Sistering Joists

[Pappa]

Hi Guys

My Structural Engineer has designed a new attic floor for me. The diagram shows only 1 joist for brevity (structural walls in red):

 

 

 

 

He has chosen 4m, 4.8m 3m lengths as they can be picked up off the shelf cheaply.

He said 6m is also available at timber merchants but substantially more expensive.

 

This is a side view as to how the joists are to be bolted together:

 

I'm not disputing the structural side of it. But not sure about the practicality of fastening two lengths of timber together while floating in mid air.

Do you have any tips or tricks that I can employ to do this?

 

If no, I might go back to SE and ask for a 6m length timber as the main beam so atleast thats supported by the the masonry wall on either side of the long span, the sister joists will then be easier to screw together.

 

Thanks in advance

 

P

Edited May 5 by Pappa

[Mr Punter]

Lay them down on some trestles to do the drilling and bolting.

[Pappa]

52 minutes ago, Mr Punter said:

Lay them down on some trestles to do the drilling and bolting.

Thanks for the reply.

How do I ensure I'm bolting them together straight?

Its hard to find straight timber, by the time I've bolted the 4th timber on, I could be way off. 

[BotusBuild]

I feel your pain

2 hours ago, Pappa said:

hard to find straight timber

 

My local timber merchant allowed me to sort through the pile to select - in fact he helped and rejected pieces I was OK with 😁

 

Bolt them together as soon as you get them onsite, as they warp as they dry out. I would start with the two larger overlaps first and then do the middle overlap last. With the larger overlaps do the two pairs of bolts at each end of the overlap and work inwards (I presume you are putting in more than just two pairs of bolts on each overlap).

[Roundtuit]

I'm guessing the sistering is to gain a bit of head room rather than use deeper joists?  What size joists are we talking? I have no expertise, but it doesn't look like a very elegant solution; any chance of getting a steel joist in there to split the span?

[Pappa]

1 hour ago, Roundtuit said:

I'm guessing the sistering is to gain a bit of head room rather than use deeper joists?  What size joists are we talking? I have no expertise, but it doesn't look like a very elegant solution; any chance of getting a steel joist in there to split the span?

 

Reasons for sistering as per SE:

1) Steel beams just aren't possible.

2) 6m lengths are premium priced, 4.8m etc are more reasonably priced and easier to handle / not as heavy.

3) Sistering joists (2x8inch) like this and bolting them together with steel screws gives more strength than what a 4x8 inch timber beam would have provided. This is important as it is a long (6m) span.

4) Deeper joists aren't an option because it messes with the staircase/attic headheight + cost of deeper joists increases exponentially.

 

 

Let me know if you've got any more advice about how to easily fasten these together.

I can do it on a trestle table and then lift the beam up, however, by the time they're all fastened together, the weight will be significantly greater and difficult to maneuvre, it will be easier if fastened in situ.

 

Cheers

Edited May 6 by Pappa

[Redbeard]

Take careful measurements, take them outside and cut to size. Clamp together, with a straight-edge to supplement the evidence of your eyes. Mark sections, drill, bolt up. Check again. Unbolt, take up to the roof, lay sections out and bolt up in situ. At this stage I'd use spiked washers.

[Roundtuit]

Maybe invest in one or two 6m lengths and install them mid span at the same height as the structural walls, just as a temporary measure to sit the joists on whilst you bolt them? 

[Spinny]

When you have fitted the first joist in situ and confirmed that it is the correct length and fits well, one possibility might be:

- move that joist across the area as a tester to help establish whether the walls are square i.e. all joists can be made to the same length

- take it apart and remove it. Then take it outside and use it as a template to cut and drill the remaining lengths of wood to make up the other joists to the same length (if it is square) 

 

(However I am not sure whether it is best to have all joists with identical overlaps and identical bolt positions. It might be arguable that different length/position of overlaps and bolts on different joists would give a more robust structure less likely to fail in the same place across all joists. But I guess your SE would say it is designed so that any one joist will never fail anyway.)

 

PS I am not a carpenter and probably don't know what I am talking about.

[Spinny]

PS Are the joists 'at the top' on your sketch done the same way with a span of 9.6m ? I guess the 2.9 wide section top-right could perhaps be joisted at right angles, instead of all joists being 9.6m long.

