Hi, I am in the process of building an intermediate floor in steel, specifically VKR (rectangular hollow section) which will be painted white and covered with glass among other things.
5 x 5 meters with support on 3 sides and no load-bearing partition walls either above or below.
3 sides because there is a staircase along the 4th wall.
(The actual dimensions are somewhat smaller with a maximum dimension of 4.8 and rhombus-shaped but have been rounded up for simplicity)
I have done a simple calculation but am just an amateur.
I think it must not be undersized due to all the glass and visually it should not look too flimsy.
I concluded that with
VKR 140x70x5 S355J2H and a point load (case 4) of 1000N, the deflection is ~9 mm
However, I am unsure if I have underestimated the weight (force).
The beams will have a center-to-center distance of about 1000mm
Can anyone spontaneously say "Yes, it is over-dimensioned and will work fine"?
Or perhaps another recommendation?
5 x 5 meters with support on 3 sides and no load-bearing partition walls either above or below.
3 sides because there is a staircase along the 4th wall.
(The actual dimensions are somewhat smaller with a maximum dimension of 4.8 and rhombus-shaped but have been rounded up for simplicity)
I have done a simple calculation but am just an amateur.
I think it must not be undersized due to all the glass and visually it should not look too flimsy.
I concluded that with
VKR 140x70x5 S355J2H and a point load (case 4) of 1000N, the deflection is ~9 mm
However, I am unsure if I have underestimated the weight (force).
The beams will have a center-to-center distance of about 1000mm
Can anyone spontaneously say "Yes, it is over-dimensioned and will work fine"?
Or perhaps another recommendation?
A beam needs supports at the ends, a floor structure with beams cannot be supported on 3 sides.
Then there will be a strong beam supporting all the others on the side towards the stairs, but it's better if you lay beams along the stairs so they have supports on two sides.
It seems spontaneous that 140mm 5mm over 5 meters freely supported is weak and you will likely get more deflection than 9 mm.
Are you planning to lay glass as the floor? What does the glass weigh? Will people then walk on this? Furniture.
Normally, you calculate a beam with all forces and support reactions to arrive at the deflection.
You have fixed load (beam, floor) and free load (people, furniture). If you input all these into various formulas, you get results for what the supports should handle and how large they should be, as well as the deflection.
This is too difficult to answer as there is no sketch or drawing to refer to.
Then there will be a strong beam supporting all the others on the side towards the stairs, but it's better if you lay beams along the stairs so they have supports on two sides.
It seems spontaneous that 140mm 5mm over 5 meters freely supported is weak and you will likely get more deflection than 9 mm.
Are you planning to lay glass as the floor? What does the glass weigh? Will people then walk on this? Furniture.
Normally, you calculate a beam with all forces and support reactions to arrive at the deflection.
You have fixed load (beam, floor) and free load (people, furniture). If you input all these into various formulas, you get results for what the supports should handle and how large they should be, as well as the deflection.
This is too difficult to answer as there is no sketch or drawing to refer to.
I think that 1000N seems little. When designing a floor or equivalent, you usually have a uniformly distributed load and a point load at the worst spot in addition to the self-weight. I don't remember how large the loads are, but they are in the order of 1000 N/m2 and 1500 N point load. Then the loads should be weighted with different safety factors or partial coefficients. This is evaluated against collapse and maximum deflection. I have never calculated for houses, so I don't know exactly what applies. I have only used parts of Eurocode for the safety verification of machines. But my feeling is that 1000N is too little.
A VKR 140x70x5 with a c/c spacing of 600 mm and a 5 m span deflects L/300 under a distributed load of just over 3 kN/sqm, which is probably OK. However, the deflection becomes too large with that dimension. I suggest VKR 160x90x5 instead. Then the question is what movements the glass can withstand?
What criterion do you use for deflection? I'm curious and have always wondered how to evaluate it. The deflection criterion, L/250 if I remember correctly, I find is quite odd and rather meaningless. I don't understand why you should relate deflection to the span.J justusandersson said:
pjapen said:
Hi Pjapen, yes you are right. Only supports on two sides are of interest. The fact that there is also a wall on one side and a stairway on the other is completely irrelevant.
Ok, I googled a bit quickly and didn't find any normal load for intermediate floors, but saw that 3000 is often used with additional load like a wall, etc., but here there is no wall. But I realize that I should include distributed load in combination with point load. I thought point load is the worst case and should be enough if you take it seriously. Guess I'll get a fail on that. Thanks for the input
Best answer
I assume that a point load of 1 kN should not cause deflection greater than 1.5 mm regardless of the span. Then you have to try to estimate how much of the point load affects adjacent beams, depending on the construction. Deflection as a proportion of the span has been included in Swedish standards since BABS 46, with slightly different values and calculation methods. The knowledge base for this can probably be questioned. Just like 4.5 m to the plot boundary, which is a remnant from the Building Statutes for the Cities of the Realm from (§34).
Interesting! I have no experience in calculating houses or floor structures, but occasionally (with a few years in between) I do calculations for platforms on machines. The issue of deflection has come up and I haven't found any guidelines.J justusandersson said: