Mechanical Engineering News
Contents
1. November 1994 article Expansion Case For Temperatures Below Ambient published in Mechanical Engineering News May 1993 By examining a pipe resting on a pipe rack this variance can be seen In the sustained cold condition the pipe is resting on the rack and there is a deadweight load imposed on the rack In order for the pipe to move off of the rack the forces due to expansion must first overcome the deadweight load Therefore the loading on the pipe will be the difference between the deadweight load and the forces due to the expansion ofthe pipe If instead of reviewing the difference the user looked at the temperature case only the deadweight effects would be ignored In order to adhere to one code s definition of secondary loadings loadings due to expansion weight stress variation differential settlement movements in the supports etc these deadweight effects must be considered 3 Ihave entered wind uniform loads in the job but my results have not changed NO OCCASIONAL LOADS ARE INCLUDED IN THE CAESARIT RECOMMENDEDLOAD CASES Although COADE has suggested guidelines Chapter 5 of the User s Manual on setting up load cases for occasional loads CAESAR II does not recommend them automatically It is the user s responsibility to edit the recommended or existing load cases in order to include these occasional effects in the analysis Additionally if wind has been included in the input the user must define t2. The Force Spectrum solutions offered by CAESAR II for simulation of relief valve water hammer slug flow etc use a frequency domain model based on a Dynamic Load Factor DLF vs natural frequency response spectrum Dynamic Load Factor is defined as the ratio of the system response i e deflection force stress restraint load during dynamic application of a load to the system response which would have resulted from the static application of the same load The DLF response spectrum for a given load is generated by solving the dynamic equation of motion for a single degree of freedom system Ma t Cv t Kx t F t Where M mass of system slug a t acceleration of system as a function of time in sec C damping of system slug sec or Ib sec in v t velocity of system as a function of time in sec K stiffness of system Ib in x t displacement of system as a function of time in F t applied force as a function of time Ib COADE Mechanical Engineering News For an idealized system where the system damping is near zero and the loading takes the form of an instantaneously applied constant force of infinite duration the system displacement solution is x t F K 1 cos t Where x t system displacement as a function of time in F magnitude of applied force Ib system angular natural frequency rad sec K M t time sec Since the maximum static deflection of the syste
3. with a stiffness of 10 Ibf in but pump loads are high when these rack supports are modeled with structural steel elements the model is sensitive to the support stiffness and accurate restraint stiffnesses are then important in this design The figure below exemplifies this process The three lines signify the limits on typical pump piping systems The area bounded by these limits indicates safe and reliable operation while results falling outside are not acceptable If a user is uncertain about some aspect of the model a second job is analyzed which changes that aspect Ifthe results move from 1 to 2 the uncertainty has no greatimpact If however the results move from 1 to 3 than the user should take the time to research the issue and build a more accurate model This sort of investigation was not feasible with earlier analysis methods due to the inherent inaccuracies or because of the large commitment in time COADE Mechanical Engineering News Figure 1 This sensitivity study serves several purposes First of all it can identify critical design parameters in the model analysis and in the actual layout A sensitivity study is used to find significant design modifications and study their impact on the piping system Another advantage to this approach is the investigation into the variability between the as designed system and as built system Troubleshooting possible variations in installation requires
4. 8 October 16 18 COADE Mechanical Engineering News German Language Files Available for CAESAR II As of November 1994 the language text files utilized by CAESAR II have been translated into German These new language files provide German users the same presentation abilities as has been available to Spanish and French users The German language files are available for download from the COADE BBS as GERMAN ZIP in the CAESAR II file area This file has also been forwarded to the COADE dealer in Germany For users running COADE software from a DOS box under Windows NT ESL drivers are available for download from the BBS The necessary file is WIN_NT ZIP located in the Miscellaneous file area This file contains drivers and instruction files for all ESLs supported by COADE software These drivers will be part of subsequent software releases Caution When Moving COADE Programs Most users eventually have the need to move a software program either from one machine to another or from one disk drive to another Up until the Summer 94 releases this was a rather simple affair for COADE products However as of the Summer 94 releases there are two new traits of COADE software products that may hamper the simple copy from here to there scenario First the products are no longer managed by a COM program The old COM loader manager has been replaced by a more powerful EXE loader manager What does this mean to
5. Check of Condition F 1 2 3 Requirement Status 1 5 FRSt2 FRDt2 5640 gt 3910 FRCa Passed 2 0 MZSt2 MZDt2 6200 gt 4338 MZCa Passed 1 5 MRSt2 MRDt2 12750 gt 12748 MRCa Passed Overall Pump Status Passed e Figure 4 With the calculated pump loads so close to their limits it is wise to take a much closer look at the model to confirm it is correct Any questionable input items could be tweaked to examine their impact on the results If the results do not suffer no additional investigation into these model details is necessary If the system is built as it was designed the pump should operate without difficulty To illustrate an additional benefit of this analysis four variations on this design will be used to determine how sensitive the pump is to the hanger load and position As itis initially designed the hanger at node 125 is placed 7 feet from the discharge riser It is a Grinnell COADE Mechanical Engineering News Figure B 286 mid range spring with a spring rate of 260 Ibf in The spring s load is set to 1200 lbf when the system is in operation Since the spring moves 1 3 inch to its operating position the installation load on the spring is set to 1286 lbf Four additional analyses will be made to test the significance of common installation variations All four involve the spring and are very simple to run through the program The changes are e Set the spring load to reach on
