Plate/Column Flange Thickness for End Plate Moment Connection 3

[JAX91]

I am performing a four bolt end plate moment connection design per AISC Design Guide 4. The design guide uses yield line theory to size the plate and column flange thickness. Reading through the text at the beginning, the design method tries to eliminate prying action by sizing the plate and column flange thick enough so that the stress does not exceed 90 percent of its capacity. The method calculates a “no prying moment” based off the strength of the connecting bolts. This “no prying moment” is then utilized to size the plate and the column flange.

I am having some issues with the design. The first step in the design examples is to calculate a required bolt diameter. In my case, the required bolt diameter is 0.545”. It is standard practice not to use anything less than a 3/4” diameter bolt, so that is what I would like to use. If I follow the design example, I would then use the strength of the 3/4” diameter bolt to calculate the “no prying moment” and then size my plate and column flange accordingly.

I have a hard time seeing why we should use the actual bolt diameter to calculate the “no prying moment” instead of using the required bolt diameter. By using the actual bolt diameter, the need for stiffeners is determined by the bolt size used instead of by the actual load. If I use 3/4" diameter bolts in my example, my column flange cannot handle the load and stiffeners would be required. A 5/8” diameter bolt would be adequate to handle the load. If I use 5/8” diameter bolts, my column flange is adequate and no stiffeners are required. This solution does not seem practical to me. Why should I have to use stiffeners just because I use a larger bolt diameter. Am I missing something? Does anyone else see an issue with using the required bolt diameter to calculate the “no prying moment” instead of the actual bolt diameter? If I do so using the same equations, I am still ensuring that the plate stress will not exceed 90 percent of its capacity, which follows the theory on which the whole design is based.

 

[ToadJones]

is the procedure the same in DG16?

 

[JAX91]

No. Design Guide 16 uses a different procedure. Design Guide 16 sizes the column flange utilizing an equation similar to that of the local flange bending equation in the steel code. For an end plate moment connection, the 6.25 from the local flange bending equation is modified to a value calculated based on the connection parameters. Design Guide 4 is a newer method and translates easier to ASD design, which is why I am using it. Some of the testing performed to come up with equations for Design Guide 16 was only done on A36 steel. They recommend using an Fy of 36 ksi for some of the checks, even if you are using 50 ksi steel.

 

[connectegr]

Your interpretation of the procedure is correct. We have had the same discussion in our office. We are currently designing hundreds of these connections for a large project and we find the column reinforcement to be unrealistic at times. In a few cases we have "cheated" and completed the calculations using a smaller bolt diameter, and substituting the standard bolt size in the detail.

I have emailed Dr. Murray for his comments. I will let you know his response.

http://www.FerrellEngineering.com

 

[ToadJones]

Is DG4 strictly for seismic and wind?

 

[connectegr]

No we use both DG 4 and DG 16 for endplate moment connections regardless of the source of the loading. When we have excessive column flange reinforcement required, we try the DG 16 method.

http://www.FerrellEngineering.com

 

[connectegr]

According to Dr. Murray, use of the actual bolt size was an oversight. Using the actual bolt force is acceptable. They are planning a MSC article on this and other common questions in the near future.

http://www.FerrellEngineering.com

 

[delagina]

this end plate moment connection is very common in o&g specially for pipe rack. thanks for the info.

 

[ToadJones]

As usual, nice input from Connect.

I have told you before Connect, your dedication to connection design is impressive and you input here invaluable.

 

[connectegr]

Delagina
That is exactly where we a using them. CISC already uses the actual force for the connection design, so we were already using it with the DG 4 procedure.

http://www.FerrellEngineering.com

 

[delagina]

@connectegr,

how come the 9th edition ASD doesnt have prying check for end plate moment connection? how come they are considering this now in 13th edition?

 

[connectegr]

DG 16 does check prying in the column flange and end plate. Designing for bending in the 9th Ed ASD was conservative. DG 4 has a pretty good explanation of the reasoning behind their procedure.

http://www.FerrellEngineering.com

 

[JAX91]

My connection is for an O&G pipe rack as well. Considering the number of columns, avoiding stiffeners wherever possible can be a big cost savings. I think I am going to proceed with connectegr's recommendation of using the actual bolt force. Thank you all for your input.

 

[delagina]

stiffeners for end plate should be avoided not only for cost but it can interfere with pipes running across.

 

[JAX91]

Delagina,
Out of curiosity, what size beams do you typically use for your racks? I am new to oil and gas, but it appears that W8 beams/columns are fairly standard. I am currently working on a job at which a W8 beam and W8 column would require both stiffeners and web doublers. To avoid using them, I designed a W10 beam with a W10 column. The W10 beam reduces the flange force, and the W10 column gives a larger capacity for web panel zone shear. This combination drew some criticism from the owner and contractor.

 

[ToadJones]

Show them a fab estimate for web doublers and stiffeners and see what they criticize!

 

[delagina]

@jax, i dont know what width of piperack you have. I usually use at minimum W10 upto W18 beams for 20 ft width piperack.

It's weird you are getting criticism from owner for beam or column sizes. You are probably involved in some "small" retrofit jobs inside US.

 

[delagina]

btw when i said stiffeners should be avoided, i was referring to 8-bolt stiffened extended end plate not the stiffeners welded to columns.

 

[JAX91]

I was wondering how stiffeners in the web would get in the way of the piping. The stiffened end plate connection explanation makes much more sense. I try to avoid those as much as possible as well. Thank you for your input! I feel much better about using the W10 beams.