Presurization - Whats it take ?

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
15,821
Location
Memphis, TN
I was working on a pressurization problem I I kept saying it’s the cockpit adjuster. They wanted me to change the outflow valve. I went to the outflow valve with a 9-16 wrench and showed them. Crack the B nut and pressure goes up. Close it goes down. Valve is working. Of course they did not have the Cockpit valve but they did have the outflow valve. Then you get to hear the management whine of buying something that’s not supposed to break. My ears hate ground pressure runs. A couple of psi and I’m done. No way I could have stood it in that KC

When taking NDT eddy current class I got to use the test probe they designed to check those 737s after the Hawaiian air convertible accident. The guy teaching was one of the group to figure a way to test for the corrosion because it was in a doubler under the primary skin.
 

rbarnes

Well-Known Member
Joined
Aug 28, 2015
Messages
405
Location
Texas
Some light reading on pressurized aircraft systems and design... NOT a trivial subject...

Cessna Training ~ Single Engine/Light Twin Pressurization ~ Student Workbook

Aircraft Pressurization - An Environmental Control System, AEM 617, Andrew Treadway

Control of Cabin Pressure - Aircraft Pressurization Systems

AC 25-20 PRESSURIZATION, VENTILATION AND OXYGEN SYSTEMS ASSESSMENT FOR SUBSONIC FLIGHT INCLUDING HIGH ALTITUDE OPERATION

What happens due to gross-over-pressurization during a ground-run test... just after Depot maintenance, prior to delivery...1999... the energy of pressurized/contained-air [gas] has-to-be-seen to be believed...
Those links are awesome.

That Cessna student manual was exactly what I was trying to find. Thanks !!!
System layouts
component diagrams
Packaging
Door and window frame seal examples
Control cable seal example
 

Matt G.

Well-Known Member
Joined
Nov 16, 2011
Messages
1,266
Location
Kansas, USA
Designing a pressurized fuselage is feasible. However my suggestion is: if the designer has little or no experience evaluating fatigue and crack propagation (fault tolerant pressure vessel), move away from aluminum. Aluminum is light, it can easily be bent, folded, stretched and riveted but it is a nightmare when it comes to crack propagation. Go for a composite vessel, pressure test it through may be some hundred cycles and you are good to go. Main manufacturers are doing this: 787 (carbon), A380 (glare), A350 (carbon), etc. but remember that pressurization stresses happen simultaneously with gust, maneuver stresses. Another suggestion: programmable outflow valves are immensely complex. Use simpler, redundant mechanical fixed-differential relief valves instead. You will also need a negative-pressure relief.
This doesn't really make the analysis any easier, it just shifts the difficulty to another skill set. A well-designed composite structure will generally be good for fatigue if it is good for ultimate, but now, the ultimate analysis is as complex as the fatigue analysis of a metallic fuselage would have been. Depending on how it is designed, the capability of some details will be difficult to predict without extensive use of finite element models, sub-component and full-scale tests, or combination of both. Composites have some different failure modes that often require point-design allowables with the materials, layup schedules, processes, etc. that are specific to the final product. With aluminum, we have nearly a century worth of analysis methods that have been validated by or derived from testing.

I have years of experience analyzing structure of pressurized fuselages. I cannot imagine someone with no structural analysis experience successfully designing and analyzing a durable, minimum-weight pressurized fuselage, either metallic or composite. A lot of the finer points of design and analysis are not going to be found in a book or on the internet. If mission requirements for my own design dictate flight at higher altitudes, I'll be installing an oxygen system

As an aside, the 787 (and likely A350 as well) still have a lot of metallic structure in certain areas of the fuselage.
 

Mad MAC

Well-Known Member
Joined
Dec 9, 2004
Messages
809
Location
Hamilton New Zealand
I can't say i have designed a pressurized fuse but i have designed repairs putting them back together after the corrosion and cracks have been removed. I have also designed / certificated clean sheet plus other peoples messes of oxygen systems. I would be very careful assuming that an O2 system is low risk, they tend to be very underrated risk wise, both at design level and in-service maintenance level.

