Rutgers test beds find weaker treated sediment than prepared samples

Rutgers University researchers reported on September 30 that cement-treated dredged sediment in outdoor test beds in Piscataway, New Jersey, developed less strength over four weeks than smaller specimens prepared by conventional methods and cured alongside them. The comparison matters because engineers use prepared samples to assess whether a ground-treatment mix can meet a project's requirements, while conditions during placement and curing can change the result.

The study, published in the International Journal of Geosynthetics and Ground Engineering, tested a way to bring some construction conditions into an early assessment. Its four small test beds, called mesocosms, were intended to resemble aspects of field placement. They were not full-scale works, and the researchers have not shown which testing approach best predicts a completed project.

How the Rutgers sediment test was set up

The team used fine-grained sediment from a New Jersey Department of Transportation maintenance dredging program in Wildwood. It mixed the sediment with Type 1L Portland cement equal to 4% of the sediment's wet mass. Researchers constructed four outdoor mesocosms in two pairs at different times of year, placing treated material in two lifts about 24 hours apart to mimic a process extending over more than one day.

They took samples from the test beds with Shelby tubes and measured unconfined compressive strength: how much pressure a specimen withstands before failing without surrounding support. The comparison specimens were made in small molds using conventional preparation methods. Crucially, those specimens also cured outdoors alongside the mesocosms under similar ambient conditions. The reported comparison therefore concerns preparation and placement as well as the conditions experienced within the larger test beds; it is not a comparison with specimens kept in an indoor laboratory throughout curing.

Why strength varied between the test beds

The larger test beds also produced less consistent strength results. Across curing durations, the average standard deviation was 32 kilopascals for mesocosm specimens and 11 kilopascals for molded specimens. Four of the 24 strength tests from the second mesocosm set contained soft spots where cement had not been fully incorporated. Removing those results reduced the apparent variation, although the remaining mesocosm strengths were generally below the molded comparison.

The first pair was built in late September 2024 and the second in mid-November 2024. Average air temperatures during their first four curing weeks were 16°C and 6°C respectively, and the colder pair generally gained less strength. The authors say temperature could explain some of the difference between the two sets, but further modeling is needed to test that explanation. Differences in mixing and other environmental conditions also complicate a simple temperature comparison.

The team additionally used a dynamic cone penetrometer, which drives a cone into the material to assess resistance within the test beds. Its readings rose as the sediment cured and revealed differences among beds. The relationship between those readings and measured compressive strength had an R-squared value of 0.67. The scatter means the readings cannot yet be treated as a precise substitute for strength tests without further calibration.

What existing deep-mixing guidance says

The Federal Highway Administration's deep-mixing design manual identifies soil characteristics, mixing efficiency, curing time and temperature among the factors that affect treated-ground strength. It also warns that laboratory-mixed material can be stronger than field-mixed material, while other site conditions, including confinement or warmer curing, can shift the comparison in the opposite direction. That makes the relationship specific to the material and the work, rather than a fixed discount to apply to a laboratory result.

The federal guidance separates bench-scale tests from field validation. It describes field trials as a way to estimate how laboratory and field strengths relate for a particular design. Its discussion of earlier deep-mixing work reports bench-scale strengths ranging from one to five times field strengths for the same mix design. That historical range is context for why validation matters; it was not measured in the Rutgers experiment and does not establish how the new test beds would perform at a construction site.

What needs testing before project use

The Rutgers authors propose mesocosms and penetrometer readings as possible aids to design and quality control. Their study covers one dredged sediment and one cement dosage, and it has not compared mesocosm outcomes with full-scale construction. Trials using other materials, binder amounts and placement conditions, followed by comparisons with actual field results, would be needed to establish when the method improves a project's strength estimate. For now, the finding is a measured difference between prepared specimens and these test beds, with an open question about which better represents finished ground.