Crane Runway Beam - Lateral Deflection

[m1ke99]

I'm evaluating an existing building for an increased crane capacity. Client wants to go from 30 Ton to 35 Ton. Overhead double girder, top running.

Everything is looking okay, except for lateral deflection on one of the crane runway beams. The runway beam on one side is on the exterior gridline of the building and laterally braced at the columns. The other runway beam is connected with a horizontal truss to the adjacent runway beam and therefore much stiffer for lateral deflection.

Is it okay to assume the lateral (side thrust) loads will be distributed to the stiffer runway beam. Or, does each runway beam need to meet the deflection criteria separately?

 

[jayrod12]

Ok, I'm on board slightly, but if that's the case and skew is the hill we're going to fight this fight on, what's the deflection for a beam subjected to a point moment? Certainly significantly less that a point load at midspan of a simply supported beam? Or if look at the skew loads as opposite equal forces acting a distance apart, the deflection again will be significantly less than the simply supported point load in one direction.

If we really wanted to sharpen our pencils, we could get an existing condition to calc out I'd bet. In a court, it doesn't matter what sounds better, it's what is acceptable. If I can work the numbers to prove something is acceptable then I don't care if it sounds like my cat's hairball at 2 am or Pavarotti.

 

[WARose]

Ok, I'm on board slightly, but if that's the case and skew is the hill we're going to fight this fight on,...

We got off on the skew tangent because Kootk was asking for a definitive source for this force. So he got one. I'm not trying to quantify every last bit of this. Just use 20% laterally (of the vertical wheel load) at each point the wheel is applied.....and you're ready to roll. (No pun intended.)

If we really wanted to sharpen our pencils, we could get an existing condition to calc out I'd bet.

We can turn just about anything into a R&D project. I'm just trying to help the OP. (And arguing with Kootk is always fun. .)

 

[KootK]

I already have have. AISC DG 7 calls it a source for lateral/horizontal forces and gives deflection criteria for the same.

I disagree that the AISC blurb should be interpreted that way. They basically just dump a bunch of lateral load sources in a bucket and then, seven pages later, rattle off some generic deflection suggestions clearly regurgitated from CMAA.

You may want to divorce the two.....but you just can't do it.

CMAA does. And I would consider that to be the more authoritative document on the subject of cranes. If skew were meant to be included in the 10% stuff for deflection, why strip it out for the force calculation?

I've drawn that/explained it too. It's very possible that one side could be engaged while the other isn't.

I disagree with this well. One wheel will indeed make contact first and that does have implications for wear. But how far do you think that first wheel impact has to push the beam laterally before the other wheel(s) engage and deflection becomes self limiting? Not very far I'd say.

At the end of the day, you have to defend the design. My way would sound a whole lot better in a deposition.

I refuse to subjugate my engineering judgment to boogeyman fears over some future court case. When and if the time comes, I'll defend myself there just as I do here. I would be utterly astounded if opposing counsel could field somebody local to put up a better fight that you do frankly. Don't cry for me Argentina.

 

[WARose]

[blue](Kootk)[/blue]

I disagree that the AISC blurb should be interpreted that way. They basically just dump a bunch of lateral load sources in a bucket and then, seven pages later, rattle off some generic deflection suggestions clearly regurgitated from CMAA.

What about the Ricker paper? Right after mentioning skew (among other things) he rattles off that pesky deflection criteria.

[blue](Kootk)[/blue]

One wheel will indeed make contact first and that does have implications for wear. But how far do you think that first wheel impact has to push the beam laterally before the other wheel(s) engage and deflection becomes self limiting?

I discussed that above/and in the sketch. If that "gap" is big enough.....and the coefficient of friction small enough.....there could be issues. Betting the farm on all that is dangerous.

[blue](Kootk)[/blue]

CMAA does. And I would consider that to be the more authoritative document on the subject of cranes. If skew were meant to be included in the 10% stuff for deflection, why strip it out for the force calculation?

You are saying CMAA separates it....but I don't agree.

[blue](Kootk)[/blue]

I refuse to subjugate my engineering judgment to boogeyman fears over some future court case.

