USING 50% UDL load For Connection Design 7

[Kishore sk]

Hello Engineers..
I have questions from connection design for AISC, In some STRUCTURAL GENERAL NOTE they have mentioned that the beam END CONNECTIONS shall develop one half the maximum ALLOWABLE UNIFORM LOAD for the beam assuming the beam is continuously SUPPORTED LATERALLY.
And the question are:
1] We have to take 50% of UDL for their respective span from CODE ?
2] What type of Load(ex. DL/LL) will be provide for Connection Software the taken UDL load?
3]For example we take (W12x14 span of 14') and the 50% of UDL ASD is 12.4K .For this we have to assign the 12.4kip as DL or any other combinations?

THANKS IN ADVANCE..​

 

[EngDM]

1. The equation for Mr for UDL simply supported beams is Mf = w*L^2/8. You set Mf to the Mr of the beam and back calculate what w is. Then Vf = w*L/2. I'm not sure if AISC provides tables for these loads, but I know CSA does based on different spans. So either you can use the tables or use the equation I've outlined. Typically Mr of the fully supported beam is used.

2. Connection software typically deals with total loads (LRFD or ASD). Special provisions are provided for seismic loading cases. Do not use LRFD loads with ASD connection design and vice versa.

3. I'm not sure how you determined 12.4kips since I use Canadian standards and am not familiar with ASD. However, unless I am mistaken the 12.4kips does not get assigned to any combinations because it is the max load that the designer of the beam could support via UDL, which the designer has already run with their combinations when sizing the beam.

 

[cliff234]

Specifying that beam shear connections be designed for a percentage of the Table 3-6 uniform load capacity is a BAD idea! If you are delegating design of the connections to an engineer working for the fabricator AISC recommends that the EOR show the reactions on the framing plans (AISC 303-22 Option #3). Parts 2 and 3 of the 16th edition AISC Steel Construction Manual specifically recommend NOT using Table 3-6 to mandate required beam shear connection strengths!

Using Table 3-6 (and applying a load multiplication factor to account for composite action) is often overly conservative, however the using the table in this manner could dangerously underestimate the required beam shear connection strength. Composite construction usually increases flexural strength (and uniform load capacity) by a factor varying from 1.2 to 2. Many engineers require that the Table 3-6 loads be increased by an arbitrary factor of 1.5 to account for the increased capacity of a composite beams. The problem is that an arbitrary factor of 1.5 can be either wildly conservative or wildly and dangerously unconservative!

Variations of Table 3-6 have been in the Manual since the first edition - however that table was never intended to be used to mandate required connection strength. The table was useful for selecting non-composite beam sizes in the "pre-computer" days.

 

[phamENG]

I agree with cliff234 1000%, and no that's not a typo.

But, it looks like the OP is working for the fabricator and has gotten stuck with one of these sub-par sets of documents from an EOR.

1) You should send an RFI to the EOR requesting the beam end reactions. It's their responsibility to provide them. They'll keep pulling this nonsense until they get sufficient push back with citations from the AISC code of standard practice.
2 & 3) We can't tell you that. We don't know what software you're using. And if you don't know how to use your own software, that's a bigger problem. You should talk with the senior engineer in your office.

 

[lexpatrie]

Y'all ain't answering the question they asked. It's fine to digress, but answer their question first.

There's a table in the steel construction manual for maximum end reaction (shorter spans are governed by the shear capacity of the section, longer spans governed by moment), I don't have a manual in front of me currently but it's in the first half of the manual.

Short spans can be particularly tricky as the shear capacity can sometimes exceed the depth for the necessary bolts.

There are drawbacks to this approach (which others will be happy to digress onto).

2 - you may have to input manually as most software isn't set up this way. Hypothetically point loads 1' from each end?

3 - If you are in LRFD I'd suggest using D+L for your load case, for floors at least, you can ask the EOR for a load ratio, too. Or just note on the submittal what ratio you used, 3 live to 1 dead is somewhat typical.

 

[EngDM]

Short spans can be particularly tricky as the shear capacity can sometimes exceed the depth for the necessary bolts.

For short spans I trypically try to get Vr to work since or, I calculate the loads by hand based on the EOR drawings and some rough tributaries and see what a reasonable connection can support. The I add that resistance - an arbitrary % to be conservative and add that value to the assembly sheet "connection has been designed for ___ EOR to verify design loading is acceptable." This kind of skips the RFI to the EOR steps and only comes back if there is a problem.

