Curved Beam vas Cambering a Beam

[SALTRAM4567777]

Dear Members,

1. Will curving a beam have the same effect as cambering a beam in terms of achieving the desired structural performance and deflection characteristics?

2. When calculating the elastic deflection of a steel beam with a natural mill camber of 10mm due to fabrication tolerances, should the natural mill camber deflection be added to the computed elastic deflection to obtain the total deflection, especially when designing for stringent deflection limits?

 

[Smoulder]

1. What do curving and cambering mean to you? They might mean different things in different countries. What structural performance are you trying to get out of it?

2. Why do you have stringent deflection limits? Mostly you wouldn't count precamber but maybe it matters in your situation.

 

[canwesteng]

The mill camber should point up. If 10mm of camber is acceptable the beam must be pretty long, so it probably isn't noticeable anyway

 

[JoshPlumSE]

2. When calculating the elastic deflection of a steel beam with a natural mill camber of 10mm due to fabrication tolerances, should the natural mill camber deflection be added to the computed elastic deflection to obtain the total deflection, especially when designing for stringent deflection limits?

First off, most of the time you tell the steel erectors to place the beams such that the natural camber is UPWARDS.
That way, when dead load its, it drops into place and becomes flatter than it was before. Therefore, you could say it DECREASES the total deflection.

Secondly, you have to ask yourself what is the REASON for the deflection limit. Say, this steel beam is on the exterior of the floor system and it supports some cladding which is sensitive to deflection. The mill camber wouldn't really come into play, only any deflection from loads applied AFTER they begin to attach the cladding.

1. Will curving a beam have the same effect as cambering a beam in terms of achieving the desired structural performance and deflection characteristics?

You're talking about a horizontal curve? I'm pretty certain that this is not analogous to camber. However, I believe there is an AISC design guide on curved beams. I'm sure this would be a good reference should you have a project with curved steel beams.

https://www.aisc.org/Design-Guide-33-Curved-Member-Design

 

[JStephen]

That made me curious how they camber beams in the first place. Answer here:

https://www.structuremag.org/wp-content/uploads/2014/08/C-InSights-Criste-Apr091.pdf

If by "curving a beam", you mean "rolling it to a specified radius", I think in most cases, that would not be suitable for very small deflections.

 

[SALTRAM4567777]

As I cannot have the natural mill camber upwards as it will be difficult to maintain uniform concrete thickness above as I need flatness of slab and uniform thickness or more.

I checked with fabricator he specified natural camber of 6mm for my case. In my design software, my deflection for beam is 8mm. So will the total deflection be 8+6mm.

 

[JoshPlumSE]

As I cannot have the natural mill camber upwards as it will be difficult to maintain uniform concrete thickness above as I need flatness of slab and uniform thickness or more.

How does a 6mm mill camber change the flatness of your slab or the uniformity of the slab thickness?

I'd even suggest that issues will be worse if you don't orient the camber upwards. What happens typically is you apply your wet concrete to the decking which loads your beam. and your beam deflects due to this load. Let's say half of our 8mm deflection is due to self weight of the beam, slab and decking. Well, now you've taken out 4mm of your 6mm natural camber and you have a section that's much closer to level, and a slab thickness that's much closer to uniform.

Of course, I'm talking about normal composite beam construction. Could your situation be something more like a "shored" deck where your beam is poured without really loading the decking (or beam) because you've got shoring down to the floor below?

 

[SALTRAM4567777]

Site teams prefer levelled concrete, but casting concrete to match a cambered beam shape is challenging. To address this, I'm considering providing a natural mill camber downwards to the beam. Here's my logic:

- By providing a natural camber downwards, I would gain an additional 6mm of concrete thickness at the center. When the beam deflects 8mm, the concrete would deflect the same amount, resulting in a total deflection of less than 10mm. Or it will be 8 + 6mm

Is my logic correct?"

Also if I erect my beam in natural camber downwards, will the therotical deflection in analysis software should be added with natural camber deflection of 6mm. I have sone concerns regarding this and believe I am wrong in this logic. If anyone give logical explanation.

