classic stressing - help me please!!! 1

[bidga1]

Hi.

I've just started a stress job and have been given a bracket 5" by 3" by 0.25" thick to stress. It is attached by 8 fasteners and has a lug around an inch from one end. I've managed to use a bolt group spreadsheet to extract the load per fastener but am a bit lost as to how to analyze the fasteners. Is it a shear check on the fasteners? and a nett tension check on the plate? (If so what do I take as the W dimension). Is there something I have missed?
I know this is all basic stuff to you guys but I'm new to this and am struggling and excuse the pun "stressing" out about it.

Thanks

 

[CoryPad]

Use Google to find NASA RP1228. It shows calculations for the lug and the fastener group.

 

[bidga1]

Thank you very much. Will I need a 3d bolt analysis?

 

[CoryPad]

Very difficult to say based on what little you have shared. I will say that 3D bolt analysis is very uncommon.

 

[desertfox]

hi bidga1

A sketch of your braclet showing dimensions and load positions would help us to help you.
That said take a look at this site, its shows how to calculate stresses in bolts with a bracket, loaded in different configurations.

http://www.roymech.co.uk/Useful_Tables/Screws/Bolted_Joint.html

desertfox

 

[bidga1]

This is a rough sketch of the scenario

 

[desertfox]

Hi bidga1

thats a good sketch, which way is the load in the lug?

desertfox

 

[bidga1]

Thanks!!

4000 lb axial 2000 lb transverse 280 out of plane

Bidga1

 

[desertfox]

Hi bidga1

I have uploaded a file to show you how to calculate the bolts holding the bracket and assuming that the load through the lug acts at right angles to the bolts, the bracket is considered to be a rigid body.

desertfox

 

[bidga1]

Thanks for all your help! Would I need to do a net tension check?
Thanks,
Bidga1

 

[desertfox]

Hi bidga1

I posted that file before I saw your last post, so what I would do is find the resultant of the 4000lb and 2000lb and proceed like I have in the file, ie find the moment about the pivot and calculate tensile loads, then consider the 4000lb load as a direct shear load shared equally on each bolt, finally the out of plane force you can treat as another shear force offset, so you need the centroid of the bolt pattern and solve as shown on the Roy Mech site I gave a link for, look for the title:-

"Strength of bolts withstanding torsion generated shear loading"

The bending calculation for the bolts ie my uploaded file can also be seen on the Roy Mech site just above the title I gave you here.
Once you have all the stresses you can obtain the resultant.

desertfox

 

[desertfox]

Hi bidga1

I would just like to emphaise that my assumption of the bracket being rigid is simplistic and that if there are any prying forces due to bracket deflection then the bolt forces will probably be higher than those calculated.
Here is a link to a site dealing with prying forces:-

http://etd.ohiolink.edu/send-pdf.cgi/GAO XIAOJIANG.pdf?ucin1134401462

desertfox

 

[bidga1]

Hi Desertfox

Thanks again for your help.
One more thing, if I was to do a net tension check on the plate would I just take it across one hole. i.e. (W - d) with W being half the width of the plate?

Thanks

Bidga1

 

[desertfox]

Hi bidga1

You should check not only the tension but also the shear out of the plate and for bearing failure of the bracket material.
Going back to the tension failure, I would take the pitch of the bolts "p" and subtract the bolt diameter from it, so
Force to cause tensile failure of plate:-

F = allow tensile stress * (p-d)*t

d = bolt dia

t = plate thickness

A word of caution, the above formula is for rivets, so it might be better to subtract the bolt hole dia from the pitch of the bolt centres.
Finally a plate with holes in would normally give rise to stress raisers, so the formula above will only give the average tensile stress across the plate.

desertfox

 

[bidga1]

Hi destertfox

Thanks again. You have been very helpful and I am extremely grateful for your help

Thanks again

Bidga1

 

[desertfox]

hi bidga1

Your welcome

 

[plasgears]

Stress analysis in aero should include LCF after you obtain max combined stress. We looked at mission profile and totaled the cycles. We were able to resolve the generalized sub-cycle to zero base so that all cycles were considered from zero stress. The analysis that separated the men from the boys was considering a crack defect in the max stress location. Then it is an iterative cyclic analysis in fracture progression.

 

[rb1957]

regarding desertfox's pic ... can the supporting structure react the "pivot point" out-of-plane load ? This is the compression 1/2 of the couple reacting the off-set moment; the tension 1/2 is distributed between the bolts, presumably in a triangular pattern.

there is a similar reaction for the sideways load (2000 lbs).

with fasteners in tension, check flange bending.

 

[bidga1]

Sorry to bother you again, but a panel analysis has been carried out on the wall the bracket attaches to and I have the x, y and shear running loads for the panels. For conservatism do I add these loads to the loads I used for the bracket analysis? And carry out a net tension, shear and bearing check again

Thanks

Bidga1

 

[rb1957]

are these "wall" (= bulkhead ?) loads going into your brkt, or around it ? are the brkt fasteners also joining piece of the wall together ?

so long as the loads are not going into your brkt, they don't affect the stressing of the brkt. they could affect the stressing of the fasteners.

the fasteners appear to be reacting loads from two sources ...
1) the brkt, as discussed, and
2) the wall, presumably they are joining pieces of the wall together so there'll be loads/reactions somewhere else on the fasteners to be considered (not just the loads from the brkts). this then suggests that the fastener reactions to the brkt loads are not simple.

we seem (IMHO) to be filling in alot of background knowledge. read up on this (Bruhn, Niu, etc); brkt stressing is rocket science. you wouldn't want to appear smarter (more knowledgable) than you really are ... try it, but i think you'll get caught out in less than 1/2 hr. i think it's better to admit your short-comings, with an attitude to learn.

good luck !