What if the plastic coating on the skin isn't removed before riveting.

Discussion in 'Sheet Metal' started by 12notes, Aug 20, 2019.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Aug 20, 2019 #1

    12notes

    12notes

    12notes

    Well-Known Member Lifetime Supporter

    Joined:
    Aug 27, 2014
    Messages:
    823
    Likes Received:
    506
    Location:
    Louisville, KY
    While taking apart the rear spar/bottom skin rivets, I noticed a small (2"x2"ish) section of the skin between one rib and the spar that I failed to remove the plastic coating on the skin before riveting it to the rear spar. What would be the effect of leaving the coating on a skin and riveting it on? Would having the plastic sandwiched in the middle of the joint significantly weaken it?

    I'm not planning on doing this, just curious.
     
  2. Aug 20, 2019 #2

    dmar836

    dmar836

    dmar836

    Well-Known Member

    Joined:
    Jun 16, 2014
    Messages:
    97
    Likes Received:
    76
    Location:
    KC
    Imo, very much so where the plastic is involved.
     
  3. Aug 20, 2019 #3

    Kyle Boatright

    Kyle Boatright

    Kyle Boatright

    Well-Known Member

    Joined:
    Nov 12, 2012
    Messages:
    779
    Likes Received:
    411
    Location:
    Marietta, GA
    I've seen more than a few RV parts that were exposed to water while still wrapped. Corrosion under the plastic is pretty common in those circumstances.

    So, it will not immediately weaken the structure, but I'd remove it out of fear of corrosion sometime down the road.
     
    bmcj likes this.
  4. Aug 20, 2019 #4

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    412
    Likes Received:
    167
    Location:
    Hamilton New Zealand
    It would have a detrimental effect on the installed strength of the rivet (how much is the hard bit) and a major reduction in the fatigue life (order of magnitude would probably be a good place to start).
     
  5. Aug 20, 2019 #5

    Kyle Boatright

    Kyle Boatright

    Kyle Boatright

    Well-Known Member

    Joined:
    Nov 12, 2012
    Messages:
    779
    Likes Received:
    411
    Location:
    Marietta, GA
    Why do you say that? It is a layer of vinyl, approximately the same thickness as a layer of primer or paint.
     
  6. Aug 20, 2019 #6

    Angusnofangus

    Angusnofangus

    Angusnofangus

    Well-Known Member

    Joined:
    Sep 29, 2015
    Messages:
    318
    Likes Received:
    136
    Location:
    Victoria, Canada
    It is not something you would want over a large area, but 2"X2" I don't see as a problem, Just the act of riveting squashes the plastic down. Obviously better to not have it there at all, but as a structural issue, I think it is a non-starter.
     
    FritzW likes this.
  7. Aug 20, 2019 #7

    12notes

    12notes

    12notes

    Well-Known Member Lifetime Supporter

    Joined:
    Aug 27, 2014
    Messages:
    823
    Likes Received:
    506
    Location:
    Louisville, KY
    Oh, I've removed the piece of plastic. Just wondering what the actual strength difference be, an what effect would it have on something like an entire wing skin or fuselage skin. I would think the plastic would squish outward from the hole when the rivet is pulled, but you'd still have the plastic between the parts outside of the rivets. I have driven a rivet with plastic still on the outside if a skin near the hole, it cut out a little donut of plastic when the rivet was pulled. Not sure if that applies if the plastic were in the middle of the sandwich.
     
  8. Aug 20, 2019 #8

    FritzW

    FritzW

    FritzW

    Well-Known Member

    Joined:
    Jan 31, 2011
    Messages:
    3,532
    Likes Received:
    3,143
    Location:
    Las Cruces, NM
    I think it's one of those theoretical vs. reality situations. In theory it could be a problem but the reality of it is it probably wouldn't make a bit of difference.
     
    Angusnofangus and Pops like this.
  9. Aug 20, 2019 #9

    narfi

    narfi

    narfi

    Well-Known Member

    Joined:
    Feb 23, 2016
    Messages:
    711
    Likes Received:
    342
    Location:
    Alaska
    Winner winner chicken dinner!

    The reality is that you might end up with some corrosion issues down the road and that is probably the only 'negative' effect.
    However you will know, and it will bug you in knowing. It is easy to correct and replace a couple rivets, so do it now while it is easy. (as you said you already did)
     
    Angusnofangus likes this.
  10. Aug 20, 2019 #10

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,781
    Likes Received:
    3,007
    Location:
    Saline Michigan
    Two possibilities:

    First is with the vinyl remaining in the joint. Instead of the rivet firmly clamping the two pieces of aluminum together, putting the joint in compression and the rivet in tension, and the preload keeping the joint from fatiguing, the springy stuff in between is allowing the rivet to see some of the cyclic load, which is the source of fatigue.