6. Dynamic Load Factors and impulse loading e The magnitude and shape of the DLF curve is independent of the magnitude of the applied force but dependent on the shape of the normalized load profile Therefore the CAESAR II user may specify the actual load or a normalized value of 1 0 as the pulse magnitude e Iftherise time is much greater than the natural period of the system the system response approaches the static response to the applied load the dynamic effects and thus the DLF are negligible 11 COADE Mechanical Engineering News e The shorter the rise time of the load profile is the greater the DLF will be If the rise time for a load of infinite duration is less than approximately one quarter of the natural period the response approximates that for an instantaneously applied load i e the DLF is 2 0 e As the duration of the load increases peak dynamic response higher DLFs will shift toward the systems with longer natural periods smaller natural frequencies Commonly Asked CAESAR IT Questions By Tim Curington The following begins a Question amp Answer series that will be continued in subsequent newsletters In addition these and other Q amp A entries can be found in the subsequent editions of the program documentation and on the COADE BBS The series is intended to provide an additional reference source from which to obtain answers to CAESAR II questions 1 Whyarethe allowable stress values z
7. angle theta is 150 degrees This results in the following matrix equation based on the forces and moments from the Global Force Moment report fx 866 302 5 564 20 5 fy 5 302 866 564 639 4 mx 866 57 5 3 47 9 my 5 57 866 3 31 1 Note that for this system the Global Y direction corresponds to the Local z direction and the Global MY corresponds to the Local mz The resulting local forces and moments agree with those reported in the Local Force Moment report These forces and moments represent the loads acting on the end of the pipe element To use these values in the WRC 107 analysis the signs must be reversed since the forces on the vessel are needed More information on local global transformations can be found in most graphics texts and in texts on matrix solutions to engineering problems The COADE BBS has recently been upgraded to the latest version of PCBOARD Suggestions from the users of the BBS are welcome Users desiring additional information about using this BBS can download the file BBS_INFO ZIP from the Information file area November 1994 Fine Tuning amp Sensitivity Studies Added Benefits of Piping System Analysis By David Diehl With today s software based analysis tools so quick and convenient their application has been expanded to include what if analysis For example what
8. is corrected in Version 1 10C CodeCalc Specifications Listed below are those bugs errors omissions in the CodeCale 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 The UCS 66 1 MDMT reduction was off for vessels whose required thicknesses were between 4 and 5 of the actual thickness This problem was corrected in Version 5 30A and was shipped to all users 2 The piping materials used by the Pipe amp Pad program were updated to the latest edition of the B31 1 piping Code These were updated in 5 30A COADE Mechanical Engineering News 3 4 The tubeside corrosion allowance was not being added to the flange thickness in the Floating Heat program This was corrected in 5 30A The allowable stresses in the summary of shellside pressure for primary type stresses were modified in the Thick Joint program The computation for the factor theta B at X YB was also corrected These were corrected in 5 30A Class 2 1 2 A file sharing conflict was discovered which kept more than one user from using the input processor at the same time on Pathworks Networks This problem was resolved in5 30A The occasional load factor was not being use
9. memory The new SYSCHK checks to see if Windows is loaded and if so what version is running CAESAR II celebrates 10 years of industry leadership this December e The new SYSCHK checks to see if a network redirector is running e The new SYSCHK also checks to see if disk caching or disk compression have been implemented COADE Mechanical Engineering News The main purpose of SYSCHK is to provide a concise summary of the machine environment as an aid to the COADE support staff The machine environment can be altered by modifying the CONFIG SYS and AUTOEXEC BAT files used during system startup What are some of the more important items on the main SYSCHK screen First and foremost is the presence of a math coprocessor All COADE products require the math chip shown on the right side of the screen near the middle This means that 486 SX processors can not be used by COADE software These chips have the math coprocessor disabled The second most important item is the amount of low free DOS RAM The User s Manual for each COADE program specifies the amount of free RAM required torun the software Attempting to run the software with less than this amount results in an abort condition A related item necessary for CAESAR Ilis the amount of free extendedmemory Version 3 21 of CAESAR II requires 2 3 Mbytes of free extended memory to run the input processor Another important item on the SYSCHK screen is the se
