And you guys are able to work from home. You can always get in touch with Bentley to give you any assistance. If you need any help with working from home. We also have the year and infrastructure awards coming up. We still have that plan. We still or hoping that that will go through. It's an exciting and prestigious global competition that recognizes the advancements in infrastructure. It's been going on for the past 16 years. The awards have recognized the extraordinary work of nearly 4000 projects, spanning infrastructure, design, construction and operations. This year the awards will be held in Vancouver from October 12th through the 15th. There's going to be three finalists from each of the 19 categories. That means a total of 50 Seven finalists will win a trip to Vancouver to present their projects before an international audience. Had leaders of industries are going to be there and the nomination deadline is May 1st. So visit Y i.bentley.com/awards to start your submission today. We are in our new. On 24 format, so this gives you the ability to engage with our presentation, so all of the windows are re sizable an movable you can check out the resource list which has some copies of the slides that I'm going to be presenting today and then there is a. Also some links there for you to follow if you want to following the presentation. You can also submit your questions to me using the Q&A window. I have a colleague that will be helping me with that, so you should be getting answers quite in a good time and then you can also submit them for on-demand viewers as well. For this presentation I suggest you keep the audience view and the live view open so that I will be going through some slides in some polls, and then I'm also going to be sharing my presentation through my screen share. Some helpful tips. This web cast is going to be streamlined through your computer, so there is no dialing numbers for the attendees. Please make sure your computer speakers or your headphones if you're using. Those are turned on and the volume is set to inaudible level so you can hear us. Keep in mind these webinars, our bandwidth intensive, so closing any unnecessary browser tabs will help conserve your bandwidth and prevent any issues on the presentation. And a hard wired Internet connection is recommended if you are not seeing the slides advance, please refresh the webpage by pressing F5 on your keyboard command. Or if you're using a Mac, Command R Keys. You can also download your certificate if you want one for this presentation. When an if you watching it as a group, you can add participants for that certificate as well. So now we're going to get onto our presentation, so let me go ahead and share my screen. We're going to be presenting the topic. We've been doing a dynamic analysis series and this is going to be, I believe, the final presentation on that series, and it's going to be on the harmonic analysis within auto pipe. We're going to be covering three topics that we're going to be giving some harmonic analysis background. We're going to review the modal analysis because the modal analysis is important for any kind of dynamic analysis and then will run through a demo of the harmonic analysis. So harmonic loads, a harmonic analysis, is used to analyze the affective of vibration due to an oscillating load. The main types of a harmonic load that can arise are from an unbalanced equipment vibration. They can also come from acoustic vibrations. And they also can come from reciprocating equipment that. Provide proof. Causes a pulsation in the flow. Generally piping vibration is due to unpredictable analysis uncertainties and those are unknown sources that are causing these, so it's difficult to take into account all of this in the design phase. So the reality is that most vibration problems are considered as they arise, an with only the predictable source is considered at the design phase. The motion of a vibrating pipe system in the film is consists of two parts. There's free vibration, which means that the pipe is oscillating with the same frequency as the natural frequency. This type of motion would damp out quickly if the disturbance discontinues, and then there's also the forced vibration. This is due to the Harmonic Applied Force. If the force persists, the motion continues at the forcing frequency. So where the vibration corresponds to a norm. Forcing frequency, then it's easy for the engineers to determine the source and then try to prevent that when Harmonic forces are insignificant than the sources. Generally random impulses, like pulsation from slag or slug flow, an water hammer. These loads are are normally handled as the time history loads and they do not tend to lead to a steady state vibration. These loads can be damaging to a piping system if their frequency is close to the systems natural frequency, thereby introducing the resident condition. Thus we need to review our modal analysis with a modal analysis. We are interested in the limits of the response of the system being analyzed. So for a certain applied load, what will the maximum displacement be and when will it occur? Our goal is to find the natural frequencies of the system or the frequencies that will make the system naturally vibrate. This is also known as the resonant frequency as there a frequency increases