Request for Shim/Filler Joint Data

[edbgtr]

Good-day Eng-Tippers,
I’m looking for statistically significant single shear joint failure strength test results for fastener sheet, or plate (metal), combinations with and without various shim/filler thicknesses.
The ideal for me would be a series of test results performed on such joints where the goal was to determine the joint reduction factor for an increasing shim to host sheet ratio.
If anyone has data of this kind that they can pass on, I’d be most grateful.
Ed.

 

[WKTaylor]

EDBGTR...

Most likely You're looking for proprietary corporate data [test reports]. Good luck.

I checked a bit. Except for liaison engineering training data [various sources], there is very little discussion regarding structural shims and fillers in open literature. I think structural engineering Texts may be Your best bet. Check-out E.F. Bruhn's "Analysis and design of flight vehicle structures". Shims [and maybe fillers] are discussed in chapter D3:

CONNECTIONS AND DESIGN DETAILS...
D3 Some Important Details in Structural Design.

Might want to check-out DoD-STD-35-5 "AUTOMATED ENGINEERING DOCUMENT PREPARATION SYSTEM WASHERS, SPACERS, AND SHIMS".

NOTE: You think finding testing data for structural shims and fillers is hard; try finding testing data for fastener installations with countersink fillers. I've found exactly (1) document [which I've lost]... that only relates to csk fillers that were installed with rivets: NO data available data for csk fillers used with bolts, screws or Hi-Loks, etc... "trust me".

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[WKTaylor]

EDBGTR...

IF You have access... this might be useful...

BAC 5430 FABRICATION AND INSTALLATION OF RESIN BONDED LAMINATED SHIMS AND SOLID FILLERS

or equivalent process standerd from any of the OEMs....

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[edbgtr]

Will…
Thank you very much for your responses, they are much appreciated.

You are right, proprietary data is about all one can hope for on this matter. It’s too trivial for anyone at a “seat of higher learning” to tackle, and companies don’t always have the budget for such testing. My methods group head agrees that this is a gap in their methodology guidelines, but cannot get the funds to perform in-house testing.

My interest in this arose because of a recent MRB requirement to “move” an item, using a shim strip, in order to recover a very high-cost item. To substantiate the affected joint we were forced to use the commonly employed so-called Peery method, based on Peery’s 1950 book. As you probably know, it is based on the guided cantilever action of the pin. This assumes that the thin sheet in the joint, joined by a pin diameter larger than the sheet thickness, can clamp the pin to such an extent as to create a zero rotation in the pin at the centerlines of the sheets. Anyone who has done any lab testing on cantilevers understands what encastré means in terms of the rotational stiffness required to achieve this condition.

This didn’t seem plausible to me, so I embarked on a reverse engineering exercise on a set of protruding head Ti Hi-lok zero-shim joint strength data. The data-set covers -5 thru -12 fastener diameters in commonly-used 2000 and 7000 series alloys of standard thicknesses. In my analysis, the joint failure was broken down into the constant stress ratio (Ra) and the bend-bearing ratio (Rb) using the allowable yield and ultimate bearing stresses for e/D = 2 joints. The calculation is done per the method used for single pin offset lug bearings from the “Stress Analysis Manual”, NTIS No. AD-759199. No fastener head/collar clamping per Bill McCombs’ Bruhn Supplement was considered.

As would be expected, the resultant bend-bearing component of the failure data indicated that even for thicker joint plates, i.e. t/D = 0.9, the maximum zero-shim clamping is in the region of 10 to 15% of the Peery assumption, and is terminated by the shear strength of the pin.

This indicates to me that the fully encastré assumption of Peery is not borne out by the joint strength test data.

It was with this background that I appealed for some test data to confirm or deny that the Liaison Engineer’s Handbook that I have in my possession is over conservative.

I shall try to get the documents you suggest.

Regards,
Ed

 

[Sparweb]

Hi Ed,

Going back to your original question, I did wonder if there is a specific shortcoming of the Bruhn text that you are trying to address. I suppose there are several limitations, particularly where author admits "the above procedure is approximate..." .

Bill McCombs may have revisited the topic in his supplementary notes (but I didn't have to tell you that).

Is it just the fact that the method is too conservative, and in some situations the extra number/size/type/etc of fasteners can't be fit in? Or are there other issues that would make the calculation less conservative, such as inadequate clamp-up, that you want to explore?


Steven Fahey, CET

 

[edbgtr]

Steve

I did consult the Bruhn text, along with all the other books and data I have accumulated over the years, and precious little seems to have been done and/or published either in the public domain or in company documentation. Using the conservative “Peery Method” during the design phase of a project will allow the original design to be modified to accommodate the conservatism, but for the maintenance mechanic or the MRB engineer, this could cause high-value items to be sent to the scrap bin. IMHO, this is not acceptable if the analysis tool does not reflect the reality of the situation. The analysis tool becomes a useless money, time and resource waster.

I think Bill McCombs realised that the clamping issue was a problem and thus attempted to quantify it. For those with access to the company data he was responsible for, you may find further information there. Be aware there is a bug in one of the formulas in the Supplement. In the second Mo formula (p 77 of the Supplement) the tf term in the brackets (parenthesis) should be tf/Vo. The basic setup of the problem, assuming the fastener head/nut takes all the end moment is correct. Some texts suggest shear plus bending interaction of the bolt as a possible failure mode for the joint. I have not explored this. My thoughts were centred on the 2% hole deformation issue. Perhaps there are other joint failure modes to be considered depending on the shim thickness to fastener diameter ratio. Certainly the little company data that I have seen, recommends a limitation be set on a percentage of the shim thickness to diameter ratio in order to avoid excess bolt bending and subsequent joint strength reduction.

I am also led to believe that acceptable shim thickness to fastener diameter guidelines can be found in the B737-300 SRM.

Ed.

 

[verymadmac]

the 737-300 SRM (51-30-01) allows floating packers for a thickness less than

21% fastener diameter for rivets
34%D shear head steel & titanium bolts, hi loks, lockbolts
41%D Tension head steel & titanium bolts, hi loks, lockbolts

I understand the numbers are lower in the 777 SRM.

 

[YoungTurk]

Do any of you using the Bruhn / McComb D3.5 formula to calculate the allowable shear load with a shim (Vo) have an opinion on how, or whether, to go about using this for a standard riveted joint?

The Vo formula requires, as an input, an allowable bending moment Mb for the fastener shank. Mb values are given in A45 for steel bolts #10 or larger (as well as several other sources), but I have not seen any similar table of Mb values for standard rivet alloys and sizes. It is possible to calculate a value of Mb for a round section given an Ftu, and a the McComb calculation gives a reasonable looking result for Vo, but without that test data...