April 02, 2015; There is different information whether Boeing will adopt new Aluminum Lithium based alloys for the fuselage of the up and coming 777X, our latest info is that this is not yet decided.
In this corner we will look at what is at stake should Boeing go the route of new materials or not. It will discuss what is to consider when choosing materials like Aluminium Lithium (AlLi) or Carbon Fiber Reinforced Plastic (CFRP) for aircraft structures, especially for fuselage structures which has many conflicting requirements.
We will see that aircraft fuselages are tricky to make light and that a stronger material not necessarily means weight gains.
To understand what gains we can get with e.g. Aluminum Lithium we need to first understand which parts of the aircraft we can change, then how large part of the aircraft’s total mass these represent and finally what gains we can achieve for the parts.
Which parts can we change
The material change we discuss is for the fuselage structure and constitutes a rather small part of an aircraft’s total weight. Lets try and understand how much that is:
Aluminum Lithium is considerably more expensive than normal Aluminium alloys and more difficult to machine, therefore it will come in qustion for parts which has a high fly/buy ratio i.e. little spill when formed to fit the aircraft. Machined parts like fuselage frames, cabin floor beams and the many fastening and fitting parts that are needed in a fuselage to attached interior items etc will therefore not be candidates for change to AlLi. The most likely candidate is the fuselage skin, it is relatively easy to form and has a high buy/fly ratio. The fuselage skin is typically half the mass of the complete fuselage structure so now we are at 18.5 tonnes which can be can be changed for a stronger or lighter material.
AlLi is about 5% lighter than normal Al alloys for the same dimensions, Lithium is a light metal and it lowers the alloys density. Rightly mixed with other alloying metals it can create alloys that has other criteria that are better than conventional Aluminium alloys. There are variants that have higher tensile strength, others that have higher toughness (capability to absorb energy while deforming but not rupturing) and still other that have good fatige characteristics like slow crack growth. Often these characteristics come at the expense of some other, therefore it is important to understand what are the key requirements for the part we want to replace.
The main requirements for the load carrying part of a fuselage structure are:
– High tensile strength for the tension loads that pull up maneuvers or gusts provoke on the upper fuselage. On the lower fuselage this result in compression loads, there we need a structure that has good compressive strengths and one that avoids risk of buckling.
– A major considerations is the hoop stresses that gets incurred in the fuselage skin by the cabins 6,000 to 8000 ft cabin altitude when the outside air pressure is much lower at cruise altitudes of 30-40,000 ft. This cyclical stress is repeated once for every sector we fly and it forces major fatigue considerations on the fuselage structure as we expect long range aircraft like 777X to withstand 20,000 flight cycles and short range aircraft to take over 60,000 pressure cycles without rupturing.
– A third major consideration is tolerance to damaging objects. Figure 1 is from a DLR paper on fuselage dimensioning and it shows the many cases of objects hitting the skin that has to be considered.
– Finally we can not make the skin arbitrarily thin, we need skin dept to e.g. countersink flush rivets. Practical minimum skin dimensions are 0.06 in or 1.6 mm.
The next picture, Figure 2, shows what happens for a medium sized aircraft when we apply these criteria. The graph is from a Doctor thesis by Jörg Fuchte which is worth a read. It deals with single or dual aisle cabin design optimization and its consequences on fuselage mass. The graph is for a 56 m aircraft fuselage i.e. the size of 787-8 or A330-200 and therefore shorter than our 777X but it shows in a very illustrative way how the different criteria we listed sets the skin depth for different parts of the fuselage.
The fuselage skin has to be sized by the highest of the different dotted lines in the graph. We can see that at the front and rear of the aircraft this is the minimum thickness criteria, then the cabin pressure differential is the sizing criteria until we get close to the wing where the compression buckling criteria (light grey curve with triangles) is dominant for the lower part of the fuselage around the wing. Only 12 m of 56 meter fuselage gets dimensioned by pull up or gust bending forces needing high tensile strength for the fuselage top part (black curve with dots).
The graph shows that for the most part we need a skin material with good fatigue characteristics more than high tensile strength. This can be achieved with special tempers (heat and mechanical treatments of AlLi material) but it might not lead to thinner dimensions, the designers might want the longer cycle life instead.
