Load sheet construction

[skyfarmer]

How do you go about constructing a load sheet. The sort with a graph where you divide the actual weight by a set figure and then move along the graph that number of divisions, then down to the next load station.

[john_tullamarine]

The devil is in the detail but the basics are -

(a) review the TCDS, POH, Weight and Balance manual, etc., to get a feel for the aircraft. Often it is a good idea to draw a picture of the load arrangement on the trimsheet so that everyone knows what is what ...

(b) choose a datum.

The datum has a significant effect on sheet accuracy and is the single most important decision the designer makes in designing a trimsheet. The typical ICAO/GAMA datum for light aircraft up near the nose is very inappropriate for any graphical loading system as the completion errors can be quite significant. Large aircraft often use a similar datum for station identification convenience but then run with a second, more appropriate position for trim calculations .. for lighties, the trimsheet designer has to do something similar.

The datum should result in a boxy, squarish shaped envelope (plotting weight against IU - remember that IU is just a non-dimensional moment). Trick is to use a datum station somewhere inside the envelope ... for the typical lightie, often the aftmost envelope limit works quite well .. although one might equally choose a loading station in the same vicinity to remove one trim line from the calculation.

To get a feel for this, try several datum positions ranging from forward through aft of the envelope. Forward datum positions result in a thin trim envelope sloping from bottom left to upper right on the sheet (typical GAMA POH envelope picture) .. central gives a fairly vertical envelope (desirable as the scale can be expanded to give maximum completion accuracy) ... aft gives an envelope similar to the forward case but sloping the reverse way .. ie from bottom right to upper left (not much point doing this as there is no benefit compared to either of the other options).

A lot of folk are uncomfortable with this idea .. there is NOTHING sacred about the manufacturer's datum .. the datum is no more than a nail on which one hangs one's tape measure. HOWEVER, once you choose your datum, you stick with it .. there is no changing horses mid-race ..

(c) pick a divisor which gives sensible numbers ... bit of a judgement call here and is worked out in conjunction with the selection of trimline division weights.

(d) revised calculation arms become

arm = load FS - datum FS

and the general IU formula for each trimline is

delta IU = (load FS - datum FS) * delta weight / divisor

(e) if the general formula is used for the entry argument at the top of the sheet .. in the form

IU = (CG limit - datum FS) * envelope weight / divisor

the zero position is at the datum, ie we get plus and minus numbers. .. bit messy, so it is usual to adopt a simple fix and add a suitable number for the entry IU to get rid of the minus numbers .. general formula for the entry line (only) - and the envelope (including MAC grid if you want to put that in) - now becomes

IU = scale shift value + (CG - datum FS) * envelope weight / divisor

(e) draw in the entry argument IU scale as a background grid.

(f) draw the envelope at the bottom of the sheet, including the admin bits (scale identification and suitable titles.

Note that constant CG lines on the weight against CG envelope become straight lines on the IU diagram .. but sloping straight lines on the CG envelope become parabolae on the IU envelope .. so you may have to plot a few points to fair a nice line on the trimsheet. Beware of CG envelope lines with negative slope (forward limit) and positive slope (aft limit) as these may NOT be simplified to straight lines on the IU envelope .. try plotting the parabola to see the problem. If the CG envelope line is a curve, then you will need to plot enough points to figure out the faired line on the IU envelope.

(g) if you want, draw in the MAC (or CG value lines - same thing but different numbers and names only) as an envelope overlay. Use the general IU formula to work out where these lines are to be plotted

(h) draw in sufficient trimlines to account for each loading position to be assessed.

(i) for each loading position, pick a suitable delta weight so that the equivalent delta IU is around 5 - 10 mm physical dimension on the trimline .. this usually works out pretty well but is up to the assessment of the designer. Alternatively, one can use "standard" things like numbers of passengers (at so many kg per passenger), volume of fuel (at such and such an SG), etc.

