Long Span Beam Design for Optimal Seismic Performance

[rodsika]

Are there any structural engineers here or know the stuff. I'd like to know this.

Can structural engineers design long span beam for example 12 meters reinforced concrete beam that is as good as standard 6 meter beam in seismic performance or are 6 meter beams always better compared to 12 meters beam even if the latter were designed for ultimate seismic performance?

 

[Studiot]

For vibration analysis you have to consider the whole structure (including the foundations), not just the individual components separately since the connections play a great part in the result.

So it depends what your 6m and 12m beams are connected to and how they are connected.

 

[rodsika]

For vibration analysis you have to consider the whole structure (including the foundations), not just the individual components separately since the connections play a great part in the result.

So it depends what your 6m and 12m beams are connected to and how they are connected.

I mentioned "seismic analysis", you wrote "vibration analysis". Unless they are identical? In what ways?

 

[Studiot]

Vibration is just the more general term as there are other sources of excitation besides seismic activity.

For example the beams of motorway bridges are subject to continual vibration and are quite a bit longer than 12m.

That is less important what is important is the connection conditions of your beams.

Another important factor is the natural frequency of vibration.

 

[rodsika]

Vibration is just the more general term as there are other sources of excitation besides seismic activity.

For example the beams of motorway bridges are subject to continual vibration and are quite a bit longer than 12m.

That is less important what is important is the connection conditions of your beams.

Another important factor is the natural frequency of vibration.

In a building with 12 meters frontage. How much generally is 12 meter beam span more expensive compared to two 6 meter span in the overall cost of the construction, materials? Is it like 20-30% more expensive?

 

[Studiot]

Two 6m span beams would require a centre support that is not present with a 12m one so would add to the cost.

You have been asking about reinforced concrete. This is usually priced at a rate per cu metre placed so the cost would likely be the same for the beams either way. I don't see that any heavier duty craneage would be required 12m is not such a significant increase in size.

 

[rodsika]

Two 6m span beams would require a centre support that is not present with a 12m one so would add to the cost.

You have been asking about reinforced concrete. This is usually priced at a rate per cu metre placed so the cost would likely be the same for the beams either way. I don't see that any heavier duty craneage would be required 12m is not such a significant increase in size.

Supposed you would be given a choice by the architect/engineer whether you want a 6 meter or 12 meters RC beam span in your house or office building? What would you choose? I'm still thinking of two 6 meters RC beam span because there are 3 connections instead of 2. In construction they don't do it perfect or sigma 5 accuracy. Hence with the 3 connections in the 6 meter span.. at least there are 3 connections instead of 2... in this latter case of 2 connections, if there is some subtle weakness in one of the beam-column joint.. it could fail. In your country. Do you commonly see 12 meter RC beam span houses or office? Isn't 6 meter the standard?

 

[Studiot]

Most building component sizes are decided by architectural considerations not structural ones. If a longer span beam is specified by the architect then the structural engineer will design a suitably stronger one.

Beams of 6m span are probably so lightly loaded that they are only just above what we call 'minimum steel'. That is the steel required to control cracking.

It is unusual to use RC beams in the UK in any case, steel is more normal.
Suspended floors are normally RC slabs or prestressed.

 

[rodsika]

Two 6m span beams would require a centre support that is not present with a 12m one so would add to the cost.

You have been asking about reinforced concrete. This is usually priced at a rate per cu metre placed so the cost would likely be the same for the beams either way. I don't see that any heavier duty craneage would be required 12m is not such a significant increase in size.

I think you forgot that the rebars of longer beam span would be quite larger and there will be much more rebars in 12 meter RC beam span versus two 6 meter RC beam span so the cost of the former would be higher.

 

[Studiot]

I think you forgot that the rebars of longer beam span would be quite larger and there will be much more rebars in 12 meter RC beam span versus two 6 meter RC beam span so the cost of the former would be higher.

No I didn't forget that.

I said that under standard building contracts reinforced concrete is priced at a rate per cu metre placed. That is the UK standards method of measurement.

Civil engineering contracts are priced differently so the reinforcement, concrete and formwork are all measured separately.

It is actually unlikely that there would be a greater number of reinforcing bars in the longer beam. Just that they would be larger diameter.

 

[rodsika]

No I didn't forget that.

