Tube/WF beam double angle connection 2

[SteelPE]

I have a friend who is a steel detailer using Tekla. Recently he was able obtain a trial version of the newest RISA connection software for review because he is interested in purchasing the product as it now links directly with the Tekla software he is using.

He contacted me yesterday with some concerns regarding a connection he was looking at on one of his jobs. In this instance he had a W24x162 beam framing into the side of a HSS8x8x1/2" column. He has the beam attached to the column with 2-L4x3x1/2" angles that are 20.5" long. RISA fails this connection stating that it is in violation of the Hb/Bb requirements of table K2.2A of the AISC 14th edition. Further investigation into the problem reveals a little explanation by RISA on their design philosophy:

http://risanews.com/how-does-risaconnection-design-hss-columns-with-shear-connections/

My question is whether or not RISA is correct in their interpretation? I'm sure they worked extremely hard on this matter and did the best that they could with the information they had. I also agree that axial loads on double angles are not really addressed well in the AISC, however, now we have a connection that technically fails. What are the options given this interpretation....

-use a single tab (not a good as a double angle in my opinion)
-increase the angle and column size such that Hb/Bb ratio is greater than 0.5 (never going to fly with the EOR or the architect)
-use a clip that is in compliance with the ratios provided by AISC/RISA, in this instance the clip would need to be less than 13-3/8" long or essentially a 11-1/2" clip with 4 bolt clip (now the clip isn't strong enough).

 

[sbisteel]

Personally, I think it is crazy to invoke HSS truss connection restrictions to a simple shear connection. Are the welds designed using AISC equation 10-1? If so, this equation completely ignores any out of plane bending in the welds.

On the other hand, with axial forces I would never use the welds shown in the RISA link. As soon as the ISL begins to flex, the root of that fillet weld will go into tension, even if there is a return at the top.

Lastly, if that connection were to go into compression the majority of force would be concentrated at the heel of the angle, rather then the toe. Without running any numbers, it feels like that design methodology might be unconservative.

 

[SteelPE]

sibsteel

When the detailer contacted me with the issue and he was describing the problem I couldn't figure out why RISA was using the provisions for a truss on a simple shear connection. It actually took about 1 hour to figure out what was going on (as I didn't have the program or a copy of the 14th edition) and once I figured it out, I thought it was absurd to be using a truss connection requirements for a simple shear connection.... but then again, I might not be a good engineer and just have never applied the provisions properly.

While you may not agree with the methodology of the connection with axial loads, every engineer I know uses HSS Column/WF Beam bracing which requires you to transfer axial loads throughout the entire connection. The only alternative to that is to use a WF column in place of the HSS. Never going to happen.

RISA has been contacted about the issue.... and this may be user error as the program is new to everyone involved.

 

[sbisteel]

Sorry if I wasn't clear, I have no problem with transferring axial loads through HSS columns, my problem is with the weld shown in the RISA website you linked to.

Other than that, I completely agree with you on all of your points. I do not think RISA is applying the provisions correctly, maybe I'm a dummy too.

 

[sbisteel]

And now I realize that when I said ISL in my original post, I really meant OSL.

 

[nutte]

The clip angle connection, bolted to the beam and welded to the column, can be used for shear. It should not be used for axial. The Risa link mentions that there are no documented examples of this connection under axial load. The reason is because it should not be used. As the connection flexes away from the column (when under tension), a torsional moment is induced on the welds at the tip of the angles. This torsional moment should be avoided.

If you have a beam to HSS column with shear and axial load, you should look at a shear tab.

 

[BadgerPE]

I am interested to see how this turns out with RISA. I like the program for simple shear/moment connections and I really like that they show the calculation procedure for each limit state. However, I have found it to be fairly "buggy" which can be quite frustrating at times.

 

[JoshPlumSE]

One of the odd things about the current edition of chapter K is that it is organized differently from other provisions in the spec. Chapter K tries to be a bit more of a recipe.

What type of connection do you want to make? Go to this table or section and follow this recipe....

The rest of the code, on the other hand deals mostly with limit states. This is how you calculate the capacity of a member for web crippling, or flange bending or such.

