
ID	0071
Class	DOCUMENTATION
Severity	1
Status	ready
Models	Trawl/TrawlDoor
Found	2026-09-26, resolving FISH-0069
Decision needed

Evidence

examples/input/6DOFTrawlDoor.xml:29-30 (the @include example of ToTrawlDoor.h:105) draws trawl_doors/traaldoer with VisualizationScale = 0.01; examples/input/Example.xml:104 draws the same mesh with 0.001.
TrawlDoorBase.cpp:106-107,742: the node scale is (Length * LengthScale, Curvature * CurvatureScale, Height * BreadthScale), with Length = sqrt(Area / AR) and Height = AR * Length, all scales defaulting to VisualizationScale. For Area = 8, AR = 3: Length 1.63 m, Height 4.90 m.
The M_MESH_BOUNDS of traaldoer.mesh (in fhsim_fishery.zip) are x -922.8 .. 615.3, y -87.3 .. 111.4, z -423.6 .. 605.7: 1538 x 199 x 1029 units. At 0.01 the drawn door is about 25 m x 6 m x 50 m, against a 1.6 m x 4.9 m door; at 0.001 it is about 2.5 m x 5 m in length and height.
Screenshot, FhVis on WSLg, camera 80 m from Door.Pos1: the red door fills most of the window (session scratchpad f68/door_traaldoer_far_after.png).
The ToTrawlDoor.h and TrawlDoorBase.h configuration excerpts copy the 0.01 (FISH-0069).

Effect

Visual only: the shipped single-door example, and anyone copying it, draws the door about ten times too large. No result is affected.

Possible fix

Set VisualizationScale = "0.001" in examples/input/6DOFTrawlDoor.xml and in the two header excerpts, as Example.xml does. The tests/in/* inputs that set the mesh run headless, so their scale is not drawn; they could follow for consistency.

Test that would prove it

A screenshot of 6DOFTrawlDoor.xml with a Visualization observer and a camera on the door, showing a door of about 2 m x 5 m.

Risk

Low: visualisation only.