Effect of galvanizing "flow" hole on beam over HSS column connection

[dauwerda]

AISC Design Guide 24 presents a connection design example of a beam over HSS column connection (Example 4.1 in section 4.5). My question is: Would there be any change in the cap plate design (thickness determined to eliminate prying) if the cap plate has a center hole cut out to allow for galvanizing, and if so, what would that be?
In the example, the HSS column is a HSS 8x8x1/4, so lets assume the fabricator wants to cut out a square hole that is 7"x7" so that the plate only extends inside the sidewall of the HSS 1/4" to allow for ease of galvanizing.
My thought is that this would affect the design in the following way:

Per the discussion in section 9 of the steel manual, the prying check is based on the assumption that the flange is in double curvature. My thought is that by adding the cutout this is no longer a valid assumption as the HSS wall will yield before the plate will (See attached sketches). If this is the case, how is this handled? Would the discussion about the asymmetric loading found in the 15th edition of the manual (below) apply in this case? If so, wouldn't the "point of application of the load" still just be the HSS wall as it carries tension induced by the moment at the joint, which wouldn't change the design?

The 15th edition of the manual includes the following discussion about the dimensions b and b': "These values are valid for tees and angles if the load is delivered symmetrically and the angle shown represents one of a pair of back-to-back angles. When the angles are not back-to-back and connected to a relatively flexible support, the effective eccentricity may be increased and a distance measured to the heel of the angle or possibly somewhat larger might be warranted. It is common to assume there is no moment transfer between the flange and the element to which it is attached, that is, all of the moment required for equilibrium of the flange is assumed to be taken at the bolt line. This discussion is not intended to be applied to asymmetrical conditions. When the load is delivered asymmetrically, an even greater moment will result. For instance, if the angle is attached to only one flange of a wide-flange member used as a hanger and the hanger is restrained from rotating about the bolt line, then the eccentricity would be measured from the centerline of the hanger or to the point of application of the load."

Or, is my overall thought here incorrect as the tension in the wall of the HSS will prevent it from bending and the original design per the example is still valid? Or does some other failure mechanism come into play (that is not addressed in the design guide) and need to be looked at?

I apologize for the long post but would greatly appreciate any thoughts on this subject.
Thanks in advance!

 

[CANPRO]

Is anyone cutting holes that large for galv? 1" diameter would be plenty. See link for some good insight into the galvanizing process. Another practical consideration - your beam is only 6" wide, so if you cut a 7"x7" hole you're going to have your column open at the top 1/2" each side.

All that aside - I agree you lose your double curvature in the plate. I think the answer really comes down to the ratio of the HSS thickness to the cap plate thickness. Extreme case - say your cap plate is 3" thick and doesn't budge under load, your HSS wall will be in pure tension. The less stiff your cap plate becomes, the more bending your HSS wall will see.

 

[dauwerda]

Thanks for the response.
I agree that the beam being smaller than the column cap hole is a potential issue. I am simply using the (existing) design example to provide the concept that I am after.
As far as the galvanizing hole size is concerned, the galvanizers I have worked with say they prefer a minimum of 30% open (really it depends on the volume/length of the tube). I recently started providing engineering for a fabricator and this has been their standard detail in the past. As soon as I saw it, it raised red flags for me. Their reason for this is they like to have all 4 corners open as it allows the galvanizer to dip in any direction and it keeps them from worrying about orientation of otherwise symmetric plates (keeping single holes diagonal from each other across the tube). They cut out all of the plates on a plasma table and it is quicker and provides less wear on the equipment to pierce once and cut a large hole than to pierce 4 times and have a hole in each corner. So, that is where the large square hole comes from.

I agree with your assessment of the really thick plate. I am trying to determine what that thickness would need to be (that is, what the calculation would be to determine required plate thickness) to make the connection perform as desired. I am trying to work with the fabricator to make their preferred detail work (I am working for them) or convince them that it results in a super thick plate and they are better off providing 4 small holes in the corners.
Thanks!

 

[KootK]

My thoughts:

1) My guttural response is that the detail sucks and ought to be done differently. That said, I get the business imperative that makes it wise to say yes to clients.

2) With such a large hole in play, it's fundamentally no longer the same connection as the one in the AISC design guide example. It more resembles flanged base plate and splice connections that are encountered in monopole construction. That might be a good space in which to search for guidance.

3) I think that it's imperative that the cap plates be designed such that they can confidently be claimed to be behaving as "stiff" elements. Otherwise, you'll induce bending in your fillet weld which is a surefire recipe for premature failure.

4) If there's an engineer of record involved, usually they intend for these joints to have a certain amount of rotational stiffness in addition to bending capacity. That, even if they haven't stated a specific requirement in the design drawings. As a potential reviewing EOR, I'd definitely have some concern for the flexural stiffness of the joint configured like this. If your fabricator has been doing this successfully for a while, perhaps I'm incorrect about the potential for that aspect of things to cause problems.

 

[dauwerda]

Thanks for the comments Koot, I was hoping you would chime in as I have been reading the forum for a few years and really value your opinions and thoughts (not to mean I don't value the opinions of other forum members). Here is my responses to your comments:

1) I agree, but as you said, I at least wanted to fully explore the options for them.

