 Welcome, everyone. My name is Marty McCann. I'm a member and chairman of the Committee on Geologic and Geotechnical Engineering, which is a standing committee at the National Research Council and the National Academies of Sciences, Engineering and Medicine. Thank you for joining us for this webinar on the state of the art in practice and the assessment of soil liquefaction and its consequences. And we wanted to just do a few brief introductory statements about the committee and the work that we do. The committee works in a number of topical areas, earth processes and materials, earth and structure, interactions, soil and rock mechanics, hazard mitigation and safe and responsible human development. I'd like to thank our standing core sponsors who support the committee's work. Currently that includes NSF, NIOSH, the National Institute of Occupational Safety and Health, and the Office of Mind Safety and Health Research. The past sponsors that we've had of our studies have included the National Forest Service, U.S. Forest Service, the Bureau of Reclamation, Nuclear Regulatory Commission, ASCE, Court of Los Angeles and Long Beach, L.A. Department of Water and Power, DOE, NASA, FEMA and NSF. We also thank our sponsors for our core support and our study support. And I also want to thank Sam Magsino and Remi Shabbata who are staff members at the NRC supporting the committee and making this webinar happen. So thank you, Sam and Remi. And last thing I wanted to mention, the liquefaction study that we're going to hear about today I think is a particularly unique example of where the academies in general can make a contribution to a particular topical area. The liquefaction, as I'm sure you all know, is complicated. There's a lot of work that's been done over the years. And the study came to the committee to try and address sort of the landscape of liquefaction assessment and present thoughts on where we go from here in the future. And Ed is going to provide that story for us. So with that, one logistical matter that I'd like to bring to your attention for those who are listening, we'd like you to, when you pose questions, use the Q&A icon at the bottom of the screen as opposed to the chat box. And you're free to post those questions during the presentation as they come to you. The chat box will take up a fair amount of the screen. So if you minimize the Zoom window, you'll still be able to see the presentation and also have your Q&A box there that you can submit questions and still see everything that's going on. So with that, I'm going to turn it over to one of our committee members, Professor Ellen Rathchie from the University of Texas, and she will moderate the session and she'll introduce our speaker. So again, thank you all for coming. Thanks, Marty. Hi, everyone. This is Ellen Rathchie. I was a member of the Liquefaction Committee that developed this report and will be presented today. And I'd like to introduce Ed Kavazanjan, Professor at Arizona State University. He was the chair of our Liquefaction Committee. And so I'm going to just turn it right over to Ed. And I'll be moderating the Q&A, so I'll be looking at those as you put them up. And then between Ed and I, we'll be trying to respond to as many as we can during the remainder of our time today. So Ed, it's all yours. Thank you, Ellen. And thank you, Marty. And welcome, everybody. In the next 20 minutes or so, I'm going to provide you a brief overview of the report that our committee put together on state of the art and practice, earthquake-induced liquefaction and its consequences. The full report is available for free. Online is a download. There will be a link to the site where you can download it from at the end of the presentation. Or I think you can just punch National Academy's Liquefaction Report into a search engine, and it should show up. First thing I want to do is thank the study sponsors. Marty, thank the people who provide core funding for the Committee on Geological and Geotechnical Engineering. That is the only independent funding the committee has. So when the committee commissions a study, they need to raise money for the study. The study is not cheap because of the expenses in bringing together a select committee and in peer-reviewing and producing a report. This particular study had funding from the Bureau of Reclamation, Federal Highways, Nuclear Regulatory Commission, ASCE, and the Geo Institute, Los Angeles Department of Water and Power, and the ports of Los Angeles and Long Beach. And we're grateful for their support without which the study wouldn't have taken place. I thought I should start with the statement of task, and I think some of the comments we've gotten on the report at times reflect perhaps not a complete understanding of what our statement of task was, what the scope of work was, and also some of the restrictions put on us by the National Academies in a study like this. But the statement of task for the committee was to examine the study on practice, an earthquake and soil liquefaction assessment. Three primary points we were supposed to address, the sufficiency, quality, and uncertainties in testing, case histories and modeling, the methods that are used to collect and analyze lab and physical modeling data for behavior analyses, and then the assessment methods, data gaps, and uncertainties for evaluating the consequences of liquefaction. And then we were asked to make some comments on future directions for research and practice in this area. This is the committee that I had the privilege of chairing. It has a very broad spectrum of subject experts, including people from practice, from academia, from agencies, public agencies, state agencies, and included engineers, geologists, and risk assessment specialists. The committee made 11 recommendations, and going back to the three bullet items in our task, three of those recommendations concerned the sufficiency and quality of the data that's available to us, for concern, uncertainty, and spatial variability, and for address the issue of improving, developing improved tools for making liquefaction assessments. So I'm just going to go through the recommendations. The background for these recommendations is provided in detail in the report. The first recommendation having to do with the data was that we needed to establish curated publicly accessible databases of relevant case histories on triggering and on the consequences of liquefaction. And I should mention that the committee, based on, in part on the statement of task, but also what made sense, basically split the subject area into liquefaction