 

Also maybe consider what apertures/fixings will be going into the ceiling and consider whether the joist positions will be compatible with these e.g. loft hatch/ladder, downlights etc.

(My carpenter was good enough to adjust a couple of joists above a kitchen to make room for specific feature downlights). It is only something you learn too late - that the structural drawing should be overlayed with the lighting plan and the room layout to resolve any clashes.

[SteamyTea]

Has the SE specified adhesive as well as bolting?

 

As for getting them straight and level, much will depend with what you start with.

Even a relatively short but of timber will be unlikely to be 'flat' in both planes. Once clamped together (bolts clamp, screws don't), they are highly likely to twist.

 

Are Posi-Joist type joist an option, they are usually made in a jig and the metal webs are more stable.

[Nickfromwales]

Timbers like these always have a ‘cup’ to them and you need to have these noted. 
 

Lay the wood down, lift up one end, and look down the length. You’ll note a depression and these need to be marked so all the depressions are ‘down’. 
 

Once installed these then need to be turned to be upside down so they deflect upwards, as these will then settle when weight is applied; weight from intersecting joists and flooring, plus stud walls etc.

 

Use crown washers, upon final assembly only, and it’ll be fine. Remember to only tighten these bolts initially just enough to get a firm purchase, then jack and level with vertical batters off the floor at the mid-point of the span (one on each side so it doesn’t cause the timber to rock) to get it all ship-shape. Once happy with the level, tighten these all up snugly. I then go down the timbers each side with 90mm ring shank nails (NOT bright nails) and nail the shit out of them. 3 nails per row and repeated every 450mm from both sides, all down the entire run. Be sure of things being in place, because once you nail this, any thought of further tweaking goes out the window. 

 

Getting these in is probably a well educated 2-man job or a 3-man job, if you wish to retain any sanity and avoid getting smashed in the head by a timber.  Usually 2 men on muscle, one running for fixings and stuff you drop when you least need it to (tools / nuts / bolts etc) so don’t be brave, be clever, get as much help as you can.

 

My 2 cents is stop being tight and buy the 6m timbers obvs, as they can’t be THAT expensive?? Compared to the grief vs results, plus the cost of fixings and logistics / manpower, that has to be the sensible option surely?

[saveasteading]

6m is more likely to be stocked at a specialist timber merchant rather than a general BM.

Phone around.

8 x 2 are not at every BM anyway.

 

But I think you are over-thinking.

 

This is heavy stuff.

 

I would splice the lhs 4.8 and 3.0 and lift to place. Then the rhs 4.8. Then add the 3.6. 

Tweak for the best fit you can manage. Structurally any twists won't matter.

in fact 2 timberz bolted could be stronger than one very wide timber. 

 

Levels for floor and ceiling could be a bit wonky but would with any solution.

You will need some clamps and perhaps props.

 

If you want to whack in some screws for control then just do it. That might make the drilling easier.

 

[Nickfromwales]

I'd also tie the left end into the wall with both timbers, not as shown with one stopping short?! That will prevent the timber from naturally wanting to roll, and this arrangement will need less 'help' from noggins then. Very odd to see that detail tbh.

[saveasteading]

2 hours ago, Nickfromwales said:

not as shown with one stopping short?! 

I don't think it reaches. And the other end isn't detailed into the wall. So is is on a hanger?

 

But yes, add another bit of wood at the left so it is the same section all the way. Not needed but tidy.

 

 

[Temp]

On 06/05/2025 at 17:09, Pappa said:

Sistering joists (2x8inch) like this and bolting them together with steel screws gives more strength than what a 4x8 inch timber beam would have provided. This is important as it is a long (6m) span.

 

Make sure you understand this. He's saying 4*8 made from 2qty of 2*8 is stronger than just one large 4*8.

 

You might need something bigger than 8" deep? I think it's to do with knots?

 

Id go with his suggestion and follow the advice above about taking it apart to get into place. If working short-handed Id look at making some "clothes props" wuth a tripod base to support one end. While you bolt the other.

[Nickfromwales]

59 minutes ago, saveasteading said:

I don't think it reaches. And the other end isn't detailed into the wall. So is is on a hanger?

 

But yes, add another bit of wood at the left so it is the same section all the way. Not needed but tidy.

