Oversized Fasteners in Repairs 2

[SuperStress]

Many structural repair guidelines specify to use a 2nd oversize fastener when replacing an existing fastener in an assembly (especially when adding a doubler).

My understanding is that this is done for two primary reasons:

1. To restore the desired fastener fit.

2. To compensate for the added flexibility in the joint resulting from the increased eccentricity.

Are there any references around that give some details about the joint flexibility aspect?

Short of building a spring model for each repair, is there a simple way to determine when a 2nd oversize fastener is actually required?

This issue is coming up because of limited availability of 2nd oversized fasteners for field repairs.

Any help is appreciated.

SuperStress

 

[koopas]

Superstress,

As a general rule,

If original rivets are replaced by hex-drives or lockbolts, a 2nd oversize fastener is used in the repair.

If hex-drive or lockbolts are replaced by the same type of fastener, a first OS is used in the repair.

Rivet replaced with rivet: same diameter


Hope that helps,
Alex

 

[SuperStress]

Thanks for the reply.

I have the guidelines for what fasteners to use when, my question has more to do with when it's okay to deviate from the standard practices, assuming your hole quality meets the installation tolerances for the smaller diameters...

Thanks,
SuperStress

 

[Sparweb]

The OEM's also assume that drilling out a fastener in the field will likely cause damage/elongation of the hole. Oversizing the hole gets rid of the damage around the edge of the hole.
When driving an aluminum rivet, the expansion of the rivet induces a compressive stress in the circumference of the hole. The exact effect of the compressive stress is unknown to me, but assuming that there is some permanent set in the hole, then the hole will be permanently larger, even if you take that rivet out. Without the rivet, the edge stress of the hole goes back to zero, but it's still wider than it originally was. Put in a replacement rivet that's the same size as the original rivet, and it expands into the hole, but does not induce the same compressive stress at the edge of the hole. The fatigue quality of the joint has just dropped.
The compressive stress around the edge of the hole helps keep cracks closed.



STF

 

[jetmaker]

SparWeb, SuperStress:

SparWeb, I'm curious about your statement "Without the rivet, the edge stress of the hole goes back to zero, but it's still wider than it originally was". Could you explain how a permanent deflection in the hole diameter can exist, yet no residual stress. According to your statement, no benefit would be derived from coldworking holes either. In general, if the fastener has not yet been loaded in service, then I just replace it with the same size.


SuperStress:

Is it possible to give us an example. At the moment, I can not think of an example of where you would add a doubler to a structure without changing the loads which pass through the fasteners.

Here is a tip tho with regards to replacing fasteners. If the hole has been loaded due to service, it is prudent to replace the fastener with an oversize one. Some believe this to "zero-time" the hole. Even though I do not believe you actually zero-time it, it does help to minimize any cracks that have started at the hole edge.

Thanks.

 

[SuperStress]

The issue that came up had to do with a doubler installed on a fuselage longeron of a military aircraft. The original engineering for this beef-up was performed by another company and we did not have ready access to the stress report.

The existing fasteners at this location were NAS1669, and NAS1751 was the replacement. The work was being done at an overseas location with limited access to oversized fasteners. We had a request from the overseas shop to re-install nominal sized fasteners if the hole dimensions were acceptable.

I never saw all of the details of the modification; this was more of a "hallway" conversation with the designer where I was trying to explain the reasons why the oversized fasteners are required in these cases in general.

In this case, without access to the original engineering, we couldn't justify going back to the nominal fastener since we couldn't be sure what the margins were and whether or not the design could tolerate the change.

What I was really looking for was some discussion of the factors that influence the fastener change, especially joint eccentricity. Is the change to a 2nd oversize partially a "rule of thumb" so that you don't have a write a margin of safety against combined shear plus bending in a single-shear joint with thick members?

SuperStress

 

[Sparweb]

Okay, jetmaker, you caught me brushing over the topic too quickly. I admit that there's a residual stress, even when the fastener is removed, but when you put a new one back in, does the stress in the material around the hole go back up to the same level that it was? I doubt it. Perhaps I'm wrong.

The biggest variable is the skill and technique of the guy with the drill taking the rivets out. I try to design with this in mind, not the laboratory case that assumes that every hole is perfectly round and straight and within tolerance. There are guys out there who won't bother to deburr a hole before pushing the rivet through, and others that carefully polish around nearly every hole they make. I have some power to tailor the repair design according to my opinion of the shop's skill level. Oversizing rivets is one of the ways.


STF

 

[jetmaker]

SparWeb, SuperStress:

SparWeb:
As far as installing a rivet back in a previous hole, the stress in the parent material would return to the original state... the concern lies with the interference of the rivet. From my experience/knowledge, a standard driven AD, D, or DD material rivet will not result in a degraded hole fill when reinstalled. However, a KE or overdriven rivet can result in a different hole fill, and therefore it is prudent to oversize these fasteners to ensure proper fit.

