John Malusky and Pete Holter join us to discuss details of the recent Soil Classification and Compaction sample rounds 185 and 186.
AASHTO re:source Q&A Podcast Transcript
Season 3, Episode 6: Digging into Soil Classification and Compaction Samples 185/186
Recorded: May 19, 2022
Released: June 14, 2022
Hosts: Brian Johnson, AASHTO Accreditation Program Manager and Kim Swanson, Communications Manager at AASHTO re:source
Guest: John Malusky, Proficiency Sample Program Manager at AASHTO re:source and Pete Holter, Senior Quality Analyst at AASHTO re:source.
Note: Please reference AASHTO re:source and AASHTO Accreditation Program policies and procedures online for official guidance on this, and other topics.
Transcribed by Kim Swanson and MS Teams.
[Theme music fades in.]
[00:00:00] Announcer: Welcome to AASHTO re:source Q & A. We're taking time to discuss construction materials, testing, and inspection with people in the know. From exploring testing problems and solutions to laboratory best practices and quality management. We're covering topics important to you. Now, here's our host Brian Johnson.
[00:00:21] Brian: Welcome to AASHTO re:source Q & A, I'm Brian Johnson.
[00:00:24] Kim: And I'm Kim Swanson. And what are we talking about today, Brian?
[00:00:27] Brian: Well, today we are going to talk about the last Soil report that came out from the Proficiency Sample Program. It's Soil Classification and Compaction 185 and 186 that came out on May 19th, 2022. So, the reason why I want to talk about this is because there were some changes to this sample round from the previous sample round and I want to talk about some of the data. Because there are some things that have changed over the years with the soil samples that people who work in standards development are keenly interested in. To talk about this, I've invited our Manager of the Proficiency Sample Program, John Malusky, on board here. Hello, John.
[00:01:13] John: Hey Brian. Thanks for having me.
[00:01:14] Brian: And our, uh…I don't know what I would say expert. On umm, data analysis. Maybe. Pete Holter. So, Pete.
[00:01:25] Pete: Expert on expert on details that no one else cares about [Brian: Yes, that's what I was looking for.] that that's an appropriate title.
[00:01:32] Brian: Expert on the emails and what else cares about. Yes, thank you. It's so much easier when you say it first because then I don't feel like I'm insulting you. So. So thank you so much, Pete. Pete gets into the weeds and it's always helpful. And I mean that from the bottom of my heart. Pete brings stuff up and finds things that always we learned something important, and we didn't even know that we were going to learn it. So, thank you, Pete. OK, so this sample round just to give us a brief overview, we have two different hydrometer tests, right? We've got the old traditional T88/ASTM D422 that's now defunct. And we have the new one, which is ASTM D7928.
[00:02:19] Brian: We also have a few other tests we're going to get into, but let's start with the hydrometer, John. We've had a couple rounds now where we were able to look at how performances between the old hydrometer and the new hydrometer. Can we draw any conclusions about whether or not this is an improvement in the variability of testing at this point? What are your feelings on that?
[00:02:46] John: Well, I think, Brian, right now it's to me it's kind of inconclusive. When I look at the data, yeah, actually have the sample round analysis pulled up here for this current round. Just taking a peek at it and looking at their traditional 422, particle diameters, right, and this is what AASHTO has specified for PSP. So, we've looked at the total material finer than the .02 millimeter, then the .002 and the .001. Now the new hydrometer analysis. Right, you look at your particle diameter at time intervals as sedimentation. And then you also look at the percent passing that particle diameter at that time frame. And when you look at the comparison, I don't see that much of a difference between the percent passing and the two, just as example. And I don't know how it's going to play out with our listeners, but at the particle diameter of smaller than .02 for traditional hydrometer, we saw an average around 47% from old 422. And if I look at the particle diameter size at 4 minutes, that was a particle diameter of .0208. So, we’re 800 thousands different and particle diameter and our percent passing. Went from 46 roughly 47 to 48.3 like I right. I mean, and the standard deviations standard deviation 6%. It's just, yeah. It's, I can't, I don't know what I.
[00:04:14] Brian: All right, John, you, I got to stop. You put Kim to sleep. You're putting Kim to sleep. We need to. [John: OK.] We need to write the ship.
