Splice beams

[Brutus76]

Hi there, i am in the process of removing 2 small walls from the corner of my little living room and i wanted to make sure that the roof and floor above does not collapse.. We have a standard edwardian house in london where a loft extension has been done above the living room, and new floor joists put in. The joists measure 220 by 47mm. I have taken some pictures of the "corner after we remobed the old lath, plasterboards, some insulation etc. the problem is that the new beams are cut above the small old beam for stud wall. Since they are cut i put in some splices also with 220@47mm timber. The question is how strong are such splices normally when we remove beam below that they currently lie on? The length of span is 4 metres on one side and 2 on the other. Thanks for input!

Untitled

• Brutus76
• 23 Aug 2013
• 1

Untitled

• Brutus76
• 23 Aug 2013
• 1

Untitled

• Brutus76
• 23 Aug 2013
• 1

Untitled

• Brutus76
• 23 Aug 2013
• 1

 

[catlad]

Them joists will certainly be taking support from that door way.

 

[i_am_fubar]

From what I'm reading, you're asking if you can take the wall out and turn what is currently a 4m and 2m span into a single 6m span?

For starters, even if it was a single, piece of 220x47, that wouldn't be enough for a floor on that span:
http://www.bsw.co.uk/uploads/files/bsw_timber_c16_span_leaflet_a4_4pp240810.pdf

The center deflection of a beam is a function of span^4. So even a short increase in length has a significant impact of how bendy your floor will be. adding 2m is not trivial and I would imagine steels will be needed.

The size of the timbers on that wall indicate it was load bearing for the roof and should have probably been upgraded when the loft was re-done.

If you have just joined two joists together with a splice, and take the wall away.... Best case, your floor will droop. Worst case, the splices will split and the entire floor will collapse. Possibly pulling the roof and outer walls as well.

Also, for structural purposes, nails have a higher shear strength than screws (I see turbo golds or similar in pics). However, joists should be bolted together (M10+) with dog collars between timbers and large washers.

Fubar.

 

[i_am_fubar]

It looks like the 220x47's are doubled up. What exactly are the splices joining?

Fubar.

 

[Brutus76]

It looks like the 220x47's are doubled up. What exactly are the splices joining?

Fubar.

Hi, thanks a lot for proper answer! Yes, understood for nails vs scews *just difficult to get a hammer in there..
Yes, understood for the length of 6meters, exactely correct that the length span would be that length, and yes it would be pretty extremely bad if it falls down of course..

The reason it looks like 2 or more joists is that the older looking one, darker, is only 11cm or so tall, vs 22cm for the newer, and used to hold up the old ceiling before a loft extension was put in.. I have screwed the old and new together to add some stiffness, i did not take the old ones out, but perhaps that does not make any difference. There are a few places where there are 2x new joists bolted together like you say, but that is not for the entire floor area.

I thought about using some metal beam before, but my wife wants the ceiling to be entirely flat, with no supporting beam cut/out below..any suggestions for how to do that^

again, thanks for having a look

 

[i_am_fubar]

Having recently learned some basic structural calcs, I'll give you the hows and whys before the whats.

A beam (lintel, joist, scaffold plank, floor board... etc) has a property called 'Second Moment Of Inertia' which is that beams shape and sizes ability to resist bending. The unit for SMOI is given as mm^4 and for a square / rectangular beam is given as:

(W x (H^3)) / 12.

As you can see from that, if you double the width (such as bolting 2 joists together), you double the SMOI. If you double the height, you increase the SMOI by 8.

Once you have a beams SMOI, you can calculate the maximum center deflection by using the calcs on this website:

http://www.engineeringtoolbox.com/beam-stress-deflection-d_1312.html

The 'Youngs Modulus' or 'Modulus of elasticity' for steel and C16 (structural) timber are around:
Timber: 6,000N/mm^2
Steel: 200,000N/mm^2

As you can see, and as makes sense, Steel is much stronger than timber.

If you sandwedge a 6mm piece of steel between 2x 220x47. You get the equivalent deflection figures as a:

47+47+((200000/6000) x 6) = 47+47+200 = 220x294

So with minimum increase in overall size of the joist, you make it over 3 times stronger.

Alternatively, if you run the maths for a piece of I beam steel or box section steel (that you could fit between the joists) using these calcs for SMOI:

http://civilengineer.webinfolist.com/str/micalc.htm

you may find you can get away with installing steels between the existing joists.

So, really, your options would be:
1) Sister the joists up with steel between them (there is a fancy word for them that escapes me.

2) Install a steel beam across the room to split it into 2x 3m spans.

3) Install additional steel beams (probably I beams) between the existing joists for the whole span

Of all of these, 3 would probably be the easiest, but even then, 6m is a damn big span and this is going to be a BIG job regardless of how you go about it.

Unless you have the money to spend, I would consider keeping a pillar next to where the door is and the horizontal beam.

As to what the wife says. I tell mine "you are in charge of how it looks, but I am in charge of how it's build". It's better to have a slightly ugly room than a pretty pile of rubble.

!!!PLEASE NOTE!!!!

I am not a structural engineer.

I'm supplying this information for informative purposes only so you understand the situation.

There WILL be other factors to consider.

Please get a structural engineer in because this is a prime opportunity for massive structural failure. This may seem hypocritical considering my current project, but worst case if mine fails, the roof will bow and leak. Worst case for you, the house could come down.

