Floor bounce - PosiJoist Floor Vibration checks

[readiescards]

So I'm still trying to get to the bottom of why we have to tip-toe across our master bedroom otherwise the floor flexes causing all manner of things to move.

 

Anyone able to tell me what the MiTek software used in designing PosiJoists 'Vibration Checks' section is trying to explain please?

[joe90]

I have no idea, but all I know is my builder spaces his engineered joists at 400mm not 600 because it’s firmer and I agree ( and glad he did) ?

[Nickfromwales]

600 and 400 are chalk and cheese. I did a job for a customer where the TF company specified 600mm centres. 1 runs of strong-back, 22mm p5 deck, glued and gas nailed, and still the dynamic deflection was horrendous. Fitted wardrobe doors seen moving etc. 

The deflection for static loading and dynamic loading are what you need to translate, as obviously it's the walking around / other activity aka dynamic that's causing issue. Can the company that specified them answer your questions ? 

 

[Jeremy Harris]

We have 253mm (PS10) Posijoists at 400mm centres over a wider span than this and the floor is rock-solid. 

 

I suspect that you have mid-span deflections that are greater than predicted for some reason, possibly the way the end-restraints are arranged on the joists, possibly because the static load is higher than anticipated, perhaps the joists were not pre-cambered during manufacture to the dead load deflection (a common failing), or maybe the floor boards were not secured as well as they might be. 

 

I doubt it's the latter and strongly suspect that it's either the lack of dead load pre-camber or greater freedom at the end attachments (like using joist hangers rather than fixing both members of the Posijoist securely to the structure).  Having them at 600mm centres will have exacerbated the problem.

 

[Nickfromwales]

42 minutes ago, JSHarris said:

We have 253mm (PS10) Posijoists at 400mm centres over a wider span than this and the floor is rock-solid. 

 

I suspect that you have mid-span deflections that are greater than predicted for some reason, possibly the way the end-restraints are arranged on the joists, possibly because the static load is higher than anticipated, perhaps the joists were not pre-cambered during manufacture to the dead load deflection (a common failing), or maybe the floor boards were not secured as well as they might be. 

 

I doubt it's the latter and strongly suspect that it's either the lack of dead load pre-camber or greater freedom at the end attachments (like using joist hangers rather than fixing both members of the Posijoist securely to the structure).  Having them at 600mm centres will have exacerbated the problem.

 

Checking to see if there is already a static, negative deflection may be somewhere to start  ?

[Jeremy Harris]

7 minutes ago, Nickfromwales said:

Checking to see if there is already a static, negative deflection may be somewhere to start  ?

 

 

Yes, it would.  If the joists have been pre-cambered to the anticipated dead load, as per the spec (and I doubt they were), then measuring the static deflection in the centre of the room, i.e. the distance the centre is lower than the edges, should give an indicator.  If the joists were pre-cambered then there should be near-zero deflection, if they weren't then I'd expect to see a small deflection, maybe a mm or two in the centre.

 

The other point is the way the Posijoists are restrained at the ends.  If they are fixed such that the top and bottom members are firmly secured then the floor will have a lot less deflection than if  they are simple bearing on something like a joist hangar.  The latter will allow the joist to pivot slightly under load and exacerbate the centre deflection under load.

Edited January 7, 2018 by JSHarris

[Nickfromwales]

Agreed. 

@readiescards, do you have pics of the floor you can upload ? 

One option here, as a retrofit fix, could be dropping the ceiling underneath and inserting some engineered, full depth noggins at say the 1/3, 1/2, and 2/3rds points along the length of the span. These can be made like a horse shoe to allow existing services to remain in place, so basically a pair of U-shape 25mm plywood baffles, one inserted with the mouth down, and the other overlaid with the mouth up. 

These then get glued and screwed together, then a pair of 50x75mm noggins get affixed to those and then to the top and bottom chords, again glued and screwed as well as practicable. Sash-clamping these so the glue is able to form a void free bond will be key to success, but a foaming PU GLUE will help a lot. 

Another belt and braces ( again retro solution ) measure will be to affix 18mm plywood to the underside of the joists before fitting the new plasterboard. That'll stop any remaining bottom chord movement, and hopefully that would stop the issues your having in the room above.  