and the amplitude of the resonant responses measured. As we approach these resonant frequencies, the response becomes infinite. The natural frequencies are found using a stiffness matrix using eigenvectors and eigenvalues and auto pipe does all this for us. I won't go into detail on this topic in this session, but our training does go into a bit more detail and so does the auto pipe help. You can review the recording also of the modal analysis sig that was presented already in the dynamic analysis series. What is most important for the user to understand is why these frequencies are important and how to set up and analyze the modal analysis in auto pipe. So for the analysis of steady state vibrations, it's generally referred to as a harmonic analysis, and as it is an analysis of loads that have signed new suits, I knew soil forms. For practical reasons, complex vibrations can be decomposed into a number of single harmonics or single degrees of freedom, and the analysis is done with one forcing function at a time. The equation for a dynamic equilibrium is associated with response of the structure subjected to a harmonic forces shown here. Since all harmonic forces acting on piping may not have a definite phase relationship, the Harmonic loads are separated into groups. Each harmonic analysis case performed for the group with a fixed phase relationship. Furthermore, to simplify, each analysis handles only one forcing frequency. Each analysis cause case calculates the displacement force and moments and the final response is obtained by combining the sum of the individual case. Results absolutely it is important to consider at least the first 3 or 4 the harmonic load dialogue can be used to show how different harmonics combined into more complex vibrations, illustrating the fact that complex vibrations can be decomposed into a number of harmonics. With these harmonic loads, there are applied a stress on the system occurs many times or many cycles. That means we have to consider the fatigue effects on the system. Auto pipe does not offer a cyclic fatigue analysis for design codes other than the ASM E nuclear code, but a simplified assessment of the effect of the vibration on the fig Teague damage can be made, which we will go in through with our example, which is next. This is a suction side of a natural gas supply system. With a reciprocating compressor. It has two 6 inch suction lines that are connected to the Reciprocating Compressor Assembly. The Reciprocating Compressor Assembly consists of 22 inch suction and discharge bottles. It also has agents, suction and discharge nozzles. And the 16 inch compressor cylinder assembly. The pressure at the suction bottle is 400 PS I which is increased to 650 PS. I at the discharge bottle acoustic shaking forces which are periodic in. Nature are produced in the operation of the compressor. If the frequency, phase and magnitude of these shaking forces are known, the harmonic analysis utility of auto pipe can be used. And we perform this. We can perform this up to 10 harmonic load cases. The acoustics shaking forces can be calculated using any acoustic simulation program, or it can be obtained from some industry guidelines. Airgo the compressors manufacturer guidelines. The program pulse, another product of Bentley, can generate these forces in the form that can be used directly within auto pipe. For this example, the acoustic shaking forces are computed using pulse. There are two critical speeds, 276 RPMS and 280 are pins of the compressor and that generates the high shaking forces. To simplify this example, only two of the most critical harmonic associated with each speed are used in the harmonic load analysis. Since the compressor operates at a single speed to harmonic load, cases are considered in the analysis. One for each critical speed. A further simplification is considered to shaking forces. A shaking forces applied at Point A5 due to an unbalanced pressure at the suction side. And a shaking force applied at a 81 due to an unbalanced pressure in the six inch suction line between the elbows A8 and A9. 81 is at the midpoint of the section between A8 and A9, just to be aware. So let's walk through our demonstration. We've got auto pipe here. We've got our model here, so our first step is to get our modal analysis. Find all those natural frequencies. So our first step is to go to the tools ribbon and head to our edit options so that we can set up our mass points for span. We're going to have it set for zero at the moment, so we're going to turn that to an automatic and we've got our cut off frequency at 200 Hertz. Now, if you're getting a little Austin this again, I would refer you to checking out the auto the. Contents that we have on the modal analysis. We have that cig and we also have training on it and then help. Content is always helpful as well. So we've set up are are cut off frequency. Keep in mind that pressure pulsations create axial loads. In order to capture the higher frequency axial modes, you gotta use a higher cut off frequency for mass discretization. While this increases the complexity of the model. Our modern computers can handle this. We will know the frequency of any measured input, so we will want to include modes of at least 50% higher. And the cut off frequency value here should always be higher than. In