Disassemble and Discovery

Engineering Analysis – Corrosion Mechanisms: Floor, Shell, Frame and Fasteners

/ Pete / Leave a comment

When we took apart the frame, we were surprised to find the frame and fasteners were significantly corroded. The frame crossmembers underneath the vanity was paper thin, the fasteners which held the sub-floor the frame were so embrittled that they snapped off when we tried to remove them.

This was a little concerning to me since I wanted to understand the mechanism to prevent it from happening again.

Galvanic Corrosion: My initial thought was that the predominant corrosion mechanism was galvanic corrosion. Galvanic (or two-metal corrosion) occurs when two metals of different potential (affinity for electrons) are in electrical contact within a corrosive environment. Simply put, the more anodic (i.e., lower EMF metal – Aluminum in this case), gives up its electrons to the more noble metal (Iron in this case). The electrons travel to a place where they can react in an oxidation reaction.

In case you were not aware, Airstreams are constructed with Aluminum and Steel in physical contact with one another, setting the stage for galvanic corrosion to occur.

In Corrosion Engineering, Fontana discusses that galvanic corrosion usually occurs at the metal-metal interface with the corrosion occurring on the more anodic metal (Aluminum). Fontana also discusses a phenomenon known as the area effect in which a large anodic surface areas (e.g., large Aluminum shell) in contact with small cathode interface area (e.g., small steel bolts) will tend to slow the reaction.

In short, what I would’ve seen if galvanic attack was the source of the corrosion was corrosion of the aluminum at the Al-Fe interface. In fact, all of the corrosion I saw was of the iron components (bolts and frame). This indicates that the corrosive mechanism was not galvanic attack.

Now that galvanic was ruled out as a source of corrosion, I started to think about other mechanisms. Looking at the frame showed general overall corrosion with areas of greater thinning. In other words, the whole thing rusted but some areas were worse than others.

Uniform Attack (aka Rust): Per Fontana, uniform attack is the most common form of corrosion and is characterized by an electrochemical reaction that proceeds uniformly over the entire exposed surface. This was the case for most of the frame and it is assumed to be this was the primary source of corrosion.

Rusted Bolt in Annual Space Between Inner and Outer Wall
General Corrosion with Increased Thinning Under the Vanity

Hmmm . . . why was the rust really bad in some spots, and not others? Doesn’t sound very ‘uniform’ to me!

Environmental Conditions Affecting Corrosion: You may be wondering why some areas corroded worse than others if the mechanism was essentially the same. The reason can be found in the abundance of oxidizers at the corrosion site.

The presence of oxygen and water will increase the oxidation reaction at the place where those chemicals can be found. The photos above show two such locations: The annular space between the inner and outer shell, and the frame area directly below the bathroom sink. Presumably the outer shell and piping to the sink both leaked.

Representative anodic reaction: Fe -> Fe2+ + 2e–

Most likely cathodic reaction: O2 + 2H2O + 4e– -> 4OH–

Wood Rot: The old sub-floor was made of plywood. Plywood is also susceptible to damage relating to water exposure. However, in this case, it’s susceptible to water absorption, rot and molding. The picture below shows the joint between the sub-floor, c-channel and shell (near the door) and is representative of the wood rot that existed throughout the trailer.

Sub-Floor As-Found Wood-Rot

Corrosion Prevention: It’s unrealistic to assume we can stop all corrosion, but it is reasonable to assume we can slow it down to so that the trailer lasts another 57 years. Here is what we are considering.

  1. Prevent oxidation from occurring at the frame’s metal/air interface by removing that interface through the use of spray-on coatings (i.e., primer & insulation).
  2. Leak-proof the exterior shell. to prevent water from getting into the annular space
  3. Select insulating material which will not hold water (i.e., won’t hold oxidizing medium)
  4. Insulate the iron fasteners from the aluminum shell using gasketing material
  5. Allow for drainage in the shell annular space.
  6. Sacrificial anode (i.e., zinc anode, or zinc galvenization) of the frame so that when oxidation occurs, the metal giving up its electrons is not the steel . . . . or the Aluminum.
  7. Different sub-floor material which is not susceptible to wood rot and/or water absorption.

Engineering Analysis – The New Sub-Floor

/ Pete / Leave a comment

One of our goals in the upgrade is to modernize the trailer and improve component reliability. One element of this upgrade effort was to improve the lifespan of the flooring. The old sub-floor had been replaced at least once and we were, I think, looking at the 3rd sub-floor when we started the rebuild.

