Common Quality Control and Assurance Methods
Written by Mike Smith
January 26, 2023
Keil Krieg is the Construction Materials Engineering and Testing Department Manager at Smith and Annala Engineering Company (SAECO). He joins his company's founder, Mike Smith, to discuss common geotechnical quality assurance (QA) and quality control (QC) tests. They also explore why these tests are valuable for contractors.
Check out what they have to say.
Meet Keil Krieg
Keil began his career washing dirt in a laboratory. He recalls, "I didn't know why, but they were willing to pay me to do it."Eventually, Keil's company made him a field technician, and he started doing density and concrete testing.
Since 2003, he's worked on projects including commercial and residential construction, storage pond levees and embankments at a nuclear power plant, and environmental cleanups at coal-fired power plants.
Keil has been with SAECO since the company's inception. Mike hired him because "he's seen just about every mistake testing labs and contractors make."
Geography and geotechnical testing
Mike has spent his whole career in Arizona, while Kiel has worked around the country. But according to Keil, most regions use the same test procedures.
"The biggest differences are the soil types. In the East and South, the soil is much finer, especially in Louisiana . . . we would be excited when we ran across a rock. We'd go wash it off and look at it," Keil says.
Rocks are a big deal because they affect geotechnical test preparation, soap times, and soil moisture. Moisture also correlates to geography. As Keil notes, "[In Arizona] we have coarse, granular material that won't absorb a lot of water. Across the country, you can end up with moisture contents and plasticities up into the hundreds."
So while test methods are very similar nationally, geotechnical engineers make different recommendations based on soil conditions that vary with geography.
Plasticity index testing
Plasticity index (PI) tests help evaluate compaction both during design and construction. Keil says, "During design, we're trying to evaluate what soils are out there and how to deal with them when we get into construction. The soils or aggregate base materials onsite need to meet a certain PI value."
Technicians use a -40 screen to filter the soil sample. Keil explains, "It's a very small screen . . . so we remove almost all the rock and sand from that sample. We're just focusing on a very fine portion." Technicians then moisture condition the soil prior to the test.
Differing PI test results
Some labs may struggle to produce results consistent with other labs. Here's why that happens, according to Mike and Keil.
Preparation methods
Compare two labs or technicians, and you'll see differences in test preparation.
For example, Keil asks, "Did they let the sample soak or not? Did they do a wet preparation—where you take a bulk sample, wash the rock off, and account for all those fine particles—or did those fine particles not get washed off the rock?"
Test prep methods can vary between ASTM standards, project specs, and state requirements. Technicians need to know what specification and method they are to perform.
Tools
Equipment also heavily influences results. Almost all equipment in the testing world must be calibrated, from the size of a dot on the bottom of a cup to the width of a groove tool.
Keil explains, "If I have a groove tool that makes a fatter groove on the soil, it will take longer to come down and make the technician think it's a higher plasticity material. If I have a really thin groove, that will close and look like the soil has a really low PI. So you need to have the properly calibrated equipment."
Technicians
Good, experienced technicians stay aware of varying prep methods and equipment calibrations. They also know how to conduct the test itself properly. For instance, they don't hold the base of the equipment, which changes the impact as they crank material through it.
Sample variability
Contractors or project owners often request samples from a specific spot, but maybe QA already sampled somewhere else. The test results may differ simply based on where the technician put their shovel. So it's important for technicians to sample similar locations to avoid variance.
Keil notes that, "If a contractor wanted to reduce some of those variables, they would have the QA and QC lab sample together. Or they could blend a sample and split it down the middle so we could be sure we're getting the same representation of the onsite product."
Sieve analysis testing
Technicians do sieve analysis tests on multiple samples during the geotechnical (design) and construction phases of a project. They typically use a -40 sieve screen to confirm that the aggregate base for a slab or pavement meets the gradation requirement.
Sometimes sieve tests can be a source of contention when labs get different results, but these discrepancies can generally be solved easier than in other tests Mike and Keil discussed.
Nuclear density testing
In addition to lab testing, geotech engineers also run field tests to help verify the level of compaction the contractor is achieving with the soil.
Often, the minimum percent compaction is between 92% and 98% of the laboratory reference value. That value comes from a laboratory proctor during the design phase. However, it's not an exact specification.
