Tension Rod End Plate Design

[chideb]

I have not done calcs in many years and was wondering what I need to check for when designing a bolted end plate for a tension rod (used as bracing for a guy tower). This rod is welded to an end plate (12" long) and the end plate will also have 2-3/4" A325 bolts (spaced 2 3/4" along the line of force) to be bolted to an existing gusset. The rod is 1.25" dia for a 26K load and the plate thickness is 3/4". I need to determine the width of the plate required and the welds for the rod to the plate. The plate on one end will be cut out to accommodate the rod diameter. These were called "hog rods" in tower hand language. I would appreciate any help/suggestions.

 

[OCI]

Do you have an AISC manual? The manual has tables and specifications for these design procedures. Your going to have to look at yeilding of the plate, block shear, bolt shear, weld type and size, ect. The 2.75" bolts sound pretty extreme for 26K to me. But hey I'm still learning too so don't take me too seriously. Good luck.

 

[ishvaaag]

You need to follow the path of loads. First from the rod to rod tip end plate; 4 aligned welds for 26 kips force. Normally the general geometrical conditions will allow on that such end plate takes well in shear the 26 kips; yet check the thickness. Then we need to know more; for I can't imagine a rod centered on an end plate be concentrically bolted to a gusset plate without a slit in the gusset. So the 26 kips force will be enforcing maybe eccentricity in two planes; the eccentricity caused by the distance between the planes of the gusset and tip end plate sometimes might, but better nor, be disregarded; and whatever eccentricity the rod has respect panel point must be accounted for the design of the bolts. If there's not such eccentricity and you are warranted by the gusset designer it can take your load, you will have merely to pass 26 kips in shear between end tip plate and gusset, with the ordinary checks, shear at the interface, block shear tearouts, bolt bearing on both plates, tensile failure on net section etc.

You always may model the thing in a 3D model like visual nastran 3D or ansys if things get overy complicated, yet better not get there.

 

[chideb]

The rod will be welded to the end plate approximately 4" from the end of the 12" long plate and from the end of the rod to the first hole it is about 3". This allows the rod to clear the existing gusset and only the bolted connection portion will have full bearing on the gusset. The bolts are 3/4".

 

[ishvaaag]

Since it seems all follows trhe geometrical axes, all above applies with th simplification of there not being in plane moments there, and those out of plane being typically small. So check your bolts capacity by AISC and give a look to the gusset welds to see if they appear undersized.

 

[BAretired]

It would be better to provide a sketch to avoid misinterpretation.

BA

 

[Lion06]

I don't think that (2) 3/4" A325 bolts are going to cut it for a 26k service shear unless you have them in double shear.

 

[OCI]

I just realized he said 2 - 3/4" bolts, for some reason I saw 2 3/4" bolts. Shouldn't this connection be detailed for ductile yeilding of the brace or at least over strength? 2-3/4" bolts seems pretty weak for a connection to a 1.25" tension rod.

 

[asixth]

Design includes everything mentioned above, rod tension, weld shear, bolt shear etc. I think it will be governed by the weld, I have not done the numbers for your example but I would provide at least 16" of weld of the rod to the end-plate.

 

[nutte]

You always may model the thing in a 3D model like visual nastran 3D or ansys if things get overy complicated, yet better not get there.

3D model?!? These are calculations to be done with pencil and paper in about 5 minutes, with 4 of those for sketching the connection and filling out the title block.

 

[slickdeals]

Well said.

 

[ishvaaag]

When things become complicated one always has such resource. I don't think out of plane effects may always be properly ascertained, even if small, and usually details are quite tipyfied, one might get interesting insight by doing such models, and one might even soundly criticize standard models for such typifications by modeling. So less horror and more appreciation of good tools.

 

[nutte]

I agree with the benefit of such tools, but this connection is not complicated. Sure, you could analyze it to death, but why? We have years of history of succesful performance using simple, "typified" procedures. You can analyze this connection by hand before your computer even boots up.