What gets lost in this format is that the equations given in Chapter K are based on specific limit states (yielding, plastification, punching shear, et cetera). The chord wall plastification that you get in a Tee connection isn't really a "truss" limit state so much as it is an HSS limit state when subjected to that type of loading. If the clip angle or end plate connects to the HSS with the same type of weld pattern and applied forces, then the limit state is valid.

Now, you can argue that there could be OTHER limit states to consider as well. Or, that you would prefer to use a different weld arrangement for that connection.

 

[JoshPlumSE]

I very much prefer to have the program make a rational attempt at treating the axial force in the connection rather than just ignoring it. The treatment of axial force, in general, has been one of the features that has generated the most positive feedback from users.

A couple more points about the program:

1) When there is no axial force in the connection, then the axial limit states are not checked. This goes will all the connections (Wide Flange or HSS). Therefore, if you don't agree with the method we use for checking axial force, then all you need to do is choose not to input any axial force for the connection design.

2) For this particular connection, when there is axial tension, RISAConnection will change the weld so that there is a full return on the bottom of the clip angle. So, the weld will look a good bit different than is shown in that RISANews article. The weld shown there is really more of a pure shear connection weld.

3) We would be happy to add to or modify our existing library of connections. But, we are somewhat limited based on the current published literature... which tends to focus on HSS truss connections.

Personally, I would like to add a connection for HSS beam end-plate connected to a WF column as I have used this in the past for industrial structures when I use an HSS beam (due to a torsional load). However, there is no published procedure that I know of. I will probably be putting together a procedure of my own at some point, but we very much prefer to follow published procedures. We don't mind supplementing the published procedures slightly, but we don't generally like building them from scratch.

 

[SteelPE]

Josh,

Please keep in mind that I was not the one using the program..... but I don't think there was any axial load in the connection that was tested in the Tekla/RISA link. We are just talking about a simple shear connection. The detailer contacted me because every connection failing in his model. The primary reason was the Hb/Bb ratio set by table K2.2a. As stated above, the only way to make the connection work using the material discussed above is use a 11-1/2" long clip or change the size of the column and clip angles such that you are in compliance with these ratios. Using these requirements ties the hands of the person trying to select the proper connection.

I know RISA is dealing with limited information and from what I can tell the competition doesn't perform any checks at the support.

 

[nutte]

… if you don't agree with the method we use for checking axial force, then all you need to do is choose not to input any axial force for the connection design.

The problem with this approach is that by providing capacities for this type of connection, the program is giving a tacit approval of this connection type and the results from the program. Tekla is a detailing software, usually operated by detailers, not engineers. When detailers use this, or SDS/2, the software gives them a sense of security that the engineering being done by the software is correct. Therefore, I think the software developers have a duty to provide safe and conservative calculations, especially when the software is going to be used by non-engineers.

Where the connections deviate from standard, published cases, a better approach in my mind is not to provide capacities for them. I know SDS/2 has done this at various times in the past, marking a connection as “graphical” or “user generated,” so as not to imply a certain capacity is being provided.

This specific case in question, a clip angle welded on the outstanding legs with axial load, should be avoided for the reasons I stated above. Welding the bottom of the outstanding leg doesn’t change my thoughts on that. (Is the top of the angle unwelded? Then you have the same problem at the top of the OSL weld. If it is welded, it no longer provides the prescriptive end rotation capabilities required for standard shear connections.)

 

[SteelPE]

nutte,

What is your opinion when axial load is not a requirement of the connection? Or should this type of connection be avoided in all instance (axial load or not)?

 

[nutte]

Never mind my opinion. Decades of successful use have shown that this connection is great for shear. Not coincidentally, there is also literature to back this up.

 

[JoshPlumSE]

SteelPE -

Yes, that is the drawback with using chapter K the way it currently exists (with tables that have specified limits of applicability). I heard someone say essentially the same thing at the last set of AISC meetings. Talking about how they were fielding a lot of questions on the subject. Apparently, a lot of people are assuming they can't use some connections because they don't meet the some of those ratios.