Mike says, "When we get into the construction phase, we go out to the jobsite to test exactly what that percentage should be. One of the primary tests we use is the nuclear density test."
How nuclear density testing works
To perform nuclear density tests, geotech engineers drill a pin into the ground and set the testing device over it. Then, they insert a nuclear particle, which gives off radioactive elements. A receiver in the nuclear gauge detects the level of radioactivity.
"The less radiation gets back to the receiver, the denser the gauge thinks the soil is,"Keil explains. "It's just like the difference between wood and concrete. The more soil you smash between the radioactive particle and the receiver, it will translate that into a higher density."
Test site preparation
For accurate results, technicians must prepare the site properly in an area that represents what the contractor will face on most of the site. Keil says, "If you have a 70,000-square-foot subgrade, you don't want to necessarily go over and test in a puddle by the water truck."
The testing surface must be flat, with no voids under the nuke gauge. Even something as small as tire tracks can result in a false reading of a "lower"density than the soil really has. The nuke gauge must have full contact with the soil.
Common nuclear density testing problems
Rocks can throw off nuclear density tests. "You'll see someone pounding a pin, and the pin struggles to get through, then breaks free. [Then] I know a rock is under there, and I might need to move my test because that just created an unseen void under the testing site," Keil says.
Technicians also need to zero their testing gauges daily, because the radioactive element decays and reduces its size daily. Failing to do so can lead to inaccurate results.
Weather conditions like heat and humidity can also be problematic. If the weather changes dramatically during the day, Keil notes, "you'll probably need to re-standardize your gauge that afternoon."
What's really wild is that even moving the gauge down into a trench can impact the reading. That's because the soil in trenches has typically been moisture-conditioned, and that moisture is evaporating.
"The gauge will read the moisture surrounding it. So we standardize the gauge in the trench to account for that,"Keil explains.
Rocks and geotechnical testing
Rocks may seem simple, but after about five minutes in construction, you know how much they can complicate your work. Here are some common ways geotech engineers help contractors combat rocks.
Rock corrections
Proctor tests provide a moisture density curve. However, engineers remove the rock from the sample prior to lab-testing it. "Since we performed the proctor without the rock, it's no longer the same material that's in the field," Keil says.
To help determine the soil's true density, technicians perform a density test onsite. Then they dig up the testing area, weigh all the soil, and break it over the same screen they used in the proctor. That lets them mathematically recount for the rock, which is typically denser than soil.
Keil explains, "When we have a proctor value that's, say, 120 pounds but we've removed the rock, we need to reinsert that rock. The rock correction will increase that proctor value [to make it more accurate]."
Specific gravity
Specific gravity helps determine the rock density onsite. As Mike points out, "All rock is not equal. Volcanic rock is more porous and less dense than quartz or river rock."
Specific gravity numbers are based on a relationship with water. At a certain temperature, water's density is 62.42 pounds per cubic foot. To find the relationship of the rock to one cubic foot of water, the technician takes the rock into the lab and weighs it underwater. That gives them the specific gravity value.
"2.65 is a typical value here in [Arizona]. That means the rock is 2.65 times heavier than one cubic foot of water . . . Gold is a 19 specific gravity," says Keil. So one cubic foot of gold weighs nearly 1200 pounds, but a typical cubic foot of rock in Arizona weighs about 165 pounds.
Mike observes, "As a contractor, there's some risk with just assuming a specific gravity because it does change."
Different sites (or even different parts of a site) can have different types of rock that affect construction. That's why geotechnical testing on every job is so important.
Sand cone density testing
Mike introduces the sand cone density test as "another in-place density test that usually marries up with the nuclear density test." As with any other geotech process, sand cone tests must be performed in a certain way to get accurate results.
Here are three ways technicians and contractors can help ensure good results.
Calibrate equipment
Technicians must calibrate every piece of geotech equipment—from scales to moisture-detection devices. They even have to calibrate their sand to meet proper grading requirements.
Even something as seemingly small as transporting the equipment can impact the test results. "Those sand cones roll around in the backs of trucks and [basically come un-calibrated]. That can change the results you get," Keil explains.
Gain field experience
"You don't just send a first-day technician onto a jobsite to run a sand cone. That normally doesn't end well. You want them to have some experience under their belt,"Keil says.
