Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Methodology for weight reduction

  1. Jan 24, 2015 #1
    Hi all,
    as you know, aircrafts are highy dependant on their weight goals to be met or exceeded to ensure the performance percieved during the design studies can be realized. However, it is nonetheless often that the resulting design still becomes overweight in some way or place because you simply cannot account for all design changes that will take place moving a design from the conceptual stage to realization. It might be that stiffeners need be added or a systems engineer shows up with another heavy enclosure.

    Enough changes and there will be a need to find and optimize components for weight. Working with aircrafts though, there is easliy a lot of systems, structures, connectors and parts that might be considered to rework.

    What i am searching for is a way or methodology that will help in searching and selecting what parts to work with that gives the most "bang for the buck". In some ways it might be considered a multi-objective optimization problem since the goal is to reduce:
    • Weight
    • Time spent redesigning
    • Cost of redesign (almost equal to time spent)
    • Performance impact
    To my question: Does any of you have any suggestion of how to set up a general method that can be applied to this? All suggestions welcome. An excel document with weighted numbers might be it. Suggestions on other typical areas of interest are also welcome. Since I'm a junior engineer, probability is high that I've overlooked something.

    Thanks and best regards,
  2. jcsd
  3. Jan 24, 2015 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    I think you will find it difficult to develop a general method for weight reduction for aircraft, primarily because different types of aircraft have different design goals. For example, a glider will need to have the lightest possible structural weight which can support its payload and the aerodynamic loads imposed on it. There will be no weight devoted for powerplants or fuel or landing gear. On the other hand, a fighter jet design will have the best compromise of speed, maneuverability, range, ability to take damage to the airframe, etc. while being able to accomplish its mission. Some aircraft may be needed to fly only a few missions; other aircraft may be expected to remain in service for decades.

    We haven't reached the point yet, IMO, where a set of design parameters can be dumped into a program and an aircraft design emerges. That's why we train engineers and give them experience trading off different aspects of design and performance and cost to develop the best compromise which meets the design goals.
  4. Feb 7, 2015 #3

    Generally speaking when designing an aircraft even at the conceptual phase an engineer will have a good "rough idea" of the weight of the aircraft. This is found by looking at historic aircraft data.

    Generally most modern aircraft will have a similar predecessor which has already been built and the weights of components already established. The new design will be an evolution on this aircraft and the engineers will consider what new technologies or configurations are being applied to the design and what the likely impact on weight will be. These estimates will be based off detailed trade studies of the new tech they might be implementing.

    Once they go into the detail design phase of the project they will establish what the exact weights are (which should not vary from the estimates by too large a degree).

    As to answer your question, for a modern aircraft to be commercially successful every component should be optimized in terms of its weight unless removing the weight from the component is prohibitively expensive.
  5. Feb 7, 2015 #4
    The problem is that general methods often only solve the technical problems. I'm currently participating in a new aircraft build at work, and the biggest reason (for us anyway) for bad design choices is schedule. The second biggest reason is politics.

    For example, everything might be going well, but the aerodynamics group may get the loads slightly wrong and then quickly correct them. That then changes the loads, and that forces designers to redesign parts, which then have to be reanalyzed by the structural analysis group, then kicked back to design to modify the design if the margins are unacceptable, and then kicked back to analysis to make sure that change is acceptable, and so on. Meanwhile, each iteration is taking up more time and making things later and later. At some point, some decision has to be reached to save schedule, so they go with a quick, unoptimized solution that works and they explain it away by saying they'll optimize that on a future build (read: never). Depending on the severity of the mistakes, several parts may need to be redesigned or another structural design concept for a certain area may have to be employed. Also, if you're trying to manufacture parts as they get completed, there might be parts already made that may have to be scrapped and then the tooling engineers may have to create new CNC tool path programs for the new parts.

    As far as politics goes, it's a game you have to play sometimes. Since there's more than one way to do something, several concepts for design/manufacturing/etc. will be brought up in meetings. Sometimes a substandard plan may be advanced by someone the group implicitly trusts (or an engineering manager/VP or something) and a lot of time is spent trying to convince the higher-ups that the plan won't work as well as plan B. They usually say things like "let's investigate the merits of both" and choose plan A anyway, or maybe they see the light and choose B. Maybe both plans might be equally as good, but half the group (or a manager/leader type) likes one over the other and it takes a while to finally downselect to a concept (either through hard numbers or just intense peer pressure). Either way, it doesn't help to get things done on schedule, which then leads to quick, unoptimized decisions as before.

    The problem with using methods like the ones you're searching for is the resistance to change in industry. To illustrate, the project I'm working on is essentially an optimized version of an aircraft already in operation. The name of the game for us is weight savings and optimization. That being said, a lot of us have been told NOT to use tools like topology optimization because it takes too long and the parts cost too much. You would think that we would be using everything at our disposal to cut out every non-essential piece of material, but nope. It's usually the older engineers who advise against using new and often unproven methods, but it's hard to argue with them because they've been around the block before and they usually know what they're talking about. Besides, no one wants to bet the millions of dollars that go into a project on something that *might* be innovative. If they lose, the company might fold.
  6. Feb 8, 2015 #5


    User Avatar
    2017 Award

    Staff: Mentor

    Those issues are not limited to airplane designs, you'll find them everywhere in complex projects. There are so many components working together that no one can have a detailed overview of all of them. Sometimes you have to follow multiple design options in parallel to explore which one is better, but you cannot develop 10 different final products, so at some point you have to decide on one even if they are not studied in all details. Sometimes you get some other constraint from somewhere else and have to redesign something, and so on.
  7. Feb 8, 2015 #6
    Obviously these problems plague any complex project. No argument there. I'm just trying to give some visibility to the fact that often the problems that plague aircraft projects are not ones that a technical method would help. It seemed to me the OP was looking for a method to help with the design process. Having studied the design process and optimization in depth and co-authored a couple journal articles on the subject, it seems to me the main issues are team dynamics. Sure some methods or software programs can help avoid design turnbacks or conflicts, but in a way you could say the people are the problem. Every aircraft company (and university aerospace program I would wager) has their own "secret sauce" method on how to develop sizing parameters quickly. If any of them are anything more than marginally better than any others, you can't really see the effect because of the other problems that occur when working with several teams stretched out in different time zones and locations.

    Maybe I'm off and the OP was looking for a technical method. In that case, ALCCA and FLOPS might be good things to look at to start.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook