STLPad

STL MESH GUIDES

How to Repair Holes in an STL File

A hole in an STL usually appears as an open boundary where faces are missing. To repair it safely, first confirm that the boundary describes a real, fillable opening. Then compare the repaired mesh's diagnostics with the original. Not every open edge can—or should—be capped automatically.

STLPad fills supported planar boundary loops and reports the boundaries it cannot safely fill. The workflow stays local to your browser, and repair produces a separate file rather than overwriting the original.

How holes appear in STL files

An incomplete export or a deleted group of faces can leave a rim in a previously closed surface. Boolean operations, damaged mesh edits, and format conversions can also leave gaps where surfaces were meant to meet.

Those causes can look similar in a preview but produce different boundaries. One missing triangle has a simple rim. A broken junction between several surfaces may branch, double back, or collapse into a line. The latter is not just a larger version of the same hole.

Also distinguish missing surface from an intended opening in the object. A cup with wall thickness can have an open top in everyday language while its mesh surface remains closed around the inner wall, rim, and outside. Capping its mouth would change the design, not fix it.

How to tell whether your STL has a hole

Boundary edges are STLPad's main signal for an open mesh. A boundary edge is used by one non-degenerate triangle; an ordinary closed surface edge is used by two. A nonzero boundary count tells you the surface has exposed edges, not how many holes need filling.

Several boundary edges may form one loop. Others may form open chains or ambiguous structures. A model with many exposed edges can therefore have fewer fillable holes—or none. Check Non-manifold edges and Degenerate triangles too, rather than interpreting every diagnostic as a missing patch.

Check your STL for open edges

The watertight STL guide explains what the closure result does and does not establish. Use a visual inspection alongside the numbers to decide whether the model needs surface repair.

How to repair holes in an STL with STLPad

  1. Open STL Repair. If the file is already in Checker or Viewer and within the supported limit, Open in Repair transfers it locally.
  2. Otherwise, use Choose STL file or drop the file into Repair.
  3. Review the original diagnostics before repair, especially boundary edges, non-manifold edges, and degenerate triangles.
  4. Run the default safe repair using Repair STL. In Advanced Options, Fill safe planar holes is enabled by default.
  5. Review the Before → After diagnostics and any skipped-boundary explanations.
  6. Confirm whether boundary edges were reduced. Do not interpret completion alone as successful closure.
  7. Use Download repaired STL to save the separate result, then inspect it in your slicer.

Repair STL

Repair supports up to 600,000 triangles. Larger models can still be inspected, but cannot bypass the local Repair safety limit. The original remains available through Reset to original during the repair session.

Advanced Options include maximum hole size and planarity tolerance. Those settings describe what the algorithm may consider, not a promise that every candidate will pass. Start with the defaults; increasing a threshold does not supply missing information about the intended shape.

Why some STL holes cannot be filled automatically

A safe cap needs a usable boundary and valid triangles. A loop can be closed in its connectivity yet have no meaningful area. If its vertices lie on a line, filling it would create degenerate faces rather than a surface.

STLPad deliberately skips boundaries such as:

Non-manifold geometry can also make the intended surface ambiguous. Adding faces around an already invalid connection is not a general solution. Read the non-manifold STL guide if that diagnostic is present.

Safe repair is better than forcing a mesh closed incorrectly. A skipped boundary is useful information: it tells you where automatic cleanup stops being a reliable substitute for an editing decision. It is not a reason to keep increasing tolerances until a warning disappears.

What happens after a hole is filled?

STLPad analyzes the repaired result using the production Analyzer. Before → After rows show the resulting topology, rather than merely listing operations attempted by the repair routine.

Hole filling adds faces, while degenerate and duplicate-face cleanup can remove others. The total triangle count may therefore rise or fall. Use boundary and non-manifold counts, the remaining diagnostics, and the visible shape to judge the result.

If the result is not suitable, Reset to original restores the starting mesh and diagnostics. Keep that original when exporting, particularly if the part has small intentional openings or disconnected details you need to preserve.

Does filling a hole always make an STL watertight?

No. Other openings may remain, or the mesh may still have non-manifold edges. A repair can remove degenerate triangles successfully while leaving every unsafe boundary untouched. That is a valid partial result, not a guarantee of closure.

Even a Watertight: Yes result covers the checked topology, not every potential geometry or printing problem. STLPad does not test general self-intersections or wall thickness. Review the repaired file rather than assuming a clean count replaces all other checks.

Should I use slicer auto-repair instead?

Slicer repair can be convenient for straightforward problems. Checking the mesh first gives you more visibility into what is wrong and a clearer basis for comparing the result. Neither approach removes the need to review the sliced layers.

If a cap changes an important feature, return to the source design or a mesh editor. For supported holes, the practical sequence is simple: identify open edges, repair conservatively, compare diagnostics, and verify the exported shape before printing.