triggering, developing close to the zero effective stress state and lost the strength to the cyclic loading, and then the consequences of triggering. And so we thought this case history, one thing that is particularly lacking with available case histories is case histories of soil structure interaction with liquefied soil. Most of the case histories available for kind of level ground free field response. We had recommendations on the level of documentation of the case histories, which in many cases is not really adequate to make a good assessment of the case history. And therefore we wanted to see this database developed with strict protocols and also include indicators of the quality of the data. You know, in liquefaction assessment, not all data is created equal. And for instance, the standard penetration test where there are energy measurements. So we know what the energy delivered to that by the hammer was is a lot higher of higher quality to us than an SPT blow count where we have no information on energy or on the equipment setup. In particular, with respect to available data, we found that the case history was lacking in data for deeper soils, sites where liquefaction occurred at greater than a 15 meter depth with an effective overburden pressure greater than 100 kPa. And also for both small and large magnitude earthquakes, earthquake magnitude is less than 5.9 earthquake magnitude is greater than 7.8. There was also a need for more data on fines on soils with fines contents greater than 35% for not for any kind of fines greater than 50% for non plastic fines. And for medium stiff soils soils with blow counts greater than 15 blows per foot. And that was particularly in regards to the consequences of liquefaction. We had a lot of data on, well, not a lot, but there is data on liquefaction triggering up the blow counts of about 30. But when you look at the lateral spreading database, which is one of the important consequences of liquefaction, there's really nothing greater than 15 blows. Some people might want to interpret that as meaning there's no lateral spreading when the blow counts greater than 15. But we were reminded repeatedly that the absence of evidence is not necessarily evidence of absence. We felt that the available data needed to be updated to include data from recent earthquakes. In particular, some of the more recent now it's been 10 years or so. But the large magnitude earthquakes that we saw at the beginning of the 21st century, we needed more data on the consistency and transparency of the information that was being collected. And the database should have open access capabilities for searching. And the histograms here kind of are meant to illustrate the narrow range over which data, in this case triggering data, is available. So if you look at the left-hand side, the number of occurrences in the most common case history database versus earthquake magnitude, you see there's basically one earthquake of less than magnitude six and at this point two earthquakes have greater than magnitude nine, although nothing between magnitude 7.7 and 9. If you look at the middle histogram, which is plotted in terms of effective overburden pressure, nothing at really shallow sites, less than 20 kPa. Of course, you can always excavate and replace when it's that shallow and very little greater than 100 kPa and only one greater than 130 kPa. So at the deeper sites, there isn't any data. The question is, do soils liquefy and the manifestations of liquefaction don't make it to the surface? Or is there something about those deeper depths that prevent liquefaction? And I don't think we know the answer to that and we need more data on that. And then finally, the last histogram is kind of similar to the middle one, but it's in terms of depth to the layer that was believed to liquefied. And the greatest depth in the database at the time of the report, at least, was 12 meters below ground surface. Second recommendation was that there were gaps in the database that needed to be filled, particularly, well, not exclusively, but particularly with respect to soil structure interaction effects and also in the areas that I highlighted in the last slide. And that we needed to fill this database by creating liquefaction observatories at well characterized and well instrumented sites that are located in areas with a high probability liquefaction. And then as I mentioned, we need case histories of soil foundation structure interaction effects. And, you know, part of the genesis of this recommendation is that a lot of our data is collected after the fact we observe surface manifestations of liquefaction and then we go out to the site and we make measurements and we do borings to get blow counts. We make measurements of lateral spreading, but the pre-earthquake conditions are not that well documented. So we either have to adjust the data that we collect to account for pre-earthquake or the effect of the earthquake or else perhaps look laterally for similar deposits to establish what the properties of the soil were at the time of the earthquake. So we need to create observatories with appropriate instrumentation, strong motion instruments at the ground motion and at depth, pool water pressure transducers in the liquefiable layers and maybe beneath the liquefiable layers, inclinometers to capture lateral movement, settlement probes, and then instrumentation and nearby structures at locations where there's a high probability of liquefaction and then wait for an earthquake basically. And so this isn't something that will improve our practice tomorrow and it's a relatively costly endeavor because these sites have to be instrumented and then importantly they need to be maintained. But we think in terms of the long-term prospects for prediction of liquefaction, this is the kind of data that we need to collect. This is a cross-section of the wildlife array. This is a good example of a free-field array. This is in the Imperial Valley in California. It's a site that has liquefied repeatedly in earthquakes since, well, we know of since 1979 but probably liquefied before then. And this cross-section shows a variety of strong motion instruments and poor pressure transducers around the cross-section to capture data on liquefaction when the next earthquake comes. So this type of array plus additional instrumentation on structures, on piles, on spread footings located on the liquefiable soil are what we think are needed. Third recommendation, the committee felt very strongly that the cone penetrometer test is the preferred method for making field-based estimates of