10. the user By default DOS looks for COM files before EXE files Therefore if one of the products is moved to a location where DOS finds an old COM before the new EXE the program will notrun Instead the user will receive an error message that the Main Menu module could not be loaded Using INSTALL instead of a COPY procedure eliminates this problem since INSTALL cleans up the installation directory This switch from a COM to a EXE can also cause the same error to occur if more than one version of the software exists on the computer DOS will find and execute whichever loader it finds first when it walks down the PATH Users with multiple versions of the same product on the same machine must adjust the AUTOEXEC BAT file to correctly run the programs 4 November 1994 The second change in the Summer 94 releases is that all products now reference a SYSTEM subdirectory beneath the installation directory for certain data files which may be subject to alteration by the user These files are now located in SYSTEM to allow network installations the option of write protecting the program directory If the software is moved the SYSTEM subdirectory must be moved also in order for the software to find the necessary data files Notes on Insulation Densities Calcium Silicate As most users know CAESAR II allows the specification of insulation density as an elemental property The CAESAR II help facility offers suggested v
11. Mechanical Engineering Nens 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 19 What s New at COADE TANK Enhancements CAESAR II Version 3 21 Released in July Seminar Schedules for 1995 German Language Files Available for Technology You Can Use Caution When Moving COADE Programs Notes on the Density of Calcium Silicate Insulation Local Coordinate Systems Revisited Fine Tuning amp Sensitivity Studies Added Benefits of Piping System Analysis The CAESAR II Pulse Table Generator Commonly Asked CAESAR II Questions CAESAR II Specifications TANK Specifications CodeCalc Specifications PC Hardware amp Systems for the Engineering User Part 19 Summer 1994 brought about a new release of all COADE software products In addition to supporting a local ESL from a new vendor these releases also support a network ESL Additional
12. alues of insulation density based on the type of insulation If the value of insulation density is not specified by the user CAESAR II will default to calcium silicate and assume a density of 11 Ib ft Recently a user pointed out that ASTM 533 states that the density of calcium silicate insulation is 15 Ib ft This statement is in fact true However several other references were checked and the following density values for calcium silicate obtained Source Density Grinnell Catalog 11 lb ft The Piping Guide 11 lb ft Intro to Pipe Stress Analysis 11 Ib ft Users should be aware of the default data used for engineering computations its origin and other possible values Local Coordinate Systems Revisited By Richard Ay The December 1992 issue of Mechanical Engineering News contains an article discussing the Global versus Local coordinate systems implemented in CAESAR II This article explains what each coordinate system represents and how they can be determined Many users have requested additional information on this subject especially for bends and skewed sections of a piping model This article is intended to provide this information The figure below shows a small piping system with the local coordinate system for each elbow sketched near the element The local element coordinate system for an elbow can be determined as follows local x is directed along the incoming COADE Mechanical Engineering Ne
13. anticipation of possible differences but when uncovered these critical construction guidelines can be established before the line is built rather than when the line is shut down for repair Finally this sort of study can actually simplify the analysis by revealing model details which prove to have no impact on results a simplification that saves modeling time and effort and improves comprehension Again this sort of analysis technique does not replace experience but works well with design experience to develop a feasible and reliable layout In many ways a sensitivity study is a cost effective check on piping systems with their tightload restrictions around rotating equipment November 1994 An Example RELY ize 125 6 a ite 105 X Li EA Tih Z x Figure 2 To illustrate the concepts promoted here and to provide an example of this sensitivity study a pump system will be analyzed using CAESAR II The system shown in Figure 2 illustrates the piping around the 10 inch suction and 8 inch discharge lines Figure 3 shows the entire 26 element system with node numbers and several of the supports The suction side runs from node 5 to node 70 the pump is constructed of rigid elements 70 to 90 and the discharge pipe carries on from node 90 to node 145 The boundary conditions are set at nodes 5 and 145 with a known operating position of node 5 and an
14. ave been generated using the CAESAR II Pulse Table DLF Spectrum Generator and are shown below demonstrating the accuracy of CAESAR IT s algorithm 2 0 i a K a 1 6 7 a E ial H 5 12 i a 4 Fin mem M Ta A 0 8 a a pt td 0 4 saat a Le m 0 0 Ee es 0 0 0 1 1 0 10 0 td T November 1994 2 0 18 a 16 tr T oO 14 1 2 1 0 0 0 1 0 2 0 3 0 4 0 tr T The ASME B31 1 Power Piping Code in Section 3 5 1 3 of Appendix II Nonmandatory Rules for the Design of Safety Valve Installations presents a discussion on dynamic load factors including a DLF curve adapted from the Biggs text Here the Code uses t the opening time of the relief valve to represent the rise time of the loading The figure below demonstrates how the ASME Code curve which is based upon an infinite load duration envelopes the DLF curves generated by CAESAR II for various finite load durations Comparison of Dynamic Load Factors 2 0 1 T s q 18 yo 1 6 Long Duration 14 12 K u F 1 0 ae Cost 7o a 06 0 4 Short Duration 0 2 0 0 py HH 0 1 1 0 10 0 t0 T In general one can reach the following conclusions about
15. causes some input data to be omitted from the listing report forjobs thatexceed roughly 1000 elements This was corrected in 3 21a 11 Documentation Applications Guide Pages 3 93 through 3 96 should have been reprinted for the 3 21 update This is necessary since the echo of two elements moved from 3 92 to 3 93 TANK Specifications Listed below are those bugs errors omissions in the TANK 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 Appendix E Seismic Computations An error has been discovered in the Appendix E seismic calculations regarding the usage of the percentage of roof weight supported by the shell This user specified percentage value entered on the Roof Details Spreadsheet should have been divided by 100 This error exists in Versions 1 00 and 1 10 of TANK The error was corrected and Version 1 10A was shipped to all users 2 Appendix E Seismic Computations An error has been discovered in the Appendix E seismic calculations in the determination of the required anchor bolt size The bolts were oversized by a factor of approximately SQRT pi 4 This error exists in Version 1 10 and was corrected in Version 1 10C and
