[Richard C]

Originally Posted by crmalcolm

In other fields that I do have experience of, the build strong and work back method makes more sense.

The testing works as follows:

A model is built that is estimated to be 1.5x the required strength. A second is built that is estimated as being 1.1x the required strength.

Both are put through the test, and then (if they pass) they are pit through a NDT process (non destructive testing).
This will now give the team two things:

A - a good indication of how accurate their modelling was in terms of strength.
B - a homologated part (ideally 2) that they know they can race with if necessary and can use for other testing purposes.

Ideally, the 1.1x model passed and the team know that the final part will be very close to this specification.

Then, a further reduced part (or two) is prepared, at 0.99x and 1.01x the estimated strength. These are then put through the test and if either pass, the team has a lighter part homologated to race with. If they both fail, the team will have to use one of the previous models until they can get a lighter model homologated (if possible).

This fits the process explained by Cranfield, fits the timeline for preseason testing and homologation, and fits more traditional R&D methodology.

Makes sense.

Richard