is up to the analyst to confirm that the load is of correct magnitude and applied at an acceptable location on the pipe CAESAR II will apply the pressure thrust load on either end of an untied joint but this is only a good approximation The properly located thrust load may be determined by imagining a position inside the joint the pipe wall seen upstream and downstream from the joint is the proper point to apply the pressure thrust load This point may be beyond a support that was assumed to contain the pressure thrust Expansion joint assemblies add flexibility to a piping sys tem that is certainly understood What may not be so clear is that these assemblies cannot supply flexibility in all directions The requirement to restrain the pressure thrust load must still be satisfied and this requirement will eliminate one or more flexible degrees or freedom from the configura tion One way then of categorizing the various assemblies is to examine the available flexibilities of these
6. in the tube or accordingly the pipe bend as the tube opens up under pressure It is COADE s contention that the Bourdon tube straightens as a conse quence of the flattened tube trying to develop a round cross section so this effect should not really apply to a forged pipe elbow which already has atruly round cross section It may have some validity when discussing bends which have been bent from straight pipe MEC 21 s derivation of the end rotation and displacement strains of a pipe elbow under pressure are shown below p Qe r NA 2 vie amp 1 5 vk R QT a QR Gn e amp c QrR Apes 61C Where a displacement in direction a as per Figure 15 c displacement in direction c as per Figure 15 Figure 15 COADE Mechanical Engineering News The full Bourdon effect considering the rotational effect of the elbows as per MEC 21 was incorporated into this case by entering 2 on the Activate Bourdon Pressure Effects entry under Special Execution Parameters of the Kaux menu NOZZLS5 This model contained more than one of the effects described above pressure stiffening of elbows pressure stiffening of straight pipes and Bourdon effect number 2 This would probably be a better model of reality in that these effects would not occur in isolation NOZZL6 This model was identical to the original case except that stiffnesses of 1E6 Ib in and 1E6 in lb deg were used for the translational an
7. joints Type Freedom Notes UNTIED axial lateral amp bending with low pressures reqs guides and thrust supports TIED lateral compact and stable HINGED bending one axis used in combina tions for lateral flexibility guides recommended GIMBAL bending both axes see above U UNIV lateral stable only for low pressure applications total offset de termined by length of center piece T UNIV lateral total allowable offset is a func tion of center piece length Press Bal axial used with a bend or tee on center piece so it may be con sidered lateral provides axial flexibility without additional pipe supports Expansion joint assemblies then can be selected based on their ability to provide flexibility in specific directions their 14 December 1993 space requirements and their support requirements This article presents a sensible way to address the variety of configurations available to the piping system designer The next issue of Mechanical Engineering News will review individual joint selection and evaluation Estimation of Nozzle Loads Using CAESAR II Software Results of pipe stress analysis or for that matter any structural analysis can vary depending upon the assump tions made when modeling the problem For example the best means of defining boundary conditions loading combi nations and material element behavior is rarely clearly defined For this reason modeling conventions often devel
8. of tie rods running over both joints to contain the pressure thrust in the line The tie rods would also have some means of attachment to the center spool piece to stabilize the entire unit Here again the tie rods eliminate any axial flexibility but permit a great range of transverse move ment through the bending of the two joints The greater the length of the center pipe the greater the transverse deflection with the same amount of joint bending Other configurations of note Swing Expansion Joint Some piping configurations require transverse flexibility in one direction but not the other In these situations a swing expansion joint is recommended A swing joint is similar to a tied univer sal joint in that it has a pair of bellows and acenter spool piece Instead of tie rods the swing joint will have hinged bars restraining the pressure thrust The parallel hinges at either end of the assembly allow bending about only one axis rather than two These joints therefore direct the transverse deflection of the joint along a defined vector perpendicular to the axis of the expan sion joint assembly Pressure balanced Expansion Joint Utilization of axial flexibility in an expansion joint usually requires the joint to be untied and heavily guided Tie rods while containing the pressure thrust forces eliminate axial flexibility ofan expansion joint Another way of keeping axial flexibility without adding extra guides and thrust r
9. 0 03 DR 03 represents the response of a typical piping system The maximum and minimum response values occur at multiples of the first natural period as in the un damped case However as can be seen in the figure the amplitude of the response continuously decays with time This behavior can be shown further by extending the analysis for a duration of 2seconds The resulting plot of the tip displacement in the 11 COADE Mechanical Engineering News Y direction is shown in Figure 11 Time History Amalusis Simple Cantil imple Cantilever er WEA ASM EOD yy yay leas ize 1490 isba isao zapa Tine ms 8 209 400 608 eee Figure 11 In Figure 10 the curve with the clover symbols for a damping ratio of 0 5 DR 0 5 represents a highly damped though still under damped system The oscillations after the first cycle are barely noticeable and in fact die out after 1 5 cycles The curve with the spade symbols for a damping ratio of 1 0 DR 1 0 represents a critically damped system The response for this run shows that the static displacement is attained in one cycle after which there are no further oscillations The curve with the heart symbols for a damping ratio of 2 0 DR 2 0 represents an over damped system As the figure shows the response for an over damped system is not much more than a slow static response These results and a study of Figure 10 show that the behavior of a cant
10. 141 67820 2711 2724 27 15 15 Table 1 December 1993 Considering only models NOZZLE through NOZZL5 some nozzle load components were seen to vary from the mini mum to the maximum by such large numbers as 117 FX 105 FZ and 96 MY Unexpectedly the ranges between the minimum and maximum code stresses were much smaller with differences of only 27 Operating 16 Sustained and 10 Expansion In case NOZZL6 where the nozzle flexibilities are intro duced the results are even more striking Selected compo nents of the maximum tabulated nozzle loads were higher by 141 FZ 67 820 MX 2711 MY and 2724 MZ vs the results with this model Meanwhile the stresses vary by only 27 15 and 15 for the Operating Sustained and Expansion cases respectively Even though this may not be a fully realistic case this effectively represents the impact that modeling of flexibilities may have on nozzle loads without necessarily providing a corresponding effect on the stress calculations Therefore it is clear that nozzle load calculations which should be independent of piping code used and its associat ed modeling conventions may be more sensitive to model ing variations than code stress calculations would be For this reason the engineer may be interested in performing more than one analysis in order to determine the extremes of loading which the equipment may be expected to endure CAESAR II Specification