[00:04:22] Kim: I was very confused on all of the things that you're talking about. I was trying to hang in there, but I got. I got nothing. [John: OK I'm sorry, I don't know how to.] So sorry about that.
[00:04:33] Brian: Alright, let's normalize this a little bit and talk about the percent 1S values. Oh, Kim is very excited now. [John: But didn't change her stance at all.] So, we can kind of see at least the relative difference in test results when we talk about that. At least that's the way I look at it. I look at the variability and I see those percent 1S values and that gives me a kind of a conceptual idea of how they are doing, when I look at individual points on that curve. Do you do you agree with that or am I just looking at the wrong way?
[00:05:00] John: I don't agree with it at this point. Like I said, when I look at the values for the particle diameters and then the percent passing that those particle diameters, the set diameters that we had and the original 422 or 288, they're still pretty close and the 1 S values are almost the same and in some cases they're even larger at that particle diameter. And I think Pete, when you and I were talking a little year two ago, we were kind of seeing that, that it seemed like the variability was larger with the new 422 method.
[00:05:28] Brian: Wait with the new. So, you talking about with D7928 [John: Yeah.], you're seeing larger variability and you determine that based on which number are you looking at?
[00:05:26] John: So, the one that I looked at for this past year was line item number 4 on the data sheet and comparing that to line item number, let me sneak here 12 on the data sheet.
[00:05:50] Brian: OK, but which column?
[00:05:51] John: So, I'm looking at the sample round analysis. So, the grand average values for 185 and 186 on view sample round analysis. But if you compare the percent passing at for line item 4. And compare this percent passing on li ne item 12. They're pretty close to the same, but that's the same particle that ammeter or very close to the same particle diameter.
[00:06:15] Kim: So, if I'm reading this correctly, we're just going to throw this out here on line number #4. We have 46.82. And then then 48.56. [John: Correct.] Those are the numbers that you are saying that are close together, that it's not that big of a difference.
[00:06:31] John: No. [Kim: No, what?] Well, they should be because that's the same sample.
[00:06:33] Kim: OK, so that should be good. OK.
[00:06:25] John: So yeah, so what you're looking at, Kim, if you look at that it says total material smaller than .02, so that .02 is the particle diameter.
[00:06:43] Kim: OK. Yeah. Mm-hmm.
[00:06:44] John: OK. So then when you Scroll down to…. Uh, so you got to take this in three bits. OK, so the particle diameter for line item 7, [Kim: OK.] this is the new Hydrometer for line item 7. The particle diameter D is being reported as .0208. So, at 4 minutes of sedimentation our particle diameter is .02 millimeters roughly.
[00:07:05] Kim: Mm-hmm.
[00:07:06] John: So, to correspond with that percent passing, you got to drop down to line item number 12, because that's the percent finer at 4 minutes of sedimentation. So now at 4 minutes of sedimentation, we know that the particle diameter is at roughly .02 and the percent passing that particle diameter is 48, which if you then scroll back up to line item number 4, it's roughly 47.
[00:07:31] Kim: Mm-hmm.
[00:07:32] John: So right, we're like our percentage is only 1% apart where we got this whole new test method that's showing us pages of stuff. To come out with a difference of a like 1% but a standard deviation that's larger than what we've seen in 422 and 88.
[00:07:49] Kim: Alright, I think I followed that.
[00:07:50] John: That's where like when I look at this and someone comes and tells me from from one of the meetings that the new Hydrometer methods are great. I'm like where show me I'm not seeing a reduction in variability and the variation is not that much and how much is this actually impacting what we do every day.
[00:08:04] Kim: And to give our listeners some kind of idea of how many labs are talking about have submitted data. I see the total number of labs that have submitted data for -we were talking about line number 4 - was 1136 so that gives people an idea of how many labs data we are analyzing for this.
[00:08:26] John: Like. Ohh yeah, it's not small. I mean it's large, it's a huge huge set of data.
[00:08:32] Brian: Yeah. So, when people say when are you going to shift away from including that? What? [John: That's not going to happen. There's no way right now. Yeah, it's not happening.] You know, there's over 1000 participants. There's clearly interest.