 

[canwesteng]

Why would you ever cast the concrete to match the camber of the beam?

 

[EngDM]

Pouring concrete will level itself (to an extent), it doesn't just follow the camber of the beam? At the highpoint of camber your concrete thickness will be less than at the ends, but the top of concrete should be consistent.

 

[Tomfh]

The initial camber will typically be irrelevant in terms of overall deflection, as the top of the concrete can be poured flat and will cancel out the camber.

However, if people can look down the length of the beam from below the slab, then the initial 6mm needs to be added to the deflection calculation.

Common example is floor systems hidden above ceilings, where metal deck fromworm (eg bondek) is allowed to have higher than normal deflections, because no one can ordinarily see the deflections from above or from below.

 

[JoshPlumSE]

Why would you ever cast the concrete to match the camber of the beam?

I've got the same question. I feel like the OP either:
a) Doesn't understand that the top of slab will be level even if the beams natural camber is not fully taken up.
b) Or, he's not explaining the situation to us very well.

 

[SALTRAM4567777]

We have a mezzanine floor within a warehouse that will support heavy racking loads, which will be operated by a robotic system. The robotics supplier has specified a stringent deflection criteria, requiring that the floor deflection not exceed 10mm at any point. This is crucial for the smooth operation of the robotic system. The total load on the floor is 30kN/m², which includes the weight of the racks, goods, and other associated loads. Meeting this deflection limit is essential to ensure the reliable and accurate functioning of the robotic system.


"We face significant challenges in achieving the stringent deflection limits due to restrictions on beam depth. Cambering the steel beam is impractical given its short length, so we proposed cambering the concrete instead. However, this solution was not accepted due to practical concerns. We then highlighted the impact of fabrication tolerances, specifically the natural mill camber of 6mm for steel beams in our case.

To ensure a minimum 200mm thick concrete slab, we must orient the beam with a downward camber, making an upward camber impossible. Our structural analysis software indicates a global deflection of 10mm. Considering the fabrication tolerances, I explained that an additional 6mm deflection should be accounted for, resulting in a total deflection of 16mm.

However, the client disagrees, arguing that the beam will not deflect more than the 10mm predicted by the software, and therefore, the 6mm fabrication tolerance should not be added to the global deflection. I am seeking clarification on this matter

 

[Tomfh]

Why do you think the natural beam camber is going to be reflected in the top surface of the slab? What is going on here?

 

[Smoulder]

Need to understand and maybe also specify construction sequence. It's normal to finish slabs flat or with a small slope for drainage if needed. So top will be flat and soffit will follow the beam shape. Sometimes you give the contractor your calculation of deflection under dead load so they can plan their pour around it. If you've assumed beams aren't propped then you also state that otherwise the dead load deflection needs to be added to live load deflection in your case because it happens when the props are removed.

Slabs have consyruction tolerance. Maybe you should reduce structural deflection limit by the tolerance. If you've used all the 10mm and there's already a low point in the slab pour at midspan then the slab surface will be more than 10mm out of flat.

Also check 10mm beam manufacture camber. Sounds too big if the beams are too short to bend.

Anyway it sounds like client is right.

 

[JoshPlumSE]

We have a mezzanine floor within a warehouse that will support heavy racking loads, which will be operated by a robotic system. The robotics supplier has specified a stringent deflection criteria, requiring that the floor deflection not exceed 10mm at any point.

Okay. That is now more clear.

Our structural analysis software indicates a global deflection of 10mm

What does this mean? Is that the deflection of the non-composite (i.e. naked) steel beam under non-composite loads? Because, that deflection doesn't really matter.

Is it the 10mm deflection of the composite beam under loading that is applied AFTER the beam has achieved it's composite stiffness / strength?

It is VERY important that you understand what this deflection is and when each component of the deflection occurs. Only then can you understand how it impacts your client's design criteria.