    Second is with the vinyl removed but the rivets not being reset. Then the joint has zero preload and the rivet will see all of the cyclic load and will fatigue the rivet.

    The springy stuff should come out. If this is a squeezed rivet, you could just set it again with a squeezer or bucking bar and air hammer. A pulled rivet will have to be replaced. That or you could inspect this join line every flight looking for cracks in the skin and popped rivets...

    Billski
     
  11. Aug 20, 2019 #11

    BoKu

    BoKu

    BoKu

    Well-Known Member HBA Supporter

    Joined:
    Aug 16, 2013
    Messages:
    2,271
    Likes Received:
    2,832
    Location:
    Western US
    Hmmm... I have a hard time believing that a layer of thin vinyl is much different in thickness or compliance from the layer of polysulfide sealant used in a gazillion wet-wing fuel tanks. Of course, leaving the vinyl in is suboptimal for a variety of reasons already discussed here. But I wouldn't lose much sleep over a couple inches of it.
     
    Angusnofangus likes this.
  12. Aug 20, 2019 #12

    Dan Thomas

    Dan Thomas

    Dan Thomas

    Well-Known Member

    Joined:
    Sep 18, 2008
    Messages:
    4,829
    Likes Received:
    2,045
    Since a significant part of the shear strength of a riveted joint is attributable to the friction between the sheets created by the rivet's compression of the sheets, the presence of a flexible layer between the rivet head and sheet will eventually result in the loss of clamping force at that point as the plastic creeps out. It's not the same as riveting with some wet sealant on the rivet, since the bucking will drive out the sealant under pressure but leave it along the shank where it seals against leakage.

    As an aircraft mechanic I've seen the result of inadequate clamping pressure. It's not nice. You get fretting between the sheets that thins the sheets. And riveted joints aren't the only place you'll see it. The Cessnas use aluminum sheet covers over the fuel tanks in the wings, and those sheets are structural: they form one side of the box that the wing structure forms, and that box makes the wing torsionally rigid. If someone takes those covers off and reinstalls them without getting the correct torque on the screws, that section of the wing twists a bit and causes fretting at those screws. Sometimes the holes get elongated. The skins get wallowed out under the screw heads.
     
    Pops, BoKu and dmar836 like this.
  13. Aug 21, 2019 #13

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,781
    Likes Received:
    3,007
    Location:
    Saline Michigan
    I would hope so too. In most of the airframe joints, things do tend to be pretty lightly loaded, and the increased fatigue exposure is probably pretty small. But if the builder is thinking about it, well, maybe it ought to be treated like it is important...
     
    BoKu likes this.
  14. Aug 21, 2019 #14

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,704
    Likes Received:
    1,075
    Location:
    NJ
    I wonder if a careful application of heat might allow the rivet to be re-set and squish out the hot plastic. It should be above the transition temperature of the plastic but below the critical temperature of the aluminum......would require some research, and IR temperature gauge.
     
  15. Aug 21, 2019 #15

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    412
    Likes Received:
    167
    Location:
    Hamilton New Zealand
    For something in-service the first question would be whats the rivet head & skin look like (cracked paint, smoking, skin around the rivet looks just like all the other rivets etc etc).

    A per spec rivet is good for approximately 100000 cycles. The additional plastic will greatly reduce the transmission of load by friction which is typical in the low loading seen in GA*, the order of this effect is about a 50% reduction in fatigue life as demonstrated by the 50% reduction you get with WD40 in a rivet joint**. Paint is well attached to 1 or both surfaces so doesn't have the same effect. The plastic is only know to be adhered to one side of the aluminum, while PRC in rivet joints is good for something like 100 to 200 lb per inch shear (although the relative stiffness of the PRC verse the rivets is the interesting bit).

    The plastic is not going to preload anything as the young's modulus is just too low and under-high loading may make those rivets ever so slightly softer but at the same time increasing the rivet bending and possibly toppling effect (this is what causes may of the issues with low end cherry rivets). These effects are all very small however crack growth is proportional to the 4th power of the stress level*** and quantifying the aculative effects would be really hard.

    The worst unknown effect would be the increased risk of any exfoliation corrosion under the plastic which by its self can be an order of magnitude effect if one can quantify it.