the modal analysis where we're going to go to next. Now, if you want to see how many mass points have been added, You can check that in our new report we've recently. Added or changed the way that our reports are created using a SQL Lite report so the ones to can step shows where the mass points have been added. You can go to the coordinates report. And this will. This has a column once it's created. That shows how many mass points were inserted. In between the node points. So between a, a one and a 8, there's been four points. Put it in so A1A8. From this point to this point, there are four miss invisible mass points to help with the modal analysis. So now we can go to the dynamic analysis. Command and verify that are cut off frequency is at or higher than the mass discretization setting and we have that correct, so we're going to go OK, and we're going to analyze our model. Now you have to ensure in our analyze all command you can either have this checkbox turned on, sometimes it is turned off and if it is turned off you need to ensure that all of your analysis are checked on, so we'll leave that check box on. Will go through our analysis. And we're going to check out our modes of vibration, so we're going to go to our results ribbon and we're going to check out our mode shapes results. We can animate those. We can see all of our mode shapes and go through. Our next And there's a few. To visualize here, I'll go through it a little bit slower. You can see all the displacements and. The results of those displacements. Now you can also go to the results grid and go to the frequency report and you can see all of those natural modes of vibration listed here. Now one thing that we talked about in our modal analysis. Is. Each mode captures a specific amount of Mass in a specific direction. You can see that in these columns here, captured modal mass X captured model mass Y. So you can see which frequency. You can see the frequencies here 8.62 in that frequency the majority of the mass is responding to the X direction and so on and so forth. And as we go through this. This table you can see the cumulative mass that's been captured. Uh. So the participating mass needs to be higher than 75% as we discuss in our modal analysis training. In this case we do not have that those numbers, but we're still going to run through it without a key trick that we could do to increase our modal mass. Are modal analysis frequencies is by going to our segments. Or not our segments, but our pipe properties. We have a few that are. We have quite a bit of properties in here, but not all of them are being used, so we're going to use a cool trick. We're going to go to our delete ribbon and we're going to delete our unused pipe identifiers, and I'm going to delete all of those. And now list leaves us with just the pipe identifiers that we have. So the only one that we are really concerned with the modal masses, the piping, all these other ones we could in effect change the pipe material to a non standard pipe material and change the density to zero. That will take that mass out of the modal analysis and help you get your model captured modal mass to above that 75%. Criteria. Now you can also use the missing Mass. Features that auto pipe as well. If you are still struggling to get above that value and again those methods are mentioned in the modal mass Cygan training. So now we're going to define our loads. We are lucky that we had the post program to create our loads, but we can go and check out our harmonic loads. You can do these by creating them manually if you know what you're dealing with with the manufacturers. Has that information available to you and it basically gives you these. This information that auto pipe needs it needs the frequency, the phase angle and then the pound load that is imposed at at those points. Now you can also see that a different frequencies that pulses given us that there's different waves. And then the sum of those and we have that for two. The two speeds that we spoke about earlier. So if you wanted to create your own new one. That's easy enough, we can specify. We want to create a new one. We're going to. Name that test is what we're going to Neymar. Our new one with and we're going to do that with an imposed acceleration and gravity going to use a damping force of .02. And then we're also going to put this at Point B 11. Frequency of six hurts. Zero phase angle Anna 1.5 acceleration in the G direction. We're going to do this again for the 12th frequency that's out of phase of 30 degrees, and that's going to provide a graph and acceleration of .8 and then again. We know, for example that at 18 hurts and out of Phase 48 degrees we get an acceleration of two. So that's just an example of creating a load. If you know these inputs, you can put make those manually, create those. Now. What this does is it puts it into our. This is the folder location for our model and it creates this file. This HM F file and you can open that file. Using any. Text, text file reader and that's This is basically the file format that it creates that hmm file. So with that done, we can now assign are harmonic loads, so we're going to go back to our analysis. Are dynamic analysis command go to our harmonic loads? These are some previous load cases, so I'm just going to delete. Those are going to create some new analysis load cases where we're going to use our datasets that we. Are in concern with. We're going to leave it as RMS combination method. Going to add another. 