Looking at TC'd pressurized fuses, single fastener rows, steel window doublers, and piano hinge for door hinges are all features that apparently are good for 35000 cycles*. The mentioned features won't make 35000 cycles and look pretty but they will make it.

A hombuilt isn't going to need more than 10000 cycles and can probably get away with 5000 cycles**, at which point good detail design (edge margins, corrosion prevention, nice load transfer, etc) for normal riveted structure is going to get you most of the way there.

*Note these features are on the SA227, which is certified to a max altitude of less than 25000 ft, which the break point in FAR23 (old version) for cabin structural design, above which a more robust cabin structure is required. SA227 was originally certified under a SFAR, then later production was certified under commuter Cat, FAR23.

** just to put that in perspective 5000 cycles is about the heavy maintenance check interval for a part 25 airliner.
 

Matt G.

Well-Known Member
Joined
Nov 16, 2011
Messages
1,266
Location
Kansas, USA
Is periodic inspection adequate to prevent catastrophic failure?
If the inspection interval and methods are sufficient to detect the cracks prior to them reaching critical length, yes.

A 'fail-safe' design would probably be a good idea...design the structure with redundant critical load paths, and ensure limit load capability with a failed structural member. Damage tolerance analysis (basically substantiation of my first statement in this post) would be required for a certified design.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
15,821
Location
Memphis, TN
If you have a pretty high differential, you might need an ndt program. Most ndt I have seen were sheet metal openings. Eddy current around passenger windows on J31s and ex ray on the cargo door opening on the SAAB 340 I remember. The SAAB passenger windows would micro crack in the flange area. That was one check with a flashlight at an angle. The parts that were billet like the windshield frames seem to be pretty care free. Think of the billet chunk on a 747 windshield frame. About the size of a pickup truck bed. My regional had a Citation for years and on one trip the pilot thought he heard the windshield pop. They did not find anything but they looked good at it.
 

ScaleBirdsScott

Well-Known Member
Joined
Feb 10, 2015
Messages
1,358
Location
Uncasville, CT
Anyone ever tried doing the SpaceX welded (stainless) steel pressure tank method?

I mean, it works for submarines, air tanks, etc where lots of high pressure is involved. And it seems to work for cryogenic pressure vessels fighting the rocket formula... so I can't begin to do the math on it but seems like there could be a there there. It seems like it should land right in the middle.

Of course for individuals, weld analysis might end up a real sticking point, but I gotta think the welding itself would be relatively trivial for someone with experience welding tanks and there's plenty of such individuals. If they had even some basic tests it seems like they could verify weld integrity, and being steel seems like pressure tests would be fairly low impact.
 

Twodeaddogs

Well-Known Member
HBA Supporter
Joined
Jan 18, 2009
Messages
1,200
Location
Dunlavin, County Wicklow,Ireland
If you have a pretty high differential, you might need an ndt program. Most ndt I have seen were sheet metal openings. Eddy current around passenger windows on J31s and ex ray on the cargo door opening on the SAAB 340 I remember. The SAAB passenger windows would micro crack in the flange area. That was one check with a flashlight at an angle. The parts that were billet like the windshield frames seem to be pretty care free. Think of the billet chunk on a 747 windshield frame. About the size of a pickup truck bed. My regional had a Citation for years and on one trip the pilot thought he heard the windshield pop. They did not find anything but they looked good at it.
I've done that torch-at-an-angle inspection. Works well...on a side note, I saw a Bombardier executive jet having both windscreens changed after one cracked (completely starred one layer) at high altitude. They said the stress on one good windscreen during the event was enough to justify changing it,as it was regarded as the equivalent of being a structural member.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
15,821
Location
Memphis, TN
Our pilots would not fly with a crack. They hated delamination on the edges because they were stuck with that. The CRJs weren’t as bad as the SAAB. Probably because of age more than quality. CRJs were new when I was there. Some spy movie came out and the spy takes down a plane by shooting through the windshield. Pilots were worried and management took one of the cracked windshields and took it to the gun range and shot it up close. At least ten times. It shattered the glass layers but barely went into the first layer of plexi or poly.
 