I said that under standard building contracts reinforced concrete is priced at a rate per cu metre placed. That is the UK standards method of measurement.

Civil engineering contracts are priced differently so the reinforcement, concrete and formwork are all measured separately.

It is actually unlikely that there would be a greater number of reinforcing bars in the longer beam. Just that they would be larger diameter.

Consider a lot size of 12 meters frontage and 15 meters depth. Supposed they put 4 beams side to side and one girder at middle front to back. This means the front and back have columns at the center. Only two RC beams of 12 meters (without columns) are put at middle of lot with beam depth of 0.8 meters. Would they make the girder at middle (front to back) also with girder depths like 0.8 meters to support the whole 15 meters front to back lot size.. or would the girder be much smaller just to support the slabs? What is usually the case?

Note I'm not constructing any of this, of course but just want to understand the design of certain office near mine. Thanks.

 

[Studiot]

Do you mean something like this?

So the upper floor is divided into 4 slabs 6m x 6.5m.

 

[rodsika]

Do you mean something like this?

So the upper floor is divided into 4 slabs 6m x 6.5m.

Yes. But with the difference there are two 12-meter beams in the middle between the front and back making up 4 side to side beams (not 3 as your drawing shows).

Also the girder is not steel i-beam but RC beam and it has same level as the side to side RC beam.. meaning it is embedded like in normal column-beam joint so the definition of girder as something supporting smaller beams don't hold. I wonder if you also defined a girder as something just to support slabs and not exactly the beams (wherein the 12 meter RC beam can stand on their own).

 

[Studiot]

OK I think I understand.

There are 3 columns at the front, 3 at the back, and 4 down each side, making 10 in all. avoiding counting corners double). They are presumably evenly spaced. They are connected by a system of rc beams at the top. This network of beams divides the support for the upper floor into 3 strips 4.3m x 12m. When complete this makes an rc frame.

I would guess that the upper floor might be then constructed by lifting precast beams or floor units to span 4.3m onto the frame and screeding over the top. That would avoid slab formwork and 4 - 5 m precast beams are not that large and readily handleable by a small crane or even a digger arm. If the floor units can be cast off site or on site whilst the rc frame is being constructed there will not be some weeks waiting time for the floor concrete to cure.

 

[rodsika]

OK I think I understand.

There are 3 columns at the front, 3 at the back, and 4 down each side, making 10 in all. avoiding counting corners double). They are presumably evenly spaced. They are connected by a system of rc beams at the top. This network of beams divides the support for the upper floor into 3 strips 4.3m x 12m. When complete this makes an rc frame.

I would guess that the upper floor might be then constructed by lifting precast beams or floor units to span 4.3m onto the frame and screeding over the top. That would avoid slab formwork and 4 - 5 m precast beams are not that large and readily handleable by a small crane or even a digger arm. If the floor units can be cast off site or on site whilst the rc frame is being constructed there will not be some weeks waiting time for the floor concrete to cure.

Thanks. In my place. We don't use precast slabs. But simply pour fresh concrete into the floor support. Anyway. The building will be two storey. They plan to use metal hip roof on the roof of the second storey. Now do you know how heavy a hip roof is compared to using pure concrete slabs or roofdeck? Like is a hip roof with full truss support 1/2 heavier or 1/4 heavier than pure concrete roofdeck or same weight given the same building size?

 

[Studiot]

Rough guess

75 to 100mm concrete slab 240 kg per sq metre = 180*12*13 to 240*156 = 25tonne to 35 tonne

metal sheet roof 6 - 10 kg /sqm plus 4 trusses at 750kg each = 3 to 5 tonnes

 

[rodsika]

Rough guess

75 to 100mm concrete slab 240 kg per sq metre = 180*12*13 to 240*156 = 25tonne to 35 tonne

metal sheet roof 6 - 10 kg /sqm plus 4 trusses at 750kg each = 3 to 5 tonnes

Thanks. I read today that irregular size lots can experience torsion as a result of unequal lateral movement. Pls see attached picture. Supposed the lot we talked about has the left side 14.57 meters and the right side 17.29 meters.. that is the right side is 2.72 meters longer. How much torsions would be produce for earthquake or seismic movement that goes up and down, side to side and diagonal? Or can the size be considered rectangular approximations and torsions negligible? If not. Can you design columns and beams that would make the left and right side equal and preventing torsions?