11. Calc asks the user for the angle between the nozzle centerline and a tangent to the vessel mean radius as also shown in Figures 4and 5 However for hillside nozzles the determination of this angle can be difficult The purpose of this article is to provide a few simple equations that can help the user determine this nozzle angle angle between vessel tangent and nozzle centerline BOLD re F IILI III III NKO COSCO ONECA NE X DIS Figure 4 Y angle Nozzle Angled in longitudinal plane of cylinder The overall goal of this calculation is simple find an angle for which the calculated diameter of the hole will match the actual diameter of the hole The finished diameter is the dimension d in Figures 4 and 5 which is called DER in the CodeCalc print outs If we had the nozzle in front of us and could measure d then we could calculate the input angle very simply using the following equation dn sin d where d inside diameter of nozzle d DLR finished diameter of hole sin A sine of angle between nozzle and vessel COADE Mechanical Engineering News December 1993 Po rl a arccos r Fm mae arccos sidi Fm Alpha 3 L 4 COG ees bxsH zA se see lt x RO 2R z gt OOO eens asis RSs serene eee ERE 2o w SEN sane sexe Se KS 0 scat RLL EES Se SRL sis
12. Guide also provides a complete index listing of each of the three manuals The complete index listing is probably the most effective way to begin a search for specific information needed for the job at hand December 1993 Network Versions of CAESAR IT and CodeCalc in Testing In August the first network versions of CAESAR II and CodeCalc were completed for in house testing These versions are designed to run on Novell s Netware Version 3 11 and utilize a special network ESL These versions are currently being evaluated at COADE and must still undergo a beta test phase at several client locations before they are generally available Users contemplating installing either CAESAR II or CodeCale on their Novell network should consider the following points The network ESL must be attached to the serial port of the file server During the startup of the file server an NLM is loaded to control communications with this ESL and monitor the number of available licenses remaining Theoperation of the software is somewhat slower on the network than on a stand alone PC due to the communi cation protocols of the network Additionally the network s serial ESL is slower than the PC s parallel ESL e Itis recommended that 1 out of every 5 licenses be provided as a local ESL This way the program can be used in the field or other off site locations which don t have access to the network There will be no l
13. Mechanical Engineering News For the Power Petrochemical and Related Industries The COADE Mechanical Engineering News Bulletin is published periodically from the COADE offices in Houston Texas The Bulletin is intended to provide information about software applications and development for Mechanical Engineers serving the power petrochemical and related industries Additionally the Bulletin will serve as the official notification vehicle for software errors discov ered in those Mechanical Engineering programs offered by COADE Please note this bulletin is published only two to three times per year TABLE OF CONTENTS PC Hardware for the Engineering User Part 17 1 What s New at COADE CAESAR II Version 3 2 Features cee 3 Using the New CAESAR II Documentation 3 Network Versions of CAESAR II and CodeCale Wi POSS sosaren Saacecds dears TO 4 Seminar Schedules for 1994 00 00 ceeeeeeceeseeereeeees 5 Technology You Can Use Mechanical Engineering News Article Index 5 API 650 Nozzle Flexibilities ececeeeeeeeeereeee 6 CodeCalc Hillside and Off Angle Nozzle ADP IES os cect eevee ce ea e E EEEE EEEE 7 An Introduction to Time History Analysis 9 Selecting and Evaluating an Expansion Joint ANSSEMDLY 5 seretii er earra aee 12 Estimation of Nozzle Loads Using CAESAR II SOfWATE 5 58 eiaeia eeann EEEa saasa riea 14 CAESAR II Specifications ssseneseseeeeneeeeee
14. S Tank Diameter srissirreneretrr anirai esien lin D 3124 000 Nozzle Height above bottom plate Cin 24 750 Nozzle Outer Diameter sscssraverssrsasaresi Cin d 24 BBB Shell Course Thickness Cin d 1 338 Max Fluid Height above bottom plate drea Linds 768 004 delta Temperature irrena cece eee irri Pos 130 000 Modulus of Elasticity Clb sq in 288 88E 8 Expansion Coefficient Cin Zin deg 667 08E 5 Fluid Specific Gravity 2 005 1 008 Reinforcement on Shell or Nozzle N COMPUTED VALUES AXIAL TRANSLATIONAL STIFFNESS C1b Zin d 114152 LONGITUDINAL BENDING STIFFNESS Cin 1b deg 1382592 CIRCUMFERENTIAL BENDING STIFFNESS Cin 1b degd 2497816 RADIAL DEFLECTION DUE TO FLUID amp TEMP Cin 2 38 LONGITUDINAL ROTATION DUE TO FLUID amp TEMP deg 1 86 MAX ALLOWED RADIAL FORCE C1b d 63476 MAX ALLOWED LONGITUDINAL MOMENT C t 1b 130649 MAX ALLOWED CIRCUMFERENTIAL MOMENT C t 1b 429899 lt C gt TO CONTINUE lt P gt TO PRINT Figure 3 CodeCalc Hillside and Off Angle Nozzle Angles There are two main categories of off angle vessel nozzles those which are off angle in the longitudinal plane of the cylinder Y angle nozzles Figure 4 and those which are off angle in the circumferential plane of a cylinder or in a head Hillside nozzles Figure 5 In order to analyze these kinds December 1993 of nozzles Code
15. a known COADE Mechanical Engineering News radius Code dimension L in Figure 1 4 of Appendix 1 For elliptical heads the spherical portion is taken to be a circle drawn on the head with a diameter of 80 percent of the head diameter The radius of the spherical portion is taken to be 0 90 times the head diameter The nozzle offset from the vessel centerline should be known from the vessel drawings The nozzle can also lie entirely in the knuckle portion of the head Figure 7 c The mean radius r is the mean knuckle radius and the offset L is the distance from the origin of the knuckle radius to the centerline of the nozzle Note that for an elliptical head the knuckle is defined as anything outside acircle drawn on the head with a diameter of 80 percent of the head diameter The knuckle radius is 0 17 times the vessel diameter Se bs SS OOOO OO Knuckle Radus Spherical Radius Spherical Portion Toroidal Portion a iM ZZA IIE EROS EEE area SE lt u SK y Z Spherical Portion Knuekle Radius Spherical Radius a rm Toroidal Portion b LIII LIII re ILIO RRR I S ERS Sp Spherical Portion Knuckle Radius Spherical Radius earl Toroidal Portion c Figure 7 Hillside nozzles in heads December 1993 Finally the nozzle may be located so that part of the nozzle
16. available to the designer and no additional thrust blocks or guides are required A simple tied expansion joint then is installed perpendicular to the plane of required flexibility HINGED Hinged Expansion Joint Tie bars on either side of a single expansion joint may also be hinged With a single joint though the only allowed motion is angulation not axial or transverse deflection and angulation about the one hinge axis only Hinged joints are quite compact and they easily contain the pressure thrust loads In many cases two or even three hinged joints work together to provide needed flexibil ity Hinged joints often require guides to drive the piping into the flexible direction GIMBAL Gimbal Expansion Joint Gimballed joints combine two perpendicular hinges across an expansion joint These bars from either end of the joint hinge off a ring at the center of the joint This articulated joint allows bending about both axes perpendicular to the axis of the joint Gimbal joints are usually used with other gimbal joints or hinged joints and with pipe guides Guides are used to force motion in a line perpendicular to the hinge axes of the joints U UNIV Universal Expansion Joint A universal expansion jointis a double joint without an anchor on the December 1993 T UNIV Universal Tied Expansion Joint As the name suggests these are assemblies that have a univer sal expansion joint configuration with a set
