# Steel beam sizing

#### twillmo

We're looking to remove a load bearing wall, which is supporting one wall of a gabel, the floor above, ceiling above and roof above. I've sent our structural engineers calcs to my dad and we're both a bit concerned.

The above image shows the wall (and chimney to be removed). The wall is supporting the joists from Room 2, and the section of Room 1 to the right of the existing steels (supporting another gable).

The values the engineer has used for his calcs is here:

And he has come up with requiring a 203x203 UC46.

Some of my/our worries are:
- The floor loads only take in to account the floor for Room 2, and not the section of Room 1 who's joists are supported between the existing steels and the wall to be removed.
- The wall dead load is for a 2m high wall. The wall is at least 2.5m tall above the steel.
- There are no calculations for the chimney. He has stated that gallows brackets are fine.
- The roof load is for just the width of the room, and doesn't take in to account the steep pitch of the roof.
- Ceilings and stored items in the loft havn't been taken in to account.
- My old man thinks the weight required for the roof is too small.
- He's calculated the span as 5.2m, when it's clearly more than that.

Now I'm no structural engineer, but some of the above points seem quite worrying, and I wouldn't imagine they're included in the 1.4 and 1.6 safety factors that are used in the calculations.

Is anyone able to put my mind at ease and let me know what steels they used for their houses? Or if any structural engineers know that the specified steel is roughle correct?

Obviously I'll follow up with the engineer, but the fact that he's clearly missed some stuff is pretty worrying.

Any help is greatly appreciated.

Edit - more calcs:

Last edited:

#### tony1851

Wall load at 3.6 kN/m² seems a bit low. If it's an old building with 9" walls, something like 5.0 - 5.3 would be more appropriate. and of course 2m high is clearly wrong. If it's a modern wall with lightweight block internally 3.6 would be about OK.

Even so and allowing for some omissions, the UC specified would almost certainly be OK, though deflection might be an issue.

Hopefully the other number-crunchers will rock up soon and perhaps shed more light on it

#### twillmo

The walls are cavity walls, build in the 90's, so hopefully not too heavy. I've tried scouring the internet for recommended loads, however haven't managed to find anything.

Even so and allowing for some omissions, the UC specified would almost certainly be OK, though deflection might be an issue.

Thanks for that, it's good to know it should be fine. I've attached the rest of the calcs for the beam below. It looks like a deflection of only 4mm, which doesn't sound like a lot to me!

#### Dereekoo

Perhaps that the SE has wide ranging assumptions to get the loads he has, it may be that he extensive experience/knowledge and knows that his figures are very conservative, but no one on the forum will be able to vouch for their accuracy without a lot more details of the structure and/or his maths.

Assuming his total loads of 14.8kn/m is multiplied by a average safety factor of 1.5 then the resultant bending moment within the beam assuming it is simply supported would be 14.8 x 1.5 x 5.2 x 5.2/8 = 75knm. From "The Blue Book" tables the moment capacity of this size of beam is 137knm which when reduced to take the buckling resistance of a 5.2m beam into account gives 97knm so on the face of it appears more than adequate. The foregoing is a really simplistic look at what you have presented and should not be taken as gospel

Edit - Would have been nice to have seen his calcs in the original posting

Last edited:

#### twillmo

Perhaps that the SE has wide ranging assumptions to get the loads he has, it may be that he extensive experience/knowledge and knows that his figures are very conservative, but no one on the forum will be able to vouch for their accuracy without a lot more details of the structure and/or his maths.

Assuming his total loads of 14.8kn/m is multiplied by a average safety factor of 1.5 then the resultant bending moment within the beam assuming it is simply supported would be 14.8 x 1.5 x 5.2 x 5.2/8 = 75knm. From "The Blue Book" tables the moment capacity of this size of beam is 137knm which when reduced to take the buckling resistance of a 5.2m beam into account gives 97knm so on the face of it appears more than adequate. The foregoing is a really simplistic look at what you have presented and should not be taken as gospel

Edit - Would have been nice to have seen his calcs in the original posting

Yes, that's what I was assuming that he was using average load figures and any small descrepancies were covered by the safety factors. But it's a bit job and I just wanted to make sure the house wasn't going to fall down! Thanks very much for talking me through the calculations, that's been really helpful.