My interpretation of your clients design criteria:
1) The robotics manufacturer cares about having a FLAT floor. Where that floor can't have any more than 10mm of variation in flatness. But, I don't think they care at all how you got there... Honestly, that's not all that a terribly strict flatness criteria.

2) Typically, whether the mill camber is up or down does not AT ALL affect whether the top surface of the slab is flat! Please try to understand this.

The slab is poured so that the top surface is flat within some construction tolerance. Gravity pretty much makes this happen since the concrete is reasonably fluid while it is being poured.

Any deflection that happens before they are finished pouring the slab will be totally irrelevant to the robotics folks.
Why? Because the robotic equipment has no idea that deflection ever occurred and will never be affected by it.

3) What the robots guy cares about is any load applied AFTER the slab has become composite. Meaning dead loads applied AFTER the slab was poured. Live loads including the weight of whatever robotic equipment is running around.

To ensure a minimum 200mm thick concrete slab, we must orient the beam with a downward camber, making an upward camber impossible. Our structural analysis software indicates a global deflection of 10mm. Considering the fabrication tolerances, I explained that an additional 6mm deflection should be accounted for, resulting in a total deflection of 16mm.

Why do you think it matters if there is a maximum of 6mm various in slab thickness?

First off, I'll suggest that much of this camber will flatten out when the concrete is poured if the camber is oriented upwards. Generally speaking, when you design a composite beam system you will try to specify a camber which is on the order of 70 or 80% of the pre-composite dead loads. In your case, you have rightly determined that cambering the beam is not economical since it is so short. That's fine. But, you still generally want to orient the natural mill camber upwards so that the beam flattens out under the weight of the slab.

This are concepts that should be familiar to anyone who understands the basics of composite beam designt. The fact that you don't seem to understand this is, to be frank, a bit of a red flag to me (and probably most experienced engineers reading this thread). I'm really hoping that that you're not the engineer of record for this project... and that there is a more experienced engineer in your office that can explain this to you.

Sorry to be so blunt about this. Perhaps there is a bit of a language barrier here that is causing the misunderstanding. Your English is actually pretty good, so my comments have presumed that this is NOT the case.

 

[SALTRAM4567777]

@Josh My understanding is same as what you have written above and explained. Maybe I am not able to explain. My question as mentioned in the beggining of the thread and I want a straightforward answer from you. My beam is non composite and with the natural mill camber erected in downward direction of 6mm( assume it's done like that). My analysis model shows a total deflection of 10mm(DL+ LL). Should I add this 6mm to my 10mm deflection if beam is erected with camber downwards. Lets forget about the flatness of slab for a moment.

 

[JoshPlumSE]

My analysis model shows a total deflection of 10mm(DL+ LL). Should I add this 6mm to my 10mm deflection if beam is erected with camber downwards. Lets forget about the flatness of slab for a moment.

The key is whether that 6mm is going to affect the flatness of the slab and affect the robotics. In my opinion, the ONLY deflection that matters (at least to the robotics) is whatever occurs after the slab has been poured flat. Therefore the 6mm doesn't matter. Neither does the deflection to to beam and slab self weight..... Only whatever loads have been poured AFTER the slab was constructed.

What I'm still not understanding is what in the world would anyone WANT to orient a beam like this with the mill camber pointed down..... You haven't explained that. I suspect that means you don't really understand this type of structure and how it should behave.

Also, if deflection is a major concern for you and your client, then why aren't you making this a composite beam floor system? It's not really all that expensive and it should dramatically improve the behavior of the floor system for whatever robotics will operate on the floor.

 

[EngDM]

I agree with Josh on this. Any deflection that occurs due to camber and the weight of the slab is irrelevant since it will self-level under gravity before it cures. The only deflection that matters in this case is whatever is imposed by the robotics equipment itself, and whatever it is carrying. Beam needs to be designed for total load in strength obviously.

 

[canwesteng]

Is there a senior engineer in your firm you can discuss this with? I'm with Josh, your understanding of deflection is not very solid, and I think you should have some senior mentorship before taking this much further. Camber has no effect on deflection, and camber has no effect on slab flatness.