liquefaction resistance. Recognizing that sometimes it may not be feasible to do the CPT because of cobbles or stiff layers that we can't penetrate. The CPT may not be the exclusive means of subsurface exploration, but in terms of applying the simplified method, in terms of getting a measure of the soil resistance to liquefaction, we believe the CPT resistance is the best available information. If you can't use the CPT, then we would use the SPT, but if the standard penetration test is to be used, the equipment setup should be well documented and as close to standard as possible. And we should make hammer energy measurements so we know how much energy is being delivered in order to make a blow count correction to the N160 value. We also think, though, it's prudent to not rely on one method to supplement the CPT or the SPT or both with other methods as appropriate. The SPT certainly has an advantage over the CPT in terms of collecting physical samples for gradation. If we have soils with fines content, the CPT only provides us an indirect measurement of the fines content, so the SPT clearly has a role in those types of soils. Shear-wave velocity measurements have become much more common and much more cost effective, and so we think it's prudent to do shear-wave velocity measurements in concert with penetration test measurements. And in gravels, we see potential for the instrumented becker penetration test. It's still a technique that's being developed, but we think in the long term, because you can't penetrate gravels with the CPT, and SPT measurements in gravel are unreliable, that the instrumented becker has great potential for evaluating the resistance of gravels in the future. Just to reiterate the advantages of the CPT, we liked it because it's less dependent, not independent, but less dependent on equipment operator or setup. It's relatively quick and cost effective. It has an ability to detect thinner layers, and if you worry about lateral spreading and you have a continuous, thin liquefiable layer that can certainly create lateral spreads. And using the CPT, it's relatively easy to measure shear-wave velocity using the seismic cone, and almost every CPT operator I'm aware of provides that option at a relatively inexpensive cost. CPT does have limitations. As I mentioned before, there's no direct measure of soil type finds content or plasticity index. For that, we need physical samples, and we can't characterize gravelly or denser soils. So those are the three recommendations on data. We then moved on to the methods of analysis and quantifying uncertainties. The first recommendation in that area is that when refining or developing new empirical relationships for liquefaction analyses, incorporate unbiased estimates for input parameters. And the genesis of this recommendation is that some of the variants of the simplified method there use biased relationships for some of the important parameters with the argument that, well, this makes it conservative. It provides a conservative assessment of liquefaction potential. It's a conservative assessment of the earthquake-induced shear stress, perhaps. The problem with that is, one, whether it's conservative or un-conservative depends on how you're doing the analysis. Are you evaluating a threshold acceleration below which you don't have to worry about liquefaction? Or are you evaluating the factor of safety against liquefaction? In one case, it might be conservative. In the other case, it might be un-conservative. But also, we think it's just good engineering practice, not the compound factors of safety and biases, that to make the best estimate of all our parameters and then apply some measure of uncertainty at the end of the analysis. So our recommendation is that we use best estimates, that we identify and where possible quantify the uncertainty associated with these estimates. And importantly, that when we're dealing with cases beyond the range in which our analysis is constrained by the data and we have to extrapolate, again, we should make a best estimate based upon principles, mechanics principles, seismologic principles, and experimental data, that we shouldn't use blind statistics to extrapolate the greater depths or higher confining pressures. But we should base those extrapolations on engineering principles. This histogram is meant to illustrate the same point that we showed before, that we have an hour range where our relationships, whichever relationship we use, is constrained by the data. And in fact, when you look at the zone in which the relationships are constrained by the data, most of the time they're in substantial agreement. Not all the time, but most of the time. However, when we use a relationship that has built in bias, it does make it difficult to assess the overall uncertainty associated with those relationships. Recommendation five, use geology to improve the understanding of case histories and project sites. Geologic context is basic to assessing liquefaction hazards. I think the genesis of this recommendation was a feeling among many members of the committee that we've moved away from considerations of geology when we're interpreting our case histories. We tend to rely more on simple soil profiles developed by borings at discrete sites. We don't necessarily take into account the types of soil deposits we have and the types of variability we're likely to have in those soil deposits. And perhaps that's a symptom of an overall trend in our geotechnical practice to move away from considering any geology when we make site assessments. Recommendation six, implement methods in the manner in which they were developed. Don't mix and match. Don't take the R sub D factor from one relationship and use that with K sigma or a fines content from another relationship. Stick with what got you there. The key relationships in a simplified method include the soil flexibility factor R sub D, the magnitude scaling factor, fines correction, and the K sigma factor. These factors are interdependent and so you shouldn't use factors from one variant for another. However, you probably should consider using more than one method that the findings of the different methods that are available are a reflection of the uncertainty in those methods. You know, and if all three methods or all the methods you use agree, then be happy, go home. If they disagree, then you need to recognize that you're probably in a zone very close to the border between liquefaction and no liquefaction. And so whichever way you go, there's going to be substantial uncertainty associated with with your evaluation. And