AFGROW Tips 1

[Burner2k]

Hi folks,
I am not new to theoretical information pertaining to Fatigue & Damage Tolerance but kinda a newbie when it comes to practical application of it. Even my theoretical exposure is just above a year and thus I don't have the exposure of even a moderately experienced F&DT practicing engineer.

With my prior employer, I have done a couple of F&DT projects of typical cases i.e. stiffened panel of a fuselage and Wing bottom skins subjected to only one loading condition i.e. in case of fuselage skin, it was just internal pressurization & with Wing, it was as limit lift load condition.

In both cases, a crack was initiated on both sides of rivet/fastener hole of a center bay with the stiffener failed to demonstrate fracture arrest, perform a Fatigue crack growth & eventually establish inspection intervals.

I used FEM to calculate SIF (K) variation vs crack length and plot the residual strength diagram manually (Excel). We had a spectrum, in which the fluctuations were already cycle counted and broken down in to a series of constant amplitude cycles. So, I just used Excel (no retardation was accounted however Crack closure i.e. (Delta K)effective was used) to determine Fatigue Crack Growth rates.

Since I used FEM result i.e. max tensile stress levels at limit load, I was able to scale down the peak stresses or K linearly down to the cycle Ratio (Smin & Smax). Also, I don't know which original VA spectrum was used to arrive at the cycle counted one which was handed over to me.

At my spare time outside work, I want to learn the usage of a F&DT software to do DT analysis. I found a free version of AFGROW and I think it will do to start with. I do a lot of static repairs and I want to try to perform a DT analysis to some of the simpler repairs to get some familiarity of the software & procedure.

I've been going through the user manual of Afgrow and I have a few questions.

1. Spectrum: Most of our repairs are with in the fuselage. So which spectrum should I use to get an accurate VA loading representation. I understand TWIST is used mainly for Wing bottom skins and some argue that it is applicable to top portion of fuselage since they see similar loads (of course in fuselage internal pressurization hoop & longitudinal tensile stresses need to be added as well). Most of our repairs go on the cargo pit of the transport aircraft and thus the bottom portion of the fuselage. Any suggestions on what spectrum I can use and how to get that spectrum?

http://www.eng-tips.com/threadminder.cfm?pid=2&page=1

2. Load Level input to AFGROW: I am a little confused about this part as well. In the user manual, the Tensile loads/stresses are entered as ratio to reference stress. What is this reference stress? For ex: At a hole in a plate, if I calculate the stress due to pin loading along with stress due to bypass load at limit load, how do I enter that in to corresponding AFGROW model?

I have a person at my current employer who I can ask a few questions, but usually he is very busy and since this is not an employer sponsored exercise, I would want to limit my interaction with him only to items which are really mind boggling.

I am hoping to get a discussion going on here with the experienced good folks on this forum.

Thanks & regards,
- B

 

[verymadmac]

Find the supporting documents for the FAA's RAPID software,this gives Methods for getting the loads along the fuselage and are considered FAA acceptable, so I have been told.

Depending on the aircraft, there are FAA docs that give the spectrum (737-400, SA227 etc).

 

[Burner2k]

Thanks verymadmac,
I will check in to FAA RAPID software and its associated supporting documents.

I would appreciate some clarity on loading input format in to AFGROW.

Thanks in advance,
- B

 

[Burner2k]

OK folks,
I figured out the loading part in to AFGROW i.e. the reference stress. The manual has explained it with a neat little example.

I did go through the RAPID supporting documents. I was able to generate a generic transport aircraft spectrum but the output is in binary or a different format which I am unable to read (cryptic characters comes up when opened with Wordpad).

Any suggestions on how I convert the FAA RAPID Transport spectrum output in to a readable format?

 

[crackman]

Hi Burner2k

Ok, so the correct way to develop a spectrum is of course to developed balanced aircraft loads and then calculate fuselage bending loads, but this of course is involved as you may well know. As verymadmac suggested however you can use the method presented in RAPID. Know that the FAA did not invent this method, actually Dr. Broek presented it in his text book and you can also find a complete description of the method on the DTIC website in a report Dr Broek wrote for the FAA. The document is entitled "Generation of Spectra and Stress Histories for Fatigue and Damage Tolerance Analysis of Fuselage Repairs" by Dr. David Broek, Report DOT-VNTSC-FAA-91-16, AD-A250-390. You can utilize the method documented there along with the load histories available at the FAA technical library website. There are numerous commercial aircraft available (727,737,747,767,DC9,DC10,A320,EJ145,etc) in reports labeled DOT/FAA/AR-XX/XX.

Of note, there are certain key items to consider when using the RAPID/Broek method and load histories:
1. It does not account for difference between wing mounted and side of body mounted engines - analyst needs to adjust based on actual aircraft
2. Assumes a neutral or near neutral balancing tail load - again analyst responsible for adjusting
3. Assumes no dynamic effects for Gust, Taxi or Landing - again analyst responsible for adjusting (see AIAA Book "Gust Loads on Aircraft - By Hoblit)
4. When using a load history from a similar aircraft (allowed by FARs), ensure the configuration and wing loading is appropriate

Good luck on your efforts

crackman