Structural Concept Development

Before CAD, Before FEA

The most important structural design decisions are made before any geometry is drawn in CAD and before any finite element model is built. They are made with sketches, free-body diagrams, and engineering reasoning. The concept stage is where the structural architecture is chosen: the arrangement of members, the type of structure (beam, truss, shell, sandwich), the location of joints, and the path that loads take through the structure. These decisions constrain everything that follows. A concept that is poorly chosen cannot be rescued by optimisation, detailed analysis or manufacturing refinement. A concept that is well chosen makes the subsequent work straightforward.

Choose the structural architecture before optimising the last few percent of its mass. The architecture determines whether the structure is good; the optimisation determines whether it is a few percent lighter.

The Six Questions

Every structural concept begins with six questions. They are deceptively simple, and the temptation is to answer them too quickly. The engineer who answers them carefully — who sketches the load entry and exit, who identifies what must remain rigid and what may flex — will produce a better concept than one who jumps straight to geometry.

1. Where does load enter? — What are the interfaces, attachments and applied forces?
2. Where must it leave? — What are the reactions, supports and downstream interfaces?
3. What spans the distance? — What structural member or form connects entry to exit?
4. What must remain stiff? — What deflections, rotations or alignments are constrained by function?
5. What can deform? — Where is flexibility acceptable, or even desirable?
6. Which failure modes matter? — Strength, stiffness, stability, fatigue, fracture, thermal — which govern?

If you cannot answer these six questions with a sketch, you are not ready to design the structure. If your concept does not address each one, it is not a complete concept.

Generating Multiple Concepts

The first structural concept that comes to mind is rarely the best. It is usually the most familiar — the one that worked on a previous project, or the one that the CAD tool makes easy to draw. A disciplined concept development process generates several plausible concepts for the same requirement and compares them before committing. The comparison is qualitative at this stage: mass, stiffness, joint complexity, manufacture, buckling sensitivity and inspection. No concept is universally superior; each has strengths and weaknesses that depend on the specific requirement.

• Generate at least three plausible concepts — not to find a universal winner, but to understand the trade space
• Compare qualitatively: mass, stiffness, joint complexity, manufacture, buckling sensitivity, inspection
• Do not declare a winner until the comparison has been written down — the act of writing exposes assumptions
• The concept that seems obvious may be the right one — but only after the alternatives have been genuinely considered

concept-comparison

STRUCTURAL CONCEPT COMPARISON

Consider a representative engineering problem: a horizontal member must carry a distributed load across a 1.2 m span, supported at both ends, with a concentrated downward load at mid-span. The mass target is aggressive; the manufacturing budget is moderate. Five plausible concepts are compared below. No concept is declared a universal winner — the comparison itself is the engineering value.

No concept is universally superior. The deep machined beam is simplest but wastes material. The truss is lightest but joint-heavy. The sandwich panel is stiffest but inspection-limited. The right choice depends on the specific requirement — mass, stiffness, manufacture, inspection and risk tolerance.

What Drives the Choice

The concept comparison does not produce a single answer. It produces a set of trade-offs that must be weighed against the specific requirements. The governing factors are listed below. Their relative importance changes from project to project, and the engineer must rank them consciously rather than defaulting to the most familiar approach.

• Mass target — how aggressive? If every gram matters, truss or sandwich; if moderate, beam or box section may suffice
• Stiffness requirement — deflection-limited? Sandwich and deep beam excel; truss may be too flexible
• Joint count and complexity — fatigue-critical? Fewer joints favour integral designs; many joints favour trusses but add fatigue sites
• Manufacturing capability — what can the factory actually build? A concept that requires a process the factory does not have is not a viable concept
• Buckling sensitivity — thin and light? Sandwich and stiffened shells are buckling-sensitive; solid sections are not
• Inspection access — can critical regions be examined? Closed sections and sandwich panels hide internal surfaces
• Cost and schedule — a lighter concept that takes twice as long to build may be the wrong choice

Avoiding Premature Detail

The most common concept-stage mistake is to add detail before the architecture is settled. Fillet radii, fastener diameters, ply thicknesses and mesh density are all detail decisions that should be made after the structural form is chosen. Adding detail early locks in decisions that should remain open, and makes it harder to compare concepts fairly. A concept sketch should show the structural form, the load path and the interfaces — nothing more. The detail comes later, once the concept has been selected and the design begins to mature.

Premature detail is the enemy of good concept development. If you are sizing fillets or selecting fastener diameters before the structural form is chosen, you are working at the wrong level. The concept sketch should show form, load path and interfaces — nothing more.

The Role of Experience

Experienced engineers generate good concepts quickly because they have a mental library of structural forms and their trade-offs. They recognise patterns — "this is a bending problem, so depth matters"; "this is a torsion problem, so a closed section is needed"; "this load path has an eccentricity, so there will be bending". Less experienced engineers can build this library by deliberately generating and comparing concepts, even when the answer seems obvious. The discipline of considering alternatives is what builds the judgement that makes future concept development faster and better.

• Experience is a library of patterns — bending needs depth, torsion needs closure, eccentricity creates moment
• The library is built by comparing concepts, not by being told the answer
• Even when the answer is obvious, sketching the alternatives confirms that it is obvious rather than merely familiar
• The engineer who always picks the first concept is not faster — they are just more likely to be wrong

From Concept to Design

Once a concept has been selected, the design process begins. The concept sketch is converted to preliminary geometry. Simple hand calculations — bending stress, shear stress, buckling load, deflection — establish whether the concept is viable at the basic level. A preliminary finite element model confirms the load path and identifies regions that need refinement. The concept is then iterated: geometry adjusted, sections sized, joints detailed, failure modes checked. This iteration is the subject of later articles in this category. The point of this article is that the concept — the architecture — must be settled before this iteration begins. Iteration refines a concept; it does not create one.

A good concept makes the design process straightforward. A poor concept makes it painful. The time spent generating and comparing concepts is the highest-leverage time in the entire design process.

Key takeaways

• The structural concept — the arrangement of material that carries the load — is the most important decision in the design process. Everything that follows is refinement.
• Choose the structural architecture before optimising the last few percent of its mass. A well-optimised poor concept is still a poor concept.
• Start before CAD detail: identify where load enters, where it leaves, what spans the distance, what must be stiff, what can deform, and which failure modes matter.
• Generate multiple concepts for the same requirement and compare them qualitatively before committing. The first concept that comes to mind is rarely the best.
• Avoid premature detail. A concept sketch with a clear load path is more valuable than a fully modelled design with a tortuous one.