 

[Studiot]

With a steel roof there would be greater twisting generated every day by solar heating and expansion.

The traditional solution (provision of movement joints and allowance in the supports) to this problem would also alleviate other movement issues.

 

[rodsika]

With a steel roof there would be greater twisting generated every day by solar heating and expansion.

The traditional solution (provision of movement joints and allowance in the supports) to this problem would also alleviate other movement issues.

I was not talking about the roofs twisting. I was talking of the RC beams and columns of one portion of building suffering torsions or rotations during seismic movement. It is said torsions contribute much to earthquake damages... something to do with center of mass and lateral movement and torsions.. you not familiar with this?

 

[rodsika]

Studiot, check this out:

http://www.ijens.org/Vol_12_I_01/125001-9494-IJCEE-IJENS.pdf

I was trying to apply it to the lot we talked about. See if you can apply it.

 

[Studiot]

No I am not familiar with the Turkish design code but it looks sensible

If the columns are more rigid than the beams, ductile
deformations will occur at the ends of the beams. Beams can
absorb a lot of energy by ductile deformations without an
important loss in the load carrying capacity. In this system, all
the beam–column connections in the building have to fail
before the collapse of the ground floor columns. Architects
should know that strong column–weak beam design is not
only advisable but also obligatory according to Codes

This is your key phrase from that link.

 

[rodsika]

No I am not familiar with the Turkish design code but it looks sensible



This is your key phrase from that link.

There is a phrase that I just can't connect with the lot drawing I attached earlier. It said "The center of building mass is generally considered as the geometrical center of the building and the center of rigidity is considered as the center of vertical elements of the structural system. The center of rigidity of a building should coincide with the center of mass. When the center of a building mass does not coincide with the center of rigidity, torsion and stress concentrations occur in the building when it is subjected to seismic loads."

So in the drawing a few message away. Where is the center of mass and center of rigidity? Do they coincide in the drawing?

Really dumb about structural...

 

[Studiot]

There is some explanation and alternative coping strategies in this american document.

http://richardpclarke.tripod.com/earthquakedesign.pdf

The basic strategy is to find some dissipate the earthquake energy within the ductility of the members - obviously without causing structural collapse.
You are now getting into significant detail which can't be answered without knowing the detailing of the structure particularly the connections as I said before.

The devil is in the detail (pun intended)

 

[rodsika]

There is some explanation and alternative coping strategies in this american document.

http://richardpclarke.tripod.com/earthquakedesign.pdf

The basic strategy is to find some dissipate the earthquake energy within the ductility of the members - obviously without causing structural collapse.
You are now getting into significant detail which can't be answered without knowing the detailing of the structure particularly the connections as I said before.

The devil is in the detail (pun intended)

Ok. Thanks.

Anyway. A separate question. I'm researching about hip roofs. It seems they rely on the perimeter walls. I'm studying about one storey building with hip roofs. Since hip roofs rely on perimeter walls. Does it mean all one storey building with hip roofs doesn't have beams that cross in the middle (I'm not referring to perimeter beams but beams in between walls)? I wonder if this is a rule or whether for larger one storey building, there must be beams in the middle to hold it.. or can one consider the hip roof enough to hold the 4 walls of the one storey building regardless of the sizes. What do you think?

 

[rodsika]

Rough guess

75 to 100mm concrete slab 240 kg per sq metre = 180*12*13 to 240*156 = 25tonne to 35 tonne

How did you get 180? 12? 13? and why 240 multiply by 156?

metal sheet roof 6 - 10 kg /sqm plus 4 trusses at 750kg each = 3 to 5 tonnes

are you saying that for every square meter of metal sheet roof, there are 4 trusses used.. so the 3 to 5 tonnes are for each square meter of metal roof?

also you said concrete slab is 240 kg per sq meter, so metal sheet weights more than concrete slabs because of the trusses requirement per square meter? remember concrete slab is 240 kg per square meter while that of metal sheet is over 3000kg (4 trusses at 750 kg each for square meter).

 

[Studiot]

Concrete weighs around 2400 kg per cubic metre.

for a slab that is 0.1 metres thick this is 0.1*2400 = 240 kg per square metre.

I estimated the roof slab as 0.075 to 0.1 so at 0.075 this figure reduces to 180 kg per sq metre.

