bolt tension question

[bf109g]

Hi all. First time here. I'm reviewing an analysis which determines bolt tension load and references Lockheed Stress Memo 79c dated 08-15-73. Problem is that I don't have the reference and can't follow what was done.

Anybody know where one of these may be lying around unattended?

Thanks.

 

[jetmaker]

Request the document from the people who did the analysis. They should be able to furnish it for you.

jetmaker

 

[Sparweb]

You could also do your own calculation - to double-check.
Try Mil-Handbook-60 (available from ASSIST website)


Steven Fahey, CET

 

[bf109g]

OK. I managed to scrounge a complete copy of the July 2000 Stress Memo Manual. Bouncing my reference of Memo 79c 8/15/73 off the copy I have, it would appear that the current Memo 79 has absolutely nothing to do with the subject at hand. It looks like Memo 65 has what I need.

From a historical standpoint, was there a major rewrite of the manual where memo numbers changed? In this kind of format, I would assume that the number would remain the same.

 

[rb1957]

i don't think lockheed ever renumbered their SMs ... that'd be very too confusing. OTTOMH i can't remember what 79 or 65 deal with, nor how they could relate to tension bolts. is the reference included in a reference section (where it might be left over from something else, and not really used) or closely linked to a particular part of the analysis ?

 

[RPstress]

79 has a bunch of detail stress methods for preloaded bolts; it also references 54 which has fastener allowables. 65 is more general waffle about joints with no formulae.

I'm not aware of Lockheed doing anything weird like a wholesale renumbering exercise; as rb says, it would be insanely confusing. However, my Lockheed data is from the '50s and '60s...not that I got it first hand!

 

[bf109g]

Here's the meat of my problem. The joint was analyzed as a tension fitting joint using the formulae in SM 65 (or 79, depending on the version you have).

I have an eccentrically loaded joint putting shear and tension due to moment into the bolt. And I'm failing the bolt with the combined tension and shear loads, including preload.

Is there something I'm missing here? I don't think the tension fitting analysis as presented in my legacy document is applicable. I believe I'm right, but this is not going to be a popular revelation.

 

[rb1957]

bathtub fittings are SM88

"combined tension load and preload" ... i don't think that happens (unless you're applying a small %age of the applied tension load). if the joint gaps the the bolt reactions all of the applied tension. if the joint doesn't gap (why not at ultimate?) then the load in the bolt is the preload + a portion of the applied load

 

[bf109g]

rb,

Thanks for the response. I guess the crux of my question is whether or not the method you describe is applicable to anything but a tension fitting. I don't have a tension fitting. It's an eccentric load which results in tension of the bolt, ie the load vector and bolt centerline are perpendicular.

One of my predecessors analyzed it using the tension fitting bolt load methodology in the SM.

 

[rb1957]

"perpendicular"? ... in bath-tubs the load is parallel to the bolt.

the main point of my post was how are you combining pre-load with applied load. most joint would be assumed to be gapped at ultimate so that preload is irrelevent.

 

[bf109g]

Again, I don't have a bathtub type fitting. Someone was performing engineering malpractice and analyzed it as such.

I have an applied up-load on a horizontal arm that reacts against a vertical leg. There are two bolts going through the leg, attaching the arm. The up load results in a tension and shear on the bolts. I am adding the preload to the induced tension load to get a total tension load. Then I use the tension/shear interaction equation and end up above the curve (BDM 6615).

The only solution I see at this point is to change from a 125 ksi bolt to a 160 ksi bolt. The 160 ksi bolt of the same diameter puts me below the curve.

 

[rb1957]

ok, no bathtub ...

you've got an upload with applies a direct shear and a monet to your bolts, the direct load is reacted in shear, and the monet as a couple (one bolt in tension).

basic question ... are you trying to show it good, or not ? it sounds like you're in some sort od dispute with the analyst.

how are you reacting the moment ? the most obvious way is a couple on the bolts. more reasonable (in an ultimate situation) would be to "heel-and-toe" about the furtherest corner (longer arm, smaller couple).

i'm still stuck on combining the preload with the applied tension ... are you sure the joint is not gapping at ultimate load ? if it is, then the preload has dissappeared and the load in the bolt is the applied load. if it hasn't, does the preload need to be that high ? maybe you could reduce the preload (check no gapping at limit load with minimum preload).

 

[jetmaker]

I agree with rb1957. There is no need to include the preload at ultimate, but you do need to include it at limit load conditions.

Now with regards to the preload, it is what it is, but the additive load due to the externally applied force is not the full value as would be calculated. It is a percentage of the value. This incremental value is a function of the joint stiffness, with a stiffer bolt in a soft joint having a greater increment for a given external load.

The basic concept of why the bolt does not take a direct superposition of the preload + external load is that the compressive force acting normal to the faying surfaces of the joint must be relieved first. To do this, the faying surfaces must be separated by the distance that they are squished by the compressive load. It is this distance that the bolt must elongate then in order to allow the gapping, and thus causes additional strain in the bolt. This strain can then be converted to a force and that force, plus the preload is the total on the bolt.

Hope that makes some sense.

jetmaker

 

[bf109g]

rb/jetmaker,

Thank you so very much for your posts. Bolt tension applications are kind of new to me as most of my experience is in sheetmetal and shear type fasteners.

This analysis is for an ultimate load condition. I already had the "heel and toe" thing figured out and it was just the preload/gapping issue I was having problems with. Jetmaker's explanation and a re-read of the SM helped with that tremendously.

Bottom line is that the previous analyst considered this bolt to be loaded in pure tension and analyzed it as such using the Lockheed SM. Truly an instance of looking at a picture of a bolt going through two thick pieces of stock and thinking that the associated analysis method was appropriate. If you look three pages before the tension bolt analysis in SM 65d, section VI. A. Preloaded Bolts, it clearly states that shear loads will be combined only with external tension loads in the analysis of the bolt.

Throwing out preload, the 125 ksi fastener is good for this shear and tension interaction per BDM 6615.

 

[rb1957]

please be carefull with that statement "throwing out preload" ... you need to check that the joint gaps at ultimate.

simplistically, the load in the bolt "can" be taken as the preload up untill the applied load = preload, after that it is the applied load. more realistically the load in the bolt increases slightly for loads less than pre-load based on the joint stiffness ... i think ESDU have a good write up on this.

 

[bf109g]

rb,

Thanks again. It does gap at ultimate.