So despite AlLi having higher strength characteristics it will be very much a trade which of its characteristics will be emphasized and what that will mean for reducing material thicknesses and therefore weight gain beyond the 5% which is gained by its lower density.
Lets now summarize our likely gain. We had 18.5 tonnes of material that could at least be 5% lighter. That is 0.825 tonnes or 0.5% of the empty weight. Should the material allow reduced dimensions for the fuselage skin that might increase to up to say the double, giving a gain of 1.85 tonnes or 1%.
The change of material for the fuselage of 777X to higher strength AlLi gives the un-informed the idea that it can mean major weight gains for an aircraft like 777-8X or -9X. It does not take a complicated reasoning chain to understand that it can bring weight advantages, but they are not in the order of what people think when they hear that a material is up to 20% stronger than conventional alloys.
Aircraft structures and especially fuselages are sized by a complex set of criteria, a materials strength being just one of many. In the end the large weight gains that were forecasted for CFRP fuselages did not materialize, very much for the reasons shown and add to these needed lightning and electrical return paths in the CFRP via in-layed metal meshes.
The above trades will have to be done for aircraft structures that changes conventional Aluminum alloys for better but more expensive Aluminium Lithium ones, to the designer he gets better characteristics at the risk of hidden issues with the new materials, something that Airbus learned the expensive way with new materials in A380 wings.
In our example with the 777-9X we could realize a weight gain of 1.85 tonnes if the net gain of our AlLi alloys were 10%, that is an empty weight improvement of 1%, a sobering result for the optimist but a gain that can be worth having for the designer if the cost-benefit trade close.
Hey, Good Morning,
just did a ballpark check for skin thickness going from those 18.5t ( L~=75m, D=6.20m,AlLi@2.58g/cm³) and get 12.66kg/m² or 4.9mm Skin Thickness. (my error?) significantly more beefy than those 2.4mm that the diagram would request for ( most of the fuselage,) A380’s Glare skin elements are said to be “flimsy” 😉
remember the diagram is for a relatively short fuselage. As you increase the length, you effectively plug the extra length in front of and behind the wing, keeping the slope of the curves constant. Thus, where in the diagram only a very short distance in front of the wing is sized by buckling, when you add length all the extra length is sized by buckling, requiring linearly increasing thickness.
On the glare, an A380 fuselage is relatively high when compared to it’s length, lowering the stresses in the extreme (upper and lower points) and thus reducing the pull up load and buckling load required thickness curves. Additionally, it’s wing is very wide, also reducing the required strength of the fuselage.
The main benefit from glare is that is is significantly lighter than base aluminum at the same thickness (by replacing some of the aluminum with glass). The graph shows most of the fuselage is thickness driven.
Additionally, the glass in glare improves fatigue performance and it is relatively easy to create local reinforcements (it’s all layers anyway)
So any gain in the material weight will be offset by the cost of the aluminum lithium and by the cost of new tooling . Is that correct
On the cost side yes.
But the performance side improves.
Boeing needs performance improvements even more than money. You can easily hide too much cost but you cannot hide performance shortfalls.
The same applies to Airbus.
Quite obvious if you wear funny glasses 😉
Quite obvious, that you have to work a little “Boeing bash” into every comment you make.
I don’t feel it is Boeing bashing but more an acceptance that going up against that
the a351 the x requires all the help it can get.
Similarities with the new a330 8/9 against the b787. Boeing have already ceded narrow seats and noisier planes all for economics. If that don’t work they will be in trouble.
@Bjorn or ikkeman,
Where does Airbus stand today with Glare? Do they still want to pursue this technology or have they abandoned it for good? As far as I know Glare has not been used on the A350XWB, but I am not sure. My understanding is that it is too expensive for the benefits it brings.
Is Glare still the mainline material, or are there other, more preferred, Al-Li constructions on the market nowadays?
Glare is new on the G650, and airbus keeps evaluating it for new and/or upgrades – but nothing definite AFAIK
leading edges for impact/tear resistance and high fatigue area’s are the primary focus area’s
“too expensive” is subjective and relative. In the current low fuel price/low interest environment better materials are not worth as much as they would otherwise be.
For just 1% over aluminium in my opnion it doesn’t worth costs are higher than the benifits.