Note: As you read this, please keep in mind that it is unique to a 1963 19′ Airstream Globetrotter trailer / sub-floor combination. Other materials, frame sizes and geometries will have different mechanical properties.

As you can see, it was rotted and in pretty bad shape.

As-Found Flooring Condition (rotted).

We determined early in the rebuild process that the original subfloor/fastener combination had a life span of something less than 57 yrs (see above).

For the new floor, we had three primary goals:

  1. First we wanted it to last pretty much forever
  2. It had to be cool . . . composite flooring sounds cool
  3. It had to last pretty much forever!

Putting the cart before the horse, We went after goal #2 first and bought 4 each 4×8 sheets of Composite boards for the new sub-floor. . . .afterwards we started to wonder if composite boards were such a good idea.

Now that we had leapt into the unknown and bought the board, we set about classic confirmation bias to prove to ourselves that the board we bought was indeed up to the job. Unfortunately, there was no readily available engineering analysis to reference, so Pete set about doing his own analysis using techniques he had learned while pursuing his Mechanical Engineering master’s degree at Texas A&M.

Side note:

MEEN 688 Advanced Solid Mechanics was a crushing but strangely enjoyable course. Imagine doing partial differential calculus with 3rd order tensors; all while 32 years out of your engineering undergrad! Pete had to re-learn how to use a calculator . . . much less all that math stuff.

The picture below shows the view of the sub-floor/frame combination showing areas of concerned where the subfloor is suspected of being insufficiently supported, either due to a lack of frame support or a panel joint between the sub-floor pieces.

Questions to answer:

  1. Will it last longer, or is it susceptible to mold, water absorption, UV radiation, . . .? (i.e., what are it’s corrosive properties?)
  2. Is composite board strong enough for this use?
  3. Is composite boarding a better choice than plywood?
  4. Do we need to install additional outriggers or cross members to the frame?

To answer #1 Pete did a little research. Plywood is actually stiffer (i.e., should deflect less), but it is susceptible to water absorption, rot and mold. Composite boarding is susceptible to UV radiation, but won’t absorb water, rot or mold. Conclusion: composite board will last longer if you can protect it from UV radiation, edge effects and if it’s adequately strong . . . which leads to question #2.

To answer #2 Pete did multiple iterations of engineering analysis. First by hand-calculating modeling analysis of small sections of the floor and then validating those calculations using engineering modeling software.

Hand Calculations: Fitting for a 1963 Airstream, we found an engineering reference from 1959 entitled Theory of Plates and Shells, which, albeit in a circuitous and confusing manner, provided sufficient technical guidance to model the sections of the floor encircled in red above. In the absence of computers in 1959, problems like this one were tackled piecemeal using finite sub-sections, formulas and look-up tables. One such analysis was the modeling of the center, left rectangle above (scenario 1). This section was modeled as a plate, clamped on 3 sides with one free edge.

The snip below shows an excerpt of the types of formulas contained in the 1959 reference.

Scenarios 2 and 2a were modeled using a method described in Theory of Plates and Shells to model overhanging floors and resulted in only deflection calculations (i.e., how much will it move up or down), but not stress. Stresses were approximated by correlating deflection with principal stresses, but this was done with low confidence.

Computer Calculations: A computer modeling simulation was next performed on the same geometries as the hand calculations. Scenario 1’s results were very close to the hand calculations. . . 🙂 . . ; while scenarios 2 and 2a results were of the same order of magnitude . . . meh.

Scenario 1 Computer Simulation
Scenario 2 and 2a Computer Simulation

That’s cool. It looks like it will work. The final step in the engineering analysis was to model the entire assembly (4 plates on the frame) to determine if a laminar coat of fiberglass was need for stiffness.

The final analysis of the sub-floor was done with computer modeling to determine the maximum stresses and deflection with and without a re-enforcing fiberglass layer. The modeled deflection went from about 1/2 ” at the cantilevered overhangs (scenarios 2 and 2a) without re-enforcement to nearly zero with a re-enforcing fiberglass layer.

Bottom line: The sub-floor should pretty much last forever as long as it is not exposed to UV radiation for extended periods of time AND the sub-floor will be sufficiently stiff if we use 3/4″ composite boarding with fiberglass re-enforcement.

