fdr,
An aerodynamic performance model is only as good as the effort which went into its development and validation .. this latter normally being a combination of wind tunnel and flight test activities.
If the data has been developed on the cheap, as it were, then one normally sees an appropriate level of conservatism. This might be the case where a particular aircraft needs a performance chart to cover a modification in the field and the local PEs do the job .. depending on the needs of the customer, cost generally dictates that a conservative approach might be the way to go.
On the other hand, for a major project line, ie the larger manufacturers' various aircraft models, one would expect to see a LOT of money get thrown at the performance estimation and validation work, as this has a big bearing on the commercial success of the project. But don't expect an ounce of needless (ie not required by the rules) conservatism .. the manufacturer is out to get the best data for commercial purposes that is feasible from an engineering point of view .. keep in mind that it has to get through the certification invigilation process and that is often an exercise in heated discussion.
If one cannot presume that the AFM data is reasonably able to be re-validated by test (ignoring the usual minor test point skewing which attends being a little selective about which test points are rejected during the test and, subsequently, the data analysis programs), then this probably amounts to suggestions of conspiracy and fraudulent activity. While mistakes and poor judgement can be found in the history of aircraft certification, I think that wilful and deliberate misdeeds are rather uncommon.
The AFM provides the bottom line requirements to meet the airworthiness Design Standards. In some cases additional levels of conservatism are mandated by operational rules. There is nothing to prevent an operator making a commercial risk assessment to add further levels of conservatism (other than commercial pressures, of course).
But none of this suggests that the AFM data is going to be faithfully reproduced day-in, day-out in the real world .. it just doesn't work that way ... unless the conditions and techniques are near identical, then there is no reason to expect the outcomes to be the same .. is that not a reasonable logic ?
And if one is going down a path of the "actual method of takeoff being totally different to that from which the performance figures are derived" then that is a problem belonging to the crew on the day .. nothing to do with the OEM I would suggest ... and the real world problems which Mutt and I push in these threads is not the province of the OEM or the airworthiness rulemaker due to the variability of the beast ... it is up to the operational rulemaker and the operators to put a finger in the wind and determine where their levels of regulatory and commercial risk comfort might lie ...
As to your "smoking evidence", of course this is the case ... if the aircraft is heavy (ie a routine commercial operation), then small real world variance from the certification assumptions makes a very critical change in the game .. but it is not an OEM or airworthiness problem .. rather an operational regulatory and commercial risk assessment problem.
Or perhaps .. your mind is made up and discussion in not going to change things as you see them ?
Quite obviously you are referring to specific investigations, events or accidents .. like Mutt, I would be very interested in the details as there is always something to be learnt along the way.
Alex,
Probably more light might be thrown on the subject were some of our appropriate specialist flight test and certification colleagues wade into the discussion as they would have the more detailed knowledge and may be able to point to specific documents of which I have no knowledge.
However, much in the rulebook is amplified in the design and certification processes and practices .. in the same way as precedent plays a major part in law.
The usual philosophy is that one ought to design and build to suit sensible engineering and commercial constraints while making sure that one doesn't miss passing the rulebook tests .. or, at least, negotiating a concession against the odd rule which provides a stumbling block.
When it comes to the rules, the explanatory detail in the ACs is essential as is that in many of the other documents which are not in general circulation. Hence the often made suggestion that one ought not to read the rulebook and try to make determinations on this and that matter on that sole basis .. only leads to confusion.
All I can suggest is that the logical intent of defining Vmcg is to "put a line in the sand" and that it would make little sense to use other than a reasonably conservative set of boundary conditions in that determination ... hence CG, thrust, and the other factors called up in the rules. It would not be useful to try and schedule a great variety of data to cover variations in many of the parameters. Unless the scheduled Vmcg covers wet and dry, then you are straight away into the arena of more cost and more data for a not very general benefit .. on most occasions, Vmcg considerations are not limiting.
If one is going to look at an aft limit cg case, then the fact of the matter is that nosewheel loading, and hence steering capability, is fairly minimal. If the added problem of wet runway is to be addressed then the problem increases .. is there any other practical way of addressing it other than by requiring the removal of nosewheel steering from the cooking pot ?
Again, there is nothing to stop an operator from adopting additional levels of conservatism .. as you would be aware, Mutt and I have a jaundiced view about blindly using min weight/speed schedules in conditions of strong crosswind if there is spare runway available to trade our way out of a potential Vmcg embarrassment ...
When you find a specific reference to support your concern, do let the rest of us know ...
Quid,
You answer your question indirectly in your final comment.
There is a set of cases to be considered in determining the limiting takeoff weight on the day. Each has to be addressed and the one yielding the lowest weight (OK, mass for the purists) gets the nod.
If, as might well be the situation, a particular aircraft doesn't ever have a problem with one case for the envelope to be scheduled in the AFM, then it may well be ignored in the data presentation ... or the OEM may choose to adopt a simplification in data presentation for whatever reason even if that might involve a conservatism. These thoughts would be relevant in respect of Mutt's comment above ...
For rules of thumb it would be a case of reviewing a sufficient sample of AFM datasets to develop a rational thesis ... I have never done this for such a purpose so I ought not to comment on what you have read elsewhere in the past. Personally I can't see much benefit to be had in rules of thumb for this discipline as there is more than enough variation to be had to complicate the issue on any given point.
This is a little different from getting a feel for a particular Type and knowing intuitively which case is probably going to be limiting in a particular set of circumstances .... If you are talking about 742 as in the thread topic, I have a few AFMs for the model in the cupboard .. but, as Mutt is au fait with the series, his implied answer above saves me the effort of playing with the books to get a feel for the situation in that case ...
I don't think that one is BFL-limited as such .. rather BFL may offer some advantages in calculation effort and data presentation .. if, indeed, the particular AFM provides data to permit both balanced and unbalanced data runs. What are your philosophical thoughts on this point, Mutt ?