What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

can ProSeal DE-cure?

brian

Well Known Member
I pulled a fuel tank that had recently gotten more pronounced weeping/staining after several years of slight localized staining at one spot.

The proseal that is exposed on the outside has varying degrees of cure. The log books tell me this tank SB was done by the previous owner in 2010, just prior to selling to me.

After pulling the tank, I initially thought the person who did the SB work either used old proseal or didnt mix it well, because it looked like what was newer proseal was soft and sticky.

After looking at the tank more and starting to remove excess proseal from the outside, I'm starting to realize that the higher up on the tank, the more cured the proseal is, even in places where all the proseal I'm seeing would have been applied at the same time. In several low places (including where the seep originated) the proseal is very soft - so soft and gooey that it's like I just mixed it up a few minutes ago, and I can just wipe it off with a paper towel.

It seems unbelievable that someone would install a tank with such obviously uncured proseal, or that an A&P would sign off on such a thing. This tank has been in use in the 8 years since this work was done and an A&P signed it off.

So can something make this stuff somehow lose its cure? Perhaps it wasn't 100% cured in 2010 and since then the fuel in contact with it has caused it to lose its cure?

Here's a picture as an example of what I'm talking about. Look at the column of prosealed rivets for this rib. At the top, the proseal is pretty firm and dry. It gets softer and stickier as you go down the column, to the point where, at the bottom, I've wiped the rivet head, and some of the bottom rivets below it, completely clean, with just a paper towel and a bit of acetone. The outboard end, especially on the bottom, where the seeping originated, is the same way - with just a paper towel, I can wipe off big gobs of the stuff.

Can this mean this tank was installed with such uncured proseal? Or could the fuel somehow make the proseal lose its cure?

And most importantly, if something has made this proseal lose its cure, does that mean that the remaining proseal will continue to lose its cure?

Img_2602a.jpg


I don't know much about ProSeal or its varieties, and I haven't worked with it before, so any help or insight you can offer will be greatly appreciated.
 
Last edited:
While you wait for others more knowledgeable to chime in, I’ll add what I’ve observed: The blue dye in avgas seems to attack the pro seal and turn it into the goo you mentioned, but only in certain instances. It doesn’t seem to affect the generally submerged sealant but only areas where there is a leak. I don’t know if it’s the exposure to the air or lack thereof or what, but it spreads outward from the leak point as a larger and larger section turns “gummy”. Have no idea why. One would think that since the inside of the tank is exposed to air, it would happen everywhere, but it doesn’t. One hunch I have is that as the fuel evaporated at the leak point, the dye is left behind. It eventually accumulates significantly enough to start the reaction... just a theory.
Only fix is to remove all the gummed up sealant inside and out and start fresh.
 
Last edited:
While you wait for others more knowledgeable to chime in, I’ll add what I’ve observed: The blue dye in avgas seems to attack the pro seal ...

I doubt the dye has anything to do with it. It is just the residue left behind when the leaked fuel evaporates.

I have seen proseal fail. In the cases I've seen, I don't believe the proseal was thoroughly mixed before being applied.

All that said, I've never seen a successful repair where new proseal was applied over failed, gooey proseal.
 
I doubt the previous owner did anything 'wrong'. The symptom has been discussed before; I've had it in my 1991 RV-4's tanks. I initially thought it was my use of E-free mogas that was causing it, but I've since seen multiple accounts of the same problem in avgas-only tanks.

One theory I've seen advanced is that once a leak starts, the fuel & oxygen in the air supposedly conspire to weaken the sealant around the leak. Never seen any science to support that, though. And you'd expect that any tank with minimal fuel for any length of time would eventually have enough air inside to soften it's proseal. You should also be able to duplicate the symptom with test samples in a jar, but I've never heard of anyone being able to do that.

Until I see some science, I just chalk it up to 'stuff happens'.
 
Search the forum archives for many discussions of this symptom.
In a nut shell, tank sealant will break down and turn to goo with extended exposure to evaporating fuel caused by a leak.
You likely see it mostly near the bottom of the tank because that is closer to the residual fuel from the leak.
 
Lots of words with no new answer.

This "reversion" has happened before as described. DanH sent me his gathered evidence and I contacted 3M and made contact with the Chemist for polysulfides. She seemed interested but based on all we information we could provide, concluded that there was no direct reason within their studies that would cause this. It seemed a reluctant conclusion that it was poor mixing. I say reluctant because it was not a 3M product that has failed. She had no evidence or reports of such failures with 3M products used as commercial tank sealant for avgas. If there is a direct example of someone who has used 3M, in date as tank sealant, please provide all information and I will reopen correspondence on this issue. We would need to provide samples, dates, codes etc for 3M.
 