 

[cliff234]

lexpatrie – You are correct. It looks like I did not answer the OP’s questions. Let me try to answer them.

Question 1: First off, the discussion should be about the UTL, not the UDL. Per AISC 303, the EOR is responsible for clarifying this. This is usually done by referring to Table 3-6 in the AISC Manual. Is the framing composite framing? If so, you can’t just use 50% of the UTL when using Table 3-6. To do so would be dangerous. Likewise uniform load tables (such as Table 3-6) are good only for uniform loads.

Question 2: Per AISC 303, the EOR is responsible for clarifying this.

Question 3: Per AISC 303, the EOR is responsible for clarifying this.

Rather than me guessing at the answers to these questions, the OP should ask the EOR.

Respectfully, both the OP both seem to have not read Section 3 of AISC 303, the Code of Standard Practice. I am only half joking when I say that most EOR’s have never read AISC 303. It’s a short document and most of the issues related to connections are in Section 3. All engineers who design steel buildings in the US have a copy of AISC 303 – it’s in the AISC Steel Manual! And if they don’t have the steel manual (which would be very strange) they can download a free pdf of AISC 303 from the AISC website (if they are members of AISC).

It is frightening to me that so many EOR’s seem to know so little about their responsibility regarding connection design. The EOR ultimately responsible for the safe design of all connections – even when delegating connection design to an engineer working for the fabricator. The law requires EOR’s to follow AISC 303! AISC 303 is referenced in AISC 360, which is referenced in most building codes. If an EOR does not follow AISC 303, they are violating the law!

I suggest that the OP send an RFI to the EOR for answers to these questions.

 

[phamENG]

I agree with cliff in everything but the statements about 303 being the law.

In the absence of specific instructions to the contrary in the contract documents, the trade practices that are defined in this Code shall govern the fabrication and erection of structural steel.

The Code of Standard Practice isn't a "law" in the same sense as an adopted building code. It's a set of standards that are implied unless specifically countermanded in the contract. So an EOR can vary from standard practice, but that's a fraught path to follow.

 

[cliff234]

phamENG: I’m (obviously) not a lawyer. What you quoted is correct as long as “…specific instructions to the contrary [are provided] in the contract documents…”. From what the OP stated, I am of the opinion that the instructions regarding connection design were neither clear, complete, nor specific enough - otherwise the OP would not have had any questions. Yes - not following the CoSP is "a fraught path to follow"!

It has been my observation that most structural failures are connection failures – and many failures (including, in part, the Kansas City Hyatt Regency skywalk hanger failure in 1981 that killed 114 people) are in part related to poor communication. Engineers who chose to deviate from AISC 303 should be darn well sure they are clearly and completely communicating their design intent. I'm going to continue taking the easy road and follow AISC's recommendations as stated in the Code of Standard Practice.

 

[phamENG]

Indeed. Most don't clearly state a deviation, and even more probably don't even know they're deviating. This is how I was taught to do steel connections. It was only after a few projects that I got sick of all the back and forth and confusion that I did my own research and found out the firm I was working for was blindly doing it wrong.

 

[dik]

FYI



-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[Kishore sk]

Thankyou Gentleman's..

Currently I started my profession as Connection design engineer. In my office there is no senior engineer in connection design and I'm in the learning process..
When I saw this platform, I just wanted to utilize this with your IDEA'S sharing.

And it's "Great"...​

 

[EngDM]

In my office there is no senior engineer in connection design and I'm in the learning process..

There is nobody in your office who will be able to spot check your work? This doesn't sound ideal when providing stamped drawings and taking liability.

 

[lexpatrie]

@EngDM, that seems reasonable but it lacks the "upstream pain" in the EOR that they made a mess for you. You could maybe do both, send the RFI and describe your approach in the absence of better direction, their lack of a response to the RFI would be helpful on the trailing liability, If you felt the urge.

 

[Once20036]

Y`all are coming down awfully hard on engineers who use the 50% rule.
I haven't bought the 16th edition, but 15th edition doesn't discourage the use of table 3-6. OP is asking about 50% criteria so I assume we're talking about a non composite beam here.