16. d for computations involving angles in the Leg amp Lug program This was corrected in 5 30A November 1994 15 COADE Mechanical Engineering News November 1994 Compuserve Access to COADE Due to the increased usage of our Bulletin Board BBS and the needs of our overseas users COADE is establishing both a forum and a mail address on Compuserve This service will be available to upload or download files or to post general questions and answers Users will need to have their own Compuserve account to take advantage of this service In the U S this costs 8 95 per month plus any service charges Most users will be able to access Compuserve via a local telephone call The COADE mail address is 73073 362 COADE Engineering Software 12777 Jones Rd Suite 480 Houston Texas 77070 Tel 713 890 4566 Fax 713 890 3301 BBS 713 890 7286 16
17. do not cause erroneous results Class 1 1 Hanger Design Modules An error has been discovered in the computation of the Actual Installed Load on Variable Spring Hangers when more than one hanger is specified at the location This error was an output error only and involved the application of the spring rate of the total installation to each individual spring This did not cause an error in spring selection calculation of hot or cold load or subsequent load case results This error exists in all versions of CAESAR II prior to Version 3 21 and was corrected in Version 3 21a which was sent to all current users 2 Refractory Lined Pipe An oversight was discovered in the element generator when generating the mass matrix for dynamic jobs The presence of refractory lining was not considered This error exists in Versions 3 19 3 20 and 3 21 and is corrected in Version 3 21a Class 2 1 Analysis Setup Module A file management error has been discovered in the setup of the wind loading data for structural only jobs For these jobs if an attempt was made to adjust the ASCE 7 wind data the program aborted back to DOS This error exists only in Versions 3 19 and 3 20 This error was corrected for the Version 3 21 release 2 WRC 107 Module A units conversion problem was discovered in the WRC 107 module which was activated if the user made an input error When the program detected the input error it returned c
18. ects the definition of the pump along with the nozzle loads to check the Appendix F criteria The API 610 report shown in Figure 4 confirms that the 10 inch end suction nozzle and the 8 inch top discharge nozzle meet the Appendix F criteria the pump is OK November 1994 API Standard 610 1989 7th Edition File RELY Date SEP 1 1994 User Entered Description Time 11 42 am AS DESIGNED Node Orientation Nominal Diameter Suction Nozzle 70 End 10 Discharge Nozzle 90 Top 8 Table 2 Allowable ratio 2 Pump Axis is in the X direction Suction Table 2 Force amp Moment Status Values Ratios X Distance 10 5 in Y Distance O in Z Distance O in X Force 1522 0 lb 1500 1 01 Passed Y Force 1931 0 Ibs 1000 13 93 Passed Z Force 674 0 Ip 1200 lt 56 Passed X Moment 4270 0 ft lb 3700 Lalo Passed Y Moment 5152 0 ft lb 2800 1 84 Passed Z Moment 3110 0 ft 1lb 1800 1 73 Passed Discharge Table 2 Force amp Moment Status Values Ratios X Distance 0 in Y Distance 15 0 in Z Distance 127 3 in X Force 313 0 lb 850 37 Passed Y Force 1499 0 lb 1100 1 36 Passed Z Force 279 0 lb 700 40 Passed X Moment 3596 0 ft 1b 2600 1 38 Passed Y Moment 2099 0 ft 1b 1900 dg Passed Z Moment 2526 0 ft 1b 1300 1 94 Passed Check of Condition F 1 2 2 Requirement Status FRSa 1 5FRSt2 MRSa 1 5MRSt2 1 756 lt or 2 00 Passed FRDa 1 5FRDt2 MRDa 1 5MRDt2 1 593 lt or 2 00 Passed
19. ero for the operating condition Operating stresses are not considered by the Power and Petrochemical Codes Therefore when reviewing operating stresses the user will first be confronted by the message that states NO CODE STRESS CHECK PROCESSED and then the user will notice that all of the allowable stress values are set to zero This is not the result of incorrect data input but rather a result of whether the individual Code being used reviews operating stresses or not 2 Whatis the difference in the expansion case D1 D2 and just running a T case The expansion case is described in the Codes as the differential between extreme conditions of the piping system e normally cold and operating cases If for example the operating load case is W P T and the sustained cold load case is W P by subtracting the sustained loads from the operating loads we are left with the temperature effects T So are the results acquired from the difference of the two loads the same as just running a T case In the event of a completely linear system the answer would be yes If however there are any nonlinear effects in the system friction gaps single directional supports etc these two results could vary In the latest addenda to B31 3 Interpretation 12 06 acknowledges that the maximum and minimum operating temperatures should be considered as one of the ranges in computing the expansion case This is in agreement with the 12
20. error the proper key to use is Shift On page 6 20 the Expansion case is defined as D1 D2 This is incorrect it should be D3 D4 7 Piping Error Checker Two errors have been discovered in the piping error check module The first error occurs when the INCLUDE feature is used with the N no option and the second intersection field of the SIF auxiliary field is used The node number increment is not applied to this second intersection node resulting in a fatal error This error exists in all CAESAR II 3 x versions The second error limits the number of intersections to 200 This limit should have been removed in Version 3 21 This was corrected in 3 21a 8 Piping Input Module An error has been discovered in the input module when attempting to use Stainless Steel pipe schedules The addition of the half pipe sizes in Version 3 21 caused the access routine to overlook the stainless thicknesses resulting in an error notification during input This was corrected in 3 21a 13 COADE Mechanical Engineering News 9 Piping Input Module A change has been made in the ordering of the bend node numbers during the node numberincrement Versions prior to 3 21 A incremented the auxiliary bend nodes backwards causing the new nodes to be ordered incorrectly This was corrected in 3 21a 10 Input Listing Module A memory management error has been discovered Version 3 21 in the Input Listing Module This error