1 using the 280 speed and leave it as RMS. So what's going to happen here? Is we're going to create two harmonic cases with these H1 and H2? The results are going to be found in those load cases. We don't want to include the missing mass or the GPA here. Those are the. We pass that model criteria. That's what we would want to do. We're not going to do that in this example, but just keep in mind that that's something that you would want to do. And now we're going to analyse our system. So we're going to go ahead. And yes, we're going to analyze our model. We don't need to review our consistency checks 'cause we know we're a good model and we don't have anything like that in our model. And again, make sure your hallmark is checked. We've got the check box to analyze everything so at all of our analysis are going to run. Going to let that run. And our first thing that we're going to do today is we're going to check out our combinations and we're going to ensure that it ran. So going to check out our code combinations and then we're going to notice that it created this sustained plus H1 and sustained plus H2. So it did work like we wanted. So now we're going to go check our code. Stress is going to check that for our sustained plus H1. And in this load case we see that are allowable for this point, a 10 is 21280, but we need to remember that the allowable stress that API 618 recommends is that we need to take half of the allowable of this combination. So what we're going to do, we're going to take note of this allowable. This 21280 will write that down, and we're going to go back to our combinations. And we're going to go back to our code combinations and we're going to Uncheck this auto update. So what this is going to allow us to do is we're going to put in our. Are Allowable that API 618 tells us to use, so we're going to divide 21280 by two, and we're going to put in 10640 for these two to load cases. And 6:40 and that's going to force are allowable to instead be codependent. It's going to be what we say it is, and that's going to our stretch ratios are going to reflect that. So now we go back to our code stresses and we're going to see that we're still over stressed. So the compressor cylinder also looks like it's highly over stress and this is a result of the high hoop stress. So we're not designing the compressors for this. So what we're going to do is we're also going to check to see if there's anything that the model is doing that's causing. These over stresses are going to check out the displacements, and we're going to look at the Harmonic load case displacements to see what's going on in this load case to see what if we can see something that's causing this problem, and sure enough, we see quite a bit of this displacement Here in this section of piping. This section between A8 and A9. So. One thing to also keep in mind, because we're doing at the root mean squared. This is always going to be a positive number. There's not going to be a negative displacement in a harmonic load case. So be aware of that. It's also going to be in this direction, and if you recall the displacements of those mode shapes, it goes positive to negative, so these results are always going to be in the positive direction. But that does not mean that it won't go in the negative direction as well, so you'll have to keep that in mind for pipe interference issues, like if you're going to hit a column or something like that. So what we're going to do in this case to solve this problem is we're going to insert a. Pipe support, so we're going to insert a line stop in this pipeline here to prevent that axial movement of that line, and we're going to hope that that fixes our problem. So with that line stop in there. We are going to analyze the system. We don't need to review our check warnings. We've checked those before. We're going to run our analysis and then we're going to look at our combinations, 'cause we know that some of our. Typing is gotta hoop. Stress we're going to take that out of consideration. We're not too concerned about our hoop stresses, so we're going to go to our maximum. We're going to Uncheck the print check box here to on print those from our results. And click OK. Then we're going to check out our code stresses and see if all of our code stresses are within range. And sure enough, we are down to a .9 one in all of our piping is within the code allowables. But now we have to take into account the effect of fatigue. So like I mentioned. Our program auto pipe does not offer a cyclic fatigue analysis for design codes other than a assuming nuclear codes, but a simplified assessment of the effects of this can be done. So the general approach and the evaluation of this is the vibration stress. The design code factor stresses using stress intensification factors, so the reported value is not the actual stress, so some codes do not factor in the girth welds. So the engineer needs to fully understand that the value of the stress needs to be applied. For high cyclic fatigue, which involves almost an infinite number of operating cycles, the acceptance criteria would be based on an allowable endurance