Look for a technician who knows how to calibrate the equipment and set up a proper testing site. (Just like with nuclear density testing, sand cones require a flat area.)
Create a proper test environment
Contractors, this is where you come in!
Operating heavy equipment near testing sites can cause inaccurate readings. The vibrations from the equipment essentially make the sand inside the testing cone consolidate.
"Mathematically, it looks like there's a larger volume when really less soil came out of [that testing site],"Keil teaches.
That may sound insignificant, which is why many contractors let machines run near the testing site. But it's a huge deal. "It's often a struggle to ask a contractor to stop rolling their water truck or other equipment, but that impacts their tests in a negative way," Keil says.
The best thing to do is keep work as far from the testing site as possible—or better, shut your machines off for a few minutes. Yes, you may lose productivity for those few minutes. But you'll avoid hours of rework due to faulty test results during construction.
"There are just so many different pieces to getting a final number for your contractor and giving them a positive result," Keil notes. Remember, you are one of those pieces. You have to help the geotech help you.
Varying density test results
Sometimes, QA and QC labs can't get their numbers to correlate well. According to Mike, "If a contractor hasn't had this issue yet, they're going to."
QA often fails more tests than QC. For example, let's say the QC technician gets passing results on a subgrade, but the QA technician says it fails. Now the contractor can't pour the concrete they need, and they're facing a two-week delay as they bring in different equipment and try different moisture conditioning.
When two tests just don't marry up, technicians and contractors have to ask a series of questions to identify the problem:
- Is the technician qualified? Can he run the test properly?
- Is the testing hole the proper size, width, and depth?
- Is the technician preparing the sample for lab testing properly?
- Is the testing equipment calibrated properly?
- Is the technician meeting ASTM or other requirements?
- Are technicians sampling from two different parts of a jobsite where there are different materials?
To answer these questions, QA and QC must work together and be as transparent as possible to root out the issue. Just like communication between geotechs and contractors, communication between QA and QC is essential to achieving the best results. Openly sharing mistakes and variables helps all the geotechs involved get past the issue and get the construction project going.
It also helps when contractors know what to look for and what to ask—you can be a uniting force to help QA and QC align.
One-point proctors
To calculate percent compaction, you divide the in-place density derived in field tests by the maximum dry density derived in lab proctors.
But Mike says, "Sometimes we don't have the luxury of having a full laboratory proctor. The shortcut method to obtain the maximum dry density is called a one-point proctor."
One-point proctors rely on a graph full of compaction curves that represent the densities and moisture contents of most soils you'll typically encounter. Those soils all behave differently during compaction—meaning clay, sand, and gravel have different compaction curves. Those curves help technicians determine the percent compaction.
Keil teaches, "We take an unknown soil, break it down a little, and pack it into a compaction mold with a known volume. We weigh that and measure the moisture. We convert that into pounds per cubic foot and graph it on this chart. It will fall on one of those curves or between these curves, which gives us that soil's typical maximum density."
Technicians can then insert that number into the equation to find the percent compaction.
That said, one-point proctors are not 100% accurate. They just give you a quick, ballpark idea—so don't base a whole project off a one-point proctor. The technician still needs to run lab tests to find the soil's exact maximum dry density.
"I would never pass or fail a test based on a one-point, but I can at least tell the contractor, 'I think you're close. You might be good.' Or, 'Let's give it another couple hits and see if we come up.' You can at least get within a couple percent of what the actual compaction will be,"notes Keil.
Test turnaround times
Contractors need QA and QC testing laboratories to turn data around quickly so they can review it. As Mike points out, that affects their decisions about equipment and sometimes even the project specs. So labs have to ensure technicians quickly, and effectively capture testing data for the project team.
According to Mike, "There's nothing worse than having a project effectively complete and everybody just now seeing issues with the test data that could have easily been corrected while the work was taking place."
Your project should list requirements for how long labs can take to turn around their test data. Some projects even have a fine mechanism in place—so if labs don't turn in their paperwork to contractors within say, 48 hours, there is some cost pressure at that point on the laboratories.
As a contractor, knowing and communicating the specified turnaround times for testing your projects is important.
Those are the biggest takeaways for contractors when it comes to any part of a geotech project: know and communicate. The more information you can get and the more you can talk to geotech firms and project owners about what's happening in the field, the smoother testing will go.