#### RonnyRaygun

The other thing here - and I’m pretty confident I’m right, although I haven’t checked the calc, (and this is based on the fact the SE has already made a few mistakes)…

4.4mm deflection seems very small for a 5.2m span carrying wall, floor, and roof - even allowing for the less than adequate loads assumed.

On TEDDS, (the software they have used) by default dead load deflection isn’t considered. I have a suspicion that the option to include dead load deflection hasn’t been ticked - quite a glaring error for a beam carrying an existing wall above, a wall that might crack if the beam deflection is too high. I usually limit deflection to around 14mm for dead and live loads combined when supporting an existing structure.

Might still be within limits but I think it’s going to be a lot tighter…

Last edited:

#### RonnyRaygun

Yeah, I think the deflection for dead and imposed load combined is more like 15mm (more when the correct loads are applied). That’s getting a bit tight and you might need a bigger beam…
Another calc shows it is indeed 4.37mm for live load deflection.

Go back to your SE and ask them to do the calcs themselves this time, not the work experience boy / girl

#### jks7492

On TEDDS, (the software they have used) by default dead load deflection isn’t considered. I have a suspicion that the option to include dead load deflection hasn’t been ticked - quite a glaring error for an existing beam carrying a wall above, a wall that might crack if the beam deflection is too high. I usually limit deflection to around 14mm for dead and live loads combined when supporting an existing structure.

Not necessarily a mistake - span/360 against imposed loads only is a valid deflection limit check.

#### RonnyRaygun

Not necessarily a mistake - span/360 against imposed loads only is a valid deflection limit check.

It is a valid check, but you should really be limiting dead and imposed load deflection to span / 250, less when carrying an existing wall above, a check that also hasn’t been carried out.
I was asked to check a very similar calc once where the beam (spanning much further than 5m) was already in and had deflected about 50mm causing significant cracking to the wall above. It was exactly the error I have described above and the client ended up having to have a post in the middle of their kitchen extension, a post they really didn’t want…

#### jks7492

It is a valid check, but you should really be limiting dead and imposed load deflection to span / 250, less when carrying an existing wall above.
I was asked to check a very similar calc once where the beam (spanning much further than 5m) was already in and had deflected about 50mm causing significant cracking to the wall above. It was exactly the error I have described above and the client ended up having to have a post in the middle of their kitchen extension, a post they really didn’t want…

Yes agreed.
My point was that just because the tedds output has been set up to check span/360 against imposed loads only, doesn’t necessarily mean that the design is wrong and the total deflection isn’t within an acceptable limit.

#### twillmo

Thanks for checking that! I'll give him a call on Monday and talk through all the points raised here.

With the additional chimney load on the gallows I imagine the deflection would be even greater.

If we needed a bigger beam, would it have to be taller, or would we be able to upgrade to a UC 60kg?

#### RonnyRaygun

If we needed a bigger beam, would it have to be taller, or would we be able to upgrade to a UC 60kg?

Heavier 203 UC should be OK. Around a third less deflection with a 203 UC 60 and even less if you go for a 71 or 86kg/m section.

#### twillmo

Heavier 203 UC should be OK. Around a third less deflection with a 203 UC 60 and even less if you go for a 71 or 86kg/m section.

Brilliant! Thanks very much. Given we want flush ceilings 203 is the maximum height we can afford. I'll update next week how the chat goes, and update with pictures hopefully next month when it's installed!

#### tony1851

For those of us anoraks fascinated by beam dead-load deflection, there's an interesting commentary here:

The main article is about how to connect twin beams when supporting a cavity wall, but there are some excellent points in the comments section on how SEs deal with deflection.

#### RonnyRaygun

Given we want flush ceilings 203 is the maximum height we can afford.

just be aware the the 46kg section is the only “true” 203mm deep section. The heavier ones get progressively deeper. Hopefully not enough that you can’t hide it in the finishes.