limit for that material. For the nature of high cycle, fatigue is somewhat different to low cycle fatigue with less than 10 to the four cycles, so so also has to keep in mind that material defects, like notches, Weld residual stresses, those are more tolerable in the low cyclic fatigue due to the local yielding. But they have high significant effects in high cyclic fatigue at the at the elastic state. So ask me defines the needs to define the endurance strength. So what we do is we are going to go to our results ribbon, go to our combinations, and we're going to. Create a code combination. We're going to call this the fatigue check. For H1, we're going to do that is a sum, and we're going to list as an occasional category and put that H1 in there. We're going to also create this fatigue check load case. H2 doing the same occasional. Category. And put H2 in there. So now we have these two low cases to do our our fatigue check. I forgot check in there but we get the point. So now we can go to our results grid when we can go to our code stresses and we can turn off all of these. Checks, except for the two that I am in. I want to check so then we double we sort these results to get the maximum stress that is created from these load cases and we can see that we are approximately in the 6000 PS. I arrange So what we can do then I'll bring this over to my other screen. We can go to an endurance. Curve for 40. For for your material and what we can see here is for this piping. So for the amount of stress that we are seeing and the number of cycles or well from the stress that we are seeing, we have to double it because that's only 1/2 of the half of the load is going to be plus 6000 plus negative 6000 so that stress range and is good needs to be about 12,000 so then we can go from that 12,000 point. Come over to our curve and then come down to see how many cycles. We're going to be experiencing here. We're going to be experiencing around 3 * 10 to the nine cycles. Is that going to be good enough for piping system? That's only a question that you can answer yourself, so you'll just if you have that issue where this isn't enough cycles for your piping system, well then you're going to have to figure out a way to reduce that amount of stress. And that's pretty much the example, I hope. You guys have enjoyed the presentation thus far. And we can now turn to any questions that may be outstanding or any questions that need to be brought to our attention. And while those questions are being brought to our attention, I do have some poll questions that I would like the audience to to answer. So go ahead and take a look at your live view and please respond to the poll question. How many years have you been an engineer? We would always like to know more information about our user base. Finding out the experience levels and seeing where we need to focus our efforts. Do we need to focus our efforts with? Our younger engineers or our more experienced engineers so please share with us. How many years you've been in the field? And just goes there. Any questions that needs to be brought up? Hey Luke, so yeah, there there is a question here, just mentioning that our limit of 10 harmonic load cases you may need more than that and just asking if there's a plan to increase the number of harmonic load cases. See and that's that's input that we like to hear, so it's always good information that we need to hear. We can always take those requests and add it to our backlog of requests. And the more people that ask for those types of things that's going to raise that priority and get those things accomplished quicker. So those are good things for us to know. So yeah, well keep will take a note down of the name of who asked for that and we will add you to our requests for that. It is something that we have. Been considering to increase that value. Um, somebody asked if there is a recording that will be available. So yes, on the you'll get an email actually, that will let you know when this is posted. It should be posted within an hour or two after the session is over and you'll be able to review the recording on the Bentley events page. Looks like it's about 5050. Well, most of our engineers are over been in the field for greater than 10 years, so that's great to know. Glad to see a good spread though of zero to five and five to 10, so that's good to see that we've got a wide breadth of experience here in the room. One more question I'd like to know is how many years you have been using auto pipe. So please let us know how long you've been using it. That's always good to know to know. Again, how much training we need to provide? Have you been using auto pipe for a long time or? Have you just been started starting to use auto pipe? So in the mean time we have some other questions we could probably discuss here. A question is will auto pipe affectively model gas transients due to valve openings? Which is flu transient topic not harmonic but. Right and go ahead we do. We do have fluid transient capability. That was another one of our. Dynamic analysis series. So I would encourage you to take a look at those past presentations. We also have it in like we stated with modal anharmonic we have training. We have helped content. We have Bentley communities. All of those are good places to go. Look for answers. For