Why take out only 5%? Since the material is 20% lighter, shouldn’t they use less of it, maybe 10% less which will still leave them with roughly 10% extra strength.
because much of the structure is not sized by loads, but by stability and/or assembly tolerances
Lighter density is the most efficient way of losing weight, if the material is used in exactly the same way, its 10% lighter is 10% weight saving.
While higher strength is a lower reduction in weight, 10% would give say 3% weight reduction.
The earliest way of reducing weight was to cut holes in structural members, which still happens.
Very interesting article, thanks Bjorn.Makes you wonder if Airbus isn’t regretting not proceeding with A350 mk 1.
In the end the large weight gains that were forecasted for CFRP fuselages did not materialize…
The author keeps using the term “weight gains” instead of “weight reductions” or “weight savings,” which is confusing to the reader.
I saw that and adjusted. Confused for a few seconds but put it in context.
Not sure if its a language thing or an engineer reference.
Well written and makes clear the conundrum in trade off.
supposedly one adder for a CFRP fuselage is the reduced maint, have to see how that works as all the need to insulate from other materials (or use titanium) to avoid a corrosion issue. Applies to 787 and A350
Of some note is that the A350 has an aluminum nose (I suspect its due to the issue with forming the composites in a pressure hull and the shaping). What that does to maint weight a supposedly less maint needed fuselage but the aluminum still needs to be checked on a normal aluminum schedule (or so it would seem)
Like the A380 wing, its still an open question in my mind until they get the experience with it all to find out if it does as well as they hoped or not as good.
Airplanes being funny that sometimes they exceed and sometimes they disappoints.
Re the A350 nose section, it has a major bird strike requirement, see the DLR impact figure (no 1). CFRP is not that great in toughness, i.e. plastically deforming while absorbing energy, it shatters and then the bird is through in a normal thin skin structure. This is the reason Airbus went for Al cockpit section, Boeing kept CFRP but adapted the design, presumably with a tougher CFRP which normally increases the resin content.
This discussion reminds me that Bombardier has chosen Al-Li for the CSeries fuselage. I assume this is only for the skin.
Are there other commercial aircraft with Al-Li fuselage?
This technological choice must translate into added manufacturing cost for the CSeries.
Has Bombardier quantified the specific weight savings associated with the use of Al-Li for the fuselage?
That is a good analysis showing why one should always look deeper instead of following the superficial hype. Thanks. Now that you have done this, what about the hypothetical? Suppose one chooses CFRP barrels for the 777X. Its strength to density ratio is much higher than Al-Li. So superficially, one would think mass savings would approach something like say 10% of the empty mass. Even ignoring the difficulty of fabrication and repair (B787 provided great examples for each of these), and cost of the autoclaves etc., and keeping in mind the need to introduce conductive paths for lightning strikes, etc. etc. what do you think that percentage is? 2%, 3%? Must be higher than Al-Li but Boeing chose not to do CFRP for the fuselage. Why? Not big enough mass savings that is worth the extra hassle? Was B787 CFRP barrel fuselage the wrong way to go in hindsight?
You have already studied Al-Li. Why not let Leeham leaders get a glimpse into CFRP for 777X?
I will leave it to Bjorn to asses the 787 choice of material for fuselage.
I do not know if the production costs are whats driving any of the Boeing issues in that regards (again Bjorn may have some answers)
As for repair, I think in general you are wrong. Fixing damage is not a problem.
The damage on the Ethiopian 787 was due to a fire. I don’t know if aluminum would have done that but it is not your typical damage.
For the average ramp rash its not a problem to repair.
Not as Airbus touted at one time would you remove a whole skin section (1/6 of an entire A350 more or less) to fix something like that.
You would do like they have done with composites forever (i.e. Fiberglass and concrete boats and then aircraft with their different material) you create a patch or patch kit and its as solid as original.
CFRP is a good material for many parts of an aircraft, the core fuselage being the most debatable for the reasons we showed. CFRP is so strong that it would propose thin skins for many areas, to thin to countersink fasteners in and for impact loads (CFRP is strong in the fibers direction, less so in others).
Therefore you don’t gain much in weight, especially for small fuselages where the skin thickness problem gets amplified (very few areas would be dimensioned by strength). What you do get is a bilge which is insensitive to corrosion and fatigue resistance which means your heavy check inspections cost much less. But you have to build your current return paths etc, it is a lot of work for little gain other than maintenance. In the long run that is important though.