17. command which can be placed in the login script An example capture command is illustrated below CAPTURE L 3 Q que_name J special TI 15 NB Each item in the above command is explained as follows Required to execute a DOS program from inside a login script If this command is manually entered from the keyboard the GP should not be used L 3 Specifies that LPT3 should be captured Q que_name Assigns LPT3 to the queue whose name is que_name Queue names are system depen dent J jcon_name Applies a queue job configuration file whose name is jcon_name to each print job This may be necessary for postscript print ers Job configuration files are usually obtained from the printer manufacturer TI 15 Indicates that the print queue should close a print file after 15 seconds of inactivity By default the Novell print queues do not close a print job file until the application EXE is exited This may cause a problem with some print jobs especially if they contain graphics The time out parameter is used to force the queue file to close Note this was required at COADE for both an HP and a postscript printer This parameter was not necessary fora dot matrix printer Note also that this option should only be used if abso lutely necessary otherwise text reports could print sooner than expected NB Specifies that the printouts are not preceded by abanner page For large networks where many users
18. d rotational stiffness respectively at the compressor connection This is not necessarily a representative value however it is certain that the compres sor connection is not infinitely rigid and this case can be used to give an indication of how much the results may be affected by modeling the compressor stiffness The true measure of compressor rigidity may be estimated through finite element analysis test manufacturer s documentation or other means RESULTS AND CONCLUSION The maximum and minimum values were tabulated for the six components of the compressor nozzle load as well as the maximum Operating Sustained and Expansion stress of any point in the system for the seven versions of the model These results are shown in Table 1 Compressor Nozzle Loads OPE Case Max System Stress Case FX FY FZ MX MY MZ OPE SUS EXP 1b ft 1b psi INOZZLE 1485 7745 45025 9279 220575 GT OS 7679 INOZZL1 1107 7777 13876 30519 201406 2683 9750 7900 3251 INOZZL2 1415 7923 16725 29702 194005 2483 9376 7662 2964 INOZZL3 1734 7751 21974 30320 322520 2687 11029 8050 3090 INOZZL4 2028 7753 25834 30340 380473 2683 11873 8878 3090 INOZZL5S 2398 7927 28396 29734 335800 2475 11605 8746 2964 IMAX MAG 2398 7927 28396 30519 380473 2697 11873 8878 3251 IMIN MAG 1107 7745 13876 29702 194005 2475 9378 7662 2964 VAR 117 2 105 3 96 9 27 16 10 INOZZL6 1859 6424 12094 45 14394 100 9366 7718 2835 VAR 32 23
19. e tip load and all other terms are defined above The value of P is computed as 403 3 pounds The objective of the analysis is to determine if the dynamic response of the beam under the time history matches the expected results i e as predicted by dynamic theory For example e Aload applied instantaneously to anun damped system should yield twice the displacement of the same load applied statically Additionally the system should cycle between plus and minus this displacement forever e The same load applied to an under damped system should show close to the same initial displacement but further cycling decays the displacement magnitude 10 December 1993 The same load applied to a critically damped system should produce a response which goes through one cycle before stabilizing at the static displacement e The same load applied to an over damped system should produce a response which does not cycle at all Rather the displacement slowly attains the static magnitude Entering the CAESAR II dynamic input module and select ing Time History as the analysis type reveals that data in the following input forms must be specified all other input forms are optional for this type of analysis 3 Spectrum Time History Definitions 7 Spectrum Time History Force Sets 9 Spectrum Time History Load Cases B Control Parameters For the purposes of this example since only motion in the vertical plane is of interest
20. ed Equating strain energies Hepa Pap Zp fe a z Rey riore Sa Since 6 N d and 6 G d Where d vector of elemental displacements G first derivative with respect to x of elemental shape function matrix N with on diagonal entries of G 0 G 6x L 6x L G 1 4x L 3x L G 0 G 6x L 6x L s 2x L 3x7 L N elemental shape matrix with on diagonal entries of N 0 N 1 3x7 L 2x L N X 2x7 L x L N 0 N 3x7 L 2x L N x L x L Carrying through the matrix multiplication and the integra tion the additional stiffness due to axial force is found to be December 1993 36 K P 3L 4P 1 30L 000 36 3L 0 36 3L 0 3L 4P Note that the upper triangular portion of the matrix is symmetric about the main diagonal Activating pressure stiffening of straight pipes in CAESAR II simply applies the stress stiffening matrix to the elemental stiffness matrices of straight pipes only using an axial force P equal to the internal pressure user selectable P1 or P2 times the internal area of the pipe Note that other internal forces due to thermal or imposed mechanical loads are not included in the P force this is not a non linear effect This option is activated by entering 1 to use pressure value P1 or 2 for pressure value P2 on the Use Stress Stiffening due to Pressure entry under Special Execution Parameters of the Kaux menu
21. ee xe Z sete Oe seca Oy 3 3 R LLL SIN Sera OOO 8 lt ans xx LS o A SOY Sy v ims Z Sees 3 EERE ES ERER EERE SERENE KKK KEK Deca RR SIII R ee PRESS K RRI ae a pote ZR SKE SON OOS ssi S x S oe s cae KOK Kere where L offset distance cylinder head centerline R een vo see lt x a oes REMY S 2S RS OIT LN IA inside nozzle radius rece Se Il lt lt see IRS SE BOSSES RS 23S Se x s S ORT Oe 3S lt sone QO RX ee sects a x L Offset Distance BQ oc L lt gt L es RLS mean vessel radius R ae X p azs bee be f Figure 5 Hillside Nozzle 1 2 in fillet When we analyze Y angle nozzles the angle is known and the result is exact this is really all the information we need dn sin d DLR However when we analyze hillside nozzles as shown in Figure 5 the angle is usually not known Instead we may know the offset distance for the nozzle This distance L is the distance between the centerline of the cylinder or head and the centerline of the nozzle A first approximation to the angle would take the cosine of the angle as L r where r L 12 in is the mean cylinder or head radius at the point of attachment However this approximation t
22. eling 7 90 11 Global vs Local Coordinate Systems 12 92 3 Expansion Joint Modeler Part 1 5 93 29 Estimation of Nozzle Loads 12 93 14 Hanger Design Discussions Part 1 3 90 4 Hanger Design Discussions Part 2 10 90 7 Large Rotation Rods and Hangers 11 87 9 Plastic Pipe Modeling 4 91 5 Relative Rigid Stiffnesses 11 88 11 Selecting amp Evaluating an Expansion Joint Assembly 12 93 12 Some Nuances of Spring Hanger Design 5 87 7 Spring Hanger Design 10 90 4 Slip Joint Modeling 4 91 9 Tees amp SIFs 3 92 4 Tee Types 3 92 14 Underground Pipe Modeling Philosophies 4 91 8 User Specified Wind Profiles 3 92 14 Subject Quality Assurance Benchmarking CAESAR II amp ANSYS 10 91 3 CAESAR II Quality Assurance Manual 5 93 Software Quality Assurance 10 90 7 nn COADE Mechanical Engineering News Subject Strength of Materials Maximum Shear Stress Intensity 8 92 4 Octahedral Shear Stress 3 87 4 Torispherical Head Equations 12 92 6 API 650 Nozzle Flexibilities Version 3 2 of CAESAR II incorporates nozzle flexibilities according to API 650 API 650 nozzles are sometimes referred to in the literature as low tank nozzles These nozzles are typically located in the bottom shell course of large storage tanks The API 650 computations performed by CAESAR II provide flexibilities at pipe terminations similar to the WRC 297 bulletin These computations are performed in accordance with the 9 edition of API 650 dated July 1993 The 9
23. eseeee 19 PC Hardware amp Systems for the Engineering User Part 17 ESL News It is a little known fact that computers of the same make and model purchased from the same vendor on the same day can be very different internally These differences range from a variety of chips on boards to a variety of boards all Volume 17 December 1993 supposedly performing the same function Recently an ESL External Software Lock access problem involving IBM PS 2 Model 77 machines has been solved For all users of COADE software running on IBM PS 2 Model 77 machines there is anew ESL available which should resolve all access problems The revision necessary to the ESL was made possible through the assistance of M K Ferguson Inc which lent a Model 77 to the ESL manufacturer Software Security Inc for testing purposes The loaner machine was necessary because both COADE and SSI had obtained Model 77s directly from IBM and they accessed the ESL correctly It was only through testing on a problem machine that the issue could be resolved Users experiencing problems with IBM PS 2 Model 77 machines should contact COADE to arrange an ESL upgrade Another parameter to check on PS 2 models is the Arbitration Level of the Parallel Port The following eight steps should be used to insure the arbitration level is disabled 1 Press Crtl Alt Del simultaneously to reboot the machine 2 When the curso