I don't view being conservative as giving up on your judgement. Like my first boss said to me (more than 20 years ago): steel and concrete are a heckuva lot cheaper than a lawsuit.

[blue](kootk)[/blue]

When and if the time comes, I'll defend myself there just as I do here. I would be utterly astounded if opposing counsel could field somebody local to put up a better fight that you do frankly.

Thank you sir......and may we all avoid that courtroom.

 

[TehMightyEngineer]

To backup a little bit; it seems that OP has two main questions to ask here:
[ol 1]
[li]Does the structure deflection exceed the requirements in the applicable specifications (CMAA, etc.) when analyzed in a traditional fashion?[/li]
[li]If it does, can we justify that the serviceability of this crane as acceptable if we perform a more rigorous analysis?[/li]
[/ol]

For item (1) many engineers would likely not step outside the cut and dry boundaries of the spec for a more grey area. If OP is correct in his analysis then it appears that he cannot utilize the stiffness of the stronger rail without entering such a grey area.

But, if OP is willing to step into this grey area, here's how I believe you could justify this:

I agree with KootK that the appropriate analysis of side thrust would be CMAA's breakdown of the various side thrust forces. I would put forth that the overall goal of the lateral deflection criteria is to avoid extreme curvature of the rails such that you had excessive wear or binding of the end trucks. With that goal in mind, I believe we can indicate which of the CMAA loads would and would not be able to distribute load to the stiffer rail:

[ol 1]
[li]Runway misalignment lateral load inherently requires both end trucks to be engaged for a lateral force to be present. Thus, we could distribute this load to the rails based on stiffness.[/li]
[li]Crane skew is independent to each rail. Thus, we can not distribute the skewing force based on stiffness.[/li]
[li]Trolley acceleration could load up only one rail depending on which wheel cheek was resting against the rail and which side of the bridge the trolley was on. Thus, we can not distribute this force based on stiffness.[/li]
[li]Trolley braking is similar to acceleration. We can not distribute this force based on stiffness.[/li]
[li]Crane steering is related to crane skewing and similarly we could not distribute this load to the rails based on stiffness.[/li]
[/ol]

If practical, I would run the calculation for these forces and distribute them to the rails based on the above. I would then see if I still violated the crane rail deflection criteria.

Here's a nice paper on free floating vs. fixed rails that's relevant to this discussion:

https://www.middough.com/Middough/files/8f/8f361fa2-3457-4395-8656-d7c088f5d548.pdf

Short version is they conclude that adding rail guide rollers (as KootK mentioned way above) is an ideal solution and done often across the pond. I would put forth that adding rail guide rollers would be a suitable alternative if OP cannot meet the lateral deflection criteria.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)

 

[KootK]

What about the Ricker paper? Right after mentioning skew (among other things) he rattles off that pesky deflection criteria.

Other than proximity, this looks the same to me as the design guide. Bucket-o-stuff --> blanket recommendation. But yeah, if my previous proximity argument held any water then I guess this does too logically.

You are saying CMAA separates it....but I don't agree.

They definitely strip out the skewing force for separate consideration. I find that suggestive and am surprised that you do not.

I discussed that above/and in the sketch. If that "gap" is big enough.....and the coefficient of friction small enough.....there could be issues. Betting the farm on all that is dangerous.

I've reviewed your comment and sketch carefully and do not see that you have addressed the issue that I'm referring too. For skewing, I'm arguing that you don't need any help from the other side. Once two wheels of the same truck come into opposing contact with the beam, meaningful deflection comes to an end and the rest is just the double curvature business that jayrod artfully described. Or are your comments and sketch in reference to other forms of lateral load than skewing?

 

[KootK]

Having reviewed WARose's sketch fifty times, I've found cause to revise my previous recommendation. I still think that one sided support is viable but, for the reason shown below, the truss should actually be held to a stricter deflection requirement than L/400. Otherwise, the beam opposite the truss may have a curvature in excess of L/400 when it does its ride along. I expect that the truss is considerably stiffer than L/400 anyhow but, still, something to consider.