so by using more than one method, it provides you with information that you can use with your professional judgment to make an appropriate decision. Recommendation seven coming back to this theme of uncertainty is we encourage the profession to explicitly incorporate uncertainties from field investigations, laboratory testing, numerical modeling, and the impact the local site conditions on the ground response in our liquefaction assessments. And that can be done in a number of different ways. It could simply be a sensitivity analysis where parameters are varied over the range of the expected range. Logic trees such as the one shown here on the right provide a very simple and direct way that we can incorporate uncertainty into our analyses. And there are probabilistic liquefaction hazard analyses that can be used both with the simplified method and with more sophisticated methods for liquefaction assessment, for triggering assessment. Recommendation eight is that the geotechnical profession needs to follow with our structural colleagues in developing performance-based design, analyses of liquefaction and its consequences that include structural damage potential and the direct and indirect losses due to that damage. Analysis that rather than just accounting for discrete levels of shaking, like the ground acceleration with 10% probability of occurrence in 250 years, we count for all levels of ground shaking. We characterize and account for uncertainties in the analysis. And we developed then basically a performance-based assessment that we can use to make rational decisions on alternatives. So this histogram is an example of an evaluation of five different cases, five different methods for remediation of the site, looking at both the construction costs, which are relatively equal for the five methods. The direct losses that were likely to occur, the indirect losses due to loss of business and commerce and other indirect factors, and then the lifecycle cost and doing these kinds of performance-based assessments to make rational decisions on engineering solutions for our problems. Recommendation nine, the committee recommended that we should use experimental data and fundamental principles to develop new analytical techniques, screening tools and models. That there's plenty of room for improvement in our existing models, and so that we shouldn't stop and be content with what we have, we should move ahead to develop new methods, new models, collect new data to improve our assessment. We recognize that empirical data, field case history data is essential, but it's really right now at least and for the foreseeable future insufficient to define behavior over the range of conditions that we might encounter in practice. Therefore, we should use models to extrapolate into those ranges. The models that we use should respect the fundamental principles of wave propagation, geology, mechanics, and soil mechanics. The geologic investigations can provide valuable information for our computational analysis. I'll come back to this in a minute. That experimental data can guide the extrapolation to new circumstances and that physical model testing, both element testing and larger model testing can provide useful data. So that would include centrifuge and shaking table tests of systems response, as well as cyclic simple shear and cyclic triaxial tests of element response. Again, coming back to the steam that perhaps we don't pay enough attention to geology and engineering geology. Geological investigations provide invaluable information on sediment structures, spatial variability, the age and other factors that are likely to influence liquefaction potential. And the consequences of liquefaction triggering. Recommendation 10 was to continue developing and validating computational models for liquefaction analysis. We're starting to see computational analysis being used more in practice, whether it's FLAC or Plaxis or some other numerical platform. These models are really kind of first generation numerical models for numerical prediction of liquefaction and lateral spreading, lots of room for improvement. Computational models are valuable tools. We can use laboratory and physical model tests to validate these computational analyses. And there are some promising new methods under development for simulation of granular flow and these should be extended. And a lot of this is being done in the engineering geology space to characterize landslides and flow slides. But we felt these models should be extended, validated and calibrated for practical prediction of lateral spreading and flow slide behavior in soils. And then finally, recommendation 11 that there's plenty of needs still for fundamental research on the behavior of soil subject to liquefaction. We should continue to devise new laboratory, physical model, experimental techniques. We need an improved understanding and quantification of post-triggering behavior of soils. And particularly the degradation in the soil fabric following triggering liquefaction is very poorly understood. And that dilation induced stiffening, which a lot of models don't take into account after the soil liquefies is poorly documented. So that very briefly in a nutshell are the 11 recommendations in the report. You can find a much more detailed discussion of those recommendations in the report itself, which is available online as a free download. And I think we're ready for the Q&A. So there's a Q&A box in the toolbar at the bottom of your Zoom screen. You can click on that and send in your questions if you haven't already. And I guess, Alan, you're going to moderate and throw the questions out. Don't throw them out, but hand them out. Thanks a lot, Ed. I was having a little bit of issues with my Zoom and I got kicked out right near the end of your talk. So I lost all of the Q&A pretty much from my screen. So we need to tell Zoom that the Q&A should stay up even if you get kicked out. But I am already going to type some of them or you can say some of them. I know that I remember some of the questions from the top because I was looking and trying to make sure I had some answers. Do you want to at least, you can see the Q&A, right Ed? I can. So the first one says UCLA is working on the next gen liquefaction database. I don't think it's just UCLA. I think it's a community effort. Well, and I wouldn't say it was in response to recommendation one. They were already working on that NGL, I think as our report was in preparation. The question is what other responses have you seen to your recommendations since the publication of the report? Good