Yes, add a piece, but would defo be a simpler way to finish the intersecting timbers to a non-staggered surface and it would prevent any rotational movement as that’s creating a pivot. 

 

Also odd that the 4m and the 3m weren’t swapped, to further stagger the joints. 

[Pappa]

Thanks guys, really good comments here.

 

On 07/05/2025 at 17:40, SteamyTea said:

Has the SE specified adhesive as well as bolting?

Waterproof PVA Glue

 

On 08/05/2025 at 12:32, Nickfromwales said:

I'd also tie the left end into the wall with both timbers, not as shown with one stopping short?! That will prevent the timber from naturally wanting to roll, and this arrangement will need less 'help' from noggins then.

I think he only wanted to one timber to be go in to the wall so as to disturb the masonry as little as possible (the masonry is to be notched 60mm deep). He feels there was going to be a lot of weight on the left hand side wall as it was supporting a long span, hence he was being more cautious on that side.

 

On 08/05/2025 at 15:14, saveasteading said:

The other end isn't detailed into the wall. So is is on a hanger?

SE said the other end can be fastened on to a timber ledger beam using joist hangers as that side is supporting a shorter span and hence less weight on the masonry.

 

 

I've got a question about how to notch 60mm in to the brick... 

Any tips as to how to achieve this? I was thinking angle grinder, multi-tool and masonry chisel?

How likely is it that I'll end up breaking the brick?

 

Thanks again

 

P

[Mr Punter]

Why not fix a ledger board to both walls?  Do it with resin fixings.

[Pappa]

4 minutes ago, Mr Punter said:

Why not fix a ledger board to both walls?  Do it with resin fixings.

SE specifically said no to ledger board on the long side. The resin anchor bolts won't be strong enough to support the weight of the floor as they needto support 4.58kN/m, that would mean having so many bolts it would be impractical/uneconomical.

The beams are to go directly in to the masonry wall.

We can use a ledger beam on the short side.

Edited Wednesday at 18:53 by Pappa

[Nickfromwales]

2 hours ago, Pappa said:

I think he only wanted to one timber to be go in to the wall so as to disturb the masonry as little as possible (the masonry is to be notched 60mm deep). He feels there was going to be a lot of weight on the left hand side wall as it was supporting a long span, hence he was being more cautious on that side.

Codswallop.....  It weighs the same, either chuffing way  !! The better job would be to increase the footprint of the timber at the point where the load is transferred to the masonry. Complete nonsense afaic, sorry.

 

2 hours ago, Pappa said:

SE said the other end can be fastened on to a timber ledger beam using joist hangers as that side is supporting a shorter span and hence less weight on the masonry.

   so different fixings change the weight of the load do they? Not what I was taught, in the few physics lessons I attended.

 

2 hours ago, Pappa said:

I've got a question about how to notch 60mm in to the brick... 

Any tips as to how to achieve this? I was thinking angle grinder, multi-tool and masonry chisel?

How likely is it that I'll end up breaking the brick?

This suggested methodology, of cutting a section of a brick out, shows that this person has zero practical skills or has ever tried this in person, as it simply cannot be done. You take out an excess amount of masonry, clean the hole and set a mix of compo in, then use an offcut of timber to create the pad out of compo (sand and cement) and allow to dry. I use a piece of synthetic slate on top of the compo to create a nice clean shelf for the timber to sit on. Once the timber is in place you fill in around the timber with more strong (3:1) compo and job done.

 

Trying to get the timbers into this wonderfully neat little pocket you are going to attempt to make would be wholly impractical and cause you no end of logistical aggravation during the install, and the suggestion of you disturbing say 100mm of masonry would compromise the strength of the wall is just make believe, sorry again.

 

Is this guy a 1st year apprentice or something?

[Gus Potter]

On 05/05/2025 at 11:40, Pappa said:

My Structural Engineer has designed a new attic floor for me. The diagram shows only 1 joist for brevity

@Pappa and all.. hope this give you possible further insight into what your SE is up to. Attic conversion are an art with often some crafty engineeering. The following I hope gives a bit of an insight.

 

The SE will have considered at least the following:

 

1/ It needs to be strong enough, the joists and connection not to break. Call this strength design.

2/ The deflections (the amount the floor bends by) need to be controlled.

