686. Hypothesis on Weight Growth During the Development Stage of Aircraft

Paper

Abstract

A weight growth from the preliminary design aircraft to the actual airplane is observed for nearly all known aircraft developments and it seems that weight growth during development is unavoidable.
In general, weight growth can be divided into two major classifications:
– in-scope weight growth is a weight change that, is independent of any change in contractual requirements
– out-of-scope weight growth is a weight change caused by a change in contractual requirenents
This paper pertains only to the in-scope weight change and presents a hypothesis which explains weight growth by the fact that various items inthe preliminary weight breakdown of the initial design are subject to errors even if the total weight estimation of the aircraft is consistent with statistically known total weights of already developed and comparable aircraft. There always will be an error distribution in the preliminary weight breakdovn. Some items will be estimated lighter and some items will be estinsted heavier than the actual weight. It is assumed that the total of all positive errors weigh up the total of all negative errors. However, parts which are estimated too heavy will have a different effect on weight growth than parts which are estimated too light. All parts estimated too light will not be strong enough to withstand the specified loads and consequently they will be discovered and strengthened. But just some of those items which would withstand more than the specified loads will be identified and made lighter. For example, when the landing gear is discovered to be overstressed and consequently too heavy, and its overweight eliminated by a redesign at the end of the development, there wi1l be still an overweight which will not be eliminated in the rest of the structure due to the original overweight of the landing gear.
The approach to calculation of weight growth uses the aircraft growth factor (a) as a tool and as a characteristic for the sensitivity of the aircraft due to a change of partial weight. A negative error (-el) of a weight item 1 leads to a decrease of total project weight equal to (-el)(a), but does not contribute to any overweight of the actual aircraft. A positive error +e2 of a weight item 2 leads to an increase of project weight equal with (+e2)(a). Due to the assumption, only the amount e2 will b eeliminated during the development thus producing a weight growth equal to (e3)(a-1). Going through the error distribution of the weight breakdown, the total weight growth can be calculated as the total of all individual weight growth contributions.
Weight growth and overweight of the final product can be influenced by adapting an arbitrary positive or negative weight to the statisticalproject weight of the preliminary design. Based on this hypothesis, overweight and weight growthof a sample aircraft and a sample weight error distribution was calculated versus arbitrary under- and overestimation of the project weight. It is shown that by a suitable Overestimation of project weight over the statistical expected weight the probability of a weight growth can be decreased but with a penalty on overweight. A suitable underestimation of project weight will decrease the probability of overweight over the statistical expected weight, but will penalize the weight growth.
Using the equations for weight growth and overweight for one development loop, it can be demonstrated what would happen, when weight growth experience of a previous development is adapted to the project weight of a second development of a similar airplane, and so on. After many loops, weight growth will approach zero indeed, but overweight is twice as big as overweight and weight growth of the first development.