Can a mild steel floor plate be considered fixed when bolted to steelwork @ 300mm centres

[CJS1878]

Hi All,

An existing structural steelwork floor has an 8mm thick mild steel floor plate bolted on top of the structural steel beams. The existing calculations under loading fails when the plate is taken as simply supported and so was assess as fixed where the check was O.K. I am unsure on how I can confirm whether or not the plate can be taken as fixed if bolted at maximum centres of 300mm? I have not been able to find much advice online. The load is a point load at the centre and not a UDL. The floor plates are rectangular and bolted on all 4 sides.

 

[GeorgeTheCivilEngineer]

Could you provide a sketch?

I can't tell whether this is a shear flow problem, a continuous beam/2-way spanning plate type problem or a catenary plate problem.

Edit - I think this is a two-way spanning type problem. Is the plate continuous over the support beams or are they all discrete plates fixed at 300mm centres?

 

[human909]

What sort of load is causing 8mm plate to fail? What span?

I'd expect bolted plate to behave in a semi rigid way however but I'd generally model it as pinned. That is what I have done myself.

 

[CJS1878]

Hi, thank you for the reply. The floor plate has conservatively been taken as a 1000mm width section and not assessed as 2 way spanning. I have attached a sketch I hope gives a better idea of what I am looking at. [URL unfurl="true"]https://res.cloudinary.com/engineering-com/image/upload/v1691502106/tips/sketch_of_floor_plate_detail_xs2q0k.pdf[/url]

 

[GeorgeTheCivilEngineer]

Thanks for the sketch. As the ratio is near 1:10 then this should be treated as a one way spanning plate.

I would treat this arrangement as as simply supported. A single line of bolts, no matter the spacing, wouldn't get to a fixed behavior - even though in reality it probably does fix it somewhat.

I would assess the span as the actual clear span (from the edges of the support beam flanges) and not the dimension of the plate or the line of the bolts. But if it still fails then that is your answer.

Edit - in what way does it fail?

 

[CJS1878]

Hi All,

@human909 - that is my concern that we cannot consider the plate as 100% fixed. Simply support is a conservative assumption but in reality we need to somehow look somewhere in the middle...

@GeorgeTheCivilEngineer - The steelwork floor is existing, the original calculations show the floor plate fails under a plate thickness check when taking into account the moment applied to the plate. As simply supported the plate also fails in deflection with a limit of span/100. The original calculations therefore check the floor plate as fully fixed where all checks are O.K. and it is stated that M12 bolts are required at max 300mm centres to make it "fixed".

The steelwork floor has a change of purpose and new loading is being applied. The point load is 10kN. the calculated moment for simply supported, only taking into account the point load and self weight of the plate, is 3.76kNm. The thickness required for this is 9.06mm which is > 8mm. Total deflection is 33.5mm which is about 310% utilised against a limit of span/100.

When considering the plate as fully fixed thickness required is < 8mm so O.K. and the deflection is 8.3mm so O.K. However, if we go against the original calculations and do not assume the plate is fixed, we cannot accept the new loading conditions without other modifications to help spread the load.

 

[Enable]

Depends on how ambitious you are in trying to get this thing to pass on paper without modifying it. I would probably just modify it since, well, it's my stamp but not my money.

But if we're in the feeling ambitious but not too silly category then I submit you look at the current loading as a sort of proof test. Reverse calculate the relative fixity of the connection by seeing what level of fixity is required under that load to achieve the observed deflection. Now apply that same relative fixity to the new loading and see if it passes muster. If so, then great, and if not, then reinforce or do some welding to make it a fixed connection. Or stiffen the plate with ribs or whatever.

 

[human909]

The steelwork floor has a change of purpose and new loading is being applied. The point load is 10kN. the calculated moment for simply supported, only taking into account the point load and self weight of the plate, is 3.76kNm. The thickness required for this is 9.06mm which is > 8mm. Total deflection is 33.5mm which is about 310% utilised against a limit of span/100.

I get a different result with Roark Formula and a second check with FEA. The stress is satisfactory ~175-200Mpa depending on your criteria and the deflection is around 20m (L/70), whether that is satisfactory or not depends on the circumstance at least in my book.

I'm not sure how you are calculating the load and the demand for the plate but given you seem to have approached using a moment I believe you are being excessively conservative. Links here for some more suitable calculators and formulas.

https://www.engineersedge.com/flat_plates_equations_calculators.htm

https://www.efunda.com/formulae/solid_mechanics/plates/calculators/SSSS_PPoint.cfm

I'd take the semi rigid connection at the edges as a comforting bonus and call it good assuming there are no other loads or risks.
*However there is potentially alot more at play here so without further information I'd want to know more.


My recent experience on this was moving a 180kN machine across an existing 6mm floor. A moving load with no guarantee that the load will spread nice and evenly across the 4 rollers we were using. I designed for a 90kN load in a concentrated area. Needless to say I added a fair bit of temporary plate. (As well having to upsize all the existing joists.) To my joy and a little surprise the whole process went super smoothly.

My calculated stresses were ~250mPa which is pretty high. But I was comfortable with that as my chosen load was pretty much an upper bound and conservative.

 

[canwesteng]

human has hit this nail on the head. You've assumed a width of plate tributary that is possibly too conservative. A yield line analysis is the least conservative way to approach this. That being said, deflection is unlikely to work for the plate if you need to use yield lines for strength.

 

[Tomfh]

I get numbers similar to human909. The plate is strong enough but does sag quite a bit under the 10kN.

Is it a real 10kN load? What's the real loading the floor sees? If the floor truly has 1tonne point loads then no way I'd be sitting that on 8mm plate spanning 1400mm..