Concept design is the stage between a specification and a detailed 3D model. Its job is to find the different ways a product could work, and to choose one of them, on evidence, to take into detail design. Feasibility is the companion question: whether a concept can meet the specification at all within its cost, size, timing and manufacturing limits, and what would have to be true for it to do so.
Time spent here is cheap compared with what follows. A concept can be dropped for the cost of some sketches and a page of calculations. The same change after tooling has been ordered costs the tooling, and often the launch date. The aim is to reach detail design with a concept whose main risks are known and judged acceptable, along with a record of why the alternatives were set aside.
Generating concepts
Start from the functions in the specification; our guide to writing a product specification covers how to state them. Breaking the product into what it has to do, such as hold, move, seal, sense or power, lets you look for several ways of achieving each function before combining them. A morphological chart is the usual tool, with functions down the side and possible solutions across, so that each concept is one path through the grid. For a powered sliding door, the rows might be the power source, the drive element, the guidance and end-of-travel detection, each with four or five options.
Other sources repay a structured look: competing and adjacent products, patent literature (for ideas, and for what you cannot use), bought-in components that already solve part of the problem, and the processes your intended manufacturer is good at. Sketches at this point should be quick and plentiful. Nobody should be attached to them yet, and an unworkable idea pinned to the wall often prompts a workable one.
Early calculations that rule concepts in or out
Each concept carries a few physical questions that first-order calculations can settle: whether the motor is big enough, whether a beam deflects too far, whether the heat can get out, whether the parts fit the envelope. Precision is unimportant at this point. An answer within a few tens of per cent separates a comfortable concept from an impossible one, so the effort belongs in getting the inputs roughly right. A handful of relations covers most of it:
Cantilever tip deflection: δ = F L³ / (3 E I)
Bending stress: σ = M y / I
Power: P = F v = T ω
Convective heat loss: Q = h A ΔT
As an example, take a sliding guard door on an industrial machine. The door weighs 60 kg and must open 1.8 m in about 6 seconds. Assume a peak speed of 0.4 m/s, an acceleration of 0.5 m/s² and a running resistance from rollers and brush seals of 2% of the weight.
- Force to accelerate the door: 60 × 0.5 = 30 N
- Running resistance: 0.02 × 60 × 9.81 ≈ 12 N
- Peak force about 42 N, so peak power is about 42 × 0.4 ≈ 17 W
- On a 40 mm diameter pulley: torque about 42 × 0.02 = 0.84 N m, at 0.4 / 0.02 = 20 rad/s, or roughly 190 rpm
With 0.5 m/s² ramps at each end, the door reaches full speed in 0.8 s and completes its travel in about 5.3 s, inside the requirement. Doubling the resistance to allow for a dirty track still leaves a load that a small geared motor handles easily. Motor size is therefore not a feasibility risk for any of the electric concepts, which moves attention to the things that are: grit in the drive, and the cost of hardware with a long stroke.
Feasibility questions to answer before detail design
A concept is ready to take forward when the answers to these are known well enough that none of them is likely to stop the project later.
- Physics. The concept can deliver the forces, speeds, stiffness, temperatures and life in the specification, with margin to spare.
- Package. It fits the envelope and the interfaces, with room for fixings and cables and access for assembly and service.
- Manufacture. The parts can be made by processes that suit the volume, and the expected unit cost sits near the target.
- Bought-in parts. Motors, bearings, seals, sensors and electronics exist in the sizes needed and can be bought at the planned volumes.
- Freedom to operate. No existing patent blocks the working principle, or there is a known way around it.
- Regulation. Nothing in the concept makes the applicable product rules much harder to meet, such as an exposed moving part that will need guarding.
Some questions cannot be answered with a calculator. The feel of a latch in the hand, or the way a brush seal behaves in metal swarf, usually needs a quick physical test, and the time to find out is now, with a rough rig. Questions about stiffness, stress, flow or heat in a defined shape are often answered faster by simulation on a simple model. Our introduction to what FEA is explains where that works well.
How many concepts, and when to stop
There is no correct number. The aim is to cover each distinct working principle at least once. In practice a wide first pass of sketches narrows to a handful worth developing with calculations and rough layouts. Developing more than five or six to that level rarely pays. Taking only one that far means nothing has been compared, and the first idea is seldom the best.
Stop generating when new concepts turn out to be recombinations of existing ones. Comparison can end when more work would not change the ranking. If the top two remain close and the deciding question is one only a test or a more detailed analysis can answer, carry both a little further and answer that question first. It is also reasonable to stop when the remaining uncertainty is the kind detail design resolves as a matter of course, such as a fixing, a seal groove or a wall thickness.
A weighted comparison worked through
The guard door drive is for a machine sold at around 300 a year into dusty workshops. A pneumatic cylinder was ruled out before any scoring, because many customer sites have no compressed air supply. Mandatory requirements work best as a pass-or-fail screen like this, with only the concepts that pass going into the weighted table. Four electric concepts remained. Each was scored from 1 (poor) to 5 (excellent) against six criteria, with the weights set before any scores were written in.
| Criterion | Weight | Rack and pinion | Toothed belt | Lead screw actuator | Roller chain |
|---|---|---|---|---|---|
| Tolerance of dust and grit | 25 | 3 | 3 | 4 | 2 |
| Unit cost at 300 a year | 20 | 3 | 4 | 2 | 4 |
| Scales to other door widths | 20 | 5 | 4 | 2 | 4 |
| Ease of fitting and adjustment | 15 | 3 | 4 | 5 | 3 |
| Development risk | 10 | 4 | 4 | 5 | 4 |
| Noise | 10 | 2 | 5 | 4 | 2 |
| Weighted score out of 5 | 100 | 3.40 | 3.85 | 3.45 | 3.15 |
Each weighted score is the sum of weight × score, divided by 100. The belt drive leads. Because the weights are judgements, test how sensitive the result is to them before accepting it. Raising dust tolerance to 35 and cutting unit cost to 10 narrows the gap, with the belt on 3.75 and the actuator on 3.65, but the order at the top holds. Had a plausible change of weights reversed it, the right conclusion would be that the two were tied and the decision rested on whichever criterion separated them.
Look at where the winner scores badly as well as at its total. The belt’s weakness is grit, so the concept taken forward is a belt drive with a full-length cover and a sealed idler, borrowing the actuator’s best feature. Folding the strengths of the runners-up into the leading concept is normal, and often the most useful result of the exercise. Keep the table and the reasons behind each score. If a customer later asks for a 3 m door, the table shows quickly whether the choice still stands, and here the belt’s good score for scaling suggests it would.
What a chosen concept should carry into detail design
A concept is ready to develop when it has a layout of the main parts with rough sizes and the interfaces between them, calculations for the main loads, speeds, power and heat, a short list of open risks, and a decision on the main manufacturing processes. The last item is easy to skip. A part drawn as a casting and later switched to a fabrication usually has to be redesigned, so the process choice belongs here, together with the design for manufacture rules that come with it.
How SixGate™ can help
At the Concept stage, SixGate™ develops several design concepts, sketched and backed by calculations, and compares them against your specification. You see the options side by side with the numbers behind each, and the reasons one is recommended, before money goes into detail design or tooling. On the offshore wellhead decommissioning project, three removal-tooling concepts were compared in the 3D model before anything was manufactured. If you already have a concept and want an independent view of whether it will work, that can be the starting point instead.