SuperStress:
I have good news and bad news.... The good news is that I just lowered my car insurance by switching to Gieco. The bad news is that there are no good hard and fast rules to use when determining the STRUCTURAL justification for oversizing without doing the analysis. I've had the shop come back numerous times complaining that they do not have specific fasteners/grip lengths available. I've told them too bad, to find a shop that has them, and buy them. Or I'll design a work around.. but that requires loads, which unfortunately are unavailable to you.

We are in a business where failure is deadly and costly.

Have a good weekend all.

 

[jetmaker]

Just another theory as to why Nui might suggest oversizing the fasteners.

Say you have a frame web splice, where per design, one frame web overlaps another frame web. You have 3 rows of fasteners connecting the 2 frame webs together. Both webs are of equal thickness.

Now, say you have a crack in the critical row of one of the webs. To repair it, you decide to convert the lap splice into a butt splice, hence not changing the single shear joint configuration. The splice strap to connect the 2 webs together will now be one gauge thicker than the webs themselves (typical repair philosophy). Also, you assume that the original splice had a zero margin (again, typical design philosophy when loads are not available). If you keep the fastener size the same as before, the maximum load to be transfered into the splice plate is the same as it was before the repair. This tends to waste the added capability of the thicker splice plate. (Actually, the capability is down somewhat because of the added eccentricity). Therefore, if you oversize the fasteners, you can transfer more load into the splice plate, thereby making the splice better than before... assuming you now do not have a net section problem in the web due to oversizing fasteners. You could achieve a similar result by adding another row of fasteners at the joint.

So to summarize, the most likely reason for oversizing the fasteners is that it is based on the conventional repair philosophy of increasing the gauge on the repair doubler/splice.

I realize this does not address the original question of when is it ok not to oversize, but if you do not gauge-up the doubler/splice, then I would see no strength reasoning for oversizing the fasteners.

Have fun.

 

[edmeister]

I understand the mathematical arguements ... but believe that the practical point is missed ...
Having installed an occasional rivet in my day ... A larger diameter fastener requires a larger impact force to properly form ("buck&quot... & naturally this additional force could cause distortion or crack initiation of the adjacent (non-supported wrt riveting) structures. Installing fasteners on a bench (or factory jig) is one thing - but installing fasteners in difficult areas with offset snaps is a talent. more impact forces .. more probability of initiating a new problem .. next size = expotential riveting impact force.

In the case of the "auto-next-oversize" ... what will happen if that structure has to be repaired again & no edge distance is remaining - will countless hours have to be invested to replace the frame?

And nothing looks more impressive then huge round rivet heads on the outside of a fugelage!! Nothing like a streamlined aerodynamic shape with chickenpox !! Impress your Customers !!

Keep it simple !! the more deviations from the original - the greater possibility something may go wrong .. increase the hole & then find out that no fasteners in stock ... screw up drilling the oversize hole .. easier to slip off the rivet head with 5X gun ... or smash the fingers of the guy holding the only tiny bar that fits .. oilcan the bay .. some inter-rivet buckling due to sheet compression from the higher impact riveting .. in the case of CSK - how about some knife-edging .. or some shaving .. oops gouged the skin ..

My belief is if the hole is reasonable (within limits) for whatever required application (solid) and the original fastener has not been fretting (loose) .. keep the original!!
In the case of blind - lean more to O/S ... but not much !!
The marginal benefit for O/S must be weighed with the mechanics of installation .. & dont forget common sense !

 

[jetmaker]

ed,

Nice addition. You brought up some good points about limitations to driving rivets, and short edge/end margins. My thoughts tend to be that if we are repairing thin structure (small gauge and small stackup), I subsitute rivets with Hi-Loks as to prevent damage arising from the bucking process. Furthermore, if necessary to install a rivet, I'll downgrade the rivet material or switch to an "ice-box" rivet to make installation easier. Finally, hand driving rivets larger than 3/8" is a no-no for many of the reasons you stated.

However, I believe that regardless of the hole apparent condition, if the rivet has been loaded due to service operation, it is prudent to run a reamer through the hole to clean it up. If you are constantly having to remove a rivet such that oversizing results in short edge margin conditions, you have other problems than need to be addressed.

jetmaker.

 

[koopas]

Jetmaker,

Nice post regarding OS fasteners and taking advantage of the full load development of the repair splice plate. You did mention that you could run into net area problems with OS fasteners.

For intance, at a row where three 1st OS fasteners will be installed, would you check for net tension problems by comparing the remaining net area x Ftu, [(Length x width - (3D + 3/64))]*Ftu, to the total joint allowable for all fasteners (3*Pall) in that cross section where the net area was calculated? Then you'd ensure that the net area strength is higher than your total joint allowable?

Is that correct?

You also mentioned using hi-loks in place of rivets to prevent damage in the bucking process. Are hi-loks available in very short shank lengths to accomodate short stackups? I've always read to use hi-loks in thick stackups.

Regards,
Alex

 

[jetmaker]

koopas,

Your question on the net tension failure mode is correct. If the joint is designed poorly, it will fail in net section tension. Therefore, it is almost always better to design a joint to be limited by fastener bearing so that a joint failure is progressive, rather than instantaneous. But yes... as you oversize the front line fasteners, your start to reduce the net tension strength.