question. Personally, I haven't seen any. Alan, are you aware of anything that's been done in response to our recommendations? I can't say that I've seen anything directly other than, you know, perhaps people submitting proposals trying to address things here or there. I think, of course, these things are all happening parallel, as you mentioned. The fact that we brought up geology, I think a lot of us have felt that and are starting to use that more in our analyses. But it's going to take a little time for that to become pervasive. Okay. This is Marty. Let me just chime in on the response to that. I have seen in the context of consulting practice where engineers out there who are dealing with a particular issue have referred to the study directly and taken to heart the recommendation to use multiple methods and to address uncertainties. Both in, I'll say, fairly straightforward assessments as well as more detailed assessments. So I haven't seen that happen a number of times here over the last year or so. Good to know. All right. Next question. If a site has liquefied repeatedly, shouldn't it have settled densified so that it is not likely to liquefied in the future? A short answer to that is no, and there are several reasons for that. One has to do with, after site liquefies, as the poor pressure migrates upwards, it can actually loosen overlying soils rather than densify them. Another is that an important factor in increasing liquefaction resistance is the age, how long the soil has been under confinement. And liquefaction resistance increases with time, but there's evidence that once the liquefaction is triggered, that clock is reset. So there'll be some loss in liquefaction resistance due to the loss of the time effect. And also, there may be multiple layers within a soil profile that are susceptible to liquefaction, and so maybe the layer that liquefied in the first earthquake will have settled and densified, but there may be other susceptible layers above it in the profile that were isolated from the shaking at the time that lower layer liquefied, but that could liquefy in a subsequent earthquake. Ellen, you want to add anything to that? No, I think those are all great. I think that is the next question about the Chinese cone. Yes. Okay, because I took some time to make sure I reminded myself what we said in our report. So for gravels, do you recommend the use of the Chinese cones as kind of one of those large diameter penetration tests? We discussed that in the report. With all those large diameter penetration tests, for instance, some of them, they're trying to relate it back to the SPT and there's uncertainty there. Probably the best approach is to actually directly use penetration resistances from those tests and develop a direct correlation with liquefaction. The problem, at least right now, is there is not a lot of data available to really fully develop a relationship as we've done with the SPT. So I think gravels is certainly a place where there's a need for more research. And I think the second what Ellen said, one of the challenges is we don't have that many examples of gravel sites that have liquefied or that have not liquefied. And so it's the same issue with the instrumented becker. There's just a limited number of sites where we can go out and do testing and use that to develop a correlation. And I'm not sure that we have enough to develop a robust correlation. All right, next question. Do you see a greater improvement in our current simplified liquefaction assessment procedures? If we fine tune some of the current factors, e.g., overburden correction, fines content, magnitude, or if we add an additional one, seismic duration, relative density, or others. So my personal opinion is I don't think we really want to add factors. Certainly we already capture relative density with the blow count. Duration is captured via magnitude. However, there may be some benefit in looking at alternative measures, particularly in terms of the earthquake loading. I know Steve Kramer at Washington has done a lot of work looking at the cumulative velocity, absolute velocity, as a measure of liquefaction resistance. And so maybe some value in looking at alternatives to the cyclic stress ratio as a measure of liquefaction. Ellen, I think you have some thoughts along that line, too. Yeah, no, I completely agree. I think as a committee, we really wanted to challenge the research community to rather than tinker, tinker, tinker with this really complicated simplified method. Rather than doing that, let's challenge ourselves to change the paradigm to really improve liquefaction assessments. And that may mean using completely different approaches, strain, energy, dynamic analysis, or things like Ed said, looking at the seismic loading a different way, looking at the full profile, not just a single layer. But that's going to take time for the research community to fully evolve and develop and assess those new assessment techniques. Okay, next question. What do you see as missing? Oh, did you want to say something? Oh, you're doing the questions. No, that's okay. You're still not to the question I can see, but I think we address that. Is there something missing? Basically, we address that. And again, to emphasize what Ellen said, you know, it's called the simplified procedure. That doesn't mean it's simple and making it more complicated isn't going to help. All right, that's the geology. I love this question because it's not really a question, but I want to echo it says as a geologist, I totally agree with recommendation five. And I'd like to add the factor of geomorphology. And I think we all should emphasize that as practitioners in geotechnical engineering, really bringing in the geologists to interpret what they can see from the landforms, the geomorphology, the geology of the area really can provide insights to the extent say of lateral spreading, etc. So, and I think that's we didn't use, we probably should have said geomorphology. We're talking about when we say we haven't used geology enough. Yep. Yep. Go ahead. The visa best approach are not usually used in practice or not recommended. I'm not sure exactly what's meant by that question as it's not in the form of a question. But visa best is used in practice in my experience quite a bit. We always use visa best as a supplement to our convention liquefaction assessments. Ed, I think I think you said that it's often not used in practice, at least for and so I think the question was, do we think it should not be used or it's just usually not used and I think all available data should be