(j) draw in the ticks along each loading trimline. It doesn't matter where you start as the trimline only gives a delta IU for the delta weight ... the important thing is the distance to be moved to account for the delta weight, not absolute position on the trimline. If you prefer to have sloping lines to simplify the interpolations when using the trimline, make a copy of the first trimline with ticks a short distance away but with the ticks slightly moved on the second line. Draw in the sloping lines so that there is an overlap .. by this I mean that a vertical line coming into the trimline must intersect one or two sloping lines.

(k) draw some vertical guide lines to make it easier to keep things aligned when completing the trimsheet.

(l) put in the admin things for each trimline (weight/division). Also add an arrow so that we know which way to move to account for the delta weight. If you choose to put slope lines instead of ticks, it is usual to slope them in the same direction as the arrow .. so that you "bounce" off the slope line in the direction of the arrow.

(m) add in the addition table to one side of the trimlines. Add any limits you might want to publish.

(n) add in any regulated information required by the jurisdiction ..

(o) some sheets may include the standard load message form but that is just a matter of adding it in as a second drawing on the one sheet.


Beyond that, it is really a matter of just making the document have a pleasing appearance, emphasising simplicity so that the poor sods who have to use it can do so quickly and accurately.


Three additional considerations ..

(a) if the fuel line has a CG varying with volume (ie a non-prismatic tank), you can address this by one of the following techniques ..

- if the variation is not too great, treat the fuel line as a constant arm and adjust the envelope lines to account for the error

- plot the trimline as a curve reflecting the fuel CG variation and then plot a series of parallel lines which form guide lines .. you will need a weight or volume scale at the side of this trimline

- if the curve is a bit complicated, draw it as a separate grid and then use straight lines in the trimline to be the guidelines

(b) for aircraft which check the basic trim at ZFW but still need to calculate a takeoff CG for stab trim (eg Gulfstreams and HS125 etc.) you can account for the resulting two envelopes by using a scaling grid for the takeoff envelope which moves and resizes the scaling for the IU line scale .. this allows you to plot the two envelopes one over the other which keeps the completion accuracy high. Just a drafting sleight of hand thing, really.

(c) keep in mind that the real world has lots of small (and not so small) errors in loading calculations. You need to do an error analysis to account for these ... the usual trick is to constrain the published envelope limits to account for the errors .. ie we acknowledge that the final plotted data is in error but constrain it so that the TC envelope is not compromised.. this often results in an error in the stab trim setting but that is not usually at all critical. A minor variation here is to use LMCs (last minute changes) envelope lines .. same philosophy but presented is a different way.

Ignoring (c) above, don't let anyone tell you that the basic trimsheet (ie not considering error analyses) is not accurate. .. reasonably designed and executed, a trimsheet loaded CG should be accurate to something in the order of several mm. This is a bit artificial, of course, as the empty weight and CG data for the aircraft usually is in error somewhat more than this.


Once you have designed your first couple of hundred sheets they become pretty routine .. but expect to have to rework the first few quite a few times until you get the hang of things ...

I have trained up work experience uni engineering students (who often knew not much about aircraft beyond the academic and nothing at all about loading) in a few days to the stage where they could design a pretty passable trimsheet .. no rocket science needed for these things .. just a bit of attention to detail.



As folk might offer any comment, I will add more words to clarify any points above which might not have been written as clearly as they might have been .... it is getting a bit late at night here, you see.

[skyfarmer]

Hi John
Thanks for that.

Will print out, digest it and start having a play with a few figures.

Have been looking at the companies sheet, that they inherited from the previous owner. it is good enough and does work, but was hoping to make it a bit easier to use. Also needs re drawing as even the master that I am photocopying from is probably a 20th generation copy.!!


Cheers

Neil

[john_tullamarine]

Neil,

By all means send me a scanned image if you get stuck ... I am only too happy to offer nuisance comments ...