24. esisting anchors is by using pressure balanced expansion joints A way of understanding this axial flexibility is to examine the center spool piece of a universal tied expansion joint This center piece is relatively free to move axially It is not affected by the pressure thrust as that load is carried across the entire joint through the tie rods The center spool piece is resisted only by the axial stiffness of the attached expansion joints If the piping ran out through the center spool piece by placing a tee in the spool piece and the pipe after the tie rods is capped then the axial or lateral flexibility remains Additional hardware is needed for the proper installation on several types of expansion joint assemblies Guides are center spool piece The lack of restraint on the center piece turns it into a linkage between the two joints This linkage assembly converts the joints bending flexibility into large transverse displacements A longer center piece produces greater transverse offsets with the same bending on each joint However the lack of restraints or other pressure containing elements limits the application of this joint to low pressure lines required to drive the piping in a specific direction so that the joint deflects in a controlled and safe manner Pressurized joints without tie rods or hinges require anchors to hold the pressure thrust load In most instances this pressure thrust load is contained by the pipe wall
25. ever beam is a 4 inch standard wall low carbon steel pipe The diameter wall thickness density elastic modulus and fluid density are shown in the figure Notice that the length is 116 544 inches and a point load of 403 3 pounds is applied at the free end These two values are COADE Mechanical Engineering News somewhat contrived so that the results and behavior of the beam are predictable The length was obtained such that the first natural frequency of the beam is exactly 10 Hz which corresponds to a period of 0 1 seconds The first natural frequency of acantilever beam modeled as a series of lumped masses can be calculated as H Agee EO B where g acceleration of gravity 386 088 in sec weight of the pipe and fluid lumped at node i Young s modulus of pipe material moment of inertial of pipe X distance of node i from fixed end i Alternatively one could turn to a reference such as Marks Standard Handbook for Mechanical Engineers 9 Edition Page 5 74 provides the following equation for the first five natural frequencies of a cantilever beam l f c GEI eu hy where w weight per unit length of the beam L length of the beam c coefficients defined in the Table below 1 2 3 4 5 0 56 3 57 9 82 19 2 31 8 Frequency c value Once the length of the beam is known the load required to produce a static deflection of 1 inch was determined from P 1 0 3 E I L3 where P is th
26. ew tutorials extra troubleshooting tips and overall more quality material The User s Manual is the starting point providing general knowledge ofall of the different capabilities of CAESAR II beginning with the installation program If the concern is simply what button to push or what procedure to use the User s Manual provides this general information The questions answered by this manual are more What do I do rather than Why am I doing this To help the first time user to get his her feet wet a Quick Start chapter has been COADE Mechanical Engineering News added to this manual This chapter gives a brief overview of a static analysis beginning with the input screens The manual also provides general model creation information as well as descriptions of each of the available menu items Additionally a Troubleshooting chapter has been added at the end of the document It is our hope that the Troubleshoot ing chapter will continue to expand to accommodate as many commonly asked questions as possible The Applications Guide should be accessed for specific modeling questions This manual answers questions of the nature of How do I do this To answer this question the manual provides single component modeling examples as well as entire example problems In the component examples specific elements such as mitered bends flexible nozzles tied expansion joints etc are addressed In the examples chapter co
27. fy them on the piping input spreadsheet Appendix P also provides equations to compute the maxi mum allowed piping loads These values are also computed by CAESAR II and provided to the user as additional information Once the CAESAR II run has been made the restraint loads computed at the nozzle should be compared to these limits to assure compliance Note the maximum piping loads computed according to API 650 assume only a single load is active i e when evaluating the radial force API 650 assumes the two moments are zero The CAESAR II implementation of API 650 Appendix P is as follows 1 The data base containing the digitized curves from Appendix P is located and made available to the pro gram 2 The stiffness coefficients based on the user s nozzle and tank parameters necessary to access these curves are determined 3 Using these stiffness coefficients and the digitized curves the dimensionless values are obtained Based on these values the modulus of elasticity and the nozzle radius the desired stiffnesses are obtained These stiffnesses are then applied by CAESAR II to the nozzle node as restraint stiffnesses Users should note that these curves from Appendix P are log log curves Not only must logarithmic interpolation be used but a slight variation in the curve values can produce stiffness variations in excess of 50 4 The unrestrained deflection and rotation of the nozzle are determined based on the p
28. he actual damping to the critical damping Values less than 1 0 are indicative of under damped systems while values over 1 0 are indicative of over damped systems and a value of exactly 1 0 indicates a critically damped system Five runs were made with the damping ratio set to 0 0 0 03 0 5 1 0 and 2 0 For each run the Y displacement at the tip of the cantilever node 1100 as a function of time was stored in a data file A plot of this data generated using the 2 D plotting module of CAESAR ID is shown in Figure 10 December 1993 Time History Analysis Results r imple Cantileve TELLE T G 1 25 me e e EE E EEEE EEE ETE AEE TO 28 48 6B 8B 168 128 140 160 1858 288 Time ms Figure 10 The curve for a damping ratio of zero DR 0 0 represents the response of an un damped system The curve shows that the maximum response peaks at just less than 2 inches which is the expected result Additionally the curve shows that a full cycle of response is attained every 100 millisec onds as expected This curve also shows that the vibration of an un damped system does not diminish with time i e the amplitude of the response for each cycle remains constant Also notice that the motion is cycling about the magnitude of the static displacement of 1 0 inches All of these characteristics of the DR 0 0 curve are as expected The curve with the diamond symbols for a damping ratio of