used. So yes, the BS approach should be part of the part of the analysis and I can now see the rest of the question. But what I meant to say is if I said it's not used is that it's not used as much as it should be used. Okay, very good. So the next question is about liquefaction of unsaturated soils. We discussed this a bit and in particular the increasing role of P wave velocity measurements. To assess whether the soil is saturated, not just assuming it's below the water table and therefore it is saturated. And so that's one thing that, you know, can be added to our site characterization in terms of P wave velocity measurements. But towards the so we did discuss the liquefaction and unsaturated soils and I think there was general acceptance among the committee members that with even very small degrees of unsaturation. The liquefaction resistance increases significantly and we pointed to the data in Christchurch where soils in the zone of water table fluctuation essentially did not liquefy and that's attributed to the fact that they were not even though they were flooded and below the water table they were not a full saturation. Okay, so the next question is about dilation of silt material and strength loss. What strength loss can be estimated for designs of dilative silt? You are absolutely right there was not significant discussion of silt in the report because there has not been as much attention perhaps as there should be on silt. The question is what exactly do you mean by dilative silt? I mean if they're dilative they will have maybe just some minor strength loss. But they should eventually dilate and therefore not be susceptible to things like flow failure, etc. But I think silt are certainly a place where people are starting to get into more research. And I think that was part of the reference we made at the end to the fact that current models don't include post triggering strength gain due to dilation. And that we don't have all the data we need on how soils behave when subject to cyclic loading and silt in particular the data is lacking to give a true understanding of the post triggering behavior of those soils. Okay, so the next question is does the study discuss or is there a source of information that discusses geological depositional mechanisms for liquefaction susceptibility? It's discussed a bit in the report. Most of that work I would say references back to the work done by Yaud in the, I think that's the late 70s. There's probably been some updates since then, but I think it's still relatively accurate what was developed back then. Ed, do you want to? I would agree, but that points to what we talked about as the deris of consideration of geology and geomorphology. It's been 30 years since anybody took a hard look at that with the exception of the age of the deposit, which has gotten a lot of attention. Thanks to the work of Ron Andrews and Ken Stokey and others. The next question is what software and method is recommended for a modern large scale liquefaction analysis using current ground motion data? That's a big one. Well, we didn't make any, we did not make any recommendations. That's exactly what I was going to say. We're certainly not in feel it's our responsibility to indicate what software is appropriate. One thing we did note that if you're really trying to look at large deformation problems, even using the finite element method in large deformation mode is not as good as being able to use techniques that can actually model flow, say the material point method or other new type of methods, and we discussed some of these kind of new ways of looking at things if you're really worried about large deformation flows, things like that. I think in a discussion of numerical modeling also there's a strong suggestion that the skill of the modeler and the calibration of the model is probably more important than what particular model you use. Different people using the same program get widely different results that any time you're using a program to predict liquefaction and consequences, you need a calibrated against benchmarks and that there's a certain amount of operator skill required to use these programs properly. They're very powerful programs and when used properly, I don't think we had a preference of one over the other. Great. The next question is about the Becker hammer. What is involved in the instrumented Becker relative to Becker that's done in the past? I think the key issue with the Becker has always been what energy is delivered actually to the sampler itself. So the instrumented Becker makes the measurements and if I remember correctly, it's not only making the measurement through what gets to the top of the drill rods, but actually what gets down there to the sampler. A couple things. One, right now it's a research tool. The only person I know has the instrumented Becker is Jason Deon at Davis. The measurement is made at the tip. In the standard Becker, the measurement is made up at the hammer level of how much energy is delivered. There are so many sources of energy loss between the top and the tip, even more than in the SPT that it's really unreliable. So in the instrumented Becker, the measurement has been made at the tip, but right now it's still a research tool and there are lots of questions on converting or correlation between the old Becker and the instrumented Becker and the database that was collected using the old Becker, which is the only thing that's available to calibrate the instrumented Becker with. So right now it's still a research tool under development. The next question, what type of laboratory testing are you interested in being developed? I don't think we discussed like new types of laboratory tests. Using laboratory tests to better understand liquefaction and in particular going beyond using simple harmonic loading, so we can look at say some of those issues related to how the seismic loading is characterized. Ed, do you want to expand? Yeah, I don't think we had anything particular in mind. I think the other development that we see is kind of local measurements of stress and strain, you know, rather than just kind of global measurements at the ends of the specimen, putting sensors on, for instance, you know, on the middle of the specimen to measure shear wave velocity and strains as the test progresses. And also some of these more modern tests where, you know, maybe they use like X-ray CT scanning to look at the segregation in the material during testing, rather than considering the sample as a homogeneous sample, taking into account shear band formation, necking and other