3/ Head room.. increasing the thickness of the floor is often not possible.

4/ Disrupting the walls where the existing joist ends rest, you often want to avoid that as they tie the walls together.

5/ If I'm going to sister up joists in this way.. where is the best place to put the joints? usually where the bending force in the timber is least onerous or at least we can achieve the best balance between the bending and shearing forces in the timbers.

6/ How to I connect the joists to each other so the forces get transferred around the lap joints.

 

@Pappa To start with lets assume the floor is uniformly loaded, no point loads from wall say. How much load might we be talking about? I'm going to simplify the factors of safety (ignore them and assume your SE loads are working loads) as they vary between the design codes. This is just a rough qualitative look at things and lets carry out a common sense check based on the info you have provide.. don't lay into your SE  if you think something is wrong ask them to explain a bit.. which they will be able to do. But for the curious let's see if @Pappa numbers roughly look credible.

 

Say the self weight (dead / permanant weight of the floor is 0.75 kN/m^2 ( about ~75kg) and we have what is called the live (imposed) load which is people, funiture etc. The design codes require us to use a value of 1.5 kN/m^2 (~150 kg/m^2) for a dwelling. Add the two values together gives us 0.75 + 1.5 = 2.25 kN/m^2. I've picked a ball park figure for the dead weight of you floor, could be less or more.

 

Take the longest left span and pretend that here is just one beam spanning between the left wall and onto the middle wall in isolation.. this is what we call a simply supported beam. Taking a 1.0m wide strip of floor with a beam span of say 6.0m gives us a total load of 2.25 * 6.0 = 13.5 kN per metre width of floor and for a simply supported beam the load per metre run of ledger would be 13.5 / 2 = 6.75 kN/m.. but see below what @Pappa's SE says.

 

On 14/05/2025 at 19:53, Pappa said:

need to support 4.58kN/m

 

The load I've just calculated is some 45% more (call that an over stress) than the SE's load, even if you have a lighter floor then on first glance there looks to be something wrong. The first obvious thing is.. can the floor be loaded with people over it's full span, is there a coomb? If so then the ledger load is going to reduce at the left wall end. Say the coomb extends in 2.0m from the left wall. You need to maybe to allow say 0.25 kN/m^2 for using that space as storage? I'll ignore that for now to keep the maths short. That will maybe knock off about 1.5 kN/m run ledger load. 6.75 - 1.5 = 5.25 kN/m run of ledger. still more than your SE's 4.58 kN/m

 

For all I worked out this qualitative reduction of 1.5 kN/ metre run of ledger by saying we have 4.0m of floor that could be loaded by people towards the middle wall. Thus 4.0m x 1.5 kN/m^2 = 6.0 kN and that 6.0 kN acts as a point load 2.0 m in from the mid wall. Thus for a simply supported beam 2/3 of the load will be carried by the mid wall and 1/3 by the left wall. So  6.0*1 / 3 = 2.0 kN/m reduction and I've added a bit back in of 0.5 kN/m to account for you storing stuff in the coomb and that is how I arrived at the 1.5 kN/m reduction I used above. But the 5.25 kN/m per run of ledger is still an over stress!

 

Here I think your SE is doing what you are paying for.. this is "crafty bit" and demonstrates engineering skill. To explain we need to look at some basics first. When designing beams etc we are primarliy interested in the bending force in the beam and the shearing forces. There is a lot of other stuff going on but let's just stick to these for now. Engineers use diagrams to repesent these forces.  

 

Below is what we call the bending moment diagram for a simply supported beam holding up a uniform load.

 

And diagram below represents the shearing forces in the beam. @Pappa we want to know a bit about this, where your SE is showing the sistered joists stopping short of the left wall.

 

 

 

But in this case the SE is lapping and joining the joists together so it's intuitive that we must be getting some transfer of forces between the left side and right side and here we can take advantage of this and while Nick below is correct about the weight not changing the way the weight influences the joist behavoir can change a lot, or enough to make things work.

On 14/05/2025 at 22:35, Nickfromwales said:

It weighs the same, either chuffing way

 

Let's now look at the doubled joists and how the bending and shear forces change even though the weight is not changing. You may have heard SE's talk about stiffness and how stiff things attract load.

