Why was the F-35 overweight?

Discuss the F-35 Lightning II
Forum Veteran
Forum Veteran
 
Posts: 578
Joined: 23 Nov 2003, 01:51

by F16VIPER » 12 Nov 2009, 03:57

I wonder how an aircraft designed in 3D with the latest CAD software, and all the computational power available to a corporation such as Lockheed Martin, could end up overweight. I am aware that that issue with the F-35 now appears to have been resolved, but you see other aviation programmes with similar problems A400M, 787, 747 etc. Is there a reason as to why tons of extra weight seem to appear out of nowhere to the outsiders.

Regards

F16VIPER
Last edited by F16VIPER on 12 Nov 2009, 08:58, edited 1 time in total.


User avatar
Elite 5K
Elite 5K
 
Posts: 28404
Joined: 05 May 2009, 21:31
Location: Australia

by spazsinbad » 12 Nov 2009, 04:48

Cannot answer your question - however whilst looking for info came across this ANTI-JSF brief PDF for Dutch Parliament 22 May 2008:

http://www.jsfnieuws.nl/wp-content/BRIE ... AY2008.pdf (1.3Mb)

Amongst other things redesign work bumped up weight (redesigning an electrical generator for example) apparently.
Last edited by spazsinbad on 12 Nov 2009, 13:09, edited 1 time in total.


Elite 4K
Elite 4K
 
Posts: 4451
Joined: 23 Oct 2008, 15:22

by wrightwing » 12 Nov 2009, 12:55

F16VIPER wrote:I wonder how an aircraft designed in 3D with the latest CAD sofware, and all the computational power available to a corporation such as Lockheedmartin, could end up overweight. I am aware that that issue with the F-35 now appears to have been resolved, but you see other aviation programmes with similar problems A400M, 787, 747 etc. Is there a reason as to why tons of extra weight seem to appear out of nowhere to the outsiders.

Regards

F16VIPER


One reason would be using less expensive but heavier materials, to help control costs, in the construction.


Forum Veteran
Forum Veteran
 
Posts: 753
Joined: 13 Nov 2004, 19:43
Location: 76101

by fiskerwad » 12 Nov 2009, 13:01

Some of the time it's requirements creep. Added capabilities without adjusting target weight.
fisk


Enthusiast
Enthusiast
 
Posts: 33
Joined: 21 Dec 2006, 03:57
Location: west of the runway

by falcon_mr.b » 12 Nov 2009, 15:29

I will try to say this in general terms, because as pointed out this is not a specific issue to the F-35. This is perhaps the oldest (and maybe most difficult to solve) problem when designing any new aircraft, as weight is one of the biggest concerns. During initial design phases, the internal structure detail parts are designed using a standard number for "factor of safety". (F.S. is a technical engineering term which is a number by which the limit load of the designed part is multiplied to get a ultimate load for that part. Limit loads are the maximum loads that the part will most likely see during its service life. Each structural part should be capable of supporting the limit loads without plastic [permanent] deformation where the defomation would interfere with the safe operation of the aircraft. The end requirement is that the aircraft handle the Ultimate loads without failure.) It may be obvious, but the higher the F.S. used in initial design, the higher the weight of the part will be due to high web and flange thicknesses. Most companies start out with an ultimate factor of safety of 1.5. As the design process progresses, certain parts are revisited by the structural design, stress analysis and service life groups to determine whether they could sustain smaller cross-sectional area (and thus, weight). Once flight test aircraft and full scale structural models are built, physical tests commence to substantiate the computer modeled data. This data is then used to adjust limit and ultimate loads and then optimize the weight of the aircraft. In addition, fiskerwad has a point that customers often add requirements and capabilities to the aircraft while it is still being designed. In all, I think this is one of those things that has always happened, but with increased scrutiny by customers, media outlets and the general public, they are brought to the mainstream light. Similar to presidential affairs and the like.
"Lethal Survivable Supportable Affordable"


Elite 1K
Elite 1K
 
Posts: 1289
Joined: 07 Oct 2007, 18:52

by Scorpion82 » 12 Nov 2009, 15:51

It should also be taken into account that this designing software is not perfect and not everything can be accurately predicted. There are certain error margins and it could simply happen that a design doesn't meet the requirements in this or that area. Fixing this can quickly add unwanted weight. I think the rest has basically been addressed in general.


Forum Veteran
Forum Veteran
 
Posts: 763
Joined: 31 May 2007, 21:46

by Tinito_16 » 12 Nov 2009, 20:38

Remember that this isn't some simple mechanical device - it's an combat aircraft! There are thousands of individual parts that need to be integrated in the design, and one change in one end of the a/c could have effects on another part(s), which you as a designer will have to modify as well. I've used AUTOCAD where I study, and I assume LM has something a bagillion times better, but still, I see no way that CAD can resolve this type of problem completely - it's up to the engineers working on the project. And, in a program which has taken over a decade to design, test, and start building production airframes, new capabilities had to be added to maintain the a/c's advantage. It is very difficult to add capability while NOT adding weight.
"Like the coldest winter chill, heaven beside you...hell within" Alice In Chains