phenomena that occurred during the test that undoubtedly affect the results. So, Ed, we've got 32 more questions, maybe 31 and 10 minutes. So I'm not sure how we want to try and, we're not going to get through all of these. What do you suggest in terms of, is there a way we can prioritize this? Or just go until two and then we'll take it from there. We can make our answers shorter. Okay. That sounds good. I'll note that problem, that, that's a great idea. So, timer frequency for measuring SPT hammer annually six months or prior to a site investigation. I think you certainly have to do it at least prior to a site investigation. And I mean, you can have it right on the system there and you could be doing it multiple times in a day, I think. So, I think you want to make sure that you stick with some good energy measurements, so as often as possible. Next question, recommendation four, why are bias relationships make conservative results? It's not that bias relationships always make conservative results, is that often researchers would select a bias data to make it conservative in their assessments. So, it's not always that way, but it seemed in this case where people made those decisions, it was always to make things conservative. For deeply, yeah. Excuse me, Ellen, this is Sam, you know, I just want to remind our panelists that we need to make sure that we're clear whether something is in the report versus your professional opinion. Okay, great, great. I think that one was clear, but thank you, Sam, for that note. For deep liquefaction, given the absence of sufficient data, what would your suggested empirical method be for lateral spreading? So, I would say we did not discuss that specifically in reports, and I think probably most of the empirical methods can't be used. I mean, if you're going to use an empirical method and completely extrapolate it well below the depth that the dataset that was used to develop it, that would be very difficult, and you would need to perhaps do some sort of more sophisticated analysis. Because as it's deeper, the material has to be able to move somewhere, so just by applying the techniques, it's hard to see how those empirical methods would be appropriate. Ed, what I'm going to do is if you want to chime in, you can chime in, but otherwise I'm going to answer as best as I can. To what extent could relative density be used as an indicative figure of liquefaction potential? And if that means relative density in lieu of penetration resistance, whether CPT or SPT, I mean, in general, it's pretty difficult to measure relative density in situ, at least at depth, other than the nuclear gauge. And so I think right now, basically, when we use SPT or CPT resistances, we are basically using relative density. Let's see. We don't see any questions in our Q&A screen. I'm not sure if that is pervasive, but I'll make sure I properly read the questions. There has been a Berkeley study of BHP, it correlates to equivalent SPT. However, Hammer Energy and Becker may need, what do you think? I think we addressed the Becker already. Thank you for that. Based on the state of the art, to what depth should liquefaction be considered? Ed, you want to take that one? I don't think we postulate any limits. There are curves that have been extrapolated out to higher blow counts that you can use to assess deeper layers, and eventually you'll get deep enough where it won't liquefy in all the methods, but we didn't put any limit on that. Okay, thanks. Did the report define liquefaction based on which data is to be collected? In other words, we do not want liquefaction definition to become a variable that we have to accommodate in analyzing and interpreting the data. I'm not sure I understand that question fully. Do you, Ed? No, all I would say is one of the things we recognize in the report is that most of the case issues for liquefaction are based on surface manifestations of liquefaction, sand boils and lateral spreading. Right. And so that's basically the definition that people have been using. And we had a more, you know, academic definition for laboratory testing, close to zero effective stress state and a loss strain level. Yeah. Has the data from the Tohoku earthquake been put together? Were some international or researchers or just the Japanese? I think it's been predominantly the Japanese with some interactions with the U.S. I think the NGL project is trying to bring some of the Tohoku data in and hopefully it will, although it's not been as pervasively shared as, say, the Christchurch data. Give some examples of built-in bias for the field and lab data. I would say going back to this built-in bias, one of the examples were if you, for the R sub D parameter, if you say I'm going to use the upper bound of the R sub D that I computed, you're providing a higher value of seismic loading. So there you go. That's a built-in bias. You're always going to overestimate the seismic loading. But then if someone else decides, well, look, I'm going to go do site response analysis to get my R sub D factor. We've got a mismatch there, and you're not, you may indicate something will not liquefy what maybe it will or vice versa. As a seismologist, I'm wondering what kind of ground motion prediction models are needed by geotechnical engineers to be able to compute the probability of liquefaction. I think Ed already indicated that, you know, CAV 5, which is something that Steve Kramer has looked at, is certainly a good potential value or parameter, and people are developing those GMPEs. But I think that there are many, it just depends on what level of analysis you're going to do. There are probabilistic versions of the simplified method that just require the peak ground acceleration. You can do probabilistic analysis with a suite of time histories, you know, doing kind of Monte Carlo type simulation or point estimates. And there are explicit probabilistic models that probably that might use something like, you know, RMS and duration. So, yeah, I wasn't sure that they meant to compute the probability of liquefaction. You're absolutely right. Any ground motion parameter could be could be used in a probabilistic model. Where do we find open databases with laboratory test results? That is a great question. There are not very many. I think NGL is trying to do that. NEES had, you know, at NEES projects that had data would have published that data. I know there are like some of the tests run by Shadi El-Motar for a NEES project he had with University of Washington was in the NEES database, which is now in DesignSafe, which is the next cyber infrastructure. Component