29. ilever beam in a CAESAR II time history analysis is as predicted by theory This article also illustrates the ease with which a time history analysis can be implemented using CAESAR II Selecting and Evaluating an Expansion Joint Assembly The basics of the CAESAR II Expansion Joint Modeler appeared in the last issue of Mechanical Engineering News The CAESAR II Expansion Joint Modeler Volume 16 This article reviewed important model characteristics of the joint and discussed how the program modeled a simple expansion joint The article did not indicate how a particular assembly was selected For example why would one recommend a simple untied expansion joint instead of 12 December 1993 universal tied expansion joint Reviewing the input matrix for the CAESAR II expansion joint modeler one finds there are at least six assemblies available see Fig 12 Each assembly must have its advantages so what are they This article will examine these and other assemblies to highlight their individual uses Pressure 15 25 58 158 388 Untied Tied Hinged Gimbal U UNIU T UMNIU JA45 31688 660Inc 625Inc Style Conu Matl Arrow Keys gt Change Highlight Options lt Enter gt Use Current Selections lt Escape gt Ignore Request DATA BASE PATHWAY DAT gt Figure 12 Pipe is quite rugged High loads and stresses in the piping do not justify the installation of an expansion joint The equipment
30. imited network version of CAESAR II All network versions will be either full purchases or monthly leases Pricing on the network versions has not been finalized however network licenses will be no less than the stand alone licenses There will be a conversion fee for users wishing to switch from stand alone licenses to network licenses COADE Mechanical Engineering News COADE Seminar Schedules The following piping seminars have been scheduled for 1994 in the COADE Houston offices January 19 21 January 24 28 Introduction to Pipe Stress Statics amp Dynamics March 7 11 Statics amp Dynamics May 2 6 Statics amp Dynamics September 21 23 September 26 30 Introduction to Pipe Stress Statics amp Dynamics November 7 11 Statics amp Dynamics The following pressure vessel seminars have been scheduled for 1994 in the COADE Houston offices February 7 9 ASME Sect VII October 17 19 ASME Sect VIII Mechanical Engineering News Article Index At the request of many COADE clients we have compiled the following index of articles from all past issues of Mechanical Engineering News This index is intended to aid clients in finding reference articles quickly Articles on Software Development Seminars and General News have been omitted for brevity Title Issue Page Subject Code Requirements AISC Unity Checks on Pressure Vessel Legs 8 92 8 Expansion Case for Temperatures Below Ambient 5 93 32 Sustained amp E
31. in addition to operating in 32 bit protected mode is now able to handle much larger models and more load cases This output processor also provides a Table of Contents of generated reports and incorporates our standard file manager for directing output re ports to disk Tank Nozzles API 650 Nozzle Flexibilities can be incor porated into a piping model similar to WRC 297 nozzles Data Bases Version 3 2 provides the Australian Structural Steel tables and spring hanger tables from China SINOPEC India BHEL and Italy Flexider Using the New CAESAR II Documentation The new release of CAESAR II includes three new manuals The User s Manual the Applications Guide and the Technical Reference Manual These three manuals repre sent a much more logical organization of the CAESAR II documentation Each guide or manual serves a specific purpose so the user will be able to access the relevant material without having to sift through all of the available documentation Hopefully this will allow for more direct searches for examples explanations etc without having to strain to lift the old manual off of the shelf The preliminary work on the new documentation dealt mainly with develop ing the logical organization that will allow for expansion of the material without simply cluttering chapters For this reason now that the groundwork has been laid we hope to improve the documentation further by including more exam ples n
32. ior by the Xircom adapter effectively hides the ESL from the software rendering it inoperable For users with a Xircom 1 adapter it is possible to utilize an ESL once three requirements are met First a parallel port multiplexer must be obtained from Xircom This device essentially provides two ports one for the network and one for the ESL Second the driver PPX COM also from Xircom must be loaded Third revised software must be obtained from COADE COADE was informed of this software change in early November after the CAESAR II Version 3 20 master disks were generated New versions of CAESAR II and proVESSEL will be available in December to address this software change to address PPX COM A revised CodeCale will be available in January Forusers witha Xircom 2 adapter there is no current work around COADE has been informed by Software Security that Xircom informed all hardware lock vendors that they Xircom will no longer make allowances for or support hardware locks Network News For those users running COADE software on networks several configuration items need to be addressed The first item for consideration is the actual network in use Novell 3 11 is implemented at COADE all testing and network development is targeted toward a Novell network The second item to consider is the setup of the workstation to properly address the printer This requires the use of the December 1993 Novell capture
33. is in the spherical portion and part in the knuckle Figure 7 b In this case the angle at the part of the nozzle in the spherical portion should be calculated as described for Figure 7 a and the angle at the part in the knuckle portion should be calculated as described for Figure 7 c That is calculate the inside angle using the spherical radius of the head and offset from the centerline Calculate the outside angle using the mean radius of the knuckle and the offset from the knuckle origin Use of these equations should yield correct nozzle angles for almost all off angle nozzle configurations An Introduction to Time History Analysis With the release of CAESAR II Version 3 20 users can now perform linear modal time history analysis of dynamic events imposed on piping systems This analysis technique provides a true simulation of the dynamic system response as a function of time due to the imposed load and is usually more accurate than the corresponding response spectrum method This article will introduce this analysis technique through the use of a cantilever beam model as well as demonstrate its accuracy in simulating dynamic response This cantilever beam is described in Figure 8 below 9 elements 1 ft in length 1 element 8 544 inches M 116 544 P E 29 5E6 psi ses 28993 Ib in wate 03611 Ib in 227 P 403 3 Ib 4 5 m Figure 8 This cantil
34. le with two expansion loops in the vertical plane before turning horizontally as it enters a compressor station The system temperature is expected to vary from an ambient of 70 F to a maximum of 100 F The line is connected restrained in six degrees of freedom to a compressor nozzle at node point 360 NOZZLE Figure 13 The initial analysis done under the B31 8 Code using CAESAR II s defaults i e pressure stiffening of elbows absence of pressure strain effects default rigid stiffnesses for all restraints and the anchor to the compressor etc yielded the results shown for the first case labeled NOZZLE in Table 1 The maximum system Operating Sustained and Expansion stresses of 9538 psi 7679 psi and 3090 psi respectively with the maxima occurring at various points along the run are acceptable against the code allowables The nozzle loads on the compressor for the Operating load December 1993 case of 1185 Ib FX 7745 Ib FY 15025 Ib FZ 30279 ft lb MX 220575 ft lb MY and 2697 ft lb MZ which should be checked against the manufacturer s allowables may be acceptable as well As noted above the stress calculation is perfectly acceptable here since the B31 8 allowable stresses take into consider ation the modeling conventions normally associated with this code The nozzle load calculations may be a different matter the analyst should be more interested in finding a less code de