The area of the roof is 12 * 13 sq metres. = 156

The sheet metal is around 6 to 10 kg per sq metre.

So weight of sheet metal is 10 * 13 * 12

You have to add to this the weight of the roof trusses I estimated 4 at 0.75 tonne each.

Strictly there will also be purlins to add; these are probably between 1 and 2 tonnes.

Does this help?

 

[rodsika]

Concrete weighs around 2400 kg per cubic metre.

for a slab that is 0.1 metres thick this is 0.1*2400 = 240 kg per square metre.

I estimated the roof slab as 0.075 to 0.1 so at 0.075 this figure reduces to 180 kg per sq metre.

The area of the roof is 12 * 13 sq metres. = 156

The sheet metal is around 6 to 10 kg per sq metre.

So weight of sheet metal is 10 * 13 * 12

You have to add to this the weight of the roof trusses I estimated 4 at 0.75 tonne each.

Strictly there will also be purlins to add; these are probably between 1 and 2 tonnes.

Does this help?

Thanks for this. But you forgot to add the weight of the beams across the walls in the case of the concrete slabs. Remember in metal sheet. They use trusses so no beams across walls are necessary. So instead of concrete slab being 25 to 35 tonnes compare to metal sheet 3 -5 tonnes. Concrete slabs with the required beams across may reach about 100 tonnes. So I guess for roof requirements. Metal sheets are sufficient. But then I wonder why many use concrete slabs as roofs which is about 30 times heavier. Eh?

 

[Studiot]

Why would an rc roofslab require additional beams?
It could easily be self supporting, I would certainly expect to design it that way.

After all it is not a floor slab that has to withstand the superimposed loading from the users.

 

[rodsika]

Why would an rc roofslab require additional beams?
It could easily be self supporting, I would certainly expect to design it that way.

After all it is not a floor slab that has to withstand the superimposed loading from the users.

are you saying there are roofslabs that are 12 meters in length that are simply put on top of the rectangular perimeter without any beam across? first time for me to hear of this. in my place, the concrete slabs are usually poured or made on site after rebars were put in the beams with wooden support for the new poured concrete. Unless you are referring to precast slabs with rebars inside that is simply put on top of the perimeter just like trusses??

 

[Studiot]

If I were the contractor I would be wanting to construct (cast) the floor slabs and cross beams all in one go at each level. That makes for economic construction.

Equally any concrete poured has to be lifted to roof level so you would want to minimise the quantity.

A concrete roof has only itself to support plus maintenance access. You cannot expect to load every square metre with filing cabinets or shop shelves as you could with a floor slab.
Don't forget that the physical form need not be a plain flat slab it can (would be?) be ribbed or otherwise strengthened by shape. And for that span it would certainly be reinforced.

But like I said we don't construct many concrete roofs in the UK.

 

[rodsika]

If I were the contractor I would be wanting to construct (cast) the floor slabs and cross beams all in one go at each level. That makes for economic construction.

Duh, if that is so, how do you connect these precast floor slabs with build in beams to the perimeter. This won't be strong. A slight jolt can disengage the connection and have the whole floors come crashing down.

Equally any concrete poured has to be lifted to roof level so you would want to minimise the quantity.

A concrete roof has only itself to support plus maintenance access. You cannot expect to load every square metre with filing cabinets or shop shelves as you could with a floor slab.
Don't forget that the physical form need not be a plain flat slab it can (would be?) be ribbed or otherwise strengthened by shape. And for that span it would certainly be reinforced.

But like I said we don't construct many concrete roofs in the UK.

Why, what is the difference between roof slab and floor slab. Arent' they the same. I assume that by using concrete slabs as the roof material, this means it is like another floors and if the owner wants to use it as floor. Then it's no longer concrete roof and ful fldege floors.

Unless there are really thin rc slabs that are only put on roofs and not to be walked on and never to function as normal floor.. maybe has got to do with the thickness?

How thick are normal floor slabs and how thick are roof slabs?

 

[Studiot]

What is your objective?

To understand the construction next door or to quibble?

What did I say about imposed loadings on a floor and a roof?

 

[rodsika]

What is your objective?

To understand the construction next door or to quibble?

What did I say about imposed loadings on a floor and a roof?

