Tension fitting analysis with filler 3

[Trajano]

Hello,

I am updating an old calculation with new loads.
It is a channel type tension fitting analysis based on Lockheed 88a. In the old calculation, in order to improve the RF, a filler was added.
The way the RF was calculated was using an *equivalent end pad thickness*:
t_eq = SQRT (t_pad^2 + E_fill / E_pad * t_fill^2)

I am trying to find out where this formula comes from, but my approachs lead to a different formula.
I consider the end pad as a simply supported beam, and the filler as another beam on top of the end pad. Two steps:

FIRST: I get the filler equiv. thickness (t_fill') as if it was made of the end pad material: assumming same deflection (E_fill*I_fill = E_pad*I_fill') I get the t_fill' = f (t_pad, t_fill, E_pad, E_fill).

SECOND: I get the end pad equiv. thickness, as the thickness (t_pad') of a single beam whose max. stress was the same as the max. stress of the original end pad beam.

K*P/[t_pad(t_pad+t_filler')] = K*P/t_pad'^2

Finally I get t_pad' , very different from the t_eq of the formula of the old calculation.

Is my approach wrong? Am I missing something?

Thanks in advance & regards

 

[rb1957]

that's a bit generous (= unconservative ?), counting the filler as effective as the endpad.

max endpad bending is under the tension load, reacted at the sides, no? doesn't sound that a filler would react this bending, unless it was tied into the sides.

 

[Trajano]

Hi rb1957, thanky ou for your answer.
Actually what you say was my first impression too. But then I found in Flabel (Chapter 5, page 321 and 322): "For higher loads applications, a tension clip can be safely beefed up by using a radius block..." (it doesn't say how the filler is helping though)
And after thinking more carefully about it, I believe it works in some way as a bending spring. In order to bend the filler it is necessary some energy, i.e. part of the applied load is used to bend it. The question is how much. Then I made the approach I described before, based on the hypothesis that both, the end pad and the filler, become deformed in the same way.

 

[rb1957]

ok, i'd buy that ...

i agree with your OP, about the endpad being a SS beam in bending (maybe fixed beam ?). I also see your story about bending the filler, it could react some of the tension load, but i think it'd need to react this as a distributed load onto the endpad (reducing the moment in the endpad).

so say the filler reacts x% of the tension load, and reacts this (against the endpad) as a UDL. Then the max moment in the filler is Pf*w/8, and the max moment in the endpad is bigger than (100-x)% of SM88a (as you have to account for the UDL from the filler.

thought ... if the endpad is dbly fixed (pretty sure SM88a assumes SS) then this'll change the displaced shape of the endpad.

 

[edbgtr]

Trajano

It is a well known fact that a thickish steel washer also improves the performance of an Alu end-pad.

For under-strength tension clips I use what the Brits call a Throat Plate, which is equivalent to the US Filler. Find attached a method for evaluating the contribution of the TP to the T-C.

You might be able to use something like this to provide reinforcemnent between the bolt and the 3 walls of the B-T fitting.

Ed.

 

[epeus]

Ed, thats a handy calculation for clips and flanges which are on the marginal side, cheers! Can you advise on which design text/manual that attached file is extracted from?

 

[edbgtr]

epeus
The item is extracted from a set of design/structural integrity notes I have accumulated over a number of years. The throat plate derivation I "picked up" while working with some Brit contractors.
I add to the notes as, and when, I acquire new data/methods or discover incorrect use of methods being applied. I then write up the misunderstanding along with a "better" (IMHO) explanation for the method and/or theory.
Cheers.
Ed.

 

[rb1957]

ok, P is reacted by the flange PF and the throat washer PT.

I see the throat washer being loaded by some distributed load, leaving PF in the flange. Given that, i'd've thought that the moment in the throat washer would be PT*(L/2) and the moment in the flange as P*L-PT*L/2.

 

[Trajano]

Firts of all, I thank you all for answering.

Please, let me sumarise what I have understood about the methods exposed here (see attached file). Please, correct me if I am wrong.

- edbgtr suggests a cantilever model with concentrated loads (model A in the attached sketch)

- rb1957 proposes a (cantilever?) model with an uniformly distributed load (model B). This is working as a washer would do (true?).

And I would even propose a variant (B') of rb1957's one in which the distribution would be triangular given that is is a prying action.

While model A is reducing the total applied load on the end pad from P to PF, model B (and B') keeps the total applied load P but reduces the arm of the moment.

And I would ask you to consider if depending on the relative stiffness of the end pad and the filler, it could be more accurate one model or the other.

 

[edbgtr]

Hi All

My apologies for not getting back to you sooner, but I had to prepare this response in between my paying work.

The attachment should be explanation enough for what I did after some questions were posed here regarding the working of the throat plate. I have attempted to clarify how I see it working and perhaps take it a little further than we had done before. The only BTF end-pad throat plate reinforcement I’ve done before was for an FTI rack, where the client “forgot” to tell us about the mass increase of the equipment the rack was supposed to support. We had angle BTFs tied to the floor of the cabin that turned out to be critical because of the mass increase, so I inserted some appropriate TPs at all four feet. Fortunately we didn’t have to deal with a “sudden arrival”, so they weren’t tested to the full. Examining the formulae, you’ll see that it pays to have TPs stiffer than the original end-pad you’re reinforcing, but you guessed that already anyway.

If you find any bugs in what I have done, please let me know.

Cheers,

Ed.