21. gure are the Global and Local Force reports for this element for the Operating case and the Restraint report for node 200 November 1994 iia 168 140 RESTRAINT REPORT Loads on Restraints CASE 3 OPE W T1 P1 FOR Forces lb Moments ft lb NODE FX FY FZ MX MY MZ TYPE 200 302 234 564 57 6352 a Rigid ANC FORCE STRESS REPORT Forces on Elements CASE 3 OPE W T1 P1 FOR DATA Forces 1lb Moments ft 1lb POINT FX FY FZ MX MY MZ 190 302 87 564 345 3153 693 200 302 233 564 56 6351 3 LOCAL FORCE REPORT Forces on Elements CASE 3 OPE W T1 P1 FOR DATA Forces lb Moments ft 1lb POINT fx fy fz mx my mz 190 20 640 88 47 5 773 2 E a PE 200 20 640 234 47 5 J 63519 COADE Mechanical Engineering News The radial force needed for the WRC 107 analysis is simply the negative of the local fx at node 200 or 20 0 pounds The circumferential moment is the negative of the local mz at node 200 or 6351 9 foot pounds The longitudinal momentis the negative of the local my at node 200 or 31 0 foot pounds These Local forces and moments can be related to the Global forces and moments at node 200 by applying the coordinate system rotation matrix This transformation matrix for a two dimensional system is shown below tf M F fs cos sin it fz lsin cos al FZ For this system the
22. he wind loading Upon entering the static processor the user will first edit the wind loading data and then edit the load cases to include the wind effects 4 How can I get CAESAR II to create larger printed plots The size of the CAESAR II printed plots is determined by your graphics resolution LaserJet printer resolution is typically 100 dots inch in the horizontal direction and 75 dots inch in the vertical direction Knowing this you can determine the size of the expected plot based on your individual graphics resolution Forexample a Color Graphics Adapter CGA has aresolution of 640 X 200 dots Therefore a horizontal plot would yield 6 4 inches in the horizontal by 2 inches in the vertical based on the horizontal printer resolution of 100 dots inch Similarly a vertical plot would yield 8 53 X 2 67 inches based on the vertical printer resolution of 75 dots inch The largest available plot is going to be with VGA graphics 640 X 480 resolution in the vertical direction which yields a plot 8 53 inches X 6 4 inches COADE Mechanical Engineering News CAESAR II Specifications 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
23. if a support is moved down the line a few feet or what if the spared pump is not heat traced One change of the input data and a few more seconds of analysis will allow the engineer to evaluate the reliability of the system under variations in installation and under unexpected operating conditions Within minutes and without paper the results from a base model s analysis can be compared to those of a modified model When only small simple model changes are made in each design iteration these modifications could serve two purposes First if the model is accurate the modification can push the design closer to the desired value This will be called fine tuning the design Second the modification is made only to test the impact of the change in the results This will be called a sensitivity study Both fine tuning the design and the sensitivity study are a natural outgrowth of computer simulation Fine tuning has a specific goalin mind For example if pump loads must be reduced then intelligent changes are made to the model and the results are examined to see if that goal is approached Of course fine tuning is only valid if the model itself is fine or detailed in the first place The sensitivity study is more of a discovery process The model is tweaked to see how sensitive the results are to the change For example if pump loads are low in a model where all rack piping supports are modeled as rigid Y restraints
24. immovable point at 145 With the drawing prepared with all the required data an experienced user can produce afirst pass analysis in under 15 minutes The program selects the spring for installation at node 125 in this case from the Grinnell catalog and then calculates the system loads and displacements for the operating and installed conditions and the B31 3 sustained and expansion stresses The pipe stresses are well within the B31 3 limits COADE Mechanical Engineering News Figure 3 The calculated pump nozzle loads are compared to the limits defined in API Standard 610 The simple check compares the component loads on the nozzle to values based on nozzle size and orientation Ifthe nozzle loads are below these limits specified in Table 2 of the standard then the pump alignment is assured A second test is available for pumps if the Table 2 limits are exceeded This secondary check is defined in Appendix F of the standard Appendix F has three checks F 1 2 1 the component loads on each nozzle which can be as large as twice the Table 2 limits F 1 2 2 the resultant forces and moments on each nozzle and F 1 2 3 the resultant forces and moments and the moment about the local Z axis of all loads resolved to the center of the pump The example here generates pump loads that exceed the API 610 Table 2 limits It is necessary then to examine the Appendix F limits defined in that standard The CAESARI API 610 processor coll