2) I actually started with this approach (treating as a base plate), making the effective bend line essentially the length between the cut-out and the edge of the plate (for the example I used, this results in only a 2" long bend line, 1" on either side of the plate) which as expected results in a very thick plate. I had the thought that this *might* be conservative in the fact that the connection is clamped together with the bolts to another "stiff" element. So, even though there isn't much plate to resist bending along the face of the column, the plate would still need to bend on the bolt line (which it doesn't have to do in a typical monopole base plate that doesn't have grout). This led me to look in other places that have a similar type connection which in turn led me to design guide 24 and the cap plate connection, but this guide also doesn't address the effects that large galvanizing holes can have. That at least was my thought process, which is why I decided to post here and get other's opinions. It is looking like my initial reaction (and yours) is that the full cutout detail sucks and it should simply be avoided (just trying to do my due diligence here).
On a side note, the design presented for an end/flange plate is the same that is used for the cap plate (based on prying). Again, the problem is the design procedures don't specifically address the effect of a large galvanizing hole. The only mention of it I have found so far is in the CIDECT Design Guide 9 (available here:

https://www.aisc.org/technical-resources/cidect-design-guides/).

Where in section 11.1.1 it states, "Blank end plates, rather than 'ring' end plates, are nearly always used, unless special conditions have to be taken into account, such as if any rebar in a concrete filled column is required to be continuous through the splice joint." However, it does not state if this changes the design procedures that are presented thereafter.

3) Agree, however this detail will actually have a full pen weld for fatigue purposes.

4) This is the Utility Industry, and while the fabricator does submit shop drawings for review to the clients (utilities) there is usually no formal EOR as most utilities/states are exempt from the stamping requirements. I believe I am the first person to throw up a red flag for this particular detail.
They have done it in the past, but I am not prepared to say it has been successful as most (if not all) of the structures they have fabricated have probably not seen their design loads (and even if they have, as long as there is no catastrophic failure, the more flexible connection would probably go unnoticed). So maybe that does make it successful, but I at least would like to have calculations that show it should perform as required.

Thanks again for providing your insights and I would certainly appreciate hearing any one else's thoughts on this.

-Dan

 

[KootK]

Thanks for the kind words Dan. A little Christmas goodwill for me!

- to confirm, moment does need to be transferred through this joint right? That wasn't just a vestigial aspect of the sketches that you shared?

- if they could keep the big hole, do you think there would be an appetite for either or both of the following:

a) beam stiffeners over the HSS walls and/or;

b) stiffener brackets under the cap plate?

 

[dauwerda]

Yes, moment does indeed need to be transferred through the joint.
Stiffener brackets may be a consideration. My ultimate goal here is to present them with a few different options for the joint (that I am comfortable with) and see what they like best and base future design off of that. So far I am exploring these options:

1) Keep joint as designed, with large hole for galvanizing. --While this would be their ideal option, I am not comfortable with it as published design information does not address the effects of the hole and intuitively it seems that it would have a large a large effect on how the joint behaves. (I believe based on both CANPRO's and KootK's comments we are all in agreement that this option is a no-go).

2) Keep joint as designed but with a small hole in each corner for galvanizing. --This is my ideal option as the small holes will have no effect on the published design equations, but they (the fabricator) want me to explore other possibilities.

3) Design joint following the typical procedures for monopole baseplates with large cut out. --This is an option I am comfortable with but it does lead itself to thick plates and/or oversizing the width of the plate past the column to increase the bendline length. It seems to me that it may be overly conservative as it does not account for the interaction between the cap plate and the beam flange, i.e. bending at the bolt line.

4) Add stiffener brackets under cap plate. --This will allow for the use of a thinner plate with the large cutout as it changes the bending mechanism of the plate but it will also require more pieces and more welding during fabrication.

5) Develop a method to determine plate thickness required that falls somewhere between options 1 and 3 above. --This is what I would like to do, although it may not be practical as I have not found any published information that specifically addresses it. I may start playing with some FEM to see how the joint behaves and if my thoughts above are correct. From this I may develop my own modified method or I may determine that option 3 is a pretty good representation of the required plate thickness. (If anyone has thoughts on this approach I would like to hear them)

Depending on what comes of my option 5, I will present them with that result(which could be option 3 or something slightly less conservative) as well as options 2 and 4 and see where they want to go from there.

 

[structSU10]

What if you left the hole clear, added a plate under the beam to close the hole (welded to bottom flange), and use angles on the faces of the tube?

 

[dauwerda]

Thanks for the input structSU10. I do like the use of angles as opposed to the cap plate, however this option was also nixed by the fabricator.

Just to be clear, all of the solutions presented are options that the fabricator can do, I am just working with them to come up with an ideal (and economic) solution that we both like. They are looking for something that can be fabricated quickly and effectively. The angles are good as it removes the need to cut out a cap plate on the plasma table. They don't like them though because it requires more time to fit up and weld as opposed to a plate, as you have to get both angles aligned correctly.

I must say it has been interesting working so closely with a fabricator and really getting to see what they like or don't like and why (although I'm sure every shop is different).