Bolts Fixed or Pinned

[jsant012]

Hi! I'm designing a cover that will be held down by several bolts onto the housing. I want to measure the deflection of the aluminum cover after the bolts are torqued. To simplify, I'm trying to use the beam deflection equation between 2 bolts, but confused if I should use treat the bolts as pinned or fixed. Advice?

 

[msquared48]

Sounds more fixed than pinned to me.

Mike McCann
MMC Engineering

 

[rb1957]

what deflection are you trying to measure ? as in "I want to measure the deflection of the aluminum cover after the bolts are torqued."

the "only" deflection i can see is that the cover will be pushed up against the support ?

 

[desertfox]

Any chance of a sketch?

desertfox

 

[TXStructural]

If this is correct: you have a plate which is bolted to some fixed surface, and you are looking at the deflection of the plate in bending along the plane parallel to the axis of the bolts.

Then: Locally, each bolt provides clamping force sufficient to be considered fixed for through-the-thickness deflection.

If the cover is longer than it is wide, short of a full-blown FEA, I'd take a strip across the middle of the cover, apply the load and consider the ends fixed for flexure, but unrestrained for axial until the bolts go into bearing (deflection sufficient to cause bearing of the bolt holes on the bolts, thus limiting deflection like a wire pulled tight.)

 

[rb1957]

why will the cover deflect (distort) "just" from torquing the bolts ? the distorsion i see would be due to the surfaces not beng perfectly flat, which is independent of the load.

 

[jsant012]

The cover is pretty thin (less than .070 inches). A gasket will be used between the cover and the housing, and it will have a compression that is equivalent to 580 psi. Just imagine that it's a rectangular cover that will be on top of a rectangular box/housing. I want to determine how many bolts and how far apart they need to be.

 

[rb1957]

so you're looking at the deflection of the edge, and you need to maintain a +ve pressure, of something like 600psi.

a 0.07" thk cover probably won't support much bending as a plate.

 

[jsant012]

Correction: The applied uniform loading will actually be 76 lbs. I wanted to see how much the cover would deflect/bend. Don't want big deflection because it will affect the signal.

 

[TXStructural]

As long as you are below the yield strength of the plate, you may be able to use simple structural formulas for the deflection, but this kind of very thin material probably requires a different method of assessment. But I suspect that if your very thin plate is very big, you may well reach yield locally near the bolts, and the cover could be a tension element.

Is the load a uniform load from pressure, or from something which might be applied over a relatively small area compared to the cover? As someone said earlier, a picture is worth a thousand guesses.

Look here for a similar discussion (found via Google):

http://www.physicsforums.com/showthread.php?t=260145

 

[dhengr]

Rb1957:

Please tell me what +ve and -ve mean, using real full words, not more shorthand. You use these terms often, and I’m not absolutely sure what they mean or where they come from.

Jsant012:

Provide some sketches with dimensions, loads, etc., so we can see what you are talking about, and also to make you think about how you have explained your problem so far. You’re looking at it, but we can’t see it from here. What the devil does “compression that is equivalent to 580 psi” mean to everyone other than you? What the devil does “The applied uniform loading will actually be 76 lbs.” mean, and how and where is it applied? You “Don't want big deflection because it will affect the signal.” but you are very likely to get big deflections with a cover sheet of aluminum .07" thick, spanning any distance. Finally, a gasket probably won’t work with that thickness sheet no matter how many bolts you use. If your housing has a fairly thick top flange to which you will be bolting, then make another flange the same size and thickness and sandwich the gasket and .07 alum. btwn. the two flanges. Then you adjust your bolt size and spacing and the flange thickness so that the gasket you are using will work. Depending upon the clamping action of the bolts and flanges, along with the interaction of the gasket used, you will have a semi-fixed boundary on your cover plate, supported on all four sides. I would check plate defections assuming simple supported boundaries; and calculate plate stresses assuming fixed edges for max. moments there. The dimensions of the cover plate are very important in the deflection calcs. You might do all your calcs. with both boundary conditions to put some bounds on the deflections and stresses, and then pick something near the max. values to actually do your design.

Right now, you are sending really bad signals, irrespective of the cover plate deflections.

 

[BAretired]

Provide a sketch and a clear description of the problem.

BA

 

[ToadJones]

I am 90% certain that:

+ve = positive
-ve = negative

Took me sometime to decipher as well.
These are common short hand terms in FEA software output....equally as annoying when used there too!

 

[rb1957]

c'mon ... of course +ve = positive
in context, positive pressure means you don't want the cover ro gap, you want it to be pushing against the body with a pressure that won't be overcome by the applied pressure (which i assume is pushing the cover off the body). the problem is more about distributing the preload from the bolts as a pressure along the edge of the cover.

at the OP, similar to above comments ... i thought i understood 580psi, but maybe i didn't, but i have a hard time cnverting that to a load of 76lbs, unless we're dealing with an area less that 0.1in2.

the way i'd look at it would be to represent the applied pressure load on the cover as a running load around the bolt PCD (=p*r lbs/in) then try to react this as bolt preloads and check what this means to bending of the cover (at mid-pitch, between the bolts). i see the seal under the cover being compressed more under the bolt, and the seal is acting as an elastic foundation to the cover, creating a distributed reaction to the bolt preload, causing bending in the cover.

 

[dhengr]

Wow, who’da ever thunk it.... thanks Toad, maybe so, maybe so, that certainly is an improvement to our lexicon and shorthand. Anything which confuses our communications and uses words and terms which everyone doesn’t understands is a vast improvement. + used to mean positive without the ve, so did pos., and - used to mean negative without the ve, so did neg. Now I realize that if I ever want to say positiver or much more positiver than something, I can use this neat shorthand; +verts = positiver than something, and mm+vert = much more positiver than, not to be confused with millimeters + vertical something .

All the same, 100% is mm+vert only 90% certain.

Pretty soon we will see whole OP’s and posts written with only acronyms and clever letter groups, in which only the OP’er will understand what he/she thought they said, irrespective of what the readers get out of it. After four hours of Googling letter groups, for some meaning, about a problem that is ill defined in the first place, we will be so pi$$ed off as to lose interest in the thread. You’ve got people wasting hours of their company’s time per week on these forums, and to sound important and in the know (re: the latest phrase-eological, term-o-graphic, obfuscational fad) and to save a second and three electrons in transmission; and to make their meaning all the lesser clearer, they’ll use some term like DMWYSCTOTMFYLOT, which stands for ‘don’t make what you say clear to others, they might follow your lack of thinking.’

Thanks Rb1957.... I do understand that + means positive, but what the hell does the ‘ve’ mean and what does it add to the clarity? I’m pullin your leg a little bit, but I really didn’t know what +ve meant in the many ways you’ve used it. Now I know and my life is ‘much fuller for it’ (MFFI).

 

[apsix]

+/- only mean positive/negative when in front of a number, otherwise it's plus/minus (or plus/hyphen).