[RandAbuild]

Ours are 5.4m long 254x122mm (TR26) Posijoists from Wolf Systems at 480mm centres, with glued 22mm thick Eggar Protect board. Like Jeremy it's rock-solid

[Jeremy Harris]

You can also pre-camber Posijoists to a small degree after fitting, by packing the ends of the lower members out by a tiny amount.  A lot depends on the end fixings, but provided both the end fixings and wall construction are up to it then packing out the ends of the lower members, perhaps with an Acro in the centre to induce the pre-camber, should work.

 

What this does is increase the compressive load in the lower Posijoist member, and so stiffen up the whole joist against bending loads.  One key thing with Posijoists is that if you are going to use them near their design limits then the end constraints become more critical.  The upper member is always in tension, so has to be secured very tightly to the wall so that is cannot pull away inwards under load at all, and the lower member is always in compression, so must be held tight to the wall or fixing so that it has no room to move outwards.  We're talking very small movements here - a mm of movement at one of the member ends will be several mm of deflection in the centre under load.

Edited January 7, 2018 by JSHarris

[joe90]

Reading the spec on the first post you have the actual dead load deflection and pre camber figures, can you not measure this and if it’s wrong you have evidence. I agree with Nick, ply noggins would help a lot ( rather than rip it out ?).

[Jeremy Harris]

In terms of the physics of stiffening a beam, ideally what you want to do is:

 

1.  Increase the area moment of inertia (not easy, as with Posijoists 99% of this is determined by the section and spacing of the top and bottom members - the metal webs or any noggins make next to sod all difference to it).

 

2.  Alter the pre-camber, to balance the compression and tension loads in the top and bottom members under the dead load.

 

3.  Change the end restraints to reduce the freedom of movement they have.

 

As a worked example, take a 4.5m long Posijoist of the same dimensions as the ones in question, and, using Beamcalc, look at the deflection under a fairly standard distributed load of 1000N/m for two different end restraint types:

 

Simply restrained ends:    centre deflection = 4.7mm, peak bending stress in top and bottom chords = 50%, average shear stress = 3%

 

Fixed ends, both top and bottom chords rigidly supported:    centre deflection =  2.8mm, peak bending stress in top and bottom chords = 50%, average shear stress = 3%

 

So, just by making the ends of the Posijoists more rigidly fixed, we've roughly halved the centre deflection for the same load condition.  This illustrates pretty well why the end restraint can make a substantial difference to the overall centre deflection and why it is is critical that this type of composite joist be properly fixed to the walls.

 

 

Edited January 7, 2018 by JSHarris

[Jml]

Is there an easy sytem to fix the ends?  Joist hangers, even I understand how to use, but am now concerned they may not be the best fixings, as we will be using posijoists I would want to avoid this potential problem.

Edited January 7, 2018 by Jml

[Jeremy Harris]

If you follow the standard details given by the joist manufacturers, like these, for example:  04 Standard Details.pdf , then you can't go far wrong.  Problems arise when composite joists like this are treated in the same way as solid timber joists, without regard for the differences in the way they need to be supported, both at the ends and when fitted over the top of internal support beams or walls.

Edited January 7, 2018 by JSHarris

[ProDave]

40 minutes ago, Jml said:

Is there an easy sytem to fix the ends?  Joist hangers, even I understand how to use, but am now concerned they may not be the best fixings, as we will be using posijoists I would want to avoid this potential problem.

With a timber frame, like mine, the joists sit on top of the downstairs walls. The upstairs walls then sit on the joists (with a laminated ring beam all around the outside). So the ends are well and truly fixed and are not going to flex as they might in joist hangers.

 

At the 2 intermediate (supporting) walls the joists sit on those walls. One continuous joist nearly 12M long right from one end of the house to the other. The upstairs walls again sit on the joists directly above the downstairs walls.

 

It was designed this way, and at the ends, and intermediate points subject to the compression load from the walls above, the joists were made with solid infill pieces between the upper and lower chords.

 

I can't measure any static deflection. I don't know if mine were pre cambered or not.

[readiescards]

On 1/6/2018 at 22:37, Nickfromwales said:

 Fitted wardrobe doors seen moving etc. 

 

Yep our wardrobe doors do exactly that :-(

 

I'm at a lost who to question builder gave the job to the merchant who used a local posi-joist manufacturer and they used software supplied by MiTek - who I think own the posi-joist design. 

[readiescards]

On 1/7/2018 at 08:49, JSHarris said:

We have 253mm (PS10) Posijoists at 400mm centres over a wider span than this and the floor is rock-solid. 