 

Here is an analogy, let your mind wander and wonder!  Say we go to the gym (cough at my end) hold out a kettle ball at arms length. The ball causes a bending force at the shoulder and a downwards shearing force. Now you have a pal, also with a perfect body, that comes up behind you with super glue on their fore arm and sticks their arm to your humerus region thus some of the load will get transferred to their arm. This will stiffen you arm and some of the  load will get transferred to your pal. We can take this analogy and put some diagrams to that as below as this is what looks to be happening in @Pappa's case.

 

Now when you get a helping hand from the right wall side timbers you can see that on the main 6.0 m span the bending force is reduced. But also the shearing force at the left wall reduces. I've just sketched the diagrams so not perfect!. I've just done this roughly and for ease and have shown the floor fully loaded. But SE's will consider the left side full loaded with minimum load on the 4.0m span as this will give the maximum shear load on the left wall where the new sistered joists stop short and we need to check the remaining short part of the original joist.

 

Now lets look at how much the floor will deflect. @Pappa's SE has sistered the joists over the longest span and will have calulated the deflection and checked it is ok. Where the joists stop short of the left wall will have little influence.

 

In this post I've not looked at how you design the bolts and so on.. it is complicated!

 

@Pappa. Hope this helps. My recommmendation is that you stick to your SE drawings and connection recommendation, don't go swapping things out without checking with your SE first.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[Gus Potter]

On 08/05/2025 at 10:07, Nickfromwales said:

Timbers like these always have a ‘cup’ to them and you need to have these noted. 

This is a fab observation... and demonstrates to me a knowledge of timber.

 

To explain SE wise. You'll often see joists or a flitch beam (commonly two bits of timber with a steel plate between) fixed together with the nails / bolts towards the outside edges. If your new timber starts to cup it pulls the nails / bolts, sets up secondary stress and causes havoc as often one of the timbers will split along it's length rendering your hard work useless. 

 

When sistering joists we need to look at the end grain. You want to put the younger wood next to the old joist so when it cups the gap is in the centre and the top and bottom of the timbers remain in contact.

 

@Nickfromwales is demonstrating good old school knowledge. Years ago a joinery apprenticeship used to be a lot longer and the kids got taught this stuff.

[Pappa]

Thanks guys

I've spoken to SE and it turns out I was mistaken.

• He wants me to notch/remove brick on both sids of the span to support the timber.
• Ledger beam is less possible because there are only a few courses of brick above the joists.
• To make the ledger beam viable we would need too many resin anchor bolts which would make it impractical.
• It is difficult to specify resin anchor bolts when the loads are this high because their rating is based on tests under laboratory conditions. The margins of error are too high due to installation disparity (dust left in the hole). The only way to be sure would be to stress test each fixing afterwards which is again not practical.
• The easiest way will be to insert the timber in to the brick wall. He said its up to me how I want to do this, I can remove a couple of bricks and place the timber beam, or I can notch a pocket if I've got the approproate tools/expertise.
• The load on the left hand side is indeed greater due to longer span; greater floor area; partition walls; heavier bathroom appliances.

 

I'll try my best to cut a couple of bricks using a combination of stitch drilling and diamond/masonry blade.

If it is too much effort, I'll just remove some bricks and fill around the timber with mortar.

 

Thanks again

[Nickfromwales]

10 hours ago, Pappa said:

I'll try my best to cut a couple of bricks using a combination of stitch drilling and diamond/masonry blade.

The secret is to NOT drill in at 90^o with the drill bit as you'll have snots of brick which are a PITA to get rid of if you try to do perfectly parallel drilling. So angle the drill outwards (sideways), 5-10^o should suffice, so the pocket is slightly wedge shaped.

 

You can try this vs just taking out the nearest whole bricks and then reinstating perfectly cut new bits of brick back in, of course, so let's see how many you do before saying "(expletive deleted) this for a game of rounders" and storming off to the pub. My bet is 2 max, but I'll have a small side bet on the first one seeing you into a straight jacket. 

 

Life is short, pick your battles well young Jedi...... and the kicker is, taking whole bricks out CAREFULLY will be a perfectly fine job; you won't find a builder milling out pockets with multi-tools etc, they'd look at you as if you're from another planet 👾 and also if you take whole bricks out you won't have a load of shite fall down into the cavities.

 

Use the force.... ( 👈 LINK)