paraplane

Well-Known Member
Joined
Jul 22, 2014
Messages
62
Location
California
In the brainstorming arena, could you create the second barrier similar to a balloon? Cheap party balloons start at about 14 psi. So could you create a strong enough/light enough membrane with lightweight bands (kevlar or carbon fiber) to keep the shape of the membrane regardless of pressure without transferring the pressure loads to the outer shell?
Yep that's what I was considering. Would weigh almost nothing. Look at Sailboat Laminated Sailcloth.
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
164
Location
Campinas, SP, Brazil
Matt,
As you have experience in analyzing airframes, you certainly know that static strength analysis methods produce more reliable results than fatigue analysis. That is why when a fatigue analysis is done, a scatter factor - typically somewhere between three and four - is used. The reason is : a crack can begin due a multitude of things: due to improper riveting, to scratch/dent or due to many factors, difficult to predict. You also know that a crack propagates very differently in a composite skin, compared to an aluminum skin. And you also know that as far as certification goes, analysis is not enough: full scale static and fatigue tests are required. As for the Aloha incident, one inportant factor in the fuselage blowoff was - mind you - epoxy resin! The 737 had a longitudinal overlapped joint with two rows of rivets. This overlapped joint was reinforced with epoxy adhesive. And so the aircraft was static and fatigue tested. At that time - middle 60's - not much was known about changes in expoxy over time. As time passed, the adhesive became brittle and sort of desintegrated, transferring all the load to the rivets. The Aloha aircraft was an extreme case of high number of cycles combined with corrosion. Multiple rivet hole cracks developed along this joint until the blow-off. As you know, transversal pressurization stresses (hoop stresses) are roughly twice the longitudinal ones in a cylindrical fuselage.
 

Matt G.

Well-Known Member
Joined
Nov 16, 2011
Messages
1,266
Location
Kansas, USA
Matt,
As you have experience in analyzing airframes, you certainly know that static strength analysis methods produce more reliable results than fatigue analysis. That is why when a fatigue analysis is done, a scatter factor - typically somewhere between three and four - is used. The reason is : a crack can begin due a multitude of things: due to improper riveting, to scratch/dent or due to many factors, difficult to predict.
Correct. There are typically layers of conservatism built into the fatigue analysis- assume lifetime for a given structural configuration is in the bottom few percent of all of the data points for the test- or fleet-based derivation of that lifetime, then the entire airplane is analyzed for two lifetimes and tested for three (as an example). In addition to this, if improper riveting, scratches, dents, or anything else not conforming to the drawing has occurred, fatigue analysis will be performed on the repair or rework of those items.

You also know that a crack propagates very differently in a composite skin, compared to an aluminum skin. And you also know that as far as certification goes, analysis is not enough: full scale static and fatigue tests are required. As for the Aloha incident, one inportant factor in the fuselage blowoff was - mind you - epoxy resin! The 737 had a longitudinal overlapped joint with two rows of rivets. This overlapped joint was reinforced with epoxy adhesive. And so the aircraft was static and fatigue tested. At that time - middle 60's - not much was known about changes in expoxy over time. As time passed, the adhesive became brittle and sort of desintegrated, transferring all the load to the rivets. The Aloha aircraft was an extreme case of high number of cycles combined with corrosion. Multiple rivet hole cracks developed along this joint until the blow-off. As you know, transversal pressurization stresses (hoop stresses) are roughly twice the longitudinal ones in a cylindrical fuselage.
Aloha 243 had also been operated for nearly twice the number of cycles it had been designed for. Not to mention that at that time period, it was common to knife-edge the countersinks on the holes for the flush rivets, and that was also a major contributor to this incident. Today, fay sealant or bonding between aluminum parts is discounted in static analysis and assumed detrimental to fatigue life in the fatigue analysis.