Having done sheets in past years for most, if not all, 310 variants, it is an easy aircraft to design for so you ought not to have too much trouble ... do try using the aft limit as your datum .. works well and gives a fairly simple, tidy sheet.

[skyfarmer]

Ok john thanks,

Will not be any time soon, busy round the farm at the moment, and flying also picking up...never a moments peace in the last few days


Cheers

[john_tullamarine]

... join the club, buddy ... sleep is just that short diurnal thing one experiences between bouts of work ...

[skyfarmer]

Hi John,
well I have been over your instructions a few times, and found i was stuck before I even got started!!

Your paragraph (b) refers to deciding on a suitable datum to use, to get the sheet a suitable boxy shape......but how do you get from the datum you decide on into a shape for the load sheet???

Am I being just plain dumb here or what ???

Also TCDS and UI, have not found any one here that can tell me what they stand for T****C**** data sheet perhaps. U**** Index? F*****Station


Our current load sheet has different divisors for each station, 180 kg for pilot, row 2, 3 and wing lockers 22kg, rear and nose bags 9kg and aux fuel tanks 45 kg, with the sloped lines for each load station being different widths

Any further help greatly appreciated, or can you poin me in the direction of a good book on the subject?


Neil P

[john_tullamarine]

Neil,

Not too much in the way of suitable books which go into the nitty gritty .... perhaps I might have to write one sometime ...


(a) datum.

If you play with various datum positions, you will see that the shape of the envelope varies from

(i) the typical GAMA shape (thin thing sloping from low weight/forward CG to high weight/aft CG) - forward datum

(ii) upright - which you can then expand to make the desired boxy shape - mid range datum (best generally will be somewhere in the back half of the envelope - for lighties, generally the most aft limit works well)

(iii) similar to (i) but sloping the other way - aft datum.

Don't waste time .. go for the (ii) option to get the best results and fine tune it from there.

For the calc side of things, with a "standard" datum, the formula is something like

IU = (mass x arm)/convenient non-dimensionalising number

If you have changed the datum and still need to refer to, say, the OEM POH (and this is the general case) then you need to "change" the arms to reflect the new datum while retaining the link to the OEM information.

If you draw a piccy, you will find that the answer becomes something like

IU = (mass x (OEM arm-datum))/number

where the new datum is defined with regard to the OEM longitudinal station diagram.

Think about it - if the standard arm is, say, 100 inches then the arm is 100 from the OEM datum. If you now decide to make the datum, say, 40 inches (referenced to the OEM FS), then the "new" arm is 60 (ie the distance from the new datum to the location). This can be summarised as

new arm = OEM arm - new datum = 100 - 40 = 60

Nothing stupid about it ... just takes a little while to get used to thinking a new way ..


(b) Sorry about the acronyms ..

TCDS Type Certificate Data Sheet (http://www.airweb.faa.gov/Regulatory_and_Guidance_Library/rgMakeModel.nsf/MainFrame?OpenFrameSet). These are an extension of the Type Certificate and, with the AFM, give you all the limiting data relating to the aircraft Type.

IU = index unit = moment/convenient number

FS = fuselage station, ie the loading arm picture in the POH


(c) re your sheet ...

The mass figures at each line define the standard (mass) load value for each division shown on that particular line.

The sloping lines are a graphical trick to improve the accuracy of the sheet's execution by removing one of the interpolations at each line. Each sloping line to the next as a distance represents the IU (or moment, if you prefer) difference (delta) relating to adding (or subtracting, if you go the other way) the standard load (ie the mass figure described above) at that station.


Suggest that you keep asking the questions here as others will be interested in the answers... and if my answers don't get the message across, ask again and I will rewrite them until I get it right in terms of your understanding what I am trying to get across.