after that, but we need people to do a better job on publishing their databases so that other people can take advantage of them. So you mentioned using different methods to assess liquefaction or even use the data from different tests. Clearly there will be discrepancy. What do you do? Go with the worst case scenario, statistical analysis. How do you want to interpret the findings? That's a great question. I think that's why we get paid the big bucks. If it was so simple, that wouldn't be the case. So I don't think going with the worst case scenario is necessarily appropriate. I think that's a matter of engineering judgment for the individual engineer to decide how he's going to wait to different methods. And she may also want to consider the consequences, what structure are you analyzing, et cetera, whether you feel comfortable with more statistical analysis or a worst case scenario. If the consequences of liquefaction would be so severe to a very critical structure, you may decide something differently. How promising are the SWV? I think that's the... Sheerway velocity. Ah, okay. Thank you. I thought it was the Swedish one. Given some inherent limitations such as small strain measurements, less sensitivity to subtle changes in velocity, aging effects. I think those are certainly limitations. In my opinion though, shear wave velocity still has that two things going for it. Number one, it is a true engineering property, not just some empirical measure of relative density. And two, you've got to get through the small strains to get to the large strains. And so I think it certainly has a role, but there needs to be more work done to better understand some of these issues and how they manifest in the prediction. Any recommendations on liquefaction of improved soils? We didn't really discuss that, Ed, right, in our report? Yeah, no specific recommendations. I guess the assumption was if you're improving it, then you'll improve it to a degree that we wouldn't worry about it. Is there any experience in taking values of VS30 from MISW to correlate with liquefaction? I would say we did not discuss that in the report. I don't think, you know, the liquefaction approaches have been developed based on this idea of a critical layer, which is obviously not the average VS over 30 meters. But I have seen some work where people who are looking at regional assessments, like rapid regional assessments of liquefaction, are using VS30. And I think that's probably the best place for using that type of value. How well is our understanding of cyclic softening of clay-like soils developing alongside liquefaction of sand-like soils? It wasn't part of the scope of the report. There you go. Thank you. That was easy. In Greece, we have reported several liquefaction manifestations in the last decade. Is there any way to upload these data to a global database? I think that's a wonderful idea because we want to go beyond just the U.S. and say Christchurch. And I think getting in contact with the folks who are doing the next generation liquefaction NGL project. John Stewart, Scott Brandenburg at UCLA are kind of leading that effort. Steve Kramer as well. And email them and they will have, they are working on a website where you can contribute data to the database. But I don't think it's quite ready yet for taking that. But it should be soon, I hope. Mention that it would be beneficial to have observatories for earthquakes to assess liquefaction during earthquakes. Technology should make this increasingly possible. Have you assessed approximately what it would cost to purchase and install such observatories? We did not scope out that cost. And that was explicitly forbidden in putting in the NRC report guidelines that we were not to address the cost. Right. But the biggest issue with having these observatories is often the maintenance once you even install them. Because things break and earthquakes don't happen that often. And so that's something to consider. What is the current thing? Yeah, this is Marty. Let me interject where we've needed our one hour duration by a minute. So maybe we should bring things to a close. Is there I'm trying to figure out it because we is there a way that we can answer these questions. Another way, Sam, like if it's recorded, we could close this off or through the text. Is that saved or what? What we might do and we can talk about how feasible this is later is we could, when we send out an email with information about where you could download slides or look at the webinar again offline. We might try and answer some of those questions in writing. I don't know. I don't think we can do that. And this is with the National Academies here and Zoom should be recording all of these questions that are being asked. And I've been trying to also capture them as well. So we do have a recording of what has currently been asked in the Q&A box. We can go back and answer these in another format. Yeah. We'll work that out. I think the most important thing, Sam, because I saw the question asked very many times. Will this be slides available, etc. This is all being recorded. But the question is where will it be posted, Sam? Remy will answer that. Yeah, so we will send an email out to all of our registrants with a link. If you search the committee on geological and geotechnical engineering, you will find a website that has also some past webinars. The webinar won't be posted immediately. We just need to do some light processing on it, get it on YouTube and upload our website or update our website. But again, we'll send an email out to our registrants with the link. Great. Wonderful. Will there be any place online, like wherever the report is? Can you put a link to something like that? I can find out about that. It'll be on the Kaga website. I am not sure if we... I'm not sure if we can post it with the report itself. All right, but that's good to know. The Kaga website as well. Okay, great. Well, I want to thank everyone, first of all, for being here and also asking so many questions. We're concerned whether we would get any questions, but obviously, liquefaction is a hot topic that you all deal with quite often and you had lots of questions. I'm sorry, we didn't get to more of them. And so I just want to say thank you to everyone. Ed, Marty, do you want to take it? Nope. Nope. Just I'll reiterate that. Thank you all for participating. I hope it was helpful and the committee will be having more webinars in the future. And those who have registered will be on our mailing list. So you'll be getting those announcements. So thank you again for attending.