Substantiation by similarity 2

[irq]

Let's assume the part, which has been substantiated by a static test. How to show properly the similarity (the components are not identical) of the new part for FAA approval? Which points should be adressed in the similarity report? I would really appreciate, if you could give here some links to adivsory circular, guidelines, or similar sources.

 

[MintJulep]

Fully describe the qualified part and the qualification process.

Fully describe the "similar" part.

Fully describe each aspect of how the "similar" part is different from the qualified part.

Identify and describe how the qualification process affects, stresses or otherwise "qualified" the original part.

Describe, with technical substantiation, how each different aspect is either not affected by the qualification process, or provide analysis to show how the different aspect would perform if subjected to the qualification process.

Argue with the FAA for several months.

Perform the qualification process on the similar part.

 

[Sparweb]

It can depend on the degree of similarity being discussed. Fasteners are a common item for "substitution", and it is often informally done in the field. For instance, an approved configuration using MS27039 screws on a panel can have a "minor" modification justified through similarity, by replacing them with a couple of handy AN3 bolts of the same diameter and length and so on. Mechanics let this through all the time, and typically it doesn't raise the flag, though sometimes it should. Minor modifications are defined by the FAR's and it's worth knowing about it, if you need to do things like drawing change notices on a design that may or may not impact the structural substantiation.

A change can easily be classed as "major" because the threshold is pretty low and there's a lot of A** covering going on. An example of a major repair is a puncture to a pressurized fuselage (which was substantiated by test) with an analysis of the patch and doubler you put on. Analysis works here because the skin structure is simple and repetitive and your repair exists within a structural shape that can be analyzed reliably. Usually your goal is to show that you have not reduced the margin of safety, either. An "equivalent strength" analysis uses the base strength of the material as "MS=0" and then proves that the changes leave you with "MS>0".

Changing your parts may lead to a slightly reduced margin of safety somewhere in the structure. Attempting a substantiation by analysis will begin with "MS=0" and there is no room to go down. Even when your original tests showed that there is a positive margin of safety, you probably don't have enough information to know how much the positive margin would be reduced by a change.

When a part or a structure was substantiated by test, changes to it cannot accurately be judged in terms of their effect on the structure, or its margin of safety, or its stiffness, unless a detailed structural analysis is done. Well, the level of structural analysis required to demonstrate such a thing can be as elaborate as a complete substantiation by analysis. Before starting down that road, you have to recognize the reasons that the thing was load-tested in the first place. The same logic or regulations will probably rule out the analysis you want to do for the change, too.




Steven Fahey, CET

 

[thgirw]

I can give some clue for engines.
When the engine is certified and you want to change something from the certification baseline, you have to prove that this change will not effect any of the certification items.
Step.1: You should prepare a report explaining why you are changing a part, what is the new part and how it will affect the system. This report should include a check list where all certification items are shown, you should "mark" all the items that are effected by this change. If it is a static part, let's say all items including strength, fatigue, etc.. should be checked. You should make sure that safety issues are not effected, so it is always good to give brief information about why/how safety is not effected. This report should also include how you are going to substantiate the part, in your case "similarity".
If above step is agreed by the authority then step.2 -
Step.2: Again you should make sure that all the "marked" items you identify in first step are covered in this report. You should clearly define how the component is similar for all those items.

I would recommend not to go and say this 2 parts are similar, this usually does not work. Proving something is "similar" means you designed , let's say compressor for 10 different engines and they all satisfied core blade off loads.

Hope this helps...

And please have a look to MIL-HDBK-1793A, not for civil applications, but the information within may help.

Regards