Considering moment In Design 2

[Xeus]

Hi there. This might sound really weired but today I got involved in a discussion with our senior engineer who has 30 years of experience under his belt. (I've joined this company just recently) He was saying that in east coast we do not consider earthquake loads in design and also in areas where basic wind speed is 90 mph (Minimum) we do not even consider connections (beam, column, base, etc.) moment resisting. We assume everything PINNED and do our design. Well, I'm not used to this and have never heard of such a thing. Particularly when it comes to a concrete building, it is automatically a full moment resisting frame and you can't just ignore it! Anyway, that's what he says and none of us could convince the other. Since I haven't been doing this for a long time (and what he says is 100% in contrary with what we learned in university) I thought I might ask you guys. Is he right? Sounds impossible to me but is it really true?

I appreciate your help. (although he wouldn't listen anyway, but I'll know the truth.)

 

[Xeus]

Well, you're right about his background! He has 30 years of experience and even he himself admits that his methods are 30 years old. To give you a better idea, I should say that he hesn't even started to use LRFD method yet!!! I don't argue actually, I've been mostly discussing the matter with him in a way like I'm trying to learn despite the fact that I myself have 10 years of experience (solid design experience with earthquake loads) but well, he's an older engineer, you know.

An example I was talking to him about was that he says although we use top and bottom rebars in continuous concrete beams that go across columns and also we have rebars in columns, there's nothing to transfer the moment from the beams to columns because we do not bend the beam rebars into columns! I argued that the core acts like a rigid body and when the end of beam rotates, the core rotates and therefore, the column end rotates. So in reality, assuming connections pinned is nowhere close to truth. What I'm afraid of right now, is that following his methods (which sound completely non-professional to me) might affect my engineering knowledge and if I join another company, saying such things make others laugh at me and ask if I'm coming from 19th century.

 

[haynewp]

The part about ignoring earthquake loads whenever the wind speed is 90 mph on the east coast is plain wrong. My buildings are on the east coast with 90 mph wind and most are controlled by earthquake in my area.

I have also recently begun a career of heavy concrete design and it is kind of weird coming from doing mostly steel, wood, and masonry. Concrete design to me is very arbitrary, you can assume all kinds of things in designing a slab and each way may work. Also, I have noticed other engineers that have been doing only concrete for years ignore compatibility torsion all the time in floor systems. And also treat floor beams and one way slabs framing into spandrel beams as pinned connections and provide very little rebar at the connection (the 20% max distribution is not considered). The reasoning is that spandrel beams aren't going to give you any significant amount of end moment resistance.

It would seem to me that always assuming a pin at such cases could be bad for shear resistance. I don't know...

 

[JAE]

The part about ignoring earthquake loads whenever the wind speed is 90 mph on the east coast is plain wrong. My buildings are on the east coast with 90 mph wind and most are controlled by earthquake in my area.

Yes

I think this is a fact that many engineers out there don't appreciate. Seismic is much more "controlling" than most think.

 

[Xeus]

Well what he says is about DC metro area where earthquake factor is 0.05 and they ignore it all the way. They even say a 90 mph wind is not controlling and just design for gravity loads. He was saying the 33% increase in allowable stresses puts LC's with wind loads off the table because wind loads are not that huge. I told him you can't use that if you go LRFD and you'll design similar to concrete with load factors and max stress ratio of 1.0.

 

[JAE]

One point about seismic is that when you apply the seismic loads to a building, for some connections and parts, the overstrength factors kick in and sometimes do govern over wind.

Also, for long narrow buildings, the wind on the short elevation doesn't control the overall lateral design due to the limited exposed area vs. the seismic demand.


Some years ago I was a young engineer working on one-story buildings in Texas. The buildings were for a grocery store chain and we had a UBC zone 0 seismic condition. As I was learning how to design them, I asked how to do the lateral wind analysis and the engineer I worked under scratched his head and said something to the effect that, "...well, we probably should check that but we just haven't in the past."

 

[haynewp]

I have had several buildings where wind goverened one direction and seismic in the other. I have found that after most engineers pass a certain age they get into that frame of mind of "this is how things are" and are not receptive of new code requirements. 30 years of experience usually qualifies. Kind of the way contractors are.

Are you serious when you say he considers everything pinned?? Even steel buildings? wtf?

 

[271828]

Xeus typed: "What I'm afraid of right now, is that following his methods (which sound completely non-professional to me) might affect my engineering knowledge and if I join another company, saying such things make others laugh at me and ask if I'm coming from 19th century."

Why would you quit doing things your way and adopt his? Is the office so dictatorial that they'd tell a 10 year guy how to design concrete beams? If so, I'd be out of there in a hurry.

This needs to be a "you can do it this way, but I'll do it that way" situation, and nothing more. He's obviously done fine and will retire soon. You're probably technically more correct, so there's no reason to start doing things in a less precise manner.

As for the DC area and ignoring lateral loads, I'd guess that he's right, regardless of whether he uses WSD or USD. My pal who went to work there is one of these ultra-precise fellows and had this "WHAT!?" reaction at first, but proved it to himself pretty quickly. You will probably do the same. It's also good engineering judgment to recognize when one is NOT in one of those severe areas and adjust one's work habits accordingly.

I've never designed anything in DC, but have designed mostly in moderate to low EQ load areas. I've found it pretty hard to make anything other than gravity control in most cases. Usually end up with girder end moments about 5-10x larger from 1.2D+1.6L than with any lateral load combo. It also shouldn't matter if it's WSD or USD load combos, BTW. You usually get the 4/3 increase with one, but 0.75 on the other.

 

[haynewp]

"Well what he says is about DC metro area where earthquake factor is 0.05 and they ignore it all the way. They even say a 90 mph wind is not controlling and just design for gravity loads."

Xeus;

I am getting the impression that he ignores the lateral system altogether and just designs for gravity?

 

[Xeus]

tunacan : Yes, he designs just for gravity and makes other simplifying assumptions that are hard for me to take as an engineer.

271828 : Well, yes, it's a dictatorial environment in our office. He's so controlling and also in bad mood every day that we have to do things the way he wants. As for leaving the company, I can't do that right now. Long story ...

everybody : Irrelevant to this topic but not worth opening a new thread for. I have a question about checking deflection. I have always checked deflection for LL and limited it below L/360. Then limited span/Depth ratio with respect to DL/LL and finalized the design. Now my question is : Is there any direct limitation for "DL only" or "DL+LL" for deflection, similar to L/360 for LL? I looked everywhere in AISC and couldn't find anything. Are the two limiting criteria I use enough or you guys check other limitations too?

 

[JAE]

Xeus,

Re: DL or TL deflection.

Answer: Understand the effects of that deflection on the usage of the structure, connecting elements, long term creep effects, visibility of the deflection....

in other words....use good engineering judgement.

 

[youngstructural]

VERY interesting discussion. Lots of thoughts, but all of them have been covered as I continued to read through the thread.

As for the deflections under live and dead combinations as well as isolation, the typical method for canadian engineers (as least those from whom I have learnt and design procedures I have read) include dead load deflection only if the system will not naturally assume all dead loads prior to being engaged by live loads. IE, if you have a system who's dead loads are mitigated (pre curve in a steel beam) or automatically assumed during construction (mass masonry construction), the live load is the only component of deflection you need consider.

This is separate from a fundamental frequency estimate of 18 over the root of the total deflection. You must always include dead loads when making this estimate. There are a couple of threads about vibration, and it's outside of this discussion.

The Canadian methods aside, and assuming they are fairly similar to the US methodology (commonly are, as our industries work quite closely together), the New Zealand code DOES contain guidelines for separate dead load and live load only, as well as combination, deflection limitations. They are contained in an appendix to Part 0: General Design Principles. See AS/NZS 1170.0 Appendix C "Guidelines for the Serviceability Limit States". So such guidance does exist, it just may not be a common approach in the US. The format is "Element type | Phemomenon Controlled | Serviceability Parameter | Applied Action | Element responce".

A good example is for wall elements: Columns...

"Column | Side Sway | Deflection at top | Ws | Height/500"
where Ws = Wind load (excluding any dead load).

Please let me know if you would like me to post a copy of the table for your consideration. However it's based on the following BRANZ report, which would be the better reference for what you want. I will post the link once's I'm back at the office (can't seem to find it right now, and the BRANZ website is down). If it's up, check for yourself...

http://www.branz.co.nz

Author: Andrew B. King.

Good luck,

YS



B.Eng (Carleton)
Working in New Zealand, thinking of my snow covered home...

 

[haynewp]

Total load deflection limits are listed under Table 1604.3 of IBC 2006 along with the L/360 live limit.

Can you tell me where the buildings he has designed are located so I won't got there? Seriously, there is always some lateral capacity even with shear tab connected steel framing. But to make this assumption as the standard under 90 mph wind? For a basic 2-story office building in 90 mph area I would expect around 15psf*200ft*(13ft+7ft)=60 kips of lateral into the framing. And seismic would likely be more.

The IBC requires the lateral system to be defined for seismic under the general notes. I would be interested to see what he lists..."pin connected moment frames"? What is the R value for those?

 

[Xeus]

He ignores both wind & seismic forces. As far as the buildings he has designed, don't come anywhere near DC metro area and you'll be safe. LOL

 

[271828]

I think the language of the thread needs changed a little.

There's a big difference between IGNORING wind and seismic and having designed a zillion bldgs of a specific type in a specific location and recognizing that 1.2D+1.6L (or more likely 1.4D+1.7L for most of his career) ALWAYS CONTROLS EVERYTHING by about 5x.

If one has done something enough times to know that it won't control, and still checks it every time ad nauseam, then I'm not sure that says anything positive about one's engineering judgment either.

I have to assume that if he designs a 10 story flat plate bldg in Memphis or anywhere else he's not familiar with, he'd check lateral loads.

And we're talking about concrete bldgs only, right? If he has a steel framed bldg, surely he includes moment frames, braced frames, shearwalls, or something to count as the LFRS.

LOL, if anybody freakin' out about going to DC, better stay out of all those 80 year old multi-wythe brick and wood framed bldgs, just as one example. Some of those suckers were built using methods that we'd never dream of today.

 

[Xeus]

One more question guys : I always limited deflections using the "span/depth" ratio given in (beam) section of AISC. Now my question is, I design the beams with factored loads per LRFD, but, if I want to actually calculate deflections in beams I have to use service loads (non-factored loads). That means I'll have to analyze each beam twice.

1- Once with factored loads and design the beam for moment;

2- Once with service loads and check actual deflection for use with L/240 & L/360.

Do you do the same thing? Analyze twice?

This comes from our latest project which is an equipment shelter platform which will be installed on the roof of an existing building and because of some sensitive equipment inside that shelter, we have been required to accurately calculate deflections of all beams and not just go with that table. Any ideas?

 

[271828]

First, how do you "limit deflections using the span/depth ratio..."?

Yes, use factored loads to come up with moments, shears, etc. for strength design.

Yes again, use service loads to check deflection.

As for the last paragraph, why is it harder than just calculating the deflection due to the new load?

 

[haynewp]

271828

I think he is saying all buildings and not just concrete. That was my question a few back. Also, I would expect the buildings that are 80 years old have a lot of redundancy built in that economized steel framed, metal stud office buildings of today don't have.

 

[Xeus]

That comes from the fast pace of our work environment. When I first started using LRFD, we were so busy that we were instructed not to waste! time on doing two series of calculations with/without facored loads. We were sorta ordered to use factored loads to design for moment and use AISC recommended span/depth ratios to limit deflection and get done with design ASAP! I know it's not professional engineering, and that's why I'm asking to see what other people do in more comfortable work environments, because I don't want to be an engineer who doesn't know what the right thing to do is, even if he weren't working in a dictatorial environment.

 

[271828]

Xeus:

Well, what he recommends won't always work, especially for girders.

You can check deflections without another load combination. Take your deflection limit and scale it up by the ratio of the factored load to the service load. For gravity bms (assuming that's all we're talking about), 1.2D+1.6L should control the strength calcs 99.9% of the time. Your D & L should be the same over your entire floor or roof, so figure out a scaling factor. Maybe L/360 goes to L/300 for example. That would be the fastest way to approach the problem without buying software.

Which brings up another question: Why not use a program like Ramsbeam? Last time I checked, it cost next to nothing, like $300 and is very handy. Then you could set it to match your standards and be done with it.

The real question to me is: "How long before you're outta there?!" There have to be other places you can work for.

 

[Xeus]

Yes, there are other places but to cut the long story short: It's an immigration problem. This company has supported me for permanent residency through employment and for that reason, I can't leave until 6 months after I receive my permanent status, which is actually, a long way to go.