Fubar.

 

[Brutus76]

Having recently learned some basic structural calcs, I'll give you the hows and whys before the whats.

A beam (lintel, joist, scaffold plank, floor board... etc) has a property called 'Second Moment Of Inertia' which is that beams shape and sizes ability to resist bending. The unit for SMOI is given as mm^4 and for a square / rectangular beam is given as:

(W x (H^3)) / 12.

As you can see from that, if you double the width (such as bolting 2 joists together), you double the SMOI. If you double the height, you increase the SMOI by 8.

Once you have a beams SMOI, you can calculate the maximum center deflection by using the calcs on this website:

http://www.engineeringtoolbox.com/beam-stress-deflection-d_1312.html

The 'Youngs Modulus' or 'Modulus of elasticity' for steel and C16 (structural) timber are around:
Timber: 6,000N/mm^2
Steel: 200,000N/mm^2

As you can see, and as makes sense, Steel is much stronger than timber.

If you sandwedge a 6mm piece of steel between 2x 220x47. You get the equivalent deflection figures as a:

47+47+((200000/6000) x 6) = 47+47+200 = 220x294

So with minimum increase in overall size of the joist, you make it over 3 times stronger.

Alternatively, if you run the maths for a piece of I beam steel or box section steel (that you could fit between the joists) using these calcs for SMOI:

http://civilengineer.webinfolist.com/str/micalc.htm

you may find you can get away with installing steels between the existing joists.

So, really, your options would be:
1) Sister the joists up with steel between them (there is a fancy word for them that escapes me.

2) Install a steel beam across the room to split it into 2x 3m spans.

3) Install additional steel beams (probably I beams) between the existing joists for the whole span

Of all of these, 3 would probably be the easiest, but even then, 6m is a damn big span and this is going to be a BIG job regardless of how you go about it.

Unless you have the money to spend, I would consider keeping a pillar next to where the door is and the horizontal beam.

As to what the wife says. I tell mine "you are in charge of how it looks, but I am in charge of how it's build". It's better to have a slightly ugly room than a pretty pile of rubble.

!!!PLEASE NOTE!!!!

I am not a structural engineer.

I'm supplying this information for informative purposes only so you understand the situation.

There WILL be other factors to consider.

Please get a structural engineer in because this is a prime opportunity for massive structural failure. This may seem hypocritical considering my current project, but worst case if mine fails, the roof will bow and leak. Worst case for you, the house could come down.

Fubar.

Hey, thanks again for this. Really appreciate it!

 

[tony1851]

The center deflection of a beam is a function of span^4.

span^3

 

[tony1851]

Having recently learned some basic structural calcs, I'll give you the hows and whys before the whats.

A beam (lintel, joist, scaffold plank, floor board... etc) has a property called 'Second Moment Of Inertia'

'Second Moment of AREA', aka 'Moment of Inertia'

 

[i_am_fubar]

Can I get away with calling a typo on the ^4?

And what is this, English class? I got my mucking words fuddled, please don't take points off

And hey, if that's the only fault, I'm rather happy with that post

Fubar.

 

[i_am_fubar]

And double checking...

For a distributed load, it's F(Length^4)

For a centre point load, it's F(Length^3)

Or are my sources off?

 

[tony1851]

Can I get away with calling a typo on the ^4?

And what is this, English class? I got my mucking words fuddled, please don't take points off

And hey, if that's the only fault, I'm rather happy with that post

Fubar.

Sorry, couldn't resist being pedantic!
You've picked it up quicker than I did - took me years

 

[tony1851]

And double checking...

For a distributed load, it's F(Length^4)

For a centre point load, it's F(Length^3)

Or are my sources off?

No.
For a uniformly distributed load, its;

5 x W x L^3/384 x E x I, where W = the total load.

(The confusion here is because most people work out the load per meter, (small 'w') so that part of the equation does indeed become w x L^4; I just work out the total load [capital 'W'] so I always think W x L^3)

For centre point load, it's W x L^3/48 x E x I.

Hope this clears it up

 

[Brutus76]

And double checking...

For a distributed load, it's F(Length^4)

For a centre point load, it's F(Length^3)

Or are my sources off?

No.
For a uniformly distributed load, its;

5 x W x L^3/384 x E x I, where W = the total load.

(The confusion here is because most people work out the load per meter, (small 'w') so that part of the equation does indeed become w x L^4; I just work out the total load [capital 'W'] so I always think W x L^3)

For centre point load, it's W x L^3/48 x E x I.

Hope this clears it up

Hey guys, no fighting, you are all great that take time to look at this! Really appreciated for this one / I think I was close to having a little disaster here.. I called over a struct.engineer that said an steel beam is definitely needed if I want to remove that wall under my little wood splices. One problem arose though since he said I need 15cm height of the H type RSJ, and I really would want to avoid that to minimize ceiling height loss. Do you think I could use any other type of beam, e.g. a solid one, or rectangular one? Thanks

 

[tony1851]

Hey, nobody's fighting!

Fubar was right, 6m is a long span and you won't do that with a timber beam.
He also mentioned deflection, and that will be the critical factor here.
The beam your SE has recommended seems to be a 152 x 152 column ('H') section.
It is dubtful that you will get any beam shallower than that and still keep deflection within limits.
I suspect you will have to live with a downstand beam.