25. ly the SYSCHK system check program Volume 19 November 1994 has been updated to reflect the variety of different operating environments under which users run COADE software The new SYSCHK main screen is shown in the figure below CAESAR II System Check COADE Engineering Softuare Graphics Device Information Secondary Board None Secondary Monitor None Primary Board Primary Monitor UGA Color Environment Data Memory Processor Data Variable CAESAR Installed RAM 655344 bytes Value G C2TEST Free RAM 589344 bytes DOS Version 6 80 HIGH Processor 80486 Windous Version Not Loaded Netuork Loaded Yes Math Chip 80487 NDP Status already Enabled Current Drive G NDP Accuracy Passed Drive Size gt 124682368 bytes Drive Free Space 21571584 bytes Disk Compression None SmartDru Loaded Yes Expanded RAM Free Exp RAM Extended RAM 7733248 bytes 6520832 bytes 6422528 bytes System Date Time 09 24 94 12 07 Strike any key to continue E The major changes enhancements to this program are e Arevised detection method for the DOS SHARE utility isimplemented This is necessary because itis illegal to load SHARE once Windows is loaded Therefore Windows always patched the standard detection method to lie about the existence of SHARE SYSCHK can now detect if SHARE is loaded even from a DOS box under Windows e The new SYSCHK checks to see if DOS is loaded in high
26. ly 70 of the operating load e Set the spring so that the spring load is 130 the design value e Move the spring towards the discharge riser by 3 5 feet and e Move the spring 3 5 feet away from the discharge riser In each case the suction loads on the pump are the same The pump discharge loads in pounds and foot pounds for each variation are shown in Table 1 below The final column of allowable lists the maximum ratio of the Appendix F criteria Case FX FY FZ MX MY MZ __ of allowable 1 light load 315 1745 279 3620 2093 2172 100 74 2 heavy load 312 1252 278 3571 2105 2882 111 00 3 closer 313 1318 278 3619 2084 2166 98 01 4 farther 316 1696 281 3592 2118 2562 101 39 Table 1 Only one of these variations moving the spring closer to the riser passes the API 610 check the loads here are much better than the original design The other three cases fail for the following reasons Case 1 The moments resolved at the pump s base point exceed the allowable limit Condition F 1 2 3 Case 2 The bending moment about the Z axis at the pump discharge nozzle exceeds the two times the Table 2 value and Case 4 The moments resolved at the pump s base point exceed the allowable limit Again why were these analyses made The analysis of the original layout showed that the pump was very close to its maximum rated loads The model was first examined to determine whether or not modeling sho
27. m a roof design according to the procedures outlined in Brownell amp Young This will provide the number and size of rafters girders and columns for supported cone roofs Other enhancements include additional input specification for anchor bolts settlement parameters more nozzles external nozzle loads and weight CAESAR II Version 3 21 Released in July In late July CAESAR II Version 3 21 began shipping to all users current on the update maintenance plan Besides the major enhancements see box there were several smaller changes many a directresult of user suggestions and requests aimed at increasing ease of use which were not given as much attention in the update documentation These subtle changes are discussed below Graphics File Viewer Many times users find themselves staring at a directory of job files wondering what is in each file or which version of a particular job contains a certain modification Entering the input processor and plotting the model is a slow and tedious process when looking for a specific job or modification If the selected job is not the one desired the input must be exited then another job selected and the input plot procedure repeated To aid in this job file search Version 3 21 incorporates a graphics file viewer directly into the File Manager of the Main Menu From the job file list the job currently selected can be viewed by simply pressing P The job is plotted showing restraints
28. m under load Fis calculated as F K the Dynamic Load Factor or the ratio of the dynamic to static response is therefore 1 cos t This DLF varies in magnitude from a value of zero to a value of 2 0 where the maximum and minimum values fluctuate along the time axis with the natural period the engineer is of course most interested in the maximum value of the DLF The response spectrum for this load is generated by plotting the maximum DLF that occurs throughout the load duration 2 0 vs the natural frequency Obviously drops out of the equation so the DLF response spectrum for an instantaneously applied constant load of infinite duration is a flat value of 2 0 DLF omega Once the Dynamic Load Factor is known a static analysis of the system or of each mode of vibration when considering multiple degree of freedom systems can be done with the static results then being multiplied by the DLF to yield the dynamic results Note that the DLF is a non dimensional value that is independent of the applied load Expanding this discussion to loads of different shapes and durations allows the application of a force spectrum solution to impulse loads of different types technically an impulse is defined as the area under the force time profile The dynamic response of a system to an impulse load is dependent upon the ratios of the load duration and rise fall rate to the 10 November 1994 natural period of the
29. ons Guide covers model building static analysis and interpretation of results Network Drive Access The File Manager incorporated into the Main Menu and both the static and dynamic output processors has been enhanced to search out and subsequently access all network drives Previous versions of the software would access all drives up to the first nonexistent drive This caused non contiguous network drives to be invisible to the software Version 3 21 specifically searches for all drives from A to Z and maintains a list of valid drives discovered Note that the installation program INSTALL also incorporates this enhancement to facilitate installation on network drives Accounting The Accounting Module has been completely replaced for Version 3 21 to streamline the operation of the system The initial release of this module allowed the generation of accounting reports toa disk file C2 ACCT OUT only At the request of several users this module has been modified to send the reports to the terminal screen or the active printer This new module is available from the BBS in the CAESAR II download area The file name is ACCOUNT EXE November 1994 Seminar Schedules for 1995 Our seminar schedule for 1995 has been set It appears below We will again have five piping analysis seminars and two pressure vessel courses These are our open attendence courses held here in our training room Other in house training can be held a