[00:08:43] Pete: Well, one thing to think about when people are comparing these numbers is I. I saw some notes from some participants who said in their data submission that they just reported T 88 or D422 numbers in the D7928 line items. They said that, so there's probably a lot of people who are simply doing that and didn't tell us. But you know, that could be one of the reasons why we don't see much of a difference between the two. Sets of results.
[00:09:20] John: Yeah. And unfortunately, the hard part about that is we'll never know, right, because it's all about integrity and what people are actually testing and how they're doing it. We...we're not standing there watching them, submitting data and doing what they're, you know what we're asking them to do, they can do what they want and hopefully, it works out for them or doesn't work out for them. We don't know. We don't control over that.
[00:09:40] Kim: Now as the newer hydrometer standards become more popular and are specified more from project owners, do you think you will see a change in that? That more people will be doing the new methods versus the old methods and that those numbers will show more information or give us more information about how they're really different the more that they're used?
[00:10:02] Pete: Yeah. I'm not sure if we're going to see a lot of push for specifying the new hydrometer at this point. There's a lot of pushback against simply the length of the standard. It's so long and it's currently under revision to add even more pages to a standard that a number of people who are already complaining about being too long. And so I don't see, you know, there's not really. And eagerness there to adopt this new standard at this point.
[00:10:34] John: Yeah, it's almost like the preparation has confused people so much that they're just ignoring it. And I think that that's kind of where it falls.
[00:10:43] Brian: Yeah, that's a challenge. When you're writing a standard, you know you want to be precise and you want it to be as repeatable as possible. But if you get too detailed, then you run the risk of people ignoring it.
[00:10:54] Pete: I think it would be nice if that happened. If there was a push from specifying agencies to stop specifying the old standard and get everyone onto the new standard. I think that would be a good thing.
[00:11:11] Kim: I do want to let listeners know if they want more information about the transition from the Hydrometer standards. We did do a podcast episode season one, episode 24. If you wanted to listen more about the transition from one standard to the other.
[00:11:25] Brian: Yeah. Thanks, Kim.
[00:11:26] Pete: And one thing that I found encouraging was that the .001 value for D422 and T88 that the average percent passing was 11.65. And to get that, you do have to extrapolate your data beyond the 24-hour readings. And so if you compare that with the the 24 hour reading of D7928, the average for D7928 was higher percentage and that's what you would expect. So it was good to see that because I've seen in the past labs performing the test per T88 and D422 and simply reporting the 24 hour reading as the...As the percent passing the for percent smaller than 001 millimeter, so that was good to see.
[00:12:24] Brian: Yeah, that's good. So not all doom and gloom. Now that's just the hydrometer. Now there is another kind of interesting change that happened this year with the soil sample, and that was the split of AASHTO T88. And ASTM G187, what happened there, John?
[00:12:41] John: So yeah, T 288 and G 187 there we had to had to make a little switch when we put these two methods into the program last year…We kind of did our own review and talked to some industry people and we thought that the methods were identical enough to keep under the same line item. So, we had them paired up last year and looked at the data and got some additional feedback from other participants, and said that maybe you should separate them just simply because of the procedure. The procedures are different enough, you know? Specifically, T 288 calls for the minimum soil resistivity, whereas G 187 just calls for the soil resistivity at saturation…the saturation of the soil.
[00:13:32] John: We've got some more feedback with G187 and people are asking what is the true definition of the term saturation? One lab's idea of saturation can be different than another, so there has been a suggestion for next year's sample that we actually specify the moisture content of the soil prior to you doing the G187 test. So, we will see how that pans out. Looking at the data right now…hopefully, it brings things down, narrows it more in line. When you look still look at the overall comparison between the two between T288 and G187 they're pretty close the the average values are very similar, but there's way more variability in the standard deviation from the G 187. So that might be a product of the whole saturation discussion that we just had. So hopefully we can nail that down a little bit, some better instructions next year, and see how things play out there. But yeah, just had to break it apart because of some test method discrepancy.