Well. I plan to move into the new building after it is finished so want to understand it.

Thanks for your distinctions. I was under the impression that all roof concrete are automatically roof deck. I even argued with the manager about this convincing him all roof slabs can be used as roof deck with future expansion to be used as floor, and I was convinced too so much that your previous statement escaped me. Now I understand.

Anyway, the original plan was to have 2-storey with roofdeck for provision up to third storey. But then only the ground floor will be important. So I suggest to him that he better just build 2-storey with metal sheet on his second floor roofing.

So you are saying a roof slab is about 35 tonne and a metal roof about 5 tonne. But then what I was asking you are not actually roof slab.. but additional floor slab or roof deck.. meaning the floor above it can have normal load for extension in the 3rd storey. This means the floor slab is not just 35 tonne but can be twice over.

Anyway. For completion of our discussions. If a roof slab is 35 tonne. How many times over are floor slab with loads of 60 lbs per square foot in the same 12 by 13 measurement you used?

 

[Studiot]

This is a two part post with information intended to help about construction practice, loadings and earthquake effects.

Firstly about the building.

UK practice we have

Ground floor
First Floor
Second floor
Roof

The floors are labelled differently in other countries, including USA.

Then we have what are known as Construction and Use regulations.

These specifiy loadings so the structure can actually be constructed as well as laodings for use.

So for an RC frame you have to provide support for the concrete before it sets and achieves its strength this support is needed for 1 to 4 weeks.

We call all the supports the falsework.

The falsework also has to support the construction plant and labour force duting construction and the 'formwork.'

The formwork is a temporary mould to hold the wet concete in shape until it sets.

Additionally the floor must be able to support the falswork and formwork for the floor aabove it until that floor has gained strength to support itself.
That is why precast units are preferred by contractors -they save the waiting time.

For loadings under the use heading we have

Floors are required to be able to support

Themselves
The weight of internal partitions etc.
The suspended ceilings etc
Possible the floor above during construction
A distributed load on every square metre of floor this allows for people, furniture shop counters etc.
A point load placed in the worst possible position on the floor.

Roofs are required to be able to support

Themselves
Snow on top of the roof
Uplift from wind forces
Maintenance staff walking on the roof, perhaps with crawler ladders.

The forces of an earthquake spread out from the source as two waves, one horizontal (P) wave and one vertical (S) wave.
These waves apply disturbing forces and movements to the building foundation points as the wave passes that particular point.
So if the building is a frame mounted on independent foundation pads or blocks the disturbance will occur at different times as the wave passes the individual pads.
this will introduce additional loading into the frame. You asked about torsion well torsion is bad because it can lead to very high stresses in directions the frame elements were not designed to be loaded and are thefore particularly weak.

Look at the sketch.

1) Is what happens when the wave arrives sqare-on ie parallel to the building grid.
As the wave passes AB at the front it lifts and drops the front pads and columns together so there is no side to dise twisting and no torsion.
So only additional bending forces aligned in the ordinary directions of the beams is introduced.

2) However there is only one direction paralle to the grid and many oblique to it so the wave is much more likely to arrive obliquely.
Immediately you can see that A is lifted before B (and CD_ so applies a twist to the frame as well as extra bending.
this is the dreaded torsion situation.

3) Shows the effect of the horizontal wave which moves the pads sideways. this applies a huge leverage on the joint at the top of the column.

It should be noted that the S and P waves travel at different speeds so arrive at different times unless the building is very close the the source.

 

[rodsika]

This is a two part post with information intended to help about construction practice, loadings and earthquake effects.

Firstly about the building.

UK practice we have

Ground floor
First Floor
Second floor
Roof

The floors are labelled differently in other countries, including USA.

Then we have what are known as Construction and Use regulations.

These specifiy loadings so the structure can actually be constructed as well as laodings for use.

So for an RC frame you have to provide support for the concrete before it sets and achieves its strength this support is needed for 1 to 4 weeks.

We call all the supports the falsework.

The falsework also has to support the construction plant and labour force duting construction and the 'formwork.'

The formwork is a temporary mould to hold the wet concete in shape until it sets.

Additionally the floor must be able to support the falswork and formwork for the floor aabove it until that floor has gained strength to support itself.
That is why precast units are preferred by contractors -they save the waiting time.