With regards to fastening parts together... you are correct in that Hi-Loks are preferable for thicker stackups. I use rivets when I join 2 parts together and the t/stackup ratio is less than 2. Once that ratio exceeds 2, I switch to Hi-Loks or even bolts. However, I will use Hi-Loks if I feel that the mechanics will do more damage bucking rivets. I am not sure what the shortest grip lengths available are.

Best luck

jetmaker

 

[koopas]

Hello Jetmaker,

Since we're still on the topic of oversizing:

Consider that you have three rows of rivets along a skin panel splice. The outside row, closest to the splice line, is the critical row.

One, perhaps two rivet holes are bad. You advise the mechanic to try to clean up the hole with a 1st OS. He comes back and says he needs to go to 2nd OS.

Q1: do you approve the 2nd OS without even going to the plane to measure the new ED or pitch? The 717 SRM 51-00-07 approves 2nd OS.

Q2: assuming that you've approved the 2nd OS, the mechanic comes back to you and says we have to go to the next size to completely remove the damage. What do you do? Typically, will ED and pitch with adjacent rivets still be okay? Will installing a next size fastener attract more load and reduce the fatigue life at that location? Could you counteract this added stiffness by going to say, a softer rivet? In such a case, would you go for OEM support and approval for the deviation?

Cheers,
Alex

 

[Sparweb]

Hi again Koopas,

Your questions tie back to an earlier question that you posed in another thread about short edge-distances. Though I recall that you weren't entirely satisfied at the time with the answer, your suggestions in this last post indicate that you're getting the hang of it. Increasing the size of one fastener by a 1/16th will have little impact on overall fatigue life, and using a softer rivet is a good idea. If you find yourself looking for something softer than an A rivet, good luck, but if you're going from an E4 to an AD5, you shouldn't be going too far wrong.

You can't make a general statement about the likelihood of the edge distance coming up short, just go check the aircraft yourself. If the ED dips below 2D at the up-sized rivet, you can always use Philcondit's advice from the other thread to substantiate it, saving a few calls (and bills) from the big "B".


STF

 

[koopas]

Hey Sparweb!

How did that big test of yours go? I've been bugging Ed quite a bit via email with my post on single & double shear joints. The man has got thick skin...lol.

You mentioned increasing the size of one fastener by 1/16th...wouldn't that constitute two sizes up? Did you mean 1/32?

Thinking about it deeper, I presume that going with a softer rivet will decrease your rivet shear strength but you're also upping the diameter so the effects somewhat cancel.

Take care,
Alex

 

[Sparweb]

Right, you are, I meant only going up by a 32nd, say from AD4 to AD5. That's what I get for replying to posts past my bedtime. (Talk about fatigue failure)

I passed my exam - not that they bothered to tell me! I'm getting a strong impression that they're not very prompt with their certification process, so I don't expect to get my little plaque until after Christmas...



STF

 

[jetmaker]

koopas, SparWeb,

Here is my take on oversized fasteners. In general, going a 1/32 O/S will not cause a significant reduction in ultimate or fatigue since the increase in diameter offsets the loss of edge margin. Note that this assumes everything was within drawing tolerance before oversizing. However, if the fastener was misdrilled in location and has a short edge margin, even a 1/32 can cause a lot of grief and needs to be checked.

I am not familiar with the 717 SRM, but I am sure it says in there that 2nd O/S is acceptable as long as you are still within 1.7D edge margin. Be careful of the use of blanket oversizing as airplanes are not all built to spec.

You ask about pitch... and my answer is that you do have to start thinking about it... especially with csk fastener in thin sheets where the average diameters start getting large. I generally check pitch Kts unless the load is inline with the fasteners or you are still maintaining about 5D to 6D spacing.

Finally... you have asked about changes in load transfer. Now here is where the situation gets complicated. In general... as a fastener gets larger, it tends to suck up more load. A general rule is to assume a weighted average based on shear areas and then multiply the standard load distibution by this number. For a 3 row joint, the standard distribution is about 35%, 30%, and 35%. Changing to a softer rivet does not really work, as the Young's Modulus for aluminum rivets is all about the same... regardless of temper. And since we are talking about fatigue mostly, the loading is on the linear portion. So.. switching to a AD rivet from a D material does nothing for you but weaken the joint.

Hope this helps.

jetmaker

 

[737eng]

Since you want to ensure the hole is crack free - you are req'd to HFEC the existing fastener hole. Typically you utilize a 0.030 notch reference (there are 0.020 notch standards available now). Therefore, even if you HFEC the hole, there could still be a small crack remaining that could not be picked up with the probe. Oversizing to 2nd over will aid in cleaning up these undetectable cracks.

 

[737eng]

I should of mentioned that oversizing 1/32" does not zero time the hole, since there is still a probability of a crack that cannot be detected by HFEC. To fully zero time a hole, you must first HFEC using 0.030 notch standard and then oversize hole by a full 1/16" (0.030 per side) to fully ensure that there is no crack.