Problems! 1

[yungenr1]

During shop drawing review, we looked at a few of our composite beam cant. designs again because of the connections used by the detailer.

During this review, we've found that the our program (RAM Steel) designs cant. beams as pinned at the column, so the only moment was that from the cant. When I fix this connection, as it actually is, the moment is greater and then my moment value provided to the detailer is no longer valid (it is lower) and the beam design is over stressed by about 5-10%.

This is for a single beam (occurs) 5 times (changing from one size (8" flange) to a 10" flange (2 weight sizes) fixes the problem. I'm not sure how to tell my boss (it is his stamp, but I did the design.)

Some of our columns don't work either in random locations due to the program not putting moment from the moment connection into the column.

What should I do? Right now I feel like just quitting and staying away from structural engineering forever.....Sigh.

 

[jike]

Most of us deal various kinds of problems everyday. It is not what happens, but how you deal with it.

Approach your boss in an honest, humble way. I am sure between the 2 of you, you can figure out the best approach to take.

 

[JAE]

I'm not sure how to tell my boss (it is his stamp, but I did the design.)

I'm with jike on this too. Simply be honest and own up to it. Public safety is far greater than your ego. There's nothing to gain and a lot more to loose if you don't talk to him.

Now..."how" you bring it to your boss is somewhat irrelevent. It does help to come with some possible solutions along with the problem...vs. just bringing the problem to your boss.

 

[sundale]

It is only a mistake if it is actually BUILT; it is simply "egg on your face" otherwise.

Fess up and take your lumps like we all have done! Every engineer has screwed something up and had to make it right. Public safety and engineering ethics demands no other choice here.

Depending on the steel fabricator, it may be cheaper to simply replace the undersized members or cover plate them where there is too much flexure. It is also legitimate to use the Fy of the steel mill certifications (like 54 ksi) in lieu of the nominal 50 ksi yield stress. You could perhaps use the live load reductions permitted by code (if you have not already)? Maybe you can squeeze in some kickers to reduced the columns' unbraced lengths? Sharpen your pencil!

To quote my master's thesis advisor: "Nothing generates wrong answers faster than a computer!"

 

[yungenr1]

Well, after my concerns I looked a little further into the program, and we have some LL reduction that I wasn't taking into effect for the beam design. So the beams work, but the moments are still a little off. I have to review this further to see what capacity we got from the moment designs submitted.

Some of the columns may still have some issues though.

I agree honest, and humble is the way to go, but I'm just scared that I will hurt the company/someone.

Also, when you design a cant. beam that passes over top of the column do you design it fixed or pinned?

 

[hokie66]

Please take the advice you have been given by jike, JAE, and sundale. When you discuss the understrength issues with your boss, also get his guidance as to how to model the beam to column joint. As you say, he is responsible for your work, so you have to be open, honest, and thorough in discussing the issues with him.

 

[SteelPE]

Not quite sure how it would hurt your company. Would only be worse if something were to happen once the building was occupied.

I have worked with many contractors and fabricators. Changing a beam size is not a problem. Since this was caught during the shop drawing review, the fabricator should be able to easily switch the beam and column sizes. There maybe a little increase in the cost but depending on how large the job is it may come out in the wash.

Last time I heard, steel was running at $1000-$1200 / ton. You should be able to estimate how much the change is going to cost your customer. This is good to know incase the fabricator tries to burry any other hidden charges.

 

[slickdeals]

I am not sure what you mean when you say the cant. beam is pinned at the column? Won't this create an instability? Are you saying the column is pinned or the beam is pinned?

 

[yungenr1]

SteelPE,
The steel supplier we're using always makes ridiculous CO prices, but for some reason still has the job. So we're trying to minimize the cost difference. The last quote we got for steel was over $2000/ton, has that changed?

Slickdeals,
You are correct, the program assumes the column is pinned at the beam, to create a simply supported cant. beam, and not a small frame.

Thanks.

 

[SteelPE]

The number I am giving you is for raw material costs (this may be slightly different based on where you are located). Fabricated costs may be higher.

However, as I see it, the only difference you have in the design would be the change in beam and column weight. The connections that need to be fabricated and erected together with the number of pieces do not change so there should only be a slight change in the overall cost due to the weight difference only.

 

[lkjh345]

The $2,000 a ton figure may be for fabricated steel. The $1,000-$1,200/ton figure SteelPE is referring to may be for the raw cost (unfabricated) of the steel.

These figures seem fairly close to the prices we are hearing form our local fabricators in the midwest.

Just changing the beam sizes should only involve a change in the raw steel cost. The fabrication required should be approximately the same for the different beam sizes, and should not change significantly.

 

[slickdeals]

If the beam is bearing on the column and is connected using a single bolt on each flange coinciding with the CG of the column, then the column will not develop any moment. Agree?

However, I am assuming that most connections will have a cap plate with 4 bolts, which will generate a couple. But again the moment that can be delivered to the column is not going to be significant.

How does everyone else model the connection as the OP asked.

 

[jike]

Depending upon how you have the eccentric parameters set up in RAM STEEL, your column moments may be over-estimated.

Try justifying the columns by a partial model in RISA or some other program. You may be surprised that they work.

 

[JedClampett]

A famous structural engineer (whose name escapes me now) designed a high rise. A student pointed out that with the building shaped the way it was, a transverse wind case would control vs. the orthogonal cases that were checked. This wasn't discovered until the building was complete and occupied.
The engineer was upfront and designed the repairs. I'm sure this cost him money, aggravation and time. The fixes had to be fitted in the already finished building.
So your issue is nothing. Don't be afraid to point out mistakes. It might seem like a big deal now, but you'll sleep better.
I'm sure someone out there knows the details and can help me out. There's a huge lesson to be learned.

 

[swivel63]

the beams are over stressed by only 5-10 percent? tell your boss, but i bet he's not going to be worried. many engineers eat 5-10 percent for breakfast.

 

[ARLORD]

Jed C.,


See the following link for citicorp bldg:

http://web.pdx.edu/~pgerst/documents/citicorp_case.pdf

 

[Lion06]

That was Bill Lemessuirer, I believe. I think I spelled his name right.

 

[hokie66]

Close, LeMessurier, I think. He passed away recently. A great example of integrity, the example should be studied by all young engineers.

 

[jechols]

I thought I read somewhere that after the building was fixed his insurance company actually lowered his premium because of the way he handled the situation.

 

[ChipB]

Fixed or pinned? Depends on column orientation and connection.
If the beam is placing weak axis bending on your column, place your cap plate bolts inside the column flanges and treat it as pinned.

If the column is oriented such that the strong axis of the column coincides with the strong axis of the beam, and the bolts connecting the cap plate to the beam are inside the column flanges: column is pinned at the top
outside column flanges: fixed (at least partially)

I haven't worked with RAMSteel in many years. What method are you designing under ASD or LRFD?
If ASD:
1989: Allowable Stress Design
or
2005: Allowable Strength Design (Dammit, they slid it under the radar!)
If you used 1989's ASD, you would likely either be ok, or alot less overstressed switching to 2005.

Depending on your structure, your boss might be able to handwave small amounts of overstress. It's a matter of personal opinion.