For loadings under the use heading we have

Floors are required to be able to support

Themselves
The weight of internal partitions etc.
The suspended ceilings etc
Possible the floor above during construction
A distributed load on every square metre of floor this allows for people, furniture shop counters etc.
A point load placed in the worst possible position on the floor.

Roofs are required to be able to support

Themselves
Snow on top of the roof
Uplift from wind forces
Maintenance staff walking on the roof, perhaps with crawler ladders.

The forces of an earthquake spread out from the source as two waves, one horizontal (P) wave and one vertical (S) wave.
These waves apply disturbing forces and movements to the building foundation points as the wave passes that particular point.
So if the building is a frame mounted on independent foundation pads or blocks the disturbance will occur at different times as the wave passes the individual pads.
this will introduce additional loading into the frame. You asked about torsion well torsion is bad because it can lead to very high stresses in directions the frame elements were not designed to be loaded and are thefore particularly weak.

Look at the sketch.

1) Is what happens when the wave arrives sqare-on ie parallel to the building grid.
As the wave passes AB at the front it lifts and drops the front pads and columns together so there is no side to dise twisting and no torsion.
So only additional bending forces aligned in the ordinary directions of the beams is introduced.

2) However there is only one direction paralle to the grid and many oblique to it so the wave is much more likely to arrive obliquely.
Immediately you can see that A is lifted before B (and CD_ so applies a twist to the frame as well as extra bending.
this is the dreaded torsion situation.

3) Shows the effect of the horizontal wave which moves the pads sideways. this applies a huge leverage on the joint at the top of the column.

It should be noted that the S and P waves travel at different speeds so arrive at different times unless the building is very close the the source.

After reading this. I googled for "does precast floor need beams" and after reading numerous pdfs. I'm still studying. Are you saying that it is possible to make 12 x 13 meter roof slab and put on top in the perimeter without beams across while for floor slabs.. maximum size seems to be 6 meters so there must be a 13 meter girder or beam at middle. Or are you saying it is also possible to make 12 x 13 meter floor slab and put on the perimeter without any beams in the middle? How. If not. So this means for 12 meter span... beams still need to be joined with the columns in poured Reinforced Concrete manner to suppor the 6 meter span precast floor.

In my place, precast buildings are so rarely used (maybe 1 in 10,000 buildings).

Btw.. as I originally mentioned (in the attachment earlier) the lot width is 12 meters and depth is 17.29 meters on the right and 14.57 on the left. But you mentioned "13" and we can use 13 for purpose of discussions.

I'll have to study the seismic angle later. Thanks.

 

[Studiot]

First thing you should know.

Nearly all building elements are 'controlled by' that is the 'determining factor is' deflection not strength. Most building elements are way too strong for the job.
This is quite important in building design.

I hope in reading the various posts you are gaining the impression the constructing a building is a blend of local availability vis a vis cost , labour, materials, lifting and other machinery, time etc. There is no one right answer.

You seem very worried about 6m limits on beam sizes. What is the maximum length of vehicles in your country.
Whatever size they are if they are made offsite they have to be delivered off loaded and subsequently handled on site.

On the other hand units made elsewhere can be erected immediately. There is no waiting time for concrete to gain strength.

Further under factory conditions you can manufacture prestressed units which have a considerably better strength/delection characteristics than those cast on site.

One thing you haven't said is how the foundations of your building were made. This also makes a difference to construction.
Does it have searate foundation pads at each column or one large foudation slab (we call it a raft foundation)?

 

[rodsika]

First thing you should know.

Nearly all building elements are 'controlled by' that is the 'determining factor is' deflection not strength. Most building elements are way too strong for the job.
This is quite important in building design.

I hope in reading the various posts you are gaining the impression the constructing a building is a blend of local availability vis a vis cost , labour, materials, lifting and other machinery, time etc. There is no one right answer.

You seem very worried about 6m limits on beam sizes. What is the maximum length of vehicles in your country.
Whatever size they are if they are made offsite they have to be delivered off loaded and subsequently handled on site.

On the other hand units made elsewhere can be erected immediately. There is no waiting time for concrete to gain strength.

Further under factory conditions you can manufacture prestressed units which have a considerably better strength/delection characteristics than those cast on site.