The point with 777X is that Boeing does not have to spend $15-20 billion to get a competitive aircraft. They will therefore spare themselves the 100,000 manyears that a CFRP fuselage would cost for very little gain.
The fact that the weight of the fuselage varies so little with the different materials answers one question I had. How can manufacturers sell an airplane, presumably with performance guarantees, before they have decided how to build it? The answer from this article is that the fuselage material doesn’t make much difference.
On the other hand, as another comment stated, why did both Boeing and Airbus go to all the trouble of introducing CFRP for the fuselage. I can understand could provide big benefits for the wings – where I would assume that CFRP gives more flexibility for shape – but not for the fuselage plain tube – unless (think lift) the fuselage is no longer a plain tube (except where the wings and landing gear fit)
The other item not discusses is the ability to swap materials using the same frame structure .
In other words, do you get the maxim benefit from LiAl if the frame was not designed with it in mind?
So the cheap classic aluminum wins the battle
All the considerations have to be balance, ease of assembly, pre assembled, fasteners used etc.
Not quite. There is a lot of research being done on the types of fibers in CFRP
Then there is the resins and methods of curing.
When they do a ‘Maxx’ on the 787 in say 15 years time, there would be a lot CFRP that would be changed into newer versions and use different techniques.
Which again will not be a simple swap “C” to “C+” and will require extensive redesign to enable potential gains.
One reason current CFRP doesn’t show large improvements in a range of use cases as the designs still have a strong taste of “Black Alu”. Fastener-less/-lean assembly processes will lead to further gains.
Your article appears to be specifically tailored for the 777X. My understanding is that Boeing initially discarded Al-Li and only came back to it recently. My reaction to this latest development was that Boeing needed Al-Li to meet its efficiency targets. In other words to make the aircraft lighter in order to remain competitive with the A350XWB. I always thought that Al-Li was, everything else being equal, 8-10% lighter than Al-Li, but you mention 5%. I must admit that I am baffled by what you say. But I am not qualified to contradict you on this.
While reading this very interesting article one question came to mind: Would a high-cycle aircraft like the 737 benefit more from Al-Li than a low-cycle aircraft like the 777? The reason why I am asking this question is that I like to think that Bombardier selected Al-Li for the CSeries fuselage because it was designed to be a very high-cycle aircraft, typically used on so-called milk runs. My reasoning is that Al-Li would be more resistant to metal fatigue as generated by the repeated take-off and landings that are associated with this category of aircraft.
One last question: How does the value of Al-Li progress with fuselage diameter? In other words would a small aircraft like the 737 benefit more, or less, than say the 777.
I am saying AlLi is around 5% lighter as a material (ie has density which is around 5% lower then classical Al alloys). To that shall be added any other characteristic that might allow use of thinner dimensions, ie strength or toughness or fatigue crack insensitivity or whatever. Then you can get your 8-10% which in marketing terms then is 8-10% lighter but as construction, not material. The article makes this distinction.
Also for all who is commenting AlLi as if it was one material, it is not. It is a group of alloys of different compositions with different characteristics just like classical aeronautical Al alloys are. The major alloying element in classical Al alloys like 2024 is Cu and it is the dominant element in most AlLi alloys as well, then Lithium is added to lower the density and to modify the properties as are other elements like Zn, Mg, Mn. By varying these amounts and the heat treatment + sometimes stretching (the alloys temper) after casting the materials final characteristics can be formed.
This whole process is important, as an example 2024 without temper has 1/3 the strength of 2024 T3 (temper 3), which is just one temper variant of many for 2024. The same goes for AlLi alloys, they are sensitive to the whole production chain and the variants that were available when the 777 was designed were not mature, they had several issues like not homogeneous (isotropic) properties and/or corrosion/toughness/manufacturing issues.
Now there are mature AlLi alloys available with DIFFERENT characteristics and depending on a parts requirements a specific AlLi alloy can be a competitive choice. What they share is lower density and higher cost. Many times one finds a non AlLi alloy that is just about as good in the for the part important characteristic like fatigue crack growth resistance but at a lower installed or life-cycle cost on an aircraft level.