 

I suspect that you have mid-span deflections that are greater than predicted for some reason, possibly the way the end-restraints are arranged on the joists, possibly because the static load is higher than anticipated, perhaps the joists were not pre-cambered during manufacture to the dead load deflection (a common failing), or maybe the floor boards were not secured as well as they might be. 

 

I doubt it's the latter and strongly suspect that it's either the lack of dead load pre-camber or greater freedom at the end attachments (like using joist hangers rather than fixing both members of the Posijoist securely to the structure).  Having them at 600mm centres will have exacerbated the problem.

 

 

They are indeed into joist hangers - I think the builder proposed these to help limit drafts into the thermal envelope
No idea if pre-cambered or not, difficult to assert now I fear.   

 

 

[readiescards]

On 1/7/2018 at 09:40, JSHarris said:

 

 

Yes, it would.  If the joists have been pre-cambered to the anticipated dead load, as per the spec (and I doubt they were), then measuring the static deflection in the centre of the room, i.e. the distance the centre is lower than the edges, should give an indicator.  If the joists were pre-cambered then there should be near-zero deflection, if they weren't then I'd expect to see a small deflection, maybe a mm or two in the centre.

 

The other point is the way the Posijoists are restrained at the ends.  If they are fixed such that the top and bottom members are firmly secured then the floor will have a lot less deflection than if  they are simple bearing on something like a joist hangar.  The latter will allow the joist to pivot slightly under load and exacerbate the centre deflection under load.

 

Ah, 

 

I've tried to measure this using a laser measure from the floor upwards some time back but all it really told me was ether the floor was sloped from one side to the other or the skimmed plasterboard was not put on with even thickness.

[Nickfromwales]

35 minutes ago, readiescards said:

 

Ah, 

 

I've tried to measure this using a laser measure from the floor upwards some time back but all it really told me was ether the floor was sloped from one side to the other or the skimmed plasterboard was not put on with even thickness.

You can only survey this with a laser and the floor exposed ( carpets up ). Also, you can't measure where the boards meet as they'll likely have cupped slightly, or a lot, with weather / exposure. 

[Nickfromwales]

52 minutes ago, readiescards said:

Yep our wardrobe doors do exactly that :-(

 

I'm at a lost who to question builder gave the job to the merchant who used a local posi-joist manufacturer and they used software supplied by MiTek - who I think own the posi-joist design. 

Simple. Pass it to the person you paid money to. 

[readiescards]

4 minutes ago, Nickfromwales said:

Simple. Pass it to the person you paid money to. 

 

I did do that and got the response 'you expect too much' 

[Jeremy Harris]

My gut feeling is that the problem is with the joist hangar detail.  It's critical that the lower chord of the joist is properly supported in the correct type of joist hangar.  The standard details given on drawing PSD05 here: http://www.mitek.co.uk/Products/Posi-Joist/Standard-Details/ is critical.

 

Note the key note to not trim the lower chord - this is the super critical bit, as it ensures that the lower chord of the Posijoist can take the compressive load without movement.

[readiescards]

Unfortunately I don't know and the various photos I have are not that clear on that detail.  

 

The joist hangers used do not seem to have the two round holes shown in PSD05 - should the joists be screwed/nailed through these holes (if they exist)?  The drawing does not seem to state if screws should be applied or not through these holes  

[Jeremy Harris]

It's hard to tell from those photos, but ideally the lower chord of every joist should be restrained so that it is held tight relative to the wall,  A tiny bit of horizontal movement at the lower chord translates into a fairly large vertical deflection, in the centre and I suspect that the joist is not held rigidly enough at the bottom, and this could be the primary cause of the problem.  It's hard to be sure without a close up of a joist end, though.

[ProDave]

What span is that? Our max span is 4.8 metres and our joists look a lot bigger than that at 100mm wide.

[Nickfromwales]

8 hours ago, readiescards said:

Unfortunately I don't know and the various photos I have are not that clear on that detail.  

 

The joist hangers used do not seem to have the two round holes shown in PSD05 - should the joists be screwed/nailed through these holes (if they exist)?  The drawing does not seem to state if screws should be applied or not through these holes  

I would expect to see, and would have fitted without being asked / instructed, full depth noggins at the joist ends just before the hangers. Even better would have been a full depth timber wall plate. 

Agree with @JSHarris a photo with more detail would have been beneficial but it doesn't look from the photos that the joists penetrate the wall ( for end restraint ) ?

 

8 hours ago, readiescards said:

 

I did do that and got the response 'you expect too much' 

I know what my reply to that would have been.....