Structural Steel Grades

[i_am_fubar]

Hey folks, quick one here I hope.

I've been calculating the deflection of a proposed lintel - 100x100mm box section steel. And the figures are working out well within limits.

Are there any regs that state that only specific section types and/or grades of steel can be used as lintels?

This steel is structural grade to EN 10219 seam welded square tube and will be treated by a spray on corrosion inhibiter. It will be bedded on a thin mortar bed 150mm either side of a 2400mm opening.

Cheers,
Fubar.

 

[rjm2k]

I would have thought that to comply with building regs etc your SE would be able to say?

 

[jeds]

No specific regs. I often specify square hollow section. It's cheaper than Catnic type box sections and you tend to get a lot more work out of them. The only criteria is calculating the loads correctly and selecting the correct section.

 

[i_am_fubar]

Cheers Jeds, just what I needed to know.

To ensure I could use 100x100 section, I ran the calcs on 10mm wall box with 10kN center point load. Deflection came out under 3mm if memory serves (so better than 1 in 600). As soon as I've finished the snow, dead and live load calcs for the rest of the roof I will re-do the maths with more correct loads and see how overkill 10mm is. But then if the price isn't that much more, I may just go with 10mm anyway (can get trade rates on steel section through work).

Fubar

 

[RonnyRaygun]

Cheers Jeds, just what I needed to know.

To ensure I could use 100x100 section, I ran the calcs on 10mm wall box with 10kN center point load. Deflection came out under 3mm if memory serves (so better than 1 in 600). As soon as I've finished the snow, dead and live load calcs for the rest of the roof I will re-do the maths with more correct loads and see how overkill 10mm is. But then if the price isn't that much more, I may just go with 10mm anyway (can get trade rates on steel section through work).

Fubar

Don't bother with snow loads, in Hampshire they will not be more onerous than the live (access) load for a roof, which is 0.6kN/m2, (although this can be reduced further if the roof pitch is over 30 degrees).

As Jeds says, any section type can be used. SHS only comes in S355 which is stronger than S275, although the steel grade makes no difference to the deflection calc.

You will get double the deflection capacity from a 100x100x10mm compared with a 100x100x4mm section.

 

[i_am_fubar]

Cheers Ronny,

The roof is for access not limited to maintenance. So live loads will be higher, and pitch at minimum to allow run-off, but I can see your point, not going to have people AND snow (deep enough to act as a substantial load at least) on there at the same time.

You say it makes no difference, but I would have though different steel grades would have different elasticities? Although deviation would be tiny and irrelevant for application, there would be some difference when the numbers are run.

Would have thought 10mm wall would give you more than double that of 4mm, or does the self weight start to give diminishing returns?

Fubar

 

[tony1851]

Would have thought 10mm wall would give you more than double that of 4mm, or does the self weight start to give diminishing returns?

'I' springs to mind

 

[RonnyRaygun]

Cheers Ronny,

The roof is for access not limited to maintenance. So live loads will be higher, and pitch at minimum to allow run-off, but I can see your point, not going to have people AND snow (deep enough to act as a substantial load at least) on there at the same time.

OK, so you need to take an imposed load of 1.5kN/m2. The British Standards and Eurocodes both agree that you only need to apply the higher load of snow or imposed.

You say it makes no difference, but I would have though different steel grades would have different elasticities? Although deviation would be tiny and irrelevant for application, there would be some difference when the numbers are run.

No, UK steel always has the same modulus of elasticity (205000N/mm^2) regardless of its tensile and compressive capacity. There will almost certainly be some variation but it's the same for all grades of structural steel.

Would have thought 10mm wall would give you more than double that of 4mm, or does the self weight start to give diminishing returns?

Self weight should be included in the calcs, but it will have little effect on the total deflection. The most important thing in increasing the deflection capacity of a member is depth, and by increasing wall thickness you aren't increasing the depth.
Double the depth of a solid section and you increase its stiffness 8 fold. Double the width and you only double the stiffness.

It's pretty simple to work out the second moment of area (Ix-x) of a square section. Basically breadth x depth cubed / 12.

So to work it out for a box section, work out Ix-x for the square, and take away the empty part. We'll ignore the radii at the corners for simplicity.

100x100x4 SHS: (100x100^3 / 12) - (92x92^3 / 12) = 2363392mm4
100x100x10 SHS: (100x100^3 / 12) - (80x80^3 / 12) = 4920000mm4

So according to my simplistic calcs the 10mm section is slightly more than twice as stiff, although I've just checked the Blue Book (which will have allowed for radii), and it gives 2320000mm4 for 4mm and 2620000mm4 for 10mm thick box. Almost exactly double.

 

[i_am_fubar]

Cheers again Ronny

I'm a bit wined up at the moment and will give a more in depth reply tomorrow. But I feel I need to query your figures for the sake of my sanity.

Agree with this as the void will be 80x80:
100x100x10 SHS: (100x100^3 / 12) - (80x80^3 / 12) = 4920000mm4

But the hollow on 4mm wall is 92, so this:
100x100x4 SHS: (100x100^3 / 12) - (96x96^3 / 12) = 2363392mm4

Should surely be this ?:
100x100x4 SHS: (100x100^3 / 12) - (92x92^3 / 12) = 2363392mm4

Also, from your little blue book you give 2320000mm4 for 4mm and 2620000mm4 for 10mm, which is only a 1.12 ratio. Should one of those 2's be a 4 at the start?