Search the forum archives for many discussions of this symptom.
In a nut shell, tank sealant will break down and turn to goo with extended exposure to evaporating fuel caused by a leak.
You likely see it mostly near the bottom of the tank because that is closer to the residual fuel from the leak.

Scott, how might a test be constructed to test this? I was thinking a dip with sample being oscillated up and down into a pool of 100LL. That should get it wet, exposed to oxygen, and allow it to dry, depending on the dip depth and cycle time .

Surely some zone of the material might experience the same a weep?? Thoughts?

Hopefully it would not take 5 years to run the test.:eek:
 
I am wondering if this may be the result of thinning the mix out with MEK or another solvent. I know a few builders mention doing this and the instructions that come with the Flame Master sealant are pretty clear not to thin out the mix.
 

That link took me to several more links, all with the exact same issue. So this isn't unique. Yes, I had a seep that I monitored for several years. It was just a blue stain, only 6" long and never wet, so I just monitored it at every preflight, and I and multiple A&Ps observed it at annual time, but it seemed stable and was never wet, so we continued to "monitor" it.

It sounds like exact same symptom as all the others - get a seep, don't fix it, then the external proseal it touches deteriorates and turns into peanut butter (good analogy from another thread) as the evaporating fuel sits on the external proseal.

As someone said in one of the other threads, the key is to fix the leak at the first sign of blue staining. Even if it isn't wet, it's still a leak. In hindsight, I should have realized it wasn't going to go away or get better on its own.
 
A couple things I have observed..

While this Proseal discussion is hard to give a direct answer to, I will tell you my experiences as an A&P working in the heavy aircraft and GA industry for 38 years, and probably have personally applied more than 100 gallons of Proseal and other trade name sealants. Currently, I manage KC10 tanker depot maintenance, and those bad boys use some sealant. I have never seen fully cured sealants "go back to pre-cured condition" unless it wasn't mixed properly to start with. I have found that improper clean/prep prior to sealing is the number one cause of leakage,and once fuel gets "under" the sealant because of this, you will inevitable continue to have leaks. Proseal and its derivatives have specific shelf life, mix times, work times, and cure times, some of which are weeks long. If the tank was re-worked by someone, he/she may have used outdated, or wrong specific kind of sealant, that you will never be aware of. In my line of work, as well as my personal aircraft, Its important to make a "test card" with a quarter size blob of sealant that is dated and has specifics of type, ect. After assembly of the tank, this test card can be checked for actual cure completion and time of cure. This is all seemingly complex,but essential. Improperly cured sealant can go for years without signs of leakage because the exterior flashes over but the core remains soft...but I have never seen properly cured sealant turn back to mush. Just my experienced 2 cents.
 
Mix by color vs weight

After a few years in the DoD aviation maintenance field, I learned that mixing ratio for "pro-seal" by weight per the instructions resulted in a 50/50 chance that the sealant would never fully cure. Most maintainers figured out pretty early that mixing the sealant by color resulted in a higher probability of curing. The typical color you get from the mixing ratio by weight is a light to medium grey color. Typically adding a little more of the black hardener gets you a dark grey / black color, which typically will cure properly. I asked our chemical engineers how important the mix ratio was and the answer was that too little of the hardener would not allow for uniform activation, but a little too much had no real affect other than ensuring the hardener was of a sufficient ratio to properly cure the sealant. The "potato masher grenade" type B-1 sealant cartridges are famous for not curing due to barely enough hardener or not being thoroughly mixed using the plunger. Using the two part can / hardener type sealant packages seems to be the best solution. The other factor that can cause the sealant to not cure properly is too low of ambient temperature. If the temps are too low, the sealant may never fully cure. There is nothing as pretty as uncured sealant on an H-60 tip cap that slings itself out in a starburst pattern on the first run up.
 