Thanks Bjorn for your enlightening reply. I have been interested in Al-Li for the 777 ever since I read the book “Twenty-Firts Century Jet: The Making and Marketing of the Boeing 777”, by Karl Sabbagh. In that book it is said that Alan Mulally considered Al-Li for the 777 because on an Airbus it had saved 2,200 pounds. But they were reluctant to use it on the 777 because, like you suggest in your reply, the early versions of the material had several weaknesses. In the end the 777 did use a bit of the material, which saved 400 pounds on the aircraft.
But since that time the material has been considerably improved and would be more practical and economical to use than CFRP in some applications. It would be interesting to know how many pounds Bombardier saved on the CSeries by making an Al-Li fuselage. To my knowledge this is the first aircraft to use this material on such a large scale.
But if we stay at 5% you gain 0.5 percent if you don’t reduce the thickness.On the other hand aspire aviation says that aluminum lithium weight savings are 12% and 6% reduction of skin friction
IMU most of the advertised savings can be achieved from design imrpovements available through the properties of AlLi.
( Like less fasteners per length of join, larger pieces, reduced numbers of stringers … . Not from just swapping out alloy “A” for Alloy “A+L” ).
Afaics going to AlLi is an “early” design decission and the effort would not be much different from a brand new design.
( Compare to the Soviet Tu4 as a B-29 replica that had all materials changed from US imperial dimensioned stock to all Russian metric stock. Looks the same but everything is different )
Re the savings see the answers above, re skin friction, how can that be affected, the material does not even touch the air, the paint layer does.
Advertised skin friction reductions are touched by less fasteners and less joins. It is an indirect gain and you would have to design for it.
As a 1:1 replacement probably no gains beyond the volumetric advantage. ( And going from there my guess is Boeing will not use AlLi on a wider scope. Maybe some select parts for the boasting value.)
Your minimum gauge assumptions are correct for fasteners however for a CFRP skin you can tailor the section thicknesses. You can put CS depth along the frame interfaces and ramp down elsewhere. This enables you to save a lot of weight. We do this all the time at my company. With Automated Fiber Placement its very easy to do, saves material costs and is faster to layup
sure, the famous Dreamliner barrell is varying the number of layers everywhere. Is your company working for aeronautical applications?
Not to detract from Adams comment but as a sobering info, the following shows the CFRP 787 which is reducing the amount of fasteners used, there are still quite a few and quite close in many places.
In other areas the change if material thickness shall be more succesful. BTW, A350 is no different and we can spare ourseleves the n’th debate barrel vs panel, it is the least that differentiate these aircraft.
Bjorn: Not to get into a debate but purely on the tech merits.
My understanding was that Airbus did not have the research in place to do what Boeing did with a spun fuselage and went with the frame and panel that they did have the capability to do in the time frame required.
I would think there was an advantage to Boeings approach though how much .
Also of interest would be cost differences, one takes a lot more equipment/expensive e materials and the other more labor to assume so any thoughts on how that factors.
the one aspect I never did buy was you would replace a panel for a ramp incident as opposed to a patch .
787 fire in London being an exception as that would hopefully be a one time incident.
I’ve never found any patents that would support you assertion. A380 tail cone is spun CFRP. My impression from following research publications here was that Airbus and its subsidiaries had and have the lead in FRP centric manufacturing for applications.
( No idea what is happening in the military design space. But I’d question the “crossability” from money is no problem military stuff to the highly cost effective civil requirements.)
Tentative guess and keeping in mind how such things are communicated by Boeing: one piece barrel sections could have been the only way of production not hampered by foreign patents.
What I have taken from comparisons is that panel manufacture will always with similar effort result in a better quality (controlable) product. Exposed surfaces have much better smoothness and the final fuselage segment assembled from panels will show much tighter tolerances at the interfaces. ( For the 787 Fitting sections together still is a problem. Strangely a similar issue existed for the 777 until a year or two ago when they fixed tolerances in manufacture.)
Autoclave volume can be used much more efficiently.
With the 777x being an upgrade to an existing base model we will have to wait much longer to see if Boeing will do another new design with barrel sections.
After experimenting with AlLi in the nineties on 330/340 prototypes, applying it on A380 wing structure and A350 nose section, many expected extensive use on the A330-800/-900NEO. Can’t see how much will be done.
Remember when GM bought out the Saturn line of cars, with ‘plastic’ panels.