 

[KootK]

Short version is they conclude that adding rail guide rollers (as KootK mentioned way above) is an ideal solution and done often across the pond.

Agreed and sounding better all the time.

1) Runway misalignment lateral load inherently requires both end trucks to be engaged for a lateral force to be present. Thus, we could distribute this load to the rails based on stiffness.

I'm not so sure that it does require both end trucks to be engaged. You could be generating your force by running at an angle to your misaligned rail on one side while having any movement be accommodated by a low friction gap on the other side. I'm not so sure there really is such a thing as a low friction condition but that seems to be what we're running with so I'll not rock that boat.

2) Crane skew is independent to each rail. Thus, we can not distribute the skewing force based on stiffness.

My argument is that there's no need to distribute the force because, once both ends of a truck are engaged, skewing doesn't actually produce a net force (other than moment) at the trucks.

3) Trolley acceleration could load up only one rail depending on which wheel cheek was resting against the rail and which side of the bridge the trolley was on. Thus, we can not distribute this force based on stiffness.
4) Trolley braking is similar to acceleration. We can not distribute this force based on stiffness.

I think that you can distribute based on stiffness once any gaps are closed. Furthermore, I contend that all systems have to be able to deal with gap closure so it's spurious to think of gap closer being a problem here. As I noted in my last post, gap closure may require truss stiffness to be greater than L/400 in order to keep the non-truss runway beam to an L/400 curvature.

I think that we're going a bit off the rails here in that we're now focused on forces. The OP's question pertains to deflection criterion for the forces which is a bit different.

 

[TehMightyEngineer]

while having any movement be accommodated by a low friction gap on the other side

Forgive me if I'm dubious; I only see significant lateral load due to rail misalignment being generated when the cheeks of the wheels on opposite end trucks start rubbing on misaligned rails.

My argument is that there's no need to distribute the force because, once both ends of a truck are engaged, skewing doesn't actually produce a net force (other than moment) at the trucks.

After some thought, I'd agree with you on this.

As I noted in my last post, gap closure may require truss stiffness to be greater than L/400 in order to keep the non-truss runway beam to an L/400 curvature.

I'm not yet on board with this but I may have simply caught a case of the Fridays and be missing something.

I think that we're going a bit off the rails here in that we're now focused on forces.

Well, forces causes the deflection and the forces are what could be distributed via stiffness of the rails per OPs question. Sorry if I got us on a tangent.

a bit off the rails

Har har.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)

 

[WARose]

Having reviewed WARose's sketch fifty times, I've found cause to revise my previous recommendation. I still think that one sided support is viable but, for the reason shown below, the truss should actually be held to a stricter deflection requirement than L/400. Otherwise, the beam opposite the truss may have a curvature in excess of L/400 when it does its ride along. I expect that the truss is considerably stiffer than L/400 anyhow but, still, something to consider.

All depends on that gap....and if it is closed.

Other than proximity, this looks the same to me as the design guide. Bucket-o-stuff --> blanket recommendation

Still holds a lot of weight with me. In the large design offices I've worked in.....this paper is well regarded.

Be back Monday.

 

[CANPRO]

I think that you can distribute based on stiffness once any gaps are closed. Furthermore, I contend that all systems have to be able to deal with gap closure so it's spurious to think of gap closer being a problem here. As I noted in my last post, gap closure may require truss stiffness to be greater than L/400 in order to keep the non-truss runway beam to an L/400 curvature.

I'm not yet on board with this but I may have simply caught a case of the Fridays and be missing something.

For the sake of conversation, lets put some numbers to this: Span = 50', allowable deflection = 50x12/400 = 1.5", assume total gap is 1" such that the less stiff side is loaded first. The stiffer side isn't engaged until the weak side is within 0.5" of its allowable deflection and now the strong side has to help prevent the weak side from exceeding its allowable deflection - this may govern the stiffness requirements of the stiffer side over the usual L/400. I believe this is was what kootk was alluding to.

 

[TehMightyEngineer]

Gotcha, makes sense to me now and I agree.

Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)