Compaction records are the earthwork documents most likely to be pulled years after a job closes. When a slab cracks, a trench settles, or a pavement section fails, the first question is always the same: show me the density tests for that area. This guide covers how compaction test results should be documented in the field so the record actually answers that question — what every test entry must contain, how test frequency is set, and how to handle failures and retests.
What a Compaction Test Actually Reports
Field compaction is judged against a laboratory reference: the Proctor test (ASTM D698 standard or ASTM D1557 modified) establishes the maximum dry density and optimum moisture content for each soil type on the project. A field density test measures the in-place dry density of a compacted lift and expresses it as a percentage of that Proctor maximum. A spec requirement of "95% compaction" means the field dry density must be at least 95% of the lab value for that specific material.
That means every field test result is meaningless without two references: which Proctor curve it was judged against, and where and at what lift the test was taken. Most documentation failures trace back to one of those two being missing.
Common Field Test Methods
Nuclear Density Gauge (ASTM D6938)
The nuclear gauge is the workhorse of field density testing. A certified technician seats the gauge on the lift, drives the source rod to test depth, and reads wet density, moisture content, and computed dry density in about a minute. Because it is fast and non-destructive, it is the standard method for lift-by-lift acceptance on structural fill, utility backfill, and pavement subgrade.
Sand Cone (ASTM D1556) and Drive Cylinder (ASTM D2937)
Slower, destructive methods used where a nuclear gauge is unavailable or as a referee test when nuclear results are disputed. The excavated-hole volume and recovered material weight yield in-place density directly.
Proof Rolling
Not a density test — a qualitative check where a loaded truck or roller crosses the subgrade while the engineer watches for rutting or pumping. Proof rolls identify soft areas that density testing missed, and the observation (pass, or limits of the soft area) should be documented like any other test event.
What Every Compaction Test Entry Must Contain
A defensible compaction log records, for every test:
- Test number — sequential, unique for the project
- Date and time
- Location — station and offset on linear work, grid or GPS coordinates on area work. "East side of building pad" does not survive a dispute; a coordinate does.
- Lift number and elevation or depth — which layer of fill was tested
- Material description and Proctor reference — which lab curve (maximum dry density and optimum moisture) applies
- Field wet density, moisture content, and dry density
- Percent compaction and the spec requirement — e.g., 96.2% vs. 95% required
- Pass/fail result
- Technician name and gauge ID (with current calibration)
- Retest linkage — if this test replaces a failed test, reference the failed test number
Test Frequency: How Many Tests Are Required
Frequency is set by the project specification, and it varies by work type. Typical patterns you will see:
- Building pads and structural fill — one test per lift per a set area (commonly one per 2,500–5,000 sq ft per lift)
- Utility trench backfill — one test per lift per a set length of trench (commonly every 100–200 linear feet, staggered lifts)
- Roadway embankment and subgrade — per station interval per lift, per the DOT spec section
Document the required frequency at the start of the job and track actual tests against it. An area with passing tests at half the required frequency is still a documentation failure at acceptance time.
Handling Failed Tests
A failed test is not a crisis — an undocumented failed test is. The defensible sequence is:
- Record the failure exactly like a passing test. Never discard failed results; a log with a 100% pass rate and a later settlement problem reads as a curated record.
- Identify the cause — usually moisture (too wet or too dry vs. optimum), insufficient passes, excessive lift thickness, or unsuitable material.
- Rework the lift — dry, wet, re-compact, or remove and replace as directed.
- Retest in the same location, and link the retest to the original failed test number.
Why Paper Compaction Logs Fail
Paper density sheets and end-of-week spreadsheet transcription lose exactly the details that make a record defensible: precise locations, lift linkage, and the failed-test-to-retest chain. Photos of the gauge reading live on someone's phone, disconnected from the test entry. When acceptance documentation is assembled months later, gaps get filled from memory — which is how records stop being evidence.
Gradelog's compaction log records each test with GPS location, lift, elevation, material, Proctor reference, moisture, density, percent compaction, and automatic pass/fail against the spec — with photos attached to the test record and PDF reports generated per area or date range for the engineer. Offline-first, so trench-bottom and remote-site tests log the same as everything else.
Field Checklist
- Confirm the Proctor curve for each material before testing starts — new borrow source means a new Proctor
- Record every test at the moment it is taken, including failures
- Use coordinates or station/offset for location, not landmarks
- Track lift numbers — a passing test on lift 4 says nothing about lift 3
- Link every retest to its failed test
- Verify gauge calibration is current and note the gauge ID
- Reconcile test count against required frequency weekly, not at closeout