With respect, sealant absolutely positively can revert. The issue is peroxide formation, observed and studied starting about 50 years ago. And it's been seen here for quite some time:

DSCN1016.JPG


Here are two studies on the subject, and each bibliography provides clues to more, if they can be located. First up, take particular note of page 4 and table III. We're interested in manganese dioxide cured polysulfide:

https://www.danhorton.net/VAF/Can Proseal Decure/Fuel Peroxides.pdf

The next is interesting because of its parallels to the sealed tank question, i.e. why reversion inside tanks is seen much less often than exterior reversion. Test sealant in capped bottles reverted at a far slower rate.

https://www.danhorton.net/VAF/Can Proseal Decure/Accelerated Peroxide Formation in Fuels.pdf

From an Exxon brochure detailing world jet fuel specifications. You'll find reference to hydroprocessing in one of the above papers:

Antioxidants
Hydroprocessing of aviation fuels removes naturally occurring antioxidants that provide protection from peroxidation. Peroxides are known to attack elastomers causing embrittlement while also contributing to gum and particulate formation. The use of antioxidants effectively prevents peroxidation from occurring and under JFSCL and Def Stan 91-91, 17 to 24 mg/L of an approved antioxidant must be added to the proportion of the fuel blend that has been hydroprocessed. All of the additives are approved by chemistry and so there may be any number of suppliers for each individual antioxidant type. The use of antioxidants is optional under ASTM D1655.


Now the wild card, as it relates to us poor folks. Polysulfide sealant was never intended for avgas, and none of the compliance standards test it with avgas. Sealant tests per AMS-S-8802B are conducted using JRF (Jet Reference Fluid), a specific mix of ingredients, and doesn't address peroxides anyway.

I feel sure there is a research document somewhere in the world which explores sealant reversion on a tank exterior subjected to slow avgas leakage. However, like everything else these days, it's probably locked in some corporate filing cabinet because the legal department said so.

So, lacking specific research (read "someone throws money at it"), we're living with the observation that if you get a weepy leak, you better fix it right away. Fully understanding sealant reversion with avgas will probably require someone in the RV community with the appropriate expertise in chemistry.

I'm no chemist, but I'll take a guess...the antioxidants lose their effectiveness or evaporate or otherwise go AWOL when outside the tank. Yes, there are anti-oxidants added to 100LL. See the list in this document, page 6:

https://www.danhorton.net/VAF/Can Proseal Decure/UK Fuel Standard.pdf
 
Last edited:
After a few years in the DoD aviation maintenance field, I learned that mixing ratio for "pro-seal" by weight per the instructions resulted in a 50/50 chance that the sealant would never fully cure. Most maintainers figured out pretty early that mixing the sealant by color resulted in a higher probability of curing. The typical color you get from the mixing ratio by weight is a light to medium grey color. Typically adding a little more of the black hardener gets you a dark grey / black color, which typically will cure properly.

I mixed Flamemaster sealant samples at 10 to 1, 10 to 2, and 10 to 1/2 ratios, i.e. standard mix, double the manganese dioxide, and half the manganese dioxide. All cured just fine, and had similar physical properties, to the extent I could observe in the home shop.

All were submerged in an MEK/lacquer thinner mix for a few years. There was some hardening and loss of elasticity, but there wasn't much difference between the samples. The solvents did appear to leach excess catalyst, as evidenced by the color change of the liquid.

A 10-1 sample was also submerged in ordinary 3% hydrogen peroxide from the drug store. No significant change was observed after years submerged. The dark color seen here was rust; the peroxide did eat the can. I put no particular emphasis on this test because I have no real understanding of the chemistry.

Soak%20600pixels.jpg
 
Last edited:
proseal turning gummy

I too have experienced this problem. Several months ago, after noticing some seeping from the port tank on my 9A, I pulled the tank and discovered that a minor seep had been occurring from a screw on the sender for a long time but apparently evaporating before getting to the outside to be seen. The Proseal on the bottom of the inboard rib seam was soft and gooey and another seep had started from the inboard rear corner where a fillet was placed on the corner on the outside as well as inside. I cleaned everything inside and out and re-sealed inside and in the rear corner on the outside. It fixed the leak. Then late last week, I noticed a leak on the starboard tank, which had never shown any signs. This leak WAS significant with a drip rate onto the gear leg fairing at a rate of a drop per second. Good thing I noticed it in the hanger just before leaving because the tank would have leaked the whole 18 gallons by morning at that rate. After removing the tank, I pressurized wih 1 PSI and did the soap test. I actually found a seep at a sensor screw like the other side and the major drip leak was coming from, you guessed it, the rear corner seam of the outer rib. All the Proseal in the seam area was gooey and just wiped off easily. The top seam, not exposed to the leak was properly hard and slightly resilient. I repaired all as on the port side tank, and tested it full of fuel yesterday and today and no more leaks. The new Proseal set up properly.
Ed
 
Back
Top