Flange with fasteners design (Am I on the right track?) 1

[rofloligist]

I am trying to design a bolted flange connection for an oddly shaped piece. There are 4 flanges that connect the piece to the ground, with the flanges being welded to the ground. Each flange has 2 bolts. What I have done is taken the load, divided it by the number of fasteners and applied the load as a shear force at each bolt. This leaves the secondary shear due to moment loads.

As this force is not acting along a line of symmetry with the fasteners, this is a case of eccentric loading. My understanding is that you take the shear load experienced by each flange (so Fshear_flange=F/4) and place it between the 2 bolts on each flange. Then you take the equivalent moment this Fshear_flange would create if it was acting at the location of F. This means you get 2 moments at the centroid of each of the bolt groups on each of the flanges. One moment is Mh=H*Fshear_flange and the other is Mw=(W/2)*Fshear_flange.

After you get these moments and loads, you distribute the Fshear_flange to the 2 bolts so Fshear_bolt1=Fshear_flange/2. Then you get an additional shear load, Fshear_bolt2, from the moment created by Mh. Fshear_bolt2 = Mh*(distance from center of bolt group to bolt). You then add these 2 components via Pythagorean theorem and get your resulting Fshear_bolt. The other moment, Mw, creates a tension in one bolt and compression on the other bolt on that flange. This tension/compression is equal to Mw*distance from center of bolt group to bolt.

I am just wondering if my initial shear vector is placed properly, between each of the 2 bolts, and I am taking the correct moments. Should the moments be taken about the center of each of the 4 flanges? Or should the moments be taken about centroid of the whole connection, IE midway between 2 and 4 and in the same plane as the bolts.

If this is too confusing let me know!

Thanks for the help!

 

[paddingtongreen]

Confusing it is.

Consider the shear and the moments separately. the shear is divided into four anchor plates (your "flanges"). The moment is HxF. This is divided by 2 sides and divided by L. The vertical load on each pair of bolts is ± H_xF/(2_xL).

You find the resultants and divide by two bolts. Then apply the resultants to designing the welds.

What worries me is the risk of having to ream the holes for fit-up. If the whole thing is bolted before being welded to "ground", okay, but if the anchor plates are to be welded prior to delivery of the package, some adjustment may be called for. I like to have some spare capacity built in to take care in case not all bolts can be in bearing at the same time.

Michael.
"Science adjusts its views based on what's observed. Faith is the denial of observation so that belief can be preserved." ~ Tim Minchin

 

[desertfox]

Hi

Take a look at this site, divide your total force in half and calculate as per the heading "strength of bolts withstanding torsional shear"

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

 

[Tmoose]

Is the load static, continuous, or used evry few months to haul treasure chests up a cliff thus subjected to the variable torque of a gasoline powered winch or the unsteady pull of a team of pirates hauling on a block and tackle, or etc ?
Will the main component be removed annually, daily, never?
If never, why not weld the the thing directly to the floor/sole plate ??

I'd design the connections to rely on frictional clamping with a big "safety factor" so -
- hole placement and bolt fit-up are no issue.
- All bolts can be counted on to do their part.
- it could be assembled by mortals on a desert island with torque wrenches and I would not have to think for 5 seconds about the variation in clamping force vs torque.
- the torque would be selected to first provide clamping force far in excess of what would be required to first bring the faces of the brackets welded to the ground and fabricated brackets into full contact despite 1/4 inch or more separation or twist, and THEN provide clamping force far in excess of the friction requirement. I'd modify the welded fabrication to allow harmless compliance of the mating faces. I'd provide a fixture to place the ground brackets during welding (if the actual bracket is not available to serve as a fixture)
- The bolts would be large enough diameter so the shear strength of one per flange was plenty, in case a completely unexpected 100 year overload exceeded the frictional clamping and the 3 or 4 bolt shoulders that could were forced into contact and become dowels.
- The bolts would have solid shoulders in the plane where the ground bracket meets the fabricated bracket.
- The mating faces of the lugs would be smooth and flat and free of paint.
- I would forbid lockwashers from the jobsite. Failsafe fastener retention would be mechanical via prevailing torque nuts or hairpin clips, safety wire, etc, etc,etc.
- I would increase the fastener grip length to at least 4 diameters using standoffs or tall spacers with square smooth ends.
- If i was really worried about it I would provide a cornered step in the fabricated brackets lugs to to resist horizontal motion or downward motion of the right hand pair of lugs.
- I'd consider making one long bracket per side with 4 carefully placed holes rather than 2 individual brackets requiring fussy 3D placement. They still will try to tip outward/inward depending on weld sequence used.