Chapter 8a: The Subfloor Removal

/ Pete / Leave a comment

Question: What kind of floors do submariners walk on?

Answer: The subfloor, of course?

OK, to be honest I had no idea what a subfloor was when we started this whole extravaganza. Stef probably did, but unlike me she’s a real engineer :). I mean, as the name implies, I figured it had something to do with the floor, duh! For those of you who share my lack of understanding of construction lexicon, the subfloor is the floor that sits right on top of the frame and forms the support structure for putting down tile.

As it turns out, the subfloor also has a very important structural function in 1963 Airstreams. It’s kinda like the glue that holds everything together. The axle attaches to the frame. The frame attaches to the subfloor, and the shell attaches to the sub-floor.

I know what you’re thinking . . . “Hmmm . . . why doesn’t the frame attach to the shell?” (at least that’s what I was thinking). I’m not quite sure, but I think a partial reason was to minimize the effects of galvanic corrosion.

In nerd-eze, galvanic corrosion is the corrosion damage which occurs when two dissimilar are brought into electrical contact and the metal with the higher affinity for electrons takes them from the metal with the lower affinity. In this case, steel is a more noble metal compared to aluminum which means aluminum will give up its electrons which will travel through the steel preferentially looking for a place to oxidize a surrounding chemical.

Follow this link to learn a little more about the corrosion we found and what we’re doing about it:

https://ps63airstreamrebuild.com/2023/03/29/engineering-analysis-corrosion-mechanisms-shell-frame-and-fasteners/

The super awesome drawing below shows how that connection was made on this trailer, while the “as found” picture gives you a feel for what we found when we started looking at the screws holding everything together. The third photo shows most of the connection depicted in the drawing (missing the subfloor and shell). Nearly every one of the “fasteners” thoroughly corroded. Some simply snapped off when we started working them with vice grips. (Side note: I’m glad we made it to TX from Northern NM without anything really big coming off.)

Artist’s rendition of subfloor/shell/frame connection
As-found badly corroded fasteners
Belly pan to channel connection after the subfloor was removed (near back bumper).

Now, I know what you’re thinking. “Pete, what about the belly pan? That’s aluminum right? And it contacts the steel frame directly, right?” Good question. As it happens, belly pan aluminum is a more corrosion resistant (5052 H32 Aluminum in case you are taking notes.)

The subfloor came out relatively easily once we found all of the fasteners and neutralized them with extreme prejudice. The most challenging part was that original subfloor was replaced piecemeal and somebody had used a ton (I mean a s-load) of sheet metal screws to connect the “after-market, non OEM” subfloor to the frame. About 1 in 10 actually connected to something and the rest were safety hazards. (OEM = Original Equipment Manufacturer).

Under the subfloor we found another former rodent habitat. This one was more akin to a country villa complete with lounging areas and extra living space.

After pic showing wild animal living quarters
What it looked like as we pulled out the subfloor
This is where the toilet used to be.

It’s interesting to note that since the trailer was permanently affixed as an apartment previously, it was hard piped to a septic tank. The “toilet” picture above shows where the toilet used to sit and the hole is where it connected to the septic tank piping. For a fleeting moment we contemplated going back with the “straight pipe” design and relying, instead, of parking over pre-dug holes or flowing bodies of water. 🙂 J/K.

Chapter 7: Pulling Windows, Panels and Insulation

/ Pete / Leave a comment

‘Mama Always Said, Life Was Like a Box of Chocolates. You Never Know What You’re Gonna Get.’

Forest Gump

Or in this case it would’ve been more useful if Mama said . . . “Refurbishing a 57 year-old Airstream is like a box of chocolates. You never know what you’re gonna get, or what critters used to live in the walls, or under the frame, or how it was put together, or how it should come apart.”

Those of you who like puzzles and surprises would’ve definitely like this. On a positive surprise standpoint, my son, Tim, surprised us with an unexpected father’s day visit and helped rip some stuff apart!

Tim (My son / Stef’s Brother) visited from Boulder for a Father’s Day surprise!

Windows and Window Frames

The windows were a definite challenge. Each window was framed by an aluminum frame that was glued in place with paint or some sort of sealant. They also had painted in locks and hand cranks which were fastened with stripped out screws. Long story short, it took a while to get the window frames out.

Note: The windows themselves were easily removed (sorry no photo) from the exterior. There is a tongue-in-groove assembly on the outside and they came out with a quick shot of WD-40.