[00:14:33] Brian: Yeah. Now one, one kind of interesting wrinkle to all this, Pete is when we have a change like this. So last year the results were combined. This year they're split out. So if someone in the future needs to test a blind sample. An extra sample that they are unaware of which sample it is for accreditation purposes, that kind of adds a little bit of complexity to that process, doesn't it?
[00:15:03] Pete: Yes, a few years down the road, then yeah, we wouldn't be able to send out older samples. You know, that's something to think about. If we have like 20 some labs getting suspended and we only have, we can only send them the new sample or we going to have enough XPS material for these.
[00:15:20] John: That's why we're only going to send them 1500 gram samples.
[00:15:23] Pete: OK.
[00:15:24] Brian: Yeah. Well, if you think about it this way, so we're not going to be processing suspensions this year. For these tests, unless they just have not been participating. But next year, if somebody had repeat low ratings. Then they would be able to test. You know, we would send them a sample that they didn't know what it was and it could be from either of the last two, not. Like the last two in the future. If that makes sense? So, this this year and next year.
[00:15:57] Pete: Although we are suspending in the case of…[John: No submittal.] No data last year and they did submit data this year, so those labs are getting suspended this year and two rounds of no data.
[00:16:13] Brian: Right. Yeah. So you do have to participate to maintain your accreditation.
[00:16:18] John: We're not concerned about supply either because we can get four of those resistivity samples out of 1 current sample. We can sub split it down since the samples are homogeneous, split it appropriately and we should be good.
[00:16:32] Brian: That's a good option as well. And I think that comes up with the pH testing. So, another test that we had on this round where we had pH testing of soils. Now, generally, there's not a lot of soil involved with that test is there and if somebody were to receive low ratings on that when they have to get an entire box of soil for that or what are we doing?
[00:16:56] John: Now we're going to prepare smaller samples. We kind of did this with the mineral filler test for the gyratory sample a few years ago. Instead of sending laboratories and entire box of gyratory materials to test a few grams of the mineral filler, we're going to kind of do the same thing with pH. So we will send participants a small bag of soil that we prepared from the previous round or current round that's homogeneous and we should be good to go. The good part is it's not going to cost the participants the full sample box. We're going to heavily discount it. I'll since it's very, very small portion of material, it won't cost that much to ship. We haven’t identified a price yet, but it will be way way less than what the full sample price is.
[00:17:43] Brian: That'll be good for anybody who wants to see how they're doing on pH. Hopefully, we won't have a lot of suspended laboratories, though, and it'll be just people who want to see how they're doing now. The next sample I want to talk about is the corrosion test. So there are two corrosion tests in this sample. We've got AASHTO T290 and AASHTO T291. Now, the reason why I wanted to talk about this is because the variability did not look so hot this time. Can you talk about that, John?
[00:18:15] John: Yeah, it actually hasn't looked very, very hot, as you would say the last two times we had coefficients of variation. Well over well actually just about 100%. So really you could be over 2 times off the average and still manage to get a satisfactory rating here from the accreditation program. So I'm not sure how we're going to evaluate that and what we're going to do in the future, we I don't know if we're going to continue to include it. The one thing that we have recently done with the assessment program is made an adjustment, so now laboratories will not have the option to use the titration method. They'll be using the meter method. I don't know what the other way to what procedure that is A or B in the method or ever but the titration method is not going to be used. So I'm not sure if that's going to change the variability. We'll have to see what happens in another year or two.
[00:19:11] John: This might be one of those things that we just kind of let fall off the books and take it away from the program after a while. If you don't see any improvement, either that or figure out a way to make appropriate suggestions to the committees that oversee the standards and maybe get some verbiage in there that will help reduce that variability.
[00:19:30] Brian: I see. I think when we see results like this, it's good that we know what they are. If they were never included in the round, we would have no idea. You could have agencies out there requiring this test and expecting results to be repeatable between testing firms and really what this data shows is that that may not be a realistic expectation to have. So I think that in itself is really valuable. And when one is one of the benefits of the proficiency sample program, now, while we're talking about this, I see a very quizzical look on Pete's face. OI, really am curious, what is rolling around in your head right now? Pete, what are you thinking?