[IFixPlanes]

http://img185.imageshack.us/img185/6910/trimmsheetuaube6fv.th.jpg (http://img185.imageshack.us/my.php?image=trimmsheetuaube6fv.jpg)

[Flap Sup]

John and Ifx,

I have always wondered why an airline would choose to create a balance sheet where the cabin is only divided into two sections? In Ifx's excample, that makes about 95 pax in each section. The error margin must be huge (in case of half full flight) if 50 pax are seated in the forward part if 0A, and 50 in forward part of 0B, instead of 100 pax spread evenly in the cabin. The answer cannot be "the cabin staff has to ensure pax are evenly distributed", as that reply would eliminate the need to divide the cabin into sections alltogether.

Anyone?

Bgrds
FS

[john_tullamarine]

Yer pays yer money and takes yer chances ... while the answer may seem flippant, it's not.

All engineering decisions involve compromise to some extent, whether that be large or small. No different with a trimsheet.

If one wants a maximally flexible and detailed trimsheet, it is simple enough to design. The advantage is accuracy at the expense of taking longer to fill it out, etc.

If, on the other hand, the loads are pretty predictable so that the errors and error analysis need not be too restrictive operationally, then the calculations can be dumbed down to increase the speed of sheet completion and apparent operational simplicity. The penalty in the case of the present sheet is that

(a) there will be company prescribed requirements regarding seating distribution to minimise the error limits on the sheet

(b) the cabin staff will, most certainly, have to ensure that the design presumed pax distribution is reasonably met. Typically, this will involve each of the two zones having either a generic desired loading distribution or a specific seating order. Generally, it is only necessary to make sure that the distribution is reasonably even and the error analysis in the sheet's design takes care of the maximum expected errors.

If the above is not practicable .. then you go for an increased number of loading zones. All a case of compromise between accuracy (and minimising error restrictions) and operational flexibility.

[212man]

John, any advice on programs to draw it with? I designed one 4 years ago and was able to calculateall the relevant numbers and draw a rough version on graph paper to check for gross errors, but had to get 'a man' to draw it accurately with Auto Cad. Ideally one would use a simpler program to make it more practical.

[Ignition Override]

We left the gate a few years ago and either the Gate, Ramp or
Load Control Agent (or a computer glitch) left out 14,000 pounds! We had 60 passengers in Coach section, whose weight was NOT included. :hmm:

This was 14,000 pounds too light, for an aircraft which would have weighed about 95-108,000 pounds for most takeoffs.

This appeared on the ACARS screen for the FINAL takeoff weight and balance, mostly displaying the pax/crew count, full plus flex. EPR setting, actual with max. runway (i.e. 18R) weight, max. weight for all conditions/runways and whether 15* or 5* flaps and STAB. setting (i.e. 5.0).

Maybe our gate agents always sends separate counts for First, then coach passengers? If so, why not just process one number, after weight and balance is processed?

[IFixPlanes]

@Flap Sup

Those Load & Trimm Sheets are accurate enough to fly absolute save.
In februar 2002 a LH B737QC does a flight wide outside the limits.
Even if you only have two zones in cabin you didn´t get so far outside the drawed limits.
Sorry, the report is in german but take a look at the Load & Trimm Sheet at the end.
Report (http://www.bfu-web.de/nn_41544/DE/Publikationen/Untersuchungsberichte/2001/Bericht__5X003-0.01,templateId=raw,property=publicationFile.pdf/Bericht_5X003-0.pdf)

[john_tullamarine]

A few things to keep in mind ..

(a) it is not difficult to automate the drawing of a trimsheet providing that you have some programming skills but why bother ? -

(i) the numbers can be generated very quickly in a spreadsheet

(ii) you can modify an existing sheet in a drawing package without too much effort

(iii) once the sheet is drawn, it doesn't need to be revised unless the configuration changes

(b) any loading calculation (manual, computer, trimsheet, whatever) is only as good as the data and care which go into its completion. We have all seen trimsheets presented with errors. It follows that any sensible system (from a pilot viewpoint) involves a pilot check at the end of the chain ....

