There is a moment on many projects where a beautifully finished 3D model lands in the inbox. Every fillet is in place, the assembly rotates cleanly, the drawings are dimensioned. It looks like the answer. And sometimes it is. But just as often, it is the question — drawn neatly, and not yet answered.

That gap, between a model that looks right and a design that has been shown to be right, is the difference between CAD delivery and engineering substantiation. For anyone buying engineering, it is worth understanding, because the two are often priced as if they were the same thing.

What CAD delivery gives you

CAD delivery produces geometry: a 3D model, an assembly, a set of drawings. Done well, it is genuinely skilled work — clean modelling, sensible tolerancing, manufacturable detail, correct GD&T. You can send the drawings to a workshop and get parts back.

What CAD delivery does not answer, on its own, is whether the thing you have drawn will actually do its job safely. Will the structure carry its load with margin? Have the right load cases even been identified? What happens under the worst realistic combination of forces, not just the nominal one? Is the bolting adequate; will the mechanism survive its service life; does the design meet the standards that apply to it?

Geometry can be perfect and the engineering still wrong.

What substantiation adds

Substantiation is the evidence that the design works — and, crucially, the reasoning that lets someone else agree. It typically means:

  • A defined design basis — the loads, duty, interfaces, environment and acceptance criteria the design must satisfy, written down before conclusions are drawn.
  • Calculations and analysis — first-principles structural calculations, and finite element analysis where it is warranted, against those defined load cases.
  • Standards assessment — showing the design meets the codes and standards that apply to it.
  • A clear report — assumptions, methods, results and conclusions, in a form a customer, a checker or a regulator can follow and challenge.

The test of substantiation is simple: could a competent engineer who was not involved pick up the evidence, follow the reasoning, and reach the same conclusion? If yes, you have something you can defend. If the only justification is “it looked about right in CAD”, you do not.

Why the distinction matters commercially

Two quotes for “designing a lifting frame” can look similar and mean completely different things. One is for the geometry. The other is for geometry plus the evidence that the frame is safe to lift the load. On a high-integrity item, the second is the whole point — and discovering the difference after the parts are made is the expensive way to learn it.

This is not an argument that every bracket needs a finite element study. Much of the skill is judging how much substantiation a given item actually warrants: proportionate to the duty and the risk, no more and no less. A guard bracket and a critical lifting beam do not need the same rigour. But deciding that, on purpose, is itself an engineering act — not something that falls out of a CAD package by default.

CAD tells you what a design looks like. Substantiation tells you whether you can rely on it.

The short version

If you are commissioning mechanical design, it is worth asking one question early: am I buying geometry, or am I buying geometry with the evidence behind it? Both are legitimate — sometimes a model is exactly what you need. But on anything where failure carries a real consequence, the model is the start of the job, not the end of it. The value is in the engineering that turns a shape into something you can put your name to.