30. ontrol to the input routine by passing the units conversion step This error is only apparent for non English operation This error corrected for Version 3 21 is in all other 3 x versions of the program November 1994 3 PenPlot Module A plotting error was discovered in the Pen Plot module which caused all elements of the model to be plotted as expansion joints when the job included thermal bowing This error exists in Versions 3 19 3 20 and 3 21 This error also exists in the Animation Module in Versions 3 19 and 3 20 This was corrected in 3 21a 4 Static Output Module An error has been discovered in the static output module which could cause the restraint hanger symbol plotting to put the symbols at incorrect locations This problem only occurred if the restraint summary report was previously requested resulting in a restraint nodal sort This plotting error exists in Versions 3 19 3 20 and 3 21 This was corrected in 3 21a 5 Graphics File Saves A compiler conversion problem was discovered which prevents appending to graphics image files This problem exists in Versions 3 20 and 3 21 in the following modules piping input structural input dynamic animation static output This was corrected in 3 21a 6 Documentation Technical Reference Manual On page 3 27 of this document reference is made to using the Alt key to plot node numbers with hangers supports anchors and nozzles This is an
31. rtcuts might have produced low values for these loads This initial review November 1994 could have indicated that model modification fine tuning was necessary But here the model looked fine so the model was manipulated in the second fashion a sensitivity study was made This sensitivity study did not focus on variations in design layout but instead on variations on the hanger installation The study clearly illustrates that the layout is sensitive to the spring the spring load and position must be properly set for reliable pump operation It is noteworthy that the spring which in so many cases is used to improve system loads can also harm the design Man time for the four analyses in this study is minimal about ten minutes but the benefits may be great Not all systems require this sort of study but here where the pump loads are very close to their limits such an examination plus a close review of the model definition is warranted If these pump loads are not reduced it would be wise to give close attention to this pump at startup This sort of formal analysis can confirm a good piping design and can assist the engineer in redesigning the piping system when it is required When used as a tool for sensitivity studies piping analysis software can indicate critical installation parameters so that safe and reliable operation is ensured The CAESAR II Pulse Table Generator By Thomas Van Laan and Richard Ay
32. ry and Computation by Mario Paz the displacements of a system due to any arbitrary load can be calculated through the application of Duhamel s integral as shown below 1 X t X cos a sin a Fl a sin dt dd T M Where t time for which displacement is solved sec Tt time at which force is applied sec When the arbitrary load can be broken into discrete linearized segments the force function can be fairly easily integrated using Simpson s rule permitting the solution of the system displacements at any point throughout the duration of the load CAESAR II provides the user with a Pulse Table DLF Spectrum Generator which performs the automatic integration of Duhamel s integral This module takes a user supplied segmented pulse and creates the appropriate equation for each segment Displacements are calculated at each terminus of the segment and the equation is differentiated in order to locate any displacement minima or maxima ocurring within the interior of the segment this assures that the maximum displacement is found without using a hit or miss approach The absolute maximum dynamic displacementis then selected from the largest of the segmental values and the DLF calculated from that This process is repeated for the number of natural frequencies specified by the user which when plotted create the response spectrum Response spectra for the same load profiles as shown in the theoretical plots above h
33. shipped to all users 3 Appendix P Nozzle Computations A data management error prevented the computation of nozzle stiffnesses and limiting loads for any nozzles after the first one specified The results for subsequent nozzles are identical 14 November 1994 to those for the first nozzle This error exists in both Version 1 00 and 1 10 of TANK This problem was corrected in Version 1 10C and was shipped to all users Class 2 1 Input Module An error has been discovered in the Version 1 10 input module regarding the automatic specification of elastic modulus and expansion coefficient for the Appendix P nozzles The data provided is correct however for nozzles 2 through 5 this information is placed in the wrong input cells This error is obvious to the user when specifying the remaining nozzle data This problem is corrected in Version 1 10C 2 Output Generation Module Several conversion errors were discovered in the output preprocessor which affected the display of user input The values of design temperature bottom plate yield stress and bottom plate thickness were not converted from the English system properly This error exists in Versions 1 00 and 1 10 of TANK and is corrected in Version 1 10C 3 Output Generation Module An error exists in Version 1 10 which prevents the nozzle input text labels from being associated with the proper input data This error produces incorrect input listings This problem