Forum Veteran
Forum Veteran
 
Posts: 579
Joined: 12 Aug 2007, 07:43

by LMAggie » 13 Nov 2009, 19:37

falcon mr.b hit it on the mark. You basically design to some guesses, which can lead to either being too strong or too weak. Being too strong means you're overweight, and it affects mission performance, but sometimes its negligble. Being too weak, means you have to add last minute patches to the airplane. Once an aircraft's detailed design is almost complete, then they re-do all of the structural analysis based on reality, not the original guess. Then you determine whether your margin of safety (strength margin) is positive or negative. Positive means you overdesigned, which means you could go back and cut weight. Negative means you have to make a last minute repair to restore strength (re: 787). This is no different than how it was done in the 70's back in F-16. However aircraft these days are stocked full of bells and whistles. Current Block 60 F-16s also have weight problems for that reason, and they'll never be able to out-turn a F-16A, but added engine performance helps them do the job. Furthermore, there are always trade-offs. The method for mating the major sections (fwd, center, wing, aft) was changed after AA:1 because it was heavy. The trade off was increased manufacturing span time for mate. Because of the dynamics of aircraft design, its often hard to know exactly how it will pan-out until the design is complete.
“Its not the critic who counts..The credit belongs to the man who does actually strive to do the deeds..”


Forum Veteran
Forum Veteran
 
Posts: 578
Joined: 23 Nov 2003, 01:51

by F16VIPER » 13 Nov 2009, 22:02

The issue is now becoming clearer to me. Is there a person in the team keeping an eye on the weight of the plane and how accurately can this person/persons do this at various stages of the production of a prototype or pre-production plane.
It seems to me that the team is fully aware that a plane is becoming overweight because I imagine the person responsible for the weight would inform the team leader the plane is fat. It seems to me that fighter planes are always putting on weight, never getting lighter!

CAD: I had the impression that a plane fully modeled on CAD would have a weight associated to each individual part, that way being able to retrieve the weight of the plane any time.

Ignoring for the time being requirement creep, it seems to me that "gut feeling" still plays a part when designing structural parts of a plane, and it is not an exact science, and the weight of those components cannot be fully established down to the last gram using computer software.

Do you think that the weight creep that has been reported in relation to the most recent airplanes is generally normal or there is a degree of incompetence and lack of capability.
Last edited by F16VIPER on 13 Nov 2009, 23:25, edited 1 time in total.


User avatar
Elite 5K
Elite 5K
 
Posts: 28404
Joined: 05 May 2009, 21:31
Location: Australia

by spazsinbad » 13 Nov 2009, 23:06

F16VIPER, IMHO you are forgetting that 'human error' is in everything and it can only be minimised. Computer software is made by humans, with the GIGO concept ('Garbage' In - 'Garbage' Out) in use. Here I mean that a 'small error input' produces a 'small error output' at the very minimum. So I'm not disrespecting the process involved in producing the JSF. Perhaps one day when SKYNET rools we will see perfect machines designed by perfect robot designers? :mrgreen:


Enthusiast
Enthusiast
 
Posts: 97
Joined: 23 Apr 2009, 07:44

by gf0012-aust » 14 Nov 2009, 00:17

F16VIPER wrote:Do you think that the weight creep that has been reported in relation to the most recent airplanes is generally normal or there is a degree of incompetence and lack of capability.


all projects - and I literally mean all projects will suffer a degree of scope creep - no matter how tight the program is declared to be managed.

JSF is a number of ways an iterative and evolutionary development - no other major capital development in US military prucrement history has undergone such radical processes - the equiv on a smaller scale would be the Virginias.

as an example, although I'm unable to give empirical detail for a number of reasons, we've had a briefing which points out that in real development terms, JSF is up to 6 times more complex and transitional than the Apollo program. Again I can't quote numbers because it was part of an overall cleared session, but I'm not trying to blow smoke up anyones r-se here.

To give practical parallels - again all major US military acquisitions such as subs, combat skimmers, even the USNS assets have all grown "fat" beyond the original design brief. In modern design, it doesn't mean that it's a fixed design, we will change things if there are new technical capabilities and competencies found relevant - and any change may well only be considered on a discrete area of interest rather than on he whole platform - thats because there are other review points where overall system issues will be looked at and revised.

in the case of JSF (and subs, which are kissing cousins to fighters in the engineering dynamic sense) there have been material sciences developments where prev materials have been replaced for a number of reasons. eg weight and thermal management, signature emission etc issues.

A platform getting "fatter" is not unique - the reason why this is getting so much attention is that the general public have been completely oblivious of such things in the past, some sections of the media sense an opportunity to run with such things (but it will fall off the attention radar as soon as a celebrity crashes their car, or Sarkozys wife wears a new dress etc... - and there is also a fair bit of ideological investment in some quarters.