    All in all, this is just two rivets, airframe fatigue life is normally a calculated number that is then divided by 6 (well untill there are some fleet hours to review or a budget crunch for those with ejector seats). Its just quicker to drill and replace them than to create a more precise answer than that previously stated

    *Gust load spectrum in GA has a very wide spread, most of the fleet stays away from the edges most of the time but not always.
    **roughly from a graph for WD40 in a rivet joint, in Riveted Lap Joints in Aircraft Fuselage: Design, Analysis and Properties.
    ** Yes I know its actually 3.9.
     
    proppastie likes this.
  16. Aug 22, 2019 #16

    proppastie

    proppastie

    proppastie

    Well-Known Member

    Joined:
    Feb 19, 2012
    Messages:
    3,704
    Likes Received:
    1,075
    Location:
    NJ
    Are not the allowables calculated based on shear of the rivet/bolt, bearing/tear-out of the sheet?....those are not effected by the plastic.
     
  17. Aug 23, 2019 #17

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    412
    Likes Received:
    167
    Location:
    Hamilton New Zealand
    Bolts are forged then installed, while rivets are installed then forged to final shape so process changes are going to effect the rivet strength.

    As for calculating rivet shear strength there is also a Shear Strength Correction Factor. The joint needs to be within specification with regard to rivet tail shape and sheet gap etc. Unlike bolts the tensile strength of installed rivets is at best 10% of shear strength and they can't apply clamping force to the joint so are much more sensitive to joint fit (than bolts), the plastic would affect the joint fit. If one is being pedantic there is also a population reduction factor* if you are using 7 or less ( I think) rivets as a group in a critical location*. OEM rivet shear allowables seem to be typically less than those published in MIL-HDBK-5H.


    *If you have a critical connection with less than 7 rivets you should be applying a knock down factor to the "B"basis design values for the rivets. Mostly covered by practical considerations like using even numbers of rivets in joints or add a rivet or 2 for luck approaches (or in practice covering the oops I know I told you it was done to the preliminary repair sketch but I cut an extra 1/2" inch out).
     
    proppastie likes this.
  18. Aug 23, 2019 #18

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,781
    Likes Received:
    3,007
    Location:
    Saline Michigan
    Yes, the several failure modes of a riveted joint are included in standard spacing and should be included in the design calcs for the joint. That is not the problem with adding a soft layer to the joint.

    In both riveted and bolted joints, the fastener ends up in tension while the joined members are put in compression. The joined materials are intended to be moderately loaded and very stiff in compression while the fastener is highly loaded and relatively springy in tension. There is supposed to be friction between the joined members, and in riveted joints the rivet is supposed to fill the hole with preload as well to further prevent the joined members from shifting. This is all analyzed for bolts in Shigley's chapter on fasteners. Then the joint is supposed to be analyzed to check that the joint does not slip nor open and close under its operating cycle. As long as the joint does not open and close, the fasteners will see only very small cyclic loads and will not fatigue. As long as the joined members do not slip, there is no shear on the fasteners.

    Design the joint with only metal to metal contact and you can actually use small factors of safety. Add a soft squishy polymer layer, and both the preload and friction are reduced, with joint opening and slipping much more likely. That is the origin of my concern, and sometimes the source of cylinders being launched at (and sometimes through) cowlings.

    Realistically, these airframe joints are supposed to have FOS of 4.0 to all failure modes, which makes any failure pretty unlikely if done correctly, even when painted or Prosealed, etc. And realistically, the vinyl layer is thin and kind of fragile, so it is likely to have been squeezed out of the immediate area of the joint when the rivet was set too, so its contribution to joint opening and slippage is probably small.

    Then to finish things off, if most of the rivets in the joint had metal to metal contact and all were set correctly, having one with the vinyl under it is most likely not an issue. But let's not make a regular practice of it, OK? If it were in my airplane, and I had this issue, and I could remove the vinyl and either reset a squeezed rivet or replace a pulled rivet, I would. Now if I would turn over heaven and earth to get at a single and otherwise inaccessable rivet with a bit of vinyl caught under it, well... I have zero plans to ever build a sheet metal airplane.:D

    Billski
     
  19. Aug 23, 2019 #19

    12notes

    12notes

    12notes

    Well-Known Member Lifetime Supporter

    Joined:
    Aug 27, 2014
    Messages:
    823
    Likes Received:
    506
    Location:
    Louisville, KY
    All plastic was removed well before I posted this, it was a question out of curiosity when i found and fixed it. At no point was I suggesting leaving it, I was just wondering what the effect would be if I hadn't, or if someone did a whole skin this way.

    I figured this had a simple, established answer, but it seems like it isn't so simple. Good discussion, I might try to do some testing on this later.
     
  20. Aug 23, 2019 #20

    David L. Downey

    David L. Downey

    David L. Downey

    Member HBA Supporter

    Joined:
    Aug 8, 2019
    Messages:
    10
    Likes Received:
    3
    Digressing a little, but all these facts are also invoked when talking about "unnecessary" correctly executed deburring of fastener holes...
     
    wsimpso1 likes this.

Share This Page

arrow_white