One thing you haven't said is how the foundations of your building were made. This also makes a difference to construction.
Does it have searate foundation pads at each column or one large foudation slab (we call it a raft foundation)?

In my place. We often used column footings below ground. In the building I'm leasing, they already have structural engineer to design the place and architect but I have the final word since I'd lease the place.

The design of the building is originally 2 storey with roofdeck (with provision up to 3 storey structural). After the architect and structure engineer already made the blue print for a month. I let the owner changed the design to 2 storey maximum with metal sheet as roofing in the second storey because I heard roofdeck always leaks and hard to maintain (I'd lease the ground floor and since there are limited parking spaces, the third storey is not important so I practially control the whole building). But then researching yesterday about metal sheet roof like garble and hip roof.. and gutters all around it. I realized metal sheet and gutter seems to need as much maintenance as roofdeck. What's your experience in this? The architect and structural engineer already charged us double for changing the plan.. do you have any reasons why roofdeck or concrete slab roof would be better in maintenance than metal sheet? But then for seismic consideration... metal sheet seems to be more advantageous, so for now metal sheet is still my choice.

 

[Studiot]

All foundations are below ground.

How about answering my question?

One thing you haven't said is how the foundations of your building were made. This also makes a difference to construction.
Does it have searate foundation pads at each column or one large foudation slab (we call it a raft foundation)?

 

[rodsika]

All foundations are below ground.

How about answering my question?

file attached is picture of the footer.

It is what we commonly use.

Why.. would precast slabs need different footings? But precast slabs are expensive so I don't consider this.

 

[Studiot]

Nice picture
So these are pad foundations and you also have the excavations for what we call a ground ring beam.

One purpose of the ring beam is to stiffen the edges of the ground slab (to prevent edge curling) and also to hold the whole thing together if there is local loss of support under the ground slab as the ground settles over the years.

I also do not see any significant seismic strengthening of the connection between the pads and the columns in your picture. In an earthquake (or even for building wind loading or thermal movement) the columns will pivot about their joints with the pads so these points should be regarded as pinned or ball joints.

One point I made earlier about the leverage of earthquake disturbance affects your roof.

I hope you understand leverage?

If you have a heavy lump waggling about on the end of a long lever you have a large effect.

The longest lever is up to the roof so a heavy (concrete) roof is most vulnerable to seismic activity. Further if it crashes down on the floor below it it will a great deal more damage than a light metal roof which might come off but could well save the structure and be easily replaced.

Please also remember this all is friendly advice, I am not acting as an unpaid structural consultant, nor would PF want me to.

 

[rodsika]

Nice picture
So these are pad foundations and you also have the excavations for what we call a ground ring beam.

One puropse of the ring beam is to stiffen the edges of the ground slab (to prevent edge curling) and also to hold the whole thing together if there is local loss of support under the ground slab as the ground settles over the years.

I also do not see any significant seismic strengthening of the connection between the pads and the columns in your picture. In an earthquake (or even for building wind loading or thermal movement) the columns will pivot about their joints with the pads so these points should be regarded as pinned or ball joints.

One point I made earlier about the leverage of earthquake disturbance affects your roof.

I hope you understand leverage?

If you have a heavy lump waggling about on the end of a long lever you have a large effect.

The longest lever is up to the roof so a heavy (concrete) roof is most vulnerable to seismic activity. Further if it crashes down on the floor below it it will a great deal more damage than a light metal roof which might come off but could well save the structure and be easily replaced.

Please also remember this all is friendly advice, I am not acting as an unpaid structural consultant, nor would PF want me to.

First. The picture is not mine. I just saw it at the net (and shared it to illustrate the idea of column footer). Second, the construction of the building I'd lease hasn't started. We haven't even started excavations yet because the final design of the building are still being done due to my changing the 3 storey to 2 storey. Anyway. Thanks for the many tips. I'll think and remember about them one by one. I'd also recommend the architect and structural engineer to use ground ring beam... as the original design is only one column footer at every corner. The picture I shared with you is from a person building a fence. I just shared it to illustrate a column footer. Thanks again for the tips. This will do for now.

 

[rodsika]

The forces of an earthquake spread out from the source as two waves, one horizontal (P) wave and one vertical (S) wave.
These waves apply disturbing forces and movements to the building foundation points as the wave passes that particular point.
So if the building is a frame mounted on independent foundation pads or blocks the disturbance will occur at different times as the wave passes the individual pads.
this will introduce additional loading into the frame. You asked about torsion well torsion is bad because it can lead to very high stresses in directions the frame elements were not designed to be loaded and are thefore particularly weak.

