Reader problem: There is a small opening where two pieces of weatherstrip meet. Does it show original construction, an incorrectly cut replacement, or a strip that has pulled back over time? The useful first step is to document the joint's location and evidence of movement, not immediately glue, stretch or cut the strip.
Four observations that change the answer
- Joint position: Mark whether it is at a corner, the top/head, a straight run or a factory-shaped transition. Photograph the whole perimeter and a close-up. A gap at a designed meeting point is different from a random break midspan.
- End appearance: Record clean square-cut ends, factory-molded shapes, uneven torn material, or old adhesive witness marks. Wear is a clue, not a proven manufacturing history.
- Retention: Can you see an already exposed kerf foot, T-slot back, adhesive edge or mechanical carrier? Keep receiver type separate from the bulb face. Do not pull intact material out to expose it.
- Context: Note whether this is a newly fitted strip, a long-installed seal or a part observed only after frame work. Record photographs and date rather than assigning a cause from memory.
| Evidence in the photo | One plausible explanation | What is still required |
|---|---|---|
| Two neat ends at a repeatable designed corner | A joint may have been intended | Assembly-specific drawing showing corner treatment |
| Short strip with offset adhesive/witness line | Possible movement or misposition | Original installation record and material/retainer identity |
| Frayed, cracked or permanently flattened edge | Aging or mechanical damage may be involved | Whole-frame contact assessment; old profile identity |
| Open corner after a recent replacement | Wrong cut, placement or part is possible | Confirmed installation method for exact system |
Measure a gap without changing it
Put a ruler next to, not inside, the opening and photograph the distance between the visible ends. Record the linear position of the joint on a simple perimeter sketch (top-left, bottom-right, etc.). Note whether the ends can be observed without moving the sash; never place fingers in a pinch point. Where the seal is compressed, do not treat a photographed contact width as its free factory width. If there is an already-detached sample, document its relaxed cross-section separately. Avoid “repair by stretching” because a strip can appear to close a gap temporarily and later retract; no treatment recommendation follows from this observation.
A comparison across two visits
For a historic problem, compare same-angle photos on two dates if they already exist. Mark the joint length, frame temperature/moisture context if known, and whether door/window operation changed. This cannot establish a material shrinkage coefficient, but it distinguishes a stable visible joint from an evolving symptom worth escalating. Never measure by heating, chilling or force-stretching a strip to prove a cause.
Distinguish the failure from a draft elsewhere
Air perceived near a corner could also arrive at a sash/frame interface or a different nearby opening. Document a visible discontinuity and exact contact surface. Do not assume a seal end gap is the only source of air, water or noise. If the perimeter is part of a fire/smoke, egress or commercial assembly, preserve its original state for qualified personnel; no trimming or sealant patch is approved here.
Bring these answers to service
Provide a whole-assembly photograph and maker/series information, the joint location, end-detail close-ups with scale, known installation date, any original document/part label, and a note for each unknown. Ask the issuer or installer for the approved joint/corner method for this model and production generation and whether any service kit is preformed, factory joined or lineal stock. Do not substitute a generic strip solely because its bulb shape is similar.
Claim boundary: Different possible causes are offered as a troubleshooting classification, not a diagnosis. No verified product-specific joinery, replacement part, expansion/shrinkage measurement or manufacturer edition underwrites a universal repair step. This classification does not prove a particular cause or identify a matching replacement.