how to handle your particular fluid transient question. And that's similar. There's another question here, can we do harmonic analysis for flow induced by Bration? We would also recommend for flow induced vibration to do time history analysis. I think as Luke mentioned in the beginning of the presentation. Right, yeah, that's the whole point of most of these analysis is you have to know the input if that's typical with any kind of finite element analysis, you have to have your boundary conditions and you have to have good definitions of your loading whether or not auto pipe has some tools to help you with that for fluid transients. Other times you may have to use third party software or a second piece of software like post to come up with these loadings, or you may have to, as we mentioned, go to vendor. The vendors of the equipment and say Hey, I need to know what kind of harmonics or transients or loads or response spectrums or what have you to so that I can apply those to my piping to make sure that this equipment is not going to damage my pipe. And we have the results to how long you been using out of pipe. So we have a lot of users that are pretty news, so that's good to see that you guys are trying to expand your knowledge of the program. That's great to see, and we've got a good number in the other age auto pipe age brackets as well, so that's good to see. Thank you for sharing your information with us. Is there any more questions that we should address? There's a few more questions in here, but I think it would be useful to be able to send some more information with those, so we will follow up with any questions that aren't answered verbally in here through and through an email. So just keep that in mind and then in here. Also, there were some recommendations for topics going forward, and we are always appreciative of those, so we will definitely take them into consideration and try to plan for. Sessions on those topics that you've recommended. Yes, please definitely please let us know what topics you're starving for. We're not scared to cover them just we just need to know what you guys want us to cover so we're always open to those suggestions. OK, with that I guess lookout for our responses via email for some of these questions and I appreciate your attention during this presentation and we look forward to seeing you in the following months. Stay safe, stay healthy and look forward to seeing you again. Hey Hey. Looks like it went pretty well people. Yeah, any any comments? Jessica know I think you did Great I. I unfortunately somebody this we have a user is kind of keeps asking a lot of questions. He's been, but I think I need to take that. Some of these aren't quick answers, so we'll take it with her. Well, I think you answered a couple of this. These from Robert, but he was asking some questions like I was chatting with Phil on the side like his first question was have have any of our auto pipe models been experimentally validated for piping system so it's like do you mean like tested in the real world and he said yeah and I said I talk to Phil and he just mentioned the verification model so I said You know not those example models that we have but we do a verification models. He said he's on the Azmi Committee committee or piping vibration standards so. Great person to attend, but he was also the one that asked if we could do more gas transients and our fluid transient and I didn't want to quickly answer that because I actually don't think we can. We can do fluid and maybe we can do kind of gas. I know I had a question on that too, but I think we can if we have like the density and all that stuff. You know, he said from what I've seen, the answer is probably not. I'm working with fluent models for such transients to explain gas pipeline explosion in fatalities, so I don't want to jump on that, but. Yeah, you just to find the density right? So I figured I can just like send him an email afterwards with like a screen. Sometimes I just prefer to do that with Screenshots. And I was trying to keep up like sometimes their first question isn't really there are enough, he said is their reaction forces on the bends. So I wrote back. Yeah, you can see there are forces at the near and far points, but then this guy came back and said sorry not clear enough. I have to look for his next question pulsation forces that are imposed on the bends. I don't understand what you mean. I think possibly he just wondering if like it's almost like a slug flow or something like that, but like I think yeah, the harmonic load is vibrating the entire system, so it's. It's affecting everywhere, so those result. _1594517093476

This AutoPIPE Special Interest Group virtual workshop continues the series on Dynamic Analysis and focuses on Harmonic Analysis. Topics for this one-hour session include:

- Types of Harmonics
- Harmonic input
- Result Review

To see videos from previous Dynamic Analysis presentations, visit **Bentley Events** online and filter All Activities>Webinars On-Demand.

The AutoPIPE SIG is open to all Bentley users, so invite your colleagues!

***The webcast will be streamed through your computer, so there is no dial-in number. Please make sure your computer speakers (or headset) are turned on and the volume is set to an audible level so you can hear the presenters. Help*