Look at the sketch.

1) Is what happens when the wave arrives sqare-on ie parallel to the building grid.
As the wave passes AB at the front it lifts and drops the front pads and columns together so there is no side to dise twisting and no torsion.
So only additional bending forces aligned in the ordinary directions of the beams is introduced.

2) However there is only one direction paralle to the grid and many oblique to it so the wave is much more likely to arrive obliquely.
Immediately you can see that A is lifted before B (and CD_ so applies a twist to the frame as well as extra bending.
this is the dreaded torsion situation.

3) Shows the effect of the horizontal wave which moves the pads sideways. this applies a huge leverage on the joint at the top of the column.

It should be noted that the S and P waves travel at different speeds so arrive at different times unless the building is very close the the source.

Hi Studiot,


I've been thinking something for half a day and needs your comment.

Please see the file attached picture. I'm deciding whether to add a column in the middle (marked in red). The architect wanted me to decide.. and I didn't have access to his structural engineer and he can't discuss with me anything structural because he said his specialty was architecture and not familiar with seismic loading or torsions.

Last I talked to him early today I wanted him to add the column at middle. He said it is ok. But then remember in the torsion article any unevenness in structural can introduce twist. Imagine "B" (in the picture) has column at middle while C doesn't. Imagine a seismic wave passing from left to right. This would make the C move differently from B and introduce torsions?

Also something perplexes me. The beam from top to bottom middle are not girder because they are same level as the horizonal beams.. meaning imbedded into each other. Supposed there was no column in the middle of "B". Would the load be distributed to the horizontal and vertical beams (in the picture)? Or would the horizonal beams be the main support and the vertical beam just to holds up the floor slabs? What is usually the case?

Note in the construction project. No precast slabs would be used. They would all be poured concrete and rebars.

So should I add the column in the middle of "B" or not? On one hand, I'm on tight budget and the architect said the "B" side beams would be smaller since they would be 6 metesr apart... and it would also support the 12 meter RC beams at "C" making cross beams. Would this make the loading at "C" distributed to the vertical and horizonal beams?

Actually one month ago. He told me it's nicer to have no columns in the middle. But now said it is ok because I'm on tight budget.

Now I'm quite undecided. I guess seismic considerations would be my primary concern in whether to put the middle column or not.. but then.. if "B" and "C" has no columns at middle... and "A" and "D" have them.. it would still introduce unbalance seismic loading for waves that come from left to right?

Lastly. The connections would be standard connections, the loading is simply one floor above or simply 2-storey with metal sheet in the 2nd floor roof.

What do you think? Column or no column at the middle of "B", that is the question?

Thanks.. this inquiry is my last question in this thread, don't worry :)

 

[Studiot]

I am not of the opinion that an extra column will do much for earthquake performance, in fact it may reduce it.

In post#21 I drew attention to the important key statement in your code.

The way to enhance earthquake performance is to make the horizontal elements more flexible. This way they can dissipate the quake energy input to the structure.

The whole matter depends upon the design of these things (again including the connections).
You should ask your structural engineer to explain.

 

[rodsika]

I am not of the opinion that an extra column will do much for earthquake performance, in fact it may reduce it.

In post#21 I drew attention to the important key statement in your code.

The way to enhance earthquake performance is to make the horizontal elements more flexible. This way they can dissipate the quake energy input to the structure.

The whole matter depends upon the design of these things (again including the connections).
You should ask your structural engineer to explain.

Problem with long span is the connections must be made perfect. Also in most homes, the distance between columns are 6 or less. It is unusual to find homes with distance between columns of 12 meters or more. Have you thought why.

Anyway. I let them added a column in middle yesterday making the span of one portion 6 meters instead of 12. This would save much steel and make the constructions cheaper.

Also for the 2-storey with metal roof. I let them design it for 3-storey foundationa and structural. I assume using 3-storey structural in 2-storey building would make the columns stronger and more earthquake proof. Any wrong with this idea? I know you would say to make it more ductile. But using 3-storey foundations and columns would add more steel making it stronger. What's the problem with this?