When you violate those "limits of applicability" what it probably means is that your connection falls outside the range of connections that were tested to develop the code procedure. The limit states are likely still valid. You just aren't assured that following the code procedures alone will result in satisfactory connections.... Other limit states not documented in the code could end up controlling. Or, the equations for the given limit states may need some minor revisions.

In the program, this should probably be a "soft" failure or warning rather than a hard failure. In it's present form, the program does not have a mechanism for a "warning" message instead of pass / fail. Even so, this is something that should definitely be better documented in the help file. The way it is written now, it implies that the if any of these "limits of applicability" are not met then all the other limit states can be considered to be invalid. And, that's just not true.

Note: I did not write that section of the help file. Though I almost certainly reviewed it and let it slide by me.

 

[wannabeSE]

Sorry I did not read all of the commentary above. Isn't this connection is similar to example K3 in AISC's Design Examples for the 14th edition (

https://www.aisc.org/WorkArea/showcontent.aspx?id=33520).

The example is for a double angle connection to an HSS column.

 

[JoshPlumSE]

Nutte -

I think you and I are more in agreement than it seems. These are all considered "shear connection" in RISAConnection. Therefore, the assumption is that the primary loading is shear, but the program does consider axial force if it is present. And, the basis is the published Chapter K formulas and tables, which seems logical.

The issue in this case, I believe, is that the connections are failing some of the limitations listed in the Chapter K tables. Therefore, we list them as "failing" connections. The reality is that they are somewhere between a fail and a pass.

These connections could be said to pass if an engineer has looked at them closely enough to say that these limits of applicability do not invalidate the other failure modes or necessitate consideration of additional failure modes. The main problem seems to be that the program isn't nuanced enough to recognize this in-between failure state and instead conservatively throws up a "fail".

 

[Ken62465]

I'm chiming in here as I'm the detailer that SteelPE was harassed by for answers to this question. Please go easy on me, even though I've been a structural steel detailer for near 30years I'm no engineer so much of what you guys say may only loose me. The one thing I spot in reading the responses to this thread deals with the axial load in this knife angle connection to HSS column check and I assure you none was entered and even in the Risa calcs produced the Axial value used says "0.00". This makes me wonder if the "Branch Aspect Ratio" is an axial force check why would it need to be performed if the connection has no axial force applied?

Don't get me wrong by any means. I think very highly of the Risa connection software and currently putting it through the test with the Risa-Tekla link. Great promise but I've run into a number of issues I'm hopeful to resolve with support and the development team.

 

[hokie66]

Just as an aside, how do you erect a beam with that type connection?

 

[JoshPlumSE]

To me, this connection is probably more related to table K3.2 (which has the same aspect ratio limitations as K2.2). But, that table shows failure states related to HSS moment connections, not truss / axial connections. And, there will be moment for the type of shear connections. Yes, they're primarily shear, but the eccentricity of the bolts creates some moment in the connection. That moment should be considered in the connection to the tube column.

Of course, there can certainly still be debate as to whether the branch aspect ratios should apply or not because we don't have a true branch.

If the reason for that limit is because they are concerned about failure in the branch members, then this would not be a reasonable limitation to enforce (for this connection which doesn't have a true branch member). If, however, they are concerned that a branch connection with a different aspect ratio is going to significantly change the chord failure states, then it would definitely apply. I haven't seen anything in the code or commentary to make this clear. I should probably go through the CIDECT or AISC design guides to see if they offer any definitive reasons. But, I don't anticipate having time to do that anytime in the next week or two.

 

[nutte]

Hokie, the bottom flange gets coped, so the beam drops in from above.

Josh, you're talking about something different than I am. You're saying a small deviation from the boundary guidelines shouldn't prompt a failure. It should alert you to look into it further and make an informed decision. I won't disagree with that.

I'm talking about providing, and thus endorsing, a connection type that lacks published literature, lacks consensus in the engineering community, and has problems with it. On top of this, you're expecting detailers, not engineers, to make an informed decision about this non-standard connection type. I think Risa overreached on this connection. It should have been left as shear-only.