An advantage for GM was lower cost of tooling and easier changes for styling purposes.
The steel industry upped its game and helped improve new ways of metal forming and lower cost dies.
The plastic Saturn died a bit before Saturn itself disappeared, so I guess steel won that one. But now its Al, so the game continues
That seems to have been a rather low tech application of plastics. ( Though fitting for the basic ancient low tech platform it was used on.)
Remember all the “5mph collision, no damage” bumpers that followed from some of Ralph Nader’s consumer protection drives?
The roof panel for the Citroen ID and DS models was some FRP material. Same for the engine and boot hoods on the BX line. Potentially some other Citroen models too.
One of the earliest uses must have been Loyd’s line of small cars nicknamed “Leukoplastbomber”.
Followed by the GDR’s Trabant ( initially Loyd derived so no surprise. )
You see similar ups and downs for wundermaterials all over the place. Manufacturing technology changes.
First you had heavy cast iron engine blocks, then lighter alloy blocks later superseeded by much thinner walled cast iron blocks that were rather competitive to the lighter alloys. A similar dance can be observed in body manufacturing technologies. advanced sheet steels, AL Alloys, FRP are still in a performace race.
CFRP uber alles was but a marketing ploy providing nothing much on a long term horizon. Every new technology will after a rage phase of hit and miss settle some (large/small/microscopic) niche or two and continue competing with other tech elsewhere.
Doing large sections in CFRP reduces the Numbers of fasterners. The same for milling skins with integral stringers. Hence the Numbers of rivets per aircraft is coming down and with it the mass added for those joints. GE switched To CFRP containment leaving the alu honeycomb and kevlar design of the CF6 and gained alot in reduces mass. I Think going to Alu-litium allors even in machining intensive parts Will be common. Moving To stickning structures with resin transfer injections instead of prepreg tapes Will be more common now When it is production for the Leap -1A fan blades. That makes for more isentropic composite structures that are more impact resistent.
GE switched To CFRP containment leaving the alu honeycomb and kevlar design of the CF6 and gained alot in reduces mass.
That is IMU largely a reduction enabled by a reduction in certification requirements: CFRP blades out containment is done on half the blade mass as those blades are “savely assumable” coming out in a fragmented state while the all metal ones have to be contained when breaking off at the blade root.
Works for the GTF too. The blades are moving slower so contain less energy. When they become crfp this will require even less containment material.
“The same for milling skins with integral stringers.”
Todays solution would be laser beam welded stringers ( as seen on forex the A318 ), the equivalent to what “cocured” represents for FRP stuff.
“To start with the aircraft’s empty weight is typically around half of its Max Takeoff Weight or MTOW. For 777-9X we are therefore talking half of 351 tonnes i.e 176 tonnes. This would fit for the existing 777-300ER but 777X uses a longer fuselage; a wider wing (though CFRP) and new high performance engines on the same MTOW as the 777-300ER. The result is a considerable weight gain and we have to count with around 185 tonnes empty weight for 777-9X.”
Not if payload is lowered e.g. 10 tonnes. The empty weight of the 777-9 is expected to be considerably higher the 777-300ER.
I think the “*airbus estimate” tells you everything you need to know about that slide….
Boeing told us the new wing won’t be lighter.
Bjorn filled in fuselage stretch weight opportunities and the engines will be larger.
Who believes in miracles?
No doubt Boeing tells us this 2.7 m stretch creates 60 extra seats and presents efficiency gains per seat. (EK already has 10 abreast, go figure..)
415klb/188.5t is what Aspire Aviation wrote in mid 2014.
Airbus estimate @190t is from somewhen in 2013. ( .7% difference between the two. Your reservations probably lack the flesh on the bone 😉 Boeing might surprise us still, but … .
To put things in perspective,
– There is a good chance that during the 2015 Paris Airshow, the 1992 all Aluminum A330 will outsell the spanking new all Carbon 787 Game Changer.
Only armchair idiots with no aircraft knowledge whatsoever where predicting that back in 2004.
& you heard it first here on Leeham.
Hi Björn, I was just about to write a lengthy answer/comment, but bottom line is: You drew the right conclusions from a not-so-right explanation 😉
Any word on 777-9 final design for fuselage? Also, how is Boeing achieving a higher cabin pressure and larger window without compromising weight?