35. lly solution times have been accelerated by an order of magnitude Time History Modal time history analysis capabilities have been added to the dynamic solution and animation modules This provides users with another method of evaluating piping system performance under impulse loads Static Load Cases The internal storage area for the static load case data has been significantly expanded Users will no longer receive errors when attempting to define many complex load cases Printing The direct printer access by CAESAR II has been significantly altered for Version 3 2 Users cannow configure the program to direct printed output text and graphics to any parallel port Additionally the user may define printer control strings thereby allowing CAESAR II to change the number of characters per line and the number of lines per page As an example of the use of this file a beta user has submitted his preferred configuration string reproduced below 27 40 49 48 U 27 40s p 49 50h 49 48 vsb T 27 amp a4948L This string sets the laser jet printer to 12 characters per inch and adjusts the left margin to the right by 10 characters to allow for binding holes December 1993 Another beta user submitted the following string 27 E27 amp 149053 4656 c 56 E 27 amp a 49 56 L 27 amp k 50 S This string sets the laser jet printer to Landscape mode 58 lines per page 16 7 characters per inch Output Processor The output processor
36. load than would the same beam subject to compressive stresses The effect of tensile stresses on bending of a beam can be determined by 15 COADE Mechanical Engineering News examining the element shown in Figure 14 ya Figure 14 Initially assuming the beam in the figure is infinitely stiff in bending so it is straight in any configuration but can deform axially the total axial strain for small lateral dis placements is constant over the cross section and is com puted as E X u Vv Where axial strain due to axial load u u L u axial displacement at node 2 u axial displacement at node 1 L length of beam g axial strain due to rotation of the bar through a small angle O while holding x motion constant 1 cos 8 1 sec 8 1 67 2 v v LP 2 v lateral displacement at node 2 v lateral displacement at node 1 If the beam is capable of bending then the axial strain is no longer constant over the cross section since bending puts the cross section in varying levels of compression and tension based upon the distance to the neutral axis For this case the axial strain is a function of the distance from the neutral axis An additional term must be added representing the rotation of the element face and the term must be modified to take into consideration the fact that it is no longer linear with 16 December 1993 respect to x e u y0
37. mplete analyses such as earthquake relief loads water hammer etc are performed In future releases additional examples will be added as well as several tutorials Preliminarily planned tutorials include those on static analysis dynamic analysis buried pipe and one on the various equipment and components analyzed by CAESAR IL The third manual the Technical Reference Manual con tains technical aspects of both the software itself and the software applications The software aspects discussed are those which we felt would needlessly complicate the discus sions in the User s Manual To begin with there is a complete discussion of each of the elements that make up the CAESAR II configuration file The User s Manual ad dresses the configuration file very briefly whereas the discussion in the Technical Reference Manual is much more thorough Also acomplete CAESAR ILupdate history and a complete listing of all of the installed files can be found inthis manual The Technical Reference Manual should also be accessed for more in depth discussions of specific topics Where the Application Guide attempts to teach by example this manual attempts to provide the significant background material In addition to the above three manuals a Quick Reference Guide will also be distributed This guide provides quick access to commonly needed information such as intersec tion types material lists and code stress equations The Quick Reference
38. n the following items must be specified TIME lt Analysis Type HARMONIC SPECTRUM MODES RANGE TIMEHIST bs lt Max No of Eigenvalues calculated 0 Not used 1 0 lt Closely Spaced Mode Criteria Time History Time Step ms 2 lt Load Duration Time History or DSRSS method sec 0 03 lt Damping Time History or DSRSS ratio of critical 40 lt ZPA Reg Guide 1 60 g s Time History Output Cases The analysis type should be set to TIME The analysis is configured to extract and use the first five natural frequen cies The value for the time step is set such that the time step multiplied by the maximum frequency is less than 0 1 which yields a value of approximately 001 seconds Actually the lowest frequency was used here since it dominates the system displacement response by a wide margin as can be seen by a review of the Participation Factor report The duration is recommended to be at least as long as one full cycle based on the lowest frequency highest period The fundamental period of this beam model is 0 1 seconds so a duration of 0 2 seconds should produce two full cycles The number of requested reporting cases was set at 40 providing reports displacements forces stresses every 5 millisec onds The value for the damping ratio was varied to show not only the effects of damping but to provide data to visualize the response of the beam The damping ratio is the ratio of t
39. ng spring hangers Figure BE400 was incorrectly labeled BE404 This error has been corrected for Version 3 2 Static Output Module Version 3 19 provided access to the intermediate data generated by the error checker flexibility factors expansion coefficients minimum wall thickness etc when an input echo was requested This data was intended to be available when the output was directed to either the printer or a disk file Unfor tunately access to the printer for this data was omitted from this version This oversight has been corrected in Version 3 2 Batch Stream Processor The Batch Stream processor provided in Version 3 19 inadvertently deleted an inter mediate job file upon completion This file the _1 file is necessary for both static and dynamic animation graphics This problem has been corrected in Version 3 2 In CAESAR II Version 3 20 an error has been discov ered in the presentation of the Spring Hanger Data spreadsheet The text data for this spreadsheet is ac quired from the file SCREEN TXT located in the installation directory For this particular spreadsheet the units labels were actually part of the text instead of variable THIS IS A TEXT ERROR ONLY IT DOES NOT AFFECT THE OPERATION OF THE PRO GRAM COADE Engineering Software 12777 Jones Rd Suite 480 Houston Texas 77070 Tel 713 890 4566 Fax 713 890 3301 BBS 713 890 7286 20 5 6 December 1993 Instructions for correcting this pr
40. oblem accompanied the distribution of the disks Alternatively users can down load a correction file F320A ZIP from the COADE bulletin board In CAESAR II Version 3 20 an error has been discov ered in the Equipment program for both the API 610 and API 617 modules This error could cause a pro gram abort and return the user to the main menu Users can obtain a corrected version of the Equipment program by down loading the file F320A ZIP from the COADE bulletin board Main Menu Module A problem has been discovered in the COADE file manager when attempting to copy files If the destination path is left blank i e not specified the program will abort when the copy process is initiated Future versions of the file manager will trap this error condition and report it to the user CAESAR II Technical Note CAESAR II Version 3 20 includes a change to the equation for computing the B31 8 Sustained Allowable Stress Prior to Version 3 20 CAESAR II computed this allowable as 0 75 S F T in accordance with codes prior to the 1990 addendum Version 3 20 utilizes the new code equation 0 75 S This new equation effectively provides an increase in the allowable of at least 28
