Testing a new aircraft: how do they do it?

[ivanotter]

It is really coming out of the discussion on the F-35.

Is testing more intense on the newer aircraft because we can test more? If we go back in time, how much testing was done and to what extent did it uncover things?

I am starting to be convinced that testing years back was more a matter of structural integrity, flying ability, workings of the instruments and some other pieces.

Insofar as any modern jet is managed by computers of many descriptions, the testing ought to be mor complex but also open up for more instances which can be tested.

Am I right in this?

Has the test plans evolved dramatically in the last few years?

The next logical question is: are we over-doing it? Not skipping things which can get people killed, that is not what I talk about.

Any comments?

Ivan

 

[michaelmaltby]

".... are we over-doing it?"

No ... you can't "overdo" for civillian AC .... technology is changing so fast ....787 Dreamliner woes of lithium ion batteries and their charging systems the most recent example. Very costly "glitch".

As for military AC - since they come from the same companies and suppliers as civillian AC - I don't see how you can have two standards for testing and/or safety.

Not to bad mouth anyone, but, look at the accident/failure record of ex-Soviet built civillian airliners ... I know, I know, lots of extenuating circumstances like slack maintenance, bad fuel, pilots letting their sons take the controls ... etc. etc. - but - I do not believe there is the legacy of public safety in these suppliers that we know is part of Boeing, Bombardier, Airbus Industries and associated companies.

The F-35 is the Perfect Storm - too many players, too many customers, tough schedule to maintain, cheeky (uninformed FTMP) media and A GLOBAL RECESSION with pressure on governments of all stripes to cut defense spending.

It's starting to look like 1930 ......

I wish the F-35 was twin engined like our RCAF F-18's and before that Voodoos and CF-100's. BUT THEY AREN"T - and modern jet engines like those in the Boeing 777 are marvels of thrift and reliability - and if they can take passengers safely across the Atlantic, I can accept the odds of bringing our pilots back from over the North Pole, or wherever

Don't switch horses mid-stream - in haste.

MM

 

[Matt308]

Development assurance is a structural methodology wherein the end game is to ensure that the methodical application of engineering sciences trends (DOES NOT GUARANTEE) that early requirement errors are discovered prior to operational fielding.

Development assurance looks like a Poisson curve wherein the discovery of development errors occur early in a temporal timeline and tail off exponentially, indicative of small residual errors encountered in fielded operations.

Any temporal change to the curve resulting in a shift to the right (time) is an epic fail.

This methodology is applied to both military and civilian projects. And for both safety and programmatic risk reasons. A good example is software development assurance (e.g., Mil Std 2167, RTCA DO-178B, ED-12C, etc)