34. system For an instantaneous rise time and an infinite load duration the ratios to all natural periods are the same which explains why the DLF was constant for all frequencies in the example discussed above A review of the textbook Introduction to Structural Dynamics by John M Biggs shows several plots of maximum DLF curves as a function of the ratio of load duration to system natural period DLF mox DLF mox How is this information useful to the pipe stress engineer who has a specific dynamic problem to solve Usually the shape of the pulse load profile is trapezoidal with the duration of each segment known For example for a relief valve load the segments of the trapezoid consist of the load rise with a duration equal to the opening time of the valve the relatively constant jet load with a duration equal to the time required to vent the excess mass from the system and the load fall with a duration equal to the closing time of the valve This type of load can be converted to a response spectrum by solving the dynamic equation of motion through each of the load segments dividing the maximum displacement throughout by the static displacement under the maximum magnitude of the load and then plotting the result vs natural frequency COADE Mechanical Engineering News According to the textbook Structural Dynamics Theo
35. t any time anywhere by appointment The CAESAR II courses will be held here every other month except for July Our entire support staff pitches in as instructors for these courses so you get to understand several different approaches to system analysis There is also ample time to meet with other students during lunch and other breaks to discuss common issues in engineering analysis and construction One change for 1995 is the elimination of the three day introductory course to pipe stress analysis Even though the people who attended the course saw the value in it participation was light To bring new CAESAR II users up to speed we will offer an optional Monday evening session to review and explain CAESAR II basics Class size is currently limited to 17 students and most classes are fully booked Sign up early to lock in your dates We find that this is an excellent opportunity for all engineers and designers both new and experienced users alike to spend a few days on the subject without interruption Understanding the concepts in modeling and analysis will produce better design strategies more efficient use of time and greater confidence in the results Piping Seminars in Houston Texas January 23 27 Statics amp Dynamics March 20 24 Statics amp Dynamics May 15 19 Statics amp Dynamics September 11 15 Statics amp Dynamics November 13 17 Statics amp Dynamics Pressure Vessel Seminars in Houston Texas February 6
36. tting of the environment variable on the left side of the screen If the environment variable for the program is not set you must run out of the program installation directory and the switch directory drive feature of the file manager is disabled Pathworks Network Users A majorincompatibility problem between Pathworks and the CAESAR II graphics hardcopy drivers has been resolved Updated modules are available from the CAESAR II file area of the COADE BBS The necessary changes are being distributed as Patch B The file to download is B321_U ZIP Some users have asked about the operation of COAD software on Pentium processors COADE has tested all software products on a Pentium 90 without any problems TANK Enhancements Work is progressing for the version 1 2 of the TANK program One item added for this release will be the ability to plot the interaction diagrams for the API 650 Appendix P limiting nozzle loads Nozzle loads are indicated on the plot as an asterisk easily showing whether the loads are within or exceed the code allowables An example plot is shown in the figure below November 1994 1 2 Lea lFp 25188E 6 a 1 8 Interaction Point at X 249 126 L 2 fFp 62212E 5 L 1 L L ii 1 2 1 0 0 5 5 1 0 1 2 7901 13949 J 1 2 COADE ENGINEERING SOFTWARE DEALER DEMO COPY IDH 108001 The most important item added for the next version will be the ability to perfor
37. valves rigids and expansion joints The next keystroke returns control back to the File Manager where another job file can be selected The entire procedure takes approximately three seconds COADE Mechanical Engineering News Note that there are no graphics controls built into the viewer distributed with Version 3 21 This decision was made to keep the viewer program as small as possible which reduces the load time Once the desired job is found the input processor should be used to access the other graphics functions A full featured viewer was developed but the load display time approaches six seconds for this reason the small viewer was released Major 3 21 Enhancements Low RAM requirement reduced to 475K Input model size limited only by amount of extended memory B31 5 piping code added UBC earthquake spectra added Direct network support and network ESL Input Echo Report Selection From either the static or dynamic output menus a user can request an input echo of the current job This report was automatically generated based on what was inthe input file Many users requested additional control over the input echo to deactivate certain reports As a result Version 3 21 incorporates into the Input Echo module the same report selection menu found in the input processor CAESAR II Tutorial Included in the update notes for the Version 3 21 release is a one hundred page tutorial This tutorial Chapter 4 of the Applicati
38. ws tangent in the From To direction local z points towards the center of the circle described by the bend local y can be found by applying the right hand rule Definition of local coordinates on bends X follows pipe into and out of bend Z points toward the center of the bend Y is defined by right hand rule Note that the figure above and the figures from the 12 92 article allrepresent systems aligned with the Global coordinate system Howcan the system forces and moments for skewed piping be resolved The figure below shows a portion of a line which terminates at a vessel nozzle The pipe running into the nozzle makes anangle of 150 degrees with the Global X axis Note that the pipe nodes are from 190 to 200 which defines the direction of the Local x axis The angle from the Global X axis to the Local x axis is 150 degrees We need to know the loads imposed on the nozzle fora WRC 107 analysis which means radial circumferential and longitudinal directions If the pipe element had been aligned with the Global X or Z directions it would be a simple matter to obtain the forces and moments from the restraint report However for a skewed system the forces and moments must be obtained from the element force moment report with a change in sign The figure below shows both the Global and Local coordinate systems for the pipe element 190 200 Below the fi
Download Pdf Manuals
Related Search