Thanks Dale — that’s very helpful. Your point about keeping the loft sections simple and structurally the same, or editing one master section, is particularly relevant to what I’m trying to do.
I’ve developed the station approach a little further and would appreciate a sanity check before I go too far with it.
Fig. 1 — Station construction
Each section is made from four open NURBS curve segments, mirrored left/right. The Top and Bottom points are constrained by the upper/lower profile guides, and the maximum-width points are constrained laterally by the plan-view guide (see Fig. 3). The remaining CVs are there purely to shape the section.
Fig. 2 — Curvature graph
This is the same section with G2 at Top, maximum width and Bottom. I’m deliberately trying to keep the CV count low and the curvature distribution clean.
Fig. 3 — Stations and guide curves
The stations are copies of the same basic section, remapped to the local guide envelope as they move along X. The plan view tells me the width at each X, but not the height at which maximum width occurs, so I optionally use a separate side-view curve to define that height. Otherwise it defaults to halfway between the upper and lower profiles.
Fig. 4 — Introducing a shoulder
This is the part I’m particularly interested in your view on. If I need an extra control to introduce a gradual shoulder, I add the corresponding control to all sections so they remain compatible, but only move it on the stations where the shoulder is developing.
Fig. 5 — Zebra on the resulting loft
Although the section joins are G2, Zebra shows a subtle longitudinal feature along the Top, maximum-width and Bottom loci, particularly at maximum width.
A few questions:
-
Does the basic four-segment station construction look sensible for this kind of loft, or would you strongly prefer each complete section to be a single NURBS curve?
-
Is copying a section, remapping it to the new guide envelope, and then reshaping the copy essentially the workflow you meant by editing one master section?
-
For something like the shoulder in Fig. 4, is progressively moving corresponding controls through several stations a reasonable way to keep it within the main fuselage loft, or would you regard that sort of feature as a point where separate surfacing becomes preferable?
-
Does the separate longitudinal guide for the height of maximum width seem like a sensible way to define the sections from orthographic references, or would you normally establish that locus differently?
-
Finally, would the Zebra feature in Fig. 5 concern you even though the section joins are G2? In this construction the curvature falls to zero at the Top/Bottom/max-width joins (see Fig. 2). I’m wondering whether the continuity is technically fine but the curvature distribution itself is creating the visible line.
The end goal is a fair RC jet fuselage with simple, controllable NURBS geometry, rather than trying to force every aircraft feature into one loft.
Thanks again — any criticism of the approach would be appreciated.