As an other example, even RAAF Mirage III's were overweight and an incentive was given for every pound of excess weight removed. As soon as you add other capabilities, as soon as you add iterative developments of extant systems with extra functionality or resilience, then there is likely to be an increase in weight somewhere in the platform. eg One of the biggest killers is cabling. start adding extra modules, start adding extra sensors or more demanding systems and the cables will start adding significant albeit subtle changes. It is a fact of development life - irrespective of whether you use a slide rule, a computer or a program like CATIA. (also progs like CATIA don't factor in weight, they are design build and functionality tools). hence why progs like CATIA will show that an access point won't open up properly, will provide an opportunity to redesign the angle of entry or access point pivot/entry points - but won't provide anyone with information that says redesigning the access point will add 100kilos of extra metal to not only brace the point, but to brace the frame etc... Aircraft are no different in that respect. Its why billets are far more aggressively used in JSF than any other aircraft as using engineered shapes will add more weight, welding points will ad more weight etc etc.....

There's also the fact that like anything, some of the supposed technical counter claims are just errant nonsense.

Its not about getting "fat" - its about capability and design intention impact issues. Its about management. Also, all aircraft programs have a degree of flex built in for weight - there are safety percentiles factored in.
Last edited by gf0012-aust on 14 Nov 2009, 00:45, edited 2 times in total.


User avatar
Elite 5K
Elite 5K
 
Posts: 28404
Joined: 05 May 2009, 21:31
Location: Australia

by spazsinbad » 14 Nov 2009, 00:30

Another point to think about (as an example from the distant past in the days of large scale meticulous hand drawings on many, many sheets of paper with slide rules for calculation) was that the original design of the A-4 Skyhawk was excellent in producing a lightweight aircraft, where weight reduction was a prime design objective. However later models or later iterations of early models had to put weight on to overcome some initial design problems due to this 'prime' objective "reduced da wait". :-) There is a balance in all the factors required as others on this thread have mentioned.

Probably no other Western aircraft has received such attention as the JSF (in this interbabble age) - probably a good thing - there are some silly anti-JSF claims with vice versa. That is why this forum exists. We sort the chaff from the wheat! :-)


Active Member
Active Member
 
Posts: 205
Joined: 07 Apr 2008, 16:52

by gtg947h » 14 Nov 2009, 02:05

gf0012-aust wrote:(also progs like CATIA don't factor in weight, they are design build and functionality tools). hence why progs like CATIA will show that an access point won't open up properly, will provide an opportunity to redesign the angle of entry or access point pivot/entry points - but won't provide anyone with information that says redesigning the access point will add 100kilos of extra metal to not only brace the point, but to brace the frame etc...


Kind of. You can program in a density for all the different parts and it will give you a pretty good weight estimate.


Active Member
Active Member
 
Posts: 231
Joined: 27 Jul 2004, 20:09

by psychmike » 14 Nov 2009, 02:09

You take 1,000 bottles, light bulbs, nails or anything else and they won't all fail at the same point. There will be a statistical mean with confidence intervals. For example, saying that a simple part may be expected to fail at 100 lbs isn't as useful as saying that 99% of the parts are expected to fail between 95 to 105 lbs. Multiply those confidence intervals by the thousands of parts that go into a complex machine and you get some idea of the tasks that aeronautical engineers face.

With well known, everyday technology, a vast sum of incremental knowledge has built up and tolerances aren't so critical. But with cutting edge military technology, there may be a lot of uncertainty with small tolerances with large consequences for failure. New materials are being used in new ways and the high required performance means you can't just 'guesstimate' on the safe side and make everything twice as strong.

I'm sure bureaucracy has a lot to do with it, but there are reasons a toilet for NASA might cost $250,000.

Mike


Enthusiast
Enthusiast
 
Posts: 97
Joined: 23 Apr 2009, 07:44

by gf0012-aust » 14 Nov 2009, 02:20

gtg947h wrote:Kind of. You can program in a density for all the different parts and it will give you a pretty good weight estimate.


you can get a really rough guestimate - but what you do in a discrete planning area may not be reflected to the end solution. eg out to the later plan or schedule. you don't actually want guestimates in aircraft as any weight changes or distributed changes will impact on flight management and potential performance issues. sooner or later the actual weights have to be used.

eg if you are redesigning the landing gear mechanisms then you can get rough estimates of what the overall traditional mech might weigh, (if anyone has bothered to gather that data which could involve maybe 200 odd separate items) or you might then cut it in a billet to reduce weight, build in webbed strength without having to add external struts etc and get a guide - that will not however get added into the overall map of the plaform.

at some point, when you add in the discrete elements they become one and then you can start lifting the eyebrows and start working out what needs to go - but its not a simple task.


in fact the use of billet for wing loading issues is a good example for JSF. prior to that the only aircraft that were using billet like components were aircraft such as the Boeing 767 series etc... it wasn't a common design element in combat aircraft at all. In the case of JSF it was deliberately used as a design construct to save weight and add strength.

the development issue is dynamic - at various points of review they will always find parts that are "fat" and need changing.

I really regard the view taken by the "professional complainers" as disingenuine (that assumes that they have some basic comprehension of how complex this process is). Of course, if they have no idea and become hysterical then that points to technical and intellectual indolence on their part. ie they should at least try to do their homework before turning on the megaphones. :)


Next

Who is online
Users browsing this forum: No registered users and 14 guests