41. op in order to simplify the decision making procedure or so the engineer need not reinvent the wheel for every new problem In the pipe stress field each piping code normally has modeling conventions associated with it which have been found to produce the best estimate of pipe stresses under loadings normally associated with the industry to which the particular code applies For example the B31 1 Power Piping Code used in an industry where pressures may not get high enough to significantly stiffen pipe does not consider pressure stiffening during the calculation of flexibility factors of elbows Likewise pressure strain effects are usually included only on very long runs of highly pressurized pipes such as might be found on cross country gas transmission pipes covered by the B31 8 Gas Transmission Piping Code Also the different piping codes have their individual conventions for calculating the stress intensification factors of local components The true behavior of the piping system under load is much more complex than that of any linear system solved by a pipe stress program Evidence that the best assumptions for modeling piping systems are not hard and fast is indicated by the diversity of the various code requirements One thing that is certain however is that the means of calculating pipe stresses for each code which vary for every code has been proven to be adequate when compared to the means of calculating the allowable
42. outthe length of the bend the bend end distortion can be modeled similarly to that for uniform thermal growth From Roark and Young s Formulas for Stress and Strain 5th Edition the end displacement and rotation of acurved beam under uniform strain is given by d 2eRsin 2 1 2v Pr Rsin 2 tE o 0 Where d displacement of bend directionally along line through ends R bend radius 6 included angle of bend radians change in angle of bend These pressure strains can be incorporated into the analysis and were for analysis NOZZL3 by entering 1 on the Activate Bourdon Pressure Effects entry under Special Execution Parameters of the Kaux menu NOZZL4 This model is similar to case NOZZL3 de scribed above except that the pressure strain effects have been applied in a slightly different way using the true Bourdon effect A number of sources such as MEC 21 the 18 December 1993 original Mare Island piping flexibility program written in 1959 and Crocker amp King s Piping Handbook 5th Edition pages 4 37 through 4 38 alternatively view pipe bends as behaving similarly to Bourdon tubes hence the popular name Bourdon effect for pressure elongation note that pressure strain of straight pipes is calculated the same under both methods Bourdon tubes are flattened tubes of steel bent into a circle which tend to straighten out when pressurized This intro duces a angular change
43. pendent result This can be done by re analyzing the system with a few different modeling conventions the true nozzle loads probably fall somewhere within the range of the results The results for anumber of modeling conditions are shown in Table 1 The analyses shown in this table were generated by making the following changes to the original model NOZZL1 This model is identical to the original except that pressure stiffening of the elbows is neglected This matches the treatment of elbow flexibilities as endorsed by the B31 1 and other piping codes Pressure stiffening reduc es the flexibility of elbows by dividing the flexibility factor by 1 6 P E C T R 9 Where wy internal pressure psi E modulus of elasticity of pipe material psi v mean radius of pipe in T wall thickness of pipe in R bend radius of elbow in For this application the flexibility factor is reduced by a factor of 1 128 Pressure stiffening of elbows can be activated de activated or used as per the specific requirements of the code in use by setting a parameter in the CAESAR II configuration file Option 9 of the Main Menu NOZZL2 This model is identical to the original except that pressure stiffening of straight pipes is included This is the phenomenon that occurs when the sag of a guy wire is reduced as itis tightened likewise a beam highly stressed in tension will deflect less under a lateral
44. print jobs banner pages may be desired to separate jobs and indicate owner ship The third item to consider is the size of the hard disk on the file server The current file manager contained in all COADE software packages has a limit of 250 files in any given directory and a limit of 155 directories on the disk partition The current file manager was not designed for 1 2 Gbyte disks configured as a single partition Additionally the list of valid disk drives assumes they are contiguous COADE Mechanical Engineering News CAESAR II Version 3 2 Features Documentation One of the major components of the 3 2 release is the rewrite of the documentation The new documentation is presented in three manuals to aid users in finding the desired information Each manual User s Manual Applications Guide Tech nical Reference Manual contains an introduction stating the purpose and intended use of the remain ing chapters Solvers Version 3 2 provides users with static solvers both in core and out of core converted to 32 bit opera tions running inextended memory Most of thejobs that previously required the out of core solver will nowrun in core The out of core solver is provided for situations where a large job must be analyzed on a machine with insufficient extended memory The static output processor has been similarly converted These changes mean that many of the past memory allocation problems have been eliminated Additiona
45. r moves to the upper right corner of the screen press Crtl Alt Ins simultaneously This will display a menu 3 From this Main Menu select Item 3 Set Configuration This will produce another menu 4 From this menu select Item 2 Change Configuration This produces a configuration screen 5 From the configuration screen choose Parallel Port Arbitration Level 6 Press F5 as many times as necessary until the setting Disabled is reached 7 Press F10 to save the configuration 8 Exit this setup program by pressing Enter F3 F3 and Enter COADE Mechanical Engineering News While this is an involved process some users have been able to access their ESL once the above steps have been implemented At this time the only other machine with known ESL access problems is the DELL 486 Unfortunately DELL has been less than cooperative so this problem will take some time to resolve However for those users with DELL 486s the following command placed in the AUTOEXEC BAT file may help the situation SET SSI_ACT 40 40 40 Additionally the updated ESL for the PS 2 problem has worked on several problem DELL machines DELL users who want to swap their ESLs should also contact COADE Users attempting to connect notebook computers to networks are typically using adapters from Xircom Unfortunately this adapter connects to and takes over the parallel port where the ESL is attached This behav
46. roduct head and thermal expansion As previously noted these values may be used as pre defined displacements if the user so desires CAESAR II will not automatically apply these values as nodal displacements 5 Finally the limiting piping loads forces and moments are determined These values are also obtained through the use of interpolation of the digitized API curves COADE Mechanical Engineering News The input screen necessary to describe an API 650 nozzle is shown in Figure 2 below This data is very similar to that required for a WRC 297 nozzle Nozzle Node Number API 65 Tank Node Number ccc cece cree cece teenies 114 BAB Nozzle Outside Diameter Nozzle Wall Thickness ccc cece een eene Tank Outside Diameter ccc cece eee ee eene Tank Wall Thickness cee cece eee e eens Reinforcing on 1 Shell or 2 Nozzle Nozzle hengnt 6 4 5 sis2sysigtert a epee an fash sua Si sie gy Aa a ses Fruta et Gh gece 5 ane mia ee ciateis PNRI NASOSNI SyS Fluid Specif 1G Gravity 60 09 ars8 s 5 sess soo we hale Wea es Tank Expansion Coefficient Delta Temperature Tank Modulus of Elasticity Figure 2 The results of the API 650 computations are shown in Figure 3 Note that the only quantities used by CAESAR II in the flexibility analysis are the three stiffness values All other data is presented for information only API 65 NOZZLE DATA FOR NODE 110 INPUT PARAMETER