Now, back to the context of a pressurized homebuilt airplane. The metallic fatigue analysis has the opportunity to be a bit less complicated, as the design lifetime and basic hoop stress is going to be a lot lower than an airliner. The nominal operating hoop stress (in areas without cutouts, bulkeads, point loads, etc) in an airliner fuselage skin is ~14 ksi and something like 70,000 cycles for a short-haul aircraft like a 737 or A320. For a homebuilt GA airplane, using the 5 psi differential someone stated in an earlier post, 30" fuselage radius (excessively large guess) and 0.032" skins, the nominal hoop stress is about 5 ksi, and a realistic design lifetime would be about 5000 cycles. Consequently, allowable stresses will be much higher than for the airliner. For someone with the necessary experience, it would not be terribly difficult to design. The methods and allowables are available and fairly well understood in this day in age.

As I stated previously, an almost entirely different skill set is required for composite analysis, and extensive use of finite element models would probably required. Depending on selection of resin, matrix, and manufacturing techniques, there may be no available allowables , so one would first have to do a bunch of allowables testing prior to even designing the structure. Fatigue analysis of composites is generally not done if the limit (or even ultimate) strains do not exceed the no-growth strain allowable. Now of course, metallic parts in the design may require fatigue analysis, and this has the potential to be more complex due to having to account for thermal stresses, as the metallic parts will have different thermal expansion rates than composites.

My opinion remains that neither of these two are going to be "easy" or appropriate for an individual with no aircraft design and analysis experience. In my experience, both are of similar difficulty for different reasons.

In the brainstorming arena, could you create the second barrier similar to a balloon? Cheap party balloons start at about 14 psi. So could you create a strong enough/light enough membrane with lightweight bands (kevlar or carbon fiber) to keep the shape of the membrane regardless of pressure without transferring the pressure loads to the outer shell?
How will you see through it? Pass controls through it? Get in and out of it?
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
15,350
Location
Port Townsend WA
For a homebuilt GA airplane, using the 5 psi differential someone stated in an earlier post, 30" fuselage radius (excessively large guess) and 0.032" skins, the nominal hoop stress is about 5 ksi, and a realistic design lifetime would be about 5000 cycles.
That doesn't sound like much additional structure is needed for pressurization.
 

Matt G.

Well-Known Member
Joined
Nov 16, 2011
Messages
1,266
Location
Kansas, USA
That doesn't sound like much additional structure is needed for pressurization.
5 ksi on top of what is already there could be, especially with how much of the area of a GA airplane fuselage is covered in windows and doors. Also, a non-round fuselage will need a lot more substantial frames, as the fuselage will want to turn into a round cross section with the application of pressure.
 

wrmiles

Member
Joined
Oct 13, 2019
Messages
24
The 210's system ran the cabin air from the turbo, through the sonic valve (which generates a sonic pressure wave to act as an increasing restriction to the air as its velocity increases, and the pressure drop of the air cools it some), through the heat exhanger (a little radiator) and into the cabin. The exchanger was in a box with a valving mechanism to duct either heated air from around the exhaust system (just like the usual cabin heat system) through it, or cold air from an air inlet in the front of the cowl. The airplane I worked on also had an aftermarket airconditioning system, with the usual compressor and evaporator and condenser exchangers. It was used mostly on the ground. It's weight was not insignificant.
It appears I was wrong on the heat exchanger. Like I said, it has been a long time and I didn't work much on piston aircraft design. Tried to find info online a 300/400 series twins without success. There also must be some sort of manual shutoff valve, but I couldn't find/don't remember details.
 
Top