Probably the best system I have seen for trimsheets was based on a circular trimsheet (looked a bit like a nav computer). Originally, I was sceptical about the presentation but the customer insisted .. after seeing it in action I was convinced of the value. Completing the calculation took only a few seconds. This operator's procedure was to have the load folk do the loading and prepare the documentation. Then the flight crew would do a recheck with one pilot reviewing the loadsheet data while the other ran the prayer wheel to check the trim. Near foolproof system.

However you do it loading is pretty critical. If the system doesn't have appropriate checks and balances then it is set up for a fright or two along the way. If the system's checks don't involve the crew .. then, if I were crew, I'd involve myself at the end regardless ..

(c) So far as the number of loading zones is concerned it really doesn't stop the careless person from running the trimsheet incorrectly. The ideal sheet is simple to use and a good balance of flexibility and speed. But, at the end of the day, it is very difficult to keep the idiot factor totally at bay. Concerning accuracy, a well designed trimsheet is as accurate as a longhand calculation for practical purposes .. but it does require appropriate technical care in the execution ... which also needs those using the system to have a reasonable idea of its requirements and limitations.

[744rules]

It's a bit off topic, but how are CG limits determined (on zfw, tow, and law). I'm a bit familiar with ahm560 (former ahm50), where the operator gives all the info required to load in a computerised system (such as the index effect for crew, deadload, pax, max wgts, catering wgts, etc...)

[john_tullamarine]

LAW ? not seen that one before .. perhaps you can elaborate ? .. perhaps landing CG limits ?

For the trimsheet designer, the AFM (POH) is the source document.

One of the final exercises in the sheet's design is to account for errors inherent in the design and, more particularly, intentional simplifications included by the designer. Once these are totted up the easiest way to address them generally is to compress the sheet envelope limits ... the astute user of such sheets may have wondered why the sheet forward and aft limits are different to the AFM's .. this is the answer to that question.

This leads to a concern with trimsheets .. I have seen many where the designer gives no (apparent) consideration to thoughts of errors ... not too sure how the user would fare at the coronial if the opposition shows that the trimsheet use (while "correct") led to a critical misbalance .. the designer, presumably, would be rather more exposed at law.

The calculated sheet may define a CG which is in error to some extent but the compressed limits guard against the actual CG's being outside the Type's prescribed limits. In practice the only problem is that the stab trim setting calculation will be a tad in error but that is not going to present a significant problem and is accepted as a reasonable practice.

For the aircraft OEM, figuring out the CG envelope is a significant part of the whole aircraft design workup and probably is a bit outside the thrust of this thread ?

Regarding computerised loading calculations, they are just a fancier (electronic) version of the manual longhand calculation system ... same as is the graphical trimsheet. Presuming that either has been designed accurately, the principal problem .. as with most things in technical life ... is GIGO.

One significant problem with electronic calculations is that the flightcrew have little means of running a final double check on the sums. Maybe I'm a little old fashioned but I would prefer that I die as a result of my own incompetence rather than in ignorance of someone else's ...


After re-reading the previous post, it may be that the question regarding CG was directed at how the trimsheet does the calculation ... if so ...

(a) as the various trimlines are worked, the moving left or right subtracts or adds (respectively) the IU (or moment) calculated for the particular trimline for the mass added/removed (if added, go in the direction specified on the line, if removed, go in the opposite direction).

(b) by the time you get down into the envelope you have figured the running total for the entire load, whether ZFW or ramp. (Keep in mind that you can run down to the envelope at any stage in the calculation to check the CG).

(c) the envelope is a plot of gross weight against gross IU (or moment) and provides all the information to determine the loaded CG.

(d) by reversing the IU equation one can calculate the CG. Generally we are not interesting in knowing the specific CG, rather the fact that we are in/out of the envelope. However, if you read off the weight and IU the sum is straightforward.

(e) many sheets have an overlay grid superimposed on the envelope so that the weight/IU intersection can be read against the overlay grid to give the CG directly.