47. s Listed below are those bugs errors omissions in the CAESAR II program that have been identified since the last newsletter These items are listed in two classes Class 1 errors are problems or anomalies that might lead to the generation of erroneous results Class 2 errors are general problems that may result in confusion or an abort condition but do not cause erroneous results Class 1 1 Piping Element Generator A restraint direction cosine tolerance error has been discovered in Version 3 19 This error only affects skewed restraints with direction cosines greater than 0 9998 angles less than 1 146 degrees The original tolerances were set at 0 999 2 56 degrees and at user requests were changed in Version 3 19 to 0 99999 0 256 degrees The limits in the Element Generator of 0 9998 were overlooked This error has been corrected in Version 3 2 19 COADE Mechanical Engineering News 2 Piping Error Checker An error has been discovered in the routine which combines models for the large job include This error caused some of the allowable stress data on the first element of the included job to be lost This error has been corrected in Version 3 2 This error only affects the first element of the included job if allowable stresses different from the preceding job were entered Class 2 1 2 3 4 Static Output Module A labeling error was discovered with regard to the Basic Engineeri
48. stresses for the same code which also vary from code to code However the conventions chosen as best for stress calcula tion purposes may not provide the best means of maximizing the accuracy of other results displacements reactions etc which are not addressed in as much detail by the codes Nozzle loadings also vary by code or more accurate ly vary by modeling conventions which may vary by code COADE Mechanical Engineering News whereas the allowables do not since they are set by the equipment manufacturers or according to standards of third parties such as the American Petroleum Institute Therefore it is important to get as accurate an estimate as possible of nozzle loads independent of the particular code s modeling conventions A single analysis conducted for code compli ance purposes may show that the nozzle loads are well within the equipment s nozzle allowables However altering the modeling conventions may cause the loadings to vary significantly in fact far more than do the maximum system stresses In some cases it may be advisable to do several analyses with different modeling assumptions in order to determine the potential range of results Consider the system of which a portion is shown in Figure 13 This system represents a 42 diameter 1 1 2 wall gas transmission pipe with an internal pressure of 1000 psig The pipeline originates at an anchor not shown and runs approximately one mi
49. the 1 Lumped Mass option was used to zero the masses in the X and Z directions For the 3 Time History Definition the impulse is defined as CANT TIME FORCE LINEAR LINEAR which states that the impulse name is CANT it is a time verse force curve and linear interpolation should be used for both axes Only two data points are necessary to describe this impulse point 1 is at 0 0 1 0 and point 2 is at 3000 0 1 0 This defines a straight line instantaneous impulse from 0 0 to 3000 milliseconds with a multiplier of 1 0 Alternatively the force of 403 3 pounds could been entered here This impulse is shown in Figure 9 gt Multiplier Time ms 3000 Figure 9 For the 7 Time History Force Set the magnitude and direction of the impulse are specified as 403 3 Y 1100 1 which states that a load of 403 3 pounds is applied in the Y COADE Mechanical Engineering News direction at node 1100 and is associated with force set number 1 For the 9 Time History Load Cases the force set is associated with the impulse as follows CANT 1 Y 1 which states that the impulse curve CANT is included in the load case with a scale factor of 1 0 in the Y direction and it acts on the load from force set 1 The total load applied to the system is the product of the force multiplier times the force set load times the load case scale factor For the B Control Parameters scree
50. to which the piping is attached is another story In many cases the piping attached to rotating equipment may be loaded to only 5 of its allowable so that the pump com pressor or turbine loads do not exceed their allowable limits This low load limit must be handled in both the cold and hot piping positions Adding or adjusting supports should reduce the cold loads on the equipment but the change between the hot and cold loads is a function of the thermal loads and the piping flexibility adding flexibility is the purpose of the expansion joint How much room is available for the joint and what sort of load must be decreased determines what expansion joint configuration should be used Clearly identifying the force or moment to be reduced is the first step in specifying an expansion joint assembly What type of joint arrangement should be used To quickly review the different expansion joint assemblies just page through any expansion joint manufacturer s catalog Ona more generic level one would refer to A Practical Guide to Expansion Joints by the Expansion Joint Manufacturers Association Inc 25 North Broadway Tarrytown NY 10591 This booklet describes the parts of an expansion joint how to design a system containing expansion joints and recommendations on proper installation and handling Closer to home the CAESAR II expansion joint modeler offers the following e UNTIED Single Expansion Joint The simplest assembly
51. urns out to be too inaccurate for normal use Figure 6 ASME Code Figure L 7 7 The ASME Code has a sample problem L 7 7 which shows These three equations can be used without any further what their preferred method is They do not explicitly information for any hillside nozzle in a cylinder However address this off angle problem in the body of the Code youneedto apply them carefully to hillside nozzles in heads Figure 6 taken from ASME Section VIII Division 1 When a hillside nozzle is in an elliptical or torispherical Addenda A92 page 512 shows this sample problem The head the nozzle may be located in the spherical portion of key to their approach is the calculation of two angles and the head the toroidal portion of the head or it may straddle a and then the calculation of the finished diameter from the the two portions This is shown in Figure 7 Each of these difference between these two angles You can follow their Cases requires a slightly different L and r to be used in the calculation on page 512 and 513 of the Code For our equations purposes we do not need to carry the calculation that far The angle we are looking for is just the average of the two When the nozzle lies entirely within the spherical portion of Code angles calculated as the head Figure 7 a L is simply the offset from the head centerline and r is the spherical radius of the head For spherical or torispherical heads this should be
52. xpansion Stress Cases 5 88 4 Sustained amp Expansion Follow Up 11 88 10 Subject Dynamics An Introduction to Time History Analysis 12 93 9 Dynamics Basics 11 87 3 Dynamics Damped Harmonic Motion 4 89 7 Dynamics The Range Check 11 88 4 Dynamic Questions amp Answers 7 90 8 Missing Mass Correction in Spectral Analysis 5 93 8 December 1993 Subject Hardware ESL s and Multiple Computers 8 92 2 Machine Times 5 88 2 Memory Requirements 5 88 2 Virus Infections 7 90 1 Virus Update 10 90 2 Subject General Information API 650 Nozzle Flexibilities 12 93 6 ASME B31G Criteria 5 93 271 ASME External Pressure Chart Name Changes 12 92 6 CodeCalc Hillside amp Off Angle Nozzles 12 93 7 FE Pipe CAESAR II Transfer Line Study 3 92 6 FE Pipe CAESAR II SIFs amp Flexibilities 12 92 8 Finite Elements in Practice 3 92 16 Flange Allowable Stresses 10 91 6 Flange Leakage 10 91 3 Flange Stresses 12 92 7 Incorrect Results From Piping Analysis 11 88 8 Numerical Sensitivity Checks 11 87 9 Static amp Dynamic Analysis of High Pressure Systems 3 90 6 What Makes Piping Finite Element Jobs Big 5 87 2 Subject Life Extension amp Failures Evaluation of Creep Stresses 5 93 18 Evaluation of Fatigue Stresses 12 92 12 Piping Failure Caused by Elastic Follow Up 8 92 10 Subject Modeling Bend Elastic Models 3 87 3 Buried Pipe Analysis 4 89 3 Buried Pipe The Overburden Compaction Multiplier 3 92 3 Cold Spring Discussion 10 90 11 Double Rod Mod
Download Pdf Manuals
Related Search