Stressed skin railcar - glued?

[trainguy]

Hi all.

Can any of you cite examples or applications of using sheet steel panels bonded to underlying stiffeners for a passenger railcar carbody? The skin is working in shear as a "conventional" stressed skin.


100% of my previous work in this area was using welding or riveting.

A client is asking to add windows to an existing welded structure, and some of my colleagues are suggesting bonding.


I personally get the creeps from this suggestion. Am I being overcautious?

tg

 

[MikeHalloran]

Structural bonding on a RAILROAD car?

No, I don't think you're being overcautious.



Mike Halloran
Pembroke Pines, FL, USA

 

[dhengr]

Trainguy:
Actually, creep (not your creeps) in the glue line would be one of the things you would want to look into. Theoretically it should be possible, and in fact be an improvement over taking the loads at discrete locations (rivets) or at just one line (welds) on thin material, since the shear flow is distributed over several sq. inches instead of an individual rivet.

The problem has always been the durability/reliability of the glue w.r.t. vibration, cycling stresses, weathering, expansion & contraction, changing glue line mechanical properties over time and the like. Then, there is the materials prep. for proper bonding, clamping pressure and its uniformity, etc. as relates to fabrication. I would talk to the glue suppliers, aircraft designers and builders, maybe truck trailer builders, and the like, makers of large stressed skin panels, Hexcel comes to mind. The people who manufacture the glue products should have some testing on things like vibration, freeze/thaw, water and UV damage, etc. The question is longevity and durability of the glue material, not initial strength which seems to be quite good with minimal special effort. I’ll bet that in the deep dark files of outfits like Pullman Standard, ACF, Bombardier, Valdunes, Hawker Siddeley, Budd, etc. there would be some very interesting reports and testing on this very subject.

If you can control the glue degradation it works well because the shear stress is only a few psi, whereas the stresses around the rivet, in the sheets, and in the rivet are quite high and susceptible to vibration, fatigue, etc. I’ve epoxied teflon sheets to steel blocks and plates to make slide bearings and am not aware that there were any problems with those designs. I also used up some of the left-over epoxy to glue some samples of .5" x 2" steel bar stock together, in lap joints of varying lengths, and these joints were tougher than hell. I never did do any serious testing to get meaningful forces or stresses. You might be able to use many fewer rivets since the glue really carries the forces/stresses, but the rivets prevent a peeling type failure. You are keeping your detail, loadings, etc. top secret, so it’s difficult to comment further. But, you want that joint to act in shear, a shear flow in lbs./sq.in., in the plane of the glue joint and sheets being joined, not in prying or peeling, or in a tension perpendicular to the plane of the joint. If you were cutting internal stiffeners in your application I probably wouldn’t do it, because of the way they concentrate loads. But, if you were cutting two skins in an area btwn. two stiffeners, I could imagine fab’ing. a strengthening/stiffening frame for that opening and maybe glueing it btwn. the skins. Then after the glue had set up, adding a few rivets at corners, etc. just to minimize out of plane flexure due to vibration, etc. Then the window could be set in this opening. It would be a mighty interesting experiment if there wasn’t too much at stake. But, I wouldn’t quote 500 cars with this detail without more to hang my hat on.

 

[trainguy]

dh,

Another well thought out informative post. The idea in the suggestion to use glue was to avoid fasteners altogether. Having glue and rivets is a promising approach to ensure good contact between the mated surfaces, but the quality control over all the things you mentioned is quite daunting, and doesn't at all fit into the cost structure of this small engineering project.

I have managed to convince my colleagues and clients that welding is the way to go.

However, I'm still VERY interested in whether or not anyone out there has experience with this kind of connection in a stressed skin application, anywhere. As I often ask people: Has it been done anywhere even once?

Thanks.

tg

 

[MintJulep]

Windows bonded to skin - yes.

Door panel skins bonded to honeycomb core - yes.

Floor panels bonded to structure- yes.

FRP skin (non-structural) bonded to structural frame - yes.

Structural skin bonded to structural frame - not that I am aware of.

 

[Tmoose]

dhengr said - "I’ll bet that in the deep dark files of outfits like Pullman Standard, ACF, Bombardier, Valdunes, Hawker Siddeley, Budd, etc. there would be some very interesting reports and testing on this very subject."

trainguy - do you have access to those folks' designs to know if they have successfully adopted bonding in production?

 

[chicopee]

Excessive temperatures could be a problem on adhesives.

 

[trainguy]

Thanks for the confirmations guys.

Tmoose:

Because I work for a small engineering consulting firm, I do not have access to most of the designs of the mentioned companies. However, in the Pullman, Bombardier and Budd designs I have touched, spot-welding and riveting were generally used for structural skin to structural frame connections.

tg

 

[MikeHalloran]

From several sources at different times, I have heard rumors that an absolutely obscene amount of maintenance work is required to keep a B-52's skins adhesively bonded to its structure. I have no first hand knowledge of adhesively bonded aircraft structure or of B-52s.

I was able to find this ancient official report on a related subject:

http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA107995

My own experience, in environments much more benign that that inhabited by railcars, typically electronic instruments in air conditioned laboratories, says that structural adhesive, or VHB tape, or whatever magic potion is being advertised, will, with the best of preparation per instructions, last about a year, then fail catastrophically, as in whatever was attached just flat falls off for no apparent reason with no obvious load applied.

The only exceptions, in my experience, are various name brand unfilled RTV silicones, and 3M 5200 (polyurethane), both of which are not self-fixturing, take a while to set up, a while longer to fully cure, and thereafter will adhere indefinitely, well enough to often require destruction of at least one of the mated pieces for disassembly.




Mike Halloran
Pembroke Pines, FL, USA

 

[berkshire]

Bonded construction has been used in aircraft for many years. One of the most notable was Fokker corp who used a rubber modified Phenolic Resin ( Aerodux) to bond the sparcaps to the wing skins.
For the most part my experience with this, in light aircraft and gliders, has not been very favourable. Slingsby aircraft used wet bonded construction on some of its models, then had to prove, to the CAA, that the rivets used to clamp the structure during glue up were sufficient to carry the load if the glue joint failed.
Grumman american corp used bonded construction on its AA1 yankee, and later tiger and Cougar lines. Several aircraft had service bulletins for disbonding of joints. Nearly all of these aircraft used rivets or other fasteners at strategic points to prevent peeling failure.
B.E.

 

[Compositepro]

Adhesive bonding in modern aircraft structures is very common (also composite marine structures) but very little of it will involve steel. You might get some interesting responses if you post your question in the Aircraft Engineering forum.

 

[chicopee]

Wasn't there a discussion about using Velcro on aircraft skin?

 

[SnTMan]

I hope not

 

[berkshire]

SnTMan,
There was,
But it was very specific and not for regular aircraft.
Certain crop dusters/sprayers use a velcro attached fabric belly skin.This is to enable the operators to drop the skin, and hose out the residues of certain corrosive herbicides, which otherwise would rot the steel tubes of the aircraft.
B.E.

 

[SnTMan]

berkshire, that I can buy