I imagine these are probably just typos, but I want to double check my methods are right.

You also say that increasing the wall thickness doesn't increase the depth. Maybe not the depth of the members envelope, and the side walls are only contributing a linear increase in moment. But wouldn't the thickening of the top and bottom of the member add towards the cubed power factor of the moment calculation? I'll run the numbers for a better understanding tomorrow, though if you have a few words on it, that'd be awesome.

Cheers,
Fubar.

 

[i_am_fubar]

Either you've edited your post a bit, or I'm going crazy at this point. So I'll have a sleep on it and check in tomorrow.

Cheers.

 

[RonnyRaygun]

Either you've edited your post a bit, or I'm going crazy at this point. So I'll have a sleep on it and check in tomorrow.

Cheers.

I'm tired! Yup, edited the 96 to 92 before you posted.

And yes, 462, not 262. You get the idea

 

[RonnyRaygun]

You also say that increasing the wall thickness doesn't increase the depth. Maybe not the depth of the members envelope, and the side walls are only contributing a linear increase in moment. But wouldn't the thickening of the top and bottom of the member add towards the cubed power factor of the moment calculation? I'll run the numbers for a better understanding tomorrow, though if you have a few words on it, that'd be awesome.

Cheers,
Fubar.

Look at it this way. In a 100x100x10 SHS, you've effectively got:

A 100x100x4mm SHS
And a 92x92x6mm SHS
Both on the same neutral axis.

The 100x100x4 section is a bit deeper but has a thinner wall, which means their Ix-x value should be similar.

So the additional steel does contribute, just not as much as you might think.

 

[i_am_fubar]

Cheers Ronny,

Knocked together a little Excel sheet and for the same load and span, but with different wall thicknesses, got the following:

Wall Thickness (mm) | Deflection (mm)
1 | 18.52
2 | 9.54
3 | 6.56
4 | 5.07
5 | 4.18
6 | 3.59
7 | 3.17
8 | 2.86
9 | 2.62
10 | 2.44

Came out as:
deflection = 17.723 x (wall thickness^-0.881)

So, as you say, 10mm only gives half the deflection of 4mm. In this situation, it's a deminishing return to the power -0.881. Definitly one to bear in mind if I need to do this again.

Don't suppose you could put my mind to rest on another matter?

I'm planning on mounting a ledgerboard to the external skin of the house with M12 epoxy fixed studs.

Ledger board will be a 150x47mm pine with a 3x150mm piece of steel sandwedged between timber and wall.

Load per joist = 2843N (1422N each end)
Sheer strength of M12 St/st studding > 10kN
Compressive strength of clay brick (worst case) = 7N/mm^2

Innitial plan was to use 2 fixings per joist. So sheer isn't an issue, but that would impose 711N vertical load on each fixing if board to wall friction is ignored.

Effective area of 12x100mm = 1200mm^2

Therefore pressure onto brick from studding = 711/1200 = 0.5925N/mm^2

So, well within the pressure rating of the bricks, and once friction is taken into account (possibly with the inclusion of resin or other adheasive between steel and bricks). The total downward load is easily managed.

However, someone has pointed out that the vertical loading on the ledger imposes a tortional moment that will try and twist it off the wall. Is this a concern, or would be it essentially negated by the presence of the joists hung off it?

Ledger X = 50mm
Ledger Y = 150mm
Downward force (per meter length of ledger) = 1422*2.5 = 3555N
Moment acting on board = 3555*0.05 = 177.75Nm
Therefore - Outward force at top of ledger = 177.75/0.15 = 1.185kN

if (top) fixings are attached 50mm from top of board:
Tensile force on fixings = 177.75/(0.15-0.05) = 1.78kN

2.5 top fixings per m
Therefore - tensile load on each fixing = 1.78 / 2.5 = 0.711kN

Seems resin bonded M12 anchour can take over 10kN in tension, so I'm really not seeing a problem here :/

Cheers for all the help,
Fubar.

 

[RonnyRaygun]

Wow Fubar, you really are going to town on this aren't you!

Why are you putting a steel plate between the timber and the wall? Usually the timber is bolted straight to the wall.

For the fixings I wouldn't worry too much about effective areas and calculations of bearing pressure onto the brick, just look in the Rawlplug catalogue (or whatever your local builders merchants stocks) and choose a suitable anchor. Just make sure you are reading from the correct column

Rawlplug give a combined tension and shear value for resin anchors into brick, and they are considered to be additive, so take the combined total and see what you need.

Not sure where you got 10kN tensile force from, but you might find it's less than that for brick or block.

If you are mounting joists on hangers, arguably your lever arm for tensile force is the thickness of the ledger board, plus the distance from face of ledger to centre of hanger...which will be a bit more than the 50mm you took.

As an aside, if calculating the pressure onto the brick, the pressure distribution won't be linear. It'll be positive at the outer face, negative at the inside end of the stud, and zero in the middle.
Although friction is ignored it will take care of much of the force.

Most builders on here would say just use M12 through bolts or resin anchors at 400mm crs. They would probably be right too, but no harm in doing what you are doing, and making sure.

 

[tony1851]

He's after your job, Ronny