Sample and Specimen Model

GeoDin's data model for samples, specimens, and the parent-child relationships between them

The Sample Data section holds every sample associated with a location. Each sample carries a reference, a depth interval, and the sampling metadata that links it back to the field operation; specimens (subsamples) reference their parent sample for hierarchical tracking.

For the related concepts, see also:

• Borehole and Location Model - the location these samples belong to.
• Layer and Stratigraphy - ground descriptions at the depths these samples cover.

Sample fields

Each sample has:

• Sample reference - sample name, often just a number; user-defined and advised to be unique within a location.
• Recovery from / to depth and penetration from / to depth.
• Sample condition - disturbed / undisturbed / etc., from a dictionary.
• Sampling method - grab sample, auger, etc., from a dictionary.
• Sample type - bag, etc., from a dictionary.
• Date, comments, blow count.

Sample condition for specific tests (e.g. undisturbed / remoulded on a UU triaxial test) is stored inside the test parameters, not directly on the sample record. The sample-record condition refers to the sample as collected; per-test conditions vary across the laboratory program.

Parent samples and specimens

Subsamples and specimens reference their parent sample via the Parent Sample Reference column. This produces a clean parent-child structure in the sample table: a single parent (e.g. a 4-inch liner tube) with multiple specimens taken from it for individual lab tests.

Onsite sample workflow

In GeoDin Onsite, the sample workflow on a drilling form is:

1. Select sampling method (e.g. liner tube, 4-inch).
2. Select condition (e.g. undisturbed).
3. Enter recovery percentage (e.g. 95 %+).
4. Select sample type.
5. Set depth reference.
6. Print a QR-coded label that can be stuck on the physical sample.

Onsite sample IDs

Onsite assigns each soil sample a unique ID with a fixed structure:

Component	Length	Notes
Prefix letter(s)	1-3 chars	Configurable
Timestamp	6 chars	Encodes the moment of sampling
Random	4 chars (default)	Shrinks if prefix grows so that prefix + random always totals 5 chars

The timestamp portion changes every second, so collisions only happen if more than ~20-100 samples are generated in the exact same second.