QAR decode/ hard landing?

[Qbix]

Recently I requested QAR data from my alleged 1,6G landings (B737NG). The thing is that these landings in my opinion were far from anything close to 1,6G landing.
I know it because in fact once I did 1,54G landing (another company) and thought that left a piece of airplane behind. Really embarrassing...

The company I fly for is crazy about soft landings (China). They are providing on request QAR data which I think is sampling at the rate of 1 frame/ sek. Pretty slow to me.

Among other things there are 8 air/grnd sensors, 8 accelerometers, RA, IAS, GS, VS, pitch,N1, etc
The landings look normal. N1 ok, descend rate reducing from 800-900 to 50-100ft/min, speed ok, pitch normal...
You can see from following frames that another A/G sensors change from AIR to GRND. Then there are 8 values of vertical acceleration. At the peak of any of them the values are:
1,63 1,37 0,87 0,68 0,91 1,18 1,21 1,02. As you can see one of them reaches 1,6G and this for the company is hard landing.

Another example at the touchdown: 0,76 1,23 1,73 1,61 1,01 0,60 0,72 1,05

I wouldn't have any doubts if I smashed the plane to the ground but again these landings were soft! The only thing I have doubts about is that I didn't fly the front gear to the ground. Just dropped it... Maybe it has something common with this?

Another strange thing about the whole system is that if you pull a really nice landing and go to the MAINT/ ALPHA through the FMC you can see VGTD of 1,37-1,41G Sometimes even 1,55G!
In the previous company I could check max G load as well. Did it different way (can't remember how) and usually the landings were around 1,05- 1,15G.

Any clues??
Does any of you have access to more detailed information about QAR system?

[STBYRUD]

I dont believe the data at all - the airplanes with ACMS display wildly random values for VGTD, which seems to be the data you are interested in - those things arent built to measure an instantaneous acceleration accurately, not even considering the low resolution. Smooth Flaps 40 landing, ACMS claims its 1.64. First time I was shocked, now I just laugh - Hard crosswind landing, lots of shuddering, 1.43. Forget about it. To my knowledge the definition of a hard landing is still that "the PIC determines that the landing was a hard one", not a box giving a soft warning that it has measured something above a certain threshold... I remember the code VACMAXLA, supposed to give the highest measured G during landing, try that for a change - its usually different from VGTD.

[punk666]

STBYRUD,

I agree also the figures are a bit wild at times! I think the nose wheel plays a part in the calculations. The reason I have this theory is because I did a smooth landing the other day but the nose came down a bit heavy so the figure I got was 1.7.

I have been told from my father (engineer) that the range is usually beween 1.4-1.7 for normal conditions anything reading 2.0+ is considered a hard landing.

[decurion]

Determining whether a landing is hard using QAR data should be done with great care. The air/ground switch is usually recorded at a low sample rate on the QAR which will affect the results if you are looking g loads at impact. Strictly hard landings should not be based on g-loads but on sink rate exceedance. However in practice g-loads are used or pilot reports. I would recommend to have a look at this article by Boeing: http://www.boeing.com/commercial/aer...onditional.pdf.

Personally I favor firm landings especially on wet/flooded runways to promote wheel spin-up. Soft landings often tend to be long landings too...

[iflytb20]

My company used to follow the same policy till we lost an airplane. It seems if you let the nose wheel drop, you will get a high VGTD and if you hold if off for a sec / fly it down to the runway, the VGTD will be a low one. If you ask me, i'd tell those chaps to stuff it

PS: i think the code you were looking for is VGMX [max G] and VGMN [min G]
Cheers

[Qbix]

decurion, thanks for the link. Looks like FDR can sample 4,8 or 16 times/sek. However not much information about ACMS which in fact airlines use. The data I got is in 1sek steps so not much to see. RA even sampled every two seconds. Crazy.

VGMX, VGMN as far as I know are the max and min from the whole flight not just landing. By the way where can I find ALPHA codes?

[A664]

Qbix:

They are providing on request QAR data which I think is sampling at the rate of 1 frame/ sek. Pretty slow to me.

Among other things there are 8 air/grnd sensors, 8 accelerometers, RA, IAS, GS, VS, pitch,N1, etc
The landings look normal. N1 ok, descend rate reducing from 800-900 to 50-100ft/min, speed ok, pitch normal...
You can see from following frames that another A/G sensors change from AIR to GRND. Then there are 8 values of vertical acceleration. At the peak of any of them the values are:
1,63 1,37 0,87 0,68 0,91 1,18 1,21 1,02. As you can see one of them reaches 1,6G and this for the company is hard landing.

Another example at the touchdown: 0,76 1,23 1,73 1,61 1,01 0,60 0,72 1,05

In the QAR systems I know a frame can contain the same parameter multiple times in order to facilitate sampling more than once per second. Therefore, I guess that these 8 values are 8 measurements from 1 accelerometer during 1 second - this would seem like a usual recording rate for vertical acceleration. It might be similar with the A/G sensor(s). However, I'm not familiar with the B737 and don't know if this interpretation is consistent with the form in which the data were presented to you.

[decurion]

There is often a big difference between the sample rates of parameters on the QAR compared to the FDR. The actual sample rates used on the QAR can also vary between different aircraft types. FDM software often interpolates the missing data to match it with the data of a higher sample rate. If the air/ground switch is sampled at a lower rate than the normal acceleration this type of interpolation will affect your results. It is always important to consider these issues when analysis QAR data: what does the FDM software do with the data, how does the raw data looks like etc.

[punk666]

All the Alpha codes can be displayed on the CDU. The LSK for the codes is on the same page that you select alpha display.