The photo on the left also shows what the long center overhead panels looked like right after we pulled them down.

Removing the Lock from the Aft Window
Every Opening Had a Frame Similar to This One Found in the Ceiling Vent

Panel Removal

Once we got the windows out, the panels came out pretty easily. We went through a ton of drill bits drilling out pop rivets and had some challenges when some rivets were glued in place with sealant. The big surprise here (for me anyway) is that all of the curved aluminum pieces flattened out once we got them removed. This means that we don’t have to shape them when we reinstall.

As we removed the aluminum panels, we came to note that there were tons of aluminum “C Channel” spacers in between the inner and outer walls. They were not connected to anything and fell out during disassembly. My best guess is that they were used to keep the space between the two walls at the correct gap.

Another surprise was that the two upper end pieces were actually fiberglass (we actually figured this out while we were stripping paint). We removed them intact with the hopes of reinstalling them later.

Aft Fiberglass Bell
Forward Fiberglass Bell
Tim As We Were Pulling Panels
In-Process Photo of the Left Side Wall as We Pulled Panels

Wildlife Refuge

This is where the surprise part comes in. As my wife will happily tell you, our family house in New Mexico is attractive to rodents. As it turns out, so are 57 year old Airstreams. As we disassembled the walls, we found evidence of spacious multi-level rodent living accommodations complete with food storage. I’m so glad we decided to pull the panels and clean this space out.

Hollowed-out nuts, and, uh, other stuff . . .fell out as we removed debris.
Spacious 3 Level Rodent Condo – Fully Furnished

Fiberglass Itches

I’ve known for quite a long time that fiberglass itches. I would, however, be remiss as a parent if I didn’t pass that knowledge on to my kids :). The fiberglass came off very quickly (matter of minutes) to expose the power distribution wiring, controls (i.e. brakes and lights) wiring, mastic and outer shell.

There was tons of mastic at the inside of every seam. Presumably for water tightness.

Looking Forward Before Pulling Fiberglass
Looking Towards the Rear After Fiberglass Removed
Stef and I as we Pulled Fiberglass

Before and After insulation interior videos.

360 View Before Stripping Insulation

360 View After Stripping Insulation

Chapter 5: OK. You got it home, what now?

/ Pete / Leave a comment

In my mind I was already rebuilding this grandiose super-cool, ultra-modern trailer with solar power, UV reflective windows and composite flooring. In reality I had no idea how to get from point A to point B . . . so we just got at it. Step 1 – try to figure out what you’re doing. Step 2 – do it. Step 3 – repeat step 1. . . .

I suppose step 1 of the rebuild started in New Mexico before we brought it home.

The first part of our rebuild was to empty and wreck out the interior. As you can see, in addition to 15 bags of stucco, it had some fine furniture, an ultra-modern beautiful utility sink, fine ply-wood trim and an state of the art hot water heater (complete with holes); all of which we kept for later reinstallation ;). The trailer had been used as a storage shed for literally decades and it needed to be emptied.

The Starting Point

Next we removed the top 3 layers of flooring to expose the sub-floor. This was to remove weight and give us something to bolt the driving controls junction box to (see previous post about electricity). The floor had tile, on top of grout, on top of a wire mesh on top of a plywood sub-floor. This was perplexing to me because it added a ton of weight and I couldn’t figure out why somebody would do this.

  • Pulling Up the Tiles
  • Stef Trying to Figure Out How to Remove the Tiles
  • Half Way Through Removing the Grout
If this looks like a lot of work . . . try doing it at 7,700 ft altitude! We were sucking wind!

We later figured out that the original sub-floor had been removed and replaced piece-meal. Per the Airstream’s original design, the sub-floor was originally a structural member and it connected the frame to the shell (more details to follow in later posts). As structural member it was no longer intact and it was no longer attached to the frame. Somebody (probably my mother) over the years had replaced the original sub-floor with a ply-wood patch floor and then smoothed it out with the grout.

Once we got it home, we set about removing everything on the interior.

  • This Is The Original Shower
  • The Side Shelves Came Off Relatively Easily

We found about $2.00 in change in the shelves, an old tablet of Benadryl and (my personal favorite) a 30+ year old stick of Wrigley’s Double Mint Gum! . . . and no we didn’t split it in two and have a bubble blowing contest.

Hey Bro, want some gum?

. . . and we’re done . . . ready to tackle the interior aluminum walls.