If you look at W8x10 x3ft long (to pick an arbitrary example) the ASD load is 53.7, the LRFD is 80.5.
Take half of that total load and design the connection. The breakout of dead load to live load is irrelevant, isn't it? As long as the ASD design force is 26.85, assign it all as DL or all as LL.

To qualify my position and help to focus on the OPs question: I agree that the 50% rule isn't a guarantee and the engineer still needs to think about snow drift, point loads, and other scenarios in which the beam design isn't governed by flexure. I agree that its a short-cut and likely leads to less economic designs.

That being said, he language used in this thread is pretty harsh... "sub par sets of documents" "have not read section 3" "neither clear, nor complete, nor specific".
Why is any upstream pain necessary? What mess was made?

 

[271828]

Once20036, there's a reason why this subject gets under people's skin. The energy for the rough comments comes from dealing with the consequences downstream.

I'll give an example. In one of my first projects, the structural drawings said to design connections for 50% of the UDL. There were four or five short beams with connections that took a total of a couple of days to deal with. A 4 ft long W10x12 would have an LRFD reaction of 93.8 kips / 2 = 47 kips. If there's a cope, then the end of the beam will need a doubler plate. These short beams probably have 5 kip or smaller reactions!

All of that to save 20 seconds of the EOR's time not putting the reactions on there. If I'm the EOR and I save myself a little time and put a big burden on others downstream, then frankly I'm being a pretty bad team member.

If it's too time intensive to put reactions on there, then 99% of the problems can be avoided by adding a little table with the reactions for W8, W10, and W12 beams. Just say design all of those for 15 kips, for example.

 

[human909]

I agree that its a short-cut and likely leads to less economic designs.

But why is this short cut necessary? Why would you waste so much of other people's time when it takes about minute or two to add reactions? Instead you could be wasting DAYS of man hours with the additional design and fabrication time.

Over here engineers generally do their own connection designs. For simple beam shear connections it is normally just a cut and past job because a shear connection is extremely simple. For more complex connections I will pick the most heavily loaded member of that type and design a connection to suit. Sure the rest get slight beefier connections, but unless one member is significantly different in nature to another then you aren't talking about significant overdesign.

Meanwhile I'll also be upsizing members to avoid doubler plates. They are expensive and it is usually cheaper to upsize the member.

 

[271828]

If you look at W8x10 x3ft long (to pick an arbitrary example) the ASD load is 53.7, the LRFD is 80.5.

You've given a pretty good example of why this approach is problematic. The LRFD reaction would be 40.25 kips.

Chances are, the typical connection on the project is a shear tab. Only two bolts will fit in the W8. A shear tab with two 3/4 in. A325-N bolts is worth about 25 kips, so an atypical connection will be needed.

The best hope is a double-angle welded to the W8 and bolted to the support. Even then, the picture doesn't get any prettier.

The W8 web shear yielding design strength is 40.2 kips. That's if there's no cope!!! If there is a 1 in. deep cope at the top or bottom, the shear yielding strength is over by 15%. If there's a cope at the top and bottom, then it's over by 34%.

If there is a 4 in. wide by 1 in. deep top cope, then the coped beam flexural strength is over by 41%.

This W8 will need expensive reinforcement.

That's what the team "gained" when the EOR saved 10 seconds by avoiding writing R = 10 kips on that W8!

I get that it would be a pain to write R = 10 kips on every short beam. The UDL approach works OK when there is a note saying something like "Design all W8 and W10 shear connections for 10 kips unless noted otherwise." Or a small table with reactions.

 

[Once20036]

I agree that its a short-cut and likely leads to less economic designs.

Again, I`m not trying to argue that this is the best method. Just that its common, and really isn't as painful as this thread makes it seem.

In my experience, everybody knows that this leads to silly results on short spans, the connection designer gives one row of bolts, posts the resulting capacity as a deviation from the contract, and moves on to the next connection with no time lost.
Again, there's no pain and no mess.

 

[271828]

Once20036, have you done any delegated design of connections?

I've used EngDM's method before. Very often it works, but I have had the EOR come back and reject what was done, causing (unpaid) rework under very tight time constraints. Even when it works, I wasted time adding all of those "EOR please verify..." notes to the calculations, and there's still a stupid open loop that has to be closed.

Why not put the reactions on there? You have them handy. It takes seconds to put them on there. I worked as an EOR for years and we always provided reactions. At this point, my mind is blown that this is even a question.