[00:20:15] Pete: Yeah, well, I haven't really had the time to look at the data for these particular tests and I'm just kind of looking at it as you're talking and. Uh, just interesting to see if you look at last year's results for T 290. There's a lot more. Sulfate ion content in the soil and it had a very large standard deviation, but the standard deviation was. Smaller than the average value, and then this year he didn't have as much. Content the average was a lot smaller, but the standard deviation compared with the average. This time it was larger. And it's just kind of curious about the. Quality of this test and like you're saying and.
[00:21:04] John: Yeah, that's one of those things, Pete, were after a few years, we get some data and we can provide a precision estimate to AASHTO and ASTM. So we have a better understanding of where this test method fits and how precise of a method it actually is. So another few years we can get some data out and get it to the appropriate bodies or standards development and get them some more information.
[00:21:27] Kim: Are those new tests like what is this? Or is this just new a new thing that happened? Like why haven't we seen the same thing you know five years ago? If it's an older test?
[00:21:38] John: Yeah, it's not an older test. We put this into the method or into the PSP last year for Soil Class. [Kim: Oh, OK.] So, it's fairly new, but I don't believe there are precision statements in the standards for them. So, it's something that may not have been looked at and I'm I don't, I can't 100% say there are not. I don't remember.
[00:21:57] Kim: I just didn't know how new they were.
[00:21:58] John: We just had a bunch of a bunch of requests from labs to say, hey, put this in the program. So, we said alright, we can accommodate that. That's pretty easy. So we want to see what happens I mean. First two years to have over 50 participants, that's pretty good. We'll take that any day.
[00:22:13] Brian: Yeah. Another interesting thing with this is that this is a test where we know that some of the state DOTs have their own versions. Of these tests, and I wonder if any of those changes that they have made on their state method, if they get incorporated into the AASHTO method, if we'll start to see some tightening of those tolerances or on the opposite side of that is if a proliferation of state methods is causing more variability right now because everybody's just doing whatever they normally do and not actually following the AASHTO method.
[00:22:48] Pete: Hey everybody, I wanted to jump back question for John. [Brian: Ah!] On the Hydrometer so T88 and D422 [John: Oh man.] of the number 10 sieve is not suppressed this year and... The last time we reported ratings for the number 10 sieve was 2014, I think so. I was very curious to see this. I'm glad it's not suppressed this year. I was kind of curious about the rationale for suppressing it for so many years. And what changed this year.
[00:23:23] John: Yeah, so this is kind of dependent on how the materials come to us, Pete. So when we order our materials for the blending proportion, sometimes the sand that we get has a more coarse material in it and sometimes it doesn't. Same way with the soil or clay, excuse me. So it's kind of dependent on what we get from the supplier and to just it just really depends on what's in the stockpile. You know we'll order a manufactured sand, sometimes it has. 98% passing the 10 and sometimes it has 80% passing the 10. So when we do the data analysis we look at what the percent passing looks like and if it's really close to 100 it's bound distribution and it really doesn't provide any benefit to the laboratory to be evaluated for it. So that's why it was suppressed. So this year? Down 85%, so there's...A reason where we can keep that in there, it's not bound.
[00:24:24] Brian: OK, great. Well, hey, I think we covered the sample round well enough. Hopefully, the people who are participating in this soil classification and compaction sample got some good information out of this. I want to thank Kim as always and Pete and John for joining us today. If you have any questions about this round, you can feel free to reach out to John Malusky at email@example.com with your proficiency sample inquiries, or Pete Holter which is firstname.lastname@example.org. Or you can get in touch with me at email@example.com. Which I am bound and determined to get emails going into that email address one day, Kim.
[00:25:12] Kim: One day we're going to get it and I'm going to say if you have further questions about this, do not email me. I will not be able to help you.
[00:25:20] Brian: Also good to know. All right. Thanks for listening.
[00:25:23] Pete: Thanks everyone.
[Theme music fades in.]
[00:25:24] Announcer: Thanks for listening to AASHTO re:source Q & A. If you would like to be a guest or just submit a question, send us an email at firstname.lastname@example.org or call Brian at 240-436-4820. For other related news and content, check out AASHTO re:source's Twitter feed or go to aashtoresource.org.
[Theme music fades out.]