August 20, 2021, ©. Leeham News: Last week, we went through our second Design Refinement and how our Processes and Tools must be mature because the next step in the Post Launch phase is Detailed Design.
If the previous phases were about research and inspirational design work, where we measured, collated, and documented the overall design data for the aircraft, Detail Design is about a massive amount of work packages and execution.
We are now entering the most resource-intensive and organizationally demanding phase of our project.
We previously worked with overall sketches of surface shapes of the fuselage, wing, and empennage; we now design these skins to the exact shape required. But the skins are a minor job. We must then create every bit and piece that sits beneath the skin. That attaches it to the aircraft and supports it, so the stressed skin construction can take the loads we envisage for the plane.
We start with the overall shape of the parts below the skin. Figure 1 shows a design situation where the overall shape and placement of wing ribs, spars, and stringers are explored, using the most common design software in the industry, CATIA.
Once the overall shape of the key parts is done, we go into the detailed design of each component, Figure 2.
We need to do all the interfaces and how the stresses shall flow. We now pipe the structural 3D design into FEM (Finite Element Modeling) software, where the parts are analyzed for the stress situation given our load spectrum on the wing from our wind tunnel, and CFD runs.
It gives a first sizing of the parts due to stresses, and we can then continue to decide how we attached them to each other. For the parts, we need to determine what Alu alloys do we use? Strong or fatigue resistant (we have decided to use conventional Alu construction, mainly for cost reasons).
If we were to choose a novel material/alloy where we don’t have existing certified design values, we must have a plan to get them before the design is finalized or we can’t optimize the structure. This includes preparing for damage tolerance assessments to the primary structure. For complex materials like composites, this would include process variation assessments and management. It can take years to certify a new material, which would not fit our schedule or our budget.
For the connection of parts, do we use rivets, fasteners, or bolts? Given the stresses, what size do we use, and what’s their weight? Can we glue any parts together (it is beneficial from a weight and fatigue angle, but then these parts are hard to separate later, during maintenance)?
Our tools department might have ideas of better design packages, but we must go with what the industry uses because if we before could do with, say, 20 structural design engineers, we now need hundreds. We might let the supplier of, for instance, the flaps do the detailed design of these. But the overall shape and design requirements shall come from us, and it’s beneficial if we use the same tools, as the transfer of design data is then straightforward.
There are converters for 3D geometry in the industry like IGES and STEP, but these never transfer a design 100% complete and fault-free between tools. The decision to allow the project to use different design packages (different versions of CATIA and some suppliers having an older Computervision tool) doomed the Airbus A380 to a two-year delay when electrical wires didn’t meet at the connectors at Final Assembly (FAL).
So we need strong discipline with what data goes where, in what form and revision, how suppliers are allowed to use the data, and how we want information back. Our organization needs to manage everything to a detailed level.
We can’t give a supplier the conceptual design package and ask him to report back when he is X% through the detailed design of the parts.
It happened in the Boeing 787 project, where partners reported back, but the maturity of their designs states were not followed up. Too late, faults in the understanding of overall design data were discovered or the lack of coherent designing rules for new areas, like CFRP designs.
It put the 787 project in a deep crisis, and it was only the strength of the Boeing organization at the time that landed the project back on a structured path. But it took some four years to get back on track, with the enormous stress of those involved as a result.
Also, we must remember that any detailed design work the supplier does must still be available to the regulator for inspection — careful agreements are needed here around intellectual property, change management, and quality management systems. All of this drives expenses into the cost of supplied parts that can be easily overlooked.
When we start detailed design, our data is of the form; shapes, volumes, the partition of volumes, fit checks of system components, and their overall shape, weight, power consumption, heat emission, etc.
For example, our estimate of the Electrical Wiring Interconnect System, EWIS, is of the rough number of connections and wires and how many thousands of meters of cables we need (we need km of wires). With a value for kg per m and the number of connectors, we estimate the weight of the EWIS.
Now we need an EWIS design package to turn the information from the Electrical System, Avionics, Cabin, and all the Systems suppliers into an EWIS design that spells out what wire needs to go where and in what wire bundle shall this be routed.
As we design the EWIS, we need to route it in the aircraft and attach it with brackets so it stays intact during G loads and doesn’t move or wear. If a box that is connected to the EWIS changes its electrical requirements, the EWIS must change.
Once again this is made complex by the aircraft regulations, where often the most efficient routings for space and weight could bring wires too close together, violating rules around system separation designed to prevent single failures from impacting redundant systems.
Once the EWIS is frozen, any changes end up in time delays. The A380 got delayed two years when the EWIS needed redesign. You would think wires, hydraulic and fuel tubes, etc., are the minor parts in an aircraft design. It’s not, mismanage it and it delays your project years.
We continue our Detailed Designing discussion next week, where we also go into how we work with the Certification authority during this intensive period.