Separate size, position and orientation
Investigate three questions independently: how long is a known element, where is a known point, and which direction does the model face? A single visual adjustment can conceal several errors. Preserve the original model and document each transformation during diagnosis.
| Concept | What to record | How to verify |
|---|---|---|
| Units | The model's length unit and the receiver's import interpretation | Measure an agreed dimension. |
| Local origin | The reference used for model coordinates | Compare a known point in local coordinates. |
| Axis convention | Up direction and any import axis conversion | Inspect a vertical member and known horizontal direction. |
| Map reference | The approved coordinate reference system and transformation | Compare documented site control points. |
| Height reference | The agreed vertical datum or project level reference | Compare a known elevation. |
Understand what georeferencing adds
In IFC, a projected coordinate reference system describes the map context, while a map conversion can relate local engineering coordinates to that context using scale, rotation and translation. Merely creating the relevant entities does not establish correct coordinates; their parameters must be supplied and checked.
Obtain the reference system, units, height reference and control-point schedule from the project's coordinate owner. Do not infer them from a nearby street address or a plausible-looking map position. Confirm whether the receiving application reads the transformation or expects coordinates already transformed.
A simple scale check
Suppose a known test element is 5 metres long. If the receiver interprets its length as 5,000 metres, the measured ratio is 1,000. Investigate a millimetre/metre interpretation mismatch before changing the model. The ratio is a diagnostic clue, not proof of the cause.
Once the size is correct, compare at least two separated horizontal control points and an elevation reference. One point can expose a translation but does not establish the orientation or scale of the entire model. Choose additional checks appropriate to the project's extent and precision requirements.
- Record expected and observed coordinates in the same units and reference system.
- Have the project team set the permitted deviation; use the agreed project tolerance for all comparisons.
- Recheck every discipline in the federation after a correction.
- Keep the approved transformation with the export settings and delivery manifest.
Use the symptom to narrow the investigation
| Symptom | Investigation |
|---|---|
| Correct shape, wrong size | Declared units, import overrides and scale applied at more than one stage. |
| Correct size, constant displacement | Origin offsets, model placement and duplicated transformations. |
| One point aligns, distant points diverge | Rotation, scale or an inconsistent coordinate reference. |
| Correct plan position, wrong height | Vertical datum, local level reference and height offset. |
| Model appears on its side | Up-axis conversion and saved object transforms. |
| Display is unstable at large coordinates | Test an approved local-origin copy and retain its documented relationship to the project reference. |
For visualisation, a local-origin delivery may be appropriate when the recipient agrees. Label it clearly and preserve the relationship to the coordination model. Do not silently replace a required georeferenced delivery with a recentered copy.